close

Вход

Забыли?

вход по аккаунту

?

Serologic and molecular characterization of a human monoclonal rheumatoid factor derived from rheumatoid synovial cells.

код для вставкиСкачать
1188
SEROLOGIC AND MOLECULAR CHARACTERIZATION
OF A HUMAN MONOCLONAL RHEUMATOID FACTOR
DERIVED FROM RHEUMATOID SYNOVIAL CELLS
DICK L. ROBBINS, THOMAS P. KENNY, MARIA J. COLOMA, JORGE V. GAVILONDO-COWLEY,
RAFAEL W. SOTO-GIL, POJEN P. CHEN, and JAMES W.LARRICK
Molecular characterization of rheumatoid factors (RF) in rheumatoid arthritis (RA) has been hampered because of their polyclonality To overcome this
problem, we generated monoclonal RF-secreting hybridomas from rheumatoid synovial cells. Among the
RF-secreting hybridomas, HAFlO secreted an IgM-RF
that was monospecific for human IgG. It bound well to
IgGl and IgG2, but not to IgG3 and IgG4. Sequence
analysis of its heavy and light chains showed that it
contained a VH1 heavy chain and a V h light chain that
did not belong to any known A light chain subgroup, and
therefore, probably represented a new A subgroup.
These results indicated that both the heavy and light
chains of a monoclonal IgM-RF from rheumatoid synovial cells were quite different from the reported variable
region sequences of several monoclonal RF derived
mainly from patients with mixed cryoglobulinemia.
.
From the Department of Internal Medicine, School of
Medicine, University of California, Davis; the Division of Exploratory Research, Genelabs, Inc., Redwood City, CA; and the Department of Molecular and Experimental Medicine, Scripps Clinic
and Research Foundation, La Jolla, California.
Supported in part by a grant from the AMA-ERF. Dr.
Robbins’ work was supported in part by NIH grant AM-32287. Dr.
Chen’s work was supported in part by NIH grant AR-39039.
Dick L. Robbins, MD: Professor of Medicine, University of
California, Davis; Thomas P. Kenny, BS:Staff Research Associate,
University of California, Davis; Maria J. Coloma, MS: Visiting
Research Scientist, Genelabs, Inc.; Jorge V. Gavilondo-Cowley,
PhD: Visiting Research Scientist, Genelabs, Inc.; Rafael W. SotoGil, PhD: Postdoctoral Fellow, Scripps Clinic and Research Foundation; Pojen P. Chen, PhD: Associate Member, Scripps Clinic and
Research Foundation; James W. Lamck, MD, PhD Director of
Exploratory Research, Genelabs, Inc.
Address reprint mquests to Dick L. Robbins, MD,Division
of Rheumatology, University of California, School of Medicine, TB
192, Davis, CA 95616.
Submitted for publication January 29, 1990; accepted in
revised form March 12, 1990.
Arthritis and Rheumatism, Vol. 33, No. 8 (August 1990)
Further studies of additional monoclonal RF from RA
patients are warranted to define precisely their genetic
basis and to further our understanding of the immuno.
pathology of RA.
The cellular component of the immune system
is undoubtedly important in the pathogenesis of rheumatoid arthritis (RA). However, RA is also an autoimmune, extravascular, immune complex disease involving synovium and interstitial tissues. The major
autoantibody present in RA is rheumatoid factor (RF),
a polyclonal autoantibody directed against the Fc
portion of IgG. Several observations indicate that
RF-IgG complexes and their subsequent activation of
complement are important in the pathogenesis of RA
(1-3). Because synovitis is central to the immunopathologic events in RA, synovial RF may have
greater pathogenetic importance than corresponding
serum RF. In fact, RF synthesis is the major characteristic of synovial tissue plasma cells in RA, and
intrasynovial immune complex formation is a predominant feature of RA synovitis (4).
RF are heterogeneous in many qualitative characteristics including complement-activating properties, antigen specificity, and avidity (2,5,6). For example, RA serum RF has cross-species reactivity and
binds well to human IgG subclasses IgG 1, IgG2, and
IgG4, but poorly to IgG3 (5,7,8). In contrast, synovial
RF binds well to IgG3 and with greater avidity than to
IgGI, IgG2, and IgG4 (5,6). Moreover, these polyclonal RF react with various antigenic sites in the
second and third domains of the y chain constant
regions (CH2 and CH3 domains), which include classand subclass-specific antigens, as well as human IgG
allotypic markers (i.e., Ga and Gm) (9). Furthermore,
RSC-DERIVED MONOCLONAL RF
the affinity of serum RF for IgG Fc is low, with an
association constant of -104-105 liters/mole (10,ll).
Thus, characterization of the molecular genetic basis
and pathogenicity of these molecules has been difficult
and unsuccessful.
