Filamentous fungi provide excellent systems for investigating the role of the AP-2 complex in polar growth. Using Aspergillus nidulans, we show that AP-2 has a clathrin-independent essential role in polarity maintenance and growth. This is in line with a sequence analysis showing that the AP-2 β subunit (β2) of higher fungi lacks a clathrin-binding domain, and experiments showing that AP-2 does not co-localize with clathrin. We provide genetic and cellular evidence that AP-2 interacts with endocytic markers SlaBEnd4 and SagAEnd3 and the lipid flippases DnfA and DnfB in the sub-apical collar region of hyphae. The role of AP-2 in the maintenance of proper apical membrane lipid and cell wall composition is further supported by its functional interaction with BasA (sphingolipid biosynthesis) and StoA (apical sterol-rich membrane domains), and its essentiality in polar deposition of chitin. Our findings support that the AP-2 complex of dikarya has acquired, in the course of evolution, a specialized clathrin-independent function necessary for fungal polar growth.https://doi.org/10.7554/eLife.20083.001
The five distinct Adaptor Protein (AP or adaptin) complexes are heterotetrameric adaptors that recruit membrane cargoes and coat proteins during vesicle formation at various subcellular locations for membrane trafficking in eukaryotes (Nakatsu and Ohno, 2003; Robinson, 2004, 2015). AP-1 is necessary for the formation of clathrin-coated vesicles that traffic between the trans-Golgi network (TGN) and early endosomes. AP-2 is involved in the formation of clathrin-coated endocytic vesicles from the plasma membrane (PM). AP-3 is involved in, apparently, clathrin-independent vesicle formation in the Golgi for traffic to endosomes and lysosomes or vacuoles. AP-4 and AP-5 seem to have more specialized roles in vesicle transport of specific cargoes from the TGN to endosomes and/or the cell surface (Hirst et al., 2013). All AP complexes are composed of two different subunits of high molecular weight ~100 kDa (called α, β or γ adaptins), one subunit of medium size (μ, 47–50 kDa), and one low-molecular-weight subunit (σ, 17–19 kDa). The five AP complexes have been shown to have a wide eukaryote distribution, supporting that all were present in the Last Eukaryote Common Ancestor (LECA) (Barlow et al., 2014). Noticeably, however, each AP complex, except AP-1, has been lost in various lineages. Fungi possess homologues of all AP-1, AP-2 and AP-3 subunits, but the great majority of them, have lost AP-4 and AP-5. This loss concerns all Dikarya (Ascomycetes and Basidiomycetes) and most primitive fungi. AP-4 or AP-5 subunits are however present in the glomeromycete Rhizophagus irregularis and chytrid fungi Spizellomyces punctatus and Batrachochytrium dendrobatidis, as well as, the sister group of all fungi, Fonticula alba. Notably, all eukaryotes studied possess clathrin, except for the Microsporidia intracellular parasites. Interestingly, the microsporidian Encephalitozoon cuniculi maintains AP subunits without having clathrin (Barlow et al., 2014), suggesting an unknown clathrin-independent role of AP complexes.
The role of the AP-2 complex in clathrin-mediated endocytosis is well established in mammals. By contrast, its role in endocytosis in unicellular eukaryotes, such as the yeasts Saccharomyces cerevisiae, Schizosaccharomyces pombe and Candida albicans has been questioned. An early systematic biochemical and genetic characterization of AP complexes in S. cerevisiae provided evidence that AP complexes are dispensable for clathrin function (Yeung et al., 1999). Similarly, Huang et al. (1999) have shown that AP deletion strains did not display the phenotypes associated with clathrin deficiency, including slowed growth and endocytosis, defective late Golgi protein retention and impaired cytosol to vacuole/autophagy function. Subsequently however, AP-2 has been localized to endocytic sites that are associated with clathrin, and have a cargo-specific function in killer toxin K28 endocytosis, making it likely that AP-2 functions with clathrin (Carroll et al., 2009, 2012). Furthermore, AP-2 has been shown to be critical for hyphal growth in C. albicans and polarized cell responses in S. cerevisiae, and in particular to be necessary for relocalization of the cell wall stress sensor Mid2 to the tip of a mating projection following pheromone addition (Chapa-y-Lazo et al., 2014). Additionally, an interaction between the mu (μ) subunit of AP-2 and the cell wall integrity pathway component Pkc1 was reported to affect recruitment of the AP-2 complex to endocytic sites (Chapa-y-Lazo and Ayscough, 2014). In another systematic screen to identify proteins required for cargo internalization, clathrin was shown to have a critical role for synaptobrevin homologue Snc1 endocytosis, but the role of AP-2 was not investigated (Burston et al., 2009). In a more recent study it has been shown that clathrin contributes to the regularity of vesicle scission and vesicle size, but is not required for elongating or shaping the endocytic membrane invagination, but again the role of AP-2 was not studied (Kukulski et al., 2016). Finally, in S. pombe, the AP-2 α component (Apl3p) was shown to interact physically with clathrin light chain (Clc1p), and genetically with both Clc1p and the endocytic factor Sla2p and additionally apl3Δ null mutants showed altered dynamics of endocytic sites associated with abnormal cell wall synthesis and morphogenesis (de León et al., 2016; de León and Valdivieso, 2016).
The apparently moderate role of AP-2 for growth in yeasts might reflect the fact that endocytosis in general is not essential for growth in unicellular fungi. In contrast to yeasts, the extreme polar growth of filamentous fungi, taking place by apical extension, is absolutely dependent on efficient and continuous endocytosis. Seminal studies on endocytosis and its relationship to growth have been performed in the basidiomycete Ustilago maydis (Steinberg, 2014) and the ascomycete Aspergillus nidulans (Peñalva, 2010), two emerging cell biology model systems.
Two aspects of endocytosis have been studied in detail in A. nidulans. The first relates to the essential role of apical endocytosis in hyphal tube extension and polar growth (Araujo-Bazán et al., 2008; Taheri-Talesh et al., 2008; Hervas-Aguilar and Penalva, 2010; Peñalva, 2010). It has been rigorously shown that polar growth maintenance is based on the spatial and functional coupling of rapid and continuous apical secretion and endocytosis in the hyphal tip region. Apically localized cargoes, as exemplified by the v-SNARE synaptobrevin homologue SynA, are internalized from a ‘collar’ region immediately below the hyphal tip and sorted in early endosomes (EEs). EEs can mature into Late Endosomes (LE) or Multivesicular Bodies (MVB) and be sorted into vacuoles, or recycle back, apparently via the trans-Golgi network (TGN) onto the PM. These post-endocytic trafficking processes taking place at the hyphal tips seem to be essential for the normal rate of secretion and proper growth. SlaBEnd4 and actin polymerization are major factors for apical endocytosis.
A second aspect of endocytosis in A. nidulans is related to its role in PM transporter down-regulation in response to physiological or stress signals (Gournas et al., 2010; Karachaliou et al., 2013). Unlike endocytosis related to apical cargoes taking place mostly at hyphal tips, transporter internalization takes place homogenously all along the hyphal membrane and necessitates cargo ubiquitination. Transporter ubiquitination requires a group of cargo-specific α-arrestin like proteins, acting as adaptors of the HECT-type ubiquitin ligase HulARsp5. Transporter ubiquitination is followed by internalization and sorting into EEs and the MVB/vacuolar pathway for terminal degradation. Transporter internalization, similar to apical endocytosis, requires actin polymerization and functional endocytic factors SagAEnd3 or SlaBEnd4.
Curiously, the role of clathrin or AP-2 in apical or transporter endocytosis had not, until very recently, been investigated in A. nidulans, or other filamentous fungi. An exception is a very recent report that studied clathrin dynamics in A. nidulans, following basically the subcellular localization and role in growth, of the clathrin heavy chain, ClaH (Schultzhaus et al., 2017b). This report showed that ClaH is localized mostly in late (trans) Golgi and secondarily to tentative endocytic sites, and although absolutely essential for growth, it does not seem to be involved in endocytosis. The later conclusion was, however, solely supported by studies concerning the internalization of the FM4-64 lipophilic marker, rather than specific endocytic cargoes. This report did not investigate the role of AP-2, or any AP complex, in relation to clathrin or to A. nidulans growth.
In this work, we investigate the role of the AP-2 complex and clathrin in polar growth and endocytosis in A. nidulans. We present evidence that the roles of AP-2 and clathrin are distinct, apparently due to an evolutionary truncation of the clathrin-binding domain in the β subunit of AP-2 in all higher fungi. We further provide evidence that AP-2 has a specialized, clathrin-independent role related to maintaining proper apical lipid and cell wall composition and thus promoting polar growth.
A recent evolutionary analysis studied the distribution and origin of all AP complex subunits in fungi (Barlow et al., 2014). The conclusion from this thorough study was that all fungi possess AP-1, AP-2 or AP-3 subunits, but most, and apparently all Dikarya, have lost AP-4 or AP-5. Given our interest on the role of AP1-3 complexes in membrane traffic and particularly of AP-2 in endocytosis in A. nidulans, we carried out a similar analysis with particular emphasis on Aspergillus species. Figure 1—figure supplement 1 shows that all Aspergilli possess highly similar homologous AP1-3 subunits, which can be unambiguously assigned into the distinct AP complexes. Aspergillus AP subunits are also phylogenetically highly related to homologues from all major fungal groups, as well as, selected model organisms (Figure 1—figure supplement 2). Importantly, AP subunits of a given complex are more similar to those of the orthologous complex in distantly related species, than to subunits of other AP complexes of the same species, in line with the acquisition of different functional roles of AP-1,–2 or −3 complexes in cargo membrane traffic.
