Plasma bioavailable interleukin-6 is elevated in human immunodeficiency virus-infected patients who experience herpesvirus-associated immune restoration disease after start of highly active antiretroviral therapy

This study compared plasma bioavailable interleukin (IL)-6 levels in 3 groups: human immunodeficiency virus (HIV)-infected patients who had a human herpesvirus (HHV)-associated immune restoration disease (IRD) during highly active antiretroviral therapy (HAART); patients who experienced an IRD initiated by Mycobacterium avium complex, hepatitis C virus, or human papillomavirus; and control patients who had uneventful immune reconstitution. Total IL-6, soluble IL-6 receptor (sIL-6R), and soluble gp130 were measured by ELISA, and levels of free IL-6 and sIL-6/IL-6R complex were modeled mathematically. Persons who had an HHV-associated IRD had increased plasma bioavailable IL-6 before HAART, compared with patients who experienced a non-HHV-associated IRD and with control patients, and their plasma bioavailable IL-6 increased progressively over 3-4 years of treatment. Increased IL-6 production may be a feature of HAART-induced restoration of immune responses to HHV infections and may have long-term immunopathologic consequences.

Growth and immortalization of human myeloma cells in immunodeficient severe combined immunodeficiency mice: a preclinical model

Human multiple myeloma (MM) purified tumour cells readily undergo apoptosis in vitro. Interleukin 6 (IL-6), a main growth factor of tumour cells, has enabled the development of IL-6-dependent MM cell lines. Recently, we developed anti-gp130 monoclonal antibodies (mAbs), two of which (B1 + I2) were able to dimerize gp130 and replace IL-6 in vitro. We show here that the injection of B1 + I2 IL-6 agonistic mAbs via the inguinal subcutaneous (SC) route efficiently produced tumours in severe combined immunodeficiency (SCID) mice grafted with IL-6-dependent myeloma cell lines compared with either the intraperitoneal (IP) or abdominal surgical bursa (SB) routes. The SC tumour graft, together with Matrigel and vascular endothelial growth factor (VEGF), leads to a strong vascularization and early detection of serum human immunoglobulins (huIgs). SCID mice treated with B1 + I2 mAbs were injected with fresh MM cells from five patients, four of whom had consistent levels of huIgs, and tumour growth was present in two. For one patient, tumour plasma cells that were passed several times subcutaneously in new SCID mice, still expressed their initial markers after several months. They remained unable to grow in vitro in the presence of B1 + I2 or IL-6. The nature of the SCID factors involved and the triggered genes are under investigation.

Optimizing therapeutic strategies to inhibit circulating soluble target molecules with monoclonal antibodies: example of the soluble IL-6 receptors

Therapeutic targeting of soluble molecules such as cytokines can be achieved with monoclonal antibodies (mAb). Anti-IL-6 mAb have been shown to form circulating complexes, resulting in the increase of the half-life of the cytokine in vivo. In IL-6-related diseases, the soluble human IL-6 receptors (shIL-6R), which have been shown to possess strong agonist activity, circulate in the plasma at a high concentration and must be neutralized. Their clearance was studied in mice that had been made to express circulating shIL-6R after i.p. grafting of mouse thymoma cells transfected with a gene coding for shIL-6R, treated with various anti-shIL-6R mAb recognizing different epitopes of the molecule. Injection of one anti-hIL-6R mAb stabilized the short-lived hIL-6R and led to their accumulation. The same result was observed when two mAb directed against two different epitopes of the hIL-6R were used. Clearance of the receptors was only achieved when three mAb specific for three different epitopes were injected. A permanent clearing of the hIL-6R could be obtained by repeated injections of the clearing mixture. No correlation was found between the ability of the mAb to clear the sIL-6R and to immunoprecipitate them in agarose gel. The F(ab')2 fragments lost the clearing ability of the intact mAb. These results clearly show that therapeutic clearance of sIL-6R by mAb need at least three mAb directed against three different epitopes of the molecule, a conclusion which is likely to apply for clearing any soluble target molecule.

Insulin-like growth factor induces the survival and proliferation of myeloma cells through an interleukin-6-independent transduction pathway

Multiple myeloma (MM) is a B-cell neoplasia that is associated with an increased level of bone resorption. One important mediator of bone remodelling, insulin-like growth factor (IGF-I), has been shown to stimulate the proliferation of human myeloma cells. However, the mechanisms of action of IGF-I in these cells have not been determined. Using interleukin (IL)-6-dependent myeloma cell lines, we show IGF-I to be as potent a survival and proliferation factor as IL-6. We demonstrated that IGF-I functions independently of the IL-6 transducer gp130 and that these two cytokines have additive effects. Moreover, inhibition of the IGF-I pathway did not modulate the proliferative effect of IL-6. Accordingly, we found that IL-6 and IGF-I activated distinct downstream signalling molecules: IL-6 activated STAT3 phosphorylation, whereas IGF-I treatment resulted in the phosphorylation of IRS-1. Interestingly, these signalling pathways appear to converge as both cytokines activated the ras/MAPK pathway. Thus, IGF-I acts as a potent survival and proliferation factor for myeloma cells by stimulating an IL-6-independent signalling cascade. These data, together with the finding that, in vivo, IGF-I is normally expressed in close proximity to myeloma cells within the bone matrix, strongly suggest a role for this cytokine in the pathophysiology of multiple myeloma.

The Mi15 monoclonal antibody (anti-syndecan-1) is a reliable marker for quantifying plasma cells in paraffin-embedded bone marrow biopsy specimens

Plasmocyte selective monoclonal antibodies (MAb) recognizing syndecan-1 have recently been described. They belong to a new cluster, CD138. Using the MAb MI15, we investigated the expression of syndecan-1 in routinely paraffin-embedded tissues. Nontumoral lymph nodes (25 cases) and bone marrow biopsy specimens (63 cases) showed strong membrane staining of plasma cells only, allowing accurate analysis of the nuclear structure. The MI15 positivity correlated with kappa and lambda light chain expression in the cytoplasm. The percentages of plasma cells calculated in bone marrow biopsy specimens after MI15 staining were, respectively, 2.1% (range, 1% to 4%) in normal bone marrows, 8.5% (range, 5 to 17) in reactive plasmocytosis, and 4.66% in monoclonal gammapathy of undetermined significance (MGUS) patients (range, 1 to 13), in the same range but slightly higher than those obtained on smears or on hematoxylin and eosin (H&E)-stained sections. In multiple myeloma (40 cases), all plasma cell types were marked, and Mi15 MAb gave additional information in 8 of 40 (20%) patients. In lymph nodes, Mi15 MAb reacted with Reed-Sternberg cells of classical Hodgkin's disease in 23 of 31 cases (74%) with variable intensity. In contrast, nodular lymphocyte predominance Hodgkin's disease (10 cases), most B cell lymphomas (88 of 107 cases) and all T cell lymphomas (30 cases) were negative. In B cell lymphomas, plasmocytomas (8 cases), plasmocytic lymphomas (2 cases), and 5 of 13 cases of immunoblastic lymphoma with plasmocytoid differentiation were stained. In lymphoplasmocytoid lymphomas (4 lymph nodes and 20 bone marrow biopsy specimens), only mature plasma cells were positive. Moreover, a wide distribution of syndecan-1 was observed in normal and tumoral epithelial tissues. Finally, Mi15 MAb appears to be a reliable marker for identifying and quantifying normal and tumoral plasma cells in paraffin-embedded bone marrow and lymph node samples.

Interleukin-6 receptor signaling. II. Bio-availability of interleukin-6 in serum

Interleukin-6 (IL-6) is used as a growth factor by various tumor cells. It binds to a gp80 specific receptor (IL-6R) and then to a gp130 transducing chain. Both receptor chains are released as soluble functional proteins which circulate in biological fluids. With a view to studying the physiological role of these soluble receptors, both proteins were purified from human plasma. Surface plasmon resonance was used to measure the kinetic constants of equilibria between IL-6 and natural sIL-6R, and between the IL-6/sIL-6R complex and soluble gp130. Kd values were found to be 0. 9 and 2.3 nM respectively. Soluble natural IL-6R and gp130 were also found to interact with a Kd of 2.8 nM in the absence of IL-6. By using these Kd values, a mathematical simulation predicted that 1) within a large range of IL-6, sIL-6R and sgp130 concentrations, free IL-6 represents 30% of the total circulating cytokine, 2) sIL-6R overconcentrations lead to dramatic changes of the concentration of free IL-6, 3) increased concentrations of sgp130 should produce an efficient buffering effect on the IL-6/sIL-6R complex without incidence on the level of free IL-6. According to this model, the IL-6/sIL-6R complex appears to be an important support of IL-6 signaling in the most commonly encountered in vivo situations. The concentration of this complex is directly under the control of the concentration of sIL-6R; its bio-availability should be efficiently buffered by increased sgp130 concentrations.

