Senescence Marker Protein 30: Functional and Structural Insights to its Unknown Physiological Function

Abstract

Senescence marker protein 30 (SMP30) is a multifunctional protein involved in cellular Ca²⁺ homeostasis and the biosynthesis of ascorbate in non-primate mammals. The primary structure of the protein is highly conserved among vertebrates, suggesting the existence of a significant physiological function common to all mammals, including primates. Enzymatic activities of SMP30 include aldonolactone and organophosphate hydrolysis. Protective effects against apoptosis and oxidative stress have been reported. X-ray crystallography revealed that SMP30 is a six-bladed β-propeller with structural similarity to paraoxonase 1, another protein with lactonase and organophosphate hydrolase activities. SMP30 has recently been tied to several physiological conditions including osteoporosis, liver fibrosis, diabetes, and cancer. This review aims to describe the recent advances made toward understanding the connection between molecular structure, enzymatic activity and physiological function of this highly conserved, multifaceted protein.

Introduction

The designation of senescence marker protein 30 (SMP30) as a marker of senescence or aging implies that its cellular levels would increase with aging, when, in fact, the opposite is true. Analysis of age-associated changes in the expression of soluble liver proteins in rats showed that expression of SMP30 decreases androgen-independently with age. SMP30 comprises 2% of soluble rat liver proteins and its levels decrease by 60–70% in aged rats (1). The molecular weight of SMP30 was initially estimated as 30 kD, and hence the 30 in its name. More accurate measurements and cloning of SMP30 later revealed that the molecular weight is actually 34 kD (2). The gene for SMP30, located on the X-chromosome, consists of 7 exons, and has an open reading frame of 897 bp coding for 299 amino acids (3). While it is primarily found in liver and kidney tissues, low amounts of SMP30 have also been detected in the lungs, ovaries, testes, epidermis, stomach, brain, mammary glands, prostate, epididymis, and seminal vesicles using both immunostaining and mRNA analysis (2–7). Furthermore, the amino acid sequence of SMP30 is highly conserved (70–90% identity) in vertebrates, which is suggestive of an important, even essential, biological function.

Although Fujita et al. (2) were the first to isolate cDNA clones and determine the amino acid sequence of SMP30, this protein was independently “discovered” in the soluble fraction of rat liver homogenates four separate times. Interestingly, the four independent discoveries correspond to four different proposed functionalities: Yamaguchi et al. (8) claimed to have found a novel Ca²⁺-binding protein involved in Ca²⁺ homeostasis, Fujita et al. (1) suggested it was an anti-aging molecule protecting cells from apoptosis and oxidative stress, Billecke et al. (9) reported the isolation of an organophosphate (OP) hydrolase, and Lehninger (10) reported the identification of an aldonolactonase thought to be involved in ascorbate biosynthesis. Isolation and sequencing of cDNA clones ultimately revealed that these four distinct functions could all be ascribed to a single gene product (11).

Over a decade before the reports of SMP30 as an aging marker, Yamaguchi et al. (8) named this rat liver protein regucalcin and proposed that it was a novel Ca²⁺-binding protein involved in the regulation of free cellular Ca²⁺ concentrations. Unlike other Ca²⁺-binding proteins, regucalcin lacks a known Ca²⁺-binding domain, such as an EF-hand motif. The Ca²⁺-binding ability of regucalcin/SMP30 was called into question when later studies failed to detect binding of ⁴⁵Ca²⁺ and indicated that the enzyme’s ability to hydrolyze diisopropyl phosphorofluoridate (DFP) was not dependent on Ca²⁺(12). However, later kinetics experiments performed in our lab using highly purified human SMP30 overexpressed from E. coli revealed that SMP30 does exhibit lactonase activity in the presence of high concentrations of Ca²⁺ such as those that would be seen in cells under stress (13). It is believed by some that SMP30 affects Ca²⁺ homeostasis indirectly through interaction with calmodulin and membrane Ca²⁺ pumps (14–16). This role of SMP30 in Ca²⁺ signaling and homeostasis has been reviewed by Yamaguchi (16, 17) and will not be covered in depth here.

The two remaining “discoveries” of SMP30 revealed enzymatic functions of the protein. SMP30 is capable of hydrolyzing DFP and other OPs such as sarin, soman, and tabun in the presence of divalent cations such as Mg²⁺ and Mn²⁺ (9, 12). Sarin, soman, and tabun are nerve agents that have been used as chemical weapons. The OP hydrolase activity of SMP30 makes this enzyme an interesting target for the development of bioscavengers. However, since OPs are man-made compounds that were not synthesized until the 1930s, this OP hydrolase activity provides little insight into the physiological function of the protein.

