Enhanced Ig production by human peripheral lymphocytes induced by aggregated C1q

Because B cells express receptors for C1q, we have investigated the role of C1q in the stimulation of B cells. When B cells were cultured in the presence of C1q that had been frozen, T cells, and suboptimal concentrations of PWM, there was a dose-dependent enhancement of IgM, IgG, and IgA by the B cells. No significant enhancement of Ig production by B cells was seen in the absence of T cells or PWM. The contribution of T cells or PWM could be replaced by supernatants of PMA and Con A-activated PBMC (T cell growth factor). C1q that had been frozen, in contrast with freshly isolated C1q, was at least 3 times more active in enhancement of the production of Ig by B cells in culture in the presence of suboptimal concentrations of T cell growth factor. The capability of C1q to stimulate B cells could be ascribed to aggregates of C1q. Monomeric C1q was only marginally active to stimulate B cell Ig production, whereas dimeric and tetrameric C1q were able to enhance Ig production by B cells in relation to their size. Furthermore, aggregation of C1q on soluble aggregates of rabbit IgM also increased its potential to enhance B cell Ig production. The interaction of C1q with the B cells occurs via the collagenous tail of C1q, as suggested by inhibition experiments with purified collagenous tails and globular heads of C1q. These results indicate that triggering of C1qR on B cells positively regulates Ig production in vitro.     

C1q induces chemotaxis and K+ conductance activation coupled to increased cytosolic Ca2+ in mouse fibroblasts

Cultured mouse fibroblasts (L cells) respond to whole C with a slow hyperpolarization. Among the C components tested, C1q was found to be most effective. In contrast, the cell did not respond to C1, in which the collagen-like region of the C1q molecule is masked. The C1q-induced hyperpolarizing response was inhibited by collagen or C1q-specific antisera. Human diploid skin fibroblasts (Flow 1,000 cells) also exhibited similar membrane potential changes in response to whole C or C1q. After repeated applications of C1q, the cell membrane became unresponsive (desensitized). The treatment of L cells with pronase E inhibited the C1q-induced response, whereas the response to ATP, which is known to interact to its own receptor, was still preserved. The reversal potential of C responses was close to the K+ equilibrium potential. The hyperpolarizing response was inhibited by a blocker of Ca2+-activated K+ channels in fibroblasts (quinine), by deprivation of extracellular Ca2+ or by a Ca2+ channel blocker (nifedipine). By means of Ca2+-selective microelectrodes, the cytosolic free Ca2+ concentration was found to increase from 126 to 206 nM upon stimulation of L cells with C1q. Using an agarose-well method, L cells were observed to migrate predominantly toward C1q or whole C. It is concluded that the fibroblasts have the C1q receptor sensitive to pronase E and that activation of C1q receptors gives rise to Ca2+ influx, triggering an increase in the cytosolic free Ca2+ ions, which in turn induces a hyperpolarizing response as a result of the stimulation of Ca2+-activated K+ channels and initiates chemotaxis to C1q.     

Participation of C1q and its receptor in adherence of human diploid fibroblast

C1q binds through its collagen-like domain to specific surface receptors of fibroblasts and to adhesive elements of extracellular matrix including fibronectin, collagens, proteoglycans, and laminin. To determine whether C1q participates in fibroblast adhesion, cells in serum-free medium were plated on surfaces coated with purified C1q at physiologic ionic strength and pH. Surfaces coated with fibronectin or collagen type I served as positive controls, and those coated with BSA were negative controls. Substratum-adsorbed C1q promoted fibroblast adhesion to a maximum of 73% of available cells within 90 min at 37 degrees C. Adhesion was C1q concentration dependent, saturable, specific, and dependent on the collagen-like domain of the molecule. De novo protein synthesis plays a role in adhesion: pretreatment of fibroblasts with cycloheximide reduced adherence about 50% of controls. Addition of exogenous fibronectin, collagen type I, or C1q as soluble mediators did not affect adhesion of the cycloheximide-treated cells to C1q substrate. Adhesion could be accounted for primarily, although not completely, by the C1q receptors. Antibodies raised against the Raji cell C1q receptors (alpha C1qR Ab) specifically inhibited fibroblast adhesion to C1q substrates about 60% of controls. The binding of fibroblasts to C1q substrates could be inhibited about 24% of controls with the GRGDTP cell recognition peptide. GRGDTP and alpha C1q Ab had an additive effect on adhesion that was inhibited 77 to 80% of controls. We conclude from these data that aggregated rather than monomeric C1q may be the natural ligand of the fibroblast C1q receptor, and the biologic function of the receptor in cells of the connective tissue may be cell adhesion.     