Interleukin-6 receptor signaling. I. gp80 and gp130 receptor interaction in the absence of interleukin-6

Interleukin-6 (IL-6) is used as a growth factor by various tumor cells. It binds to a gp80 specific receptor (IL-6R) and then to a gp130 transducing chain. Both receptor chains are released as soluble functional proteins which circulate in biological fluids. To study the physiological role of these soluble receptors, both proteins were purified from human plasma and the kinetic constants of equilibria between IL-6 and its natural soluble IL-6R (sIL-6R) and gp130 receptor (sgp130) were measured using surface plasmon resonance analysis. Unexpectedly, natural sIL-6R and natural sgp130 were found to interact (Kd = 2.8 nM) in the absence of IL-6. No interaction was seen between the recombinant soluble receptors or between either natural soluble receptor and its recombinant partner. This binary complex was not due to copurification of IL-6 and was detected in human plasma of healthy donors. It results from either direct interaction between the two natural soluble receptors or indirect binding mediated by a yet unidentified copurified plasma molecule playing the role of an IL-6 antagonist. Once formed, the binary complex was found to be unable to bind IL-6. Soluble gp130 had already been shown to inhibit IL-6 signaling by inactivating the IL-6/IL-6R complex. In addition we show that, in the absence of IL-6, circulating natural sgp130 is able to inhibit directly the circulating sIL-6R that is a strong synergic molecule of IL-6 signaling.

Dimerization and activation of the common transducing chain (gp130) of the cytokines of the IL-6 family by mAb

The objective of our research was to study the mechanisms of activation of mAb against the gp130 transducer chain common to the IL-6 cytokine family. It has been found that among the 56 anti-gp130 available worldwide, none was able to activate the growth of IL-6-dependent myeloma cell lines. When certain of them were associated in pairs they allowed the cells to grow; alone, they were inhibitory. The same activation was also obtained by cross-linking certain anti-gp130 mAb on the cell membrane with a goat anti-mouse Ig antiserum. A bispecific mAb was prepared by the somatic fusion of two hybridomas secreting two mAb whose association was able to activate gp130 signaling; the bispecific mAb was inactive. The activating mAb were able to support long-term proliferation of the IL-6-dependent myeloma cell lines, which indicates that they are potential valuable growth factors of tumor cells and hematopoietic stem cells. When they were injected into SCID mice, they allowed human IL-6-dependent myeloma cell lines to grow, develop tumors and metastasize. By studying the functional epitopes of the cell membrane gp130 receptors, it was shown that the activating mAb induced gp130 dimerization and STAT3 activation, as did IL-6.

Functional interaction of the gp80 and gp130 IL-6 receptors in human B cell malignancies

IL-6, or cytokines of the IL-6 family using gp130 as transducer chain receptor, have been suggested to play a role in certain B lymphoid neoplasia. The presence of cell membrane gp80 and gp130 IL-6 receptors was studied in 98 patients with various leukaemia and non-Hodgkin's malignant lymphoma using flow cytofluorometry and immunohistology. Except neoplasia of immature B cells which expressed neither of the receptors, the majority of B cell tumours expressed one or both of them, mantle cell lymphoma being found to express the highest density of receptors. Using IL-6-dependent XG myeloma cell lines and mAb recognizing various gp80 and gp130 functional epitopes, it has been shown that IL-6 activation leads to a modified expression of some epitopes. In particular, the decrease or the disappearance of a gp130 epitope called A1 signed gp130 dimerization which is the first step of the gp130 activation pathway. Gp80 and gp130 epitope analysis was achieved in 17 of the patients. In four, an epitope phenotype compatible with a cytokine-induced activation was found. The cells of five B-CLL patients which expressed both gp80 and gp130 receptors were incubated with IL-6 to induce activation. In three of the cases they were found to rearrange their receptors in activated forms but not in the two others, showing that cells able to be activated or not can be found. These results confirm that gp130 signalling might play an important role in the pathogenesis of certain B cell neoplasia.

IFN-alpha is a survival factor for human myeloma cells and reduces dexamethasone-induced apoptosis

IFN-alpha is used as a maintenance therapy in patients with multiple myeloma, but its benefit is a matter of controversy. In vitro studies show that IFN-alpha can both stimulate and inhibit myeloma cell proliferation. We have tested the effect of IFN-alpha on the survival of myeloma cell lines and primary plasma cells. IFN-alpha significantly reduced the apoptosis induced by removal of IL-6 in four IL-6-dependent myeloma cell lines. It also reduced the level of apoptosis induced by dexamethasone in these cell lines as well as in purified primary myeloma cells from seven patients. IFN-alpha promoted the survival of myeloma cells, which, following removal of IL-6, were blocked in G1 and died. However, unlike IL-6, IFN-alpha-treated cells remained mainly blocked in the G1 phase of the cycle. While the effects of IL-6 are mediated through stimulation of its gp130 receptor subunit, the IFN-alpha-induced survival of myeloma cells was independent of gp130 transducer activation (as demonstrated using a neutralizing anti-gp130 Ab). However, the signal transduction cascades activated by these two cytokines share at least some common elements, since stimulation with either IFN-alpha or IL-6 resulted in STAT3 phosphorylation. These results indicate that IFN-alpha promotes the survival, but not the proliferation, of myeloma cells, preventing the apoptosis induced by removal of IL-6 or addition of dexamethasone. This survival factor activity may explain the conflicting reports on the effects of IFN-alpha on myeloma cell proliferation.

A gp130 interleukin-6 transducer-dependent SCID model of human multiple myeloma

Agonist antihuman gp130 transducer monoclonal antibodies (MoAbs) were used in SCID mice to grow myeloma cells whose survival and proliferation is dependent on gp130 transducer activation. The agonist anti-gp130 MoAbs neither bound to murine gp130 nor activated murine cells and, as a consequence, did not induce interleukin-6 (IL-6)-related toxicities in mice. They have a 2-week half-life in vivo when injected in the peritoneum. The agonist antibodies made possible the in vivo growth of exogenous IL-6-dependent human myeloma cells as well as that of freshly explanted myeloma cells from 1 patient with secondary plasma cell leukemia. Tumors occurred 4 to 10 weeks after myeloma cell graft and weighed 3 to 5 g. They grew as solid tumors in the peritoneal cavity and metastasized to the different peritoneal organs: liver, pancreas, spleen, and intestine. Tumoral cells were detected in blood and bone marrow of mice grafted with the XG-2 myeloma cells. Tumoral cells grown in SCID mice had kept the phenotypic characteristics of the original tumoral cells and their in vitro growth required the presence of IL-6 or agonist anti-gp130 MoAbs. Myeloma cells from 4 patients with medullary involvement persisted for more than 1 year as judged by detectable circulating human Ig. However, no tumors were detected, suggesting a long-term survival of human myeloma cells without major proliferation. These observations paralleled those made in in vitro cultures as well as the tumor growth pattern in these patients. This gp130 transducer-dependent SCID model of multiple myeloma should be useful to study various therapeutical approaches in multiple myeloma in vivo.

The myeloma cell antigen syndecan-1 is lost by apoptotic myeloma cells

Syndecan-1 is a cell membrane proteoglycan that binds extracellular matrix components and various growth factors. It is expressed only on malignant plasma cells in bone marrow samples from patients with multiple myeloma (MM). Several reports have suggested that syndecan-1 was present only on a part of the myeloma cells. By using either IL-6-dependent myeloma cell lines or primary myeloma cells stained by annexin V, we report here that syndecan-1 was rapidly lost by myeloma cells undergoing apoptosis. In the same experimental conditions, expression of other cell membrane antigens such as CD38, HLA class-I or CD49d on apoptotic myeloma cells was not affected. In addition, we show that syndecan-1 loss was independent of activation of the gp130 IL-6 transducer. Dexamethasone induced a strong apoptosis of myeloma cells associated with the loss of syndecan-1. Finally, by using freshly-explanted tumoural samples, we show that syndecan-1 rapidly disappeared from myeloma cells in association with induction of apoptosis. In conclusion we showed that syndecan-1 is a marker for viable myeloma cells which is rapidly lost by apoptotic cells. These results emphasize the usefulness of anti-syndecan-1 antibodies to purge tumoural cells from haemopoietic grafts or to purify these cells for further manipulations for immuno or gene therapies.