In contrast, the second observed enzymatic function of SMP30 has clear biological significance in non-primate mammals. SMP30 also functions as a lactonase and catalyzes the penultimate step in the ascorbate (vitamin C) biosynthetic pathway. While studying the conversion of D-glucuronic acid to L-ascorbate, Lehninger purified an aldonolactonase from rat liver and demonstrated its ability to catalyze the reversible interconversion of L-gulonate and L-gulono-γ-lactone (10, 18, 19). In the formation of ascorbate, L-gulonate is closed to form L-gulono-γ-lactone which is then converted to L-ascorbate by gulonolactone oxidase (20). Lehninger’s aldonolactonase was determined to be SMP30 in a study showing that SMP30 knockout mice fed a vitamin C deficient diet developed scurvy-like symptoms such as brittle bones, low body weight, and shortened lifespan. In addition to gulono-γ-lactone, SMP30 also displayed lactonase activity with other aldonolactones (21). Most mammals synthesize their own ascorbate via the pathway including SMP30, however, for some species, including primates of the Haplorrhini suborder and guinea pigs, this compound must be obtained through diet. Primates and guinea pigs no longer synthesize ascorbate because they lack a functional copy of gulonolactone oxidase, the final enzyme in the pathway. Yet, the amino acid sequence of SMP30 is remarkably well conserved, even in these species that do not synthesize ascorbate; the protein sequence of the human form is 88% identical and 93% similar to the mouse form. Thus, the primary physiological function of SMP30 in humans remains unclear.

Recent studies describe the relationship of SMP30 with a number of physiological effects. Counter to other aging observations, overexpression of SMP30 appears to cause bone loss and osteoporosis (22, 23). On the other hand, SMP30 deficiency leads to decreased glucose tolerance and abnormal lipid accumulation in the liver (22, 24–28). SMP30 has also been associated with control of cell proliferation and is down-regulated in human prostate and breast cancers (29, 30). As the effect of SMP30 in these conditions appears to be largely independent of vitamin C, they are likely clues to the physiological relevance of this protein in humans and other mammals lacking the capacity to synthesize ascorbate. Although SMP30 has been implicated in bone loss, abnormal lipid metabolism, decreased glucose tolerance and certain cancers, its role in these conditions has not been described on a metabolic or molecular level. The crystal structure of SMP30 was recently solved and described (13). Further structural characterization of this protein and the elucidation of its reaction mechanism should help to identify and further elaborate on each of the physiologic functions of SMP30 in humans.

Highly Conserved in Vertebrates

Evolutionary conservation of protein sequence often indicates that a protein has an important biological function. SMP30 homologs have been identified in at least 16 different species ranging from vertebrates to insects, bacteria and fungi (Figure 1). Vertebrate forms of SMP30, which include human, monkey, orangutan, cow, hamster, rabbit, pig, mouse, rat, and frog homologs, are 70–90% identical to one another. Sequence similarity drops to 25–45% in insects, bacteria and fungi. No known homolog exits in C. elegans however a 51-amino acid segment of the SMP30 protein exhibits homology with a yeast RNA polymerase (31–33).

Figure 1.

Amino acid sequence alignment of SMP30 homologs in vertebrates and invertebrates. Red indicates completely conserved, blue indicates identical, and yellow indicates similar residues. Identity between vertebrates ranges from 70–90%, and drops to about 30% between vertebrate and invertebrate forms of SMP30. The amino acid sequence of SMP30 is highly conserved among all mammals, regardless of ability to synthesize ascorbate. Latin and common names as well as gi accession numbers for the above represented species are as follows: Hs, Homo Sapiens, human, 23111021; Mf, Macaca fascicularis, monkey, 115502619; Pa, Pongo abeilii, orangutan, 197101437; Bt, Bos Taurus, cow, 13633941; Oc, Oryctolagus cuniculus, rabbit, 20178120; Ss, Sus scrofa, pig, 122131846; Mm, Mus musculus, mouse, 15215231; Rn, Rattus Norvegicus, rat, 68067383; Xl, Xenopus laevis, frog, 147905135; Lt, Lampyris turkestanicus, firefly LRE, 301068495; Dm, Drosophila melanogaster, fruit fly, 23171287; Xc, Xanthomonas campestris, bacteria, 21233020. This sequence alignment was prepared using the Biology Workbench (81).

