| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
Immunoglobulin Light Chains in Myelomas with Nonamyloid (Randall-Type) Light Chain Deposition Disease
From the Laboratoire d'Immunologie,*
Centre National de
la Recherche Scientifique (CNRS) EP118, Faculté de
Médecine, and Institut Universitaire de
France,**
Limoges, France; Département de
Néphrologie,
Centre Hospitalier
Universitaire La Milètrie, Poitiers, France; Laboratoire
d'Immunologie,
CNRS ESA 6031, CHU La
Milètrie, Poitiers, France; Service d'Anatomie
Pathologique,¶
INSERM U423 Hôpital
Pitié-Salpètrière, Paris; Service
d'Immunochimie,||
Hôpital
Pitié-Salpêtrière, Paris, France; and Department of
Pathology,§
New York University Medical Center,
New York, New York
| Abstract |
|---|
|
|
|---|
type responsible for nonamyloid
light chain deposition
disease. Both patients were affected with severe forms of myeloma
complicated with renal failure. The pathological presentation typically
featured Congo red-negative deposits along tubular basement membranes
but differed somewhat from the typical "Randall-type"
light
chain deposition disease: they lacked the prominent glomerulosclerosis
pattern often featuring nonamyloid
deposits and were associated
with cylinders or myeloma casts. Both protein sequences were deduced
from those of the corresponding complementary DNAs in the bone marrow
plasma cells. For each chain, products of three independent
amplifications by polymerase chain reaction were sequenced and found to
be identical. BOU and RAC
mRNAs had a
normal overall structure consisting of V
2 segments rearranged to
J
2C
2 but displayed a number of unusual features within their
primary sequences. These substitutions are likely responsible for
changes in light chain conformation that promote their aggregation and
deposition along renal tubule basement membranes.
| Introduction |
|---|
|
|
|---|
LCs highly
organized in ß-pleated sheet fibrils. By contrast, LCDD deposits
(mostly of the
type) are amorphous, predominate along the outer
part of basement membranes in the distal tubule and the loop of
Henlé, and often associate with marked nodular
glomerulosclerosis.1-14
To date, many LCs implicated in
AL-amyloidosis have been described, and a large amount of sequence data
has allowed the formulation of hypotheses concerning the special role
of some amino acid substitutions in the process of fibril
formation.15-20
By contrast, only a few LCDD chains have
been studied, all of the
type,4,5,7-9,11-13
and the
process by which LCs aggregate is still largely
unknown.
We herein report on the first two
LCs (BOU and
RAC) implicated in nonamyloid LCDD. BOU and
RAC belonged to the V
2 subgroup and included J
2/C
2
segments. In both cases, the renal lesions differed somewhat from those
seen in most cases of
nonamyloid deposits, because they did not
reveal prominent glomerulosclerosis. In addition, for both patients,
deposits along the tubular membrane basements were accompanied by
tubular lesions resembling cast myeloma.
| Patients and Methods |
|---|
|
|
|---|
Patient BOU, a 59-year-old French man, was referred in
September 1993 for renal failure with anemia (hemoglobin, 8 g/100 ml).
The creatininemia was 750 µmol/L, and proteinuria was abundant (8
g/24 hours) with a predominant
Bence-Jones protein that was also
detectable in serum in small amounts. Hypo-
-globulinemia was marked
(4 g/L). A moderate microscopic hematuria was found. Bone marrow smears
showed 50% plasma cells. Light microscopic examination of a kidney
biopsy showed myeloma cast nephropathy. Eighteen glomeruli were normal
without nodular sclerosis. The interstitium was focally infiltrated by
lymphocytes and plasma cells. Congo red staining was negative.
