The objective of this dissertation was to study B cell development and the mechanism of immunoglobulin gene diversification in cattle. The bovine immune system differs from that of the mouse and human in several respects. First, the bovine have an unique lymphoid tissue, the ileal Peyer's patch (IPP). Secondly, at least 90% of the immunoglobulin molecules produced by bovine are of the lambda ($\lambda$) isotype. Northern blot and FACS analysis were performed in order to document the distribution of B cells found in calves of various ages. B cells were located in the IPP, jejunal Peyer's patch, spleen and peripheral blood. It was demonstrated that $>$90% of the IPP lymphoid cells were IgM$\sp+$ B cells. In contrast to the IPP, bone marrow of young calves did not appear to be a site for B cell development. To determine the mechanism by which bovine create a diverse repertoire of antibody molecules, cDNA clones encoding $\lambda$ light chains from IPP B cells were sequenced. A comparison of these sequences indicated that all of the V$\sb{\lambda}$ regions were very homologous to one another, and there appeared to be three different V$\sb{\lambda}$ regions expressed. This correlated with Southern blot analyses which showed a limited number of V$\sb{\lambda}$ gene segments in the germline. The bovine $\kappa$ light chain gene was also investigated. Based on mRNA expression and cDNA sequence analysis, it was found that cattle appear to possess a functional $\kappa$ locus. Preliminary studies suggest that the bovine $\kappa$ locus may have a limited number of V$\sb{\kappa}$ gene segments in the germline DNA. These data suggest the following: (1) IPP tissue, not bone marrow, is the major site of B cell development in a young calf. (2) Bovine B cells do not depend upon a large number of V$\sb{\lambda}$ gene segments to produce their immunoglobulin repertoire. Thus, diversity must be generated by another mechanism, perhaps somatic mutation or gene conversion. (3) Cattle are able to transcribe the $\kappa$ light chain gene, although the number of V$\sb{\kappa}$ gene segments may be limited in number.

This project dissects the transcellular membrane protein polarity of bovine aortic endothelial cell monolayers in vitro with respect to the roles that the submembranous cytoskeleton and tight junctional complexes may play in controlling this phenomenon. Endothelial cells, isolated from micro- and macrovascular beds, were examined morphologically and biochemically for their ability to retain in vivo characteristics in culture. Macrovascular endothelial cells from bovine aorta resisted the greatest detectable variation in vitro when compared with microvascular endothelium from brain or adrenal medulla and were therefore selected as the endothelial system on which to study transcellular membrane protein polarity. Apical and basolateral membrane domains of confluent bovine aortic endothelial cell monolayers were isolated using the cationic colloidal silica technique and their proteins resolved by two-dimensional gel electrophoresis. Construction of an isoelectric point-molecular weight database identified domain specific membrane proteins. Domain specific membrane proteins were assessed for cytoskeletal interaction by determining if they co-isolated in a Triton X-100 detergent-resistant cytoskeletal fraction. The maintenance of polarized membrane protein segregation by tight junctional complexes was determined by comparing apical and basolateral protein patterns of confluent monolayers to patterns generated by subconfluent monolayers. Membrane proteins found to be asymmetrically distributed in the absence of tight junctions were assessed for cytoskeletal interaction by their inability to be extracted by Triton X-100 from subconfluent monolayers. Cross referencing of data obtained from the above fractionation procedures catagorizes a subset of plasma membrane proteins with respect to their apical or basolateral position, their interaction with the cytoskeleton and their ability to remain in the proper membrane location in the absence of tight junctional complexes. Many polarized proteins resisted randomization in subconfluent monolayers, regardless cytoskeletal association. Other membrane proteins became equally distributed in the subconfluent state, in spite of cytoskeletal interaction. The data suggests that membrane proteins obey different mechanisms in establishing and maintaining transcellular membrane protein polarity in endothelial cells.