Expression of gelatinase B in trachomatous conjunctivitis.

BACKGROUND/AIMS: Gelatinase B is a matrix metalloproteinase involved in extracellular matrix (ECM) breakdown often associated with scarring and other pathological disorders. It was investigated whether gelatinase B is involved in the pathogenesis of ECM degradation associated with trachomatous conjunctivitis. METHODS: Conjunctival biopsy specimens obtained from six patients with active trachoma, six patients with active vernal keratoconjunctivitis (VKC), and seven control subjects were studied. Immunohistochemical techniques and a specific monoclonal antibody against human gelatinase B were used, and a monoclonal antibody against macrophage CD68 to identify mononuclear cells with gelatinase B immunoreactivity. In addition, quantitative zymography was used to compare the activity of gelatinase B in conjunctival biopsy specimens from seven patients with active trachoma and seven control subjects. RESULTS: Gelatinase B was detected by immunohistochemistry only in polymorphonuclear cells located in the vascular lumens in three normal conjunctival biopsy specimens. In all trachoma specimens and in five VKC specimens, gelatinase B was localised in monocyte/macrophage cells, positive for the CD68 marker, and in polymorphonuclear cells. The majority of the latter cell type was located in intravascular spaces. Compared with VKC specimens, trachoma specimens showed significantly more immunoreactive gelatinase B monocyte/macrophage cells (52.3 (21.9) v 8.2 (6.4); p <0.001) and polymorphonuclear cells (23.2 (14.2) v 6.3 (5.4); p = 0. 013). Activated macrophages with giant cell morphology clearly stained with the gelatinase B specific monoclonal antibody were observed in trachoma specimens. Zymography revealed that gelatinase B levels in trachoma specimens were significantly higher than the levels found in normal conjunctiva (1739.6 (1078.3) v 609.3 (395.9) scanning units; p = 0.0127). CONCLUSIONS: The increased activity of gelatinase B and numbers of inflammatory cells containing gelatinase B in trachoma specimens suggest that this enzyme plays a part in the pathogenesis of conjunctival scarring in trachoma.

Resistance of young gelatinase B-deficient mice to experimental autoimmune encephalomyelitis and necrotizing tail lesions.

Regulated expression of matrix metalloproteinases (MMPs) and their inhibitors (TIMPs) plays a role in various physiological processes. To determine in vivo how unbalanced expression of these factors can promote or affect the course of pathologies, we knocked out the mouse gelatinase B gene by replacing the catalytic and zinc-binding domains with an antisense-oriented neomycin resistance gene. Adult gelatinase B-deficient mice and wild-type controls could be induced to develop experimental autoimmune encephalomyelitis (EAE) with similar scores for neurologic disease, blood-brain barrier permeability, and central nervous system histopathology. However, whereas diseased control animals showed necrotizing tail lesions with hyperplasia of osteocartilaginous tissue, adult gelatinase B-deficient mice were resistant to this tail pathology. Gelatinase B-deficient mice younger than 4 weeks of age were significantly less susceptible to the development of EAE than were age matched controls and, even as they aged, they remained resistant to tail lesions. These data illustrate that gelatinase B expression plays a role in the development of the immune system and that, in ontogenesis, the propensity to develop autoimmunity is altered by the absence of this MMP.

Prevention of interleukin-8-induced mobilization of hematopoietic progenitor cells in rhesus monkeys by inhibitory antibodies against the metalloproteinase gelatinase B (MMP-9).

Previously, we demonstrated that IL-8 induces rapid mobilization of hematopoietic progenitor cells (HPC) from the bone marrow of rhesus monkeys. Because activation of neutrophils by IL-8 induces the release of gelatinase B (MMP-9), which is involved in the degradation of extracellular matrix molecules, we hypothesized that MMP-9 release might induce stem cell mobilization by cleaving matrix molecules to which stem cells are attached. Rhesus monkeys were treated with a single i.v. injection of 0.1 mg/kg human IL-8, which resulted in a 10- to 100-fold increase in HPC within 30 min after injection. Zymographic analysis revealed a dramatic instantaneous increase in the plasma levels of MMP-9, followed by the increase in circulating HPC. Enzyme levels decreased at 2 h after injection of IL-8, simultaneously with the decrease in the numbers of circulating HPC. To test the hypothesis that MMP-9 induction was involved in HPC mobilization, rhesus monkeys were treated with a highly specific inhibitory monoclonal anti-gelatinase B antibody. Anti-gelatinase B at a dose of 1-2 mg/kg completely prevented the IL-8-induced mobilization of HPC, whereas a dose of 0.1 mg/kg had only a limited effect. Preinjection of inhibitory antibodies did not preclude the IL-8-induced production and secretion of MMP-9. Pretreatment with an irrelevant control antibody did not affect IL-8-induced mobilization, showing that the inhibition by the anti-gelatinase B antibody was specific. In summary, IL-8 induces the rapid systemic release of MMP-9 with concurrent mobilization of HPC that is prevented by pretreatment with an inhibitory anti-gelatinase B antibody, indicating that MMP-9 is involved as a mediator of the IL-8-induced mobilization of HPC.