Major role of the soluble interleukin-6/interleukin-6 receptor complex for the proliferation of interleukin-6-dependent human myeloma cell lines

Interleukin-6 (IL-6) is a proinflammatory cytokine which possesses a central growth factor activity for certain tumor cells such as plasma cells in multiple myeloma (MM). Upon binding of IL-6, soluble IL-6 receptor (sIL-6R) has been shown to retain its affinity for IL-6 and to associate with the signal-transducing gp130 chain. Therefore, contrary to the majority of soluble cytokine receptors, it plays an agonist role in IL-6 signaling. In order to test its physiological importance as compared to that of its membrane counterpart, we studied cells from two myeloma cell lines which need exogenous IL-6 to proliferate and release sIL-6R into their culture supernatant. Using a new culture system where the supernatant recirculated permanently through an anti-IL-6R affinity column, all sIL-6R was removed from the culture medium throughout the culture period. Under these conditions IL-6-dependent cells were unable to grow in the presence of physiological concentrations of IL-6, showing the major role of the sIL-6R for sustaining the proliferation of these cell lines. Increasing IL-6 concentrations well over the physiological values allowed the cells to proliferate again. No effect was seen when sIL-6R was removed from the supernatant of an IL-6-independent myeloma cell line. These results show that the levels of circulating sIL-6R (and thus those of IL-6/sIL-6R complex) are worth looking at in pathologies involving IL-6 hyperactivity.

Expression of the interleukin 6 receptor in primary renal cell carcinoma

AIMS

Interleukin 6 (IL-6) is expressed in the majority of renal cell carcinomas and has an important role in the proliferation of some renal cell carcinoma cell lines. This action is mediated by two membrane proteins, gp80 (the IL-6 receptor; IL-6R), which binds IL-6, and gp130, which transduces the signal. The soluble form of gp80 (sIL-6R) is able to activate gp130 when complexed to the IL-6 molecule. These considerations prompted an investigation of IL-6R expression in this malignancy. IL-6, C reactive protein (CRP), and sIL-6R were also measured in serum and correlated to clinical and pathological features.

METHODS

Immunostaining was performed on cryostat sections from renal cell carcinoma tumours with M91, an anti-IL-6R monoclonal antibody, using the alkaline phosphatase antialkaline phosphatase technique. The proliferation index was measured using the KI-67 monoclonal antibody. CRP, IL-6, and sIL-6R were measured in serum before nephrectomy, using an immunoenzymatic or immunoradiometric assay.

RESULTS

There were significant differences in survival in patients with tumours larger than 8 cm, metastasis at diagnosis, high nuclear grade tumours, detectable serum concentrations of IL-6 (correlated to CRP serum concentration), more than 4% proliferating cells, and the presence of the IL-6R in situ. Furthermore, the serum IL-6 concentration correlated with tumour size and stage. The mean serum sIL-6R concentration was not significantly different from that observed in 40 normal subjects. Tumour IL-6R expression was present in 10 samples. There was a significant association between the presence of the IL-6 receptor in tumours and tumour stage, nuclear grade, proliferation index, and serum IL-6.

CONCLUSIONS

This study revealed the importance of IL-6/CRP and IL-6R expression in situ as potential new prognostic factors and opens the way to new therapeutic strategies in renal cell carcinoma.

Large scale and clinical grade purification of syndecan-1+ malignant plasma cells

For cancer immunotherapy, it is usually necessary to obtain a large number of tumor cells from patients. We have previously reported that syndecan-I is present only on malignant plasma cells in samples from patients with multiple myelomatosis. We report here that this antigen is cleaved by chymopapain. This makes it possible to develop a rapid and clinical grade procedure to purify large numbers of myeloma cells using anti-syndecan-1 mAb, magnetic beads and chymopapain.

Specific inhibition of IL-6 signalling with monoclonal antibodies against the gp130 receptor

A family of cytokines [IL-6, IL-11, oncostatin M (OM), leukaemia inhibitory factor (LIF), ciliary neurotrophic factor (CNTF) and cardiotrophin-1] involved in various inflammatory or tumoral diseases share the same gp130 signal transducer chain. The complex formed with their specific receptors associates with a common transducing gp130 membrane protein (gp130) resulting in the formation of high avidity receptor and activation of tyrosine kinases. With the view of identifying gp130 domains specifically involved in IL-6 signalling, the authors prepared 37 new anti-gp130 mAb and analysed the structure-function relationship of the molecule. By cross-competition ELISA, the mAb were classified in 10 subgroups called A to J. By ELISA and BIAcore analysis, the mAb were found to recognize at least 18 antigenic specificities of the gp130 chain. The mAb reacted against the soluble and the membrane forms of gp130 as well. Their ability to inhibit the proliferation of the human myeloma cell line XG-4 of which the growth is strictly dependent on the presence of either exogenous IL-6, or LIF, or OM, or CNTF was studied. Besides mAb with no evident neutralizing effect (G and H) and mAb which neutralized equally well the activity of all tested cytokines (all mAb of groups A, I and J), some showed a selective effect. Those of group F inhibited also the proliferation induced by the 4 cytokines, but more specifically that dependent on the CNTF. mAb of groups B and E specifically inhibited the growth induced by IL-6, whereas those of group C inhibited that induced by LIF and OM. These results show the presence of different gp130 epitopes specifically involved in the signaling induced by the cytokines of the gp130 family. In ELISA, only mAb of group B and E were found to inhibit the binding of the IL-6-IL-6R complex to gp130, showing that they identified one or two domains of gp130 involved in its interaction with the IL-6-IL-6R complex. Precise identification of this(ese) epitope(s) would be useful to better understand the mechanisms of the IL-6 signalling.

Activation of the gp130 signaling pathway by monoclonal antibodies directed against the gp130 molecule

Six cytokines of the interleukin (IL)-6 family involved in various inflammatory or tumoral diseases share the same gp130 signal transducer chain. We made a panel of anti-gp130 monoclonal antibodies (mAb) to study the structure and function of the gp130 molecule. These mAb recognized different epitopes of the gp130 that we called A to J. Most of the mAb were found to be inhibitors and we studied whether some of them could also induce gp130 activation. When used alone, none of them was able to initiate the proliferation of IL-6-dependent cell lines. However, some particular associations of the mAb were able to induce a proliferative response. mAb B1 could activate the lines in association with F1 or with I2 but not with I1, which in ELISA was similar to I2. In contrast mAb B2, which in ELISA appeared to be very similar to B1, was able to activate the cells in association with I1 but not with F1 or I2. Two other mAb belonging to specificities A and C were found to be activators either in association with I1 only, or with I1 or B2, respectively. These associations of mAb appeared to be nearly as potent activators as IL-6 itself. Although we still have no precise idea of the mechanisms involved, they are interesting tools to study the molecular interactions leading to gp130 activation and, from a practical point of view, valuable growth factors of hematopoietic stem cells.

Mononuclear cell retention in rheumatoid synovial tissue engrafted in severe combined immunodeficient (SCID) mice is up-regulated by tumour necrosis factor-alpha (TNF-alpha) and mediated through intercellular adhesion molecule-1 (ICAM-1)

The aim of this study was to assess regulation of mononuclear cell (MNC) traffic to human synovial tissue by TNF-alpha and IL-1 and the involvement of ICAM-1 in MNC retention in rheumatoid synovial tissue. Human rheumatoid arthritis synovium was engrafted subcutaneously in 6-8 week-old SCID/CB17 mice. Three weeks later, we injected 20 x 10(6) human peripheral blood mononuclear cells (PBMC) previously labelled with 111indium intraperitoneally into mice containing control or cytokine-injected grafts. Total body scintigraphy was performed 72 h postinjection. The graft was removed and immunochemical analysis carried out to assess ICAM-1, vascular cell adhesion molecule-1 (VCAM-1) and E-selectin expression. In some experiments, mice were treated intravenously with 500 micrograms MoAb anti-ICAM-1 (BIRR-1) or an isotype-matched control MoAb before introduction of MNC. TNF-alpha, but not IL-1 alpha, enhanced MNC retention in the rheumatoid synovial graft 72 h post-injection (graft activity 989 +/- 1227 ct/min per 200 pixels or 3.36 +/- 4.16% of initial injected activity versus 411 +/- 157 ct/min per 200 pixels or 1.13 +/- 0.45% in controls; P < 0.03). TNF-alpha enhanced ICAM-1 expression by synovial cells and endothelial cells, whereas VCAM-1 or E-selectin expression was not enhanced on either cell type. After MoAb treatment of ICAM-1, synovial lymphocyte recruitment of TNF-alpha-treated mice decreased significantly to levels below that of control mice (160 +/- 97 ct/min per 200 pixels, 0.54 +/- 0.33%; P < 0.01). Mononuclear cell retention in rheumatoid synovial tissue engrafted into SCID mice was up-regulated by TNF-alpha and blocked by MoAb to ICAM-1. These results suggest that ICAM-1 is involved in mononuclear cell retention in rheumatoid synovium.

