A

oaoan©

K

1

ti(uvstgwtp

CT

0d

Leukemia Research 31 (2007) 231–234

Brief communication

VH3–48 and VH3–53, as well as VH3–21, gene rearrangements defineunique subgroups in CLL and are associated with biased lambda lightchain restriction, homologous LCDR3 sequences and poor prognosis

Christine Matthews a,b, Mark A. Catherwood a,T.C.M. ‘Curly’ Morris a, H. Denis Alexander a,b,∗

a Department of Haematology, Level C, Belfast City Hospital, Belfast, Northern Ireland, UKb School of Biomedical Sciences, University of Ulster, Coleraine, Northern Ireland, UK

Received 13 March 2006; received in revised form 29 March 2006; accepted 31 March 2006Available online 22 May 2006

bstract

This study determined IgVH gene usage in 228 chronic lymphocytic leukaemia patients to investigate associations between gene usage andther biological or clinical characteristics. VH3–48 [N = 8] and VH3–53 [N = 4] gene rearrangements showed biased � light chain restrictionnd were predominantly found in female patients with short lymphocyte doubling time but without adverse prognosis cytogenetics. Overuse

f VL3–21(V�2–14) gene and highly homologous LCDR3 sequences were found in VH3–48 patients. VH3–21 gene usage [N = 18, 7.9%] wasssociated with poor prognosis, overuse of VL3–21(V�2–14) gene and highly homologous heavy- and light-chain CDR3 sequences, but wasot associated with poor prognosis chromosomal aberrations.

2006 Elsevier Ltd. All rights reserved.

ukaem

bt

mrfaaa

eywords: VH gene usage; IgVH mutational status; Chronic lymphocytic le

. Introduction

In recent years much interest has surrounded IgVH muta-ional status, VH gene usage, and the potential role of antigensn the leukomogenesis of chronic lymphocytic leukaemiaCLL) [1–3]. In particular, the identification of VH3–21 genesage as defining a unique subset of CLL has lead to a re-isitation of the prognostic limitations of IgVH mutationaltatus, as VH3–21 is associated with poor prognosis irrespec-ive of the fact that two thirds of patients have mutated IgVHenes. Furthermore, this specific gene usage has been linked

ith highly homologous heavy- and light-chain complemen-

arity determining region 3 (CDR3), indicating that theseatients possess virtually identical B cell receptor (BCR)

∗ Corresponding author at: Department of Haematology, Level C, Belfastity Hospital, Belfast BT9 7AB, Northern Ireland, UK.el.: +44 2890236733; fax: +44 2890263870.

E-mail address: hd.alexander@bch.n-i.nhs.uk (H.D. Alexander).

2

fabit

145-2126/$ – see front matter © 2006 Elsevier Ltd. All rights reserved.oi:10.1016/j.leukres.2006.03.028

ia; Prognosis

inding sites and suggesting a common antigenic progeni-or.

In this study, patients with the VH3–21 gene rearrange-ent were investigated to determine if chromosomal aber-

ations were the source of poor prognosis and ascertain therequency of this specific gene usage, known to vary region-lly. Additionally we wanted to ascertain whether usage ofny other specific VH genes was associated with unique char-cteristics similar to VH3–21.

. Methods

Two hundred and twenty eight patients were recruitedrom Belfast City Hospital Haematology Outpatient Clinic

nd surrounding regional hospitals over a 3 year periodetween 2003 and 2006. Clinical staging (Rai and Binet),mmunophenotyping, lymphocyte doubling time (LDT) andime to treatment (TTT) were available on all patients.

2 mia Res

Fn7n≥(r

2C

Pvd[angnSgr

di(

2

m1c(i

2

sa

3

gtV(

racfac(

3

OaacpVvAawtimoc

3

p

TC

FMGILCLT

32 C. Matthews et al. / Leuke

ollow-up varied from a few months to three years for theewly diagnosed patients [N = 153], to 4–26 years (median

years) for those diagnosed previously [N = 75]. Diag-ostic criteria included peripheral blood lymphocytosis of5 × 109/l, typical CLL morphology and immunophenotype

CD19+ CD5+ CD23+ CD43+ kappa or lambda light chainestricted) and absence of t(11;14) as determined by FISH.

