Eisosomes and plasma membrane organization

Abstract

Membrane compartmentalization allows the spatial segregation of different functions, such as signal transduction and protein trafficking, and ensures their fidelity and efficiency. Eisosomes constitute nanoscale furrow-like invaginations of the plasma membrane where proteins and lipids segregate. The intense interest elicited by eisosomes over the last few years has led to the identification and molecular characterization of their key constituents. This review addresses eisosome structure, functions and its implications for the mechanistic understanding of curvature-induced membrane nanodomains formation and signaling compartmentalization in living cells.

Appendices

Box 1: Eisosome conservation and organization in other fungi

There is currently no evidence for eisosome presence in organisms outside the fungi kingdom. However, structurally comparable domains called caveolae are found in the plasma membrane of mammalian cells. Caveolae are 60–80 nm flask-shaped stable invaginations formed by scaffolding of the integral membrane proteins caveolins and the membrane-associated proteins cavins (Hansen and Nichols 2010). Like eisosomes, caveolae concentrate sterols and phosphoinositides and have been implicated in plasma membrane stress-mediated signaling and regulation of lipid homeostasis (Bastiani and Parton 2010; Hansen and Nichols 2010). Besides these appealing similarities, the repertoire of known core components and mechanism of assembly of both domains are quite dissimilar (Moreira et al. 2009; Hayer et al. 2010). Eisosomes and caveolae might, therefore, represent a case of convergent evolution.

Conservation of different eisosomal components across all fungi is disparate. Most eisosomal components are present in ascomycetes, Pil1 and Lsp1 and few others are found in basidiomycetes and no eisosomal proteins were found in zygomycetes or chytridiomycetes (Olivera-Couto et al. 2011). A phylogenetic analysis across the ascomycota phylum shows that most eisosomal proteins share a history filled with multiple events of gene duplication and gene loss, a common feature in this phylum (see Online Resource 1). In addition to extensive work in budding yeast, eisosomes are being actively characterized in four other ascomycetes, annotation of eisosomal proteins and their phylogenetic relationships are summarized in Table 2 and Online Resource 1:

Aspergillus nidulans: In ungerminated conidia of this filamentous fungus, eisosome organization is similar to budding yeast: Pil1/Lsp1 orthologs (PilA and PilB) and Sur7 ortholog (SurG) all colocalize at the spore periphery forming a dense net of foci (Vangelatos et al. 2010). In contrast, in actively growing hyphae eisosome organization is markedly different: PilA forms foci, whereas PilB is cytoplasmic and SurG shows vacuolar and endosomal localization. PilA is required for organization of SurG peripheral foci but not of PilB foci. On the other hand, SURG deletion leads to loss of PilB (but not PilA) peripheral foci (Vangelatos et al. 2010).

Ashbya gossypii: Eisosomes in this filamentous fungus share several common features with their S. cerevisiae counterparts: Pil1 and Lsp1 form static and stable foci that are assembled de novo at active sites of cellular growth (Seger et al. 2011). Also, eisosomes organization depends on the presence of Pil1 (but not of Lsp1). Unlike S. cerevisiae, deletion of NCE102 in A. gossypii does not affect eisosome organization, suggesting that there is no connection between Nce102 and Pkh kinases signaling. In addition to loss of eisosome organization, PIL1 deletion in A. gossypii leads to severe reduction of polar surface expansion and formation of abnormal bulged hyphae (Seger et al. 2011). The ortholog of YMR086w/YKL105c, SEG1, is needed to maintain eisosome stability (Seger et al. 2011).