During the last few years, our understanding of
RF, particularly their molecular genetics, has advanced significantly. However, because of the polyclonal nature of RF in RA patients, previous molecular
studies involved mainly the monoclonal RF from patients with mixed cryoglobulinemia (12). Since the
relationship between monoclonal RF in patients with
cryoglobulinemia and polyclonal RF in RA patients is
unclear, it is important to define precisely the genetic
basis of polyclonal RF in RA patients, and to compare
such results with those in patients with cryoglobulinemia. Toward this end, we have generated RFsecreting hybridomas from unstimulated rheumatoid
synovial cells (RSC), and analyzed these hybridomas
in detail. We report herein the binding specificity and
the variable (V) region sequences of a monoclonal RF
from 1 hybridoma. These results show that both heavy
and light chain sequences of a synovial RF are quite
differentfrom those of monoclonal RF from cryoglobulinemia patients. Further studies of RSC-derived
monoclonal RF are warranted and should reveal the
genetic basis of RF production in RA synovium.
MATERIALS AND METHODS
Preparation of synovial cells. Synovial tissue was
obtained incidental to clinically indicated surgical procedures (synovectomy, total joint replacement)on seropositive
RA patients, as previously described (5,6). These patients
had definite or classic RA, according to the American
Rheumatism Association criteria (1 3). Tissue was collected
during surgery, under conditions of local, temporary ischemia, so that contamination of the tissue with peripheral
blood was minimal. In all cases, synovial cells were washed
during processing to remove serum RF. Specimens were
collected in Caz+- and Mg*+-free Hanks’ balanced salt
solution (HBSS; Gibco, Grand Island, NY) containing 2.5
IU/ml of heparin and antibiotics (50 IUlml penicillin and 50
mg/ml streptomycin) and were processed immediately. The
tissue was minced and dissociated by treatment with 0.5
mg/ml of crude collagenase and 0.15 mg/ml of DNase (both
type I; Sigma, St. Louis, MO) at 37°C for 60 minutes,
followed by filtration through nylon mesh (single layer, pore
size 200 pm). The eluted cell suspension was centrifuged
(150g for IS minutes at 4°C) and the cells resuspended in
HBSS. The cells were washed 3 times in HBSS at ISOg for 10
minutes at 4°C and resuspended in RPMI 164&10% fetal calf
serum (FCS; Gibco) and antibiotics. Cells were counted in a
hemacytometer and viability assessed by trypan blue dye
exclusion. Cell viability was consistently 290%.
1189
Generation of human RF-secreiing hybridomas. RSC
were fused as described (14.15) with slight modifications.
Briefly, RSC and mouse/human hybrid F3B6 cells were
washed separately in HBSS without calcium and containing
2 mM magnesium (HBSS-/+). The cells were mixed at a I: 1
ratio and 2 X lo7 cells/well were added to peanut agglutinincoated 6-well plates (Falcon, Oxnard, CA). The plates were
centrifuged at 500g for 6 minutes at room temperature. The
supernatant was aspirated from the monolayer and 2 ml of
warm (37°C) 40% polyethylene glycol (PEG)fusion mixture
(8 gm PEG 4000, 2 ml DMSO, and sufficient HBSS-/+ to
give a final volume of 20 ml) was added down the side of the
wells. After 1 minute, 4 ml of warm HBSS-/+, 5% DMSO
was slowly added over the next 2 minutes. Four milliliters of
HBSS-/+, 5% DMSO was then added over the next minute.
The wells were aspirated and 2 ml of HBSS-/+ was added
and the plate was centrifuged at 400g for 5 minutes. The fluid
was aspirated, 2 ml of HBSS-/+ was added, and the plate
was centrifuged at 400s for 5 minutes. The HBSS-/+ was
aspirated and 3 ml of growth media (Iscoves modified
Dulbecco’s medium supplemented with 10% prescreened
and heat-inactivated FCS, 0.005 IUlml bovine insulin, 5
&ml human transferrin, 5 ng/ml sodium selenite, 5 Fg/ml
human low density lipoprotein, 50 IUlml penicillin, and 50
kg/ml streptomycin) was added to the wells. The plate was
incubated overnight at 37°C in an atmosphere of 5% CO, and
95% air. The cells were then diluted with growth media
hypoxanthine and 2 pdml azaserine. The
containing 100
cells were gently resuspended and seeded into 96-well flatbottom plates at a density of 1 x lo5 cells/well. The cells
were given fresh hypoxanthine and azaserine-supplemented
growth media every third day. Hybrids were usually ready to
assay by day 14 after fusion. The hybrids were screened
initially for IgM and IgM-RF production by an enzymelinked immunoabsorbent assay (ELISA).
Once RF-secreting cells were identified, these cells
were subcloned at 30, 3, and 0.3 cells/O.l ml of growth media
containing 15% FCS in U-bottom wells. Cells were given
fresh media every 4 days and assayed on day 12. Cells at the
highest dilution showing <36% positivity for hybridoma
growth (Poisson formula) were presumed to be monoclonal.
Several of these hybridoma samples were recloned until
100% of the wells containing growing cells secreted monoclonal RF.
Preparation of human and rabbit IgG. Human and
rabbit Cohn fraction I1 were passed through a DEAEcellulose column (DE-52; Whatman, Springfield Mill, Kent,
UK) in 0.01M phosphate buffer, pH 8.0, as described (5,6).