Immunofluorescence study (Figure 1A)
showed linear deposits along the tubular basement membranes, which only
stained for
chain. Electron microscopic study of a renal medulla
fragment was essentially normal and did not show granular osmiophilic
deposits along tubular basement membranes.
|
Patient RAC, a 48-year-old Guyanese woman, was referred in
September 1996 for persistent renal failure 2 months after a
Plasmodium falciparum infection. The creatinine plasma level
was 547 µmol/L, and there was an abundant proteinuria (7 to 9 g/24
hours) essentially consisting of monoclonal free
chains that were
also detectable in serum by immunofixation. There was no monoclonal IgD
or IgE by immunofixation, and serum polyclonal Ig levels were
moderately affected (9.7 g/L of IgG, 0.3 g/L of IgA, and 0.2 g/L of
IgM). Bone marrow smears showed 13% dystrophic plasma cells and
allowed the diagnosis of myeloma. Light microscopic examination of a
kidney biopsy mainly showed severe and diffuse tubular lesions with a
moderate inflammatory cell infiltrate (Figure 1B)
. Some tubules were
lined by a flattened epithelium; other tubules were obstructed and
dilated by proteic casts; some tubular basement membranes were enlarged
with Congo red-negative material; nodular glomerulosclerosis was not
observed. By immunofluorescence (Figure 1C)
, tubular basement membrane
deposits only stained with anti-
LC antibodies and not with other
anti-Ig chains or anti-complement antisera.
The patient was included in a protocol composed of two autologous bone
marrow transplantations. One month after the first transplantation
(January 1997), creatininemia dropped to 159 µmol/L and proteinuria
to 0.39 g/24 hours. The second transplantation was performed in April
1997; 2 months later the serum creatinine level was 120 µmol/L and
the proteinuria 0.26 g/L, without free LC detectable in the serum.
However, 10 months later, a severe relapse occurred with a heavy plasma
cell infiltrate in bone marrow and with high levels of free
LC in
urine.
RNA Preparation and Reverse Transcription-Polymerase Chain Reaction (PCR) Experiments
Total RNA was prepared by lysis of bone marrow cells in 4 mol/L
guanidine isothiocyanate followed by centrifugation at 170,000 x
g for 18 hours on a 5.7 mol/L cesium chloride pad. Total RNA
was analyzed on a 1% agarose, 1.7 mol/L formaldehyde gel in comparison
with RNA from the human lymphoma cell line IARC 518 producing
normal-sized
mRNA and from the plasmacytoma cell line RPMI 8226
producing a normal-sized
mRNA.21
They were transferred
to nylon sheets and hybridized with either a C
probe, a 2.5-kb
EcoRI genomic fragment containing the human C
exon, or a
C
probe, a 3.5-kb EcoRIHindIII fragment
containing the human C
2 exon. Total RNA was used as a template for
synthesizing single-stranded cDNA using reverse transcriptase and an
oligodeoxythymidylic acid primer (Boehringer Mannheim, Mannheim,
Germany). PCR primers were a 5' primer corresponding to a
V
1-V
2-V
3 consensus leader region (5'-ATGGCCKGSWYYSYTCTCCTC-3')
and a 3' primer complementary to the consensus upstream part of the
C
exons (5'-CTCCCGGGTAGAAGTCACT-3'). Amplification of the cDNAs by
PCR was performed with Taq polymerase (Pharmacia, Uppsala, Sweden)
through 35 cycles consisting of denaturation at 94°C for 30 seconds,
annealing at 53°C for 30 seconds, and elongation at 72°C for 30
seconds.22
After amplification, PCR products were
fractionated on 1.2% agarose gels and sequenced by the dideoxy
termination method,23
using Taq polymerase and an automated
laser fluorescent DNA sequencer (Perkin-Elmer, Branchburg, NJ).
Protein Analysis
Serum and urinary proteins were analyzed by conventional agarose gel zonal electrophoresis in nondenaturing conditions and by immunofixation.
For patient RAC, urinary proteins were purified using diethylaminoethyl-Trisacryl chromatography in 10 mmol/L Tris, pH 8.0, on a 0 to 0.3 mol/L sodium chloride gradient, followed by a second chromatography on Sephadex G100 in 0.1 mol/L Tris-0.5 mol/L NaCl, pH 8.0. An electrophoresis in nondenaturing conditions then showed that the protein was mainly present in dimeric form.
Purified urinary RAC protein (3 mg) was dissolved in 1 ml of 0.2 mol/L NH4CO3 and digested for 1 hour at 37°C with trypsin (modified, sequencing grade; EC 3.4.21.4) (Boehringer Mannheim) at an enzyme/protein ratio of 1:100 (w/w).