Candida albicans: Studies in this human pathogen, which exhibits both budding yeast and hyphae morphologies, uncovered novel functional roles for Sur7. Like in S. cerevisiae, in C. albicans Pil1, Lsp1, Sur7 and Fmp45 form static foci in budding cells and also in hyphae (Alvarez et al. 2008; Reijnst et al. 2011; Wang et al. 2011). C. albicans eisosomes are absent at the tips of growing hyphae or buds, suggesting that de novo assembly is also restricted to areas of active growth (Reijnst et al. 2011). In contrast to S. cerevisiae, deletion of C. albicans SUR7 ortholog leads to several phenotypes, including lack of septin localization at bud necks, defective growth polarization, ectopic growth of cell wall, defective biofilm formation and, more importantly, decreased virulence in a mouse model of infection (Alvarez et al. 2008; Bernardo and Lee 2010; Wang et al. 2011; Douglas et al. 2012). Since septins control the correct positioning of actin patches and cell wall-synthesizing enzymes, it has been proposed that many of the phenotypes observed in sur7Δ cells are due to defects in septin organization (Alvarez et al. 2008). How Sur7 is implicated in septin organization remains to be elucidated. Notably, SUR7 deletion does not affect Lsp1 foci organization (Alvarez et al. 2008). Currently, there is no published evidence about the phenotypes caused by deletion of PIL1 or LSP1 in C. albicans. Thus, it is currently unknown whether C. albicans eisosomes mediate plasma membrane organization and if they are needed to sustain Sur7 functions.

Schizzosaccharomyces pombe: Both frozen-etch EM and fluorescent microscopy data showed that fission yeast has 1–2 μm long eisosomes, much larger than those present in S. cerevisiae (Moor and Muhlethaler 1963; Streiblova 1968; Takeo 1984; Kabeche et al. 2011). Like S. cerevisiae, eisosomes are formed de novo behind the active sites of growth. Eisosomes are also stable and static (Kabeche et al. 2011). However, during cell division eisosomes are actively removed from the future zone of septation. This clearance includes breakage, disassembly and even directional movement of eisosomes away from the cell division zone (Kabeche et al. 2011). Gene swapping experiments suggest that there is a Pil1-independent mechanism that regulates eisosome assembly in both fission and budding yeasts. In an S. pombe pil1Δ pil2Δ strain, expression of S. cerevisiae PIL1-GFP leads to formation of 1–2 μm long filaments that are indistinguishable from those observed in wild-type fission yeast cells. Conversely, expression of S. pombe Pil1-GFP in a pil1Δ S. cerevisiae strain renders Pil1-GFP puncta that colocalize with endogenous Lsp1 (Kabeche et al. 2011).

Like Nce102 in S. cerevisiae, S. pombe Fhn1 presents a dual localization being dispersed along the plasma membrane and concentrated at Pil1 filaments. Similarly, in pil1Δ cells Fhn1 disperses homogeneously in the plasma membrane and Pil1 filaments are less numerous and shorter in fhn1Δ cells (Kabeche et al. 2011). Surprisingly, S. pombe Slm1 and Sur7 orthologues do not colocalize with Pil1 or depend on Pil1 for their localization stressing the functional divergence that exists in budding and fission yeast’s Pil1 (Kabeche et al. 2011). There is a third, more distantly related Pil1/Lsp1 ortholog, Meu14, which is required for maturation of the forespore membrane (FSM) during sporulation (Okuzaki et al. 2003). The FSM is a double-membrane envelope that extends and engulfs each of the haploid nuclei produced by meiosis during sporulation. The extending FSM adopts a cup-like shape and Meu14 localizes at the rim, which is called the leading edge. In Meu14-depleted cells, the leading edges are abnormally assembled and FSMs are morphologically aberrant (Okuzaki et al. 2003). A high sensitivity search for distant orthologs suggested that, like Pil1 and Lsp1, Meu14 is a BAR domain-containing protein (our unpublished results). Thus, it is tempting to speculate that the molecular function of Meu14 is to maintain the high curvature of the FSM leading edge.

Box 2: Outstanding questions

• Is there a physiological or genetic context in which eisosome domain organization is essential for cell growth?

• Are Pil1 and Lsp1 sufficient to segregate proteins and lipids?

• What are the molecular features that distinguish eisosome targeting of H⁺ -symporters from proteins of the Sur7 and Nce102 families?

• Which are the molecular functions of still uncharacterized eisosomal proteins?

• Which are the Pil1 and Lsp1 phosphatases?

• What keeps eisosomes static?

• How is eisosome assembly controlled?

• Which are the molecular features that distinguish Pil1 from Lsp1?

• Is there an eisosome-dependent endocytic pathway?

• What is the molecular mechanism that releases Slm1 from eisosomes upon plasma membrane stress?