Influenza A virus (IAV) employs multiple strategies to manipulate cellular mechanisms and support proper virion formation and propagation. In this study, we performed a detailed analysis of the interplay between IAV and the host cells' proteostasis throughout the entire infectious cycle. We reveal that IAV infection activates the inositol requiring enzyme 1 (IRE1) branch of the unfolded protein response, and that this activation is important for an efficient infection. We further observed the accumulation of virus -induced insoluble protein aggregates, containing both viral and host proteins, associated with a dysregulation of the host cell RNA metabolism. Our data indicate that this accumulation is important for IAV propagation and favors the final steps of the infection cycle, more specifically the virion assembly. These findings reveal additional mechanisms by which IAV disrupts host proteostasis and uncovers new cellular targets that can be explored for the development of host -directed antiviral strategies.

Metadata for the manuscript entitled "ATG9A regulates dissociation of recycling endosomes from microtubules leading to formation of influenza A virus liquid inclusions" - MAIN FIGURES

AbstractInfluenza A virus (IAV) is one of the main causes of annual respiratory epidemics in humans. IAV employs multiple strategies to evade host immunity and hijack cellular mechanisms to support proper virion formation and propagation. Some of these strategies encompass the manipulation of pathways involved in protein homeostasis, leading to changes in the host proteome and protein distribution within the cell. In this study, we performed a detailed analysis of the interplay between IAV and the host cells’ proteostasis mechanisms throughout the entire infectious cycle. We reveal that IAV infection induces the activation of the inositol requiring enzyme 1 (IRE1) branch of the unfolded protein response (UPR), at an infection stage that coincides with high rates of viral protein translation. This activation is particularly important for infection, as attenuation of virus production was observed upon IRE1 inhibition. Concomitantly to UPR activation, we observed the accumulation of virus-induced insoluble protein aggregates, which contain both viral and host proteins and are associated with a dysregulation of the host cell RNA metabolism. We demonstrate that this accumulation is important for IAV propagation, as its prevention using a quinoline-steroid hybrid compound significantly reduces the number of produced infectious virus particles. Our data suggests that the formation of these insoluble protein aggregates favors the final steps of the infection cycle, more specifically the virion assembly. Our findings reveal additional mechanisms by which IAV disrupts the host cell proteostasis to favor infection and uncover new cellular targets that can be explored for the development of host-directed antiviral strategies.

It is now established that many viruses that threaten public health establish condensates via phase transitions to complete their lifecycles, and knowledge on such processes may offer new strategies for antiviral therapy. In the case of influenza A virus (IAV), liquid condensates known as viral inclusions, concentrate the 8 distinct viral ribonucleoproteins (vRNPs) that form IAV genome and are viewed as sites dedicated to the assembly of the 8-partite genomic complex. Despite not being delimited by host membranes, IAV liquid inclusions accumulate host membranes inside as a result of vRNP binding to the recycling endocytic marker Rab11a, a driver of the biogenesis of these structures. We lack molecular understanding on how Rab11a-recycling endosomes condensate specifically near the endoplasmic reticulum (ER) exit sites upon IAV infection. We show here that liquid viral inclusions interact with the ER to fuse, divide, and slide. We uncover that, contrary to previous indications, the reported reduction in recycling endocytic activity is a regulated process rather than a competition for cellular resources involving a novel role for the host factor ATG9A. In infection, ATG9A mediates the removal of Rab11a-recycling endosomes carrying vRNPs from microtubules. We observe that the recycling endocytic usage of microtubules is rescued when ATG9A is depleted, which prevents condensation of Rab11a endosomes near the ER. The failure to produce viral inclusions accumulates vRNPs in the cytosol and reduces genome assembly and the release of infectious virions. We propose that the ER supports the dynamics of liquid IAV inclusions, with ATG9A facilitating their formation. This work advances our understanding on how epidemic and pandemic influenza genomes are formed. It also reveals the plasticity of recycling endosomes to undergo condensation in response to infection, disclosing new roles for ATG9A beyond its classical involvement in autophagy.