Myeloma proteins were isolated using ion-exchange chromatography, followed by gel permeation chromatography
where indicated, as previously described (5,6). In addition,
all IgG3 preparations were applied to a protein A-Sepharose
column (Pharmacia, Piscataway, NJ) in 0.1M phosphate
buffer, pH 7.0, to remove any possible contaminating subclasses. All of the subclass proteins used in these studies
were 98-99% pure, as determined by an inhibition ELISA.
Monomeric preparations of the respective IgG proteins were prepared by ultracentrifugation at 40,000revolutions per minute at 4°C to remove polymeric forms (5,6). All
IgG preparations were negative for RF activity, as determined by ELISA.
ROBBINS ET AL
1190
Table 1. Sequences of the 5' and 3' regions of synthetic oligonucleotide polymerase chain reaction
primers*
Human A light chains
SEco RI/Sign.HL
S'--GG-ATG(
AG)CCTG(CG)(AT)C(CT)CCTCTC(CT)T(CT)CT(CG)(AT)(CT)C--3'
3' Hind IWHL Constant
5'--CCmGAAGCTCCTCAGAGGAGGG--3'
Human p heavy chains
5'Eco RI/Sign.HHI
S'--GGGAATTCATGGACTGGACCTGGAGG(AG)TC(CT)TCT(GT)c--3'
5'Eco RI/Sign.HHZ
S'--GGGAATTCATGGAG(CT)TTGGGGCTGA(CG)CTGG(CG)~(CT)T--3'
5'Eco RIISign.HH3
5 ' - G G m A T G ( AG)A(AC)(AC)(AT)ACT(GT)TG(GT)(AT)(CG)C(AT)(CT)--
-(CG)CT(CT)CTG-3'
3'Hind W H H Constant
5'--CCmAGACGAGGGGGAAAAGGGTT-3 '
_
_
_
_
~
* Bases in parentheses represent substitutions at a given position, i.e., (AT)means both A and T were
present in equimolar amounts during the synthesis of a particular position. Eco RI (5' end) and Hind
Ill (3' end) sites are underlined. An additional 2 bases (G or C) were added 5' to the restriction
endonuclease site to facilitate enzymatic cleavage. Sign. = signal peptide region; H = heavy; L =
light.
The pFc' fragment of IgGl was prepared by pepsin
digestion, followed by chromatography of the digest on
Sephadex G-150 (6). Fab and Fc fragments were obtained
from 1-hour papain digestion hydrolysates by DEAEcellulose chromatography, as described previously ( 16). The
F(ab'), fragment was obtained by pepsin digestion and
Sephacryl S-200 chromatography ( 17). These proteolytic
fragments were examined for homogeneity by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (6).
ELISA for RF. Plastic microtiter plate wells (Falcon
#3912; Becton Dickinson, Oxnard, CA) were precoated with
0. I mllwell of the indicated IgG diluted in phosphate buffered
saline (PBS) for 2 hours at room temperature. The wells
were washed 3 times with PBS, 0.05% Tween 20, and then
quenched for 2 hours with PBS supplemented with 1%
bovine serum albumin (BSA). The buffer was removed, and
0.1 ml of a predetermined concentration of monoclonal
IgM-RF and standards containing IgM-RF was added. After
1-2 hours incubation at room temperature, the wells were
washed as described above; then 0.1 mi of Ftab'), goat
anti-human IgM labeled with horseradish peroxidase (Tago,
Burlingame, CA) diluted 1:2,000 in PBS-BSA was added to
the wells. After I hour of incubation at room temperature,
the wells were again washed as described above. A solution
of 0.05% ABTS (Sigma) diluted in 0.05M citrate buffer, pH
4.2, was freshly prepared for each assay. Immediately before
addition to the wells, 0.1 ml of 3% H,O, was added to 10 ml
of ABTS. The color was allowed to develop for I5 minutes,
and was stopped by the addition of 0.05 ml of 5% sodium
dodecylsulfate. The absorbance at 412 nm (optical density)
was measured in a Titertek Multiscan Meter (Flow Laboratories, Rockville. MD) and was used to calculate the amount
of R F bound per microgram of IgG. The absorbance due to
background binding (derived from wells that were not coated
with IgG) was subtracted.
Amplification and cloning of the heavy and light chain
variable region complementary DNA (cDNA). Based on the
reported sequences of human Ig molecules (181, 5' and 3'
primers were constructed, respectively, from the conserved
sequences of the signal peptide regions and the constant
regions of p heavy and A light chains (Table I). Mixed
oligonucleotides were constructed with an Applied Biosysterns (Foster City, CA) 380B DNA synthesizer. All primem
were designed with Eco R1 and Hind 111 restriction endonuclease sites at the 5' and 3' ends, respectively. Additionally,
2 bases (Go r C ) were added 5' to the restriction enzyme site
to improve enzymatic digestion.
Total RNA was prepared from 10" hybndoma cells
by a microadaptation of the guanidinium thiocyanatelcesium
chloride procedure (19). and was used to synthesize the first
strand of cDNA, using the Boehringer Mannheim (Chicago,
IL)cDNA kit (20). Ten microliters of cDNA and 5 p1 of each
primer ( I pM final concentration) were added to 80 4 of the
polymerase chain reaction mixture (Perkin-Elmer Cetus,
Norwalk, CT). The mixture was amplified in a Perkin-Elmer
thermal cycler for 30 cycles, with each cycle consisting of
melting at 94°C for 1 minute, annealing at 55°C for I minute,
and extension at 72°C for I minute. with a 1-minute ramp
time between each temperature step. The amplified products
were analyzed on a 2% agarose gel with ethidium bromide.