.1. IgVH and IgVL mutational status, gene usage andDR3 sequence

High molecular weight DNA was isolated using theuregeneTM DNA isolation kit. Multiplex PCR methods pre-iously described in the BIOMED-2 study were used toetermine IgVH and IgVL mutational status and gene usage4]. Products were purified and sequenced directly using anutomated ABI 310 DNA Sequencer with Big-Dye termi-ators. Mutations were identified by comparison with theerm-line sequence using Ig Blast (http://www.ncbi.nlm.ih.gov/igblast) and IGMT (http://imgt.cines.fr) databases.equences containing more than 2% deviation from theermline sequence were considered as somatically mutatedather than genetic polymorphisms.

HCDR3 and LCDR3 sequences in patients wereetermined using forward and reverse primers in sequenc-ng reactions and submitted to IMGT and V-BASEhttp://vbase.mrc-cpe.cam.ac.uk) databases.

.2. FISH analysis

Patients were tested for the presence of recurring chro-osomal aberrations, including trisomy 12, del 13q14, del

1q23, del 17p13 and t(11;14)(q13;q32) using commer-ial FISH probe, according to manufacturer’s instructionsAbbott Diagnostics). One hundred nuclei were examinedn each subject.

.3. Statistical analysis

Fisher’s exact test was utilised because of the relativelymall numbers of patients in each VH subgroup. All statisticalnalyses were performed using SPSS v13.0.

t(ut

able 1haracteristics of patients displaying biased lambda light chain restriction

VH3–21

requency 18/228 (7.9%)edian age, years 70 (48–94)ender, M:F 12M:6F

gVH, M:UM 14M:4UMight chain, �:� 6�:12�

hromosome abnormalities, good/poor prognosis 13:3ymphocyte doubling time (median) 5 monthsime to treatment (median) 8.5 months (N =

earch 31 (2007) 231–234

. Results

Of the 228 CLL patients, 132 (58%) had mutated IgVHenes while the remaining 96 (42%) were unmutated. Inhe cohort of patients studied, the most frequently expressed

H genes were VH4–34 (13.5%), VH1–69 (12.3%), VH1–27.9%), VH3–21 (7.9%) and VH1–3 (7.5%).

Immunophenotyping, including light chain restriction,evealed that those patients displaying VH3–21, VH3–48nd VH3–53 gene usage had a bias towards � lighthain restriction (p < 0.001) (Table 1). These patients wereurther analysed to reveal their HCDR3 and LCDR3mino acid sequences, and by FISH analysis to rule outhromosomal aberrations as a source of poor prognosisTable 2).

.1. VH3–21 patients

Eighteen patients showed VH3–21 gene usage (7.9%).f these 18 patients, 14 (78%) had mutated IgVH genes

nd 4 (22%) unmutated IgVH, (ratio 3.5:1). Light chainnalysis revealed that 12 patients showed clonal � lighthain restriction, whereas 6 expressed � light chains. Tenatients (56%) exhibited identical light chain gene usage,L3–21/J3 (V�2–14/J3). Nine of these displayed identical orery similar HCDR3 and LCDR3 amino sequence (HDCR3-RDANGMDV, LCRD3-QVWDSSSDHPWV). This group

s a whole represented a poorer prognosis subgroup in CLLith short LDT (median LDT 5 months) (p = 0.003), time to

reatment (TTT) (median 8.5 months), and progressive clin-cal course with 10 out of 18 patients already requiring treat-

ent. FISH analysis was performed on 16/18 patients. Onlyne of the 10 VH3–21/V�2–14 cases displayed an adversehromosomal aberration (del17p13).