Full text Figures and data Side by side Abstract Editor's evaluation eLife digest Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract In biological systems, liquid and solid-like biomolecular condensates may contain the same molecules but their behaviour, including movement, elasticity, and viscosity, is different on account of distinct physicochemical properties. As such, it is known that phase transitions affect the function of biological condensates and that material properties can be tuned by several factors including temperature, concentration, and valency. It is, however, unclear if some factors are more efficient than others at regulating their behaviour. Viral infections are good systems to address this question as they form condensates de novo as part of their replication programmes. Here, we used influenza A virus (IAV) liquid cytosolic condensates, AKA viral inclusions, to provide a proof of concept that liquid condensate hardening via changes in the valency of its components is more efficient than altering their concentration or the temperature of the cell. Liquid IAV inclusions may be hardened by targeting vRNP (viral ribonucleoprotein) interactions via the known NP (nucleoprotein) oligomerising molecule, nucleozin, both in vitro and in vivo without affecting host proteome abundance nor solubility. This study is a starting point for understanding how to pharmacologically modulate the material properties of IAV inclusions and may offer opportunities for alternative antiviral strategies. Editor's evaluation Etibor and collaborators have performed a series of well-thought and careful experiments to understand some of the physical and thermodynamic properties of liquid condensates produced by the infection with the influenza A virus. However, their approach and rules could be easily applied to any other cellular phenomena that involve the formation of intracellular liquid condensates. Finally, this article is setting up the basis for an in-depth theoretical analysis of the physical phenomena described here and their correlation with the biology of intracellular liquid condensates. https://doi.org/10.7554/eLife.85182.sa0 Decision letter Reviews on Sciety eLife's review process eLife digest Cells are organized into compartments that carry out specific functions. Envelope-like membranes enclose some of those compartments, while others remain unenclosed. The latter are called biomolecular condensates, and they can shift their physical states from a more liquid to a more solid form, which may affect how well they function. Temperature, molecular concentration and molecular interactions affect the physical state of condensates. Understanding what causes physical shifts in biomolecular condensates could have important implications for human health. For example, many viruses, including influenza, HIV, rabies, measles and the virus that causes COVID-19, SARS-CoV-2, use biomolecular condensates to multiply in cells. Changing the physical state of biomolecular condensates to one that hampers viruses’ ability to multiply could be an innovative approach to treating viruses. Etibor et al. show that it is possible to harden condensates produced by influenza A virus. In the experiments, the researchers manipulated the temperature, molecular concentration and strength of connections between molecules in condensates created by influenza A-infected cells. Then, they measured their effects on the condensate’s physical state. The experiments showed that using drugs that strengthen the bonds between molecules in condensates was the most effective strategy for hardening. Studies in both human cells and mice showed that using drugs to harden condensate in infected cells did not harm the cells or the animal and disabled the virus. The experiments provide preliminary evidence that using drugs to harden biomolecular condensates may be a potential treatment strategy for influenza A. More studies are necessary to test this approach to treating influenza A or other viruses that use condensates. If they are successful, the drug could add a new tool to the antiviral treatment toolbox. Introduction Central to the spatiotemporal control of reactions in many viral infections is the formation of biomolecular condensates that facilitate key steps of viral lifecycles (Etibor et al., 2021). In influenza A virus (IAV) infection, this is key for assembling its segmented genome, a complex composed of eight different viral RNA segments (vRNA) (Pons, 1976). Each vRNA is encapsidated by molecules of nucleoprotein (NP) along its length, with one unit of the RNA-dependent RNA polymerase (RdRp, consisting of PB2, PB1, and PA) bound to the base-paired RNA termini, forming viral ribonucleoproteins (vRNPs) (Alenquer et al., 2019). How the eight vRNP complex self-assembles is unknown. It is known that genomic complex formation relies on RNA-RNA interactions between distinct vRNPs and is a selective process because most virions contain exactly eight different vRNPs, as reviewed elsewhere (Hutchinson et al., 2010). After export from the nucleus where vRNPs are synthesised, vRNPs reach the cytosol and induce the formation of cytosolic condensates, known as viral inclusions (Amorim et al., 2011; Avilov et al., 2012; Chou et al., 2013; Eisfeld et al., 2011; Lakdawala et al., 2014; Momose et al., 2011), which we postulated to be sites dedicated to IAV genome assembly (Alenquer