For many years now, several studies in steroid-producing systems have led to the postulation of an induced stimulatory protein originally designated as the “labile protein factor,” since after stimulation with ACTH in the presence of a protein synthesis inhibitor as cycloheximide (CHX), the rate of steroidogenesis rapidly returned to its unstimulated level (Haynes et al. 1959; Garren et al. 1965; Schulster et al. 1970).
Using two-dimension gel electrophoresis, one of the first papers in the matter identified a protein of 28,000 Da, present as different isoforms in hormone-stimulated adrenal cells but not in quiescent cells. Furthermore, 35S-methionine pulse-chase experiments showed that this protein is not produced from a preexisting one, even if protein synthesis is not inhibited (Krueger and Orme-Johnson 1983).
After that work, Pon, Hartigan, and Orme-Johnson showed in 1986 that these forms of the approximately 28 kDa protein, which appear after ACTH stimulation in adrenal cells, correspond to two phosphorylation species of the protein (Pon et al. 1986). The pool of experiments performed by Alberta and coworkers demonstrated that these phosphorylated proteins are localized predominantly in the mitochondria and are tightly associated with this organelle. They also found that the inhibition of mitochondrial protein synthesis affects neither the accumulation of these proteins nor the stimulation of steroidogenesis. Thus, the protein and its phosphorylated counterpart are synthesized in the cytosol and transported to the mitochondria (Alberta et al. 1989).
Years later, studies made by Stocco and colleagues in MA-10 cells, a murine Leydig tumor cell line, showed that these mitochondrial proteins are synthesized in a dose-dependent manner of both trophic chorionic gonadotropin hormone (hCG) and Bt2-cAMP (a permeant analogue of cAMP) and that they are sensitive to CHX inhibition (Stocco and Sodeman 1991). A crucial paper from Clark and Stocco showed the purification, cloning, and expression of this labile factor and called this protein StAR, as an acronym for steroidogenic acute regulatory protein (Clark et al. 1994).
StAR Structure and Mechanism of Action
Full-length human StAR is expressed as a 285-residue protein with a molecular mass of 37 kDa having a mitochondrial leader sequence that is cleaved to yield a 30 kDa protein upon entering the mitochondria to the matrix, where StAR is degraded (Clark et al. 1994; Lin et al. 1995). Deletion of 62 N-terminal residues yields a protein that remains in the cytoplasm yet retains full biologic activity (Arakane et al. 1996).
Besides the mitochondrial leader peptide, an evolutionarily conserved module of approximately 200 amino acids implicated in lipid/sterol binding is found in StAR and other related proteins. This motif is called steroidogenic acute regulatory protein (StAR)-related lipid transfer (START) with an α/β helix-grip structure, consisting of nine-stranded twisted antiparallel β-sheet and four α-helices. The most striking feature of the START domain structure is a predominantly hydrophobic tunnel extending 26 Å, nearly the entire length of the protein that allows the binding and entrance of one molecule of cholesterol in StAR protein (Tsujishita and Hurley 2000; Lavigne et al. 2010).
The three main classes of START domains – classical (CSD, from mammals), birch antigen (BA, from plants), and bacterial (BAC) – share a common topology: a C-terminal α-helix packed against a core β-sheet provides support for a hydrophobic tunnel, which has been shown to accommodate lipid molecules (Lee et al. 2010).
Group 1 contains the name-giving family member, steroidogenic acute regulatory protein (StAR/STARD1), as the most intensively studied of this subclass, and STARD3. Inspection of the ligand cavity of ligand-free STARD1 proposed glutamic 169, arginine 188, leucine 199, and histidine 220 as key residues in cholesterol binding. These side chains will likely change conformation upon ligand binding. Cholesterol binding to STARD1 involves a hydrogen bond between the cholesterol hydroxyl and either the arginine 188 side chain or the backbone carbonyl of leucine 199. Either of these ligand-binding modes is consistent with the present STARD1 crystal structure (Thorsell et al. 2011).
It is undoubtedly demonstrated that StAR activity is obligatory in the passage of cholesterol from the outer mitochondrial membrane (OMM) to the inner membrane, where cytochrome P450 side-chain cleavage resides rendering cholesterol to pregnenolone. This is the well-described rate-limiting step in total steroid synthesis in all steroidogenic tissues (Stone and Hechter 1954; Simpson 1979; Privalle et al. 1983).
