The chaperonin GroEL is a double-ring structure with a central cavity in each ring that provides an environment for the efficient folding of proteins when capped by the co-chaperone GroES in the presence of adenine nucleotides. Productive folding of the substrate rhodanese has been observed in cis ternary complexes, where GroES and polypeptide are bound to the same ring, formed with either ATP, ADP or non-hydrolysable ATP analogues, suggesting that the specific requirement for ATP is confined to an action in the trans ring that evicts GroES and polypeptide from the cis side. We show here, however, that for the folding of malate dehydrogenase and Rubisco there is also an absolute requirement for ATP in the cis ring, as ADP and AMP-PNP are unable to promote folding. We investigated the specific roles of binding and hydrolysis of ATP in the cis and trans rings using mutant forms of GroEL that bind ATP but are defective in its hydrolysis. Binding of ATP and GroES in cis initiated productive folding inside a highly stable GroEL-ATP-GroES complex. To discharge GroES and polypeptide, ATP hydrolysis in the cis ring was required to form a GroEL-ADP-GroES complex with decreased stability, priming the cis complex for release by ATP binding (without hydrolysis) in the trans ring. These observations offer an explanation of why GroEL functions as a double-ring complex.

The synthesis, proof of structure, and the absorption and fluorescence properties of two new unsymmetrical cyanine dyes, thiazole orange dimer (TOTO; 1,1'-(4,4,7,7-tetramethyl-4,7- diazaundecamethylene)-bis-4-[3-methyl-2,3-dihydro-(benzo-1,3-thiaz ole)-2- methylidene]-quinolinium tetraiodide) and oxazole yellow dimer (YOYO; an analogue of TOTO with a benzo-1,3-oxazole in place of the benzo-1,3-thiazole) are reported. TOTO and YOYO are virtually non-fluorescent in solution, but form highly fluorescent complexes with double-stranded DNA (dsDNA), up to a maximum dye to DNA bp ratio of 1:4, with greater than 1000-fold fluorescence enhancement. The dsDNA-TOTO (lambda max 513 nm; lambda maxF 532 nm) and dsDNA-YOYO (lambda max 489 nm; lambda maxF 509 nm) complexes are completely stable to electrophoresis on agarose and acrylamide gels. Mixtures of restriction fragments pre-labeled with ethidium dimer (EthD; lambda maxF 616 nm) and those pre-labeled with either TOTO or YOYO were separated by electrophoresis. Laser excitation at 488 nm and simultaneous confocal fluorescence detection at 620-750 nm (dsDNA-EthD emission) and 500-565 nm (dsDNA-TOTO or dsDNA-YOYO emission) allowed sensitive detection, quantitation, and accurate sizing of restriction fragments ranging from 600 to 24,000 bp. The limit of detection of dsDNA-TOTO and YOYO complexes with a laser-excited confocal fluorescence gel scanner for a band 5-mm wide on a 1-mm thick agarose gel was 4 picograms, about 500-fold lower than attainable by conventional staining with ethidium bromide.

Recent studies of GroE-mediated protein folding indicate that substrate proteins are productively released from a cis ternary complex in which the nonnative substrate is sequestered within the GroEL channel underneath GroES. Here, we examine whether protein folding can occur in this space. Stopped-flow fluorescence anisotropy of a pyrene-rhodanese-GroEl complex indicates that addition of GroES and ATP (but not ADP) leads to a rapid change in substrate flexibility at GroEL. Strikingly, when GroES release is blocked by the use of either a nonhydrolyzable ATP analog or a single-ring GroEL mutant, substrates complete folding while remaining associated with chaperonin. We conclude that the cis ternary complex, in the presence of ATP, is the active state intermediate in the GroE-mediated folding reaction: folding is initiated in this state and for some substrates may be completed prior to the timed release of GroES triggered by ATP hydrolysis.