Oligosaccharides of recombinant mouse gelatinase B variants.

Gelatinase B (matrix metalloproteinase-9, MMP-9) contains three N-glycosylation sites and a Ser/Thr/Pro-rich type V collagen domain with repetitive attachment sites for O-linked sugars. Recombinant mouse gelatinase B was expressed in the yeast Pichia pastoris and the N-linked oligosaccharides of the truncated glycoprotein variants were analysed by in gel enzymatic release followed by mass spectrometry and normal phase HPLC. This technology, despite of the limiting amount of material, allowed the analysis of the formula of N- and O-linked sugars of the different glycoprotein variants. The 112/99- and 88-kDa gelatinase B forms each contained an oligomannose series (Man8GlcNAc2 to Man15GlcNAc2). Analysis of the hydrazine-released sugars showed that the O-linked oligosaccharides contained alpha1-2, alpha1-3 or alpha1-6 linked mannoses. These results were confirmed by lectin blot analysis of intact and glycosidase-treated enzyme variants.

Synergistic induction of MCP-1 and -2 by IL-1beta and interferons in fibroblasts and epithelial cells.

Monocyte chemotactic protein (MCP)-1 and MCP-2, two closely related CC chemokines, are important mediators of monocyte and lymphocyte migration. These chemokines are secreted by various normal cell types, including fibroblasts, epithelial cells, and leukocytes, as well as by tumor cells. After stimulation with different cytokines and cytokine inducers the MCP-2 production levels are always lower than those of MCP-1. In human diploid fibroblasts cytokines differentially regulate chemokine induction, interleukin (IL)-1beta and interferon (IFN)-gamma being potent stimuli of MCP-1 and MCP-2, respectively. Co-stimulation of fibroblasts by 10 U/mL IL-1beta and 20 ng/mL IFN-gamma resulted in a synergistic induction of MCP-2, whereas the combined effect on MCP-1 and IL-6 production was rather additive. These findings were confirmed at the mRNA level by Northern blot analysis. In contrast, in human MG-63 fibroblastoid cells and HEp-2 epithelial cells, selected for their poor responsiveness to IL-1beta and IFN-gamma, MCP-2 as well as MCP-1 and IL-6 were synergistically induced, yielding protein levels that were increased 3- to 30-fold above the additive levels. When IFN-beta was used as a co-stimulant of IL-1beta, a similar synergistic induction of MCP-1 and MCP-2 was measured both at the protein and the mRNA level. It can be concluded that, when synergy occurred, the MCP-1 and MCP-2 expression levels reached a comparable maximum, indicative for an equal contribution of these chemokines in normal and pathological conditions.

Cloning and expression in Escherichia coli of a human gelatinase B-inhibitory single-chain immunoglobulin variable fragment (scFv).

The murine monoclonal antibody REGA-3G12 selectively and specifically inhibits the activity of human gelatinase B. The cDNA fragments which encode the variable regions of the light and heavy chains were isolated by PCR-mediated cloning and sequenced. Single-chain Fv expression constructs for Escherichia coli were generated in which c-myc tag sequences were encoded. Inducible expression of the scFv and secretion to the periplasm were obtained with higher yields when the c-myc tag sequence was positioned at the amino-terminal side. The inhibitory activity of purified scFv on neutrophil gelatinase B was tested in a gelatin degradation assay and it was found to possess a similar specific activity as that of the intact monoclonal antibody and of the pepsin-clipped F(ab')2 derivative. This shows for the first time that inhibition of soluble enzymes with scFv is possible and opens new perspectives for the treatment of diseases with excessive and detrimental enzyme production in the host.

Expression of gelatinase B and the extracellular matrix metalloproteinase inducer EMMPRIN in benign and malignant pigment cell lesions of the skin.

