Insight into the biochemical and cell biological function of an intrinsically unstructured heat shock protein, Hsp12 of Ustilago maydis

Abstract

Small heat shock proteins (sHsps) play diverse roles in the stress response and maintenance of cellular functions. The Ustilago maydis genome codes for few sHsps. Among these, Hsp12 has previously been demonstrated to be involved in the pathogenesis of the fungus by our group. In the present study we further investigated the biological function of the protein in the pathogenic development of U. maydis. Analysis of the primary amino acid sequence of Hsp12 in combination with spectroscopic methods to analyse secondary protein structures revealed an intrinsically disordered nature of the protein. We also carried out detailed analysis on the protein aggregation prevention activity associated with Hsp12. Our data suggest Hsp12 has trehalose‐dependent protein aggregation prevention activity. Through assaying the interaction of Hsp12 with lipid membranes in vitro we also showed the ability of U. maydis Hsp12 to induce stability in lipid vesicles. U. maydis hsp12 deletion mutants exhibited defects in the endocytosis process and delayed completion of the pathogenic life cycle. Therefore, U. maydis Hsp12 contributes to the pathogenic development of the fungus through its ability to relieve proteotoxic stress during infection as well as its membrane‐stabilizing function.

The involvement of Hsp12 from Ustilago maydis in prevention of protein aggregate formation, stabilization of artificial lipid membranes in vitro, and endocytic processes in U. maydis is shown.

1. INTRODUCTION

Heat shock proteins (Hsps) constitute important cellular components involved in maintaining intracellular protein homeostasis (Lindquist & Craig, 1988). Accumulation of unfolded or misfolded proteins within a cell leads to damage due to the irreversible formation of protein aggregates. Hsps therefore function as molecular chaperones and contribute to protein folding in either energy‐dependent or ‐independent ways (Richter et al., 2010), thereby relieving the host cells of the load of misfolded proteins. Unlike higher‐molecular‐weight Hsps, small Hsps (sHsps) are incapable of refolding a protein on their own owing to the lack of an ATPase domain. Nevertheless, sHsps can associate with partially misfolded proteins to keep them in a folding‐compatible state (Lee, 1997; Ungelenk et al., 2016). At this stage the misfolded proteins are either transferred to ATP‐dependent Hsps that carry out the refolding process (Ehrnsperger, 1997; Haslbeck et al., 2005) or are degraded through autophagy or the proteasome pathway (Carra et al., 2014; Parcellier et al., 2003, 2006). The human genome has been shown to code for 10 sHsps, each of which contains a conserved stretch of 80 amino acid residues that constitute the α‐crystallin domain (ACD) (Kappé et al., 2003). ACDs are therefore considered to be one of the characteristic features of sHsps. ACDs have also been demonstrated to be essential and sufficient for the oligomerization of sHsps (Jehle et al., 2011). However, despite this initial characterization of sHsps as proteins with ACDs, some of the sHsps from different organisms have since been found to lack ACDs. Saccharomyces cerevisiae Hsp12 is one such sHsp that lacks an ACD (Welker et al., 2010). Human Hsp16.2 is another example of a protein that exhibits inducible expression at elevated temperatures, chaperone‐like properties, and small monomer size similar to sHsps but, unlike conventional sHsps, it lacks an ACD (Bellyei et al., 2006; Kappé et al., 2009). Besides contributing to protein folding, sHsps have also been demonstrated to regulate the structural integrity of biological membranes through their interaction with membrane lipids. For instance, Synechocystis Hsp17 and α‐crystallin exert stabilizing effects on model membranes formed of synthetic and cyanobacterial lipids (Török et al., 2001). Likewise, in the lactic acid bacterium Oenococcus oeni the sHsp Lo18 has been shown to contribute to maintaining order in the lipid bilayer during heat stress (Coucheney et al., 2005). A recent study demonstrated that HspB1, an sHsp from mammals, functions as a membrane chaperone. In that study, HspB1 was found to alter both the rotational and the lateral mobility of the lipids in the interacting lipid membranes (Csoboz et al., 2022). sHsps lacking an ACD, like human Hsp16.2, have also been demonstrated to stabilize lipid rafts through involvement of Hsp90 (Bellyei et al., 2006). Although not much is yet known about the detailed mechanisms that facilitate membrane interaction of sHsps, few studies indicate correlations between the oligomeric status and the membrane insertion abilities of some of the sHsps. For instance, HspB5 oligomerizes at higher temperature, which aids its insertion into the membranes of vertebrate lenses (Tjondro et al., 2016). In a recent study, HspB1 and HspB5 were shown to be inserted into liposomes primarily through their respective ACDs (De Maio et al., 2019). HspB5 exhibits association with diverse membranes, including lenses, mitochondria, and Golgi bodies (Boyle & Takemoto, 2009; Gangalum & Bhat, 2009; Whittaker et al., 2009). Evidence of HspB5 participation in exosome assembly and release has also been obtained in some instances (Gangalum et al., 2016). Taken together, these studies revealed the ability of sHsps to modulate membrane lipid polymorphism. In the present study we have investigated the biological function of Hsp12 from the biotrophic plant‐pathogenic fungus Ustilago maydis. Hsp12, unlike the human HspB family of sHsps, does not possess an ACD but exhibits signatures of an intrinsically disordered protein (IDP). In a previous study we showed the contribution of Hsp12 to the pathogenicity of U. maydis (Ghosh, 2014). However, the nature of the involvement of the protein in the pathogenic life cycle of the fungus remained unexplored. Here we demonstrate the role of Hsp12 in the prevention of stress‐induced protein aggregation in the presence of trehalose. In addition, the protein has also been found to have stabilizing effects on unilamellar vesicles made of amphipathic lipids. In this study we also demonstrate for the first time the involvement of a fungal sHsp in the endomembrane trafficking system of the fungus. Our data revealed the ability of Hsp12 to fragment larger lipid vesicles into smaller ones upon its association. However, further investigation is needed to decipher any link between the membrane interaction properties of Hsp12 and its role in the proper functioning of the endomembrane trafficking system of the fungus.

