The research lines in our laboratory have aimed to uncover the molecular principles of protein folding in the cell and the cellular mechanisms that maintain and control this process. While fundamental in nature, this research bears direct relevance to biotechnology, health and disease. Every protein in a cell (whether viral or cellular) needs to acquire its proper structure to become biologically active; failure to do so causes loss of function via misfolding, aggregation, or degradation. The cell controls folding processes via networks of molecular chaperones, which are regulated by cellular stress responses.
Our research aims to uncover processes and mechanisms at the molecular level, using mammalian cells in culture complemented with in-vitro translations and purified proteins. We pair molecular biology, biochemistry, and pharmacological interventions with cell biology and live-cell imaging, and embrace opportunities from virology, immunology, medicine, to organic chemistry and industry. When suitable we exploit yeast, insect, and bacterial cells.
Establishing folding-function relationships of proteins
The usual term is “structure-function relationships”, and instead of structures at the atomic level we probe conformation (all changes) and stability of proteins in intact cells. This research line forms the basis for the other lines. Because of our in-depth knowledge of the folding of our model proteins we exploit them to study cellular processes and responses. We have characterized the basic folding pathways for several proteins in the endoplasmic reticulum (ER), the first compartment of the secretory pathway: the glycoproteins of influenza virus and HIV-1, the LDL receptor (a repeat protein), and the ABC-transporter CFTR, mutated in cystic fibrosis patients. Disulfide bonds, resistance to proteolytic digestion, antigenic epitopes, and glycosylation are examples of the features we use to probe folding. We for instance found that incorrect, inter-repeat disulfide bonds are obligatory intermediates during efficient folding of the LDL receptor rather than a road to misfolding, and that calcium incorporation is essential for the native phase of folding. We have several manuscripts in preparation that report on the interplay between the repeats within the LDL receptor and the role of this folding path for LDL receptor function, which suggests similar complex but essential folding pathways for other repeat protein families. Aim is to establish how the complete ER environment (from calcium, ATP, redox to folding factors) manages this folding pathway, preventing initial native folding of the repeats and supporting it at a later stage.
An exciting project in progress is the folding of CFTR and the defect in cystic fibrosis mutants. We showed before that the individual membrane-spanning (MSDs) and nucleotide-binding domains (NBDs) fold co-translationally, followed by a largely post-translational domain assembly phase. Rather than in the expected (as it contains the major patient mutation F508del) subdomain in NBD1, the defect appears in the integration of discontinuous up- and downstream subdomains, a folding pathway reminiscent of our yet unpublished influenza virus hemagglutinin folding path. This and other similarities suggest a general principle of proteins folding as hairpins, zippering up from core to periphery, irrespective of protein sequence and structure, a suggestion we will follow up with experiment. Understanding the folding path now allows an understanding of the mechanism of action of corrector drugs and molecular chaperones on the folding protein (see research line below). Our future focus is the role of intramembrane, ER-luminal, and cytosolic chaperoning during (multimembrane-spanning) domain assembly and the importance of the membrane (thickness) for proper assembly in the ER. For both LDL receptor and CFTR we have started examining the role of local translation rate on protein folding, by changing silent codons.
To relate folding to function we have used reverse genetics for HIV-1 (inserting mutant Env into the virus) in combination with in-vitro evolution, a powerful approach that teaches us how the virus solves a folding problem and escapes the defect. This marks events on the folding pathway, and also identifies the insults the virus can nót escape from. We intend to use the same approach for influenza virus. Next to common features, each protein shows peculiarities, such as the late, folding-dependent cleavage of the signal peptide that targets HIV Env to the secretory pathway. We are preparing 2 manuscripts that will describe the mechanism and function of this conserved process. It is a fascinating story of timing and regulation through conformational changes. Whereas signal peptide cleavage of Env bears all hallmarks of Signal Peptidase, we will examine its role as well as that of Signal Peptide Peptidase and other candidate intramembrane chaperones (see below).
One of the novel folding assays we established bridges the gap between in vitro refolding and folding in intact cells, with the aim to reconstitute the folding process in vitro. We have now managed to successfully fold influenza virus hemagglutinin in a diluted detergent cell lysate through addition of the proper redox buffer and excess protein disulfide isomerase, and plan to use this for further reconstitution and characterization of the folding process at the level of the folding molecule.
This research line offers detailed insights into folding of proteins in intact cells, its requirements, and its relation to protein function. It generates new (folding) assays, tools, and reagents and so far has allowed us to determine the molecular target of a corrector drug in the clinical pipeline for cystic fibrosis, design of assays suitable for high-throughput screening for antiviral or CFTR corrector compounds, and the partial reconstitution of folding of a complex protein without an intact ER.