A major and somehow surprising finding concerning the β subunits of not only Aspergillus species, but most fungi (all dikarya) came from the sequence analyses we performed. We observed that the β subunits of all three AP complexes do not contain the C-terminal part, of approximately 332 amino acid residues, which however is present in some primitive fungi, such as the chytrids Rozella allomycis, Allomyces macrogynus Batrachochytrium dendrobatidis, Spizellomycetales punctatus and Macrophomina phaseolina and in the fungal sister group Fonticula alba (all sharing 35–40% identity with human β subunits) (Figure 1A). The C-terminal domain of the β subunit is conserved in several Amoebozoa (25–36% identity), Rhizaria (31% identity), Alveolata (29% identity in Plasmodium or Toxoplasma species), Choanoflagellates (34–44% identity), Mesomycetozoea (44–50% identity), Algae (30% identity), and in the great majority of plants (38% identity) and Metazoa (58–60% identity) (Figure 1B and data not shown). Some prominent losses, except fungi, were identified in Trypanosoma and Leishmania species and in Excavata. Importantly, the lost part corresponds to an essential part of β appendage, which includes two C-terminal domains (pfam02883 and pfam09066) involved in clathrin binding and polymerization, and consequently necessary for the translocation of endocytic accessory proteins to a membrane bud site (Mousavi et al., 2004; Lemmon and Traub, 2012). A hydrophobic patch present in the lost domain also binds to a subset of DΦ[F/W] motif-containing proteins that are bound by the alpha-adaptin appendage domain (epsin, AP180) (Lemmon and Traub, 2012). Thus, fungal AP complexes might be incapable of interacting, at least via a ‘canonical’ association, with clathrin. Microsporidia, which are probably the only known eukaryotes lacking clathrin, also lack the clathrin binding domains in their AP complexes (Barlow et al., 2014). A major prediction from these observations is that, in fungi, clathrin-dependent processes, such as endocytosis and vesicle budding, might be AP-complex independent, while any role of the AP complexes might in turn be clathrin-independent. Obviously this prediction does not exclude that AP-2 or AP-1 complexes might still have the potential to interact with clathrin through unknown binding sequences in the β subunit or domains in other subunits, as for example reported for AP-1 in S. cerevisiae (Yeung and Payne, 2001). However, subsequent experimental evidence presented herein strongly support a clathrin-independent role of AP-2 in A. nidulans polar growth.
To investigate the role of AP complexes in cargo membrane traffic and endocytosis in A. nidulans we constructed three null mutants, each carrying a deletion of the gene encoding the σ subunit of AP-1 (AN7682; apaS), AP-2 (AN0722; apbS) or AP-3 (AN5519; apcS) (http://www.aspergillusgenome.org/ and Osmani and Goldman, 2008). Our primary interest was to examine the role of AP-2 in endocytosis, which however also necessitated the knock-out of the σ subunits of AP-1 and AP-3 in addition to that of AP-2, for excluding any possible functional complementation due to the similarity of the three σ subunits. Genetic disruption of the σ subunits, similarly to disruption of any of the four AP subunits, has been reported to inactivate the function of the full adaptor complex (Robinson, 2004, 2015). Isogenic null mutants were constructed and named thiAp-ap1σ, ap2σΔ or ap3σΔ (see Materials and methods). For the ap1σ gene, given its knock-out proved lethal (results not shown), we constructed a conditional knock-down strain using the thiamine-repressible promoter, thiAp (Apostolaki et al., 2012). The σ subunit null mutant strains were analyzed phenotypically and microscopically.
Knocking-down ap1σ or deleting ap2σ severely affected growth rate and colony morphology, whereas the ap3σΔ mutant exhibited only a very minor delay in growth rate (inserts in Figure 2A). The double mutant strains ap2σΔthiAp-ap1σand ap2σΔap3σΔ, constructed via standard genetic crossing, had growth rates similar to the single thiAp-ap1σ or ap2σΔ mutant, showing that the defect in ap2σΔis independent of AP-1 or AP-3 complexes. All mutants tested remained sensitive to selected antifungals, such as 5-flourouracil, 5-fluorocytosine or 8-azaguanine, similar to wild-type isogenic controls, suggesting that the expression and/or turnover of transporters specific for these antifungals is not affected (not shown).
We examined the microscopic morphology of the mutants constructed. thiAp-ap1σand ap2σΔ, but not ap3σΔ, showed altered hyphal morphology, characterized by anomalous polar tube hyper-branching and shorter coenocytic compartments, that is, reduced septa distances (Figure 2A). This effect was confirmed by quantification of the distinct microscopic morphological phenotypes and septa frequencies, as shown in Figure 2B and C. We also tested calcofluor deposition at the apex of polar tips in AP complex mutants, relatively to wild-type. Calcofluor stains chitin and the absence of intense calcofluor-staining at the tip indicates abnormal cell wall synthesis, related to problematic recycling of chitin synthases. This has been observed in mutants blocked in apical endocytosis (Higuchi et al., 2009). Calcofluor localization was totally depolarized in thiAp-ap1σand ap2σΔ, similar to what was observed in the endocytic mutant SagA (Figure 2D). Thus, the overall picture obtained with thiAp-ap1σ and ap2σΔ mutants was that of defective maintenance of polarity (Momany, 2002; Steinberg, 2014).
We investigated the role of the AP complexes on membrane cargo traffic and endocytosis by following the subcellular localization of a GFP-tagged transporter, namely UapA. UapA is an extensively studied uric acid-xanthine/H+ symporter, which has been used as a model cargo for studying the regulation of exocytosis and endocytosis of polytopic membrane proteins in A. nidulans (Pantazopoulou et al., 2007; Gournas et al., 2010; Karachaliou et al., 2013). After synthesis and translocation in the ER, fully functional GFP-tagged UapA molecules follow a vesicular secretion pathway to localize homogenously and stably in the PM of hyphal cells. However, UapA molecules undergo ubiquitylation-dependent endocytosis and vacuolar degradation in response to substrate excess or to the presence of rich nitrogen sources (NH4+ or glutamine). This picture remained unaffected when we followed the subcellular localization of UapA-GFP in isogenic thiAp-ap1σ, ap2σΔ or ap3σΔ mutant backgrounds (Figure 3A). The same picture is obtained under conditions of de novo or continuous UapA-GFP synthesis, or when signals triggering endocytosis are imposed after repression of UapA-GFP synthesis (data not shown). The non-essentiality of the AP-1 and AP-3 complexes in UapA traffic was somehow expected, because transporter exocytosis occurs by direct fusion of post-Golgi vesicles with the PM, and seemingly does not involve sorting from the Golgi to the endosomal compartment. However, the dispensability of AP-2, an adaptor involved in the formation of endocytic vesicles in other systems, for UapA-GFP endocytosis was somehow unexpected and contrasted the picture observed in standard endocytic mutants (sagAΔ or thiAp-slaB), where UapA internalization from the PM was totally blocked (see right panels in Figure 3A and Figure 3—figure supplement 1 for the construction of thiAp-slaB). The non-involvement of AP-2 in UapA endocytosis was further confirmed by time-lapse experiments, performed in the ap2σΔ mutant strain and an isogenic wild-type control (ap2σ+), which showed that UapA-GFP internalization initiates at 15 min and becomes more prominent 20–30 min after NH4+ addition in both strains (Figure 3B).
The non-involvement of AP-2 in UapA endocytosis raised issues worthy to be investigated. Firstly, we tested whether the endocytosis of other evolutionary, structurally and functionally distinct transporters is also AP-2 independent. Figure 3C,E and F show that this is indeed the case, as all six transporters tested are internalized with the same efficiency and rate, in both ap2σΔ and ap2σ+ backgrounds, in response to endocytic signals (Tavoularis et al., 2001; Vlanti and Diallinas, 2008; Apostolaki et al., 2009; Krypotou et al., 2015). Noticeably, different transporters exhibit different stabilities and endocytic ‘sensitivities’, but as quantification confirmed (Figure 3E and F), in all cases AP-2 proved dispensable for their proper internalization rate and turnover. We also examined whether the genetic elimination of a different AP-2 subunit, other than σ2, also leads to similar results in respect to transporter endocytosis. We deleted the gene encoding the μ2 subunit of the AP-2 complex, and subsequently constructed ap2σΔ ap2μΔ double mutants and strains expressing UapA-GFP in the ap2μΔ background. The ap2σΔ, ap2μΔ and ap2σΔ ap2μΔ mutants had identical phenotypes, characterized by loss of polarity maintenance and poor growth rate (Figure 3—figure supplement 2). UapA-GFP endocytosis operated normally in the ap2μΔ background, similar to what was observed in the ap2σΔ or wild-type isogenic backgrounds (Figure 3D). Overall, these results confirmed that AP-2 is dispensable for transporter endocytosis.
We showed that AP-2 is essential for polarity maintenance, but dispensable for transporter trafficking and endocytosis. To investigate the role of clathrin in these processes, we constructed relevant null or conditional mutants of the heavy (ClaH) and light (ClaL) chains of clathrin. The ClaH null mutant (claH) was not viable as it could only be rescued in heterokaryotic transformants (Osmani et al., 2006). The ClaL null mutant (claL) was viable, but severely affected in growth (Figure 4A). The corresponding conditional mutants were based on the use of the highly repressible thiAp promoter (Apostolaki et al., 2012). In the absence of thiamine (repressor), strains expressing thiAp-claH or thiAp-claL grew similar to wild-type isogenic control strains (not shown), whereas in the presence of thiamine thiAp-claH did not grow to form colonies, while thiAp-claL formed very small compact colonies, compatible with a severe growth defect (Figure 4A). Schultzhaus et al. (2017b), recently reported very similar growth phenotypes. We further examined the thiamine-repressible mutants under the microscope (Figure 4B). The thiAp-claL strain showed normal hyphae morphology and polar growth in the absence of thiamine, but prominent hypha tip swelling under ab initio repression of ClaL, that is, in the continuous presence of thiamine, especially at 37°C. This phenotype, which was confirmed by quantification (Figure 4C–E), might be associated with genetic blocks in apical endocytosis of specific cargoes. Knock-down of ClaL did not however modify the polar depositioning of calcofluor staining, which suggests normal chitin synthesis (Higuchi et al., 2009; Takeshita et al., 2012), and neither affected polarity maintenance. The thiAp-claH strain also showed normal hyphae morphology and polar growth in the absence of thiamine, but contrastingly, hyphae morphology was severely affected in the continuous presence of thiamine, becoming flattened and often showing swelling of tips or conidiospore heads, or becoming thinner and smaller in overall size, an indication of progressive lethality (Figure 4B and F). Still however, polarity maintenance was preserved when ClaH was repressed. Thus, in no case the phenotypes resulting from the absence of clathrin, either ClaH or ClaL, resembled those obtained in the absence of AP-2.