The mechanism by which StAR can accomplish cholesterol transport from the cytoplasmic side of the OMM to the intermembrane space is not completely understood, and the search for this issue has led to a great number of publications in the field.
One model suggests that the mechanistic model of StAR action involves a partially unfolded active form of StAR, with the N-terminal domain entering the mitochondria and a partially unfolded C-terminal interacting with the OMM. Direct evidence has been presented showing that StAR exists as a molten globule. While certain native structure is retained at the N-terminal domain, the C-terminal domain folding appears to be less tight at the low pH that StAR may undergo on the mitochondrial membrane (Yaworsky et al. 2005). Then, the tightly folded N-terminal domain could make StAR stops as it enters the mitochondria, extending the time window for the C-terminus to transport cholesterol.
Elegant work from Miller’s group, by means of import experiments using a modified leader peptide (N-62 StAR), demonstrated that mitochondrial StAR protein exclusively acts on the OMM, with no activity when localized to the intermembrane space or to the matrix, as reflected by a negative correlation between the time of StAR mitochondrial entry and its activity (Bose et al. 1998). Thus, N-62 StAR associated to the OMM could transport several cholesterol molecules before entering mitochondria to be processed.
Regulation of StAR Expression
The transcriptional regulation of the Star gene is mediated by multiple upstream DNA elements with recognition motifs for a myriad of sequence-specific transcription factors. To date, many regulatory elements have been found in the StAR promoter which has been identified by the presence of their consensus-binding sequences.
In general, regulation of the StAR gene is highly complex and is subjected to both positive and negative regulation. StAR expression is upregulated by luteinizing hormone, insulin, chorionic gonadotropin, pregnant mare’s serum gonadotropin, adrenocorticotropin, angiotensin II, growth hormone, oxysterols, estradiol, and calcium and potassium ions (Cherradi and Capponi 1998).
The disruption of StAR expression is provoked by several effectors including lipopolysaccharide, actinomycin D, heat shock, overexpression of DAX-1 protein (dosage-sensitive sex reversal, adrenal hypoplasia congenita x-linked), tumor necrosis factor, and pesticides.
It has been demonstrated that the protein kinase C (PKC) pathway plays an essential role in IGF-1-mediated StAR expression and steroid synthesis (Stocco et al. 2005).
One of the most important regulators of StAR expression is the steroidogenic factor 1 (SF-1). Knockdown of SF-1 induced lipid accumulation in Leydig cells by the reduction of StAR expression triggering an increase in unmetabolized cholesterol (Hatano et al. 2016).
StAR expression is also regulated by lipidic molecules as leukotrienes, which are produced through the metabolism of arachidonic acid (AA) (Wang et al. 2001), mediated by the activity of two enzymes, an acyl-CoA synthetase 4 and an acyl-CoA thioesterase. These two enzymes act in a concerted manner to perform a hormone-dependent cycle of mitochondrial AA generation/exportation in steroidogenic tissues (Castillo et al. 2006; Duarte et al. 2007). This particular system acts under different hormonal stimuli in steroidogenic cells, since it has been demonstrated in cAMP-dependent (Maloberti et al. 2005) as well in cAMP-independent signal transduction pathways (Mele et al. 2012).
Several reports have studied the role of MAP kinases (MAPKs) in the regulation of steroid synthesis, both at genomic and non-genomic regulation levels. One of the first reports indicates that cAMP-induced steroid synthesis is dependent upon the phosphorylation and activation of ERK, one of the most studied MAPK family members. ERK activation results in enhanced phosphorylation of SF-1 and increased steroid production through increased transcription of the Star gene. These findings correlate with an increase in StAR mRNA levels, StAR protein accumulation, and steroidogenesis (Gyles et al. 2001). In addition, ERK activity induces Nr4a1 gene expression, which codes for the transcription factor NUR77 required for StAR cAMP-dependent expression in steroidogenic cell lines (Mori Sequeiros Garcia et al. 2015).
It is well known that StAR expression requires cAMP-dependent protein kinase (PKA) activity. Nevertheless, the promoter region of Star gene lacks an easily recognizable cAMP response element for the binding of the cognate-binding protein (CREB), suggesting that StAR-induced expression by PKA occurs through other noncanonical transcription factors.