Chromosome 5q deletion and epigenetic suppression of the gene encoding alpha-catenin (CTNNA1) in myeloid cell transformation

Interstitial loss of all or part of the long arm of chromosome 5, or del(5q), is a frequent clonal chromosomal abnormality in human myelodysplastic syndrome (MDS, a preleukemic disorder) and acute myeloid leukemia (AML), and is thought to contribute to the pathogenesis of these diseases by deleting one or more tumor-suppressor genes. Although a major commonly deleted region (CDR) has been delineated on chromosome band 5q31.1 (refs. 3-7), attempts to identify tumor suppressors within this band have been unsuccessful. We focused our analysis of gene expression on RNA from primitive leukemia-initiating cells, which harbor 5q deletions, and analyzed 12 genes within the CDR that are expressed by normal hematopoietic stem cells. Here we show that the gene encoding alpha-catenin (CTNNA1) is expressed at a much lower level in leukemia-initiating stem cells from individuals with AML or MDS with a 5q deletion than in individuals with MDS or AML lacking a 5q deletion or in normal hematopoietic stem cells. Analysis of HL-60 cells, a myeloid leukemia line with deletion of the 5q31 region, showed that the CTNNA1 promoter of the retained allele is suppressed by both methylation and histone deacetylation. Restoration of CTNNA1 expression in HL-60 cells resulted in reduced proliferation and apoptotic cell death. Thus, loss of expression of the alpha-catenin tumor suppressor in hematopoietic stem cells may provide a growth advantage that contributes to human MDS or AML with del(5q).     

Biosynthesis of normal and low-molecular-mass complement component C1q by cultured human monocytes and macrophages.

High levels of low-molecular-mass complement component C1q (LMM-C1q), a haemolytically inactive form of C1q, are found in serum of individuals with inherited complete (functional) C1q deficiency and in serum of patients with systemic lupus erythematosus, whereas lower levels are present in normal serum [Hoekzema, Hannema, Swaak, Paardekooper & Hack (1985) J. Immunol. 135, 265-271]. To investigate whether LMM-C1q is a (by-)product of C1q synthesis or the result of degradation of C1q, cultures of blood monocytes and of alveolar macrophages, which secrete functional C1q, were studied. A considerable portion of C1q-like protein secreted by these cells was found to be LMM-C1q. In contrast with the C1q fragments that resulted from degradation of normal C1q during phagocytosis, culture-derived LMM-C1q appeared to be identical with LMM-C1q found in serum, as judged by sedimentation behaviour, subunit structure and recognition by poly- and mono-clonal antibodies raised against C1q. The presence of LMM-C1q in cytoplasmic organelles compatible with the Golgi apparatus and the inability to generate LMM-C1q by impeding hydroxylation and triple-helix formation of C1q further argues against degradation as its source. Monocyte cultures of homozygous probands from two families with complete functional C1q deficiency reflected the abnormalities in serum, i.e. absence of functional C1q, but increased levels of LMM-C1q. By contrast, secretion of C1q and LMM-C1q by cells from healthy individuals was clearly co-ordinate, indicating that LMM-C1q in serum may provide a unique marker of C1q synthesis in vivo.     

Enhanced phospholipase C stimulation and transformation in NIH-3T3 cells expressing Q209LGq-alpha-subunits.

NIH-3T3 cells were transfected with cDNA encoding the native alpha-subunit of the G protein Gq(alpha q) or a mutant (Q209L) form of alpha q. Cells expressing Q209L-alpha q showed greatly enhanced basal phospholipase C activity. Stimulation of phospholipase C activity by prostaglandin F2 alpha or fetal calf serum was increased up to 10-fold in Q209L-alpha q-transfected cells. Continuous expression of Q209L-alpha q or overexpression of alpha q in NIH-3T3 cells resulted in formation of foci after 3 weeks. The number of foci was proportional to the number of transfected cells and was greater in cells expressing the Q209L-alpha q than in cells that overexpressed the wild type alpha q. Q209L-alpha q-transfected NIH-3T3 cells also formed colonies in soft agar indicating their capacity to grow in an anchorage-independent manner. Expression of Q209L-alpha q in Rat-1 cells resulted in enhanced basal and fetal calf serum-stimulated phospholipase C activity, but these cells were not transformed as assessed by either the focus formation or the soft agar colony formation assays. These results indicate that expression of continuously activated Gq-alpha can result in transformation in a cell type-specific manner.     

