A 2

To gain an understanding of the molecular events underlying the evolution of C, photosynthesis, we have undertaken a detailed study of the NADP-malic enzyme gene family in C, and C, species of Haveria. Three genomic clones from the C, species Haveria bidentis were characterized and found to encode two highly similar chloroplastic forms of NADP-malic enzyme, termed ME1 and ME2. Genomic Southern blotting with gene-specific probes showed that both Me7 and Me2 are found in Haveria trinervia (C,) and fiaveria pringlei (C,) as well as in F. bidentis. Northern blots demonstrated that M e l expression in leaves parallels the degree of C, photosynthesis in seven Haveria species. Furthermore, whereas M e 2 was expressed at a low level in both roots and leaves of f . bidentis, M e l expression was seen only in leaves and was light-regulated. We discuss these results in the context of the evolution of C, photosynthesis in Flaveria.The C, photosynthetic pathway is a recently evolved modification of the C, pathway in which plants spatially separate the initial fixation and subsequent conversion of CO, to 3-phosphoglycerate and other carbohydrates. Through the use of this system, C, plants effectively concentrate CO, in the vicinity of Rubisco (located in bundle-sheath cells) and, thus, avoid the wasteful process of photorespiration. The efficient operation of C, photosynthesis requires the strict compartmentation of a suite of enzymes in either mesophyll or bundle-sheath cells. This intercellular compartmentation is a distinguishing feature of C, plants: a11 of the key enzymes utilized in C, photosynthesis are present in C, plants; however, their expression patterns are distinct from those seen in C, plants. It is generally accepted that C, plants evolved from C, ancestors and that the specific enzymes utilized in C, photosynthesis derived from corresponding enzymes present in the C, ancestors (Cockburn, 1983). In the evolution of C, photosynthesis, the expression programs of these ancestral proteins were Present address:

Neff et al. (1982, J. Cell Biol., 95:654-666) have described a monoclonal antibody, CSAT, directed against a cell surface antigen that participates in the adhesion of skeletal muscle to extracellular matrices. We used the same antibody to compare and parse the determinants of adhesion and morphology on myogenic and fibrogenic cells. We report here that the antigen is present on skeletal and cardiac muscle and on tendon, skeletal, dermal, and cardiac fibroblasts; however, its contribution to their morphology and adhesion is different. The antibody produces large alterations in the morphology and adhesion of skeletal myoblasts and tendon fibroblasts; in contrast, its effects on the cardiac fibroblasts are not readily detected. The effects of CSAT on the other cell types, i.e., dermal and skeletal fibroblasts, cardiac muscle, 5-bromodeoxyuridine-treated skeletal muscle, lie between these extremes. The effects of CSAT on the skeletal myoblasts depends on the calcium concentration in the growth medium and on the culture age. We interpret these differential responses to CSAT as revealing differences in the adhesion of the various cells to extracel[ular matrices. This interpretation is supported by parallel studies using quantitative assays of cell-matrix adhesion. The likely origin of these adhesive differences is the progressive display of different kinds of adhesion-related molecules and their organizational complexes on increasingly adhesive cells. The antigen to which CSAT is directed is present on all of the above cells and thus appears to be a lowest common denominator of their adhesion to extracellular matrices.Adhesive differences among embryonic cells are thought to play a prominent role in directing morphogenesis (2, 32). Numerous examples of such differential adhesions have been described and evidence has been presented that relates them to morphogenic phenomena. The histotypic sorting out of embryonic cells is rationalized as reflecting the graded, differential adhesive interactions of the constituent cells (24,26,30,31). In the nervous system, the high specificity of retinotectal interactions appears to reflect the adhesive gradients generated by the retinal and tectal cells (12,24). The migration of neuronal processes to their peripheral targets is also highly specific and is hypothesized to reflect the differential affinities of the processes for extracellular substrates (7,23).Recently membrane proteins have been identified and isolated using adhesion-perturbing monoclonal antibodies whose antigens localize in the region of cell-substratum adhesions (13, 27, 28). One of these antibodies, called CSAT, is directed against an antigen that participates in the adhesion of embryonic skeletal muscle to extracellular matrices (17,18,27). T The active determinant appears to be part of an integral membrane protein complex located in the vicinity of transmembrane assemblies involved in adhesion. The identification of such adhesion-related molecules allows an investigation of their role in regulating the...

The efficient functioning of C4 photosynthesis requires the strict compartmentation of a suite of enzymes in either mesophyll or bundle sheath cells. To determine the mechanism controlling bundle sheath cell-specific expression of the NADP-malic enzyme, we made a set of chimeric constructs using the 5[prime] and 3[prime] regions of the Flaveria bidentis Me1 gene fused to the [beta]-glucuronidase gusA reporter gene. The pattern of GUS activity in stably transformed F. bidentis plants was analyzed by histochemical and cell separation techniques. We conclude that the 5[prime] region of Me1 determines bundle sheath specificity, whereas the 3[prime] region contains an apparent enhancer-like element that confers high-level expression in leaves. The interaction of 5[prime] and 3[prime] sequences was dependent on factors that are present in the C4 plant but not found in tobacco.