By the degradative effect on basement membrane collagen type IV, matrix metalloproteinases (MMPs) or gelatinases are important in the early invasion of malignant tumors. These enzymes may be released by the tumor cells themselves or may be derived from nearby fibroblasts that have been stimulated by the extracellular MMP inducer EMMPRIN. We studied the distribution of 92-kd gelatinase B (MMP-9) and of EMMPRIN in 33 benign and 41 malignant, paraffin-embedded pigment cell lesions using immunohistochemistry and monoclonal antibodies. In benign pigment cell lesions, EMMPRIN but not gelatinase B was expressed in cellular blue nevi whereas all other benign lesions, including common blue nevi, were negative. In malignant melanomas (MMs), both gelatinase B and EMMPRIN were variably expressed in the pure and invasive radial growth phase but not in the vertical growth phase. All lentigo maligna cases and all metastatic lesions were negative. Of MMs with thickness < 1.6 mm, 63% expressed gelatinase B and 70% expressed EMMPRIN, whereas in MMs with > 1.6 mm thickness, only 10% expressed gelatinase B and only 25% expressed EMMPRIN. We conclude that early invasion of MM is associated with de novo expression of gelatinase B and EMMPRIN by neoplastic melanocytes. Expression of EMMPRIN and MMP-9 may be partly responsible for the stromal changes observed in thin MM. Their absence in the vertical growth phase and in metastatic lesions suggests that other factors are involved in tissue degradation during later stages of tumor progression in MM. The lack of both gelatinase B and EMMPRIN in lentigo maligna may contribute to the indolent behavior of this type of pigment cell lesion.

Gelatinase B in chronic synovitis: immunolocalization with a monoclonal antibody.

Gelatinase B is a matrix metalloproteinase (MMP-9) involved in the remodelling of extracellular matrices of connective tissues. With the use of specific monoclonal antibodies against human gelatinase B, the producer cell types were pinpointed in histopathological sections of a number of arthritic diseases. In cases of acute joint trauma, chondromatosis, villonodular synovitis and a cyst of a bursa, high numbers of strongly immunopositive neutrophils were observed in addition to weaker staining macrophages. Activated macrophages with giant cell morphology clearly stained with the gelatinase B-specific monoclonal antibody in the case of villonodular synovitis and in an epidermoid cyst. However, in the sections from patients with rheumatoid arthritis, no immunostaining was seen. In other cases of chronic synovitis, however, within the lymphocyte nodular aggregates a strong gelatinase B expression was observed in morphologically identified dendritic cells. In conclusion, gelatinase B production in joint disease seems to be predominantly by neutrophils and cell types of the macrophage/antigen-presenting cell lineage.

Induction of gelatinase B and MCP-2 in baboons during sublethal and lethal bacteraemia.

Intravenous injection of sublethal or lethal doses of Escherichia coli in baboons resulted in increased serum levels of the matrix metalloprotease gelatinase B and the chemokine monocyte chemotactic protein 2 (MCP-2). In both animal models, gelatinase B appeared faster than MCP-2. After sublethal challenge, serum levels of gelatinase B and MCP-2 were found to be correlated, reaching peak levels between 2 and 4 h after bacterial challenge. After lethal challenge, however, MCP-2 tended to increase until 10 h. The kinetics of appearance suggest induction of release of gelatinase B and de novo synthesis and secretion of MCP-2, both by endotoxin.

False positive immunostaining of Locusta neurosecretory cells with a variety of preimmune sera.

A large number of antisera directed against vertebrate neuropeptides have been reported to yield positive staining when applied to insect brains. In most cases, the preimmune serum of the same animal in which the antiserum was developed is not available for testing in control experiments. We have experienced that a large percentage of preimmune sera, as well as a culture medium for hybridomas, stain cell populations and fibers in the central nervous system of the insect Locusta migratoria. Purification of these preimmune sera on a Protein A and Protein G support indicates that the reaction is due to preexisting antibodies of the IgG class. Western analysis of brain and nervous tissue extracts indicates the presence of two immunoreactive 27-kDa bands. These bands could also be visualized in other tissue extracts such as muscle, midgut, Malpighian tubules, and fat body of Locusta. The brain of other insect species, such as Periplaneta americana, Leucophaea maderae, and Neobellieria bullata were devoid of the false immunopositive reaction. There is no easy way to eliminate this type of immunoreaction. It follows that when affinity chromatographic purification of the antibody is not feasible, it is essential to include in the control procedure, the preimmune serum of the animal that was used for the production of the antiserum. This means that it should become common practice to sell or exchange sera together with their corresponding preimmune sera.