2. RESULTS

2.1. Hsp12 shows features of an intrinsically disordered protein

Hsp12 is an 80‐amino‐acid residue protein in U. maydis with an N‐terminal Hsp9‐Hsp12 domain (Pfam ID: PF04119) (Figure 1a). Hsp9‐Hsp12 domains are present in Hsps that are strongly expressed during the stationary phase of growth in S. cerevisiae cells (Orlandi et al., 1996; Praekelt & Meacock, 1990). In silico bioinformatics analysis of the amino acid composition of Hsp12 revealed a number of features suggesting a intrinsically disordered nature of Hsp12. IDPs have been shown to exhibit a biased amino acid composition with increased abundance of disorder‐producing amino acids like alanine, arginine, glycine, glutamine, serine, proline, glutamic acid, and lysine (Dunker et al., 2001). Likewise, in Hsp12 more than 50% of the amino acid residues within the primary sequence belong to the disorder‐producing amino acid category (Figure S1). Yet, when a structure‐based amino acid sequence alignment of the protein was carried out in ClustalW and EsPript 3.0 using Hsp12 sequences from closely related fungi and the S. cerevisiae Hsp12 structure as the template, four conserved helical regions were predicted in U. maydis Hsp12 (Figure 1b). We next analysed the topographical distribution of the polar and nonpolar amino acid residues within each of the predicted helices of U. maydis Hsp12 using the helical wheel projection method of protein structure representation (Schiffer & Edmundson, 1967). Interestingly, three out of the four predicted helices of Hsp12 showed an amphipathic nature, with the hydrophobic and hydrophilic residues clustered on the opposite sides of the helices (Figure 1c). Previously, the S. cerevisiae Hsp12 secondary structure also showed similar amphipathic helices but only in the presence of the hydrophobic environment of certain lipids. In the absence of a hydrophobic environment, the protein was found to be completely unfolded (Welker et al., 2010). To further assess the structural properties of U. maydis Hsp12, the secondary structure associated with the protein was assessed through circular dichroism (CD) analysis of purified recombinant Hsp12 with a 6×His tag at the N‐terminus. The recombinant Hsp12 was expressed and purified from Escherichia coli BL21 (DE3)‐RIL cells using Ni‐NTA affinity chromatography to achieve more than 90% purity (Figure 2a). The CD spectrum of purified recombinant Hsp12 primarily revealed a random coil structure associated with the protein. However, upon introduction of a hydrophobic environment through the addition of 8 mM SDS, α‐helical structures were observed (Figure 2b). Deconvolution of the CD data revealed up to 55.3% helix in the protein in the presence of 8 mM SDS, as opposed to only about 12.8% in the absence of any hydrophobic agent. This indicated that like S. cerevisiae Hsp12, U. maydis Hsp12 also attains structure in the presence of a hydrophobic environment. We further investigated the spatial orientation of the amphipathic helices of Hsp12 by analysing the changes in the fluorescence properties of 8‐anilinonaphthalene‐1‐sulfonic acid (ANS) when added to the protein. ANS is a fluorophore that selectively interacts with hydrophobic patches, resulting in an increase in its fluorescence intensity (Gasymov & Glasgow, 2007). In the presence of increasing concentrations of SDS, the fluorescence intensity of ANS increased significantly, indicating structural perturbation in Hsp12 leading to surface exposure of the hydrophobic patches associated with the helices (Figure 2c).

FIGURE 1.

Structural features evident from the amino acid sequence of Hsp12. (a) Domain architecture of Hsp12 as obtained from SMART domain analysis of the primary amino acid sequence of the protein. The N‐terminal 53 amino acids comprise the Hsp9‐Hsp12 domain. (b) Structure‐based amino acid sequence alignment of Hsp12 with Hsp12 homologues from closely related fungi using ClustalW and EsPript 3.0. α1–α4 represent four possible α‐helical regions and ƞ1 represents a 3₁₀ helix within the Hsp12 amino acid sequence. The NCBI accession numbers of the Hsp12 homologues are Moesziomyces antarcticus (XP_014659768.1), Melanopsichium pennsylvanicum (CDI51149.1), Chichorium endivia (KAI3477947.1), Testicularia cyperi (PWZ02823.1), Sporisorium graminicola (XP_029742406.1), Pseudozyma hubeiensis SY62 (XP_012190373.1), Kalmanozyma brasiliensis GHG001 (XP_016294819.1), Pseudozyma flocculosa PF1 (XP_007876336.1), Thecafora frezii (UOP57153.1), and Saccharomyces cerevisiae (NP_116640.1). (c) Helical wheel projections of the three helices, helix 1, helix 2, and helix 4, of the Hsp12 secondary structure created through NetWheels. The amino acids are colour‐coded as per their chemical nature. Red represents basic, blue represents acidic, green represents uncharged, and yellow represents nonpolar amino acids.

FIGURE 2.

SDS‐induced changes in the secondary structure of Hsp12. (a) Purification of recombinant Hsp12 from Escherichia coli BL21 (DE3)‐RIL using Ni‐NTA affinity purification. Lane 1: protein molecular weight marker; lane 2: total lysate from uninduced E. coli cells; lane 3: total lysate from E. coli cells induced with 500 μM isopropyl β‐d‐thiogalactopyranoside; lane 4: total protein from the soluble fraction of induced E. coli; lane 5: flowthrough after binding; lanes 6 & 7: column wash 1 and wash 2 with 10 mM and 20 mM imidazole, respectively; lane 8: purified Hsp12 eluted with 100 mM imidazole. (b) Circular dichroism spectra of purified recombinant Hsp12 in the absence (brown) and presence (orange) of 8 mM SDS. (c) Graph showing the fluorescence intensity over a wavelength range of 400 to 600 nm of the fluorophore 8‐anilinonaphthalene‐1‐sulfonic acid (ANS) when added either to Hsp12 in the presence of indicated concentrations of SDS or to SDS alone. The fluorescence values at each data point represent the average calculated from three independent biological replicates for the experiment.