Molecular mechanisms of assisted protein folding
Protein folding is neither effective nor efficient without assistance by folding enzymes and molecular chaperones. We focus on the folding proteins –clients or substrates– and aim to understand mechanism of action. A decade ago we had set out to identify novel chaperones and folding enzymes in the ER of plasma cells, and our proteomics screen in activated B cells identified seven. Using a combination of cell-free and in-cell approaches we aim to find mechanism of action and function of two novel protein families, which may functionally interact with Grp94 in the ER. For these novel families we aim to obtain funding to move into a model organism, preferably Drosophila, zebra fish, or mouse.
Particularly interesting is the pERp1 family of 5 proteins, low-activity redox enzymes that likely need activation by a partner protein. Of the 5 homologs only pERp1 is immune-cell specific. It is essential for efficient antibody secretion, but the family has broad substrate clientele. It shows homology with saposins, opening avenues to connect lipid metabolism and protein chaperoning. pERp1 resides in large chaperone complexes that include Grp94, which is a relatively poorly studied molecular chaperone with a limited client list, which includes our extremely well-characterized LDL receptor. New in our research questions will be the role of calcium, as the ER is the major cellular calcium store, Grp94 is one of its major calcium-binding resident proteins, Grp94 and pERps regulate and are subject to changing calcium concentrations, and last but not least, because calcium is essential for the LDL receptor to acquire and maintain its functional conformation.
Similar questions on function and mechanism of action we address for the effect of Hsp90, Grp94’s cytosolic paralog, on CFTR folding and maturation. Here we are close to coupling bottleneck co-translational folding steps of CFTR with this chaperone, a surprise as Hsp90 is considered to act late in the maturation of its clients. Major future focus will be the identification and characterization of intramembrane chaperones during CFTR’s domain assembly. Calnexin, Derlins, Bap31/29, rhomboids/iRhoms, signal peptidase, SPPase, and Sec61 are candidates we are exploring.
Next to general chaperones and folding enzymes, we work the hypothesis that many of the proteins that interact with our model proteins during their functioning are likely to interact early in the secretory pathway, allowing mutual chaperoning. CFTR is a stellar example of this, as some 80 proteins have been reported to physically interact with this protein and so-called genetic modifiers determine the clinical phenotype of the F508del patient genotype. The majority of modifying gene products associate with or function during CFTR channel functioning at the cell surface. We are examining various candidates for chaperone activity and will characterize their mutual action with CFTR. Amongst these are other multimembrane-spanning channels, testifying to the importance of the membrane-spanning domains and the role of the ER membrane during CFTR folding.
Although we are a major fan of the reductionist approach to establish molecular mechanisms, folding factor networking is essential for their functioning. None work alone and chaperones and folding enzymes are particularly adaptive (see below). We have been examining this issue by characterizing proteome and transcriptome of cells with stable decreased expression of Grp94. The adaptation turned out to be broad and small, broad because all disulfide-exchanging enzymes changed ~1.2x, considered not significant but striking, and small with only the ER Hsp70 BiP increasing two-fold. We plan to map individual folding factor interactions with clients using the split-GFP approach we used for our peroxisome studies (see below). We started using this approach in the yeast cytosol for proof of principle, and are in the process of transplanting it to mammalian cells and the ER. As this approach traps interactions, we will in parallel use BRET (a biosynthetic alternative to FRET) to examine interactions in living cells.
Organelle biogenesis and maintenance
To accommodate secretion, ER size is regulated by demand. When excess misfolded or folding protein accumulates in the ER (as in an antibody-secreting plasma cell or in viral infection), a stress signal is activated, leading to increased transcription of genes for ER-resident folding factors as well as ER membrane synthesis. This stress-related signaling cascade is used for physiological regulation in the absence of a stressor, and is exploited by viruses to modulate innate immune signaling and virus propagation. Alternative stress responses upregulate the cellular chaperone systems, in particular Hsp90, Hsp70, and their co-chaperones. We aim to establish and manipulate signaling routes and interaction networks that regulate the growth, shrinkage, and maintenance of the ER volume, in ER stress, in physiological ER expansion, as well as upon viral infection.
A prime example of physiological expansion of the ER occurs upon differentiation of B cells into plasma cells. The initial phase involves a largely unknown signaling path from LPS-bound TLR4 to transcription factors for membrane expansion and ER folding factors, and the second phase is thought to be a canonical ER stress response, which can be elicited as well by ER stressors such as (drugs that induce) misfolded proteins. We recently found in pilot experiments, however, that also the second phase deviates from a canonical ER stress response. We moreover found that a proteotoxic ER-stress response in non-B cells occurs in waves of activation and silencing unrelated to ER function, and that the timing and outcome depend on the presence of the ER-resident Grp94. The adaptation of cells to new demands uses the stress response sensors and signaling cascades in more subtle ways, which we aim to uncover, following Ire1, ATF6, PERK, and their downstream cascades.