Production and characterization of recombinant active mouse gelatinase B from eukaryotic cells and in vivo effects after intravenous administration.

Gelatinase B is a matrix metalloproteinase involved in tissue remodelling. When mouse cells are triggered in vitro with interleukin-1, bacterial endotoxin, virus-mimicking double-stranded RNA or cytokine inducers, they produce gelatinase B. To test the effects of gelatinase B in vivo, the enzyme was expressed in Chinese hamster ovary (CHO) cells. Hybrid genomic DNA-cDNA constructs under the control of two constitutive viral promoters were generated by PCR-mediated exon amplification. In vitro transcription and translation of the mRNA in reticulocyte lysate yielded the correct 79-kDa protein, and expression in CHO cells resulted in an intact glycosylated 110-kDa gelatinase B which was enzymically active. However, the production yields of recombinant enzyme from 50 tested clones were low and cell-culture supernatants contained significant amounts of copurifiable endogenous CHO gelatinase B. Therefore, the enzyme was expressed in the yeast Pichia pastoris. Recombinant proenzyme was secreted and recovered from the yeast culture medium at 10 mg/l. Amino-terminal sequence analysis indicated that affinity purification of the recombinant protein on gelatin-Sepharose yielded the expected N-glycosylated proenzyme form (110 kDa) in addition to an amino-terminally truncated unglycosylated variant (69 kDa). Both forms had gelatinolytic activity on zymography. The recombinant mouse gelatinase B was used to determine its pharmacokinetics and its haematological effects in vivo. After intravenous injection in rabbits, gelatinase B disappeared from the circulation within 6 h. In addition to a transient leukopenia, we observed a rapid increase in leukocytosis, which indicates that gelatinase B might be a factor involved in the desorption of adherent leukocytes from the vascular bed and in the release of leukocytes from the bone marrow. Gelatinase B secretion and activation might well be one of the crucial molecular mechanisms explaining leukocytosis which is associated with infections and almost all types of inflammation.

The expression of tissue-type plasminogen activator, matrix metalloproteases and endogenous inhibitors in the central nervous system in multiple sclerosis: comparison of stages in lesion evolution.

The expression of tissue-type plasminogen activator (t-PA) and a number of metalloproteases as well as plasminogen activator inhibitor-1 (PAI-1) and tissue inhibitor of metalloproteases-1 (TIMP-1) was analyzed in the central nervous system (CNS) of normal control and multiple sclerosis (MS) cases by immunohistopathology. The expression of t-PA was detectable only in the blood vessel matrix in control white matter, but positive infiltrating mononuclear cells were also observed in MS white matter and primary lesions. In active plaques this pattern converted to strong positivity of foamy macrophages in areas of demyelination, declining in chronic lesions. In general PAI-1 expression paralleled that of t-PA. Gelatinase A and B were detected predominantly in astrocytes and microglia throughout normal control white matter, with additional positive mononuclear cells in perivascular cuffs in MS white matter. In the demyelinating lesion there is widespread prominent expression of gelatinase B in reactive astrocytes and macrophages, which persists in astrocytes in the chronic lesion. TIMP-1 was also present in the vessel matrix and in lesional macrophages. These observations on the coexpression of enzymes and inhibitors of the matrix degrading cascade in CNS tissue pinpoint t-PA, a rate-limiting enzyme, and gelatinase B as therapeutic targets in MS.

The gelatinase inhibitory activity of tetracyclines and chemically modified tetracycline analogues as measured by a novel microtiter assay for inhibitors.

A quantitative nonisotopic solution assay for gelatinases and inhibitors was developed using biotinylated gelatin as enzyme substrate. In this assay, residual biotinylated substrate is sandwiched between avidin-coated plates and streptavidin-peroxidase and is quantified by the peroxidase reaction. This assay was useful for measuring gelatinase activities and defining the activities of gelatinase inhibitors. When 23 tetracycline analogues were compared, significant differences in gelatinase B inhibition were found between various compounds. 4-epioxytetracycline base, 4-epichlortetracycline, meclocyclinesulfosalicylate, and unmodified metacycline and minocycline proved to be the most potent gelatinase B (EC 3.4.24.35) inhibitors. The gelatinase B inhibitory activity of tetracyclines was clearly dissociated from their antimicrobial activity. The effect of high-molecular-weight inhibitors, such as monoclonal antibodies, was also demonstrable in the microtiter plate assay. In view of the pathophysiological function of gelatinases, the definition of gelatinase inhibitors with known efficacy, safety, and side effects is crucial for the treatment of diseases such as rheumatoid arthritis and multiple sclerosis. Particular tetracyclines fulfil these criteria and the described assay is useful for defining other gelatinase-inhibiting lead compounds.