We also asked whether clathrin has a role in transporter endocytosis. Knockdown of ClaL expression blocked ammonium-elicited endocytosis of the UapA transporter, but knockdown of ClaH expression showed a more complex picture (Figure 4G). More specifically, in the latter case, when thiamine was added from the beginning of growth, UapA-GFP showed a rather diffuse and cytoplasmic appearance, instead of the normal cortical marking of the PM (notice the magnified area in the relevant insert in Figure 4G). This picture was obtained independently from the presence or absence of the endocytic trigger. Under the light of the recent publication of Schultzhaus et al. (2017b), which showed that ClaH is principally localized in the late Golgi, but also based on our own independent results on the effect of ClaH on other apical markers, described later in this work, it seems that ab initio knockdown of ClaH severely affects secretion of cargoes, including transporters. This leads to apparent Golgi and endosome collapse and progressive lethality, which could easily explain why we detect UapA-GFP in cytosolic structures, rather than in the PM. In addition, a block in transporter secretion would also create a cellular stress signal resulting in further internalization and turnover of transporters.
To allow secretion and proper localization of UapA-GFP to the PM to take place, and subsequently to check the role of ClaH in endocytosis, we decided to repress claH expression by adding thiamine after a period of pre-growth (14 hr) in the absence of thiamine. For this, we directly tested whether 5 or 10 hr repression of claH via the thiA promoter is sufficient for depleting ClaH, by performing and quantifying immunoblot assays using anti-GFP or anti-actin antibodies against total protein extracts from a strain expressing ClaH-GFP. Results shown in Figure 4—figure supplement 1 confirmed that addition of thiamine for ≥10 hr dramatically reduced ClaH. In contrast, 5 hr addition of thiamine had no effect on ClaH steady state levels, apparently because the relevant polypeptide is quite stable. Based on these immunoblots, we examined the effect of ClaH on UapA-GFP endocytosis by using at least a 10 hr period of claH repression, after pre-secretion of UapA-GFP into the PM. In this case, a significant fraction of UapA-GFP remained PM-associated despite the appearance of UapA-specific cytoplasmic structures and increased vacuolar degradation, which are however expected outcomes resulting from a block in Golgi functioning and transporter secretion (Figure 4G, extreme right panel).
Overall, our results strongly suggest that although both chains of clathrin seem essential for transporter (UapA-GFP) endocytosis, they do have distinct roles in transporter secretion. ClaL is dispensable for Golgi functioning and transporter secretion towards the PM, while ClaH seems absolutely essential for Golgi functioning and cargo secretion. These results further support that clathrin and the AP-2 complex have distinct roles in respect to both cargo secretion and transporter endocytosis.
Notably, we also found that FM4-64 endocytosis is not affected by deletions of genes encoding AP-2 complex subunits, clathrin or the endocytic factors SagA and SlaB, suggesting still another pathway of endocytosis in A. nidulans (Figure 4—figure supplement 2).
Guided by the fact that lack of a functional AP-2 complex (ap2σΔ or ap2μΔ mutants) leads to defective polarity maintenance and dramatically reduced growth, we tested whether AP-2 has a specific role in the polar localization of well-studied apical cargoes, such as the v-SNARE SynA (Valdez-Taubas and Pelham, 2003; Pantazopoulou and Peñalva, 2011) or the DnfA and DnfB flippases (Schultzhaus et al., 2015). These are the only presently known recycling cargoes of the endocytic collar shown rigorously to be involved in polar growth (Peñalva, 2015). In parallel, we examined the role of AP-2 in the polar localization of the endocytic markers, SlaB (Araujo-Bazán et al., 2008; Hervas-Aguilar and Penalva, 2010) and SagA (Karachaliou et al., 2013), or of AbpA, which marks the sites of actin polymerization (Araujo-Bazán et al., 2008). Appropriate strains were constructed by standard genetic crossing. Figure 5A shows that in the ap2σΔ genetic background there is total depolarization of DnfA and DnfB, as these proteins now mark the hyphal PM rather homogenously (see also Figure 5E, Figure 5—figure supplement 1A). In contrast, absence of a functional AP-2 did not detectably affect the polar localization of SagA, SlaB, SynA or AbpA. Quantification of polarized versus non-polarized localization of these markers confirmed our conclusions (Figure 5B). The same overall results were obtained when we used the ap2μΔ mutant rather than ap2σΔ (results not shown).
We tested the role of clathrin (ClaL) in the polar localization of DnfA, SlaB and SagA. Figure 5C shows that in a thiAp-claL strain repressed ab initio for ClaL expression, all markers tested remain polarly localized, marking principally the collar region behind the tip (see also Figure 5E). We also tested the role of ClaH in DnfA or DnfB localization (Figure 5D and Figure 5—figure supplement 1B). In a thiAp-claH genetic background, ab initio (that is, o/n) addition of thiamine led to cytoplasmic distribution of DnfA-GFP or DnfB-GFP, which in this case mark large, often bulbous, bodies and a membrane-like network (upper and lower left panels in Figure 5D), resembling the picture obtained when we followed UapA-GFP in the absence of ClaH (see Figure 4G), or when Golgi/endosomes collapse in response to the presence of Brefeldin A (BFA) (middle panels in Figure 5D). We further showed that addition of BFA in the culture medium of a strain expressing ClaH (de-repressed thiAp-claH) leads to reversible de-polarization of DnfA-GFP (lower panel in Figure 5—figure supplement 2), but to irreversible de-polarization of DnfA-GFP and progressive apparent hyphae death (that is, hyphae cells become thinner) when ClaH expression is repressed (upper panel in Figure 5—figure supplement 2). This picture is in very good agreement with the reversibility of the negative effect of BFA on Golgi function, which has been rigorously studied in Pantazopoulou and Penalva (2009). Importantly, our results show that normal Golgi secretion is necessary for maintaining the polar localization of DnfA-GFP.
When ClaH was depleted, by thiamine addition for 10 hr, following a period of pre-growth (16 hr) that allows a fraction of DnfA-GFP or DnfB-GFP to be secreted and polarly localized, a degree of polarization of both flippases was preserved in some hyphae, but in general the polar localization of DnfA-GFP or DnfB-GFP was disrupted (see upper and lower right panels in Figure 5D and also Figure 5—figure supplement 1B). This picture, however, did not constitute a surprise, given the effect of BFA in DnfA-GFP polar localization. It is rather compatible with the idea that a severe block in conventional cargo secretion, obtained either by BFA or by the depletion of ClaH, leads to an inability to maintain the apical localization of markers, such as DnfA, and apparently DnfB.
In summary, our results strongly suggest that the depletion of clathrin does not directly affect the pre-established polarization of flippases. While the dispensability of ClaL in apical cargo localization is clear, the non-essential role of ClaH on flippase polarization establishment is more difficult to formally confirm because depletion of ClaH affects Golgi function and secretion, and thus polarization maintenance. Our results are in agreement with results reported in Schultzhaus et al. (2017a); (2017b), which establish that ClaH is localized principally in late Golgi, only weakly in the apical collar region and is excluded from the hyphal tip. Overall, these results further showed that AP-2 and clathrin have distinct roles in cargo subcellular localization in A. nidulans.
We performed a series of co-localization studies examining the subcellular positioning of AP-2 in relationship to several membrane cargoes, including endocytic proteins (SagA and SlaB), actin polymerization markers (AbpA), apical markers (DnfA, DnfB and SynA) or transporters (UapA). For that we constructed strains expressing functional Ap2σ-GFP or Ap2σ-mRFP, and crossed these with strains expressing GFP- or RFP-tagged apical markers. Figure 6A shows that, in a wild-type background, Ap2σ-GFP or Ap2σ-mRFP has a polar cortical distribution, marking mostly the endocytic collar of growing tips, but also the septa. The apical and cortical localization of Ap2σ-GFP was ‘replaced’ by a rather diffuse cytoplasmic fluorescent signal in an ap2μΔ genetic background, showing that disruption of the full AP-2 complex by deleting a single subunit also disrupts its physiological localization. Noticeably, apical localization of Ap2σ-GFP remained unaffected in a genetic background lacking clathrin light chain (repressed thiAp-claL). Kymograph analysis further showed that the AP-2 complex marking the apical region, including the collar, is rather static. This also suggests that AP-2 is not localized in highly motile early endosomes (EEs). Additionally, no significant co-localization of Ap2σ-GFP was obtained with a marker of the TGN (PHOSBP) (Figure 6—figure supplement 1).
AP-2 co-localized cortically, with a high statistical significance (p<0.001), with SagA (80.7%), SlaB (77.4%), DnfA (70.1%), DnfB (77.0%), AbpA (75.8%) and SynA (68.8%). We also quantified and statistically confirmed that co-localization of AP-2 with apical markers (DnfA and SynA) occurs basically in the collar region, rather than the apical tip (lower panel in Figure 6C). Noticeably, co-localization of AP-2/SlaB or AP-2/SagA was modified in genetically deleted sagAΔ or thiAp-slaB backgrounds, respectively, becoming less polarized and more extending away from the tip (Figure 6D). Importantly, AP-2 did not co-localize with clathrin (ClaL) or UapA (Figure 6B,C). In fact ClaL and UapA, in contrast to all other apical markers, do not localize significantly in the collar region, where apical endocytosis occurs.
Given the importance of confirming the non-colocalization of AP-2 and clathrin for the present work, we obtained additional evidence to support this finding. Figure 6—figure supplement 2, panel A, shows that strains expressing the chimeric ClaL-mRFP or ClaL-GFP proteins grow similar to wild-type, contrasting the severe growth defects associated with the lack of ClaL expression. This constitutes good evidence that GFP- or RFP-tagged versions of ClaL are functional. Additionally, panel B shows that the localization of ClaL-GFP or ClaL-mRFP is identical and compatible with the expected localization for clathrin, as in both cases ClaL marks Golgi-like structures and cortical foci. Panel C further shows a rather weak association of ClaL with subcortical patches close to the collar endocytic region. The apparent prominent localization of ClaL in Golgi-like structures is in full agreement with results in Schultzhaus et al. (2017b), who have shown that ClaH-GFP localizes principally in the late Golgi, and has only a weak association with the sub-apical collar region. In the same article, the authors have further shown that ClaH and ClaL co-localize. Finally, the functionality of ClaL-mRFP was also confirmed by showing that strains expressing ClaL-mRFP are fully active in respect to UapA-GFP endocytosis (panel D), contrasting the block of UapA endocytosis when ClaL is not functional (see Figure 4G).
To further test the distribution of AP-2 and clathrin between plasma membrane and internal vesicles, as well as, to assess more rigorously the degree of possible co-localization between these two protein complexes, we analysed the localization of Ap2σ-GFP and ClaL-mRFP by TIRF microscopy, followed by relevant quantification analysis (Figure 7). Practically no co-localization of the two polypeptides was observed (PPC = 0.12) in the PM. Only a minor fraction of the two polypeptides colocalized (PPC = 0.34) intracellularly. These data are in perfect line with the rest of our findings, confirming that the function of AP-2 in apical endocytosis is clathrin-independent.
