The rapid and active transport of polymeric IgA from the blood to the bile (1, 2) is selective (3, 4), which suggests the existence of a specific receptor for this molecule. The liver cells responsible for the transfer have been identified as the hepatocytes by autoradiography and electron microscopy of the livers of rats killed 5, 30, and 60 min after the injection of radiolabeled IgA (5). 5 rain after being injected, the radiolabeled IgA appeared to be bound to the sinusoidal membrane of the hepatocytes. We have found that intravenously injected radiolabeled IgM, IgG1, and monomeric IgA did not appear in the bile at all, and that the transport into bile of s-IgA (i.e., polymeric IgA bound to secretory component [SC]) was both considerably slower and quantitatively less than that of polymeric IgA devoid of SC. Studies of the transport into the bile of IgG2a, IgE, IgM, monomeric and polymeric IgA, and s-IgA with isolated perfused rat livers gave similar results: only polymeric IgA that lacked SC was actively transported (6). These facts suggest that SC may be the receptor for IgA on the surface of rat hepatocytes as it is on the epithelial cells from human gut (7). The availability of hepatocytes in short-term monolayer culture (8), together with that of specific antibodies to rat SC, has enabled us to test this hypothesis by experiments in vitro. Materials and MethodsPreparation of lmmunoglobulins, Free SC, and Specifically Purified Antibody to SC. The preparation and properties of the polymeric rat IgA formed by the IR 461 plasmacytoma have been described (1). Rat IgG was the fraction of normal rat serum eluted from DEAE cellulose with 0.01 M phosphate buffer at pH 7.2. Normal rat bile is a good source of both s-IgA and free SC (9). Pooled bile was first treated with 15-20 mg/ml of eholestyramine resin (Cuemid; Merck, Sharpe and Dohme, Ltd., Hoddesdon, Herts., U. K.), to reduce the amount of pigment. After 10-fold concentration by ultrafiltration, followed by centrifugation, the bile was fractionated on DEAE cellulose. All protein, reactive with antibody to either a-chain or to SC, eluted together with a phosphate buffer (0.05 M, pH 6.2) that contained 0.2-0.3 M NaCI. Gel filtration of this material on a column of Ultrogel AcA34 (LKB Instruments, Croydon, Surrey, U. K.) gave an excluded peak that contained IgA and a second peak that contained free SC and albumin. Further gel filtration of these two fractions on Ultrogel AcA22 and AcA34, respectively, yielded a preparation of pure polymeric s-IgA and one of free SC that still contained some serum albumin; the antigenic properties of the biliary SC are shown in Fig. I a. Rabbit antiserum obtained after three injections of a crude SC preparation was absorbed first on an immunoadsorbent column of normal rat serum proteins coupled to AH-Sepharose (Pharmaeia Fine Chemicals, Inc., Piscataway, N. J.) and then on a similar column of IR 461 myeloma IgA. The specificity of the anti-SC serum is shown in Fig. 1 b. Specifically purified antibody was made from this antiserum...

Tuberculosis (TB) of the central nervous system (CNS) is a deadly disease characterized by extensive tissue destruction, driven by molecules such as Matrix Metalloproteinase-2 (MMP-2) which targets CNS-specific substrates. In a simplified cellular model of CNS TB, we demonstrated that conditioned medium from Mycobacterium tuberculosis-infected primary human monocytes (CoMTb), but not direct infection, unexpectedly down-regulates constitutive microglial MMP-2 gene expression and secretion by 72.8% at 24 hours, sustained up to 96 hours (P < 0.01), dependent upon TNF-α. In human CNS TB brain biopsies but not controls the p38 pathway was activated in microglia/macrophages. Inhibition of the p38 MAP kinase pathway resulted in a 228% increase in MMP-2 secretion (P < 0.01). In contrast ERK MAP kinase inhibition further decreased MMP-2 secretion by 76.6% (P < 0.05). Inhibition of the NFκB pathway resulted in 301% higher MMP-2 secretion than CoMTb alone (P < 0.01). Caspase 8 restored MMP-2 secretion to basal levels. However, this caspase-dependent regulation of MMP-2 was independent of p38 and NFκB pathways; p38 phosphorylation was increased and p50/p65 NFκB nuclear trafficking unaffected by caspase 8 inhibition. In summary, suppression of microglial MMP-2 secretion by M.tb-infected monocyte-dependent networks paradoxically involves the pro-inflammatory mediators TNF-α, p38 MAP kinase and NFκB in addition to a novel caspase 8-dependent pathway.

No abstract