.2. VH3–48

Eight patients expressed VH3–48 gene (3.5%). Six of theseatients were female resulting in a 3F:1M ratio. In addition,

his group showed a bias towards � light chain restriction7�/1�). Equal numbers of patients displayed mutated andnmutated IgVH genes (4M/4UM). Further analysis revealedhat these patients showed bias towards VL3–21/J3 gene

VH3–48 VH3–53

8/228 (3.5%) 4/228 (1.75%)70 (60–91) 79 (76–90)2M:6F 1M:3F4M:4UM 3M:1UM1�:7� 1�:3�

7:1 4:09 months 1 month

10) 18 months (N = 3) 2 months (N = 2)

C. Matthews et al. / Leukemia Research 31 (2007) 231–234 233

Table 2VL gene usage, CDR3 sequence results and clinical data of patients expressing VH3–21, VH3–48 and VH3–53 gene usage

VH gene HCDR3 sequence VL gene LCDR3 sequence LDT Treated Stage Rai/Binet

VH3–21 ARDANGMDV VL3–21/J3 QVWDSSSDQPWV TBR T RII/BBVH3–21 ARDANGMDV VL3–21/J3 QVWDSSSGSSMGV 14 months T RI/BAVH3–21 ARDANGMDV VL3–21/J3 QVWDSSDDHPWV 6 months NT RII/BAVH3–21 ARDRDGMDV VL3–21/J3 QVWDSSDDHPWV NR NT R0/BAVH3–21 ARDQNGMDV VL3–21/J3 QVWDSSSDHPWV 18 months T RIV/BCVH3–21 ARDANGMDV VL3–21/J3 QVWDSSSDQPWV <1 months T RIV/BCVH3–21 ARDRNGMDV VL3–21/J3 QVWDSSSGSSMGV 3 months T RIV/BCVH3–21 ARDQTGMDVWG VL3–21/J3 QVWDSSSDHPWV 3 months NT R0/BAVH3–21 ARGSQDVTLFDY VL3–21/J3 QVWDSSDDHPWV 4 months T RII/BBVH3–21 ARDQNGMDV VL3–21/J3 QQSYSTFHF 3 months T RIV/BCVH3–21 ARDQSGMDV VK1D–39/JK3 QQSYSTFHF 16 months NT R0/BAVH3–21 ARVSVPAAAWTNYYYYIDV VK4–1/JK2 QQDYTIPPT 25 months NT R0/BAVH3–21 VRGSLESLGAAMGAFDI VK4–1/JK5 QQYYSTPPT 3 months T RI/BAVH3–21 ARDEIS VK3–20/JK4 QQYPSSPS <1 months NT RI/BAVH3–21 ARDGMSSGWYVSYYYYMDV VK3–20/JK4 QQDYTIPPT 6 months T RIII/BBVH3–21 ADRKYCSGGSC VL1–47/J3 QQSYSTPGF 6 months T RII/BAVH3–21 ARWASYYDSSGYYYPVGN VL7–46/J3 QQYPSSPS NR NT RI/BAVH3–21 ARGHRGRFSHIDY VK1D–39/JK3 QQSYSTPGF 9 months NT RI/BAVH3–48 ARDPGLD VL3–21/J3 QVWDSSSDHPWV 11 months NT RI/BAVH3–48 ATLYMPRGDPPFDH VL3–21/J3 QVWDSGSDHPWV 19 months NT R0/BAVH3–48 ASGSEGYDSSGYFFDY VL3–21/J3 QVWDSTSDHPWV 21 months NT RI/BAVH3–48 ARASDD QVGDY VL3–21/J3 QVWDSSSDHPWV 6 months NT R0/BAVH3–48 CARPTQQLVQD VL3–21/J3 QVWDSSSDHPYV <1 months T RI/BAVH3–48 N/A VL NA 7 months NT R0/BAVH3–48 ARGHKPSPSSGWYEVDY VL3–21/J3 QVWDSSSDHPYV TBR T RII/BAVH3–48 ARPTQQLVQDY VK3–11/JK2 QQYGSSPVHF <12 months NT RIV/BCtVH3–53 ARAPXTNTFDH VL3–21/J3 QVWDSSSDHPWV 12 months NT R0/BAVH3–53 ANSGSYPSGYF VL2–1/J2 AAWDDSLRGVF TBR T RII/BBVH3–53 ATDRSPGVFGT VL2–1/J2 AAWDSSLSGVF TBR T RIV/BCVH3–53 ATAFYYDINTYGPDY VK4–1/JK1 QQYYITPRT 1 months DBT RIV/BC