Overall, our results are also in line with the following notions. First, AP-2 seems to synergize with endocytic factors SlaB and SagA at the endocytic collar, at sites of actin polymerization, marked by AbpA. Second, disruption of SagA or SlaB, but not of AP-2, somehow depolarizes the localization of this endocytic complex. Third, AP-2 is very probably the cargo-recognition (for example, DnfA and DnfB) partner of this complex. Finally, AP-2 and clathrin are involved in mutually exclusive endocytic or secretion pathways, also supporting a cargo-centric view of membrane trafficking (Maldonado-Báez et al., 2013).
To further establish the role of AP-2 we tested its genetic interactions with SagA, DnfA and DnfB, but also with proteins involved in apical lipid maintenance, namely StoA and BasA. StoA is a stomatin homologue involved in the maintenance of apical sterol-rich membrane domains (SRDs) and polarity in A. nidulans (Takeshita et al., 2012). In metazoa, stomatins are oligomeric, lipid raft-associated proteins with scaffolding functions necessary for the maintenance of specific lipid composition in membranes, but their molecular function is still unclear (Lapatsina et al., 2012). BasA is required for phytosphingosine biosynthesis and is essential for fungal viability. A previously reported thermosensitive missense mutation in basA (basA1) resulted in an aberrant cell wall thickening and growth arrest at 42°C (Li et al., 2007). For the present work, we used null mutants of SagA (Karachaliou et al., 2013), DnfA, DnfB (Schultzhaus et al., 2015) and StoA (Takeshita et al., 2012), and either basA1 or a conditional knock-down mutant of basA, constructed herein using the thiAp promoter. These mutants were all crossed with the ap2σΔ strain. Growth phenotypes of the resulting double mutants compared to single mutants are shown in Figure 8A and D. In all cases, double mutants showed reduced growth compared to single mutants (that is, synthetic negative phenotypes), strongly suggesting that the function of AP-2 is related to that of SagA, DnfA and DnfB, as expected based on the subcellular localization experiments, but also to StoA and BasA. Microscopic examination of the morphology of the mutants confirmed that the defects in AP-2 and SagA, DnfA, DnfB, StoA or BasA were additive (Figure 8B,C and E). Finally, Figure 8F shows that the localization of AP-2σ-GFP in a basA1 genetic background is not anymore in the collar region, but instead marks internal structures just behind the tip. All above observations were confirmed by relevant quantitative analysis, as shown in Figure 8G and H.
Given that AP-2 was related to BasA, and thus sphingolipid localization, we tested whether AP-2 also affects the localization of ergosterol, the partner of sphingolipids in lipid rafts. To do so, we used filipin III, an established fluorescent ergosterol marker (Van Leeuwen et al., 2008). In wild-type A. nidulans and other filamentous fungi, filipin stains the PM, but with a predominant polar deposition at the hyphal apex (Li et al., 2006). We observed significant depolarization, often associated with the appearance of discrete cortical foci and abnormal filipin staining in the ap2σΔ, as well as, in thiAp-slaB, stoAΔ and thiAp-basA mutant backgrounds. In contrast, we observed a polar localization of filipin in thiAp-claL, dnfAΔ, dnfBΔ and sagAΔ mutant strains, similar to wild-type (Figure 9A). Quantification of these results, by measuring the strength of the filipin fluorescent signal along the tip, confirmed the critical role of AP-2, as well as, of BasA, SlaB or StoA, in the apical depositioning of ergosterol (Figure 9B). These results further supported that AP-2 has a specific role in lipid apical localization, distinct from clathrin, necessary polar growth.
In mammals, the AP-2 appendages of α2 and β2 subunits orchestrate endocytosis through hierarchical interactions with clathrin, other endocytic proteins and PM lipids. The appendage of the β2 subunit specifically interacts with clathrin (Keyel et al., 2008; Thieman et al., 2009). Lower eukaryotes, like free living filamentous fungi, face the challenge of rapid and polarized growth via apical extension, a process that is absolutely dependent on efficient endocytosis and recycling of chaperones and enzymes, related to PM and cell wall deposition at the growing tip (Peñalva, 2010, 2015; Peñalva et al., 2012). Fungi also need both basal and conditionally elicited endocytosis all along their hyphal PM, serving the renewal or modification of membrane components in response to changing nutritional or stress conditions, best exemplified by the endocytic turnover of transporters (Gournas et al., 2010; Karachaliou et al., 2013; Diallinas, 2014). The role of AP-2 and that of clathrin in the endocytosis of apical cargoes or transporters had not been studied in filamentous fungi, until recently.
Here we showed that AP-2 is dispensable for transporter endocytosis and conventional apical secretion, but essential for polarity maintenance and the polar localization of membrane lipid or cell wall components. In parallel, we showed that clathrin (both ClaH and ClaL) is also essential for A. nidulans growth and that ClaH has a critical role in Golgi function and cargo secretion, which is probably the main reason why the ClaH null mutation is lethal. The importance of ClaH in conventional secretion via the Golgi was also recently reported by Schultzhaus et al. (2017b). An essential role of clathrin in the endocytosis of transporters is also well supported by our results. Importantly, we also showed that the function of ClaL is unrelated to polarity maintenance or the polar localization of lipid or cell wall- related apical cargoes, which contrasts its essential role in transporter endocytosis. Our results also supported that the effect of ClaH depletion in maintaining the polar localization of apical cargoes is due to malfunctioning of Golgi-dependent secretion, rather than a direct effect on the polar localization establishment, similarly to what is observed when BFA is added to growing cells. These results, together with the absence of any significant overlap in the PM localization of clathrin and AP-2, and the distinct phenotypes of clathrin and AP-2 mutants, confirmed that clathrin and AP-2 are recruited in distinct, cargo-dependent, trafficking pathways.
Led by the observation that genetic knock-out of AP-2 resulted in phenotypes compatible with loss of polarity maintenance, we followed the effect of deleting AP-2 on the subcellular localization of well-established apical markers. We thus showed that AP-2 co-localizes apically with endocytic factors SlaB and SagA at sites of actin polymerization in the sub-apical collar region. Subsequently, we provided direct evidence that AP-2 has a role on the apical localization and function of DnfA and DnfB flippases, and affects ergosterol (filipin) and cell wall (calcofluor) deposition at the tip. We finally showed that AP-2 interacts functionally with other proteins involved in apical lipid maintenance and scaffolding (BasA and StoA). The key role of AP-2 in apical endocytosis is in line with previous results showing that proper membrane lipid and cell wall composition is essential for fungal polar growth (Cheng et al., 2001; Takeshita et al., 2012). A speculative model on the role of AP-2 and other relevant factors in apical endocytosis and polarity maintenance is shown in Figure 10. This model also considers recent findings on the recycling of apical markers described in Schultzhaus et al. (2015); (2016); (2017b) and Peñalva (2015).
An AP-2 independent role of clathrin in endocytosis is not novel, as it has been reported before in mammals and yeast (Conner and Schmid, 2003; Motley et al., 2003; Sorkin, 2004; Traub, 2009; Brach et al., 2014). To our knowledge, however, the opposite is a novel finding, as no other report has shown a major clathrin-independent role of the AP-2 complex. A role, albeit minor, for AP-2 in maintaining normal post-endocytic trafficking of major histocompatibility complex class I (MHCI) proteins and beta1 integrin is the only reported case of a clathrin-independent role of AP-2 (Lau and Chou, 2008). Interestingly, Microsporidia possess AP complexes, but not clathrin, which further shows that AP complexes can function without clathrin. The fact that no canonical clathrin binding domains were identified in the β1 and β2 subunits of primitive fungi supports the notion that Dikarya have lost clathrin-binding domains during their evolution. The experimental support of a clathrin-independent role of AP-2 in endocytosis was in good agreement with the in silico observation that the β2 subunit of AP-2 of all higher fungi lacks the entire C-terminal β appendage, which includes known clathrin binding domains. Interestingly, clathrin binding domains are also missing from the AP-1 β subunit (β1) of all Dikarya, suggesting that in fungi AP-1 might also function independently of clathrin, very probably being involved in the recycling via the TGN compartment of specific apical proteins. Thus AP-2 and AP-1 might function in the same pathway for apical cargo recycling. Interestingly, AP-1 has been reported to function independently of clathrin in phagocytosis in murine macrophages (Braun et al., 2007). This has been correlated with the presence of two isoforms of the γ-adaptin subunit of AP-1 in macrophages (Santambrogio et al., 2005), and it has been proposed that the cleaved form of AP-1 is the one associated with vesicles present under phagocytic cups (Braun et al., 2007).
Our findings strongly suggest the existence of clathrin-independent endocytosis (CIE) related to polar growth in A. nidulans. Over the last years, there has been an increasing interest concerning CIE, not only because it is the mode of entry of bacterial toxins and cell surface proteins (Maldonado-Báez et al., 2013; Mayor et al., 2014), but also because it has raised strong debates concerning its physiological significance. In a recent report, Bitsikas et al. (2014), provided evidence that clathrin-independent pathways do not, in fact, contribute significantly to endocytic flux. However, in cases where it has been supported experimentally, CIE seems to be specialized for the maintenance of PM lipid composition (Shvets et al., 2015) and the endocytosis of proteins anchored to the membrane by glycosyl phosphatidylinositol (GPI) (Nichols, 2009; Bitsikas et al., 2014). Additionally, in some cases CIE has also been reported to depend on specific PM lipid rafts and/or microdomains, such as caveolae or flotillins (Glebov et al., 2006), despite some recent contradicting reports supporting that caveolae and flotillins function in organizing PM domains, rather being bona fide endocytic factors (Parton and del Pozo, 2013). Finally, some CIE pathways also depend on dynamin (Lamaze et al., 2001). An apparent conclusion from the above and other studies is that distinctions and variations in both clathrin-dependent endocytosis (CME) or CIE seem to arise from the endocytic cargo being examined (Maldonado-Báez et al., 2013). Our work suggests that AP-2 is a key cargo-specific recognition factor in apical CIE in fungi.