Insect homologs of SMP30 include firefly luciferin-regenerating enzyme (LRE), flesh fly and Drosophila anterior fat body protein, and Drosophila SMP30. Sequence identity between the insect homologs and vertebrate SMP30 ranges between 32 and 38% (34). Firefly LRE, a 34kD protein composed of 308 amino acids, catalyzes the recycling of oxyluciferin to luciferin in firefly lanterns. Light is produced when luciferase converts luciferin to oxyluciferin in the presence of ATP and molecular oxygen. Oxyluciferin is thought to be a competitive inhibitor of luciferase. Oxyluciferin is converted back to luciferin in a two-step process; LRE first catalyzes the conversion of oxyluciferin to 2-cyano-6-hydroxybenzothiazole which then combines with D-cysteine to form luciferin (34–36). An SMP30 homolog in Drosophila was found to be up-regulated after exposure to cold temperatures (15 °C). Interestingly, expression of the Drosophila SMP30 homolog increased in senescence, in contrast to the vertebrate homologs, for which expression decrease with age (37). Significant protein sequence homology (34% identity, 50% similarity) to mammalian SMP30 was also seen in anterior fat body protein (AFP) of Sarcophaga peregrina, a 34 kD protein found in the larval fat body. AFP levels in the anterior fat body decrease after pupation in a similar trend with mammalian SMP30 homologs, and like SMP30, it does not appear to have a strong affinity for Ca²⁺ (38).

Inferences about the biological significance of SMP30 can be made from sequence analysis of homologs from different species. The role of SMP30 in ascorbate biosynthesis was first discovered in mice (21). Although most invertebrates do not synthesize ascorbate, a large number of vertebrates do (39–41). The trait of ascorbate biosynthesis most likely appeared before the divergence of plant and animal species, and well before the emergence of terrestrial vertebrates (21, 42). Cold-blooded vertebrates, such as amphibians, reptiles, and some fish, synthesize ascorbate in their kidneys, while most warm-blooded mammals synthesize ascorbate in their livers (21, 42–44). Organisms that have lost the ability to synthesize ascorbate include teleost fishes, some, but not all, passeriform birds, some bat species, guinea pigs, and Haplorrihini primates, which includes monkeys, apes, and humans. It was commonly assumed that all primates lack the ability to synthesize ascorbate, however studies have shown that primates of the suborder Strepsirrhini (prosimians) including lemurs, lorises, and galagos synthesize ascorbate (42, 44). Additionally, ascorbate is also synthesized in plants, and fermented in bacteria (21). Thus, of the organisms with identified SMP30 homologs, all produce their own ascorbate except humans, monkeys, orangutans and flies. SMP30 is the penultimate enzyme and gulonate oxidase (GULO) the final enzyme in the ascorbate biosynthetic pathway as shown in Figure 2. The gene for gulonolactone oxidase in Haplorrhini primates has undergone genetic drift, as expected for a protein evolving without functional constraint; however, SMP30 sequences remained highly conserved, indicating the existence of another biologically significant function.

Figure 2.

The final 3 steps of the biosynthesis of L-ascorbate from D-glucuronate. First, D-glucuronate is converted to L-gulonate by either aldose reductase (AR, 15%) or aldehyde reductase (GR, 85%) (67). SMP30 then closes L-gulonate to the 5-membered lactone, L-gulono-γ-lactone. In the final step, gulonolactone oxidase (GULO) converts L-gulono-γ-lactone to L-ascorbate. In mammals lacking the ability to synthesize ascorbate, GULO is non-functional and highly mutated.

SMP30 and Apoptosis

A protective effect of SMP30 against apoptosis has been observed and partially attributed to the ability of SMP30 to regulate cellular Ca²⁺ concentrations (14, 45). Sustained, elevated cellular Ca²⁺ levels induce apoptosis. Human Hep G2 cells and pig renal tubular epithelial cells overexpressing SMP30 showed enhanced Ca²⁺ efflux following a transient spike in cellular free Ca²⁺ levels induced by ATP (14, 45). When pig kidney cells transfected with SMP30 were treated with a calmodulin inhibitor, the rate of Ca²⁺ efflux was reduced, indicating that SMP30 indirectly regulates the activity of a calmodulin-dependent membrane Ca²⁺ pump by interacting with calmodulin (45). Interaction of SMP30 with calmodulin also appears to regulate the activation of survival factor Akt (46). The ability of SMP30 to stimulate membrane Ca²⁺ pumps was also demonstrated in vivo using isolated rat liver plasma membranes and microsomes (47, 48).

