How misfolded forms are recognized and distinguished from the myriad of folding substrates represents a fascinating problem in discrimination. Cells must maintain a balance between overly promiscuous destruction of slow-folding proteins, while preventing escape of toxic forms. For most luminal proteins, this discrimination depends not only on the folding status of a polypeptide, but also on its glycosylation state.

We, and others concurrently, identified the conserved luminal protein Yos9 as the lectin sensor responsible for probing the glycosylation state of misfolded proteins. This allowed us, in a series of papers, to work out the elaborate mechanism used to correctly mark and degrade terminally misfolded proteins. This strategy of marking potential substrates by a sugar modification and decoding the signal in a second recognition event provides a proofreading mechanism for enhancing substrate specificity. We are now recapitulating this modification and recognition cascade in vitro from purified components. I am particularly excited about this work as it represents a terrific opportunity to provide our most in-depth structural and mechanistic understanding to the fascinating question of how the ER tells right from wrong.

Elegant work pioneered by my UCSF/HHMI colleague and collaborator Peter Walter identified Ire1 as a key ER sensor of misfolded proteins. These studies revealed the canonical activation pathway in which unfolded proteins sensed in the ER lumen induce the cytoplasmic nuclease domain of Ire1 to cleave the mRNA encoding XBP-1 (Hac1 in yeast), enabling splicing and production of active transcription factor. This allows for the remodeling of the ER in response to stress. But transcriptional responses are slow. We found that in metazoans, activation of the Ire1 endonuclease independently induces the rapid turnover of many mRNAs encoding membrane and secreted proteins through a pathway we call regulated Ire1-dependent decay (RIDD). This response is well suited to complement other UPR mechanisms by selectively halting production of proteins that challenge the ER and clearing the translocation and folding machinery for the subsequent remodeling process. More recently, we, and independently the Papa group, provided evidence that cells use a multi-tiered mechanism by which different conditions in the ER lead to distinct outputs (transcriptional versus RIDD) from Ire1. This allows for a nuanced response to different stresses and provides the potential for therapeutics that "surgically" intervene with specific Ire1 outputs, preventing proapoptotic effects while keeping protective outputs intact.

Tail-anchored (TA) proteins, defined by the presence of a single C-terminal transmembrane domain (TMD), play critical roles throughout the secretory pathway and in mitochondria. Because of its position near the C terminus, the TMD of TA proteins is occluded by the ribosome until translation is completed. Thus, TA proteins cannot exploit the classic cotranslational SRP/Sec61 translocation mechanism used by most secretory pathway proteins. Starting from our Genetic Interaction Map approach (see below), we identified the Guided Entry of Tail-anchor (GET) proteins as three components of a conserved machinery responsible for the biogenesis of TA proteins. Further genetic interactions maps identified two additional ribosome-associated components (Get4 and Get5). These and other observations suggest a pathway for the biogenesis of TA proteins in which Get4/5 recognize TA proteins as synthesis is completed. The TA is then delivered to a soluble ATPase, Get3, which is subsequently recruited to the ER membrane by a receptor composed of the Get1/2 proteins, thereby allowing proper membrane insertion.

Despite their numerous essential functions, our understanding of how very long-chain fatty acids (VLCFAs) are made was highly limited. From an academic standpoint, the chain length diversity of VLCFAs, which enables their functional diversity, represents a fascinating example of how polymerization of simple building blocks can be used to build chemically complex macromolecules. From a practical standpoint, there are potential therapeutic and commercial benefits to be gained from manipulating VLCFA synthesis. A key obstacle to these endeavors has been the fact that VLCFAs are made by a collection of detergent-labile, multi-enzyme complexes embedded in the ER membrane. Consequently, even though the basic outline of the chemistry underlying VLCFA synthesis was established in the 1960’s, the minimal set of proteins, including the key dehydratase component, needed for VLCFA synthesis was not known.

We identified Yjl097w (Phs1p) as the founding member of a new family of membrane proteins that act as the VLCFA dehydratase. This missing piece allowed us to reconstitute VLCFA synthesis in proteoliposomes containing purified Phs1p and three other membrane proteins. We used this in vitro system to address what is arguably the most important question regarding VLCFA synthesis: how does the biosynthetic machinery instruct the precise number of two-carbon additions that yields a defined length product. Our studies revealed a fascinating mechanism, akin to a caliper, in which the chain length is determined by the distance between the cytosolic active site that mediates two-carbon addition and a lysine near the luminal end of a transmembrane helix. By stepping this lysine residue along one face of the helix toward the cytosol, we engineered novel synthases with correspondingly shorter VLCFA outputs.

Although recent studies have shed light on the critical roles of sphingolipids both as structural components of membranes and critical signaling molecules, how cells sense and regulate the levels of sphingolipids has remained a mystery. We found that members of the highly conserved but previously uncharacterized Orm protein family inhibit the first and rate-limiting step of de novo sphingolipid production. Moreover, the activity of Orm proteins is dynamically regulated by a key feedback loop: when sphingolipid synthesis is disrupted, Orm proteins are inactivated by phosphorylation, thus providing cells with an elegant mechanism to couple sphingolipid production to metabolic demand.

Our findings are of particular interest in light of the recent identification of a major genetic risk factor for asthma near the locus of the human ORM gene homolog ORMDL3. Polymorphisms which alter ORMDL3 expression account for up to ~20% of childhood asthma cases and have also been implicated in other inflammation-based diseases such as Type I diabetes. Thus, our finding that alterations in ORM gene activity profoundly affect sphingolipid metabolism raises the testable hypothesis that misregulation of sphingolipids contributes directly to the pathogenesis of asthma. We are now extending our initial studies to define the mechanism by which Ormdl proteins are regulated in mammalian cells. Recent progress in identifying upstream components of the yeast phosphorylation feedback pathway will guide these efforts and may also enable us to define a molecular sensor of sphingolipid levels. Thus, our work on Orm family proteins promises to inform a fundamental question in cell biology and may also reveal novel roles for sphingolipids in inflammatory diseases.