For long, yeasts were thought to depend solely on CME. However, recent evidence demonstrated the existence of a CIE pathway that depends on the GTPase Rho1, Rom1/2, formin Bni1 and α-arrestins (Prosser and Wendland, 2012; Prosser et al., 2011, 2015). In addition, a CIE pathway was discovered in Candida albicans (Epp et al., 2013). Interestingly, in yeast, α-arrestins are important for cargo selection in both the CME and CIE, but seem to function by distinct mechanisms in each pathway, as in the CIE pathway, unlike CME, their function is independent of the Rsp5 ubiquitin ligase (Lin et al., 2008; Nikko et al., 2008; Prosser et al., 2015). Although the involvement of AP-2 in CIE in yeast has not been examined, the involvement of α-arrestins differentiates it from the apical CIE pathway we identify herein, as none of the A. nidulans arrestin knock-out mutants shows defects in polarity maintenance and growth (Karachaliou et al., 2013). This however does not exclude the possibility that the apical CIE pathway in A. nidulans shares other factors of the CIE pathway identified in yeast (for example, Rho1, Rom1/2 or formin Bni1). In fact, SepA (orthologue of Bni1; Harris et al., 1997; Sharpless and Harris, 2002), is necessary for septum formation and polarity maintenance, and RhoARho1 plays a role in polarity, proper branching pattern, and cell wall deposition (Guest et al., 2004), phenotypes related to the proposed AP-2 function.
The present work does not intend to explore the mechanistic details of the AP-2 pathway in A. nidulans or identify the entire set of factors involved. However it shows, for the first time, the existence of a clathrin-independent pathway related to polarity maintenance in filamentous fungi, and most importantly identifies AP-2 as a key factor for the apical recycling of enzymes, lipids and cell wall components necessary for fungal growth. From the evolutionary point of view, our work suggests that in higher fungi the function of the AP-2 complex has been uncoupled, probably via a specific gene truncation, from clathrin-mediated endocytosis, and that this uncoupling seems to serve specific cellular challenges of fungal polar growth.
Standard complete and minimal media (MM) for A. nidulans were used. Media and supplemented auxotrophies were at the concentrations given in FGSC (http://www.fgsc.net.) (RRID: SCR_008143). Media and chemical reagents were obtained from Sigma-Aldrich (Life Science Chemilab SA, Hellas) or AppliChem (Bioline Scientific SA, Hellas). Glucose 1% (w/v) was used as a carbon source. NaNO3 at 10 mM was used as a nitrogen source. Thiamine hydrochloride was used at a final concentration of 5–10 μM. A. nidulans transformation was performed as described previously in Koukaki et al. (2003). An nkuA DNA helicase deficient strain (TNO2A7; Szewczyk et al., 2007; Nayak et al., 2006) was the recipient strain for generating ‘in locus’ integrations of tagged gene fusions, or gene deletions by the A. fumigatus markers orotidine-5’-phosphate-decarboxylase (AFpyrG, Afu2g0836) or GTP-cyclohydrolase II (AFriboB, Afu1g13300), resulting in complementation of auxotrophies for uracil/uridine (pyrG89) or riboflavin (riboB2) respectively. Transformants were verified by PCR and Southern analysis. Combinations of mutations were constructed by standard genetic crossing. A. nidulans strains used are listed in Supplementary file 1.
Genomic DNA extraction from A. nidulans was as described in FGSC (http://www.fgsc.net) (RRID: SCR_008143). Plasmid preparation from E. coli strains and DNA bands gel extraction were done using the Nucleospin Plasmid kit and the Nucleospin Extract II kit (Macherey-Nagel, Lab Supplies Scientific SA, Hellas). DNA sequences were determined by VBC-Genomics (Vienna, Austria). Southern blot analysis using specific gene probes and upstream or downstream fragments in the case of verifying gene deletions, was performed as described in Sambrook et al. (1989). [32P]-dCTP labeled molecules of gene specific probes were prepared using a random hexanucleotide primer kit following the supplier's instructions (Takara Bio, Lab Supplies Scientific SA, Hellas) and purified on MicroSpin S-200 HR columns (Roche Diagnostics, Hellas). Labeled [32P]-dCTP (3000 Ci mmol−1) was purchased from the Institute of Isotops Co. Ltd, Miklós, Hungary. Restriction enzymes were from Takara Bio (Lab Supplies Scientific SA, Hellas). Conventional PCR reactions, high fidelity amplifications and site-directed mutagenesis were performed with KAPA Taq DNA and Kapa HiFi polymerases respectively (Kapa Biosystems, Lab Supplies Scientific SA, Hellas). Gene deletions and ‘in locus’ integrations of tagged gene fusions were generated by one step ligations or sequential cloning of the relevant fragments in the plasmids pBluescript SKII, or pGEM-T using oligonucleotides carrying additional restriction sites. These plasmids were used as templates to amplify the relevant linear cassettes by PCR. For primers and information related to these constructs see Supplementary file 2.
Cultures for total protein extraction were grown in MM supplemented with NaNO3 at 25o C. Thiamine hydrochloride was used at a final concentration of 5–10 μM. Total protein extraction was performed as previously described (Galanopoulou et al., 2014). Equal sample loading was estimated by Bradford assays. Total proteins (30–50 μg) were separated in polyacrylamide gels (8–10 % w/v) and electroblotted (Mini PROTEAN Tetra Cell, BIORAD) onto PVDF membranes (Macherey-Nagel, Lab Supplies Scientific SA, Hellas). Immunodetection was performed with a primary mouse anti-GFP monoclonal antibody (Roche Diagnostics), a mouse anti-actin monoclonal (C4) antibody (MP Biomedicals Europe) and a secondary goat anti-mouse IgG HRP-linked antibody (Cell Signaling Technology Inc, Bioline Scientific SA, Hellas). Blots were developed using the LumiSensor Chemiluminescent HRP Substrate kit (Genscript USA, Lab Supplies Scientific SA, Hellas) and SuperRX Fuji medical X-Ray films (FujiFILM Europe). Quantification of ClaH-GFP or actin levels using ImageJ were estimated separately and relative to each other and are given as ratios, where in each case the lowest value was arbitrarily set as 1.
BLASTp searches were performed on the NCBI database (RRID: SCR_004870) to identify which organisms acquire a C-terminal domain on β AP-2 subunit, with H. sapiens β AP-2 as query. The selected sequences were retrieved and, with the use of UniProt database (RRID: SCR_002380), the C-terminal domains were identified and selected from each sequence. The phylogenetic tree reconstruction of the C-terminal domains was performed on MEGA6 software (RRID: SCR_000667) (Tamura et al., 2013) with the maximum-likelihood method and bootstrap testing for 150 replications. The substitution model was WAG and the ML heuristic method selected was Nearest-Neighbor-Interchange (NNI). NCBI sequences used for figure illustrations; H. sapiens (NP_001273: βAP-2, NP_001118: βAP-1 and NP_004635:β AP-3), F. alba (XP_009492690: β AP-2, XP_009492179: β AP-1 and XP_009495754: β AP-3), R. allomyces (EPZ36209: β AP-2, EPZ33551: β AP-1 and EPZ35993: β AP-3), S. punctatus (KNC98413: β AP-2, KNC96576: β AP-1 and KND02292: β AP-3), A. nidulans (CBF70501: β AP-2, CBF83537: β AP-1 and CBF90059: β AP-3), S. cerevisiae (NP_012538: β AP-2, NP_012787: β AP-1 and NP_011777: β AP-3), and S. pombe (NP_596435: β AP-2, NP_595274: β AP-1 and NP_593796: β AP-3). The illustrations were taken from HMMER (RRID: SCR_005305) homology (search mode) of EBI database (RRID: SCR_002872) where the sequences were subjected and subsequently manipulated.
Samples for wide-field epifluorescence microscopy and Total Internal Reflection Fluorescence Microscopy (TIRF-M) were prepared as previously described (Evangelinos et al., 2016). Germlings were incubated in sterile 35 mm μ-dishes, high glass bottom (ibidi, Germany) in liquid MM for 16–22 hr at 25° C. Filipin III and Calcofluor white were used for 5 min prior to observation at final concentrations of 1 μg ml−1 and 0001% (w/v) respectively. FM4-64 and CMAC staining was according to Peñalva (2015) and Evangelinos et al. (2016), respectively. Brefeldin A was used at a final concentration of 100 μg ml−1. Images were obtained using a Zeiss Axio Observer Z1/Axio Cam HR R3 camera. Contrast adjustment, area selection and color combining were made using the Zen lite 2012 software. Live imaging of plasma membrane ClaL was accomplished by TIRF-M. Hyphae were analyzed using a Leica AM TIRF MC set up on a Leica DMI6000 B microscope and a Leica 100X HCX PL APO 1.4 NA objective. Biological replicates correspond to different samples, while technical replicates correspond to different hyphae observed and/or photographed within each sample. For kymograph generation and all measurements of fluorescence intensity, the Reslice and Plot profile commands in ImageJ (RRID: SCR_003070) were used, respectively (https://imagej.nih.gov/ij/). For the statistical analysis in Figure 2, Tukey’s Multiple Comparison test was performed (One-way ANOVA), using the Graphpad Prism software (RRID: SCR_002798). Confidence interval was set to 95%. For quantifying colocalization (Dunn et al., 2011), Pearson's correlation coefficient (PCC) above thresholds, for a selected Region of interest (ROI) was calculated, using the coloc2 plugin of Fiji (RRID: SCR_002285). Costes P-value was 1.00 for all the images tested (Costes et al., 2004). PSF was set to 1.2 and the number of iterations was 100. One sample t-test was performed to test the significance of differences in PCCs, using the Graphpad Prism software. Confidence interval was set to 95%. Colocalization results were also confirmed using the ICY (RRID: SCR_010587) colocalization studio plugin (pixel-based method) (http://icy.bioimageanalysis.org/) for ROIs that included either the foci and the tip area which were selected using the Spot Detector plugin, or the extreme apex area, selected using the Area Selection tool. The same tool was also used for the measurement of Vacuolar Surface (Total surface of vacuoles containing GFP/Hypha) and Vacuolar GFP Fluorescence (Total fluorescence intensity of vacuoles containing GFP/Hypha) in Figure 3E–F, while Tukey's Multiple Comparison Test (One-Way ANOVA) using Graphpad Prism was performed to test the statistical significance of the results. Scale bars were added using the FigureJ plugin of the ImageJ (RRID: SCR_003070) software (Mutterer and Zinck, 2013). Images were further processed and annotated in Adobe Photoshop CS4 Extended version 11.0.2 (RRID: SCR_002078).