et al., 2019). Interestingly, IAV cytosolic inclusions exhibit liquid properties (fuse and divide, dissolve upon shock and are dynamic) (Alenquer et al., 2019), providing the first indication that defined material properties are critical for the formation of influenza epidemic and pandemic genomes. As the list of viruses utilising liquid biomolecular condensates rapidly increases, including reoviruses, human cytomegalovirus, HIV, rabies, measles, SARS-CoV-2 (reviewed in Etibor et al., 2021; Lopez et al., 2021), it becomes pertinent to ask whether targeting the material properties could constitute a novel antiviral approach. Recently, the Sonic hedgehog pathway antagonist cyclopamine and its analogue A3E were demonstrated to inhibit human respiratory syncytial virus (hRSV) replication by altering the material properties of viral condensates (Risso-Ballester et al., 2021). However, compounds targeting hRSV-related (Risso-Ballester et al., 2021) and cancer-associated condensates exhibited off-target effects (Klein et al., 2020). Therefore, a critical advance in condensate disease therapy, including in the case of viral infection, requires defining how to efficiently and specifically target selected biomolecular condensates. In several studies, it was demonstrated that the properties of biomolecular condensates respond to many factors in a system-dependent manner (Alberti et al., 2019; Falahati and Haji-Akbari, 2019; Hyman et al., 2014; Riback and Brangwynne, 2020; Mittag and Parker, 2018; Snead and Gladfelter, 2019; Milovanovic and De Camilli, 2017; Perdikari et al., 2020). Entropic free energy (Quiroz and Chilkoti, 2015), concentration (Riback et al., 2020), type, number and strength of interactions (Sanders et al., 2020) have all been demonstrated to affect the properties of biomolecular condensates. In vivo, it is unknown if these changes affect equally the material properties (and function) of biomolecular condensates. This knowledge is the basis to understand which pathways could be manipulated to modulate the material properties of selected condensates, ultimately targeting their behaviour and function. For example, pathways affecting local energy production, consumption, or metabolism will alter the free energy landscape of biomolecular condensates (Patel et al., 2017). Similarly, pathways that regulate the local density of condensate drivers could affect concentration (Banani et al., 2016; Riback et al., 2020). Finally, pathways regulating post-translational modifications (Rai et al., 2018), local pH (Kroschwald et al., 2018; Munder et al., 2016), or ionic strength (Yang et al., 2020), as well as strategies promoting aggregation or dissolution of condensate interactomes could affect the type, number (valency), and strength of interactions (Bracha et al., 2019a, Bracha et al., 2019b, Zhu et al., 2019). Viral replication programmes rely on cellular pathways, and as such, their condensate biology is more complex than attainable with in vitro reconstituted systems. For this reason, studying them in their native intracellular environment will more accurately define how they respond to specific stimuli. Our study depends on introducing a disturbance and evaluating its effect on the material properties through two different approaches. The first is to map the intracellular biophysical traits of the perturbations on viral condensates measuring thermodynamic parameters, as done in Riback et al., 2020; Shimobayashi et al., 2021; Wei et al., 2020. As liquids tend to be spherical to minimise surface tension (Elbaum-Garfinkle et al., 2015; Lee et al., 2023), we assessed their morphology and topology, and as molecules move freely in liquids, we also measured the Gibbs free energy of partition (henceforth called free energy, ΔG) to define the degree of molecular stabilisation (Shin et al., 2017; Riback et al., 2020). The lower the value of ΔG, the more stabilised the system becomes. The calculation of ΔG requires assessing the ratio of concentration of material inside a condensate (Cdense), relative to the concentration dissolved in the cytosol (Cdilute), which provides the value of the partition coefficient (K), and allows the use of the formula ΔG = -RTInK and evaluate how the system progresses. In addition, the nucleation density determines how many viral condensates are formed per area of cytosol. Overall, the data will inform us if changing one parameter, e.g. the concentration, drives the system towards condensates of different sizes with the same or more stable properties (e.g. induced by phase transitions), or changes their abundance on account of additional available nucleation centres or dissolves the condensates (Riback et al., 2020; Snead et al., 2022). These are the most relevant parameters for our scientific question, which is to define how IAV liquid viral inclusions can be hardened. However, other values are retrieved from these analyses, revealing, for example, the types of interactions required to maintain the properties of the system. This can be illustrated by reports showing that with bulk concentration, Cdilute is constant in a binary mixture (Klosin et al., 2020) but increases in multi-component systems (Riback et al., 2020). This type of information has implications about the condensates formed during influenza infection. If the system is binary, the eight vRNPs could behave as a single component, on account of their