T BT, die

u(pppTl

3

altiiociTfmat

ndp

4

pnaergrlsILt

BR, treated before reached; NR, not reached; T, treated; NT, not treated; D

sage and highly homologous LCDR3 amino acid sequencesQVWDSSSDHPWV). Only one VH3–48/VL3–21 patientossessed a chromosomal aberration associated with poorrognosis (del11q23). Clinical data revealed that all VH3–48atients had a short LDT (<1–21 months, median 9 months).o date, three patients have been treated, 12–36 months fol-

owing presentation.

.3. VH3–53

Four patients expressed VH3–53 gene rearrangement withpreponderance of females to males, (3:1) and a bias towards

ambda light chain restriction (3�:1�.) Despite the rarity ofhis particular gene usage (4/228, 1.8%), a distinct patterns evident. The three females presented with advanced clin-cal stage (Binet stage B-C), regardless of mutational statusf IgVH genes (all mutated) and absence of any detectablehromosomal abnormality. A short LDT was also recordedn this cohort of cases (<1–12 months, median 1 month).o date, two patients have been treated, both 2 months

ollowing presentation, one patient died before commence-ent of treatment and one remains untreated. Light chain

nalysis revealed identical VL gene usage (V�2–1/J2) inwo patients and these patients also had highly homoge-

bffe

d before treated; NA, not available.

ous LCDR3 sequence (AAWDDSSLSGVF). One patientisplayed V�2–14/J3 usage similar to VH3–21 and VH3–48atients.

. Discussion

VH3–21 gene usage was found in 7.9% of our cohort ofatients. This frequency lies between that reported in Scandi-avian countries (9–10%) [5] and in Mediterranean regionsnd the US (0–3%) [6]. The increased frequency in North-rn European countries, including Northern Ireland, mayeflect a racial divide as well as geographical variation in VHene usage, with Northern European countries showing littleacial diversity in comparison to other countries, in particu-ar the US. In keeping with previous studies we found thispecific VH rearrangement showed a bias towards mutatedgVH genes, � light chain restriction, and highly homologousCDR3 and HCDR3 amino acid sequences. This indicates

hat these cases effectively have an identical antibody com-

ining site, implying that the clonal progenitors all developrom B cells responding to the same antigen. Equally, weound this arrangement to be associated with progressive dis-ase, irrespective of mutated IgVH genes and lack of poor

2 mia Res

pu

regrir

hamapheuasasp

ouurrttapaTpcaocr

A

c

cma

adtmano2si

R

[

[

[

[

[

[

[

34 C. Matthews et al. / Leuke

rognosis chromosomal aberrations, and thus comprising anique entity in CLL.

In keeping with expectations, immunophenotypingevealed a predominance of � light chain restriction in thentire cohort of CLL patients studied. However, certain VHene rearrangements did show a bias towards � light chainestriction. An overrepresentation of � light chain was evidentn VH3–21 (13�/5�), VH3–48 (7�/1�) and VH3–53 (3�/1�)earrangements (p < 0.001).

Additional investigation of specific VH rearrangementsighlighted that patients displaying VH3–53 gene usage haddvanced clinical stage, short LDT and TTT, irrespective ofutational status and lack of poor prognosis chromosomal

berrations, thus indicating that this may represent a poorrognosis VH rearrangement. Additionally, VH3–48 patientsad an equal number of mutated to unmutated cases. How-ver the mutated cases had similar or shorter LDT than thenmutated cases. Intriguingly, these VH3–48 patients showedn overuse of VL3–21 gene and highly homogenous LCDR3equences, identical to those witnessed in VH3–21 patients,lthough HCDR3 sequences showed little homogeneity. Thisuggests that these patients represent a further unique poorrognostic subgroup in CLL.