DNA sequence analysis. The amplified cDNA was
sequenced either directly or after M 13 subcloning. For direct
sequencing (21). amplified cDNA was electrophoresed on a
low-melting agarose gel. and the relevant bands were isolated. The isolated samples were asymmetrically amplified,
using 0.2 pM of the 5' end primer and 0.01 pM of the 3' end
primer. The products were extracted with phenol/
chloroform. spin-dialyzed in Centricon-30 microconcentrators (Amicon. Danvers. MA), and sequenced with 0.5 pM of
the 3' end primer. For MI3 subcloning, the amplified cDNA
RSC-DERIVED MONOCLONAL RF
1191
BSA
mR0 =DNA
GLOBUJN
dSDNA
HlSfONE
COLL-
OVALBUMIN
HubG
Antlgens
HulgM
HulgG
Kappa
Lambda
Mouse Ig'a
ANTISERA TO
Figure 1. Characterization of HAFIO monoclonal rheumatoid factor (mRF) by a direct-bindingenzyme-linked immunosorbent assay.
This mRF was used to coat microtiter plate wells and tested against
the indicated antisera. Northern blot analysis of F3B6 cells demonstrated that human chromosomes 2, 14, and 22 were absent (data not
shown). O.D. = optical density; HuIgM = human IgM.
was digested with Eco RI and Hind 111 and was ligated into
M13mp18/19 (22,231. Appropriate clones were isolated and
sequenced by the dideoxynucleotide chain termination
method. At least 2 independent M13 clones of opposite
orientation were used to validate each sequence. The computer programs of the Genetics Computer Group (24) were
used to analyze the sequence data.
RESULTS
Generation and characterization of the HAFlO
RF-secreting hybridoma. RSC from patient WP were
fused with F3B6 cells and hybridoma cell growth was
observed in 450 (94%) of 480 culture wells. The
supernatants from these wells were analyzed for RF
activity; 16 were found to be positive and each exhibited its own IgG subclass reactivity profile. These cell
lines were subcloned to ensure the monoclonality of
each hybridoma. Among the RF-secreting hybridomas, HAFlO secreted only a human monoclonal
IgMh-RF (Figure 1). Of greater importance, the
HAFlO RF was monospecific, reacting only with human IgG, but not with any other antigens tested,
including BSA, single- and double-stranded DNA,
histone, collagen, thyroglobulin, and ovalbumin (Figure 2). Furthermore, the binding of this RF to IgG was
species specific and subgroup specific. It reacted with
human and monkey IgG, but not with IgG from rabbit,
chicken, mouse, rat, and goat (data not shown).
Figure 2. Binding specificity of HAFIO monoclonal rheumatoid
factor (mRF) to a panel of antigens and to human IgG (HulgG)
measured by a direct-binding enzyme-linked immunosorbent assay.
O.D. = optical density; BSA = bovine serum albumin; ssDNA =
single-stranded DNA; dsDNA = double-stranded DNA.
HAFlO RF also showed subclass specificity. It
bound well to many IgGl and IgG2 proteins, but not to
any IgG3 or IgG4 proteins (Figure 3). For example,
HAFlO RF bound efficiently to IgGl molecules MCC,
LOW, PEI, GER, DRI, and WIE (Figure 3a), but it
bound CAR, HIL, HEI, FEN, and VAN to a far lesser
degree. HAF10 RF also reacted with IgG2 molecules
REY and POS, but not with TIL and DUP (Figure 3b).
This R F did not react significantly with any of the IgG3
or IgG4 molecules examined (Figures 3c and d).
Since HAFlO RF reacted well with some monoclonal IgGl and IgG2 paraproteins, but poorly with
other monoclonal IgGl and IgG paraproteins, we
determined the Gm allotypes of a few IgG1 and IgG2
paraproteins (Figures 3a-d). It appeared from these
limited data that HAFlO RF was not allotype specific.
To define the epitope recognized by the HAFlO
RF, the RF was analyzed for reactivity against various
fragments of IgG1. HAF10 RF did not bind to the Fab,
F(ab'),, or pFc' fragments of IgGl ,suggesting that the
epitope was contained in the CH2 domain, or at the
juncture of the CH2 and CH3 domains (data not
shown).