similar structure, and interact with the host Ras-related protein in the brain 11 (Rab11), reported to be part of inclusions (Avilov et al., 2012; Eisfeld et al., 2011; Amorim et al., 2011). If IAV inclusions are multi-component systems, each vRNP could behave as an independent entity, on account of differences in length, RNA sequence, and valency or, as an alternative, viral inclusions could contain yet unidentified components other than Rab11 and vRNPs. The second type of approach to accurately define how IAV liquid inclusions respond to specific stimuli is to assess the effect of the perturbations in the dynamics and kinetics of viral condensates, thus inferring changes in elasticity and viscosity (Wang et al., 2022; Alberti et al., 2019). This relies on using live-cell imaging approaches to assess if two condensates can fuse, internally rearrange, or exchange material, which are all bona fide traits of liquids but not of solids (Wang et al., 2022; Alberti et al., 2019). The information collected will offer insight into how the material properties of IAV inclusions are regulated and maintained. In addition, understanding how IAV viral inclusions can be hardened can be therapeutically relevant if they rely on their liquid character for function. The objective of this project is to determine the most effective and precise approaches for hardening IAV liquid inclusions. We find that the stabilisation of intersegment interactions is more efficient at hardening IAV inclusions than varying the temperature or the concentration of the drivers of IAV inclusions. Importantly, we show that the hardening topological phenotype is observed in the lungs of infected mice. We also report that it is possible to affect viral inclusions without imposing additional changes in host protein abundance and solubility using solubility proteome profiling (SPP) of infected cells (Sridharan et al., 2022). In sum, our data support the development of strategies targeting the material properties of cellular condensates in viral infections and provide a critical advance in how these structures may be regulated. Results Framework to identify perturbations that harden IAV liquid inclusions We previously demonstrated that viral inclusions formed by IAV infection display a liquid profile in the sense that they drip, acquire a spherical shape upon fusion, and dissolve in response to hypotonic shock or brefeldin A treatment (Alenquer et al., 2019). Here, we seek to identify the best strategies to harden viral inclusions to investigate if altering their material properties may be a novel antiviral therapy. For this, we systematically probed and compared the impact of temperature, concentration, and number/strength of ligations on the material properties of liquid viral inclusions, as a proxy of entropic, molecular, and valency contributions, respectively. We selected these parameters given the deep understanding on how they regulate the interactions amongst components and the material properties of condensates (Riback et al., 2020; Sanders et al., 2020; Quiroz and Chilkoti, 2015; Figure 1A). Methodologically, we employed established protocols for imposing the selected thermodynamic perturbations. We quantified the impact of these perturbations on the number, nucleation density (ρ=number of inclusionCytoplasm Area, µm-2), size, shape, dynamics, supersaturation (S=InCdiluteCsaturation, in which Csaturation is the concentration above which molecules demix from an homogenous system), and the Gibbs free energy of partition (ΔG = -RTInK, in which K=CdenseCdilute is the partition coefficient) to define how our system adapts to the perturbations. Material concentrations inside (Cdense) and outside (Cdilute) viral inclusions were measured using the analytical strategies described in Riback et al., 2020; Shimobayashi et al., 2021, and shown in Figure 1B (and validated as described in Materials and methods and in Figure 2—figure supplement 1). For this, we used the mean fluorescence intensity (MFI) of NP as proxy of vRNP concentration (Amorim et al., 2011; Vale-Costa et al., 2016), as it is well established that the majority of cytosolic NP is in the form of vRNPs (Avilov et al., 2012; Momose et al., 2011; Eisfeld et al., 2011; Amorim et al., 2011). Figure 1 Download asset Open asset Framework applied in this study to define the hierarchy in effectiveness in hardening influenza A virus (IAV) liquid inclusions or other condensates. (A) To compare the contributions of entropy, concentration, and valency/strength/type of interactions, we subjected infected cells to the different perturbations: temperature, concentration of viral inclusion drivers (viral ribonucleoproteins [vRNPs] and Ras-related in brain 11a [Rab11a]) and number or strength of interactions between different vRNPs using the well-studied vRNP pharmacological modulator, nucleozin, that increases the number and strength of intersegment interactions. (B) Our aim is to determine which amongst these perturbations impact more dramatically the material properties of viral inclusions and for this we measured number, shape, size, and Gibbs free energy of partition (free energy, ΔG). Methodologically, we segmented circa 20 cells under the different conditions to measure the above-mentioned parameters and the amount of material inside (Cdense) and outside (Cdilute) viral condensates. With this, we calculated the partition coefficient