Identification of a novel antigenic structure of the human receptor for interleukin-6 involved in the interaction with the glycoprotein 130 chain

The receptor for interleukin-6 (IL-6) is characterized by a ligand-binding glycoprotein 80 (gp80) transmembrane chain (IL-6R) which associates with a signal-transducer gp130 chain. We previously raised a series of monoclonal antibodies (mAb) recognizing different epitopes of the human IL-6R and interfering with the function of the receptor. One of them, M182, was able to diminish the proliferation of IL-6-dependent plasmacytoma cell lines although it was found unable to inhibit the binding of IL-6 to its receptor. Using an enzyme-linked immunosorbent assay for measuring the binding of IL-6 IL-6R to the gp130 chain, we showed that M182 was directed against a structure directly involved in the IL-6R gp130 interaction. M182 was able to potentiate the inhibitor effect of anti-IL-6R mAB which interfere with the binding of IL-6, leading to complete inhibition of the proliferation of IL-6-dependent cell lines. M182 was also found to synergize with inhibitory anti-IL-6 mAb. Therefore this structure appears to be an important regulatory domain of the IL-6R and a valuable target for inhibiting IL-6 signalling.

Participation of two Ser-Ser-Phe-Tyr repeats in interleukin-6 (IL-6)-binding sites of the human IL-6 receptor

The alpha-subunit of interleukin-6 (IL-6) receptor is a member of the hematopoietin receptor family. The alignment of its amino acid sequence with those of other members of this family (human somatotropin receptor/murine IL-3 receptor beta and human IL-2 receptor beta) has suggested that amino acids included in two SSFY repeats found in each of its hematopoietin receptor domains, contribute to the binding of the ligand. The involvement of these amino acids in IL-6 binding and signal transduction was studied by site-directed mutagenesis and molecular modelling. We present a computer-derived three-dimensional model of the IL-6/IL-6 receptor complex based on the structure of the human somatotropin/human somatotropin receptor complex. This model allowed the location of distinct regions important for IL-6 and gp130 binding. We show that some of the residues included in the SSFY repeats located in our IL-6 receptor model in the loops between beta-strands E and F of domain-I and B' and C', of domain-II, participate in the formation of a major IL-6-binding site. These residues are necessary for IL-6 and gp130 binding and for signal transduction. Using our IL-6 receptor mutants we mapped the epitopes of our anti-(IL-6 receptor) neutralising monoclonal antibodies to these residues. Our results demonstrate that a generic hematopoietin receptor family structural module can be used for the study of both alpha and beta receptor subunits belonging to this family.

IgA isotype rheumatoid factor in rheumatoid arthritis: clinical implications

OBJECTIVES

To study the clinical, biological and radiological characteristics of RA with a predominant increase of IgA isotype rheumatoid factor (IgA-RF) over IgM-RF.

METHODS

The presence of IgA-RF was determined by a sandwich-type ELISA with an antibody against the human IgA used to capture the immunoglobulin. Associated RF activity was revealed with a peroxidase-conjugated human IgG Fc fragment. Forty-nine RA patients were studied, of whom 19 had an increase in IgA-RF (38%). The control group comprised 30 RA patients without IgA-RF.

RESULTS

None of the patients had isolated IgA-RF. In the selected 19 RA patients, the OD of IgA-RF (0.971 +/- 0.62 U) was higher than the IgM-RF (mean OD: 0.675 +/- 0.522 U). A statistically significant correlation was found between IgA-RF and IgM-RF (r = 0.64, p < 0.0001). No correlation was noted between the IgA concentration and the IgA-RF titer. The two groups were comparable for age, disease duration, sex ratio and previous DMARD use. We observed that patients with RA associated with increased IgA-RF more often had the sicca syndrome, but no other extra-articular features. RA patients with IgA RF also had more erosive disease: the mean Larsen score at the hand and wrist was 76 (SD = 68) versus 54 (SD = 60) in the controls, p < 0.02. Replacement surgery for the hip or knee was necessary in 47% of the RA patients with IgA-RF, versus 13% in the controls, p < 0.01. No association of IgA-RF with disease activity was noted.

CONCLUSION

Our study showed that RA patients with a predominant increase of IgA-RF had a more erosive disease and a high frequency of associated sicca syndrome.

A monoclonal antibody directed against a granule membrane glycoprotein (GMP-140/PADGEM, P-selectin, CD62P) inhibits ristocetin-induced platelet aggregation

P-selectin (also called CD62, GMP-140, PADGEM, CD62P) is a recently described member of a family of vascular adhesion receptors expressed by activated platelets and endothelial cells that are involved in leucocyte cell adhesion. The aim of this study was to characterize a new monoclonal antibody (LYP7) directed against activated human blood platelets that inhibits ristocetin-induced platelet aggregation. Immunoadsorbent affinity chromatography and immunoprecipitation studies showed that LYP7 (IgG1) bound a surface-labelled glycoprotein (GP) which changed its apparent molecular mass (M(r)) on reduction from 138 kD (situated below GPIIb) to 148 kD (above GPIIb alpha). LYP7 and S12, a monoclonal antibody directed against P-selectin immunoprecipitated the same band. Using ELISA assay, purified P-selectin was shown to bind LYP7 and S12 monoclonal antibodies. Binding sites of 125I-labelled LYP7, which was greatly increased on thrombin-stimulated (2 U/ml) washed platelets (10825 +/- 2886, mean +/- SD) Kd = 1.5 +/- 0.5 nM) compared to resting platelets (2801 +/- 1278, mean +/- SD) (Kd = 1.5 +/- 0.6 nM), was found to be normal on thrombin-stimulated platelets taken from a patient with grey platelet syndrome or a patient with Glanzmann thrombasthenia. LYP7 (IgG1, F(ab')2 or Fab fragments) inhibited ristocetin-induced platelet aggregation of platelets in a dose-dependent fashion without affecting the binding of von Willebrand (vWf) factor. However, agglutination of formaldehyde-fixed platelets induced by ristocetin was not affected by monoclonal antibody LYP7. In addition, the binding of thrombin-activated platelets to neutrophils was inhibited by monoclonal antibody LYP7. These results strongly suggest that P-selectin, by promoting cell-cell contact, may play an active role in platelet-platelet interactions.

In vivo migration of radiolabelled lymphocytes in rheumatoid synovial tissue engrafted in SCID mice: implication of beta 2 and beta 7-integrin

OBJECTIVE

Integrin-adressin binding is a critical step in lymphocyte attachment to target tissues. The lymphocyte function associated antigen (LFA-1)/intercellular adhesion molecule-1 (ICAM-1) pathway has been shown to be involved in the homing of lymphocytes to arthritic joints in animal models. The mucosal recognition system [alpha E beta 7/E-cadherin, alpha 4 beta 7/mucosal vascular adressin cellular adhesion molecule 1 (MADCAM-1)] has been implicated in the autoimmune process of nonobese diabetic mice and in rheumatoid arthritis (RA). We developed a model for in vivo study of radiolabelled lymphocyte circulation and attachment to human engrafted rheumatoid synovium, and studied the involvement of LFA-1 and alpha E beta 7 integrin.

METHODS

We engrafted human RA or osteoarthritis (OA) synovium subcutaneously in 6-week-old SCID/CB17 mice. Three weeks later, we injected intraperitoneally 20 x 10(6) human peripheral blood lymphocytes (PBL) labelled with 3 mCi 99mtechnetium hexamethyl propylenamine oxime. A mouse total body scintigraphy was obtained 20 h postinjection. The same protocol was performed after pretreatment of the PBL with monoclonal antibodies (Mab) against CD11a (25-3) or against alpha E beta 7 human mucosyl lymphocyte marker 1.

RESULTS

PBL migrated in the rheumatoid synovial graft 20 h postinjection (activity in the region of interest of the graft: 7699 +/- 4383 cpm/200 pixel or 4.43 +/- 2.65% of initial activity) versus OA engrafted synovial tissue (1453 +/- 1137 or 0.74 +/- 0.6% of initial activity), p = 0.007. The homing to the engrafted rheumatoid synovial tissue of PBL from healthy subjects was not significantly different from the migration of PBL from patients with RA. A Mab against alpha E beta 7 significantly decreased lymphocyte attachment to rheumatoid synovial tissue (3094 +/- 3808 cpm/200 pixel or 2.65 +/- 2.4% of injected activity), p < 0.03. The same results were obtained with Mab against CD11a (5007 +/- 4190 cpm/200 pixel or 2.27 +/- 1.2%), p < 0.01. Our results show increased blood lymphocytes homing to rheumatoid synovial tissue engrafted in SCID mice versus OA tissue.

CONCLUSION

The data suggest that both LFA-1 and mucosal recognition integrin alpha E beta 7 are involved in lymphocyte binding to target tissues in RA.