Interestingly, mitochondrial fusion involved in steroidogenesis through the induction of mitofusin 2 participates in StAR mRNA expression and localization of this protein in mitochondria (Duarte et al. 2012).
Regulation of StAR Function: Phosphorylation
Protein phosphorylation is an integral component of signal transduction pathways within eukaryotic cells and is regulated by the fine interplay of protein kinases and phosphatases. Numerous cellular responses are regulated by the reversible phosphorylation of serine, threonine, and tyrosine residues.
The StAR protein contains two consensus sequences for phosphorylation catalyzed by PKA, which are conserved across all species in which the amino acidic sequence of StAR has been determined, namely, serine 55/56/57 and serine 194/195. Particularly in human StAR, the sites comprise serine 57 and 195, and the phosphorylation of serine 195 by PKA has been identified as an obligatory event in the regulation of human StAR function (Arakane et al. 1997; Fleury et al. 2004).
LeHoux and coworkers have shown that mutations in serine 55, 56, and 194 of hamster StAR dramatically decrease the incorporation of phosphate, as assessed by antibody against phospho-(serine/threonine) PKA substrate site and that they correlate with the disappearance of lower pI StAR species observed by two-dimension electrophoresis. Then, the relevance of these PKA consensus phospho-sites on basal and stimulated StAR activity was certainly assessed (Fleury et al. 2004).
The mutation of serine 195 to alanine (S195A) resulted in an approximately 75% reduction in pregnenolone production with the remaining activity independent of this phospho-site. Taken together, these data demonstrate that phosphorylation of the 194/195 serine residue of StAR may account, at least in part, for the immediate increase in cholesterol side-chain cleavage as a result of enhanced activity of the StAR protein.
By now, StAR phosphorylation has been proven undoubtedly necessary for the activity and functional relevance of StAR protein in the mitochondria, and its significance has been clearly established. In the same line of studies, work from Stocco’s group explored the mechanism of action of the ERK signaling pathway in regulating StAR expression and steroidogenesis in conjunction with PKA and PKC pathways (Manna et al. 2007). This work included the analysis of StAR phosphorylation under different stimuli in MA-10 Leydig cells. The activation of PKC markedly increased StAR expression but not phospho-StAR, with only a modest increase in progesterone production. On the other hand, Bt2-cAMP-activated PKA provoked a substantial augmentation in the band of StAR phosphorylated in serine 194.
Examining the StAR protein structure, a consensus sequence was found that might allow the docking of StAR protein to ERK1/2, as well as a consensus site for ERK1/2 phosphorylation. A typical docking site known as the D domain (KTKLTWLLSI) was found between amino acids 235 and 244. This site is conserved among ERK1/2 upstream kinase MEK, ERK substrates, and its phosphatases (MKPs) (Zhou et al. 2006). Although the database does not reveal ERK as a possible kinase of StAR, the manual alignment of StAR sequence indicates that there are two serine-proline motifs, targets for ERK phosphorylation, detectable in the mature form of the murine StAR protein at serine 232 and serine 277. In accordance with the database Expasy Prosite, serine 232 has a 90% probability of being phosphorylated by ERK and is adjacent to the docking D domain, while the probability of serine 277 is only 5%. Besides, serine 277 is relatively less conserved among species. Also, a canonical binding site included in ERK1/2 substrates in StAR has been found (Poderoso et al. 2008).
StAR phosphorylation by ERK has been shown both in vitro and in vivo, and it is strictly dependent on the presence of cholesterol in the environment of the reaction (Poderoso et al. 2008; Poderoso et al. 2009). This result indicates that StAR phosphorylation by kinases other than PKA is modulated by the endogenous ligand of the protein. In agreement with these results, Baker and colleagues have shown that StAR phosphorylation by PKA is neither necessary nor dependent on the presence of cholesterol in the reaction media (Baker et al. 2007).
It was demonstrated that ERK directly interacts with StAR, it phosphorylates this protein in the presence of cholesterol, and it is suggested that this phosphorylation is required for StAR appropriate association to the OMM (Duarte et al. 2014). Up to date, no phosphatases have been described to dephosphorylate StAR protein. It is conceivable to think that StAR phospho/dephosphorylation could be a hormone-dependent cycle to allow several cholesterol molecules to be transported to the mitochondria per unit of StAR.