K and extrapolated the ΔG. (C) When ΔG dramatically changed, we assessed how perturbations altered the material properties of IAV inclusions by comparing how fast and how much they moved (using coarsening assays, particle tracking, photobleaching to inspect internal rearrangements, and fluorescence loss after photoactivation [FLAPh]). (D) We also assessed whether the phenotype could be recapitulated in vivo using mice infected with influenza A virus reassortant X31. The overall goal of this framework is to determine, for IAV, how liquid inclusions may be efficiently hardened to prioritise research and development of strategies with that activity. Additionally, the framework may be applied to other systems, including other viruses, for informed decisions on how to harden condensates. Our goal was to identify which perturbations translated into significant shifts in ΔG to further explore whether these resulted in dramatic alterations in the material properties of viral inclusions, by assessing their kinetics and dynamics (Figure 1C) and determine how they impact viral replication in vivo (Figure 1D). Changes in temperature mildly perturb IAV inclusions Cellular steady state is maintained at a narrow permissive physiological range, including of temperature. However, biomolecular condensates respond to fluctuations in temperature, and we took advantage of this to assess the entropic contribution of free energy and evaluate whether regulating host cell metabolism could offer future solutions to harden IAV liquid inclusions (Figure 2A). We quantitatively analysed the viral inclusions formed in cells incubated at 4°C, 37°C, and 42°C for 30 min at 8 hr post-infection (hpi) (representative images in Figure 2B). This short duration in temperature shift did not alter the levels of cytosolic vRNPs, as expected (Figure 2C). Increasing the temperature from 37°C to 42°C did not significantly change the size (Figure 2D), aspect ratio, or number of viral inclusions (Figure 2D–G), but decreased the concentration of vRNPs in condensates (Cdense), and increased the nucleation density despite not altering the concentration of vRNPs in the milieu (Cdilute) (Figure 2H–M, Supplementary file 1 (Sheet 1)). This means that increasing the temperature up to 42°C still maintains the system in a two-phase regime, but affects the nucleation capacity, increasing the number of inclusions. Of note, vRNPs become homogenously distributed in the cytosol at 43.5°C (data not shown), pointing that the entire system may undergo regulatory processes. Importantly, this increase in temperature modestly destabilised the structure, as observed by an increase in Gibbs free energy (–2167.3±2361 J/mol @ 37°C to –1477.9±228 J/mol @ 42°C, mean ± SD, Figure 2N–O, Supplementary file 1 (Sheet 1)). Conversely, decreasing the temperature until 4°C leads to an increase in the size of inclusions that is statistically significant considering 42°C to 4°C shifts only (shift in area from 0.2896±0.02 µm2 at 42°C to 0.3474±0.05 at 4°C), rounds up liquid inclusions, and decreases their nucleation capacity and abundance (the latter significant only considering 42°C and 4°C, Figure 2D–H and Supplementary file 1 (Sheet 1)). A drop in temperature increases the concentration of vRNPs in inclusions (Cdense at 37°C of 3116.0±0.05 AU, mean ± SD, and at 4°C of 2144.5±0.04 AU, Supplementary file 1 (Sheet 1)), and does not significantly change the stability of IAV inclusions as determined by Gibbs free energy (–2415.6±273 J/mol @ 4°C, Figure 2N and O). Overall, the data indicate that the temperature increasing from 4°C to 42°C shifts our system to smaller inclusions that have less vRNPs. Cdilute did not change but there is an increase in nucleation density which indicates that heat disruption of weak molecular interactions leads to alterations in nucleation, fusion, and fission, as reported previously (Iserman et al., 2020). However, we did not observe significant alterations in the stabilisation of our system, supporting that alterations in the conditions tested for temperature do not affect the material properties of viral inclusions. Figure 2 with 1 supplement see all Download asset Open asset Thermal changes mildly perturb the material properties of inclusions. A549 were infected at a multiplicity of infection (MOI) of 3 with PR8 virus for 8 hr, incubated at different temperatures (4°C, 37°C, 42°C) for 30 min, fixed, and analysed by immunofluorescence using antibody staining against Rab11 and nucleoprotein (NP) as a proxy for viral ribonucleoprotein (vRNP). The biophysical parameters were extracted from immunofluorescence images (n=15–20), adapting the method published by Riback et al., 2020; Shimobayashi et al., 2021, to determine concentration Cdense as the mean fluorescence intensity of vRNPs in the segmented influenza A virus (IAV) inclusions, while concentration Cdilute was extrapolated from the cytoplasmic vRNP intensity outside the inclusions. Each dot is the average value of a measured parameter within or outside IAV inclusions per cell. Also, size and shape of inclusion were extracted from inclusions after image segmentation. Parameters that were normalised to an infection state without IAV inclusions (3 hr post-infection [hpi]) are indicated by a dashed horizontal line. Above each boxplot, same letters indicate no significant difference between them, while different letters indicate a statistical