Production and characterization of a murine monoclonal IgM antibody to human C1q receptor (C1qR)

A hybridoma cell line that produces a monoclonal antibody (MAb) to cell surface C1q receptor (C1qR) has been produced by fusion of the P3 X 63-Ag8.653 mouse myeloma cell line with the spleen cells of a CD-1 mouse that had been hyperimmunized with viable Raji cell suspensions (5 X 10(7) cells/inoculum). This MAb, designated II1/D1, is an IgM antibody with lambda-light chain specificity. Radiolabeled or unlabeled, highly purified II1/D1 was used to determine that: this antibody competes for C1q binding sites on C1qR-bearing cells; the molecule recognized by this MAb is the C1qR; and cells that are known to bind C1q also bind II1/D1 in a specific manner. Western blot analysis of solubilized Raji, or U937 cell membranes, showed that the 125I-MAb detected a major protein band of approximately 85,000 m.w. in its unreduced state, indicating that the C1qR is similar, if not identical, in both types of cells. Analyses of 125I-II1/D1 binding experiments revealed that the antibody bound to Raji cells or U937 cells in a specific manner. Uptake of the antibody was saturable, with equilibrium virtually attained within 35 min. Scatchard analysis of the binding data using the intact MAb suggests that the affinity constant KD is 2.9 X 10(-10) M, and at apparent saturation, 24.6 ng of the antibody were bound per 2 X 10(6) cells, giving an estimated 7.8 X 10(3) antibody molecules bound per cell. That the II1/D1 antibody is specifically directed to the C1q was further evidenced by an ELISA in which the ability of C1qR-bearing cells to bind the MAb was abrogated by c-C1q in a specific and dose-dependent manner. These results indicate that the II1/D1 is a specific antibody directed against the C1q and can be a useful tool in studying the biologic interaction of human C1q with its receptors on a variety of cells.     

Human pulmonary surfactant protein (SP-A), a protein structurally homologous to C1q, can enhance FcR- and CR1-mediated phagocytosis

C1q, a subunit of the first component (C1) of the classical complement pathway, and the pulmonary surfactant protein SP-A are structurally homologous molecules, each having an extended collagen-like domain contiguous with a non-collagenous domain. It is the collagen-like region of C1q that binds to mononuclear phagocytes and mediates the enhancement of phagocytosis of opsonized particles by these cells. Because SP-A enhances the endocytosis of phospholipids by alveolar type II cells and alveolar macrophages, we examined whether these two molecules were functionally interchangeable. The phagocytosis of sheep erythrocytes opsonized with IgG or with IgM and complement was enhanced by the adherence of monocytes or macrophages, respectively, to SP-A. The enhanced response was dependent on the concentration of SP-A used for coating the surfaces, similar to that seen when monocytes were adhered to C1q-coated surfaces. Both the percentage of cells ingesting the opsonized targets and the number of targets ingested per cell increased with increasing concentrations of SP-A. No such enhancement was seen with cells adhered to albumin, iron-saturated transferrin, or uncoated surfaces. However, SP-A did not substitute for C1q in the formation of hemolytically active C1. C1q did not stimulate lipid uptake by alveolar type II cells or alveolar macrophages and had only a slight inhibitory effect on the binding of SP-A to alveolar type II cells. Thus, these results suggested that a function which requires interactions of both the collagenous and the non-collagenous regions (i.e. initiation of the classic complement cascade) could not be mimicked by a protein sharing structural macromolecular similarity but lacking sequence homology in the non-collagen-like region. However, SP-A could substitute for C1q in stimulating a function previously shown to be mediated by the collagen-like domains of the C1q molecule.     