In vivo migration of tonsil lymphocytes in rheumatoid synovial tissue engrafted in SCID mice: involvement of LFA-1

Integrin-adressin binding is a critical step in lymphocte attachment to target tissues. The mucosal recognition systems (alpha E beta 7, alpha 4 beta 7, MADcam-1) have been implicated in the autoimmune process in rheumatoid arthritis. We developed a model for in vivo study of radio-labelled lymphocyte circulation and their attachment to human rheumatoid synovium. We studied the homing of tonsil lymphocytes, considered as mucosal lymphocytes, and the involvement of alpha E beta 7 integrin and LFA1 in the homing of tonsil lymphocytes. We engrafted human rheumatoid synovium subcutaneously in 6 week old SCID CB17 mice. Three weeks later, we injected intraperitoneally 20 IO6 human peripheral blood or tonsil mononuclear cells, previously labelled with 3 mCFi HMPAO-99mTc. A mouse total body scintigram was obtained 20 h postinjection. The same protocol was performed after treatment of the MNC and mAb against LFA-1 (CD11a) or alpha E beta 7 (CD103). Tonsil MNC retention in the rheumatoid synovial graft 20 h post-injection was enhanced compared to blood MNC (12731 +/- 8297cpm/200 pixel) versus 5982 +/- 4713cpm/200 pixel, p < 0.05). A monoclonal antibody against LFA 1 decreased the activity in the graft (4152 +/- 1287 cpm/200 pixel), p < 0.05. No significant difference in tonsil MNC attachment to rheumatoid synovial tissue was observed with a mAb against alpha E beta 7 (8057 +/- 5009 cpm/200 pixel). Our results showed an increase in radiolabelled mucosal MNC migration in synovial tissue engrafted in SCID mice compared with blood MNC. Moreover, the date suggest that LFA-1 but not the alpha E beta 7 integrin is involved in tonsil MNC binding to synovial tissue in RA.

Detection of membrane and soluble interleukin-6 receptor in lymphoid malignancies

We studied the membrane expression of the gp80 chain of IL-6 receptor (IL-6R) by quantitative flow cytometry in chronic lymphocytic leukaemia (CLL) and leukaemic centrocytic lymphoma using a panel of seven monoclonal antibodies. IL-6R was detected in 18/26 CLL cases and 4/7 lymphoma cases, with a mean antigen density < 3000 molecules/cell. Multiple labelling experiments confirmed the IL-6R expression by neoplastic cells. Specific mRNA was found by RT-PCR in neoplastic cells. A specific ELISA test was designed using two anti-IL-6 receptor MAbs to measure the serum soluble IL-6R (sIL-6R) in CLL (n = 48). B-cell non-Hodgkin's lymphoma (NHL; n = 40), and monoclonal gammopathy (MG; n = 32). SIL-6R was higher in CLL (170 +/- 12.6 ng/ml) in NHL (160 +/- 12 ng/ml) and MG patients (183 +/- 23 ng/ml) than in age-matched controls (100 +/- 5.6 ng/ml; P < 0.001) and higher in high-grade than low-grade NHL. No correlation was noted with a previous treatment. Among CLL cases the patients classified as stage B according to the Binet's staging of the disease had the highest sIL-6R values, thus suggesting a link with tumour cell mass.

Biologic effects of anti-interleukin-6 murine monoclonal antibody in advanced multiple myeloma

In patients with advanced multiple myeloma (MM) there is an excess of production of interleukin-6 (IL-6) in vivo, and elevated serum levels are associated with plasmablastic proliferative activity and short survival. These data prompted us to perform a clinical trial with a murine anti-IL-6 monoclonal antibody (MoAb) to neutralize the excess of this putatively deleterious factor in these patients. Ten MM patients with extramedullary involvement frequently were treated with anti-IL-6 MoAb. The MoAb was administered intravenously to 9 patients; 1 patient with malignant pleural effusion received intrapleural therapy. Of the 3 patients who succumbed to progressive MM after less than 1 week of treatment (including the only 1 treated locally), 2 with evaluable data exhibited marked inhibition of plasmablastic proliferation. Among the 7 patients remaining more homogeneous receiving the anti-IL-6 MoAb for more than 1 week, 3 had objective antiproliferative effect marked by a significant reduction of the myeloma cell labelling index within the bone marrow. One of these 3 patients achieved a 30% regression of tumor mass. However, none of the patients studied achieved remission or improved outcome as judged by standard clinical criteria. Of major interest, objective antiproliferative effects were associated with complete inhibition of C-reactive protein (CRP) synthesis and low daily IL-6 production in vivo. On the other hand, the lack of effect in 4 patients was associated with a higher IL-6 production and inability of the MoAb to neutralize it. Anti-IL-6 was also associated with resolution of low-grade fever in all the patients and with worsening thrombocytopenia and mild neutropenia. The generation of human antibodies to Fc fragment of the murine anti-IL-6 MoAb observed in 1 patient was associated with dramatic progression. These data show that anti-IL-6 MoAb can suppress the proliferation of myeloma cells and underscore the biologic role of IL-6 for myeloma growth in vivo. Furthermore, suppression of CRP and worsening of neutropenia/thrombocytopenia both indicate that IL-6 is critically involved in acute-phase responses and granulopoiesis/thrombopoiesis.

Immunomodulating IL-6 activity by murine monoclonal antibodies

The human anti-mouse immunoglobulin antibody (HAMA) response, which occurs frequently after injection of murine monoclonal antibodies (MAb) directed against cellular targets, has been reported extensively in several studies. We analysed here HAMA in 12 patients (six with multiple myeloma, MM, and six with metastatic renal cell carcinoma, MRCC) who were treated with B-E8, an IgG1 MAb against interleukin-6 (IL-6). Efficiency of the treatment was evidenced by the drop in the serum levels of C reactive protein (CRP), of which the in vivo production is under the control of IL-6. Three patients with MM and the six patients with MRCC became immunized to the injected MAb. HAMA appeared between days 7 and 15 after the beginning of the treatment. The nine patients made IgG antibodies; four also made IgM. All of immunized patients made anti-idiotype antibodies specific to B-E8. Two of them also developed HAMA directed to murine IgG1 isotype; in these two patients B-E8 MAb cleared rapidly from the circulation with loss of treatment efficiency. In the patients who developed only anti-idiotype antibodies, serum levels of B-E8 remained unchanged and CRP production remained inhibited, indicating that treatment efficiency was not affected by the presence of HAMA. Circulating B-E8 MAb were still able to bind to IL-6 and to inhibit IL-6-independent proliferation despite the presence of anti-idiotypic HAMA. Therefore, in contrast to HAMA against MAb directed against cellular targets, HAMA against anti-IL-6 MAb idiotopes led neither to clearance nor to functional inactivation of the injected MAb.(ABSTRACT TRUNCATED AT 250 WORDS)

IL-6-induced changes in synthesis of alpha 1-acid glycoprotein in human hepatoma Hep3B cells are distinctively regulated by monoclonal antibodies directed against different epitopes of IL-6 receptor (gp80)

The synthesis of the human acute-phase alpha 1-acid glycoprotein (AGP) is primarily controlled by IL-6 and IL-1 in liver cells. In the present study, monoclonal antibodies against human gp80 interleukin-6 receptor (IL-6R) were utilized to study the role of the IL-6R in the control of the IL-6-induced AGP synthesis in the human hepatoma Hep3B cell line. Two of the 4 MAbs used in this study, M164 and M195, identified 2 different epitopes involved in IL-6 binding and two others, M91 and M182, recognized epitopes not involved in IL-6 binding. Dose-response experiments indicated that up to 55% of AGP synthesis was inhibited by 10(5) ng/ml of MAbs 164 or 195 when Hep3B cells were treated by IL-6 for 48h. Kinetics of the inhibition of AGP synthesis after addition of anti-IL-6R indicated that the decrease of the IL-6-induced AGP synthesis by Hep3B cells was obtained immediately after the addition of the anti-IL-6R MAbs. Of the two MAbs not involved in IL-6 binding, M91 was unable to interfere with the IL-6-induced AGP synthesis whereas, surprisingly, M182 decreased it by about 25%. Since M182 was also able to interfere with the proliferative response of an IL-6 dependent plasma cell line, our results suggested that M182 may be directed to a structure involved in the IL-6/IL-6R gp130 complex formation.(ABSTRACT TRUNCATED AT 250 WORDS)

Human anti-mouse antibody response to the injection of murine monoclonal antibodies against IL-6

We analysed human anti-mouse antibodies (HAMA) in 12 patients (six with multiple myeloma (MM) and six with metastatic renal cell carcinoma (MRCC) who were treated with B-E8, an IgG1 MoAb against IL-6. Efficiency of the treatment was evidenced by the drop in the serum levels of C-reactive protein (CRP), the in vivo production of which is under the control of IL-6. Three patients with MM and the six patients with MRCC became immunized to the injected MoAb. HAMA appeared between days 7 and 15 after the beginning of the treatment. The nine patients made IgG antibodies; four also made IgM. All immunized patients made anti-idiotype antibodies specific to B-E8. Two of them also developed HAMA directed to murine IgG1 isotype; in these two patients B-E8 MoAb cleared rapidly from the circulation with loss of treatment efficiency. In the patients who developed only anti-idiotype antibodies, serum levels of B-E8 remained unchanged and CRP production remained inhibited, indicating that treatment remained efficient in the presence of HAMA. Circulating B-E8 MoAbs were still able to bind to IL-6 and to inhibit IL-6-dependent proliferation despite the presence of anti-idiotypic HAMA. Therefore, in contrast to HAMA produced against MoAb directed against cellular targets, HAMA against anti-IL-6 MoAb idiotopes led neither to clearance nor to functional inactivation of the injected MoAb. This was further shown by resuming the B-E8 treatment with success in a patient who still had anti-idiotypic HAMA.