Human gelatinase B, a marker enzyme in rheumatoid arthritis, is inhibited by D-penicillamine: anti-rheumatic activity by protease inhibition.

The direct and indirect inhibitory potential of D-penicillamine toward human neutrophil and synovial fluid gelatinase B, a marker enzyme for disease severity in RA, was investigated. Gelatinase and plasminogen activator activities were assessed by SDS-polyacrylamide gel electrophoresis zymography. D-penicillamine significantly inhibits purified and synovial fluid gelatinase B in vitro at concentrations attainable in vivo and also blocks in vitro plasminogen activation. Protease inhibition may be a mechanism of action for D-penicillamine as DMARD.

The presence and effects of mammalian signal molecules in immunocytes of the insect Leucophaea maderae.

Opioid peptides activate immunocytes and opiate alkaloids inhibit this activation in the mussel, Mytilus edulis. Here we present evidence that cells of another invertebrate, Leucophaea maderae, can be influenced in a similar way by the Met-enkephalin analogue D-Ala2-Met5-enkephalin (DAMA) and morphine. Effects of different signal molecules on Leucophaea hemocytes were evaluated by computer-assisted image analysis of their conformational state. A small percentage of the untreated cells were found to display spontaneous conformational changes after 25 min of incubation without pharmacological agents which was noted as a decrease in both circularity factor and shape factor values. Activation caused the cells to become elliptical, a feature that appears to be characteristic of Leucophaea immunocytes. Administration of DAMA induced a similar activation of most of the cells. After 30 min these DAMA-activated cells started to display distinct locomotory activity not seen in the controls. alpha-Melanocyte-stimulating hormone (MSH, 10(-7)) added to the incubation medium after DAMA-activation caused the cells to return to their original "rounded" conformation. In addition, the presence of immunoreactive interleukin (IL-1), adrenocorticotropin (ACTH) and tumor necrosis factor (TNF) in the hemolymph was demonstrated. These data suggest an interaction between both vertebrate-type immunological signal molecules and neuropeptides in the regulation of immunological cells in Leucophaea.

Prevention of acute autoimmune encephalomyelitis and abrogation of relapses in murine models of multiple sclerosis by the protease inhibitor D-penicillamine.

The in vitro activity of gelatinase B, an enzyme whose appearance in the cerebrospinal fluid is associated with inflammatory diseases of the central nervous system, was dose-dependently inhibited by the antirheumatic D-penicillamine. Inhibition of gelatinase B in electrophoretically pure preparations and in cell culture supernatants and human body fluids was obtained at dosages reached in the circulation of patients treated with a peroral dosis of 750 mg D-penicillamine per day. In mice, developing acute demyelination, D-penicillamine significantly reduced the mortality and morbidity rates of experimental allergic encephalomyelitis (EAE). In chronic relapsing EAE in Biozzi AB/H mice, an animal model for relapses in multiple sclerosis (MS), it attenuated the exacerbations, even when the treatment was started after the primary full-blown disease had developed. We infer protease inhibition as the mechanism of action of D-penicillamine and suggest that its use may be effective as peroral treatment for MS.

Monoclonal antibodies specific for natural human neutrophil gelatinase B used for affinity purification, quantitation by two-site ELISA and inhibition of enzymatic activity.

