An initial exposure to lipopolysaccharide (LPS) induces a transient state of hyporesponsiveness to a subsequent challenge with LPS. The mechanism underlying this phenomenon, termed endotoxin tolerance, remains poorly understood despite a recent resurgence of interest in this area. We demonstrate herein that SHIP(-/-) bone marrow-derived macrophages (BMmphis) and mast cells (BMMCs) do not display endotoxin tolerance. Moreover, an initial LPS treatment of wild-type BMmphis or BMMCs increases the level of SHIP, but not SHIP2 or PTEN, and this increase is critical for the hyporesponsiveness to subsequent LPS stimulation. Interestingly, this increase in SHIP protein is mediated by the LPS-induced production of autocrine-acting TGFbeta and neutralizing antibodies to TGFbeta block LPS-induced endotoxin tolerance. In vivo studies with SHIP(+/+) and SHIP(-/-) mice confirm these in vitro findings and show a correlation between the duration of endotoxin tolerance and elevated SHIP levels.

We recently reported that SHIP restrains LPS-induced classical (M1) activation of in vitro differentiated, bone marrow-derived macrophages (BMMPhis) and that SHIP upregulation is essential for endotoxin tolerance. Herein, we show that in vivo differentiated SHIP-/- peritoneal (PMPhis) and alveolar (AMPhis) macrophages, unlike their wild-type counterparts, are profoundly M2 skewed (alternatively activated), possessing constitutively high arginase I (ArgI) and Ym1 levels and impaired LPS-induced NO production. Consistent with this, SHIP-/- mice display M2-mediated lung pathology and enhanced tumor implant growth. Interestingly, BMMPhis from SHIP-/- mice do not display this M2 phenotype unless exposed to TGFbeta within normal mouse plasma (MP) during in vitro differentiation. Our results suggest that SHIP functions in vivo to repress M2 skewing and that macrophage polarization can occur during differentiation in response to TGFbeta if progenitors have elevated PIP3.

Mcl-1 protein expression was found to be up-regulated during infection with virulent Mycobacterium tuberculosis strain H37Rv. Mcl-1 induction in THP-1 cells was optimal at a multiplicity of infection of 0.8–1.2 bacilli per macrophage and was independent of opsonin coating of the bacteria. Mcl-1 expression was elevated as early as 4 h, peaked at 5.8-fold above control cells at 24 h, and remained elevated at 48 h after infection. In THP-1 cells, mMcl-1 mRNA was induced by infection with live H37Rv but not with attenuated M. tuberculosis strain H37Ra, heat-killed H37Rv, or latex beads. In THP-1 cells and monocyte-derived macrophages (MDMs), Mcl-1 protein was induced by infection with live H37Rv but not with attenuated M. tuberculosis strain H37Ra, heat-killed H37Rv, or latex beads. Treatment of uninfected, H37Ra-infected, and H37Rv-infected THP-1 cells and MDMs with antisense oligonucleotides to mcl-1 reduced Mcl-1 expression by >84%. This resulted in an increase in apoptosis of both MDMs and THP-1 cells that were infected with H37Rv, but not cells that were uninfected or infected with H37Ra. Increased apoptosis correlated with a decrease in M. tuberculosis CFUs recovered from antisense-treated, H37Rv-infected cells at 4 and 7 days after infection. In contrast, CFU recoveries from sense-treated, H37Rv-infected cells or from antisense- or sense-treated, H37Ra-infected cells were unchanged from controls. Thus, the antiapoptotic effect of the induction of Mcl-1 expression in H37Rv-infected macrophages promotes the survival of virulent M. tuberculosis.

Macrophages infected with Mycobacterium tuberculosis undergo increased rates of apoptosis. Important objectives are to define the microbial factors that cause apoptosis, the mechanisms involved and the impact on infection. The 19-kDa M. tuberculosis glycolipoprotein (p19) is both cell wall-associated and secreted and is a candidate virulence factor. We investigated the potential of recombinant, His-tagged p19 lacking the secretion/acylation signal to induce macrophage apoptosis. The TUNEL assay and annexin V binding to membrane phosphatidylserine were used to measure apoptosis. The results show that p19 does act to induce apoptosis in differentiated THP-1 cells and monocyte-derived macrophages and that this effect is both dose- and time-dependent. Furthermore, this effect of p19 is Toll-like receptor (TLR)-2-mediated because preincubation of either THP-1 cells or TLR-2-expressing CHO cells with anti-TLR-2 mAb inhibited apoptosis induced by p19. Apoptosis of macrophages in response to p19 was found to be caspase-8 dependent and caspase-9 independent consistent with a transmembrane pathway signaling cell death through TLR-2. The viability of M. tuberculosis in cells undergoing apoptosis induced by p19 was significantly reduced suggesting the possibility that this may favor containment of infection. Although native p19 is a mycobacterial glycolipoprotein, based upon the use of recombinant p19 where the acylation signal had been removed, we conclude that it is the polypeptide component of p19 that is responsible for signaling through TLR-2 and that the lipid moiety is not required.