Recent Progress in the Topology, Structure, and Oligomerization of Caveolin: A Building Block of Caveolae

Abstract

Caveolae are cholesterol-rich plasma membrane invaginations that are found in a plethora of cell types. They play many roles including signal transduction, endocytosis, and mechanoprotection. The most critical protein in caveolae is the integral membrane protein, caveolin, which has been shown to be necessary for caveolae formation, and governs the major functions attributed to caveolae. Caveolin is postulated to act as a scaffold in the high molecular weight striated coat that surrounds the caveolar bulb, stabilizing it. Caveolin interacts, both directly and indirectly, with a large number of signaling molecules, and presides over the endocytosis of molecular cargo by caveolae. However, many of the key biophysical aspects of the caveolin protein, its structure, topology, and oligomeric behavior, are just beginning to come to light. Herein is an up-to-date summary and critique of the progress that has been made in understanding caveolin on a molecular and atomic level.

Introduction

Caveolae are 50–100 nm invaginations in the plasma membrane that are found in many cell types such as adipocytes, endothelial cells, fibroblasts, and smooth muscle cells (Mobley and Eisenberg, 1975, Palade, 1953, Thorn et al., 2003, Wheaton et al., 2001). Using electron microscopy, caveolae were first observed in the 1950s as flask-shaped domains within the heart and also the gall bladder (Palade, 1953, Yamada, 1955). Caveolae are proposed to have three major functions in the cell: mechanoprotection; recruiting, concentrating, and effecting signaling molecules; and endocytosis (Chang et al., 1994, Lisanti et al., 1994, Liu et al., 1997, Parton et al., 1994). Caveolae act as membrane reservoirs that allow cells to sustain mechanical stresses (Sinha et al., 2011). Facilitated by their highly curved nature, caveolae can bud off from the membrane in an endocytic manner and form vesicle-like structures that can carry cargo from the cell membrane to other parts of the cell (Kiss, 2012).

At the center of all of these functions is the integral membrane protein caveolin. Initially, there were two key proteins that were associated with caveolae, VIP-21 and caveolin. Caveolin was first isolated as a marker protein for caveolae, while VIP-21 was identified as part of the trans-golgi network (Kurzchalia et al., 1992, Rothberg et al., 1992). Shortly afterward, VIP-21 was sequenced and shown to be the same protein as caveolin (Glenney, 1992). This led to the consensus of calling the protein VIP-21/caveolin and now the protein is referred to simply as caveolin.

Caveolin interacts and effects, both directly and indirectly, a large number of signaling molecules including Src-like kinases, Ha-Ras, endothelial nitric oxide synthase (eNOS), and G-proteins (Carman et al., 1999, Couet et al., 1997, Ju et al., 1997, Lisanti et al., 1994, Okamoto et al., 1998, Roy et al., 1999, Sargiacomo et al., 1993). One well-studied interaction is between caveolin (specifically caveolin-1) and eNOS (García-Cardeña et al., 1997, Ghosh et al., 1998, Ju et al., 1997). The disruption of the binding between caveolin-1 and eNOS has negative effects on cellular function by causing a buildup of nitric oxide within the cell, as caveolin acts as a negative regulator of eNOS (Feron, Saldana, Michel, & Michel, 1998). Initially, many of the caveolin-signaling protein interactions were proposed to occur directly through a consensus sequence dubbed the caveolin-binding motif (CBM). However, recent structural data and bioinformatic analyses suggest that this motif (CBM) is not indicative of direct caveolin-partner interactions (Byrne et al., 2012, Collins et al., 2012). Although these studies call into question the direct interaction of caveolin with signaling molecules through a well-defined consensus motif, they do not preclude a specific nonconsensus interface for caveolin-partner interactions. Clearly, more work will be needed to establish which caveolin-partner interactions are direct versus indirect.

Caveolin-1 has been shown to play an integral role in regulating and mediating the endocytosis of caveolae. For example, the overexpression of caveolin-1 appears to inhibit the endocytosis of caveolae (Le, Guay, Altschuler, & Nabi, 2002). Caveolin-1 has also been implicated in the endocytosis of virus particles through caveolae (Pelkmans, Kartenbeck, & Helenius, 2001). Clearly, caveolin is involved in how molecules are transported from the plasma membrane to other parts of the cell through non-clathrin-mediated pathways.

Caveolae allow for a cell to experience abrupt mechanical stress without being deformed or lysed (Sinha et al., 2011). For example, when a cell is subjected to mechanical force, caveolae rapidly dissipate to allow the cell to accommodate the additional force. When the force is removed, caveolae are reformed.