Human gelatinase B was produced from peripheral blood neutrophils and purified by affinity chromatography on gelatin sepharose. This material was used as an antigen to prepare mouse monoclonal antibodies (mAb). The resulting hybridomas were selected on the basis of binding to biotinylated antigen and by a sandwich ELISA using gelatinase-B-specific polyclonal rabbit antiserum and pure natural antigen. Five of these mAb were selected for further characterization. They all displayed variable epitope specificity, binding capacity and inhibitory activity. Whereas mAb REGA-2D9 and REGA-3G12 showed the strongest binding to biotinylated gelatinase B and natural gelatinase B, respectively, mAb REGA-2F9 did not bind biotinylated antigen. None of the mAb displayed cross-reactivity to gelatinase A in a direct ELISA. The mAb REGA-1G8 was found to cross-react with human serum albumin. The binding capacity of the other four mAb with leukocyte gelatinase B was compared and a sensitive sandwich ELISA was developed with the antibodies REGA-3G12 and REGA-2D9 (detection limit 0.5 ng/ml). The mAb REGA-3G12 was unique in that it inhibited catalysis by gelatinase B. This was shown by assaying the degradation of nasal septum type II gelatin in the presence and absence of each of the five mAb. Furthermore, mAb REGA-3G12 inhibited the degradation of biotinylated gelatin in a microtiterplate solution assay. In addition to the potential use of the inhibitory mAb REGA-3G12 in the treatment of diseases with excessive gelatinase B production, several of the described mAb are useful as diagnostic probes to detect gelatinase B in body fluids and tissue samples of patients with multiple sclerosis, rheumatoid arthritis and cancer.

Expression of a human mutant monocyte chemotactic protein 3 in Pichia pastoris and characterization as an MCP-3 receptor antagonist.

The cDNA encoding human monocyte chemotactic protein 3 (hMCP-3) was cloned in pHIL-S1, a vector designed for inducible secreted heterologous expression in the methylotrophic yeast Pichia pastoris. After transformation of P. pastoris by electroporation, several clones with the human MCP-3 gene integrated at the alcohol oxidase (AOX-1) locus were isolated. One of these clones (M30) expressed the mature MCP-3 protein with three additional amino acids at its NH2 terminus as a secretion product in the supernatant. The recombinant protein comigrated on SDS-PAGE and cross-reacted immunologically with synthetic hMCP-3. Intermediate-scale production in shake flasks was obtained at expression levels of approximately 1 mg per liter. The recombinant mutant MCP-3 was purified to homogeneity by adsorption on silicic acid, affinity chromatography on heparin-Sepharose, and reversed-phase HPLC. At the amino terminus of the purified recombinant protein, the presence of the additional sequence Arg-Glu-Phe was confirmed by direct protein sequence analysis. The recombinant hMCP-3 mutein was not glycosylated, as evidenced by deglycosylation experiments and by mass spectrometry. In analogy with MCP-1, the amino terminus of MCP-3 is crucial for its agonistic effect on receptive cells. At concentrations up to 3.5 micrograms/ml, the recombinant mutein was not active in vitro as a chemotactic factor for monocytes. However, the mutant MCP-3 acted as an MCP-3 receptor antagonist in a competition chemotaxis assay at 100- to 1000-fold excess over the synthetic MCP-3 agonist. It might thus be a useful tool to study antagonism of MCP-3 action in vitro and in disease models of cancer and inflammation.

Plasma levels of the chemokines monocyte chemotactic proteins-1 and -2 are elevated in human sepsis.

Because of their effects on monocytes, monocyte chemotactic proteins-1 and -2 (MCP-1 and MCP-2) may participate in the pathophysiology of sepsis. We measured circulating MCP-1 and MCP-2 levels in 42 septic patients having positive local or blood cultures. MCP-1 and MCP-2 levels were elevated in 24 (57%) and 25 (59%) of 42 septic patients, respectively, compared with healthy volunteers. Both patients with gram-positive and gram-negative infections had elevated MCP-1 plasma levels (P = .0001) and P < .0001), respectively; Mann-Whitney-U test), whereas patients with gram-positive infection, but not those with gram-negative infection, had increased MCP-2 plasma levels (P= .0182). No relative differences in MCP-1 and MCP-2 plasma levels were observed between several subgroups of patients (sepsis v septic shock; survivors v nonsurvivors), although levels of MCP-1 were the highest in patients with the more severe forms of sepsis, ie, those with shock or a lethal outcome. Serial observations showed that MCP-1 and MCP-2 plasma levels remained elevated for at least 48 hours. MCP-1 correlated weakly with interleukin-8 and MCP-2, the correlations for which were most pronounced in patients with septic shock. MCP-2 correlated with interleukin-8, and surprisingly, with the complement activation product C3a; these correlations further improved when analyzing patients with septic shock or when applying gram-positive infections. Thus, our results not only show increased MCP-1 and MCP-2 levels in patients with sepsis, but also suggest that the synthesis and release of MCP-1 and MCP-2 in sepsis are differently regulated in part.