Cloning and sequencing of the heavy and light
chains of HAFlO RF. RNA from the HAFlO cells was
prepared and used to synthesize the first strand of
cDNA. Heavy and light chain cDNAs were then
specifically amplified with designed oligomers. Based
on the leader sequences of the p heavy and A light
chains, we constructed and synthesized mixed oligomers for p heavy chain V genes and mixed oligomers
for A light chain V genes (Table 1). These oligomers
ROBBINS ET AL
1192
1 0
.a
lgG1 PROTEINS
o a
4
o_
f
pm
0 1
o a
0 s
' 4
0.3
a
E
0 1
0.9
0 0
MCC CAR LOW PEI GER DRI HlL MA1 PED E R A HE1 FEN VAN WIE
K
K
K
K
K K
K
L
L
L
L L
L
ND ND ND NO a NO NO NO NO 1
I
ND NO
LTCHAINK
ALLOTVPE 2
d
-
1
0
IQGZ PROTEINS
lQG3 PROTEINS
lpG4 PROTEINS
..*
4
a
I
1
nated HumlaflO) is considerably different from all 6 or
7 families of A light chain sequences defined either by
Kabat et a1 (18) or by the National Biomedical Research Foundation (NBRF) library database. The main
differences between these 2 classification systems are
I) the A 3 and A 4 subgroups defined by the NBRF are
termed A 4 and A 3, respectively, by Kabat et al; 2) 2 of
3 h 5 light chains (i.e., Bo and McG) defined by Kabat
et al are classified in the A 2 subgroup by the NBRF;
and 3) the single A 5 light chain defined by NBRF is
classified in the A 3 subgroup of Kabat et al. Because
the A light chain sequences of the NBRF are in the
computer database form, we will use the NBRF terminology in all subsequent discussions. Using the
GAP program of the Genetic Computer Group sequence analyses package, the HumlaflO amino acid
sequence shares a 68%, 72%, 67%, 63%, 62%, 63%,
65%, and 56% similarity with Llhumm, L2humc,
L2hung. L3hush, L4hukn, LShudl, L6hul1, and
L7humt, respectively. Although HumlaflO is more
similar to L2humc than to L2hung, it should be noted
that L2humc of the NBRF database is classified as A 5
by Kabat et al, indicating that L2humc may not be a
I
REV TIL WSDUPLIEICRO HUS GRE WD L A 1HER TEASED H A S
LTCHAlNK K
K L
K K K
L
L L
L
K K
L
ALLOTWE ND ND n - ND b g ND NO b
Q
4, 4b ND
-12
Figure 3. Human IgG subclass specificities of HAFIO monoclonal
phe cys l e u l e u a l r Val a l r pro g l y Val h i s ser
TTC TGC TTG CTG GCT GTA GCT CCA GGT GTC CAC TCC
rheumatoid factor (mRF) measured by a direct-binding enzymelinked immunosorbent assay. IgG subclass molecules were used to
precoat microtiter plate wells at a concentration of 50 pglml. In each
case, anti-human IgG was used to confirm that approximately equal
concentrations of subclass protein adhered to the well. a, lgGl
proteins. b, IgG2 proteins. c, IgG3 proteins. d, lgG4 proteins. O.D.
= optical density; LT = light; K = kappa; L = lambda: ND = not
determined.
S I W
-1
1
FRl
g l n Val g l n
-------- > CAG
GTG CAG
4
FR1
LEU Val glu SER g l y 81r g l u Val l y s 1yS pro g l y 818 ser Val l y r v r l ser
CTG CTG CAG TCT GGG GCT GAG GTG AAG AAG CCT GGG GCC TCA GTG M G GTT TCC
________._
-- -
22
FR2
CDRl-cys lys a11 ser g l y t y r t h r phe t h r ser t y r t y r net h i s t r p v r l rrg g l n
TGC AAG GCA TCT GGA TAC ACC TTC ACC AGC TAC TAT ATG CAC TGG GTG CGA CAG
40
FR2
--------------.--CDRz---------ala pro g l y gln GLY LEU GLU TRP nat g l y 110 v 8 l r r n pro ser g l y g l y i l e
GCC CCT GGA CAA GGG CTT GAG TGG ATG GGA ATA GTC AAC CCT AGC GGT GGT ATC
----------CDRL--------------------58
FR3
t h r ser t y r a11 g l n net phe gln gly ARC Val t h r met t h r 8rg asp t h r SER
ACA AGC TAC GCA CAG ATG TTC CAG GGC AGA GTC ACC ATG ACC AGG GAC ACG TCC
were used in conjunction with the oligomers that were
complimentary to the 5' end of the p and A chains to
amplify the heavy and light chain V region cDNAs.
The amplified cDNA was sequenced directly,
and was also sequenced after subcloning into M13.
The resulting sequences of the HAFlO RF heavy chain
and light chain V regions from both approaches were
identical and are presented in Figures 4 and 5 , respectively. The HAFlO RF heavy chain nucleotide sequence (designated Humhalflo) is most similar to the
HG3 germline VHI gene, sharing 96% homology over
a region of 305 basepairs. At the protein level, these 2
VH genes are 95% homologous (18).