From all to (nearly) none: tracing adaptin evolution in fungiCellular Logistics 4:e28114.https://doi.org/10.4161/cl.28114
AP-1 and ARF1 control endosomal dynamics at sites of FcR mediated phagocytosisMolecular Biology of the Cell 18:4921–4931.https://doi.org/10.1091/mbc.E07-04-0392
Regulators of yeast endocytosis identified by systematic quantitative analysisThe Journal of Cell Biology 185:1097–1110.https://doi.org/10.1083/jcb.200811116
Analysis of yeast endocytic site formation and maturation through a regulatory transition pointMolecular Biology of the Cell 23:657–668.https://doi.org/10.1091/mbc.E11-02-0108
Cell cycle progression and cell polarity require sphingolipid biosynthesis in Aspergillus nidulansMolecular and Cellular Biology 21:6198–6209.https://doi.org/10.1128/MCB.21.18.6198-6209.2001
Differential requirements for AP-2 in clathrin-mediated endocytosisThe Journal of Cell Biology 162:773–780.https://doi.org/10.1083/jcb.200304069
The long life of an endocytic patch that misses AP-2Current Genetics 62:765–770.https://doi.org/10.1007/s00294-016-0605-3
Phylogeny.fr: robust phylogenetic analysis for the non-specialistNucleic Acids Research 36:W465–W469.https://doi.org/10.1093/nar/gkn180
A practical guide to evaluating colocalization in biological microscopyAJP: Cell Physiology 300:C723–C742.https://doi.org/10.1152/ajpcell.00462.2010
Transport-dependent endocytosis and turnover of a uric acid-xanthine permeaseMolecular Microbiology 75:246–260.https://doi.org/10.1111/j.1365-2958.2009.06997.x
Aspergillus nidulans RhoA is involved in polar growth, branching, and cell wall synthesisFungal Genetics and Biology 41:13–22.https://doi.org/10.1016/j.fgb.2003.08.006
The AP-2 adaptor beta2 appendage scaffolds alternate cargo endocytosisMolecular Biology of the Cell 19:5309–5326.https://doi.org/10.1091/mbc.E08-07-0712
A novel improved method for Aspergillus nidulans transformationJournal of Microbiological Methods 55:687–695.https://doi.org/10.1016/S0167-7012(03)00208-2
Origin, diversification and substrate specificity in the family of NCS1/FUR transportersMolecular Microbiology 96:927–950.https://doi.org/10.1111/mmi.12982
Statistical analysis of molecule colocalization in bioimagingCytometry Part A 87:568–579.https://doi.org/10.1002/cyto.a.22629
Distinct ceramide synthases regulate polarized growth in the filamentous fungus Aspergillus nidulansMolecular Biology of the Cell 17:1218–1227.https://doi.org/10.1091/mbc.E05-06-0533
Clathrin-independent endocytosis: a cargo-centric viewExperimental Cell Research 319:2759–2769.https://doi.org/10.1016/j.yexcr.2013.08.008
Polarity in filamentous fungi: establishment, maintenance and new axesCurrent Opinion in Microbiology 5:580–585.https://doi.org/10.1016/S1369-5274(02)00368-5
Clathrin-mediated endocytosis in AP-2-depleted cellsThe Journal of Cell Biology 162:909–918.https://doi.org/10.1083/jcb.200305145
Adaptor protein complexes as the key regulators of protein sorting in the post-Golgi networkCell Structure and Function 28:419–429.https://doi.org/10.1247/csf.28.419
Endocytosis of lipid-anchored proteins: excluding GEECs from the crowdThe Journal of Cell Biology 186:457–459.https://doi.org/10.1083/jcb.200907119
The Aspergilli: Genomics, Medical Aspects, Biotechnology, and Research MethodsCRC Press.
Organization and dynamics of the Aspergillus nidulans golgi during apical extension and mitosisMolecular Biology of the Cell 20:4335–4347.https://doi.org/10.1091/mbc.E09-03-0254
Maturation of late golgi cisternae into RabERAB11 exocytic post-Golgi carriers visualized in vivoMolecular Biology of the Cell 25:2428–2443.https://doi.org/10.1091/mbc.E14-02-0710
Caveolae as plasma membrane sensors, protectors and organizersNature Reviews Molecular Cell Biology 14:98–112.https://doi.org/10.1038/nrm3512
A lipid-managing program maintains a stout spitzenkörperMolecular Microbiology 97:1–6.https://doi.org/10.1111/mmi.13044
Endocytosis in filamentous fungi: cinderella gets her rewardCurrent Opinion in Microbiology 13:684–692.https://doi.org/10.1016/j.mib.2010.09.005
Existence of a novel clathrin-independent endocytic pathway in yeast that depends on Rho1 and forminThe Journal of Cell Biology 195:657–671.https://doi.org/10.1083/jcb.201104045
α-Arrestins participate in cargo selection for both clathrin-independent and clathrin-mediated endocytosisJournal of Cell Science 128:4220–4234.https://doi.org/10.1242/jcs.175372
Molecular Cloning: A Laboratory ManualCold Spring Harbour, New York: Cold Spring Harbour Press.
Clathrin localization and dynamics in Aspergillus nidulansMolecular Microbiology 103:299–318.https://doi.org/10.1111/mmi.13557
The flippase DnfB is cargo of fimbrin-associated endocytosis in Aspergillus nidulans, and likely recycles through the late golgiCommunicative & Integrative Biology 9:e1141843.https://doi.org/10.1080/19420889.2016.1141843
Phospholipid flippases DnfA and DnfB exhibit differential dynamics within the A. nidulans spitzenkörperFungal Genetics and Biology 99:26–28.https://doi.org/10.1016/j.fgb.2016.12.007
Functional characterization and localization of the Aspergillus nidulans formin SEPAMolecular Biology of the Cell 13:469–479.https://doi.org/10.1091/mbc.01-07-0356
Cargo recognition during clathrin-mediated endocytosis: a team effortCurrent Opinion in Cell Biology 16:392–399.https://doi.org/10.1016/j.ceb.2004.06.001
Endocytosis and early endosome motility in filamentous fungiCurrent Opinion in Microbiology 20:10–18.https://doi.org/10.1016/j.mib.2014.04.001
Fusion PCR and gene targeting in Aspergillus nidulansNature Protocols 1:3111–3120.https://doi.org/10.1038/nprot.2006.405
MEGA6: molecular evolutionary genetics analysis version 6.0Molecular Biology and Evolution 30:2725–2729.https://doi.org/10.1093/molbev/mst197
Clathrin regulates the association of PIPKIgamma661 with the AP-2 adaptor beta2 appendageJournal of Biological Chemistry 284:13924–13939.https://doi.org/10.1074/jbc.M901017200
Tickets to ride: selecting cargo for clathrin-regulated internalizationNature Reviews Molecular Cell Biology 10:583–596.https://doi.org/10.1038/nrm2751
Adaptor complex-independent clathrin function in yeastMolecular Biology of the Cell 10:3643–3659.https://doi.org/10.1091/mbc.10.11.3643
Randy SchekmanReviewing Editor; Howard Hughes Medical Institute, University of California, Berkeley, United States
In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.
Thank you for submitting your article "The AP-2 complex has a specialized clathrin-independent role in apical endocytosis and polar growth in fungi" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Randy Schekman who served as the Reviewing and the Senior Editor. The reviewers have opted to remain anonymous.
The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.
This manuscript reports a clathrin-independent function for the AP-2 complex in filamentous fungi. In a bioinformatics analysis, they show that the β subunit of AP-2 of A. nidulans, as well as of other fungi, does not contain the C-terminal domain with the clathrin-binding site. They then investigate the role of AP-2 in A. nidulans endocytosis, and find that it is critical for polarized growth, but not for endocytosis of various plasma membrane-resident transporters; it has been known before that clathrin-mediated endocytosis does not necessarily require AP-2. However, they then show that polar localization of particular cargo requires AP-2 but, surprisingly, not clathrin. Furthermore, they show that polarized AP-2 colocalizes with a number of endocytic factors and polarized cargo, but not with clathrin or non-polarized cargo. Thus, they suggest that AP-2 has a (endocytic) role in polarization of the hyphal tip in filamentous fungi, which is independent of clathrin. A clathrin-independent endocytic role for AP-2 is a novel finding of broad interest to cell biologists. Although other AP complexes have been attributed clathrin-independent roles, these are not well understood. The here presented finding can contribute to our understanding of how endocytic adaptors function and of their specificity in cargo selection. Furthermore, it is interesting with regard to the role of clathrin function, especially its dispensability for some endocytic pathways. The manuscript is well written, the data well presented and the experiments appear to be carefully done.
1) One of the major findings is presented in Figure 5; the loss of polar localization of the flippases in absence of AP-2, but not in absence of clathrin. Given the importance of these results, they require quantification, and perhaps a higher number of representative example images. Would it be possible to quantify the fluorescence at the tips vs. fluorescence along the hyphae for mutants and wild type cells, for example by measuring the fluorescence intensities along line profiles at defined distances from the tips?
Throughout the manuscript, the data often consists of single fluorescence images of cell(s) from an experiment. When at all possible, quantification of multiple cells in multiple images should be carried out to support interpretations based on the images presented, as was done in Figure 2. For example, in the analysis of transporter uptake in Figure 3C, the authors conclude that ap2∆ does not inhibit internalization (subsection “AP complexes are dispensable for membrane traffic and endocytosis of transporters”, last paragraph), but there is little apparent difference between control and endocytic trigger images of AgtA-GFP or FurA-GFP, suggesting that little internalization occurred in response to the endocytic trigger in these strains. Quantification of internalization or the level of transporter remaining at the cell surface could address this issue. In Figure 7C-E, the authors conclude that the effects of double mutants on morphology is additive (subsection “Further genetic and cellular evidence that AP-2 is involved in apical lipid maintenance”, first paragraph) but most of the double mutant images resemble the single ap2∆ mutant. Some measurement of morphological phenotype is needed to support the conclusion that mutant effects are additive. In addition, Figure 4 would benefit from quantification of morphology and types of septa to allow comparison to ap2∆ mutants in Figure 2; Figure 5 would benefit from quantitation of the ratio of hyphae in each strain with polarized vs non-polarized protein distribution; Figure 8 would benefit from quantification of polarized distribution of filipin staining.
2) The reviewers are particularly concerned about the functionality of the ClaL-GFP construct and the extent of the depletion of the clathrin heavy and light chains in the experiments with thiamine. The author should know that C-terminal tagging of the clathrin light chain in S. cerevisiae results in a partially miss-functional protein, so they should provide evidence that ClaL-GFP is a fully functional protein. There is the possibility that ClaL depletion does not have a strong effect on triskelia, so they should heavy chain depletion. However, if the heavy chain depletion also does not affect flippase polarization, the clear genetic distinction between the pathways would be sufficient for an initial report.