Caveolin is an approximately 22 kDa protein that can be described by four domains: the N-terminal domain, the scaffolding domain, the intramembrane domain (IMD), and the C-terminal domain (Figure 1). Caveolin has three isoforms, caveolin-1, -2, and -3 (Figure 2). Caveolin-1 is the most ubiquitous, being found in most terminally differentiated cell types, and the protein has been shown to be necessary for caveolae formation (Drab et al., 2001, Park et al., 2002, Razani et al., 2001, Razani et al., 2002). Caveolin-1 also has two specific isoforms referred to as caveolin-1α and caveolin-1β. The caveolin-1β isoform can be thought of as a truncated construct of the α-isoform lacking the first 31 amino acid residues (Scherer et al., 1995). Caveolin-2 is coexpressed with caveolin-1 and the two are thought to form heterooligomeric complexes that are part of the striated coat (Scherer et al., 1996). However, caveolin-2 appears to lack the ability to form caveolae in the absence of caveolin-1 (Mora et al., 1999, Parolini et al., 1999). Caveolin-3 is expressed mostly in muscle tissue and also has the ability to form caveolae (Song et al., 1996, Tang et al., 1996, Way and Parton, 1995). This is supported by the high degree of sequence identity between caveolin-1 and -3. In contrast, the sequence of caveolin-2 deviates from that of the other isoforms, particularly in its scaffolding domain (Figure 2). In addition, the lengths of the N-terminal domains vary significantly between the isoforms, with caveolin-1 having the longest and caveolin-3 having the shortest (Figure 2).

Caveolin-1 has been shown to be palmitoylated at all three of its cysteine residues located in or near the C-terminal domain (C133, C143, and C156). The removal of this posttranslational modification by mutation of the cysteines to serines has been shown to have no effect on the trafficking of caveolin-1 to the plasma membrane (Dietzen, Hastings, & Lublin, 1995). Caveolin-2 has five cysteines and it has recently been shown that three of its cysteines are palmitoylated (Kwon, Lee, Jeong, Jang, & Pak, 2015). Caveolin-3 contains nine cysteines and has been shown to be heavily palmitoylated, but the exact sites have not been experimentally verified (Galbiati, Volonte, Minetti, Chu, & Lisanti, 1999). However, because of the sequence similarity to caveolin-1, it has been assumed that caveolin-3 is palmitoylated at the analogous three sites (Figure 2) (Dietzen et al., 1995, Galbiati et al., 1999).

The composition of caveolae is distinct compared to the bulk membrane because it is enriched in cholesterol and sphingomyelin (Ortegren et al., 2004). This has led to the postulation that caveolae have significantly less fluidity than the bulk membrane, a composition not unlike that of lipid rafts. The relationship between cholesterol and caveolin appears to be very strong and has been implicated in the proper folding and behavior of caveolin (Murata et al., 1995). In particular, the scaffolding domain of caveolin contains a CRAC (cholesterol recognition amino acid consensus) motif that is purported to directly bind cholesterol (Epand, Sayer, & Epand, 2005).

Mutations in the caveolin proteins as well as cellular expression levels have been linked to many diseases such as cancer, diabetes, cardiovascular disease, artherosclerosis, pulmonary fibrosis, and many muscular degenerative diseases (Capozza et al., 2005, Cohen et al., 2003, Fernández-Hernando et al., 2010, Komers et al., 2006, Sotgia et al., 2005, Weiss et al., 2008, Williams et al., 2004, Xia et al., 2010). The literature on the links between caveolin and disease is extensive and is beyond the scope and purpose of this review; therefore, only a few examples will be highlighted (For extensive reviews of caveolin and disease, see Engelman et al., 1998, Razani and Lisanti, 2001; Williams & Lisanti, 2004, and others). The knockout of caveolin-1 in mice can cause an increase in tumor proliferation indicating that caveolin-1 is a tumor suppressor (Sloan et al., 2004, Williams et al., 2003). However, in prostate cancer it is found that caveolin-1 is upregulated and that it assists in the metastasis of the cancer (Williams et al., 2005). This dual role of caveolin-1 in cancer highlights that normal levels of caveolin-1 expressed within specific cell types is critical for maintaining proper cell function. Other studies have also shown that the overexpression of caveolin-1 can mirror the phenotype of a putative disease mutant (P132L) implicated in breast cancer by causing a disruption of caveolin migration from the Golgi to the plasma membrane (Bonuccelli et al., 2009, Hanson et al., 2013). However the true role and extent that this mutation appears in affected patients is unclear (Koike et al., 2010, Lacroix-Triki et al., 2010). Caveolin-3 has been linked to many muscular diseases such as heart disease and muscular dystrophy (Horikawa et al., 2008, Weiss et al., 2008). P104 in caveolin-3 which is analogous to P132 in caveolin-1 is also implicated in increased disease states and misregulation of the protein (Ohsawa et al., 2004, Stoppani et al., 2011).

Despite caveolin's preeminent role in cell homeostasis, many of the biophysical details surrounding caveolin function remain elusive. Much like other proteins whose improper function has been linked to disease states, a complete understanding of the determinants of cellular disruption will only come to fruition once a clear picture of the caveolin protein is obtained in molecular and atomic detail. This review will strive to report the progress that has been made in the field of caveolin structure, topology, and oligomerization and to highlight the work that still needs to be done.