In contrast, the HAFlO RF light chain (desig-
FR3
76
t h r ser t h r Val h i s net g l u l e u r e r r e r l e u arg ser g l u asp thr ALA v r l
ACG AGC ACA GTC CAC AT6 GAG CTG AGC AGC CTG A M TCT 6AG 6AC ACG GCC GTG
------------------------tDR)-----------------------91
ijr
t y r cvs a l r rrg asp r e r rrg g l y g l y asp l e u l e u t h r g l y h i s h i s cys
TAC TAC TGC GCG AGA GAT TCG CGC GGC GGC 6AT CTT TTG ACT GGT CAT CAT TGC
__-____---112
FR4
i l e asp t y r t r p g l y gln GLV t h r l e u v r l t h r VAL ser SER
A l l 6AC TAC 166 GGC CA6 GGA ACC CTG GTC ACC GTC TCC TCA
125
---------- > CH1
Figure 4. Complete nucleotide and inferred amino acid sequence of
the HAFIO monoclonal rheumatoid factor heavy chain variable
region. Framework regions (FR) and complementarity-determining
regions (CDR)are indicated. Amino acids different from the database for human immunoglobulin heavy chain subgroup I (IgM) (18)
are capitalized. SIGNAL = signal peptide sequence; CHI = constant heavy 1 region.
RSC-DERIVED MONOCLONAL RF
-5
SIGNAL
-1
1 SER
FR1
thr gly ser leu ser
gln leu Val LEU thr GLN ter pro SER ala
ACA GGG TCT CTC TCC ------->cAG
CTT GTG CTG ACT CAA TCG ccc TCT GCC
12
FR1
-------CDRI-------SER ala SER leu GLY ala ser Val lys leu thr CVS thr leu ser ser gly
T C T GCC TCC CTG GGA GCC TCG GTC AAG CTC ACC TGC ACT CTG AGC AGT GGC
FR2
his ser ser tyr ala i l e ala TRP his gln gln lys PRO asp lys gly PRO
29-----------CDRl----------
CAC AGC AGC TAC GCC ATC GCA TGG CAT CAG CAG AAG CCA GAC AAG GGC CCT
arg phe leu met lys leu asn i l e asp gly ser his ser lys ly asp gly
CGG TTC TTG ATG AAG CCT AAC ATT GAT GGC ACT CAC AGT AAG ~ G GGAC GGG
62
FR3
I l e pro asp ARG phe SER GLY SER ser ser gly a11 glu cys tyr LEU THR
ATC CCT FAT CGC TTC TCA GGC TCC AGC TCT GGG GCT GAG TGC TAC CTC ACC
_____
CDRJ79
Ila SER ser leu gln ser glu asp GLU ALA asp TYR phe CYS gln thr trp
ATC TCC AGC CTC CAG TCT GAG GAT GAG GCT GAC TAT TTC TGT CAG ACC TGG
96---------------------
FR4
gly thr gly thr phe val val phe gly gly gly thr
GGC ACT GGC ACC TTT GTG GTA TTC GGC GGA GGG A c t
107
--- ----- > CL
Figure 5. Complete nucleotide and inferred amino acid sequence of
the HAFlO monoclonal rheumatoid factor A light chain variable
region. Amino acids different from the database for human immunoglobulin A light chain subgroup I1 (18) are capitalized. CL =
Constant light region. See Figure 4 for other definitions.
typical A 2 light chain. These data suggest strongly that
the HumlaflO Vh sequence represents a new VA
subgroup.
DISCUSSION
To delineate the genetic basis of the diseaserelevant RF in RA patients, we generated a hybridoma
that secreted a monospecific RF and determined the
nucleotide sequence of its V region. It should be
pointed out that the current sequence data are the first
set of V region nucleotide sequences of an RF derived
from the rheumatoid synovium. A major characteristic
of the R F in RA patients is that some RF are produced
locally in the rheumatoid synovium and are implicated
in the chronic tissue damage of involved joints (1-3).
Thus, it is of critical importance to define the genetic
basis of RF production.
We generated RF-secreting hybridomas from
RSC rather than from peripheral blood lymphocytes of
RA patients. Furthermore, we used unstimulated RSC
in our cell fusions so that we might select only the in
vivo activated RF-secreting B cells from the joints,
instead of irrelevant resting B cells activated and
transformed by Epstein-Barr virus. Thus, the RFsecreting hybridomas described here should provide
the best samples for delineating the molecular basis of
R F production in RA patients. As is evident in Figure
2, the HAFlO monoclonal IgM-RF is monospecific for
1193
human IgG. a distinctive feature of autoantibodies in
patients with various autoimmune diseases. In contrast, most so-called “natural” or “physiologic” autoantibodies found in nonautoimmune states are
polyspecific (25). Furthermore, HAFlO did not bind to
the Fab and F(ab’)*fragments of IgG 1, indicating that
HAFlO is a classic RF that reacts with the Fc portion
of IgG molecules (1-3).
Detailed sequence analyses of Ig V region gene
sequences in several murine antibody responses have
revealed that the extent of somatic diversification in
the heavy and light chains of an antibody molecule is
often similar, reflecting a phenomenon that has been
termed “parallel diversification” (26). As described
earlier, the Humhalflo amino acid sequence is 95%
homologous to the HG3 amino acid sequence (181,
indicating that the Humhalflo was derived either from
HG3 with 6 amino acid residue changes or from other
more closely related VH genes with fewer than 6
amino acid changes. Accordingty, the HumlaflO amino
acid sequence is likely to deviate from the putative
germline VA gene by approximately 6 amino acid
residues. When these findings are related to the sequence comparison of HumlaflO with A light chains of
all 7 known A light chain families, it can be concluded
that HumlaflO is encoded by a Vh gene that represents
a new A light chain family.