2.2. Hsp12 exerts protein aggregation prevention activity

sHsps commonly exhibit protein aggregation prevention activities (Bakthisaran et al., 2015). Therefore, in order to test whether U. maydis Hsp12 can also contribute to the prevention of protein aggregation, a nonspecific substrate, lysozyme, was treated with dithiothreitol (DTT) in the presence or absence of Hsp12 and separated into soluble and insoluble fractions. In the absence of Hsp12 a major population of the lysozyme was detected within the insoluble fraction, indicating DTT‐induced denaturation (Figure 3a). Most interestingly, the presence of Hsp12 alone could not rescue the DTT‐induced aggregation of lysozyme. Instead, it enhanced the extent of the aggregation process and itself was found in the pellet fraction (Figure 3a). In previous studies, trehalose has been shown to have a positive influence on the aggregation prevention activities of Hsps (Viner & Clegg, 2001). We therefore tested whether Hsp12 can exhibit protein aggregation prevention activity in the presence of trehalose. Surprisingly, trehalose indeed induced the protein aggregation prevention activity of Hsp12, and almost equal proportions of lysozyme were detected in the soluble and insoluble fractions (Figure 3a). To further confirm the aggregation prevention activity of Hsp12 on DTT‐treated lysozyme, the protein aggregates were assessed spectrophotometrically by measuring the UV light scattering at a wavelength of 360 nm. Lysozyme treated with DTT when assessed in this way showed a stable increase in scattering intensity over the entire time range studied (Figure 3b). However, the presence of Hsp12 and trehalose significantly reduced the scattering, confirming that Hsp12 can prevent protein aggregation in the presence of trehalose. Neither Hsp12 nor trehalose was found to exert protection against DTT‐induced aggregation of lysozyme independently. We further tested any role of Hsp12 in providing protection against protein denaturation stress within U. maydis sporidial cells. The expression of hsp12 in SG200 wild‐type (WT) cells when treated with DTT was increased to about 20‐fold compared to untreated conditions (Figure 3c). The induced expression of hsp12 upon exposure of U. maydis cells to DTT indicated an important role of Hsp12 in regulating protein denaturation stress in U. maydis. Accordingly, SG200∆hsp12 showed reduced growth in the presence of DTT compared to the SG200 WT strain. However, the observed growth defect was fully restored upon expression of a functionally active hsp12 gene under the control of a constitutive promotor (Figure 3d).

FIGURE 3.

Protein aggregation prevention activity associated with Hsp12. (a) Coomassie brilliant blue‐stained SDS‐PAGE gel showing the effect of Hsp12 on the distribution of lysozyme into soluble and insoluble fractions upon treatment with 20 mM dithiothreitol (DTT) either in the absence (left panel) or in the presence (right panel) of trehalose. In both panels, lane 1 shows the protein molecular weight marker (kDa), lanes 2 and 3 represent soluble (S) and pellet (P) fractions from the reaction mix with Hsp12 alone, lanes 4 and 5 represent soluble and pellet fractions from the reaction mix with lysozyme alone, and lanes 6 and 7 represent soluble and insoluble fractions from the reaction containing both Hsp12 and lysozyme. Mn‐SOD from Bacillus aryabhattai was used as an internal control. (b) Left panel: graph showing the scattering of UV light at 360 nm over a time span of 3500 s representing the levels of insoluble protein aggregates formed during treatment of lysozyme with DTT in the absence of Hsp12 (brown), in the presence of Hsp12 alone (orange), in the presence of Hsp12 and trehalose (yellow), and in the presence of trehalose alone (green). In addition, control reactions including lysozyme without DTT treatment (light blue), Hsp12 in the presence of trehalose and DTT (blue), and Hsp12 in the presence of DTT alone (purple) are also shown in the graph. The scattering values at each data point represent the average calculated from three independent biological replicates for the experiment. Error bars represent standard error. Right panel: region of the scattering graph (left panel) showing the controls with increased resolution through stretching the y axis. (c) Graph showing the expression of hsp12 in SG200 wild‐type (WT) cells in the absence and presence of 5 mM DTT relative to the expression of ppi, which was used as a constitutive expression control. The y axis presents relative expression of hsp12 with respect to ppi in logarithmic scale. Data are presented as the average calculated from three independent biological experiments. Error bars represent standard error. Statistical difference was examined by paired two‐tailed t test (*p < 0.05). (d) Growth of Ustilago maydis SG200 WT (upper row), the hsp12 deletion mutant (middle row), and the respective complementation strain (lower panel) in complete medium (CM) without (left panel) or with (right panel) 5 mM DTT. The adjacent columns in each of the panels represent serial dilutions of the axenic cultures of each of the strains in the order of 10.