Characterization of C1q receptor expression on human phagocytic cells: effects of PDBu and fMLP

The receptor-mediated binding of C1q to human phagocytic cells was investigated in this study. By using a C1q binding assay, we determined that purified, elutriated monocytes expressed an average of 4.6 X 10(5) C1q receptors (C1qR) per cell, with an equilibrium binding constant (Keq) of 0.91 X 10(7) (M-1). When analyzed in an identical manner, the polymorphonuclear leukocytes (PMN) expressed an average of 4.2 X 10(5) C1qR per cell, with a Keq for C1q of 1 X 10(7) (M-1). Fluorescent flow cytometric analysis showed that C1q was bound by 98% of the monocytes studied. Further, the pattern formed by these cells was consistent with a normal log distribution, indicating that this was a homogeneous population of cells. When PMN were assayed with flow cytometry, however, we found that C1q was bound by an average of only 45% of the PMN analyzed. Further, these PMN were not dispersed in a normal log distribution, indicating some heterogeneity among the cells that bind C1q. We examined the abilities of the chemoattractant N-formylmethionylleucylphenylalanine (fMLP) and the phorbol ester phorbol dibutyrate (PDBu) to modulate expression of C1qR as compared to the receptor for C3b (CR1). Pretreatment of the monocytes and the PMN with either 10(-6)M fMLP or 10 ng/ml of PDBu significantly increased cell surface CR1 expression, as reported previously by other investigators. In contrast, no significant increase in the number of C1qR on the monocytes or the PMN was observed with any of the concentrations of fMLP or PDBu used during pretreatment. However, with certain pretreatment doses of these agents, some reduction was noted in the amount of 125I-C1q bound to the monocytes and the PMN. This study characterizes the binding of C1q to purified monocytes and confirms previously published values for PMN. The distribution of cells expressing C1qR is shown to be significantly different between identically treated populations of monocytes and PMN. Finally, the abilities of fMLP and PDBu to modulate the binding of C1qR are examined. Our results indicate that the control of C1qR expression differs markedly from that of CR1.     

Complement subcomponent C1q stimulates Ig production by human B lymphocytes

The regulation of Ig production by human B lymphocytes is a complex process involving interactions among B cells, APC, T lymphocytes and soluble factors including activation, growth, and differentiation factors. Components of the complement system, including C3a, C3b, C3d, and C5a, have been shown to influence various stages in this process. In this study, we demonstrate that the C1q subcomponent of complement binds to both small resting and large activated B cells and stimulates immunoglobulin production by Staphylococcus aureus Cowan-activated tonsillar B lymphocytes. This effect is present whether C1q is added to the B cells either at the beginning or near the end of a 7-day culture period and is not associated with enhancement of proliferation. The C1q stimulation of Ig production is, however, associated with increased steady state levels of mRNA for the mu Ig H chain. Furthermore, C1q stimulated IgM production by the human B cell line SKW 6.4, which is capable of secreting IgM in response to B cell differentiation factors (BCDF). SLE is a disorder frequently associated with polyclonal activation of B lymphocytes. We studied the effect of C1q on B cells from two patients with this disorder and one with an SLE-like illness, all selected for the predominance of either IgM or IgG in serum. Spontaneous or BCDF-stimulated Ig secretion was of the isotype predominant in vivo, whereas C1q selectively stimulated B cells to produce the other isotype (IgG vs IgM). Thus, C1q interacts with B lymphocytes in a manner distinct from that of BCDF found in mixed lymphocyte supernatants. C1q may be an important factor influencing the production of Ig by B lymphocytes in normal individuals and in patients with abnormalities of B cell activity.     

Reconstitution of the complement function in C1q-deficient (C1qa-/-) mice with wild-type bone marrow cells

Besides Ab-independent and Ab-dependent activation of the complement classical pathway in host defense, C1q plays a key role in the processing of immune complexes and in the clearance of apoptotic cells. In humans, C1q deficiency leads to systemic lupus erythematosus-like symptoms in over 90% of the cases, thus making this defect a strong disease susceptibility factor. Similarly, C1q-deficient mice (C1qa-/-) develop systemic lupus erythematosus-like symptoms, such as autoantibodies and glomerulonephritis. We have previously provided evidence that C1q is produced by cells of the monocyte-macrophage lineage. In this study, we have tested whether transplantation of bone marrow cells would be sufficient to reconstitute C1q levels in C1qa-/- mice. C1qa-/- mice received a single graft of 10(7) bone marrow cells from wild-type (wt) donors after irradiation doses of 6, 7, 8, or 9 Gy. Engraftment was monitored by a Y chromosome-specific PCR and a PCR that differentiated wt from C1qa-/- genotype. Serum levels of C1q Ag and C1 function increased rapidly in the recipient mice, and titers reached normal levels within 6 wk after bone marrow transplantation. In wt mice that received C1qa-/- bone marrow, serum levels of C1q decreased constantly over time and became C1q deficient within 55 wk. These data clearly demonstrate that bone marrow-derived cells are the source of serum C1q and are competent to reconstitute normal C1q serum levels in C1q-deficient mice. Therefore, stem cell transplantation could be a therapy for patients with hereditary C1q deficiency.     

