Models for regulation of the eukaryotic heat shock response typically invoke a negative feedback loop consisting of the transcriptional activator Hsf1 and a molecular chaperone. Previously we identified Hsp70 as the chaperone responsible for Hsf1 repression and constructed a mathematical model that recapitulated the yeast heat shock response (Zheng et al., 2016). The model was based on two assumptions: dissociation of Hsp70 activates Hsf1, and transcriptional induction of Hsp70 deactivates Hsf1. Here we validate these assumptions. First, we severed the feedback loop by uncoupling Hsp70 expression from Hsf1 regulation. As predicted by the model, Hsf1 was unable to efficiently deactivate in the absence of Hsp70 transcriptional induction. Next, we mapped a discrete Hsp70 binding site on Hsf1 to a C-terminal segment known as conserved element 2 (CE2). In vitro, CE2 binds to Hsp70 with low affinity (9 µM), in agreement with model requirements. In cells, removal of CE2 resulted in increased basal Hsf1 activity and delayed deactivation during heat shock, while tandem repeats of CE2 sped up Hsf1 deactivation. Finally, we uncovered a role for the N-terminal domain of Hsf1 in negatively regulating DNA binding. These results reveal the quantitative control mechanisms underlying the heat shock response.

20Models for regulation of the eukaryotic heat shock response typically invoke a negative 21 feedback loop consisting of the transcriptional activator Hsf1 and a molecular chaperone 22

Reaction of [(Ph2C2S2)2M] (M = Ni 2+ , Pd 2+ , Pt 2+ ) with 2 eq of RN≡C (R = Me (a), Bn (b), Cy (c), t Bu (d), 1-Ad (e), Ph (f)) yields [(Ph2C2S2)M(C≡NR)2] (M= Ni 2+ , 4a-4f; M = Pd 2+ , 5a-5f; M = Pt 2+ , 6a-6f), which are air-stable and amenable to chromatographic purification. All members have been characterized crystallographically. Structurally, progressively greater planarity tends to be manifest as M varies from Ni to Pt, and a modest decrease in the C≡N bond length of coordinated C≡NR appears in moving from Ni toward Pt. Vibrational spectroscopy (CH2Cl2 solution) reveals νC≡N frequencies for [(Ph2C2S2)M(C≡NR)2] that are substantially higher than for free C≡NR and increase as M ranges from Ni to Pt. This trend is interpreted as arising from increasing positive charge at M that stabilizes the linear, charge-separated resonance form of the ligand over the bent form with lowered C-N bond order. UV-vis spectra reveal lowest energy transitions that are assigned as HOMO (dithiolene π) → LUMO (M-L σ*) excitations. Oneelectron oxidations of [(Ph2C2S2)M(C≡NR)2] are observed at ~+0.5 V due to Ph2C2S2 2-→ Ph2C2S -S • + e -. Chemical oxidation of [(Ph2C2S2)Pt(C≡N t Bu)2] with [(Br-p-C6H4)3N][SbCl6] yields [(Ph2C2S -S • )Pt(C≡N t Bu)2] 1+ , identified spectroscopically, but in the crystalline state [[(Ph2C2S -S • )Pt(C≡N t Bu)2]2] 2+ prevails, which forms via axial Pt•••S interactions and pyramidalization at metal. Complete substitution of MeNC from [(Ph2C2S2)Ni(C≡NMe)2] by 2,6-Me2py under forcing conditions yields [(2,6-Me2py)Ni(μ2-η 1 ,η 1 -S',η 1 -S"-C2Ph2)]2 (8), which features a folded Ni2S2 core. In most cases, isonitrile substitution from [(Ph2C2S2)M(C≡NMe)2] with monodentate ligands (L = phosphine, CN -, carbene) typically leads to [(Ph2C2S2)M(L)(C≡NMe)] n (n = 0 or 1 -), wherein νC≡N varies according to the relative σ donating power of L (9 -21). Use of 1,3-bis(2,6diisopropylphenyl)imidazol-2-ylidene (IPr) provides [(Ph2C2S2)M(IPr)(C≡NMe)] for M = Ni (18)or Pd ( 19), but for Pt , attack by IPr at the isonitrile carbon occurs to yield the unusual η 1 ,κCketenimine complex [(Ph2C2S2)Pt(C(NMe)(IPr))(C≡NMe)] (20).

No abstract