VH3–48 gene usage has been reported previously to beverused in CLL patients in comparison to healthy individ-als [7]. However the incidence of VH3–48 and VH3–53sage in this study was relatively low (8/228 and 4/228,espectively) These intriguing initial results however war-ant additional investigation, as recent reports have suggestedhat specific VH genes appear to have a strong influence onhe gene expression pattern, pointing to antigen recognitionnd ongoing BCR stimulation as a pathogenic factor [8]. Thereferential use of certain VH gene rearrangements may occurs a result of B cells responding to specific antigenic stimuli.he differences in frequency of VH3–21 gene usage, betweenopulations, suggests a regional variation in frequency of aommon antigen. Additionally, this study identified VH3–48nd possibly VH3–53 to be unique entities in CLL, with anverrepresentation of female patients showing lambda lighthain restriction, homogenous LCDR3 sequences and infe-ior outcome irrespective of IgVH mutational status.

cknowledgements

The authors wish to acknowledge the patients whoonsented to take part in this study and colleagues who

[

earch 31 (2007) 231–234

ontributed, including Consultant Haematologists for per-ission to study their patients, Mr M Stevenson for statistical

dvice and Ms Tina Tomany for secretarial assistance.Contributions. HDA designed the study, obtained ethical

pproval and supervised the practical work of the PhD stu-ent, as well as providing a major contribution to writinghe paper. CM is a PhD student who carried out the IgVHutational status and gene usage investigations, and FISH

nalysis. She also collated all the clinical data from patientotes and produced the early drafts of the paper. MAC devel-ped the IgVH mutational status method from the BIOMED-primers and protocol and was responsible for the day to day

upervision of the molecular work. TCMM provided clinicalnput and critically reviewed the manuscript.

eferences

1] Damle RN, Wasil T, Fais F, Ghiotto F, Valetto A, Allen SL, et al. Ig Vgene mutation status and CD38 expression as novel prognostic indicatorsin chronic lymphocytic leukemia. Blood 1999;94:1840–7.

2] Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. UnmutatedIg V(H) genes are associated with a more aggressive form of chroniclymphocytic leukemia. Blood 1999;94:1848–54.

3] Tobin G, Thunberg U, Johnson A, Eriksson I, Soderberg O, Karls-son K, et al. Chronic lymphocytic leukemias utilizing the VH3–21gene display highly restricted Vlambda2–14 gene use and homologousCDR3s: implicating recognition of a common antigen epitope. Blood2003;101:4952–7.

4] van Dongen JJ, Langerak AW, Bruggemann M, Evans PA, Hummel M,Lavender FL, et al. Design and standardization of PCR primers and pro-tocols for detection of clonal immunoglobulin and T-cell receptor generecombinations in suspect lymphoproliferations: report of the BIOMED-2 Concerted Action BMH4-CT98-3936. Leukemia 2003;17:2257–317.

5] Thorselius M, Krober A, Murray F, Thunberg U, Tobin G, BuhlerA, et al. Strikingly homologous immunoglobulin gene rearrangementsand poor outcome in VH3–21-using chronic lymphocytic leukaemiapatients independent of geographic origin and mutational status. Blood2006;107:2889–94.

6] Ghia P, Stamatopoulos K, Belessi C, Moreno C, Stella S, Guida G, et al.Geographic patterns and pathogenetic implications of IGHV gene usagein chronic lymphocytic leukemia: the lesson of the IGHV3–21 gene.Blood 2005;105:1678–85.

7] Duke VM, Gandini D, Sherrington PD, Lin K, Heelan B, Amlot P, et al.V(H) gene usage differs in germline and mutated B-cell chronic lym-phocytic leukemia. Haematologica 2003;88:1259–71.

8] Kienle DL, Korz C, Hosch B, Benner A, Mertens D, Habermann A,et al. Evidence for distinct pathomechanisms in genetic subgroups ofchronic lymphocytic leukemia revealed by quantitative expression anal-ysis of cell cycle, activation, and apoptosis-associated genes. J Clin Oncol2005;23:3780–92.