2.3. Hsp12 interacts with and stabilizes lipid membranes

Hsps, including the sHsps, have previously been demonstrated to interact with biological membranes through binding to the membrane lipids (De Maio & Hightower, 2021). These interactions have also been shown to stabilize membrane structures. In order to test any possible role of Hsp12 in stabilizing lipid membranes, the interaction of Hsp12 with large unilamellar vesicles (LUVs) formed with 1,2‐dimyristoyl‐sn‐glycerol‐3‐phospho‐(1′‐rac‐glycerol) (DMPG) lipids was studied. Anisotropy measurements of 1,6‐diphenyl‐1,3,5‐hexatriene (DPH) in solution containing DMPG vesicles and Hsp12 revealed that Hsp12 associates with the lipid vesicles in a cooperative manner. Furthermore, unlike the positive effect on the protein aggregation prevention activity of Hsp12, trehalose showed nearly no effect on the binding of the protein with DMPG lipid vesicles (Figure 4a). Therefore, the interaction of Hsp12 with lipids is independent of the presence of trehalose. Besides, the increase in the anisotropy value of DPH with increasing concentrations of Hsp12 also demonstrated a positive effect of Hsp12 on the stabilization of the vesicle structure (Figure 4b). We also tested the ability of Hsp12 to resist a temperature‐induced perturbation of the vesicle structures. The anisotropy values of DPH were reduced significantly upon an increase in temperature irrespective of the presence of Hsp12 in the solution. However, the degree of reduction differed significantly. While in the absence of Hsp12 the anisotropy value reduced to 30% at 25°C, upon a temperature rise of 5°C, in the presence of Hsp12 the value reduced to only 60% (Figure 4c). This showed that Hsp12 can indeed protect against the loss of membrane rigidity at increased temperatures. Also, like lipid binding, structural stabilization of lipid vesicles by Hsp12 was independent of the presence of trehalose (Figure 4c). Likewise, the fluorescence intensity of ANS also increased with increasing concentrations of LUV in a solution containing Hsp12, indicating the surface exposure of the hydrophobic patches associated with the protein (Figure 4d). Hsp12 thus undergoes a transition from an unstructured to a structured conformation in the presence of a hydrophobic environment created through either SDS or lipids. Following on from the in vitro studies, we next investigated the ability of Hsp12 to associate with biological membranes in U. maydis cells. U. maydis SG200∆hsp12‐hsp12‐ilov‐HA strains were generated that express Hsp12 with a C‐terminal fusion of the LOV (light, oxygen, or voltage) domain‐based fluorescent protein named iLOV (Chapman et al., 2008) under the control of a constitutive otef promoter. A significant proportion of the iLOV fluorescence in SG200∆hsp12‐hsp12‐ilov‐HA was detected within the periphery of sporidial cells, indicating association with the plasma membrane (Figure 4e). A similar distribution of the fluorophore was also obtained in the filamentous cells of the pathogen during in planta growth of the fungus, further confirming the plasma membrane association of Hsp12 within U. maydis cells (Figure 4f).

FIGURE 4.

Interaction of Hsp12 with lipid membranes. (a) Graphs showing the titration of 20 μM 1,2‐dimyristoyl‐sn‐glycerol‐3‐phospho‐(1′‐rac‐glycerol) (DMPG) large unilamellar vesicles (LUVs) with increasing concentrations of Hsp12 ranging from 0 to 5000 nM either in the absence (left panel) or in the presence (right panel) of trehalose. (b) Graphs showing the anisotropy values of 1,6‐diphenyl‐1,3,5‐hexatriene (DPH) associated with DMPG LUVs when treated with indicated concentrations of Hsp12 either in the absence (left) or in the presence (right) of trehalose. (c) Graph showing the percent anisotropy values of DPH associated with DMPG LUVs with increasing temperature. The anisotropy at 20°C was considered as 100%. Graphs representing different reaction compositions are colour‐coded. Brown represents LUVs alone, orange represents LUVs in the presence of Hsp12, yellow represents LUVs in the presence of both Hsp12 and trehalose, and green represents LUVs in the presence of trehalose alone. (d) Graph showing the fluorescence intensity scan associated with ANS over a wavelength range of 400 to 600 nm when added to Hsp12 in the presence of increasing concentrations of DMPG LUVs as indicated in the colour codes. (e, f) Confocal microscopy images showing the distribution of Hsp12‐iLOV‐HA in Ustilago maydis SG200Δhsp12‐hsp12‐iLOV‐HA sporidia (e) and filament (f). In each figure the left panel represents iLOV fluorescence, the middle panel represents the differential interference contrast (DIC) image, and the right panel represents the merge of iLOV fluorescence and the DIC image. In (a), (b), (c), and (d), data are presented as the average calculated from three independent biological experiments. Error bars represent standard error.

2.4. Hsp12 interaction leads to vesiculation of LUVs into smaller vesicles

In a previous study, S. cerevisiae Hsp12 has been demonstrated to exert membrane remodelling functions whereby large liposomes are vesiculated into smaller ones (Kim et al., 2018). Because U. maydis Hsp12 also exhibits membrane interaction, we tested whether the protein fragments LUVs into smaller vesicles. The diameters of the LUVs upon treatment with increasing concentrations of Hsp12 were measured using dynamic light scattering (DLS) and plotted against the percentage of the total population with a specific size for each individual treatment. Compared to untreated vesicles, which showed a predominance of LUVs with a diameter of around 100 nm, the Hsp12‐treated vesicles were found to be much smaller in size. The most prominent effect was observed at 5 μM Hsp12, where the diameters of the majority of the LUVs were reduced to about 20 nm (Figure 5a). On the contrary, when the negative control bovine serum albumin was used to treat LUVs, no significant reduction in vesicular size was observed. Transmission electron microscopy images of the LUVs, either untreated or following treatment with 5 μM Hsp12, further confirmed the sizes of the vesicles (Figure 5b,c). These observations indicate that the association of Hsp12 with lipids is linked to membrane vesiculation in U. maydis.

FIGURE 5.

Hsp12 induced vesiculation of 1,2‐dimyristoyl‐sn‐glycerol‐3‐phospho‐(1′‐rac‐glycerol) (DMPG) large unilamellar vesicles (LUVs). (a) Graph showing the hydrodynamic size of the DMPG LUVs when treated with increasing concentrations of Hsp12 as determined by dynamic light scattering (DLS). The diameters of the LUVs at different concentrations of Hsp12 and bovine serum albumin (BSA) as a negative control are plotted against the percentage of the LUVs representing the size distribution at each of the Hsp12 concentrations. (b) Transmission electron microscopy image of DMPG LUVs in the absence (left panel) and presence (right panel) of Hsp12. Size bars represent 50 nm. (c) Graph showing the diameters of the LUVs when treated with or without Hsp12 as calculated from transmission electron microscopy images. At least 60 LUV images were analysed for the calculation of the average diameter, which is plotted on the y axis. The red bar represents the mean diameter of liposomes. Statistical difference was examined by paired two‐tailed t test (****p < 0.0001).

