These studies have focused on the role of Gi-proteins and their regulators in mitosis and cytokinesis. In model organisms such as Caenorhabditis elegans and Drosophila receptor-independent heterotrimeric G protein function is vital for the orientation of mitotic spindle, generation of microtubule pulling force, aster-induced cytokinesis, and centration of the nucleus-centrosome complex. This new paradigm is now being extended to mammalian cells. We and others have shown that Gi proteins and their regulators such as AGS3, LGN, and RGS14 localize in centrosomes, at the mitotic cell cortex, and at the midbody region. At these sites AGS3, LGN, and RGS14 likely bind Gi alpha proteins and function similar to G beta/gamma subunits. Whether G alpha proteins also pair with G beta/gamma at these sites or can only interact with proteins such as LGN and AGS3 is not known. Indicating that G beta/gamma association may occur, we have shown the dynamic localization of endogenous G beta/gamma subunits to centrosomes/spindle poles, the mitotic cell cortex, the mitotic spindle, the central spindle, and the midbody. Bimolecular fluorescence complementation of a split YFP-tagged G beta/gamma confirmed targeting of these dimmers to these sites. Exogenous expression of G beta/gamma subunits, siRNA-mediated knock-down of G beta1, and expression of a G beta/gamma scavenging protein, GRK2-ct, all caused cell division defects. To prove the phenotype observed is due to GRK2-ct-mediated sequestering of G beta/gamma, a mutation known to abolish the interaction between GRK2 and G beta/gamma has been introduced to the GRK2-ct. Whether the mitotic defects can be rescued by the mutated GRK2-ct or overexpression of G beta/gamma is currently being examined. In addition, simultaneous knock-down of G beta1 and G beta2 using siRNA is being considered. If phenotype is found, rescue of the phenotype by expression of siRNA-insensitive G beta1 and G beta2 constructs will be attempted. We have also have shown a role for a non-GPCR activator of Gi protein termed Ric-8A in human cell division. Ric-8A expression occurs in most human cells and at high levels in lymphocytes. We have evidence that Ric-8A is important for recruiting a signaling complex to the metaphase cell cortex consisting of NuMA, LGN, dynein, p150 glued, and Gi alpha1. Interference with the localization of this complex caused defects in mitotic spindle orientation and normal cell division. In collaboration with Zhen Huang at the University of Wisconsin we are initiating studies to examine mice in which Ric-8A has been conditionally deleted from B or T lymphocytes. The non-conditional disruption of Ric-8A causes embryonic lethality. We have also begun to examine whether Ric-8A is involved in the Gi activation in interphase cells. Not only is Ric-8A, but also the LGN like protein AGS4 is found in lymphocytes. In preliminary experiments antigen receptor mediated activation of B cells results in the phosphorylation of AGS4 and a RIC-8A knock-down in a human T cell line impaired responses to chemokine stimulation. RGS14 and RGS12 contain an RGS domain and a GoLoco motif. Both the RGS domain and the GoLoco motif of RGS14 target members of the Gi subclass. We have co-localized RGS14 with Gi alpha subunits in centrosomes and in the midbody during cytokinesis. To further our studies of Rgs14, mice in which Rgs14 can be conditionally deleted have been developed. Despite a report to the contrary, germline deletion of Rgs14 did not cause embryonic lethality. Studies are in progress to phenotype these mice and to examine the role of RGS14 in immune cell function. C. elegans RGS7 functions in early cell divisions and RGS7 mutants show hyper-asymmetric movementof mitotic spindles. Among the mammalian RGS proteins, RGS3 most closely resembles C. elegans RGS7. We have shown that one isoform of RGS termed PDZ-RGS3 functions to regulate microtubule dynamics and cytokinesis. Two independent mouse lines each with a targeted disruption of Rgs3 have been identified, however, one line is an embryonic lethal while the other is viable. Extensive back-crossing of the two lines onto a C57/Bl6 background has not resolved the differences between the two lines.
Showing the most recent 10 out of 11 publications