S. cerevisiae.3) Another major concern is that the evolutionary significance of the findings is somewhat uncertain. The authors propose that clathrin-independent function applies to AP-2 in all dikarya (e.g. Abstract), and suggest that the AP-1 complex in dikarya also functions in a clathrin-independent manner (Discussion, third paragraph). These proposals are based primarily on the phylogenetic analysis in Figure 1, which shows that higher fungal species encode β subunits that are shorter than mammalian β subunits, lacking the folded domains in mammalian β appendages that are involved in binding clathrin and clathrin-accessory proteins. However, there is direct evidence that the S. cerevisaie AP-1 complex binds clathrin through the C-terminal regions of the β and γ subunits and functions in clathrin-dependent transport, even though the β subunit also lacks the folded appendage domains of mammalian β subunits (e.g. Yeung and Payne. 2001. Traffic 2(8):565-76). By extension, other fungal AP complexes with β subunits lacking folded appendage domains could have the potential to interact with clathrin through clathrin binding sequences in the non-conserved β "hinge" region and α appendage domain. Furthermore, S. cerevisiae AP-2 localizes to endocytic sites that are associated with clathrin (Carrol et al. 2009. Dev Cell 17(4):552-60, 2011; Carroll et al., 2012. Mol Biol Cell 23(4):657-68), making it likely that AP-2 functions with clathrin even though it does not exhibit strong clathrin binding (Yeung et al. 1999. Mol Biol Cell 10(11):3643-59). For these reasons, the authors' interpretation of the phylogenetic analysis in Figure 1 is questionable and it is difficult to predict whether clathrin-independent function described for A. nidulans will be a general feature of AP-2 in other higher fungal species. At a minimum, the manuscript should be rewritten to consider the data from S. cerevisiae and moderate any conclusions about other fungi.
4) A key indication for a spatial uncoupling of clathrin- and AP-2-dependent endocytosis is the colocalization data shown in Figure 6. ClCl-GFP (clathrin) does not appear to localize at the plasma membrane. What is seen in the micrograph is probably clathrin on intracellular membranes, which is likely to obscure weaker possible patches on the plasma membrane. However, to support the conclusions made, it is crucial to see where on the plasma membrane clathrin performs its endocytic function. Given that it is involved in transporter uptake like SlaB, one would expect, besides the expected intracellular localization, similar PM-patches for ClcL as for SlaB. Could the authors try to overcome this problem by using different imaging techniques, for example by TIRF microscopy, to visualize the plasma membrane only rather than the intracellular membranes? If not, it may be sufficient to assess colocalization of AP-2 and clathrin at endocytic sites by analyzing higher magnification/resolution images of the apex and collar regions.
The full comments of the reviewers follow. However, only the essential revisions listed above will be required for your revised submission. Additional major questions below may be addressed at your discretion.
This manuscript reports a clathrin-independent function for the AP-2 complex in filamentous fungi. In a bioinformatics analysis, they show that the β subunit of AP-2 of A. nidulans, as well as of other fungi, does not contain the C-terminal domain with the clathrin-binding site. They then investigate the role of AP-2 in A. nidulans endocytosis, and find that it is critical for polarized growth, but not for bulk endocytosis of various plasma membrane-resident transporters. On the other hand, endocytosis of the same transporters appears to be clathrin dependent, which is also important for polarized growth. So far this is not surprising; it has been known before that clathrin-mediated endocytosis does not necessarily require AP-2. However, they then show that polar localization of particular cargo requires AP-2 but not clathrin. Furthermore, the show that polarized AP-2 colocalizes with a number of endocytic factors and polarized cargo, but not with clathrin or non-polarized cargo. Thus, they suggest that AP-2 has a (endocytic) role in polarization of the hyphal tip in filamentous fungi, which is independent of clathrin. A clathrin-independent endocytic role for AP-2 is a novel finding of broad interest to cell biologists. Although other AP complexes have been attributed clathrin-independent roles, these are not well understood. The here presented finding can contribute to our understanding of how endocytic adaptors function and of their specificity in cargo selection. Furthermore, it is interesting with regard to the role of clathrin function, especially its dispensability for some endocytic pathways. The manuscript is well written, the data well presented and the experiments appear to be carefully done. The following points should be addressed:
1) One of the major findings is presented in Figure 5; the loss of polar localization of the flippases in absence of AP-2, but not in absence of clathrin. Given the importance of these results, they require quantification, and perhaps a higher number of representative example images. Would it be possible to quantify the fluorescence at the tips vs. fluorescence along the hyphae for mutants and wild type cells, for example by measuring the fluorescence intensities along line profiles at defined distances from the tips?
2) The involvement of well-characterized endocytic markers in the two proposed pathways should be established more thoroughly, as it is important to know if the two pathways share part of their machinery. The best-characterized fungal endocytic machinery is the one of budding yeast S. cerevisiae. Firstly, does SlaB, which the authors use as endocytic marker, correspond to Sla2 in S. cerevisiae? What is the S. cerevisiae ortholog of SagA?
If SlaB corresponds to Sla2, this would a good marker for endocytosis, especially as it is known to be essential for endocytosis in budding yeast (Kaksonen et al., Cell 2003). Are SlaB and SagA involved in both, the AP-2- and clathrin-dependent pathways in A. nidulans? The colocalization studies may suggest this, but are not very strong evidence. The authors show that UapA internalization depends on SlaB and SagA, but not whether polar localization of the flippases DnfA and DnfB does. They could use the SlaB and SagA deletion/knockdown mutants to test this.
Along the same lines, it is a bit confusing that FM4-64 uptake is neither affected by lack of AP-2, clathrin, SagA nor SlaB. This would actually suggest that SagA and SlaB are not involved in both pathways, or that there is a third pathway for uptake. Can the authors experimentally test these possibilities?
3) A key indication for a spatial uncoupling of clathrin- and AP-2-dependent endocytosis is the colocalization data shown in Figure 6. ClCl-GFP (clathrin) does not appear to localize at the plasma membrane. What is seen in the micrograph is probably clathrin on intracellular membranes, which is likely to obscure weaker possible patches on the plasma membrane. However, to support the conclusions made, it is crucial to see where on the plasma membrane clathrin performs its endocytic function. Given that it is involved in transporter uptake like SlaB, one would expect, besides the expected intracellular localization, similar PM-patches for ClcL as for SlaB. Could the authors try to overcome this problem by using different imaging techniques, for example by TIRF microscopy, to visualize the plasma membrane only rather than the intracellular membranes?
4) The Calcofluor staining in Figure 2D is important, as it shows depolarization of apical endocytosis in the ap2 mutant. However, the images are not entirely convincing, and the conclusions would be much better supported by a quantification similar to that suggested in point 1 (for Figure 5).
This manuscript characterizes the role of the Aspergillus nidulans AP-2 adaptor complex in endocytosis using genetic and localization approaches. The results provide evidence that, in this filamentous fungus, AP-2 functions in cargo-selective apical endocytosis independently of clathrin. This is a novel observation, which suggests for the first time (to my knowledge) that AP-2 has evolved in an organism to function independently of clathrin.
One major concern is that the evolutionary significance of the findings is somewhat uncertain. The authors propose that clathrin-independent function applies to AP-2 in all dikarya (e.g. Abstract), and suggest that the AP-1 complex in dikarya also functions in a clathrin-independent manner (Discussion, third paragraph). These proposals are based primarily on the phylogenetic analysis in Figure 1, which shows that higher fungal species encode β subunits that are shorter than mammalian β subunits, lacking the folded domains in mammalian β appendages that are involved in binding clathrin and clathrin-accessory proteins. However, there is direct evidence that the S. cerevisaie AP-1 complex binds clathrin through the C-terminal regions of the β and γ subunits and functions in clathrin-dependent transport, even though the β subunit also lacks the folded appendage domains of mammalian β subunits (e.g. Yeung and Payne. 2001. Traffic 2(8):565-76). By extension, other fungal AP complexes with β subunits lacking folded appendage domains could have the potential to interact with clathrin through clathrin binding sequences in the non-conserved β "hinge" region and α appendage domain. Furthermore, S. cerevisiae AP-2 localizes to endocytic sites that are associated with clathrin (Carrol et al. 2009. Dev Cell 17(4):552-60, 2011; Carroll et al., 2012. Mol Biol Cell 23(4):657-68), making it likely that AP-2 functions with clathrin even though it does not exhibit strong clathrin binding (Yeung et al. 1999. Mol Biol Cell 10(11):3643-59). For these reasons, the authors' interpretation of the phylogenetic analysis in Figure 1 is questionable and it is difficult to predict whether clathrin-independent function described for A. nidulans will be a general feature of AP-2 in other higher fungal species. At a minimum, the manuscript should be rewritten to consider the data from S. cerevisiae and moderate any conclusions about other fungi.
Other major comments:
1) Throughout the manuscript, the data often consists of single fluorescence images of cell(s) from an experiment. When at all possible, quantification of multiple cells in multiple images should be carried out to support interpretations based on the images presented, as was done in Figure 2. For example, in the analysis of transporter uptake in Figure 3C, the authors conclude that ap2∆ does not inhibit internalization (subsection “AP complexes are dispensable for membrane traffic and endocytosis of transporters”, last paragraph), but there is little apparent difference between control and endocytic trigger images of AgtA-GFP or FurA-GFP, suggesting that little internalization occurred in response to the endocytic trigger in these strains. Quantification of internalization or the level of transporter remaining at the cell surface could address this issue. In Figure 7C-E, the authors conclude that the effects of double mutants on morphology is additive (subsection “Further genetic and cellular evidence that AP-2 is involved in apical lipid maintenance”, first paragraph) but most of the double mutant images resemble the single ap2∆ mutant. Some measurement of morphological phenotype is needed to support the conclusion that mutant effects are additive. In addition, Figure 4 would benefit from quantification of morphology and types of septa to allow comparison to ap2∆ mutants in Figure 2; Figure 5 would benefit from quantitation of the ratio of hyphae in each strain with polarized vs non-polarized protein distribution; Figure 8 would benefit from quantitation of polarized distribution of filipin staining.