Investigations on the C1q-calreticulin-phosphatidylserine interactions yield new insights into apoptotic cell recognition

Both C1q and calreticulin (CRT) are involved in the recognition of apoptotic cells. CRT was initially characterized as a receptor for the C1q collagen-like fragment (CLF), whereas C1q was shown to bind apoptotic cells through its globular region (GR). Using purified CRT and recombinant CRT domains, we now provide unambiguous experimental evidence that, in addition to its CLF, the C1q GR also binds CRT and that both types of interactions are mediated by the CRT globular domain. Surface plasmon resonance analyses revealed that the C1q CLF and GR domains each bind individually to immobilized CRT and its globular domain with K(D) values of (2.6-8.3) × 10(-7) M. Further evidence that CRT binds to the C1q GR was obtained by electron microscopy. The role of CRT in the recognition of apoptotic HeLa cells by C1q was analyzed. The C1q GR partially colocalized with CRT on the surface of early apoptotic cells, and siRNA (small interfering RNA)-induced CRT deficiency resulted in increased apoptotic cell binding to C1q. The interaction between CRT and phosphatidylserine (PS), a known C1q ligand on apoptotic cells, was also investigated. The polar head of PS was shown to bind to CRT with a 10-fold higher affinity (K(D)=1.5 × 10(-5) M) than that determined for C1q, and, accordingly, the C1q GR-PS interaction was impaired in the presence of CRT. Together, these observations indicate that CRT, C1q, and PS are all closely involved in the uptake of apoptotic cells and strongly suggest a combinatorial role of these three molecules in the recognition step.     

C1q production and C1q-mediated immune complex retention in lymphoid follicles of rat spleen

Summary Involvement of C1q in retaining immune complexes in germinal centers in rat spleen was studied in vivo and in vitro. C1q production was found in fibroblastic reticulum cells in the peripheral mantle zone, in follicular dendritic cells in germinal centers, and in transitional forms between these two cells in the inner mantle zone. In passively immunized animals, immune complexes were found transiently on fibroblastic reticulum cells, then on the transitional forms and follicular dendritic cells. Extracellular C1q was detected by the presence of immune complexes on both the transitional forms and follicular dendritic cells, but not on fibroblastic reticulum cells. Thus, the fibroblastic reticulum cell appeared to trap immune complexes but not to retain either immune complexes or C1q. The morphology and function of the fibroblastic reticulum cell and the follicular dendritic cell suggest that they belong to the same lineage. Immune complexes were bound in vitro to germinal centers in cryostat spleen sections in the same manner as those retained in vivo. The binding required no complement in the incubation medium and was inhibited by C1q-suppressing factors. The extracellular C1q originating from the follicular cells may therefore play a role in retaining immune complexes in the germinal center.     

Human and Pneumococcal Cell Surface Glyceraldehyde-3-phosphate Dehydrogenase (GAPDH) Proteins Are Both Ligands of Human C1q Protein

C1q, a key component of the classical complement pathway, is a major player in the response to microbial infection and has been shown to detect noxious altered-self substances such as apoptotic cells. In this work, using complementary experimental approaches, we identified the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a C1q partner when exposed at the surface of human pathogenic bacteria Streptococcus pneumoniae and human apoptotic cells. The membrane-associated GAPDH on HeLa cells bound the globular regions of C1q as demonstrated by pulldown and cell surface co-localization experiments. Pneumococcal strains deficient in surface-exposed GAPDH harbored a decreased level of C1q recognition when compared with the wild-type strains. Both recombinant human and pneumococcal GAPDHs interacted avidly with C1q as measured by surface plasmon resonance experiments (K_D = 0.34–2.17 nm). In addition, GAPDH-C1q complexes were observed by transmission electron microscopy after cross-linking. The purified pneumococcal GAPDH protein activated C1 in an in vitro assay unlike the human form. Deposition of C1q, C3b, and C4b from human serum at the surface of pneumococcal cells was dependent on the presence of surface-exposed GAPDH. This ability of C1q to sense both human and bacterial GAPDHs sheds new insights on the role of this important defense collagen molecule in modulating the immune response.     

