Separation of copurifying GroEL from glutathione-S-transferase fusion proteins.

The purification of overexpressed fusion proteins using bacterial expression systems is a useful tool for the study of many proteins. One problem that can occur is the formation of stable interactions between the expressed fusion protein and certain endogenous bacterial proteins, such as the molecular chaperone GroEL. Such interactions may result in the copurification of contaminating bacterial proteins. Here we describe an efficient and inexpensive method for the removal of contaminating GroEL from a bacterially expressed GST fusion protein. In this method, denatured bacterial proteins are added to the bacterial lysates prior to the addition of glutathione Sepharose resin. The denatured proteins compete for GroEL binding, thereby releasing the GroEL contaminants from the expressed fusion protein.

RGS4 and RGS2 bind coatomer and inhibit COPI association with Golgi membranes and intracellular transport.

COPI, a protein complex consisting of coatomer and the small GTPase ARF1, is an integral component of some intracellular transport carriers. The association of COPI with secretory membranes has been implicated in the maintenance of Golgi integrity and the normal functioning of intracellular transport in eukaryotes. The regulator of G protein signaling, RGS4, interacted with the COPI subunit beta'-COP in a yeast two-hybrid screen. Both recombinant RGS4 and RGS2 bound purified recombinant beta'-COP in vitro. Endogenous cytosolic RGS4 from NG108 cells and RGS2 from HEK293T cells cofractionated with the COPI complex by gel filtration. Binding of beta'-COP to RGS4 occurred through two dilysine motifs in RGS4, similar to those contained in some aminoglycoside antibiotics that are known to bind coatomer. RGS4 inhibited COPI binding to Golgi membranes independently of its GTPase-accelerating activity on G(ialpha). In RGS4-transfected LLC-PK1 cells, the amount of COPI in the Golgi region was considerably reduced compared with that in wild-type cells, but there was no detectable difference in the amount of either Golgi-associated ARF1 or the integral Golgi membrane protein giantin, indicating that Golgi integrity was preserved. In addition, RGS4 expression inhibited trafficking of aquaporin 1 to the plasma membrane in LLC-PK1 cells and impaired secretion of placental alkaline phosphatase from HEK293T cells. The inhibitory effect of RGS4 in these assays was independent of GTPase-accelerating activity but correlated with its ability to bind COPI. Thus, these data support the hypothesis that these RGS proteins sequester coatomer in the cytoplasm and inhibit its recruitment onto Golgi membranes, which may in turn modulate Golgi-plasma membrane or intra-Golgi transport.

A 102 kDa subunit of a Golgi-associated particle has homology to beta subunits of trimeric G proteins.

We have identified a 102 kDa protein, p102, which is found on the cytoplasmic face of Golgi membranes, exocytic transport vesicles and in the cytosol. A monoclonal antibody that cross-reacts with p102 is able to immunoprecipitate a 500-600 kDa protein complex containing p102 and additional subunits. The composition of this p102-containing protein complex resembles that of the Golgi coatomer complex, which constitutes the coat of non-clathrin coated vesicles. One of the subunits of the p102 complex reacts with a monoclonal antibody that detects beta-COP, a subunit of the Golgi coatomer complex. Like beta-COP, p102 exists in a brefeldin A-sensitive association with Golgi membranes. The sequence of p102 contains an N-terminal domain composed of six repeats which are similar to those found in the beta subunit of trimeric G proteins and other regulatory proteins. We suggest that p102 may be involved in regulating membrane traffic in the constitutive exocytic pathway.

A cosmid clone map derived from a small region of human chromosome 11.

We have used cosmid "fingerprinting" to construct an overlapping DNA clone "map" of the human DNA in a mouse/human hybrid cell line, E65-9, that contains about 4 x 10(6) bp, including the H-Ras gene, as its human component. We have additionally used 32P-labeled RNA probes to establish linkage of particular sets of clones, and the final map comprises about 300,000 bp and is contained in three nonoverlapping segments. The reasons for failure to close the gaps by direct probing are discussed. We have developed techniques to search for rare cutting restriction enzyme cleavage sites in large numbers of cloned DNAs and have positioned sites for EagI and BssHII on our clone map. The methods we used are capable of considerable scale-up and are currently being applied to the short arm of human chromosome 11.

Cloning and regulation of Erwinia herbicola pigment genes.

The genes coding for yellow pigment production in Erwinia herbicola Eho10 (ATCC 39368) were cloned and localized to a 12.4-kilobase (kb) chromosomal fragment. A 2.3-kb AvaI deletion in the cloned fragment resulted in the production of a pink-yellow pigment, a possible precursor of the yellow pigment. Production of yellow pigment in both E. herbicola Eho10 and pigmented Escherichia coli clones was inhibited by glucose. When the pigment genes were transformed into a cya (adenylate cyclase) E. coli mutant, no expression was observed unless exogenous cyclic AMP was provided, which suggests that cyclic AMP is involved in the regulation of pigment gene expression. In E. coli minicells, the 12.4-kb fragment specified the synthesis of at least seven polypeptides. The 2.3-kb AvaI deletion resulted in the loss of a 37K polypeptide and the appearance of a polypeptide of 40 kilodaltons (40K polypeptide). The synthesis of the 37K polypeptide, which appears to be required for yellow pigment production, was not repressed by the presence of glucose in the culture medium, as was the synthesis of other polypeptides specified by the 12.4-kb fragment, suggesting that there are at least two types of gene regulation involved in yellow pigment synthesis. DNA hybridization studies indicated that different yellow pigment genes exist among different E. herbicola strains. None of six pigmented plant pathogenic bacteria examined, Agrobacterium tumefaciens C58, Cornyebacterium flaccumfaciens 1D2, Erwinia rubrifaciens 6D364, Pseudomonas syringae ATCC 19310, Xanthomonas campestris 25D11, and "Xanthomonas oryzae" 17D54, exhibited homology with the cloned pigment genes.