Because of their frequent occurrence in patients
with mixed cryoglobulinemia, monoclonal IgM-RF
have been studied in detail in these patients (12). In
1973, Kunkel and colleagues found that approximately
60% of monoclonal RF shared a cross-reactive idiotype, termed Wa (27). Recently, combined serologic,
structural, and molecular analyses have shown that
the light and the heavy chain variable regions of most,
if not all, idiotype positive monoclonal RF are encoded
by the germline Humkv325 Vk3 gene and the Humhallr gene, respectively, the latter being a rearranged
VHI gene for the surface IgM of a chronic lymphocytic leukemia (12). In addition, it has been shown that
some monoclonal RF light chains are encoded by the
germline Humkv328 Vk3 gene and the germline Vg
Vk3 gene (12.28-30). When compared with these Ig V
genes, the current V gene sequences of our RAderived RF are considerably different. The Humhalflo
sequence differs from the Humhallr sequence in that
they share an 86% homology, while Humhalflo is 96%
homologous to the germline HG3 VHI gene. Interestingly, the 21/28 VH sequence of the only anti-DNA
antibody derived from a patient with systemic lupus
erythematosus (31) is most closely related to HG3
ROBBINS ET AL
1194
among all the reported functional human V H l genes.
The meaning of these relationships is unclear at this
point. Similarly, the HumlaflO is rather different from
all reported RF-related light chain V gene sequences.
In conclusion, we generated RF-secreting hybridomas from unstimulated RSC in an effort to understand the roles of RF in the pathogenesis of rheumatoid synovitis, and showed that these RA-derived
monoclonal RF were monospecific and displayed binding properties similar t o those of the polyclonal R F in
RA patients. Molecular analysis of 1 RF revealed that
both the p heavy and A light chain variable region
sequences were quite different from previously reported V gene sequences of monoclonal R F from
non-RA patients. Thus, it is clear that further studies
of the RSC-derived monoclonal RF will allow us to
define precisely qualitative characteristics of R F such
as isotype, affinity, and fine specificity, and to delineate the molecular basis of pathogenic autoantibodies.
This information will help elucidate the induction and
the sustained production of RF in RA patients, particularly in the rheumatoid synovium, and, in turn, reveal
the role of R F in the pathogenesis of RA.
Addendum, Since this manuscript was submitted for
publication, a recomparison of the HumlaflO sequence with
the latest release of Genbank database (December 1989)
revealed that lafl0 was 89% homologous with the k6h6
sequence derived from the PK follicular lymphoma (32).
This finding clearly establishes the new Vh8 family. Accordingly, HumlaflO has been renamed Humla8flO.
ACKNOWLEDGMENTS
The authors thank Cameron Hoover for oligonucleopeptide synthesis, Kirk Fry for helpful discussions and
computer analysis, Richard Wistar, Jr., for IgG subclass
protein, Nikki Rojo for excellent secretarial assistance, and
Kim Robbins for excellent technical assistance.
REFERENCES
1. Carson DA: Rheumatoid factor, Textbook of Rheumatology. Edited by WM Kelley, ED Harris JR, S Ruddy,
CB Sledge. Philadelphia, WB Saunders, 1989
2. Robbins DL, Feigal DW Jr, Leek JC, Shapiro R,
Wiesner K: Complement activation by 19s IgM rheumatoid factor: relationship to disease activity in rheumatoid
arthritis. J Rheumatol 13:33-38, 1986
3. Zvaifler NJ: The immunopathology of joint inflammation
in rheumatoid arthritis. Adv Immunol 16:265-336. 1973
4. Fehr K, Velvart M, Rauber M, Knopfel M, Baici A,
Salgam P, and Bdni A: Production of agglutinators and
rheumatoid factors in plasma cells of rheumatoid and
nonrheumatoid synovial tissues. Arthritis Rheum 24:
510-519, 1981
5. Robbins DL, Wistar R Jr: Comparative specificities of
serum and synovial cell 19s IgM rheumatoid factors in
rheumatoid arthritis. J Rheumatol 12:437443, 1985
6. Robbins DL, Benisek WF, Benjamini E, Wistar R Jr:
Differential reactivity of rheumatoid synovial cells and
serum rheumatoid factors to human immunoglobulin G
subclasses 1 and 3 and their CH3 domains in rheumatoid
arthritis. Arthritis Rheum 30:489-497, 1987
7. Butler VP Jr, Vaughan JH: The reaction of rheumatoid
factor with animal gammaglobulins: quantitative considerations. Immunology 8: 144-147, 1965