Administration of an anti-interleukin-6 monoclonal antibody to patients with acquired immunodeficiency syndrome and lymphoma: effect on lymphoma growth and on B clinical symptoms

Increased interleukin-6 (IL-6) production and expression by malignant cells of the IL-6 receptor has been evidenced in a subgroup of non-Hodgkin's lymphomas, suggesting that this cytokine plays a role in lymphoma growth and in B clinical symptoms. In this study, the effect of the administration of an anti-IL-6 monoclonal antibody (MoAb) was analyzed in 11 patients seropositive for human immunodeficiency virus-1 and suffering from an immunoblastic or a polymorphic large-cell lymphoma. The antibody (BE-8, 10 to 40 mg/day) was administered for 21 days. Neutralization of in vivo IL-6 effect was assessed by monitoring C-reactive protein levels in the serum. In 5 patients, the lymphoma progressed during treatment. Among them were the 2 patients in whom endogenous IL-6 effect was not neutralized. Five patients experienced a stabilization, and 1 a partial remission. This effect on lymphoma growth lasted for 8 to 28 weeks. The anti-IL-6 MoAb had a clear effect on lymphoma-associated fever and cachexia. The mean body weight increase was 1.4 +/- 0.5 kg between day 1 and day 21, and reached 12 kg in 120 days in 1 patient who received three courses of treatment. Side effects were a consistent but moderate thrombocytopenia, and an occasional and moderate decrease of neutrophil counts. Immunization against the MoAb was observed in only 2 patients. These results indicate that in some cases of lymphomas growth of malignant cells may be partially IL-6-dependent and that neutralizing endogenous effect of IL-6 completely abrogates B clinical symptoms.

Epitope analysis of human IL-6 receptor gp80 molecule with monoclonal antibodies

Gp80 human IL-6R was studied using 7 murine mAb (M37, M91, M113, M139, M164, M182 and M195) obtained after fusion of splenocytes of Balb/c mice immunised with a mixture of recombinant IL-6 receptor (rIL-6R) and cells from 2 cell lines expressing IL-6R. These were U266, which is IL-6 independent and XG-1 which is IL-6-dependent. In ELISA the 7 mAb reacted against the rIL-6R and against the natural soluble form found in plasma (nIL-6R), which both lack transmembrane and cytoplasmic domains. However, M195 reacted less with the natural than with the recombinant soluble IL-6R. Using FACS analysis, the 7 mAb were shown to bind to U266 cells but not to the Namalva cell line which is deprived of IL-6R. This showed that they all recognised the membrane form of the IL-6R. Three of the anti-IL-6R mAb reacted with rIL-6R by Western blotting. Four different epitopes of the molecule were identified, either by cross-blocking experiments of mAb binding to IL6R in ELISA or by the biosensor Biacore technology. A group of 4 mAb (M37, M113, M139 and M164) and another mAb (M195) identified 2 different epitopes involved in IL-6 binding. These antibodies were able to inhibit the binding of IL-6 to IL-6R and the proliferation of the IL-6-dependent XG-1 cell line. M91 and M182 recognized 2 other epitopes that were not involved in IL-6 binding. As expected, M91 did not inhibit XG-1 proliferation; in contrast, M182 interfered with the proliferative response of the XG-1 cell line.(ABSTRACT TRUNCATED AT 250 WORDS)

Quantification of CD24 and CD45 antigens in parallel allows a precise determination of B-cell maturation stages: relevance for the study of B-cell neoplasias

Quantitative expression, i.e. absolute number of monoclonal antibody molecules bound per cell, was evaluated for CD24 and CD45 by flow cytometry with standards of fluorescence intensity on a panel of normal and neoplastic B cells. The CD24 antigen was expressed at homogeneous high level in fetal bone marrow and liver. Its density decreased progressively in the other normal tissues in parallel with the B-cell maturation. The ratio between CD24 density measured on fetal bone marrow B cells and that seen on adult peripheral B cells was 6:1. The CD45 antigen density was lower on fetal bone marrow cells than in the more mature stages. Fetal spleen lymphocytes and all the mature B lymphocytes displayed the same CD45 density than that seen on normal adult peripheral T cells. The CD24/CD45 antigen density ratio was precisely related to the stage of B-cell maturation. The same pattern of variation of CD24 and CD45 antigen density was seen on B-cell neoplasias, with a significantly higher value of CD24 and lower value of CD45 in acute lymphoblastic leukemias than in chronic malignancies. CD24 and CD45 antigen levels were frequently out of the range observed in the corresponding normal population. Among ALL patients, a low CD24/CD45 antigen density ratio was associated with a good prognosis. These data confirm the interest of an absolute quantitative study for widely distributed antigens.

Immunoassay for functional human soluble interleukin-6 receptor in plasma based on ligand/receptor interactions

Soluble forms of most cytokine receptors, able to bind effectively to their respective ligands, have now been described. A soluble interleukin-6-binding molecule derived from the gp80 component of the multichain IL-6 receptor can be detected in biological fluids, and can act as an agonist of IL-6 activity. The clinical significance of the soluble receptor levels still remains to be explored. We took advantage of the characterization of an anti-IL-6 monoclonal antibody and of an anti-IL-6R monoclonal antibody that both bound to IL-6/IL-6R complexes to design an immunometric assay for the measurement of soluble IL-6R complexed to IL-6. This reaction scheme was designated as ELIA (enzyme-ligand immunoassay). When exogeneous IL-6 was added in excess to an sIL-6R containing sample, all sIL-6R was present in a complexed form. Thus, the reaction scheme could also be used to determine total sIL-6R concentrations. A recombinant sIL-6R standard was prepared from the supernatant of murine thymoma cells transfected with a gene coding for an extracellular portion of the IL-6 receptor. The assay permitted the precise and reproducible measurement of sIL-6R in serum or plasma. This approach is of general relevance for the determination of soluble cytokine receptors in biological fluids, provided that adequate anti-cytokine and anti-receptor antibodies are available.

High amounts of circulating interleukin (IL)-6 in the form of monomeric immune complexes during anti-IL-6 therapy. Towards a new methodology for measuring overall cytokine production in human in vivo

A patient with plasma cell leukemia was treated with anti-interleukin (IL)-6 monoclonal antibodies (mAb) for 2 months. Using chromatography on protein A-Sepharose, anti-murine-IgG-Sepharose, anti-IL-6-mAb-Sepharose and gel filtration at pH 2.3, we have demonstrated that the anti-IL-6 mAb, by preventing the binding of IL-6 to its cell membrane receptor and its renal elimination, has induced huge amounts of IL-6 to circulate in the form of monomeric immune complexes. By using this observation, we have developed a mathematical modelling that allows the determination of the overall daily production of IL-6 in this patient, which was in the range of 15 micrograms per day. Overall in vivo production of cytokines has never been evaluated in animals or in humans before.

CD5 B lymphocyte antigen in monoclonal gammopathy

Because B lymphocytes bearing the CD5 antigen have been involved in many B-cell malignancies, we have investigated the presence of the CD5 B-cell antigen on B and plasma cells in monoclonal gammopathy. Quantification of CD5 B cells was made in the peripheral blood of seven individuals with monoclonal gammopathy of undetermined significance (MGUS) and in that of 21 patients with multiple myeloma (MM). The bone marrow of ten patients with MM was also studied. Patients with progressive MM presented a significant reduction in both B and CD5 B lymphocytes (i.e., percentages and absolute numbers), when compared with individuals with MGUS and patients with stable MM. These latter individuals and patients did not differ from healthy donors. No CD5 B cells were found in the bone marrow of patients with MM. Moreover, no CD5 antigen could be detected on eight freshly established human myeloma cells lines including six totally dependent on interleukin-6. However, it was weakly expressed on two standard myeloma cell lines not requiring exogenous interleukin-6 (i.e., RPMI 8226 and U 266). In conclusion, our data show mainly an overall reduction of the polyclonal CD5 B lymphocytes similar to what is observed for the other polyclonal B lymphocytes in patients with active MM. Finally, the expression of the CD5 antigen human myeloma cell lines is not constant.