Proteolysis was terminated by freezing and freeze-drying. The resulting peptides were separated by reverse-phase high-performance liquid chromatography using a Vydac C18 (The Separations Group, Hesperia, CA) column (218TP52; 0.21 x 25 cm) and a 60-minute 0 to 60% linear gradient of acetonitrile-water (pH 2.1) at a flow rate of 0.2 ml/min. Automated Edman degradation sequence analysis of purified peptides was carried out on a 477A protein-peptide sequencer, and the resulting phenylthiohydantoin amino acid derivatives were identified using the online 120A PTH analyzer (Applied Biosystems, Foster City, CA).
| Results |
|---|
|
|
|---|
LC mRNAs by Northern blotting:
blots hybridized with a C
probe gave a strong signal at the expected
correct size in both cases, although hybridization with a C
probe
yielded no signal in the lanes corresponding to myeloma samples (Figure 2)
-producing cells.
|
+21, +45
+60, +66
+88, and +156
+166),
which fell in complete agreement with the protein sequence deduced from
RAC cDNA. From BOU and RAC cDNA
nucleotide sequences, we could deduce complete amino acid sequences and
assign both LCs to the V
2 subgroup. In both cases, the V region was
normally rearranged to the J
2C
2 segment, which did not present
any sequence abnormality. On the contrary, comparison of V
sequences
with previously reported
chains pointed out several unusual
features consisting of the appearance of residues never or rarely found
at these positions (Figure 3)
|
2
proteins were noticed: Ser+2
Ala, Val+27F
Leu, Leu+46
Ile,
Ile+48
Leu, Ser+65
Phe, Ser/Cys+89
Gly, Ser+90
Leu,
Ala+92
Val, Ser/Asn+95
Arg, and Thr95A
Leu. Some other
substitutions have been rarely observed among V
2 proteins:
Ala+43
Val and Thr+70
Ala. Two other rare substitutions were
encoded at the VJ junction (Val+96
Trp) and within J
2
(Lys103
Arg). Another feature of the BOU sequence is a
three-codon deletion within CDR1 involving codons +28 to +30.
In protein RAC, unique substitutions included Thr+5
Val,
Pro+7
Leu, Thr/Ser+27
Gly, Ser+29
Thr, Val+33
Leu,
Ile+48
Leu, Thr+70
Ser, and Ala+84
Gly.
Rare substitutions included Gly+28
Thr, Asn+31
Lys, and
Lys+42
Ile. Two other rare substitutions were common with BOU,
Val+96
Trp at the VJ junction and Lys103
Arg within J
2.
| Discussion |
|---|
|
|
|---|
type and
with an overrepresentation of the V
4 subgroup, have been sequenced
to date.4,5,7-12
We herein describe two
chains forming
amorphous deposits in patients' kidneys. The BOU and
RAC LCs were responsible for nonamyloid LC deposits stacked
to tubular membrane basements; both cases lacked the prominent
glomerulosclerosis that often features Randall-type
nonamyloid
LCDD. Indeed, it has been previously noticed that morphological changes
were often less severe and nodular sclerosis was less frequent in
LCDD than in
LCDD (13% versus 49% among LCDD cases
reported in the literature).12
Deposits evidenced only by
immunofluorescence, without gross morphological alterations of the
kidney structures, have also been reported in an experimental model of
LCDD that represented an initial step of the disease.24
The constellation of LCDD and myeloma casts that we report in patient
BOU also appear to be unusual and was only present in 2 of 24 patients
with LCDD in the study of Strøm et al.12
Strikingly, both chains are related to the same V
2
subgroup25
and encoded by a V
segment normally
rearranged to J
2 and spliced to C
2.26
However, given
their numerous differences, BOU and RAC
nucleotide sequences likely derived from two different members of the
V
2 gene subgroup: BOU is mostly related to the V
gene V1-7
(90.8% identity),27
whereas RAC is mostly related to DPL12
(90.5% identity).25
None of them encodes any potential
N-glycosylation site, but they display a number of features
unique among V
2 chains: RAC and BOU share a
few amino acid substitutions, and both display several replacements
introducing hydrophobic residues.