8. Normansell DE, Young CW Jr: The IgG subclass specificity of anti-IgG immunoglobulin rheumatoid factors.
Immunochem 12:187-188, 1975
9. Chen PP, Fong S, Carson DA: Rheumatoid factor.
Rheum Dis Clin North Am 13545-568, 1987
10. Dissanayake S, Hay FC, Roitt IM: The binding constants of IgM rheumatoid factors and their univalent
fragments for native and aggregated human IgG. Immunology 323309-318, 1977
1 I . Steward MW, Turner MW, Natvig JB, Gaarder PI: The
binding affinities of rheumatoid factors interacting with
the C gamma 3 homology region of human IgG. Clin Exp
Immunol 15:145-152, 1973
12. Chen PP, Silverman GJ, Liu M-F, Carson DA: Idiotypic
and molecular characterization of human rheumatoid
factors. Chem Immunol48:63-8 I , 1990
13. Ropes MW. Bennett GA, Cobb S, Jacox R, Jessar RA:
1958 revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9:175-176, 1958
14. Robbins DL, Kenny TP, Larrick JW, Wistar R Jr:
Production of human monoclonal rheumatoid factor
secreting hybridomas derived from rheumatoid synovid
cells. Med Sci Res 17:157-159, 1989
15. Larrick JW, Raubitschek AA, Dyer B, Senyk G, Hart S,
Lippman D, Jahnsen M, Wang J, Weintraub H: In vitro
expression of human B cells for production of human
monoclonal antibodies, Human Hybridomas and Monoclonal Antibodies. Edited by E Engleman, S Foung, JW
Larrick, AA Raubitschek. New York, Plenum Press,
1984
16. Mage MG: Preparation of Fab fragments from IgG’s of
different animal species. Methods Enzymol70: 142-150,
1980
17. Michaelsen TL, Natvig JB: Unusual molecular properties of human lgG3 proteins due to an extended hinge
region. J Biol Chem 249:2778-2785, 1974
18. Kabat EA, Wu TT, Reid-Miller M. Perry HM, Gottesman KS, editors: Sequences of Protein of Immunological Interest. Fourth edition. Washington, DC, US Dept.
Health and Human Services, 1987
19. Rappolee DA, Brenner CA, Schultz R,Mark D, Werb Z:
Developmental expression of PDGF, TGF-alpha, and
RSC-DERIVED MONOCLONAL RF
20.
21.
22.
23.
24.
25.
TGF-beta genes in preimplantation embryos. Science
24 1 :1823-1 825, 1988
Larrick JW, Danielsson L, Brenner CA, Abrahamson
M, Fry KE, Borrebaeck CAK: Rapid cloning of rearranged immunoglobulin genes from human hybridoma
cells using mixed primers and the polymerase chain
reaction. Biochem Biophys Res Commun 160:12501256, 1989
Mihovilovic M, Lee JE: An efficient method for sequencing PCR amplified DNA. Biotechniques 7: 14-16,
1989
Messing J , Groneborn B, Muller-Hill B, Hofschneider
PH: Filamentous coliphage MI3 as a cloning vehicle:
insertion of a Hind111 fragment of the lac regulatory
region in M13 replicative form in vitro. Proc Natl Acad
Sci USA 74:3642-3647, 1977
Vieira J, Messing J: The pUC plasmids, a MI3 mp7derived system for insertion mutagenesis and sequencing with synthetic universal primers. Gene 19:259-264,
1982
Devereux J, Haeberli P, Smithies 0:A comprehensive
set of sequence analysis programs for the VAX. Nucleic
Acids Res 12:387-395, 1984
Burastero SE, Casali P, Wilder RL, Notkins AL:
Monoreactive high affinity and polyreactive low affinity
rheumatoid factors are produced by CD5+ B cells from
1195
26.
27.
28.
29.
30.
31.
32.
patients with rheumatoid arthritis. J Exp Med 168:19791992, 1988
Perlmutter RM, Crews ST, Klotz J , Livant D, Siu J,
Hood L: Molecular genetics of anti-carbohydrate antibodies. Ann Immunol (Paris) 135(3:83-88. 1984
Kunkel HG, Agnello V, Joslin FG, Winchester RJ,
Capra JD: Cross-idiotypic specificity among monoclonal
IgM proteins with anti-gammaglobulin activity. J Exp
Med 137:331-342, 1973
Chen PP, Olsen NJ, Yang P-M, Soto-Gil RW, Olee T,
Siminovitch KA, Carson DA: From human autoantibodies to the fetal antibody repertoire to B cell malignancy:
it's a small world after all. Int Rev Immunol (in press)
Pech M, Zachau HG: Immunoglobulin genes of different
subgroups are interdigitated within the Vk locus. NUcleic Acids Res 12:9229-9236, 1984
Newkirk MM, Capra JD: Restricted usage of immunoglobulin variable-region genes in human autoantibodies,
Immunoglobulin Genes. Edited by T Honjo, F Alt, T
Rabbits. New York, Academic Press, 1987
Dersimonian H. Schwartz RS, Barrett KJ, Stollar BD:
Relationship of human variable region heavy chain germline genes to genes encoding anti-DNA antibodies. J
Immunol 13924962501, 1987
Levy S , Mendel E. Kon S, Avnur Z, Levy R: Mutational
hot spots in Ig V region genes of human follicular
lymphomas. J Exp Med 168:475-489, 1988
Документ
Категория
Без категории
Просмотров
0
Размер файла
763 Кб
Теги
factors, molecular, serological, monoclonal, characterization, synovial, derived, human, rheumatoid, cells
1/--страниц
Пожаловаться на содержимое документа