Taken together, these results give a new insight into StAR as a substrate of several kinases involved in steroid biosynthesis. It could be concluded that the additional negative charges in the StAR molecule due to phosphorylation at serine 232 may contribute to the retention of the mature form of StAR in the OMM, e.g., through the interaction with the voltage-dependent anionic channel (VDAC) or with cholesterol.
Implications of StAR in Diseases
StAR has been implicated in the development of one serious endocrine disease, congenital lipoid adrenal hyperplasia (lipoid CAH), the most fatal form of adrenal hyperplasia, seriously disrupts adrenal and gonadal steroidogenesis by a defect in the conversion of cholesterol to pregnenolone. Affected patients show salt loss from impaired mineralocorticoid and glucocorticoid synthesis and typically present signs of severe adrenal failure in early infancy; 46,XY genetic males are phenotypic females due to disrupted testicular androgen secretion. Given the relevance of this disease and the fact that novel StAR variants arise periodically, bibliography databases continuously show papers on the field (Albarel et al. 2016; Huang et al. 2016; Kaur et al. 2016). Lipoid CAH has been reported in most ethnic groups but is common among the Japanese, Korean, and Palestinian-Arab populations. To date, 48 different mutations in the StAR gene have been reported in various ethnic groups (http://www.hgmd.org/). Genetic clusters consistently contain the p.Q258X mutation in the Japanese and Korean populations, the p.R182L mutation in Palestinian Arabs, the p.R182H mutation in eastern Saudi Arabians, and the p.L260P mutation in the Swiss population (Kim 2014).
In addition, other tissues apart from steroidogenic ones may be affected by StAR alterations. For example, StAR is expressed in monocytes, macrophages, and human aorta, albeit at levels far lower than those expressed in adrenal or gonadal tissues. The expression of StAR is regulated by inflammatory cytokines, increased during macrophage differentiation and following exposure to liver X receptor α (LXRα), peroxisome proliferator-activated receptor γ (PPARγ), and retinoid receptor (RXR) agonists but repressed following exposure to oxidized or acetylated LDL. Notably, overexpression of StAR impacts on macrophage lipid and inflammatory phenotype, promoting cholesterol efflux to apoA-I and/or HDL; moreover, tail vein administration of a viral vector expressing StAR to apoE(−/−) mice reduced aortic lipids and atheroma (Graham 2015).
However, overexpression of StAR was also associated with increased lipogenesis in macrophages, presumably via induction of Srebp1c; this side effect may limit the therapeutic usefulness of targeting StAR to limit macrophage lipid accumulation and atherogenesis. It is also clear that under pathophysiological conditions involving oxidative stress, macrophage StAR can transfer pro-oxidant cholesterol hydroperoxides to mitochondria, triggering lipid peroxidation and membrane depolarization.
Once imported into the mitochondrial matrix, the accumulation of inactivated 30 kDa StAR must undergo rapid degradation in order to avoid organelle damage. The presence of undergraded StAR excess in this compartment as an apparent “nuisance protein” triggers a novel mitochondria to nucleus signaling event, named the “StAR overload response (SOR),” which activates nuclear factor-like 2 (NRF-2)-dependent genes transcription encoding the mitochondrial proteases needed to degrade StAR. In steroidogenic tissues, this process is regulated by the inner membrane potential, and mediated by mitochondrial proteases, including Lon protease, and the inner membrane complexes of several metalloproteases (Granot et al. 2007). The Lon protease is involved in degradation of oxidized proteins within the mitochondrial matrix, which is essential for maintaining mitochondrial homeostasis (Graham 2015).
StAR is a unique protein that binds and transports cholesterol, from the outer to the inner mitochondrial membrane, participating in the rate-limiting step of the biosynthesis of all the steroids molecules, as well as bile salts in liver. StAR is undoubtedly a key regulatory protein in steroidogenesis, depicted by the fact that there is no other protein that could replace it or exert its function (with one exception in placenta). The structure of StAR is also quite particular since it presents a hydrophobic pocket fitting just one molecule of cholesterol. Nevertheless, StAR protein remains active as long as it is attached to the OMM probably transferring more than one cholesterol molecule per unit of StAR.