The BOU sequence harbors several unusual features within the
variable region, both in framework (FR) and in complementary
determining regions (CDRs). A striking feature of BOU is a
three-residue deletion within the CDR1 between the +28 and +30
positions (according to the numbering of Kabat et al28),
which had never been previously found in any
2 chain. This
three-codon deletion shortens the L1 binding loop and may change its
conformation, which is normally helical in
chains.29
Apart from the CDR1, important changes were also noticed in the CDR3
region, with four unique and two rare amino acid substitutions. In
particular, Ser+90
Leu, Ala+92
Val, Ser/Asn+95
Arg, and
Thr+95a
Leu may disturb the CDR3 structure by strongly increasing the
hydrophobicity of the L3 binding loop.29
As for the FR
regions, unique or rare substitutions altogether increasing
hydrophobicity include: Ser+2
Ala in the FR1, Ala+43
Val in the
FR2, and Ser+65
Phe and Thr+70
Ala in the FR3. The Leu+46
Ile and
Ile+48
Leu unique substitutions probably have a limited effect on the
protein structure. However, it is noticeable that the Ile+48
Leu
substitution is common to BOU and RAC and affects
an invariant Ile supposedly implicated in the maintenance of the L2
loop conformation.29
The presence in both proteins of an
Asn+60 residue, germinally encoded in BOU and resulting from
a substitution in RAC, may also be destabilizing: most LCs
carry a negatively charged residue at this position, in the vicinity of
a critical Arg+61 to Asp+82 salt bridge usually stabilizing V
domains.30
Two other unusual features common to
BOU and RAC are the presence of a Trp+96 residue
at the end of the CDR3 encoded at the VJ junction and a Lys+103
Arg
substitution in the FR4; strikingly, a solvent-exposed Trp features all
LCDD proteins sequenced to date.
All of the other unusual features of protein RAC differ
from those of BOU. Although the CDR2 and CDR3 were
almost completely canonical, the CDR1 presents three unique
substitutions: Ser/Thr+27
Gly; Gly/Ser+29
Thr; and Val+33
Leu and
two rare residues, Thr+28 and Lys+31. Hydrophobic amino acids appear in
the FR1, Thr+5
Val and Pro+7
Leu, the latter affecting an invariant
Pro residue and likely destabilizing the FR1 conformation. Finally, the
RAC chain presents an additional Lys+42
Ile substitution
in the FR2 and two substitutions within the FR3: Thr+70
Ser and
Ala+84
Gly.
The herein reported BOU and RAC
chain
sequences are the first to be documented in Randall-type LCDD, in which
monoclonal
chains strongly predominate. The presence of V region
peculiarities in LC BOU and RAC was expected from
previous sequence analysis of LCDD
chains4,5,7,8,11,12
and from direct evidence for the involvement of V region abnormalities
obtained in a murine model of human
LCDD.24
It is
striking that the herein reported
chains are closely related. By
analogy to the strong implication of
6 chains in AL-amyloidosis, it
may be tempting to speculate that germinally encoded residues of some
V
2 domains may promote LC aggregation and deposition. However, it is
known that the V
2 subgroup is overexpressed in myeloma and
Waldenström macroglobulinemia (28% of monoclonal Ig
, instead
of 3% of Ig
from normal sera),31,32
and it is clear
that not all
2 chains are responsible for LCDD. In addition, a dozen
AL-amyloidosis cases implicating V
2 subgroup LC have been
reported.31,33-35
A likely hypothesis is thus that the
2 germinally encoded sequences may somehow favor LC aggregation (and
in some instances lead to amyloidosis) but that a definitive role in
the process of tissue deposition in patients BOU and
RAC is played by specific amino acid replacements resulting
from somatic mutations and facilitating hydrophobic interactions
between LC monomers or dimers. Such replacements would further promote
the destabilization and deposition of LC. It is indeed striking that
BOU and RAC
chains carry a number of unusual
replacements, several of them introducing hydrophobic residues.