Major immunoglobulin capping deficiency in the peripheral blood B cells of patients with Sjögren's syndrome

The capping of surface immunoglobulins (sIg) is a major characteristic of normal B lymphocytes. Thus, we have investigated sIg capping by peripheral blood (PB) B cells of patients with Sjögren's syndrome (SS) and we have found a major deficiency in these patients. In 12 healthy donors (HD), 8 +/- 2.8% of PB mononuclear cells were B cells (i.e. expressing the B-cell antigens CD19, CD20 and CD21 simultaneously) and more than 90% of these PB B cells were able to cap their sIg. In 12 experiments performed using PB lymphocytes from seven patients with SS, a major capping deficiency was noted with only 30% of PB B lymphocytes being able to cap sIg. This defect was not related to an expansion of the B-cell subpopulation expressing the CD5 antigen and was not observed in five patients with rheumatoid arthritis lacking SS. Capping of sIg via antigen binding (i.e. antigenic modulation) constitutes the initial signal for B-cell activation. This process is involved in anti-viral defence and could have a potential pathogenetic role in autoimmune diseases. This impaired B-cell function presently described represents an immune defect which could be important in the pathogenesis of SS.

Analysis of monoclonal rheumatoid factors obtained from the B-cell repertoire in rheumatoid arthritis

We have sought to determine whether rheumatoid factors (RF) produced in rheumatoid arthritis (RA) were different from physiological RF produced in normal, healthy adults. RF-secreting clones were established following Epstein-Barr virus (EBV) stimulation of peripheral blood lymphocytes. Ten RF-secreting clones were established from seven RA patients and 16 from six healthy controls. All monoclonal RF (MRF), except two in each group, were monoreactive and ten of these were shown to have low to medium affinity for IgG,Fc, irrespective of their origin. A majority (74%) of the MRF bound to protein A, indicating that genes of the VHIII family were preferentially used for synthesizing these autoantibodies. The expression of cross-reactive idiotypes (CRI) by the MRF did not allow distinction between those derived from RA patients and controls. The VHI-associated CRI G8 and VHIII-associated CRID12 were expressed at low frequency in both panels of RF. These CRI have been shown to be expressed at high frequency in RF paraproteins. However, the idRQ idiotype was expressed within both panels of RF. A possible distinction between polyreactive and monoreactive MRF appeared to be light chain usage since all (four) polyreactive RF used lambda chains while the normal kappa/lambda ratio was observed for monoreactive RF. The frequency of EBV-activated cells secreting IgM bearing CRI or secreting RF was determined and showed that CRI expression occurred with a higher frequency than did RF, suggesting a dissociation between CRI expression and RF activity.

The B cell repertoire in rheumatoid arthritis. III. Preferential homing of rheumatoid factor-producing B cell precursors in the synovial fluid

OBJECTIVE

We sought to compare the frequencies of precursors producing IgM rheumatoid factors (IgM-RFs) in synovial fluid and peripheral blood B cells from patients with rheumatoid arthritis (RA).

METHODS

We used limiting-dilution analysis of Epstein-Barr virus-activated B cells from seropositive and seronegative patients. B cell precursors producing IgM against insulin, an irrelevant autoantigen, were also assessed for comparison.

RESULTS

On average, IgM-RF-producing precursors were 15-fold higher in the synovial fluid than in the peripheral blood of seropositive RA patients, but not in seronegative RA patients. The frequency of B cell precursors producing IgM against insulin was lower in the synovial fluid than in the peripheral blood of both the seropositive and the seronegative patient groups; moreover, the concentrations were similar in both groups.

CONCLUSION

The findings provide evidence against a nonspecific accumulation of IgM-producing cells in the synovial fluid, and suggest that there is an active attraction of the RF-producing B cell precursors toward sites of inflammation in RA.

Murine anti-interleukin-6 monoclonal antibody therapy for a patient with plasma cell leukemia

A patient with primary plasma cell leukemia resistant to chemotherapy was treated for 2 months with daily intravenous injections of anti-interleukin-6 (IL-6) monoclonal antibodies (MoAbs). The patient's clinical status improved throughout the treatment and no major side effects were observed. Serial monitoring showed blockage of the myeloma cell proliferation in the bone marrow (from 4.5% to 0% myeloma cells in the S-phase in vivo) as well as reduction in the serum calcium, serum monoclonal IgG, and the serum C-reactive protein levels. The serum calcium and serum monoclonal IgG corrected by approximately 30%, whereas the C-reactive protein corrected to undetectable levels during treatment. No major side effects developed, although both platelet and circulating neutrophil counts decreased during anti-IL-6 therapy. A transient immunization was detected 15 days after the initiation of the treatment, which could explain the recovery of myeloma cell proliferation after 2 months of treatment (2% myeloma cells in the S phase). In conclusion, this first anti-IL-6 clinical trial demonstrated the feasibility of injecting anti-IL-6 MoAbs, and also a transient tumor cytostasis and a reduction in IL-6-related toxicities. It gave insight into the major biologic activities of IL-6 in vivo and may serve as a basis for further development of anti-IL-6 therapy in myeloma and other IL-6-related diseases.

The B cell repertoire in rheumatoid arthritis. II. Analysis of rheumatoid factors bearing the IdRQ cross-reactive idiotype

With the view of studying whether rheumatoid factors (RFs) produced in rheumatoid arthritis (RA) were different from those synthesized in physiological situations, we analyzed the usage of a cross-reactive idiotype (IdRQ) previously reported to be specific for RA RFs. Using EBV immortalization of circulating B cells, we prepared monoclonal RFs from patients with RA and matched controls. In both groups between 1/2 and 2/3 of the monoclonal RFs bore IdRQ. Using limiting dilution analysis, we studied the frequencies of the EBV-activated B cells able to synthesize immunoglobulins bearing IdRQ. In patients and in controls, on average, 1/3 of the RF-secreting cells used IdRQ and around 2/3 of the synthesized IgM bearing IdRQ were devoid of RF activity. These results show that precursor cells containing the germline gene encoding IdRQ are present in similar quantities in RA patients and healthy individuals, and that the IdRQ cross-reactive idiotype, although interesting for the study of the B cell repertoire, is probably not useful as a marker for susceptibility to RA.

The B cell repertoire in rheumatoid arthritis. I. Frequency of EBV-inducible circulating precursors producing autoantibodies

The frequency of B cell precursors producing antibodies against various autoantigens (Fc fragment of IgG, F(ab')2 fragment of IgG, type II collagen, cytoskeleton filaments and insulin) was determined in patients with rheumatoid arthritis (RA) using immortalization of peripheral blood B cells by the Epstein-Barr virus (EBV) and limiting dilution analysis. Equally large numbers of B cell precursors producing IgM-rheumatoid factors (RFs) were present in the peripheral blood of seronegative and seropositive RA patients and of controls. On average, 1 out of 15,000 B cells could be induced by EBV to secrete IgM-RFs, which represents 0.5-1% of the EBV-induced proliferating clones. By cloning or somatic hetero-hybridization of EBV cell lines derived from patients and controls, we obtained two types of monoclonal RFs: one polyreactive, reacting with Fc but also with the other autoantigens tested, and the other monoreactive, reacting with Fc only and that previously had only been found in the RA B cell repertoire. Moreover, patients and controls had similar numbers of circulating B cell precursors secreting IgM antibodies against other autoantigens that might be regarded as specific targets of RA (F(ab')2 fragment of IgG and type II collagen), and against cytoskeleton filaments that are targets of natural autoantibodies, increased in RA. The frequencies of EBV-induced B cells producing antibodies against all these autoantigens were of the same order of magnitude as the frequency of EBV-induced B cells producing RFs. The patients also possessed a similar number of precursors producing antibodies against insulin, an autoantigen irrelevant to the pathogenesis of the disease, taken as control. These data tend to demonstrate no abnormality in the autoantibody repertoire of B cells activable by EBV in RA, especially those secreting RFs. In vitro spontaneous RF secretion by circulating B cells was observed in seropositive RA patients but not in seronegative patients and in the controls tested. We enumerated the number of B cells spontaneously secreting RFs in seropositive RA patients and found that it correlated with the serum RF titer, but not with the number of RF-secreting B cells activated by EBV. The mean frequency values of B cells secreting RFs either spontaneously or after EBV infection were of the same order of magnitude, showing that the expanded population of in vivo-activated B cells was not (at least partially) infectable by EBV. This raised the possibility that EBV triggers a repertoire which may not reflect the status of B cells secreting autoantibodies in autoimmune diseases.

