CUET-ting edge research

Keywords:

• lysosome
• macroautophagy
• phagophore
• receptor
• stress

Ubiquitinated aggregates of aggregation-prone proteins are a common feature in many neurodegenerative diseases. In mammalian cells receptor proteins that deliver these aggregates to the lysosome for degradation via selective autophagy are well established. The common feature of these receptors, namely SQSTM1/p62 and NBR1, is the ability to bind ubiquitin and LC3, thus connecting the ubiquitin pathway and autophagy. However, it remained unknown if a counterpart of this pathway exists in yeast. A recent study identified the CUET-proteins as a new class of conserved ubiquitin-Atg8 receptors.

In the paper by Lu et al.¹ (and as highlighted in a punctum in this issue of the journal) the authors specifically sought to address the question of whether yeast utilize a ubiquitin-dependent cargo recognition system² for selective autophagy. After initially establishing the autophagy-dependent degradation of (poly)ubiquitinated proteins upon autophagy induction in S. cerevisiae, the authors immunoprecipitated HA-tagged Atg8 from starved cells and thereby coprecipitated ubiquitin conjugates. Aiming to determine potential ubiquitin-Atg8 receptors, the authors used a mass spectrometry approach and thus identified Cue5 and the E3 ubiquitin ligase Rsp5 among the coprecipitating proteins. Cue5, a previously uncharacterized protein, drew their particular attention as it potentially contains an Atg8-family interacting motif (AIM) and a ubiquitin-binding (CUE) domain, thereby fullfilling the characteristics of a ubiquitin-Atg8 receptor. They subsequently verified that Cue5 directly binds Atg8 via its C-terminal AIM, with a possible preference for Atg8–PE, while binding ubiquitin via its CUE-domain. The deletion of CUE5 as well as the expression of Cue5 mutants with a nonfunctional AIM- or CUE-domain result in the accumulation of ubiquitinated proteins, leading to the conclusion that Cue5 functions as a ubiquitin-Atg8 receptor similar to SQSTM1 in mammals.

Using a SILAC-based mass spectrometry approach combined with a GFP-processing assay, the authors subsequently identified 24 proteins as bona fide substrates of Cue5-dependent autophagy. Despite the fact that Cue5-substrates did not show any similarities, a subset of these proteins displayed a tendency to aggregate even under normal (i.e., 30°C) growth conditions. The protein aggregates further accumulated under stress-inducing conditions (37°C) and their degradation depended on Cue5-mediated autophagy. Next, the authors provided evidence that the ubiquitination of Cue5 substrates through Rsp5- and Cue5-dependent autophagy are tightly coupled: Rsp5, Cue5, and Atg8 form a complex, and a population of Cue5 is ubiquitinated by Rsp5 and degraded in the vacuole.

In yeast, the degradation of HTT-96Q with an expanded polyQ-repeat, a well-studied aggregation-prone and cytotoxic variant of the HTT (huntingtin) protein,³ depends in part on the same Cue5-Rsp5-dependent mechanism. Encouraged by the possible medical implications of this result, Lu et al. expanded the study to human cells and thus identified TOLLIP as the functional human ortholog of Cue5. Even though TOLLIP differs from Cue5 in its overall domain organization, it also binds ubiquitin through a CUE-domain and LC3 cooperatively via 2 C-terminal LC3-interacting regions (LIRs, the mammalian equivalent of the AIM). TOLLIP colocalizes with ubiquitin and LC3-positive autophagosomes, and cofractionates with LC3 together with protein aggregates. In yeast cells where CUE5 is deleted, TOLLIP expression suppresses the partial sensitivity against HTT-96Q and restores autophagic degradation of ubiquitin conjugates. In human cells, TOLLIP plays a role in the ubiquitin-dependent degradation of overexpressed HTT-103Q. Notably, the heterologous expression of Cue5 in human cells is able to fulfill this function, although less efficiently.

The authors convincingly establish the Cue5/TOLLIP-dependent autophagic degradation of ubiquitinated substrates as an additional pathway separate from SQSTM1-dependent clearance. This newly discovered pathway is especially potent in the clearance of highly aggregation-prone polyQ-containing proteins and might thus be a promising therapeutic target.

Funding

This work was supported by NIH grant GM053396 to DJK.