significance at α=0.05. All data are displayed in Supplementary file 1 (Sheet 1). Abbreviations: AU, arbitrary unit. (A) Representative depiction of the experimental analysis workflow. (B) Representative images of fixed A549 cells infected with PR8 virus showing alterations in viral inclusions at different temperatures. (C). Boxplot depicting the fold change in cytoplasmic to nuclear vRNP concentration; p=0.684 by one-way ANOVA followed by Tukey multiple comparisons of means. (D) Boxplot of mean viral inclusion area (µm2) per cell; p=0.00234 by Kruskal Wallis Bonferroni treatment. (E) Boxplot of aspect ratio of inclusion; p<0.001 by one-way ANOVA followed by Tukey multiple comparisons of means. (F) Scatter plot of inclusions circularity versus roundness. (G) Boxplot showing number of viral inclusions per cell; p<0.001 by one-way ANOVA, followed by Tukey multiple comparisons of means. (H) Scatter plot of nucleation density (ρ = numberofinclusioncytoplasmArea µm–2) versus degree of supersaturation (S=In CdiluteCsat), as a measure of propensity to remain dispersed in the cytoplasm. (I) Boxplot of of vRNP concentration within inclusions (Cdense (AU)) p<0.001 by one-way ANOVA, followed by Tukey multiple comparisons of means. (J) Scatter plot of Cdense (AU) versus surrounding cytoplasm (Cdilute, AU). (K) Scatter plot Cdense (AU) versus its total cytoplasmic vRNP concentration (Ccytoplasm, AU). (L) Boxplot showing Cdilute (AU); p=0.203 by one-way ANOVA followed by Tukey multiple comparisons of means. (M) Scatter plot of Cdilute (AU) versus Ccytoplasm (AU). (N) Scatter plot of fold change in free energy of partition (ΔΔG, J/mol) where ΔG = -RTInK, and K = (CdenseCdilute), and ΔΔG = ΔG – ΔG3 hpi, versus Cdilute (AU). (O) Boxplot of ΔΔG (J/mol); p<0.001 by one-way ANOVA followed by Tukey multiple comparisons of means. Changes in concentration of viral inclusions’ drivers do not impact their liquid profile Two factors were shown to drive the formation of IAV inclusions – vRNPs and Rab11a (Amorim et al., 2011; Eisfeld et al., 2011; Lakdawala et al., 2014; Vale-Costa et al., 2016; Alenquer et al., 2019; Veler et al., 2022). In fact, vRNP accumulation in liquid viral inclusions requires its association with Rab11a directly via the viral polymerase PB2 (Veler et al., 2022; Amorim et al., 2011), and the liquid character may be maintained by an incompletely understood network of intersegment interactions bridging several cognate vRNP-Rab11 units on flexible membranes (Vale-Costa et al., 2016). As the concentration of material is a key determinant for the physical properties of condensates (Riback et al., 2020; Weber and Brangwynne, 2015; Hernández-Vega et al., 2017), we evaluated how concentration of these two drivers impacts the behaviour of IAV inclusions. For this, we took advantage of the fact that vRNP levels increase during infection (Kawakami et al., 2011), and analysed viral inclusions over a time course, in two conditions: with endogenous levels of Rab11a (using cells expressing GFP, as in Alenquer et al., 2019), and overexpressing Rab11a (in the form of GFP-Rab11a, as in Alenquer et al., 2019; Figure 3A–B, Figure 3—figure supplement 2). With this approach, we aimed at analysing whether the material properties of viral inclusions changed over time and whether increasing the levels of Rab11 would alter these properties. This strategy would reveal if regulating Rab11a activity could harden IAV liquid inclusions. Figure 3 with 2 supplements see all Download asset Open asset Changes in concentration of viral ribonucleoproteins (vRNPs) and Ras-related in brain 11a (Rab11a) modestly alter the material properties of viral inclusions. (A–H) A549 cells stably expressing GFP, or GFP-Rab11a-WT were infected at an MOI of 3 with PR8 virus and, at the indicated timepoints, were fixed, and analysed by immunofluorescence using an antibody against nucleoprotein (NP) (as a proxy for vRNPs). (Number of cells (n)=14–18 for GFP and 19–25 for GFP-Rab11-WT). (C–H) Each dot is the average value of measured parameters per cell. Above each boxplot, same letters indicate no significant difference between them, while different letters indicate a statistical significance at α=0.05 using one-way ANOVA, followed by Tukey multiple comparisons of means for parametric analysis, or Kruskal Wallis Bonferroni treatment for non-parametric analysis. All thermodynamic related values are displayed in Supplementary file 1 (Sheets 2 and 3). Abbreviations: AU, arbitrary unit. (A) Representative depiction of the experimental analysis workflow. (B) Immunofluorescence images of infected cells at different hours post-infection (hpi) in cells overexpressing GFP (left) or GFP-Rab11 (right) (both in green); NP (red, as a proxy of vRNPs), and nucleus (blue). Scale bar = 10 µm. (C) Boxplot depicting the fold change in the ratio of cytoplasmic to nuclear vRNPs concentration at different times of infection, with endogenous or overexpressed Rab11a; p<0.001; Kruskal Wallis Bonferroni treatment. (D) Boxplot of mean inclusion area (µm2) per cell; p<0.001 by one-way ANOVA, followed by Tukey multiple comparisons of means. (E) Scatter plot showing nucleation density (ρ, µm–2) versus degree of supersaturation (S). (F) Boxplot of Cdense (AU); p<0.001 by Kruskal Wallis Bonferroni treatment. (G) Boxplot of Cdilute (AU); p<0.001 by Kruskal Wallis Bonferroni treatment. (H) Boxplot of ΔΔG (J/mol); p<0.001 by