StAR expression is subjected to numerous regulatory pathways, induced by trophic hormones in steroidogenic tissues; a great variety of transcription factors modulate StAR’s level in a positive or negative manner. Mitochondrial fusion exerts a transcriptional regulation through the presence of mitofusin 2 by a yet unknown mechanism.
One of the most important and first-described regulatory mechanisms is phosphorylation of the protein. StAR is phosphorylated by PKA in two serine residues, being the serine nearest to the C-terminus completely necessary for StAR activity. Although this phospho-residue does not confer StAR complete activity since mutated StAR in this serine still remains some residual activity. This issue was later clarified when StAR was found to be substrate of ERK1/2, which directly phosphorylates StAR in mitochondria leading to StAR full activation along with PKA.
The relevance of StAR activity is embodied by the fact that several and different mutation on its sequence provoke one of the most lethal endocrine pathologies, lipoid CAH. But StAR presence and function is not just circumscribed to steroidogenic tissues but also to macrophages, skin, lung, adipose tissue, liver, endothelium, and diverse cancer cell lines, opening a wide spectrum of possibilities in StAR studies.
- Arakane F, Sugawara T, Nishino H, Liu Z, Holt JA, Pain D, et al. Steroidogenic acute regulatory protein (StAR) retains activity in the absence of its mitochondrial import sequence: implications for the mechanism of StAR action. Proc Natl Acad Sci USA. 1996;93:13731–6.PubMedPubMedCentralCrossRefGoogle Scholar
- Castillo AF, Cornejo Maciel F, Castilla R, Duarte A, Maloberti P, Paz C, et al. cAMP increases mitochondrial cholesterol transport through the induction of arachidonic acid release inside this organelle in Leydig cells. FEBS J. 2006;273:5011–21. doi: 10.1111/j.1742-4658.2006.05496.x.PubMedCrossRefGoogle Scholar
- Duarte A, Castillo AF, Castilla R, Maloberti P, Paz C, Podesta EJ, et al. An arachidonic acid generation/export system involved in the regulation of cholesterol transport in mitochondria of steroidogenic cells. FEBS Lett. 2007;581:4023–8. doi: 10.1016/j.febslet.2007.07.040.PubMedCrossRefGoogle Scholar
- Maloberti P, Castilla R, Castillo F, Cornejo Maciel F, Mendez CF, Paz C, et al. Silencing the expression of mitochondrial acyl-CoA thioesterase I and acyl-CoA synthetase 4 inhibits hormone-induced steroidogenesis. FEBS J. 2005;272:1804–14. doi: 10.1111/j.1742-4658.2005.04616.x.PubMedCrossRefGoogle Scholar
- Men Y, Fan Y, Shen Y, Lu L, Kallen AN. The steroidogenic acute regulatory Protein (StAR) is regulated by the H19/let-7 axis. Endocrinology. 2017;158(2):402–9. doi: 10.1210/en.2016-1340.
- Poderoso C, Converso DP, Maloberti P, Duarte A, Neuman I, Galli S, et al. A mitochondrial kinase complex is essential to mediate an ERK1/2-dependent phosphorylation of a key regulatory protein in steroid biosynthesis. PLoS One. 2008;3:e1443. doi: 10.1371/journal.pone.0001443.PubMedPubMedCentralCrossRefGoogle Scholar
- Rone MB, Midzak AS, Issop L, Rammouz G, Jagannathan S, Fan J, et al. Identification of a dynamic mitochondrial protein complex driving cholesterol import, trafficking, and metabolism to steroid hormones. Mol Endocrinol. 2012;26:1868–82. doi: 10.1210/me.2012-1159.PubMedPubMedCentralCrossRefGoogle Scholar
- Schulster D, Tait SA, Tait JF, Mrotek J. Production of steroids by in vitro superfusion of endocrine tissue. 3. Corticosterone output from rat adrenals stimulated by adrenocorticotropin or cyclic 3′,5′-adenosine monophosphate and the inhibitory effect of cycloheximide. Endocrinology. 1970;86:487–502. doi: 10.1210/endo-86-3-487.PubMedCrossRefGoogle Scholar