2.5. Hsp12 influences endocytosis in U. maydis

One of the key steps in the process of endocytosis‐mediated vesicular trafficking is the formation of endocytic vesicles. Membrane curvature‐inducing proteins play key roles in the formation of such vesicles. Because of its ability to modulate membrane stability and induce membrane vesiculation, we evaluated whether Hsp12 is involved in the endocytosis processes of U. maydis. Accordingly, the generation and trafficking of endocytic vesicles in U. maydis SG200 WT and SG200∆hsp12 were followed using FM4‐64. FM4‐64 is a styryl dye that initially stains the plasma membrane by intercalating between the constituent lipids (Vida & Emr, 1995). As time progresses, the fluorophore can be visualized within the membranes of the endocytic vesicles, thereby facilitating monitoring of the endocytic pathway. We stained both SG200 WT and SG200∆hsp12 with FM4‐64 for 5 min and chased the fluorescence for another 45 min to compare the functioning of the endocytic pathway between the two strains. The endocytic vesicles in both the WT and the mutant strains were stained similarly throughout the chase period. Interestingly, a significant difference was noted in the fluorescence that remained associated with the plasma membrane of the two strains as early as after 5 min. Unlike SG200 WT, in SG200∆hsp12 the FM4‐64 stain within the plasma membrane was clearly visible. The difference became even more prominent at later time points of the chase experiment (Figure 6a), when the fluorescence within the plasma membranes of SG200 WT was barely visible. The ratio of the fluorescence intensities within the plasma membrane to that within the vesicular membrane in the stained cells was also calculated to estimate the relative efficiencies of the endocytosis processes in SG200 WT and SG200∆hsp12 strains. Compared to SG200 WT, endocytosis was found to be about 5 times less efficient in SG200∆hsp12 (Figure 6b).

FIGURE 6.

Contribution of Hsp12 to the endocytosis process of Ustilago maydis. (a) Confocal microscopy images showing the uptake of FM4‐64 by either SG200Δhsp12 (left) or SG200 wild‐type (WT) (right) following an initial incubation with the dye for 5 min and subsequent chase periods of 5 min (upper panel), 15 min (middle panel), and 45 min (lower panel). Size bars represent 5 μm. (b) Graph showing the relative intensities of fluorescence associated with FM4‐64 in the plasma membrane with respect to the vacuolar membrane calculated from the confocal microscopy images of SG200 WT and SG200Δhsp12 with FM4‐64. Images of at least 20 cells in each category were analysed for the calculation of the average relative fluorescence intensities. Error bars represent standard error. Statistical differences were examined by unpaired Student's t test (**p < 0.01, ***p < 0.001, ****p < 0.0001).

2.6. Pathogenic development of the hsp12 deletion mutant is significantly delayed

We have previously shown that the hsp12 deletion mutant of U. maydis exhibits reduced pathogenicity (Ghosh, 2014) with significantly compromised vigour of infection. Any defect associated with the progression of infection involving transition from one developmental stage to the other in this strain leading to the observed reduced pathogenicity was not investigated. In the light of the present study, which shows a significant role of Hsp12 in stabilizing biological membranes and its involvement in the endocytosis process of the fungus, we compared the pathogenic development of SG200∆hsp12 with that of SG200 WT cells. Infected maize tissue was stained with both wheat germ agglutinin‐Alexa Fluor 488 (WGA‐AF) and propidium iodide (PI) to visualize the fungal and the plant cell walls, respectively. Confocal microscopy analysis of the stained tissue showed that like SG200 WT, SG200∆hsp12 attained all the stages of development. However, the rate of disease progression was much reduced in the hsp12 mutant strain compared to the WT strain. While SG200 WT showed mature spore formation at 10 days postinfection (dpi), in the case of SG200∆hsp12 earliest traces of mature spores could be noticed only at 16 dpi (Figure 7). While in the case of SG200 WT abundant small black round spores could be noticed at 16 dpi, in the case of SG200Δhsp12 very few and immature spores could be observed (Figure 7). This indicated that the absence of Hsp12 in U. maydis led to a significant delay in the completion of the pathogenic life cycle of the fungus, which was marked by the formation of mature spores.

FIGURE 7.

Confocal microscopy images of Zea mays infected with either SG200Δhsp12 (upper panel) or SG200 wild type (WT) (lower panel). The infected leaf tissues were stained with WGA‐Alexa Fluor‐488, which stains the fungal cell wall, and propidium iodide (PI), which stains the plant cell wall. The infection was assessed at different time points (indicated as days postinfection [dpi]). The differential interference contrast (DIC) images of 16 dpi tissue show teliospores as small round black bodies. Size bars represent 50 μm.