Kruskal Wallis Bonferroni treatment. Conditions were normalised to an infection state without IAV inclusions (3 hpi) that is indicated by the dashed black line. In GFP expressing cells, as the progeny vRNP pool reaches the cytosol (Figure 3B and C), viral inclusions augment in size (from 0.172±0.04 to 0.289±0.06 µm2, mean ± SD, Figure 3D), with similar aspect ratio (Figure 3—figure supplement 2A and B). There is a mild reduction in the number of inclusions from 8 hpi onwards, as measured by the nucleation density (ρ) (Figure 3E, Figure 3—figure supplement 2C, all topological data in Supplementary file 1 (Sheet 2)). As infection progresses, the concentration of vRNPs inside condensates increases until 8 hpi (Figure 3F and Figure 3—figure supplement 2D and E), accompanied by an increase in the diluted cytosolic phase (Figure 3G and Figure 3—figure supplement 2D and F, Supplementary file 1 (Sheet 2)), and both parameters stabilise thereafter, indicating that the critical concentration occurs around 8 hpi. This indicates that the liquid inclusions behave as a multi-component system and allow us to speculate that the differences in length, RNA sequence, and valency that each vRNP may be key for the integrity and behaviour of condensates. Importantly, Gibbs free energy (normalised to 3 hpi) is lowest at 6 hpi (–1799.0±623 J/mol) and destabilises mildly onwards (–1139.8±382, –1131.2±444, and –833.8±342 J/mol @ 8, 12, and 16 hpi, respectively) (Figure 3H, Figure 3—figure supplement 2G, H, Supplementary file 1 (Sheet 2)). These results are consistent with the increase in cytosolic vRNP leading to bigger sized inclusions that overall maintain the same concentration although becoming modestly destabilised, suggesting that the material properties are also modestly affected. When overexpressing Rab11a (right side of each graph), cytosolic vRNPs also accumulated in viral inclusions that increased with infection (Figure 3C–D, from 0.243±0.03 to 0.385±0.04 µm2), but were significantly bigger than viral inclusions in GFP expressing cells. In addition, the nucleation density was higher (Figure 3E and Figure 3—figure supplement 2C), despite having similar aspect ratio (Figure 3—figure supplement 2A and B), Cdense (Figure 3F and Figure 3—figure supplement 2D and E) and Cdilute (Figure 3G, Figure 3—figure supplement 2D and F). The lowest value of Gibbs free energy occurs at 8 hpi (–1337.7±331 J/mol) and destabilises from then onwards (–1145.3±443 and –895.3±394 J/mol @ 12 and 16 hpi, respectively, Figure 3H, Figure 3

In biological systems, liquid and solid-like biomolecular condensates may contain the same molecules but their behaviour, including movement, elasticity, and viscosity, is different on account of distinct physicochemical properties. As such, it is known that phase transitions affect the function of biological condensates and that material properties can be tuned by several factors including temperature, concentration, and valency. It is, however, unclear if some factors are more efficient than others at regulating their behaviour. Viral infections are good systems to address this question as they form condensates de novo as part of their replication programmes. Here, we used influenza A virus (IAV) liquid cytosolic condensates, AKA viral inclusions, to provide a proof of concept that liquid condensate hardening via changes in the valency of its components is more efficient than altering their concentration or the temperature of the cell. Liquid IAV inclusions may be hardened by targeting vRNP (viral ribonucleoprotein) interactions via the known NP (nucleoprotein) oligomerising molecule, nucleozin, both in vitro and in vivo without affecting host proteome abundance nor solubility. This study is a starting point for understanding how to pharmacologically modulate the material properties of IAV inclusions and may offer opportunities for alternative antiviral strategies.

Multiple blood-borne pathogens infecting mammals establish close interactions with the host vascular endothelium as part of their life cycles. In this work, we investigate differences in the interactions of three Trypanosoma species: T. brucei, T. congolense and T. vivax with the blood vasculature. Infection with these species results in vastly different pathologies, including different effects on vascular homeostasis, such as changes in vascular permeability and microhemorrhages. While all three species are extracellular parasites, T. congolense is strictly intravascular, while T. brucei is capable of surviving both extra- and intravascularly. Our knowledge regarding T. vivax tropism and its capacity of migration across the vascular endothelium is unknown. In this work, we show for the first time that T. vivax parasites sequester to the vascular endothelium of most organs, and that, like T. congolense, T. vivax Y486 is largely incapable of extravasation. Infection with this parasite species results in a unique effect on vascular endothelium receptors including general downregulation of ICAM1 and ESAM, and upregulation of VCAM1, CD36 and E-selectin. Our findings on the differences between the two sequestering species (T. congolense and T. vivax) and the non-sequestering, but extravasating, T. brucei raise important questions on the relevance of sequestration to the parasite’s survival in the mammalian host, and the evolutionary relevance of both sequestration and extravasation.

