To study mechanisms of peripheral self-tolerance, we injected small numbers of naive CD4+ TCR-transgenic T cells into mice expressing the MHC/peptide ligand under the control of an MHC class II promoter. The donor T cells expand rapidly to very large numbers, acquire memory markers, and go out into tissues, but the animals remain healthy, and the accumulated T cells are profoundly anergic to restimulation with Ag in vitro. Provision of a costimulatory signal by coinjection of an agonist Ab to OX40 (CD134), a TNFR family member expressed on activated CD4 T cells, results in death of the mice within 12 days. TCR-transgenic T cells recovered at 5 days from anti-OX40-treated mice have a unique phenotype: they remain unresponsive to Ag in vitro, but they are larger, more granular, and strongly IL-2R positive. Some spontaneously secrete IFN-γ directly ex vivo, and the majority make IFN-γ in response to PMA and ionomycin. Although they are anergic by conventional tests requiring Ag recognition, they respond vigorously to cytokines, proliferating in response to IL-2, and secreting IFN-γ when TCR signaling is bypassed with IL-12 and IL-18. We conclude that the costimulatory signal through OX40 allows otherwise harmless, proliferating, autoreactive T cells to acquire effector cell functions. The ability of these T cells to respond to cytokines by synthesizing additional inflammatory cytokines without a TCR signal may drive the fatal pathogenic process in vivo.

Bioactive glasses form a surface apatite layer in vivo that enhances the formation and attachment of bone. Sol-gel Bioglass graft material provides greater nanoscale porosity than bioactive glass (on the order of 50-200 A), greater particle surface area, and improved resorbability, while maintaining bioactivity. This study histologically and biomechanically evaluated, in a rabbit model, bone formed within critical-sized distal femoral cancellous bone defects filled with 45S5 Bioglass particulates, 77S sol-gel Bioglass, or 58S sol-gel Bioglass and compared the bone in these defects with normal, intact, untreated cancellous bone and with unfilled defects at 4, 8, and 12 weeks. All grafted defects had more bone within the area than did unfilled controls (p < 0.05). The percentage of bone within the defect was significantly greater for the 45S5 material than for the 58S or 77S material at 4 and 8 weeks (p < 0.05), yet by 12 weeks equivalent amounts of bone were observed for all materials. By 12 weeks, all grafted defects were equivalent to the normal untreated bone. The resorption of 77S and 58S particles was significantly greater than that of 45S5 particles (p < 0.05). Mechanically, the grafted defects had compressive stiffness equivalent to that of normal bone at 4 and 8 weeks. At 12 weeks, 45S5-grafted defects had significantly greater stiffness (p < 0.05). At 8 and 12 weeks, all grafted defects had significantly greater stiffness than unfilled control defects (p < 0.05). In general, the 45S5-filled defects exhibited greater early bone ingrowth than did those filled with 58S or 77S. However, by 12 weeks, the bone ingrowth in each defect was equivalent to each other and to normal bone. The 58S and 77S materials resorbed faster than the 45S5 materials. Mechanically, the compressive characteristics of all grafted defects were equivalent or greater than those of normal bone at all time points.

This study tested the hypothesis that the rate and the extent of bone formation adjacent to porous, coated Ti-6Al-4V implants are differentially affected by the type of bioactive ceramic coating. Forty-eight rabbits received cylindrical Ti-6Al-4V intramedullary distal femoral implants bilaterally. Implants for the right limbs were coated with 45S5 Bioglass (45S5). Implants used for the left limbs either were coated with tricalcium phosphate/hydroxyapatite (HA) or were left uncoated as controls (CTL). The 45S5-coated implants histologically and biomechanically were compared to HA-coated and CTL implants at 4, 8, 12, and 16 weeks. After 12 and 16 weeks of healing, more bone and thicker trabeculae were measured histomorphometrically within the implant pores for the 45S5-coated implants compared to the HA-coated and CTL implants (p < 0.05). With time the HA-coated and CTL groups exhibited a significant decline in percent of bone and of trabecular thickness (p < 0.05) while the 45S5-coated implants did not. Biomechanical analyses indicated similar shear strengths for all treatment groups. In summary, 45S5-coated implants exhibited greater bone ingrowth compared to HA-coated and CTL implants, and they maintained their mechanical integrity over time.

No abstract

The immunologic requirements for generating long-lived protective CD8 T cell memory remain unclear. Memory CD8 populations generated in the absence of CD4 Th cells reportedly have functional defects, and at least a subset of CD8 T cells transiently express CD40 after activation, suggesting that direct CD4-CD8 T cell interactions through CD40 may influence the magnitude and functional quality of memory CD8 populations. To ascertain the role of CD40 in such direct T cell interactions, we investigated CD8 T cell responses in CD40−/− mice after infection with Listeria monocytogenes, an intracellular bacterium that induces APC activation and thus priming of CD8 T cells independently of CD4 Th cell help through CD40. In this study we show that memory CD8 T cells generated in CD40-deficient mice show in vivo cytotoxicity and cytokine production equivalent to CD8 memory T cells from wild-type mice. Upon secondary Listeria infection, CD40−/− memory CD8 T cells expand to greater numbers than seen in wild-type mice. These results indicate that CD40 ligation on CD8 T cells, although reportedly a part of CD8 T cell memory development in an H-Y-directed response, is not needed for the development of functional memory CD8 T cell populations after Listeria infection.