3. DISCUSSION

In contrast to the ACD, sHsps exhibit a predominantly disordered structure (White et al., 2006). Accordingly, U. maydis Hsp12 also showed characteristics of an IDP. However, in the presence of SDS the protein could attain a more structured conformation. This has previously been shown for other IDPs as well. For instance, a study of the conformational changes together with solvation dynamics of bovine β‐casein in the presence of SDS demonstrated that SDS‐induced crowding led to coil‐to‐helix transition of the protein (Mubashira et al., 2022). The U. maydis Hsp12 amino acid sequence, when arranged in the form of an α‐helix, could form four amphipathic helices. We therefore propose that the presence of the hydrophobic environment of SDS is the guiding force for further folding of Hsp12 into α‐helical segments. When we tested whether any aggregation prevention activity is associated with Hsp12, initially it was found that the protein could not prevent the aggregation of lysozyme upon treatment with DTT. At this point we considered evaluating the effects of organic solutes and osmolytes that are synthesized in large amounts during cell stress (Somero & Yancey, 1997) on the protein aggregation prevention activity of Hsp12. Some of these solutes and osmolytes are involved in stabilizing native proteins (Hottiger et al., 1994). The disaccharide trehalose is one such organic osmolyte that has previously been demonstrated to be involved in stress resistance and virulence in U. maydis (Cervantes‐Chávez et al., 2016; Salmerón‐Santiago et al., 2011). When trehalose was included in the experiment, significant levels of aggregation prevention activity of Hsp12 were noted after DTT‐induced denaturation of lysozyme. The underlying mechanism explaining the contribution of trehalose to the functioning of sHsps is currently being studied. Several hypotheses have been proposed to date. According to one of the hypotheses, trehalose is a release factor that hinders the formation of a stable complex between the substrate and sHsp. It even helps in releasing the folding intermediates of the substrate from the complex (Viner & Clegg, 2001). In the case of U. maydis Hsp12, we found that in the presence of trehalose a significant proportion of lysozyme can remain in the soluble fraction together with Hsp12 under the DTT‐induced denaturation conditions. It is therefore quite likely that trehalose can provide an environment that is favourable for the interaction of Hsp12 with lysozyme, so that most of the complex remains soluble and does not form aggregates. In the case of plant dehydrins, which are a group of IDPs with chaperone activities, it was proposed that a part of the disordered structure of the proteins helps to prevent collision among the substrate molecules by entropically filling the intermolecular space within the substrate (Tunnacliffe & Wise, 2007). This accounts for the associated chaperone function of the dehydrins. In the case of U. maydis Hsp12, it might happen that trehalose together with Hsp12 acts in a similar way, thereby minimizing collision between the misfolded substrate molecules and preventing aggregation.

U. maydis undergoes a number of morphological changes during its life cycle (Banuett & Herskowitz, 1996). Most of these developmental stages, including appressorium formation, filament fragmentation, and sporulation, involve extensive modulations to the plasma membrane structure of the pathogen. It is therefore most likely that U. maydis has an elaborate mechanism to stabilize its plasma membrane integrity. Our observation of the membrane‐stabilizing effects of Hsp12 on DMPG lipids in vitro prompted us to evaluate whether Hsp12 plays a part in this mechanism. Accordingly, we investigated any involvement of Hsp12 in the pathogenic development of the fungus. Although the deletion mutant displayed all the developmental stages, the relative time taken was significantly longer than for the WT cells. This indicates a compromised efficiency in the progression of the pathogenic life cycle of U. maydis in the absence of Hsp12. Besides influencing the morphological development of U. maydis, Hsp12 also showed involvement in the vesicular trafficking processes of the fungus. Involvement of Hsps in endomembrane trafficking has been evidenced earlier. Hsp70 has been demonstrated to increase the rate of endocytosis of transferrin and its receptor in human hepatoblastoma cells during heat stress (Vega et al., 2009). It has also been shown to accelerate the phagocytic process in macrophages (Vega & De Maio, 2005). Hsp70 has also been demonstrated to interact with lysosomal membranes. This interaction has been shown to stabilize the lysosomes, thereby aiding in the reversion of the lysosomal pathology associated with Niemann–Pick disease (Kirkegaard et al., 2010). An sHsp from human, HspB5, has also been shown to be associated with a variety of membranes including Golgi bodies (Gangalum et al., 2004; Gangalum & Bhat, 2009). Therefore, the reduced efficiency of endocytosis observed in SG200Δhsp12 points towards a possible role of Hsp12 in the stabilization of the membrane vesicles generated during trafficking processes. In U. maydis endocytosis plays a major role during polar growth and following infection (Wedlich‐Soldner, 2000). It has also been demonstrated to be involved in the pheromone perception and in the formation of conjugation hyphae leading to a successful mating process linked to the pathogenic development of the fungus (Fuchs et al., 2006). Therefore, smooth functioning of the endomembrane trafficking system in U. maydis is a prerequisite for the proper development of the fungus. Considering a possible role of Hsp12 in this process, improper functioning of the endomembrane trafficking system can also contribute to the delayed life cycle of U. maydis that is observed in the absence of Hsp12. Further detailed studies focusing on the endomembrane trafficking system in relation to Hsp12 is necessary to decipher the nature of the function of Hsp12 in this process.

4. EXPERIMENTAL PROCEDURES

4.1. Strains

U. maydis SG200 (Kämper et al., 2006) and its derivatives were used throughout the study. For cloning and expression of recombinant Hsp12, E. coli XL1‐Blue and BL21 (DE3)‐RIL were used, respectively. Plant experiments were carried out on Zea mays ‘Early Golden Bantam’.

4.2. U. maydis growth conditions

Axenic cultures of U. maydis sporidia were grown at 28°C on a rotary shaker (200 rpm) in liquid YEPSL medium (0.4% sucrose, 0.4% peptone, 1% yeast extract). For testing the growth of different U. maydis strains in the presence of DTT, the respective strains were first grown in YEPSL medium until OD₆₀₀ reached 0.8. The cells were then collected by centrifugation and resuspended in fresh YEPSL medium at a final OD₆₀₀ of 1.0. Following this, the suspensions were serially diluted and spotted on solid complete medium (CM) supplemented with 2% glucose (Holliday, 1974) and 5 mM DTT. The growth of the tested strains was then compared with growth in solid CM without DTT.

4.3. Construction of plasmids

Standard molecular cloning techniques were used for plasmid construction. PCRs were performed using Phusion High‐Fidelity DNA polymerase (Thermo Scientific). All restriction enzymes were procured from New England Biolabs. Detailed cloning strategies for constructing all the plasmids used in this study are listed in Table S1 and the primer sequences used for the construction of each of these plasmids are listed in Table S2.