Although crystallization of the proteins will be needed for a
definitive assignment of substitutions to solvent-exposed portions of
the molecules, several such hydrophobic residues located in the
amino-terminal part of the molecule (Ala+2 in BOU and Val+5
and Leu+7 in RAC) or in CDR regions (Val+51, Leu+90, Val+92,
and Leu+95A in BOU, and Leu+33 in RAC) are likely
to altogether increase hydrophobicity at the surface of both LCs and
may play a destabilizing role. Hydrophobic residues may participate in
interchain hydrophobic interactions leading to aggregation and thus
promote tissue deposition. Similarly, substitutions introducing
hydrophobic residues in solvent-exposed portions of LCs have been
previously pointed out for several
LCs implicated in
LCDDs4,8,11,12
as well as for LCs involved in amyloid
fibril formation.15-17
These new sequences extend the
short list of characterized LCs involved in LCDDs cases and will
hopefully help to understand the process by which LCs of decreased
solubility or stability can interact, aggregate in high-order polymers,
and form deposits in tissues. It also appears likely that structural
properties of a given LC are directly correlated to the extent of
visceral deposition occurring in the patient and to the prognosis. In
that sense, the herein reported
LCDD cases with high-level
secretion of the LC and minimal pathological alterations strikingly
contrast with many reported
LCDD cases in which low or even
undetectable secretion of the LC is associated with extensive visceral
deposition and renal morphological
alterations.5,7-9,12-14
| Acknowledgements |
|---|
LC sequences. | Footnotes |
|---|
Supported by grants from the Association pour la Recherche sur le Cancer (Grant 9121), Fondation contre la Leucémie, Ligue Nationale contre le Cancer, and Conseil Régional du Limousin and by National Institutes of Health Grant AR02594 to Dr. Blas Frangione. CD is a recipient of a fellowship from the Association pour la Recherche sur le Cancer.
Accepted for publication April 15, 1998.
| References |
|---|
|
|
|---|
Sci, the Bence Jones protein isolated from a patient with light chain deposition disease. Biochim Biophys Acta 1991, 1097:177-182[Medline]
chain of the V
IV subgroup in the kidney and plasma cells in light chain deposition disease. J Clin Invest 1991, 87:2186-2190
chain in myeloma with light chain deposition disease. Clin Exp Immunol 1992, 87:122-126[Medline]
I subgroup (ISE) in light chain deposition disease. Clin Exp Immunol 1993, 91:506-509[Medline]
IV subgroup in light chain deposition disease. Immunol Lett 1994, 42:63-66[Medline]
III immunoglobulin light chain in myeloma with light chain deposition disease. Clin Exp Immunol 1996, 106:357-361[Medline]
gene segments. Eur J Immunol 1993, 23:1456-1461[Medline]
light chains. J Immunol 1987, 139:824-830[Abstract]
-light chain variable region subgroups in multiple myeloma, AL amyloidosis and Waldenström macroglobulinemia. Clin Immunol Immunopathol 1994, 71:183-189[Medline]
light-chain amyloid fibril protein AR. Biochem J 1981, 195:561-572[Medline]
This article has been cited by other articles:
![]() |
S. M. Korbet and M. M. Schwartz Multiple Myeloma J. Am. Soc. Nephrol., September 1, 2006; 17(9): 2533 - 2545. [Full Text] [PDF] |
||||
![]() |
C. Sirac, F. Bridoux, C. Carrion, O. Devuyst, B. Fernandez, J.-M. Goujon, C. E. Hamel, J.-C. Aldigier, G. Touchard, and M. Cogne Role of the monoclonal {kappa} chain V domain and reversibility of renal damage in a transgenic model of acquired Fanconi syndrome Blood, July 15, 2006; 108(2): 536 - 543. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Decourt, H. R. Galea, C. Sirac, and M. Cogne Immunologic basis for the rare occurrence of true nonsecretory plasma cell dyscrasias J. Leukoc. Biol., September 1, 2004; 76(3): 528 - 536. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. Vidal, F. Goni, F. Stevens, P. Aucouturier, A. Kumar, B. Frangione, J. Ghiso, and G. Gallo Somatic Mutations of the L12a Gene in V-{kappa}1 Light Chain Deposition Disease : Potential Effects on Aberrant Protein Conformation andDeposition Am. J. Pathol., December 1, 1999; 155(6): 2009 - 2017. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Decourt, A. Rocca, F. Bridoux, F. Vrtovsnik, J.L. Preud'homme, M. Cogne, and G. Touchard Mutational Analysis in Murine Models for Myeloma-Associated Fanconi's Syndrome or Cast Myeloma Nephropathy Blood, November 15, 1999; 94(10): 3559 - 3566. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |