Membrane Biology of Infectious Diseases and Developmental Disorders

Dysferlinopathy constitutes a group of rare, genetic diseases, including limb-girdle muscular dystrophy (LGMD)2B/R2/DYSF–related and Myoshi myopathy, classified as adult-onset of progressive weakness and wasting of skeletal muscle (SM). The sarcolemma endures transient injuries resulting from normal stresses associated with activity. In healthy individuals, these wounds are repaired. However, mutations in dysferlin result in excessive wounding and/or compromised healing. To better understand molecular drivers of dysferlinopathies, we identified age-dependent changes in the proteomic profile of SM in wild-type (WT) and dysferlin-deficient mice, using novel analytical techniques developed in part in our laboratory, i.e., mass spectrometry (MS)–based proteomic analysis [Reference 1]. It is currently revolutionizing our understanding of human disease. What began as a discovery-based approach to identify and quantify proteins can now be used in a targeted manner; the rapid expansion of this technology and methodology permits structural and functional proteomics through the identification of post-translational modifications, cellular localization, and protein-protein interactions. As a result, significant advances have been made in biomarker and drug discovery, as well as personalized medicine. Furthermore, the comprehensive integration of omics data, such as genomics and transcriptomics, can provide a summary of how an entire system is behaving. The aim of this study was to identify and relatively quantify proteins across the lifespans of, and between, WT and dysferlin-deficient mouse quadricep SM.

To quantify progressive changes in healthy and dysferlin-deficient mouse SM proteomes, quadriceps were isolated from 6-, 18-, 42-, and 77-wk-old C57BL/6 (WT, Dysf^+/+) and BLAJ (Dysf^–/–)mice (n = 3, 2 male/1 female or 1 male/2 female, 24 total). Whole-muscle proteomes were characterized using liquid chromatography-mass spectrometry with relative quantification using TMT10plex isobaric labeling. Principle component analysis was used to detect age-dependent proteomic differences over the lifespan of, and between, WT and dysferlin-deficient SM. The biological relevance of proteins with significant variation was established using Ingenuity Pathway Analysis. Over 3,200 proteins were identified between 6-, 18-, 42-, and 77-wk-oldmice. In total, 46 proteins varied significantly in aging WT SM, while 365 varied in dysferlin-deficient SM. However, 569 proteins varied between aged-matched WT and dysferlin-deficient SM. Proteins with significant variation in expression across all comparisons followed distinct temporal trends. Proteins involved in sarcolemma repair and regeneration underwent significant changes in SM over the lifespan of WT mice, while those associated with immune infiltration and inflammation were overly represented over the lifespan of dysferlin-deficient mice. The proteins identified herein are likely to contribute to our overall understanding of SM aging and dysferlinopathy disease progression [Reference 1].

The receptor tyrosine kinase (RTK) KIT and its ligand stem cell factor (SCF) are essential for human mast cell (huMC) survival and proliferation. HuMCs expressing oncogenic KIT variants secrete large numbers of extracellular vesicles (EVs) [Reference 2]. EVs are cell-secreted nanovesicles loaded with selected molecular cargo such as lipids, nucleic acids, and proteins that may reflect the status of the cell of origin; last year we reported a new class of EV from human umbilical vein [Reference 3]. After their secretion, EVs may transfer donor-cell cargo to recipient cells and potentially modify functional responses. As such, EVs are intercellular communicators conveying messages to proximal or distal tissue environments. While there is a vast heterogeneity regarding their biogenesis, size and composition, the most studied types of EVs are ectosomes (often also referred to as microvesicles) and exosomes. Ectosomes directly bud from the plasma membrane and, although there are no definite consensus protein markers for EV subtypes yet, annexin A1, ARF6 and basigin are considered characteristic of ectosomes. Exosomes originate from the endosomal compartment by the initial formation of intraluminal vesicles (ILVs) from endosome membranes, which are released as exosomes when multivesicular bodies (MVBs) fuse with the plasma membrane. The role KIT plays in regulating EV secretion has not been examined. The goal of this project is to investigate the effects of stimulation or inhibition of KIT activity on the secretion of small EVs (sEVs).

In huMCs expressing constitutively active KIT, the quantity and quality of secreted sEVs positively correlated with the activity status of KIT. SCF–mediated stimulation of KIT in huMCs or murine MCs, or of transiently expressed KIT in HeLa cells, enhanced the release of sEVs expressing exosome markers. In contrast, ligand-mediated stimulation of the RTKEGFR in HeLa cells did not affect sEV secretion. The release of sEVs induced by either constitutively active or ligand-activated KIT was remarkably decreased when cells were treated with KIT inhibitors, concomitant with reduced exosome markers in sEVs. Similarly, inhibition of oncogenic KIT signaling kinases such as PI3K, and MAPK significantly reduced the secretion of sEVs. Thus, activation of KIT and its early signaling cascades stimulate the secretion of exosome-like sEVs in a regulated fashion, which may have implications for KIT–driven functions.

The fully assembled IAV has on its surface the highest density of a single membrane protein found in nature: the glycoprotein hemagglutinin (HA), which mediates viral binding, entry, and assembly [Reference 4]. HA clusters at the plasma membrane of infected cells, and the HA density (number of molecules per unit area) of these clusters correlate with the infectivity of the virus. Dense HA clusters are considered to mark the assembly site and ultimately lead to the budding of infectious IAV. The mechanism of spontaneous HA clustering, which occurs with or without other viral components, has not been elucidated. Because the interruption of viral assembly halts viral disease, a better understanding of the interaction of viral components with host-cell membranes is crucial to the development of new antiviral therapies. The matrix protein of the influenza virus (M1) can bind to lipid bilayers through electrostatic interactions and can form virus-like particles (VLPs) in the absence of other viral components, demonstrating its ability to interact with lipids, including cholesterol, in the plasma membrane (PM). Upon such binding, M1 multimerizes. This formation of M1 multimers or clusters is suspected to play a crucial role in the viral life cycle, specifically in the control of membrane curvature, viral morphology, and budding. However, the role of HA in the clustering of M1 during interaction with the PM requires further elucidation. Our recent finding that HA and phosphatidylinositol 4,5-bisphosphate (PIP2) can interact to modulate HA clustering in the PM raised the possibility that HA interactions with M1 could be mediated by PIP2.

We focused on PIP2, a minor component of the PM in mammalian cells (about 1% of total lipid). Yet as the most abundant polyphosphoinositide it plays an outsized role in cell function. PIP2 mediates many intracellular processes, such as endo- and exocytosis, actin cytoskeleton regulation, cytoskeleton PM adhesion, and many others. PIP2 clusters at the PM, and these clusters are known to interact with various cytoskeletal proteins, such as actin and actin-binding proteins. Proteins that have been purified from IAV are also known to interact with PIP2, and IAV exploits several PIP2–dependent pathways. Because other viruses, such as HIV and Ebola, use PIP2 to mediate the membrane binding of their capsid proteins, and because IAV HA and nucleoprotein have known interactions with PIP2, it is reasonable to investigate whether PIP2 and M1 might also interact. We recently showed that the quaternary ammonium compound cetylpyridinium chloride (CPC), at non-cytotoxic concentrations, modulates the membrane association and clustering of PIP2–binding proteins such as myristoylated alanine-rich C-kinase substrate (MARCKS) and phospholipase C (PLC) through its pleckstrin homology (PH) domain. CPC contains a positively charged head group and a hydrophobic tail, allowing it to efficiently associate with membrane bilayers and micelles. The detergent action of CPC is observed when the concentration of CPC is above its critical micelle concentration (CMC) in water, which has been quantified by various methods in different buffers and temperatures to be in the range about 600–900 M. Although CPC has been previously shown to possess antibacterial and antiviral properties through micelle formation at millimolar concentrations, there is much less information on how low micromolar concentrations of CPC could affect interactions between IAV components. Our recent CPC study showed that HA and PH (PIP2) co-clustering were modulated by CPC and suggested a mechanism for antiviral properties of CPC at non-cytotoxic concentrations. In this study, we are building on these findings and testing the hypothesis that CPC modulates the assembly of HA and M1. Using the localization-based super-resolution microscopy technique of fluorescence photoactivation localization microscopy (FPALM) with total internal reflection fluorescence (TIRF) excitation, we also shed light on the effect of M1 on the PM distribution of PIP2 and the effect of HA in the clustering of M1 adjacent to the PM. The interaction of M1 with lipid bilayers is primarily electrostatic, and M1 can bind lipid bilayers through multiple residues. M1 is known to interact with PS in cells and cell models, but M1 interaction with other lipids has not been extensively studied. If interactions between M1 and lipid bilayers are primarily electrostatic, it is plausible that M1 could interact with lipids that are more negatively charged than PS. We tested this hypothesis and found that M1 colocalized with the PLC–PH domain, an established marker of PIP2. The finding that PIP2 clustering was significantly enhanced in the presence of M1 is consistent with the response of PIP2 clustering to other positively charged membrane-associated molecules and ions and the positively charged face of M1, supporting the notion that M1 interacts with PIP2 electrostatically at the PM. The quaternary ammonium compound CPC, found previously to disrupt PIP2 clustering and reduce the PM association of PIP2–binding proteins at low micromolar concentrations, significantly disrupted the colocalization, clustering, and co-clustering of HA and M1, as well as the colocalization, clustering, and co-clustering of M1 and PIP2. CPC also improved the outcomes of IAV infection in vivo. The implicated role of PIP2 in HA–M1 interaction and M1 association with the PM is novel, suggesting that it may be fruitful to explore strategies that modulate host-cell PIP2 or other phosphoinositides to prevent or inhibit IAV infection.

Viral spike proteins mutate frequently, but conserved features within these proteins often have functional importance and can inform development of anti-viral therapies that circumvent the effects of viral sequence mutations, leading to inefficacy of vaccines or drug resistance. Because of their involvement in viral binding and entry, spike-protein ectodomains have been well characterized. However, intracellular portions of spike proteins can interact with host cell components and other viral components for trafficking to appropriate locations for assembly and budding. Despite the typically small relative size of viral spike intracellular domains (vSIDs), the domains can potentially be anti-viral drug targets. Anti-viral drug designs that target conserved viral features are highly desirable. The cytoplasmic tails of some spike proteins have a cysteine-rich motif. For example, the 1011–amino acid intracellular domain of the HA from IAV is known to contain three highly conserved cysteines, C555 (technically within the transmembrane domain), C562, and C565, which can be palmitoylated or sometimes stearoylated, and whose removal has been shown to alter viral morphology and replication. Site-directed mutagenesis has shown that these cysteines are crucial for budding and infectivity, particularly C565. Proper folding of HA in the endoplasmic reticulum (ER) also requires palmitoylation. Mutating these final cysteines can inhibit HA fusion activity, alter viral morphology, or impair release, ultimately resulting in a non-infectious virus. Host cell proteins with clusters of basic amino acids and acylated residues could interact with phosphoinositides, which mediate their trafficking to the PM. We therefore set out to test whether such amino acid patterns are found in viral intracellular domains, to test whether such features are conserved in viral intracellular domains, and to test whether they can interact with host-cell phosphoinositides.

We studied large numbers of viral spike-protein sequences from several viral families. We analyzed the degree to which viral spike proteins contain a particular motif within their vSID: one or more basic amino acids near a potentially acylated cysteine, serine, glycine, or hydrophobic amino acid. We abbreviate this generalized feature as (C+/CX+), with the reverse of the features also being included (i.e., +XC or + C). We chose to analyze a subset of viruses with relevance to human health: influenza viruses A and B, with particular attention to avian influenza virus strains, SARS-CoV-2 and other β-coronaviruses, Ebolavirus, HIV1, RSV, and measles. While gag is not technically a spike protein, it does have known interactions with anionic lipid membranes and is thus of interest. HIV1 envelope protein was also included as a comparison. We used several approaches: sequence analysis to identify patterns with putative phosphoinositide interaction; all-atom molecular dynamics to test for interactions between phosphoinositides and these residues in viral proteins; and super-resolution microscopy to test for nanoscale interactions between these proteins and phosphoinositides in fixed cells [Reference 5]. We found highly (over 99%) conserved patterns within their intracellular domains. The patterns generally consist of one or more basic amino acids (arginine or lysine) adjacent to a cysteine, many of which are known to undergo acylation. These patterns were not enriched in cellular proteins in general. Molecular dynamics simulations show direct electrostatic and hydrophobic interactions between these conserved residues in HA from influenza A and B and the phosphoinositide PIP2. Super-resolution microscopy shows nanoscale colocalization of PIP2 and several of the same viral proteins. We propose the hypothesis that these conserved viral spike-protein features can interact with phosphoinositides such as PIP2. In conclusion, we analyzed viral spike proteins from a number of viral families and found that they contain highly conserved sequence features in their cytoplasmic tails, typically consisting of one or more basic amino acids in proximity to one or more cysteines. We hypothesized that these basic amino acids, together with the cysteines, when acylated, could interact with phosphoinositides, as many cellular proteins are known to do. Using molecular dynamics simulations, we showed that, in HA from influenza A and B, these features do interact directly with PIP2 through a combination of electrostatic interactions and association of the acylated cysteines with the PIP2 tails. Experimental evidence from super-resolution microscopy verifies that several of these same spike proteins colocalize with PIP2 at the nanoscale. Taken together, these results suggestive a mechanism of interaction between viral spike proteins and host cell phosphoinositides. Further work is needed to test the generality of this proposed mechanism.