4.4. Construction of U. maydis strains

For generation of the hsp12 deletion mutant of U. maydis SG200 a 6.2‐kb fragment representing the hsp12 knockout cassette was PCR‐amplified from p∆hsp12 using the primer pair 5_∆hsp12/3_∆hsp12. SG200 protoplasts were then transformed with the PCR product and the resulting transformants were screened for hygromycin resistance to obtain SG200∆hsp12 candidates. The strain SG200∆hsp12_P_otef‐hsp12‐iLOV‐HA was constructed through the transformation of SG200∆hsp12 with P123‐P_otef‐hsp12‐iLOV‐HA linearized at AgeI to facilitate integration within the ip locus of the genome as described in Loubradou et al. (2001). Successful transformations were confirmed through PCR and Southern blot analysis as described in Method S1.

4.5. Plant infection

To infect maize plants U. maydis strains SG200 WT, SG200∆hsp12, and SG200∆hsp12‐hsp12‐iLOV‐HA were grown in YEPSL medium at 28°C under constant shaking at 200 rpm until OD₆₀₀ reached 0.8. The cells were then collected by centrifugation, washed once, and resuspended in sterile distilled water to a final OD₆₀₀ of 2.0. The resulting cell suspensions were used to infect 7‐day‐old maize seedlings through syringe infection. Leaf sections showing infection symptoms were collected at different time points after infection and prepared for microscopic examination.

4.6. Confocal microscopy

4.6.1. WGA‐AF/PI staining of infected leaf tissue

In order to visualize and compare the pathogenic development of SG200 WT and SG200∆hsp12, leaf sections of infected maize showing disease symptoms were excised and stained with WGA‐AF and PI as described in Redkar et al. (2018). Whereas WGA‐AF stains the fungal cell wall, PI stains the plant cell wall. The leaf samples thus prepared were mounted on water and visualized under a confocal microscope with appropriate illumination (WGA‐AF: excitation/emission at 495/519 nm; PI: excitation/emission at 493/636 nm).

4.6.2. Intracellular localization of Hsp12‐iLOV‐HA

To visualize the intracellular localization of Hsp12 in sporidial cells, SG200∆hsp12‐hsp12‐iLOV‐HA expressing Hsp12‐iLOV‐HA under the control of a constitutive promoter were grown in YEPSL medium at 28°C under constant shaking until OD₆₀₀ reached 0.6. The cells were then collected, washed in phosphate‐buffered saline, and mounted on a 2% agarose layer in a microscope slide. The fluorescence associated with iLOV was visualized under a confocal microscope with excitation at 405 nm and emission at 480 nm. To visualize Hsp12 in hyphae, leaf sections from maize plants infected with SG200∆hsp12‐hsp12‐iLOV‐HA were excised, mounted with water, and visualized with appropriate illumination (excitation/emission at 405/480 nm).

4.6.3. FM4‐64 staining

To assess any effect of Hsp12 on endocytic membrane trafficking in U. maydis SG200 WT and SG200∆hsp12, the respective cells were stained with the lipophilic styryl dye FM4‐64 (Invitrogen), which stains the plasma membrane and the endomembranes in eukaryotic cells (Vida & Emr, 1995). FM4‐64 (1 μM) was added to exponentially growing sporidial cells of SG200 WT and SG200∆hsp12 in YEPSL medium. The cells were then grown further for 5 min in the dark. Following this, the cells were washed with YEPSL twice at 4°C and finally resuspended in fresh YEPSL. The internalization of FM4‐64 was monitored at 5, 15, and 45 min. Cells collected at different chase points were placed on a thin 2% agarose layer and immediately visualized under a confocal microscope with appropriate illumination (excitation/emission at 515/640 nm). To assess the relative efficiencies of FM4‐64 internalization in SG200 WT and SG200∆hsp12, the plasma membrane‐to‐endomembrane ratio of the FM4‐64 fluorescence signal was compared between the respective strains as described in Method S2.

All confocal images were captured using a Leica Stellaris 5 confocal laser scanning microscope equipped with a 63× oil immersion objective and a Hamamatsu camera with Suite X (LAS X) software.

4.7. RNA extraction and reverse transcription‐quantitative PCR

To determine the relative transcript levels of hsp12 in U. maydis cells upon treatment with DTT, SG200 WT cells were grown at 28°C to the exponential phase and treated with 5 mM DTT (Promega) for 4 h. Following incubation, the control cells and DTT‐treated cells were collected through centrifugation, snap‐frozen in liquid nitrogen, and stored at −80°C. RNA isolation, subsequent cDNA synthesis by reverse transcription, and quantitative PCR were carried out as described in Mukherjee et al. (2020) and detailed in Method S3.

4.8. Purification of recombinant Hsp12 from E. coli

In order to purify recombinant Hsp12, Luria–Bertani (LB) growth medium was inoculated with E. coli BL21 (DE3)‐RIL cells harbouring the pET28a‐hsp12 plasmid and grown at 37°C under constant shaking at 200 rpm until OD₆₀₀ reached 0.6. At this point the expression of Hsp12 was induced with 500 μM isopropyl β‐d‐thiogalactopyranoside for 3 h at 37°C, following which the cells were collected by centrifugation and resuspended in lysis buffer (50 mM Tris pH 8.0, 100 mM NaCl) supplemented with EDTA‐free protease inhibitor cocktail. Lysis was carried out through sonication with an ultrasonicator (Cole Parmer). The lysed cell suspension was then centrifuged at 24,303 g for 30 min to remove cell debris. The cleared lysate containing 6×His‐tagged recombinant Hsp12 was passed through a Ni²⁺‐NTA affinity column (Takara) to facilitate binding of the recombinant protein to the column. Proteins nonspecifically bound to the column were washed away with lysis buffer supplemented with 10 mM and 20 mM imidazole in two subsequent washing steps. The purified recombinant Hsp12 was finally eluted from the column with lysis buffer supplemented with 100 mM imidazole, dialysed in lysis buffer overnight, and finally visualized in a 12% SDS‐PAGE gel.