AbstractMultiple blood-borne pathogens infecting mammals establish close interactions with the host vascular endothelium as part of their life cycles. In this work, we investigate differences in the interactions of threeTrypanosomaspecies:T. brucei, T. congolenseandT. vivaxwith the blood vasculature. Infection with these species results in vastly different pathologies, including different effects on vascular homeostasis, such as changes in vascular permeability and microhemorrhages. While all three species are extracellular parasites,T. congolenseis strictly intravascular, whileT. bruceiis capable of surviving both extra- and intravascularly. Our knowledge regardingT. vivaxtropism and its capacity of migration across the vascular endothelium is unknown. In this work, we show for the first time thatT. vivaxparasites sequester to the vascular endothelium of most organs, and that, likeT. congolense, T. vivaxY486 is largely incapable of extravasation. Infection with this parasite species results in a unique effect on vascular endothelium receptors including general downregulation of ICAM1 and ESAM, and upregulation of VCAM1, CD36 and E-selectin. Our findings on the differences between the two sequestering species (T. congolenseandT. vivax) and the non-sequestering, but extravasating,T. bruceiraise important questions on the relevance of sequestration to the parasite’s survival in the mammalian host, and the evolutionary relevance of both sequestration and extravasation.

SummaryTrypanosoma bruceiis responsible for lethal diseases in humans and cattle in Sub-Saharan Africa. These extracellular parasites extravasate from the blood circulation into several tissues. The importance of the vasculature in tissue tropism is poorly understood. Using intravital imaging and bioluminescence, we found that gonadal white adipose tissue and pancreas are the two main parasite reservoirs. We show that reservoir establishment happens before vascular permeability is compromised, suggesting that extravasation is an active mechanism. Blocking endothelial surface adhesion molecules (E-selectin, P-selectins, or ICAM2) significantly reduced extravascular parasite load in all organs and delayed host lethality. Remarkably, blocking CD36 had a specific effect on adipose tissue tropism that was sufficient to delay lethality, suggesting that establishment of the adipose tissue reservoir is necessary for parasite virulence. This works demonstrates the importance of the vasculature in aT. bruceiinfection and identifies organ-specific adhesion molecules as key players for tissue tropism.

Trypanosoma brucei is responsible for lethal diseases in humans and cattle in Sub-Saharan Africa. These extracellular parasites extravasate from the blood circulation into several tissues. The importance of the vasculature in tissue tropism is poorly understood. Using intravital imaging and bioluminescence, we found that gonadal white adipose tissue and pancreas are the two main parasite reservoirs. We show that reservoir establishment happens before vascular permeability is compromised, suggesting that extravasation is an active mechanism. Blocking endothelial surface adhesion molecules (E-selectin, P-selectins, or ICAM2) significantly reduced extravascular parasite load in all organs and delayed host lethality. Remarkably, blocking CD36 had a specific effect on adipose tissue tropism that was sufficient to delay lethality, suggesting that establishment of the adipose tissue reservoir is necessary for parasite virulence. This works demonstrates the importance of the vasculature in a T. brucei infection and identifies organ-specific adhesion molecules as key players for tissue tropism.

Plasmodium parasites, causative agents of malaria, scavenge host nutrients to sustain their intracellular replication. Modulation of the host's nutritional status can potentially help control infection by limiting the parasite's access to nutrients, or by boosting the immune system. Here, we show that dietary supplementation of mice employing a combination of arginine (R) with two additional amino acids, lysine (K) and valine (V), termed RKV, significantly decreases Plasmodium liver infection. RKV supplementation results in the elimination of parasites at a late stage of their development in the liver. Our data employing genetic knockout mouse models and in vivo depletion of specific cell populations suggest that RKV supplementation boosts the host's overall innate immune response, and that parasite elimination is dependent on MyD88 signaling in immune cells. The immunostimulatory effect of RKV supplementation opens a potential role for dietary supplementation as an adjuvant for prophylaxis or immunization strategies against Plasmodium infection.

Iron refractory iron deficiency anemia (IRIDA) is an autosomal recessive ferropenic anemia. Its hypochromic microcytic pattern is associated with low transferrin saturation, normal-high ferritin, and inappropriately high hepcidin level. This entity is caused by mutants of the TMPRSS6 gene that encodes the protein matriptase II, which influences hepcidin expression, an iron metabolism counterregulatory protein. We report two 29-year-old dizygotic female twins with ferropenic, hypochromic microcytic anemia with 20 years of evolution, refractory to oral iron therapy. After exclusion of gastrointestinal etiologies, IRIDA diagnosis was suspected and a novel mutation in the TMPRSS6 gene was identified. It was found in intron 11 (c.1396+4 A>T) and seems to affect the gene expression. In addition, 3 polymorphisms already associated with a higher risk of developing iron deficiency anemia were also found (D521D, V736A, and Y739Y). Our case reports an undescribed mutation causing IRIDA and supports the hypothesis that this clinical syndrome may be more common than previously thought and its genetics more heterogeneous than initially described.