Genome-wide examination of chromosomal aberrations in neuroblastoma SH-SY5Y cells by array-based comparative genomic hybridization

Most neuroblastoma cells have chromosomal aberrations such as gains, losses, amplifications and deletions of DNA. Conventional approaches like fluorescence in situ hybridization (FISH) or metaphase comparative genomic hybridization (CGH) can detect chromosomal aberrations, but their resolution is low. In this study we used array-based comparative genomic hybridization to identify the chromosomal aberrations in human neuroblastoma SH-SY5Y cells. The DNA microarray consisting of 4000 bacterial artificial chromosome (BAC) clones was able to detect chromosomal regions with aberrations. The SH-SY5Y cells showed chromosomal gains in 1q12 approximately q44 (Chr1:142188905-246084832), 7 (over the whole chromosome), 2p25.3 approximately p16.3 (Chr2:18179-47899074), and 17q 21.32 approximately q25.3 (Chr17:42153031-78607159), while chromosomal losses detected were the distal deletion of 1p36.33 (Chr1:552910-563807), 14q21.1 approximately q21.3 (Chr14:37666271- 47282550), and 22q13.1 approximately q13.2 (Chr22:36885764-4190 7123). Except for the gain in 17q21 and the loss in 1p36, the other regions of gain or loss in SH-SY5Y cells were newly identified.     

Functional complement C1q abnormality leads to impaired immune complexes and apoptotic cell clearance

C1q plays a key role in apoptotic cell and immune complex removal. Its absence contributes to the loss of tolerance toward self structures and development of autoimmunity. C1q deficiencies are extremely rare and are associated with complete lack of C1q or with secretion of surrogate C1q fragments. To our knowledge, we report the first case of a functional C1q abnormality, associated with the presence of a normal C1q molecule. Homozygous GlyB63Ser mutation was found in a patient suffering from lupus with neurologic manifestations and multiple infections. The GlyB63Ser C1q bound to Igs, pentraxins, LPSs, and apoptotic cells, similarly to C1q from healthy donors. However, the interaction of C1r(2)C1s(2) and C1 complex formation was abolished, preventing further complement activation and opsonization by C3. The mutation is located between LysB(61) and LysB(65) of C1q, suggested to form the C1r binding site. Our data infer that the binding of C1q to apoptotic cells in humans is insufficient to assure self-tolerance. The opsonization capacity of C4 and C3 fragments has to be intact to fight infections and to prevent autoimmunity.     

Multiple-interactions among EMILIN1 and EMILIN2 N- and C-terminal domains

EMILIN1 and EMILIN2 belong to a family of extracellular matrix glycoproteins characterized by the N-terminal cysteine-rich EMI domain, a long segment with high probabilty for coiled-coil structure formation and a C-terminal gC1q domain. To study EMILIN1 and EMILIN2 interaction and assembly we have applied qualitative and quantitative two hybrid systems using constructs corresponding to the gC1q and EMI domains. The identified interactions were further confirmed in yeast extracts of co-transfected cells followed by co-immunoprecipitation. The data indicated that gC1q domains are able to self-interact as well as to interact one each other and with the EMI domains, but no self interactions were detected between the EMI domains. Furthermore EMILINs interactions were studied in 293-EBNA cells co-transfected with full lenght EMILIN1 and EMILIN2 constructs. Specific antibodies were able to co-immunoprecipitate EMILINs, indicating that also full-lenght proteins can give rise to non-covalent homo- and hetero-multimers even if reduced and alkylated before mixing. Immunofluorescence analysis on mouse cell cultures and tissues sections with specific antibodies showed co-distribution of EMILIN1 and EMILIN2. Thus, we can hypothesize that EMILINs multimers are formed by head-to-tail interaction between C-terminal and N-terminal domains of EMILIN1 and/or EMILIN2 but also by tail-to-tail interaction between gC1q domains. These multiple interactions may regulate homo-typic and/or hetero-typic linear and eventually lateral branching assemblies of EMILIN1 and EMILIN2 in tissues.