4.9. CD spectroscopy

The far‐ultraviolet (UV) CD spectrum of Hsp12 was measured at 25°C from 200 to 260 nm in a quartz cell with a path length of 1 mm in a 150 CD‐spectropolarimeter (JASCO). Data were produced with a pitch of 1 nm and a scan rate of 100 nm/s. The far‐UV spectrum was also recorded after treating Hsp12 with 8 mM SDS at 25°C. Presented far‐UV spectra are the means of three acquisitions.

4.10. Aggregation protection assay

Aggregation protection activity of Hsp12 was tested on the nonspecific substrate lysozyme. Lysozyme (10 μM) was incubated with 20 mM DTT at 25°C for 1 h to induce protein aggregation, which was monitored through measurement of light scattering at 360 nm in a UV‐vis spectrophotometer (UH5300; Hitachi). To assess any role of Hsp12 in protecting lysozyme against DTT‐induced aggregation, either Hsp12 alone or Hsp12 in combination with 5 mM trehalose at a molar ratio of 1:1 was added to the lysozyme solution before adding DTT. The light scattering at 360 nm of the solutions was measured and compared with the light scattering in the absence of Hsp12. In addition, the effect of Hsp12 on DTT‐induced aggregation of lysozyme was also monitored by 12% SDS‐PAGE of the reaction solution. Following incubation of 10 μM lysozyme with 10 μM Hsp12 either in the absence or in the presence of 5 mM trehalose at 25°C for 1 h, the solutions were centrifuged at 18,000 g for 30 min. The distribution of lysozyme within the supernatant and pellet fractions was then visualized on a 12% SDS‐PAGE gel. ImageJ software was used to quantify the intensities of protein bands corresponding to lysozyme. Mn‐SOD from Bacillus aryabhattai was used as an internal control.

4.11. Preparation of DMPG LUVs

LUVs are a suitable biochemical model to study the interaction between proteins and membrane bilayers under in vitro conditions. Analysing this interaction might shed light onto the contribution of proteins to the overall fluidity and stability of lipid bilayers. Therefore, in order to study the interaction of Hsp12 with lipid, in vitro LUVs were made out of DMPG lipid as described in Mohid et al. (2022) and detailed in Method S4.

4.12. Measurement of structural perturbations of Hsp12 using ANS

To assess any structural perturbation of Hsp12 in the presence of a hydrophobic environment, 20 μM ANS was mixed with 10 μM Hsp12 in the presence of different concentrations of either SDS (3 to 8 mM) or DMPG LUVs (10 to 200 μg/ml) at 25°C. The resulting solutions were excited at 390 nm and the fluorescence emission spectra were recorded in the wavelength range from 400 nm to 600 nm with a data pitch of 1 nm using an F‐8500 fluorimeter (JASCO). The excitation and emission slits were set to 2.5 and 5 nm, respectively. The scan rate used was 240 nm/min. The experiment was performed in triplicate and the average fluorescence intensities at each data point in each sample were plotted against the range of emission wavelength scanned.

4.13. DLS

To test any role of Hsp12 in further vesiculation of 90–95 nm LUVs, the size distribution of the LUVs in the presence and absence of Hsp12 was measured through DLS using a Zetasizer Nano S (Malvern Instruments) as described in Method S5.

4.14. Assay to detect membrane stabilization using DPH

In order to test the effect of Hsp12 on the stabilization of lipid membranes, DMPG LUVs were labelled with DPH at 16°C for 15 min in a reaction mix containing 20 μg/mL DMPG LUV and 4 μM DPH. Following labelling, LUVs were titrated with increasing concentrations of Hsp12 either in the absence or in the presence of 5 mM trehalose at 25°C as described in Method S6.

4.15. Transmission electron microscopy

To visualize structural perturbations in DMPG LUVs in the presence of Hsp12, DMPG LUVs (0.5 μg/ml), either untreated or treated with 5 μM Hsp12 at 25°C for 1 h, were spread on AGS160‐3 copper grids (Agar Scientific) and stained with 2% uranyl acetate. The stained LUVs were then observed with a JEM‐2100 HR transmission electron microscope (JEOL) at 200 kV.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflicts of interest.

Supporting information

Figure S1. Graphs showing the relative number of order‐producing and disorder‐producing amino acids within the primary amino acid sequence of Ustilago maydis Hsp12.

Table S1. Cloning strategies for different plasmid constructs used in the present study.

Table S2. Primers used in the present study.

Method S1. PCR and Southern blot confirmation of different Ustilago maydis strains generated.

Method S2. Calculation of the relative ratio of fluorescence intensities associated with FM4‐64 between the plasma membrane and the endomembranes of different Ustilago maydis strains.

Method S3. RNA extraction and reverse transcription‐quantitative PCR.

Method S4. Preparation of 1,2‐dimyristoyl‐sn‐glycerol‐3‐phospho‐(1′‐rac‐glycerol) (DMPG) LUV.

Method S5. Dynamic light scattering (DLS).

Method S6. Titration of 1,2‐dimyristoyl‐sn‐glycerol‐3‐phospho‐(1′‐rac‐glycerol) (DMPG)‐labelled large unilamellar vesicles (LUVs) with Hsp12.

Mitra, A. , Bhakta, K. , Kar, A. , Roy, A. , Mohid, S.A. , Ghosh, A. et al. (2023) Insight into the biochemical and cell biological function of an intrinsically unstructured heat shock protein, Hsp12 of Ustilago maydis . Molecular Plant Pathology, 24, 1063–1077. Available from: 10.1111/mpp.13350

DATA AVAILABILITY STATEMENT

The data from the experimental work described herein are available on request from the corresponding author.