Search

This thesis investigates the causes of front back confusion in two systems that make use of digital signal processing for the production of virtual images. Front back confusion is well in known psychoacoustics and describes the phenomenon occurring when a source placed behind a listener is perceived to be in front, or vice versa. The two virtual imaging systems studied are designed to use respectively two and four channels to give listeners the impression that sources of sound exist at locations that are other than those used for reproduction. The four channel system is shown to be robust with regard to the production of convincing images at any location in the horizontal plane around the listener. The two channel system is able to achieve good images in front of the listener, but not behind. Both reproduction systems are examined with reference to the signals received at the ears of a listener. These signals are compared to those produced by real sources placed at different locations around the listener. This investigation is carried out by using a simple model of the listeners head based on the sound field scattered by a rigid sphere. These models indicate that the difference between the two systems could be explained in terms of the interaural time delay ( ITD ). The ITD was measured by calculating the interaural cross-correlation function ( IACC ). Head movement has been shown by previous work to play a very important part in resolving front back confusion. It is concluded that the rate of change of the IACC with head rotation is an important possible means of resolving these confusions, and that the four channel system is capable of closely replicating this measure. The changes in the IACC resulting from head movements were also modelled and are in accord with the results of the subjective experiments carried out to test the hypothesis from the modelling work. It is believed that this is the method used by the hearing system to interpret the localisation cues contained within the ITD's.

Rat bone marrow-derived mast cells (BMMCs), cultured in the presence of a T lymphocyte conditioned medium, are analogous to MMCs as defined by the granule content of the soluble chymase, rat mast cell protease-II (RMCP-II); by the granule proteoglycan chondroitin sulphate; and by their secretory characteristics. To investigate the secretory response of BMMCs to IgE-dependent stimulation, a sensitive, specific and repeatable enzyme-linked immunospot (ELISPOT) assay was developed to detect the release of RMCP-II from individual cells. Within populations of BMMCs, only 6-24% of the cells responded to challenge with either anti-IgE or specific antigen, leaving a large residual refractory population. Pre-incubation of mature BMMCs with the multi-functional cytokine, stem cell factor (SCF), significantly increased (≈ 2-fold) the proportion of cells responding to IgE-dependent stimulation without directly causing mediator release. Furthermore, SCF enhanced the total percentage release of RMCP-II and β-hexosaminidase from populations of mature BMMCs in association with an increased proportion of cells secreting RMCP-II as detected by ELISPOT. These results suggest that SCF augments IgE-dependent secretion from rat BMMCs primarily by activating previously unresponsive cells. To further characterise the functional phenotype of rat BMMCs, the electrophysiological properties of the cells were investigated using the whole-cell configuration of the patch-clamp technique. Rat BMMCs had a mean membrane potential of -25.9 mV and a mean whole-cell capacitance of 4.8pF. With the amphotericin B perforated-patch technique, both inwardly rectifying (IR) and outwardly rectifying (OR) currents were observed in rat BMMCs. The reversal potential and conductance of the IR current depended on the extracellular K⁺ concentration, indicating that the channel was K⁺ selective. The OR current was reversibly decreased both by lowering the extracellular Cl^- concentration and by the Cl^- channel blocker DIDS, indicating a Cl^- conductance.

This thesis encompasses two separate aspects of the biology of osteoclasts. Firstly, proteins were purified from bone matrix that modulate osteoclast formation and function. The effects of two of the most abundant growth factors insulin-like growth factors-l (IGF-I) and -II (IGF-II) on osteoclast differentiation and function were also studied. The aim of the second area of research was to determine the roles of individual members of the cysteine proteinase (CP) and matrix metalloproteinase (MMP) family on bone resorption using selective enzyme inhibitors. Bone matrix is a rich source of polypeptide growth factors and evidence suggests that osteoclasts do not mature functionally until they reach bone. Since bone resorption occurs at discrete sites throughout the skeleton, it is likely that an osteoclast activating factor(s) could be sequestrated in bone matrix where it may regulate osteoclast formation from haemopoietic precursors. Matrix proteins were extracted from bovine cortical bone with EDTA/TrisHGI under non-dissociative conditions at neutral pH. Four distinct bone resorptive proteins of 14, 25, 29, and 40 kDa were purified and partially characterized using an in vitro neonatal calvarial assay. The 14 and 25 kDa proteins possessed insulin-like growth factor (IGF) and transforming growth factor-6 activity respectively, whilst the 40 kDa protein enhanced the formation of osteoclast-like cells in a mouse bone marrow àssay. The 29 kDa protein stimulated bone resorption in an isolated osteoclast assay, suggesting a direct action on osteoclasts. This is the first report of (i) a protein isolated from bone matrix that stimulates osteoclast differentiation and (ii) a protein that directly stimulates osteoclast function. IGF-I and -II were found to enhance both osteoclast differentiation and function but they were less effective than 1,25 Dihydroxyvitamin D3. The stimulatory effects of IGFs on osteoclast activity were mediated via an interaction with the type I IGF receptor present on osteoblastic cells. Osteoclasts solubilize mineralized bone by means of acid pH generation and proteinase action, notably enzymes of the métallo and cysteine classes. To determine the roles and sites of action of the individual proteinases in bone resorption both natural and synthetic enzyme inhibitors were utilized. Four CP inactivators were tested: Ep475, a compound with low membrane permeabilty that selectively inhibits cathepsins B,L,S and calpain; Ep453, the membrane permeant prodrug of Ep475; CA074, a compound with low membrane permeability that selectively inactivates cathepsin B; and CA074Me, the membrane permeant prodrug of CA074. Ep475, Ep453 and CA074Me inhibited both calvarial and osteoclast lacunar resorption, whilst CA074 was without effect. These results suggest that cathepsin L and/or S act extracellularly and possibly intracellularly whilst cathepsin B acts intracellularly perhaps through the activation of other proteinases involved in bone collagen degradation. Six MMP inhibitors were used: initially the tissue inhibitors of metalloproteinases (TIMP) -1 and -2 were tested followed by four synthetic inhibitors. These were CT435, a general MMP inhibitor; CT543 and CT1166, compounds that are concentration dependent selective inhibitors of gelatinases-A and -B; and Ro 31- 7467, a concentration-dependent selective inhibitor of collagenase. All the compounds produced a dose-dependent inhibition in calvarial bone resorption. However, CT1166 and CT543 only produced a 40% inhibition and Ro 31-7467 a 50% inhibition at concentration selective for the inhibition of gelatinase and collagenase respectively. The compounds also produced a partial reduction in the resorptive activity of interleukin-1 -stimulated osteoclast cultures. Immunofluoresence studies demonstrated that rabbit osteoclasts synthesize gelatinases-A and -B, collagenase, stromelysin and tissue inhibitor of metalloproteinases-1. Human recombinant TIMP-1 and TIMP-2 were as effective at preventing calvarial bone reorption as the synthetic MMP inhibitors. It is suggested that the MMPs are produced by osteoclasts and participate with the CPs in osteoclast lacunar resorption.