2) Based on the images in Figure 6B, the measured levels of co-localization in some cases are somewhat surprising. For example, whereas the patterns of SagA or SlaB are quite similar to that of AP-2, the same is not true of DnfA, DnfB, and SynA. In these cases, the proteins localize at the apex while AP-2 concentrates at the collar. However, the quantified levels of AP-2 colocalization with these proteins are very similar to those of AP-2 with SagA or SlaB (subsection “AP-2 co-localizes with DnfA, DnfB, SagA and SlaB, but not with clathrin or UapA”, third paragraph). Considering the discordance between the images and quantitation, there is some concern that the resolution/quality of the images, and/or the co-localization methodology is not sufficient to distinguish between actual colocalization and diffuse overlap between proteins localized to distinct but nearby sites. The authors should address this issue and also provide more detail about how the colocalization analysis in Figure 6 was carried out, including how the background was set in each case.
3) The pattern of clathrin fluorescence in Figure 6B is distinct from all other plasma membrane and endocytic proteins analyzed, and probably reflects predominant localization at the Golgi, as is observed in S. cerevisiae. There do appear to be weak cortical clathrin puncta at the apex and collar regions, which could represent endocytic clathrin coats. Consequently, it may be more informative to assess colocalization of AP-2 and clathrin at endocytic sites by analyzing higher magnification/resolution images of the apex and collar regions. Alternatively, TIRF microscopy, if possible, might be a more rigorous assay for colocalization of AP-2 with all of the proteins analyzed in Figure 6, including clathrin.
4) The functional significance of the double mutant phenotypes in Figure 7 is not clear. Using deletion mutations, negative genetic interactions could reflect inactivation of functionally distinct pathways or could be due to inactivating two factors in the same pathway that when inactivated alone only partially affect the pathway. The significance of the tested genetic interactions would be more convincing if a mutation in a functionally distinct pathway, such as thiA-claL, does not exhibit genetic interactions with ap2∆.
5) The double mutants in Figure 8B are not referred to in the text. Most resemble the single ap2∆ mutants, so if the authors have included the double mutants to show additive effects, some quantitation is needed. Also, it is curious that the ap2∆ stoA∆ double mutant displays polarized filipin staining similar to wild-type, whereas the single mutants exhibit clear defects. The authors should address this phenotype.
The manuscript by Martzoukou et al. provides evidence suggesting that in the filamentous yeast Aspergillus nidulans: 1) clathrin plays endocytic roles in membrane transporter internalization, which are independent of the AP-2 adaptor; 2) the AP-2 adaptor is required for polarity maintenance and proper hypha growth and; 3) AP-2 plays a role in maintenance of lipid polarity, which is clathrin-independent and which probably implies the clathrin-independent internalization of the lipid flippases DnfA and DnfB. From these three main conclusions, the last one is most novel, provided that AP-2-independent clathrin functions in endocytosis have already been demonstrated in mammalian cells and yeast and that a role of AP-2 in maintenance of polarity has already been demonstrated in C. albicans and S. cerevisiae. The 3rd conclusion is novel and provides an interesting research avenue but the data supporting it is rather preliminary. The third conclusion is mainly supported by three observations: 1) the sequence known to mediate direct clathrin binding is missing from the β2 adaptin subunit of higher fungi; 2) depletion of the clathrin light chain with a repressible promoter does not cause depolarization of the lipid flippases or ergosterol, whereas depletion of the AP-2 µ or σ subunits does and finally; 3) the AP-2 complex barely co-localizes with a ClaL-GFP construct. There are some concerns regarding all these experiments or the conclusions derived from there. For example, even AP-2 might have lost its capacity to directly bind clathrin in dikarya, it might still be indirectly recruited to clathrin pits and vesicles through binding to other endocytic adaptors. Thus, one would need to demonstrate that AP-2 assembled at the plasma membrane do not colocalize at any time (in time lapse movies) with clathrin. Figure 6 intends to address this issue by showing that the degree of co-localization of AP-2 and the clathrin light chain is lower than the co-localization with other endocytic markers. This could be a valid argument if the authors could assure that the ClaL-GFP construct is competent for its endocytic functions. This construct shows a somehow strange pattern showing big intracellular clumps, while staining at the plasma membrane seems to be completely absent. The author should maybe know that C-terminal tagging of the clathrin light chain in S. cerevisiae results in a partially miss-functional protein. Analysis of the co-localization with the endogenous protein or with an N-terminal GFP-ClaL chimera should be included. Also, functional analysis of the GFP and mCherry chimeras used should be provided. Finally, most experiments that show phenotypical differences between the knock out of the AP-2 subunits and clathrin are performed with the constitutive deletions of the AP-2 subunits but with a thiamine repressible construct for the ClaL. It is not clearly defined in the figure legend or the Materials and methods for how long and to what extend the expression of the clathrin light chain is repressed. It might well be that under the conditions used, the ClaL is not fully depleted. Further, the levels of the clathrin heavy chain and its capacity to form triskelions should be analyzed under these conditions. In S. cerevisiae, it is known that constitutive deletion of the clathrin light chain cause strong membrane traffic phenotypes because it results in concomitant depletion of the clathrin heavy chain and its inability to form triskelions. However, most clc1∆ phenotypes can be overcome by overexpression of the heavy chain (including triskelion assembly). It might well be that under the experimental conditions used in this work, enough triskelions are still present in the cell to support endocytosis of the lipid flippases. To provide strong evidence that the DnfA and DnfB flippases are internalized in an Ap-2 dependent and clathrin-independent manner, the authors would need to analyze the localization of the flippases in the constitutive ClaL mutant and in clathrin heavy chain depleted cells. Further, the authors would need to demonstrate that the flippases can be crosslinked to AP-2 but not to clahtrin. Finally, it would be interesting to investigate the localization of DnfA in a double AP-2-mu and ClaL Knock out. It could be that clathrin deficiency reestablished DnfA polarization by compensating an endocytic defect with slower secretion or recycling rates. If this is the case, depletion of clathrin in an AP-2 mutant would restore polarity. If this is not the case, the double mutant should show a defect similar to the AP-2 mutant.
[Editors' note: further revisions were requested prior to acceptance, as described below.]
Thank you for resubmitting your work entitled "The AP-2 complex has a specialized clathrin-independent role in apical endocytosis and polar growth in fungi" for further consideration at eLife. Your revised article has been favorably evaluated by Randy Schekman (Senior and Reviewing editor) and two reviewers.
The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below:
The reviewers would very much like to give the authors another chance to address the weaknesses identified by #3 (full comments below). It should be possible to convincingly show that ClaH is efficiently knocked down in the thiamine experiments that target depletion of ClaH! Quantitative immunoblots should be a good control for that. We are not convinced that endocytic assays such as FM4-64 uptake or LatA treatments are able to discriminate between the clathrin-dependent and the AP-2-dependent pathways, as it is likely that the two share parts of the machinery (It appears that what differs between the two is cargo and localization – both addressed by the authors).
We ask that you do the following:
1) Immunoblots of ClaH in the thiA-ClaH experiments, at the different time points and conditions; +/- thiamine, o/n, 5h and 10h after regrowth).
2) Colocalization of ClaL-RFP with AP2-GFP by TIRF, although, as mentioned before, we do expect a certain degree of colocalization, which does not exclude different functions – but the degree of co- or distinct localization of AP2 and ClaL-RFP on the PM will influence our judgement about the validity of your conclusions.
In the revision of the manuscript, the authors have addressed all relevant points raised by the reviewers. As far as I can judge, they have achieved that in a satisfactory manner. In particular, the fluorescence imaging results have been quantified and the authors have obtained more and clearer experimental evidence for distinct roles and localizations of clathrin and AP-2. Thus, I think the manuscript in its current form is well written, and the conclusions made are adequately backed up by data. Most importantly, the major claim that AP-2 has a clathrin-independent role in A. nidulans is supported by quantitative experimental data. The manuscript is a very interesting addition to the general recent importance given to clathrin-independent pathways in the endocytosis field, as it emphasizes the variability of endocytic pathways that have evolved from similar sets of protein components.
The authors have made an effort to answer several of the points raised by the reviewers but in my opinion, the major concerns have not been addressed satisfactorily. The main point of the manuscript is that AP-2 has a clathrin-independent role maintaining the polarity of DnfA and B localization. In this context, the negative results showing that clathrin does not co-localize with AP-2 and that clathrin does not play a role in the polarization of the lipid flippases are both essential and need to be properly controlled.
Regarding the first point, all reviewers asked for evidence demonstrating the functionality of the ClaL-GFP construct. The authors now add a figure that actually shows that indeed the ClaL-GFP does not fully complement growth. The authors argue that the localization of the ClaL-GFP is fine because the ClaL-GFP pattern resembles somehow the localization of ClaL-mRFP (in terms of showing some clumps in the cytosols) and because the ClaL-mRFP complements the Uap-GFP uptake defect of ClaL mutants. They should then use the ClaL-mRFP to check co-localization with AP-2, not ClaL-GFP (which shows some obvious defects in growth). Even if they use the ClaL-mRFP (that seems to complement better than the ClaL-GFP), most CLaL signal appears in the Golgi and will mask possible co-localization analysis with AP-2 at the plasma membrane. The authors still need to apply TIRF to analyze the co-localization of CLa and AP-2. They only show TIRF to demonstrate that some CLa-GFP seems to be recruited to the plasma membrane. Indeed, in the images shown in Figure 6, one can see a clear co-localization of AP-2 and clathrin in the subapical region, if the putative Golgi signal is subtracted.
With regard to the second point, the authors did not analyzed the extent of depletion of the clathrin heavy and light chains in the CLaL depletion experiments, as requested. This is particularly important given that they cannot used CLaH depletion to show lack of a defect in the DnfA and B localization, since the strain shows major alterations in the secretory pathway upon long thiamine exposure. The experiment depleting CLaH for 5 hours is rather useless because clathrin is a quite stable protein and therefore, most of it, might still be there. The authors would at least need to demonstrate that under the experimental conditions used, other clathrin-dependent functions are significantly affected.
Finally, even though the experiments are maybe not essential to the major point of the article, it is unacceptable that the experiments shown in Figure 3B, C, E and F as well as Figure 7F do not show the wild type control under the same experimental conditions.https://doi.org/10.7554/eLife.20083.024
- George Diallinas
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
We thank Brian Shaw for the DnfA and DnfB strains and Norio Takeshita for the StoA strain. We also thank Spiros Efthimiopoulos for Brefeldin A. This work was supported by the Fondation Santé, to which we are grateful.
- Randy Schekman, Reviewing Editor, Howard Hughes Medical Institute, University of California, Berkeley, United States
© 2017, Martzoukou et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.