Quantitative analysis of chaperone network throughput in budding yeast

Abstract

The network of molecular chaperones mediates the folding and translocation of the many proteins encoded in the genome of eukaryotic organisms, as well as a response to stress. It has been particularly well characterised in the budding yeast, Saccharomyces cerevisiae, where 63 known chaperones have been annotated and recent affinity purification and MS/MS experiments have helped characterise the attendant network of chaperone targets to a high degree. In this study, we apply our QconCAT methodology to directly quantify the set of yeast chaperones in absolute terms (copies per cell) via SRM MS. Firstly, we compare these to existing quantitative estimates of these yeast proteins, highlighting differences between approaches. Secondly, we cast the results into the context of the chaperone target network and show a distinct relationship between abundance of individual chaperones and their targets. This allows us to characterise the ‘throughput’ of protein molecules passing through individual chaperones and their groups on a proteome-wide scale in an unstressed model eukaryote for the first time. The results demonstrate specialisations of the chaperone classes, which display different overall workloads, efficiencies and preference for the sub-cellular localisation of their targets. The novel integration of the interactome data with quantification supports re-estimates of the level of protein throughout going through molecular chaperones. Additionally, although chaperones target fewer than 40% of annotated proteins we show that they mediate the folding of the majority of protein molecules (∼62% of the total protein flux in the cell), highlighting their importance.

Quantitative proteomics and network analysis of SSA1 and SSB1 deletion mutants reveals robustness of chaperone HSP70 network in Saccharomyces cerevisiae

“…BC has also been applied to topological analysis of mammalian transcription networks where it was recognised to be the most representative parameter with regards to node biological significance 50. Notably, a clear positive correlation between the chaperone workload estimated previously 8, 32 and its BC exists (Spearman coefficient for top 15 chaperones = 0.89). Chaperones with higher BC index experience higher workloads, and could be regarded as functionally more important.…”

“…We modelled the workload placed on individual nodes (chaperones) in the network, using two theoretical frameworks; BC and chaperone synthetic flux (referred to here as workload). The latter was defined in our previous work 8, 32, estimating the number of molecules per unit time passing through each chaperone en‐route to the native state. This estimate of chaperone workload predicted that SSB1 and SSA1 handle the biggest total substrate volume and total protein flux.…”

“…The folding flux ( F c ), of a chaperone ( c ) in terms of molecules per min per cell, is calculated over all n substrates as follows: $F_c=\sum _{i=1}^n{k}_{syn,i}$where k syn is calculated from known protein concentrations and protein degradation rates, $k_{syn,i}=cp{c}_i\times k_{deg,i}$assuming steady state where protein synthesis and degradation rates are equal ($\frac{dCP{C}_i}{dt}=0$). To reduce bias from potential false positives we performed these calculations with the high‐quality subset of 3649 interactions defined previously 32 rather than the full network. The substrate flux values where then divided equally pro rata among each chaperone.…”

“…This relatively modest overlap between paralogue pairs suggests that simple sequence identity alone is insufficient to explain the phenotypes observed. We note this modest overlap is further reduced when the network is filtered and limited to 3649 chaperone‐substrate pairs present in multiple datasets 32, shown in Fig. 2D–F, a high quality dataset designed to reduce the level of false positives.…”

“…Substantial numbers of substrates are unique to the individual chaperones. Panels (D)–(F) show the same data, restricted to a filtered, high‐quality subset of the chaperone network 32, constituted by 3649 chaperone‐substrate interactions.…”