The CD24 antigen discriminates between pre-B and B cells in human bone marrow

When bone marrow (BM) lymphoid cells from 12 adult healthy donors were labeled by CD24 antibodies and analyzed by flow cytometry, two positive populations of cells were demonstrated in each sample (by a separated bimodal specific immunofluorescence). One population had intermediate CD24-Ag density (termed CD24+ cells) whereas the other had high CD24-Ag density (termed CD24(2+) cells). CD24+ cells represented 5.8 +/- 2.7% of the total lymphoid BM cells and CD24(2+) cells 5.6 +/- 2.5%. Using dual fluorescence analysis on eight samples, all CD24+ cells expressed the CD21 and CD37 mature B cell Ag and also surface IgM (sIgM), but this population lacked CD10 Ag. These cells also expressed CD19 Ag, and at a higher density than CD24(2+) cells. They were also positive for HLA-DR Ag. Conversely, CD24(2+) cells were shown to be early cells of the B cell lineage. While all the CD24(2+) cells were HLA-DR+ and CD19+, 64 +/- 16% of them expressed CD20 Ag (at a lower density than CD24+ cells), 65 +/- 21% CD10 Ag, and 22 +/- 8% were positive for cytoplasmic mu-chains (c mu). None of these cells expressed the CD21 and CD37 mature B cell Ag or sIgM. Additional experiments on four different healthy donors demonstrated that 30 +/- 9% of the CD24(2+) cells expressed the CD34 Ag and that the CD24+ cells did not express it. Thus, the CD24 Ag permits discrimination between two populations of the B cell lineage present in adult BM: 1) A CD24(2+) cell population including "pre" pre-B cells (HLA-DR+, CD19+, CD10+/-, CD20-, CD21-, CD34+, CD37-, c mu-), "intermediate" pre-B cells (HLA-DR+, CD19+, CD10+, CD20+, CD21-, CD34-, CD37-, c mu-), and "true" pre-B cells (HLA-DR+, CD19+, CD10+, CD20+, CD21-, CD34-, CD37-, c mu+). 2) A CD24+ cell population including B cells of the standard phenotype (HLA-DR+, CD19+, CD10-, CD20+, CD21+, CD34-, CD37+, c mu-, sIgM+).

The CD24 antigen discriminates between pre-B and B cells in human bone marrow

When bone marrow (BM) lymphoid cells from 12 adult healthy donors were labeled by CD24 antibodies and analyzed by flow cytometry, two positive populations of cells were demonstrated in each sample (by a separated bimodal specific immunofluorescence). One population had intermediate CD24-Ag density (termed CD24+ cells) whereas the other had high CD24-Ag density (termed CD24(2+) cells). CD24+ cells represented 5.8 +/- 2.7% of the total lymphoid BM cells and CD24(2+) cells 5.6 +/- 2.5%. Using dual fluorescence analysis on eight samples, all CD24+ cells expressed the CD21 and CD37 mature B cell Ag and also surface IgM (sIgM), but this population lacked CD10 Ag. These cells also expressed CD19 Ag, and at a higher density than CD24(2+) cells. They were also positive for HLA-DR Ag. Conversely, CD24(2+) cells were shown to be early cells of the B cell lineage. While all the CD24(2+) cells were HLA-DR+ and CD19+, 64 +/- 16% of them expressed CD20 Ag (at a lower density than CD24+ cells), 65 +/- 21% CD10 Ag, and 22 +/- 8% were positive for cytoplasmic mu-chains (c mu). None of these cells expressed the CD21 and CD37 mature B cell Ag or sIgM. Additional experiments on four different healthy donors demonstrated that 30 +/- 9% of the CD24(2+) cells expressed the CD34 Ag and that the CD24+ cells did not express it. Thus, the CD24 Ag permits discrimination between two populations of the B cell lineage present in adult BM: 1) A CD24(2+) cell population including "pre" pre-B cells (HLA-DR+, CD19+, CD10+/-, CD20-, CD21-, CD34+, CD37-, c mu-), "intermediate" pre-B cells (HLA-DR+, CD19+, CD10+, CD20+, CD21-, CD34-, CD37-, c mu-), and "true" pre-B cells (HLA-DR+, CD19+, CD10+, CD20+, CD21-, CD34-, CD37-, c mu+). 2) A CD24+ cell population including B cells of the standard phenotype (HLA-DR+, CD19+, CD10-, CD20+, CD21+, CD34-, CD37+, c mu-, sIgM+).

Interleukin-6 is the central tumor growth factor in vitro and in vivo in multiple myeloma

When bone-marrow cells from patients with multiple myeloma (MM) were seeded in short-term cultures, a spontaneous proliferation of the myeloma cells occurred for most of the patients with active disease and proliferating myeloma cells in vivo. In all cases, this spontaneous proliferation was inhibited by anti-IL-6 monoclonal antibodies (mabs). Moreover, myeloma cell lines, completely dependent upon exogenous IL-6 for their growth, could be reproducibly established by initially stimulating the myeloma cells with both IL-6 and GM-CSF. These results demonstrate that IL-6 is a major paracrine myeloma-cell growth factor in vitro. High serum IL-6 levels were observed in MM patients with active disease, especially patients with terminal disease. High IL-6 mRNA levels were found in bone-marrow cells of MM patients, mainly in myeloid and monocytic cells, in vivo. The myeloma cells did not express IL-6 mRNA. Injection of anti-IL-6 mabs to MM patients with terminal disease and extramedullary proliferation, completely blocked the myeloma-cell proliferation in vivo and completely inhibited the serum IL-6 bioactivity and the serum CRP levels. One patient with plasma cell leukemia and hypercalcemia was treated for two months with anti-IL-6 mabs and maintain in remission for 2 months without major side effects. Interestingly, the serum calcium levels also decreased in these patients. All these results show that IL-6 is the main cytokine responsible not only for the myeloma-cell proliferation in vivo, but presumably also for the large bone resorption processes observed in human MM.(ABSTRACT TRUNCATED AT 250 WORDS)

Anti-F(ab')2 antibody response to the injection of anticlass II HLA alloantibodies in patients with rheumatoid arthritis

Placenta eluted gamma globulins (PEGG) contain antibodies against class II HLA antigens and have been used for treating patients with rheumatoid arthritis (RA). In view of the potential use of antibodies to class II HLA for treating autoimmune diseases we looked for the immunobiological effects of PEGG injections in patients. No modulation of class II HLA was seen at the surface of circulating mononuclear cells after one week of daily PEGG injections. In some patients, antibodies to F(ab')2 fragments of PEGG-IgG were produced. These antibodies reacted against F(ab')2 of any IgG as well and did not prevent anticlass II HLA antibodies from binding to class II HLA, thus showing no characteristics of classical antiidiotypic antibodies. The appearance of anti-F(ab')2 antibodies was not correlated with the clinical course of the disease. Their significance is discussed.

Cell lyophilization for safe conservation and transportation in view of DNA analysis

Images

The dendritic reticulum cell pattern in B cell lymphomas of the small cleaved, mixed, and large cell types: an immunohistochemical study of 48 cases

An immunohistochemical study was designed to study the dendritic reticulum cell (DRC) patterns in 48 cases of B cell non-Hodgkin's lymphomas of the small cleaved, mixed, and large cell types, both follicular (20 cases) and diffuse (28 cases), in order to evaluate the possible influence of DRCs on homing and the differentiation of neoplastic B cells. Three DRC patterns were observed. In the follicular lymphomas, DRCs constituted nodular networks of variable density. In the diffuse lymphomas, DRCs were present either as isolated and scattered cells (17 cases) or constituted irregular meshworks of variable sizes (11 cases). These DRC patterns correlate with B cell immunophenotypes. Like follicular lymphomas, and unlike diffuse lymphomas without DRC networks, diffuse lymphomas with DRC networks constantly expressed the pan B antigens and one marker characteristic of normal germinal center cells, CD21 antigen, the C3d receptor. The finding of organized DRC networks in a significant number of diffuse lymphomas does not substantiate the hypothesis that DRCs may play a role in the homing of neoplastic B cells. The correlations observed between DRC patterns and B cell immunophenotypes suggest that the persistence and/or the development of DRC networks within follicular center cell-type lymphomas are related to the degree of functional differentiation of neoplastic B cells.

Phenotypic analysis of human myeloma cell lines

Multiple myeloma (MM) is a B-cell malignancy characterized by the accumulation, primarily in bone marrow, of a clone of plasma cells. The nature of the stem cells feeding the tumoral compartment is still unknown. To investigate this special point, we have studied the phenotypes of nine well-known human myeloma cell lines (HMCLs) and compared them with those of normal lymphoblastoid cell lines (LCLs). Twenty-four clusters of differentiation involved in B lymphopoiesis were investigated using a panel of 65 monoclonal antibodies (MoAbs). For each cluster, the percentage of positive cells and the antigen density were determined, giving rise to a "quantitative phenotype". We thus classified the HMCLs into two different groups: those with cytoplasmic mu chains (c mu+) and those without (c mu-). In the first (c mu+) group, comprising seven cell lines, the HMCLs had a phenotype of pre-B/B cells close to that of Burkitt's lymphoma cell lines. They expressed low densities of surface mu chains, without detectable cytoplasmic or surface light chains. Three of them were infected with the Epstein Barr virus (EBV). These c mu+ HMCLs bore most of the B-cell antigens except CD23. They expressed the CALLA antigen (CD10) and lacked the plasma-cell antigen PCA1. In contrast, LCLs expressed surface light chains, high densities of CD23, low densities of PCA1 antigen, and no CD10 antigen. The c mu- HMCLs had a plasma-cell phenotype, lacking most of the B-cell antigens and expressing high densities of PCA1 antigen.(ABSTRACT TRUNCATED AT 250 WORDS)