In addition to preventing apoptosis induced by Ca²⁺ influx, SMP30 also exhibits a protective effect against apoptosis induced by other signaling pathways. Tumor necrosis factor-α (TNF-α) induces apoptosis by binding to membrane receptors that activate multiple intracellular signaling pathways. Primary cultured hepatocytes from SMP30 knockout mice are more susceptible to apoptosis induced by treatment with tumor necrosis factor-α (TNF-α)/actinomycin D than wild type hepatocytes (15). Likewise, rat hepatoma H4-II-E cells overexpressing SMP30 were significantly less affected by treatment with TNF-α than control cells (49). The Fas ligand is another member of the TNF family that induces apoptosis by binding to membrane death receptors. Livers from SMP30 knockout mice display an increased sensitivity to Fas-mediated apoptosis, an observation that supports a protective role of SMP30 against apoptosis (15).

Other important events in the induction of apoptosis include activation of the caspase cascade and protein kinases. Lipopolysaccharide (LPS) activates many proapoptotic genes including caspase-8 and caspase-3 (50). Dibucaine is a Ca²⁺-dependent protein kinase inhibitor and has been shown to activate caspase-3, -6, -8, and -9. PD98059 is an ERK inhibitor that induces apoptosis by inactivating Bcl-2. Rat hepatoma H4-II-E cells overexpressing SMP30 were shown to be less likely to undergo apoptosis when treated with LPS, PD98059, and Dibucaine (51). Furthermore, rat kidney NRK52E cells overexpressing SMP30 were less likely to undergo apoptosis when treated with TNF-α, LPS, Bay K 8644, and thapsigargin than wild type rat kidney NRK52E cells. Addition of a caspase-3 inhibtor to wild type NRK52E cells prevented apoptosis caused by TNF-α, LPS, and Bay K 8644. Expression of caspase-3 mRNA was increased in wild type NRK52E cells treated with TNF-α; however no such change was observed in NRK52E cells overexpressing SMP30 (52). These results indicate that SMP30 prevents apoptosis by inhibiting the caspase cascade.

SMP30 and Oxidative Stress

A protective effect of SMP30 against oxidative stress has also been observed. Levels of protein carbonyls (a known marker of oxidative stress) were higher in SMP30 knockout mice than in wild type mice of the same age. After an eight week exposure to cigarette smoke SMP30 knockouts developed pulmonary emphysema due to elevated oxidative stress, whereas wild type mice did not (53). Elevated levels of oxidative stress were also observed in the brains of SMP30 knockout mice (54). Furthermore, overexpression of SMP30 in P19 cells (from a multipotent mouse embryonic carcinoma) protects the cells from t-butylhydroperoxide induced cytotoxicity (55).

Discovery of SMP30’s role as a lactonase involved in the biosynthesis of ascorbate provided an explanation for the increased oxidative stress observed in tissues from knockout mice. Ascorbate, more commonly known as vitamin C, is a powerful, water-soluble antioxidant. In 2006 Kondo et al. (21) discovered that SMP30 was the enzyme that catalyzes the penultimate step in the biosynthesis of ascorbate (Figure 2). Therefore, observed antioxidant effects of SMP30 can largely be attributed to its role in ascorbate production. SMP30 knockout mice have recently been used as a model system in studies concerning the role of ascorbate (56–61). In most cases, dietary vitamin C supplementation restores a normal phenotype to these SMP30 knockouts. However, one study indicates that the antioxidant effect of SMP30 may be due to more than ascorbate production. Handa et al. (62) overexpressed SMP30 in human hepatoma cells, Hep G2, and observed a decrease in reactive oxygen species (ROS), lipid peroxidation, superoxide dismutase activity, and GSH levels compared to control Hep G2 cells. They also found that recombinant human SMP30 was not able to directly scavenge radicals in vitro. As noted above, human cells lack gulonolactone oxidase, the final enzyme in the ascorbate biosynthesis pathway, and cannot produce vitamin C. Consequently, ascorbate production does not explain the antioxidant effect of SMP30 in this study which utilized human hepatoma cells. Handa et al. (62) proposed that altered Ca²⁺ modulation in SMP30 transfectants played a role in the decrease of ROS levels.

Gluconolactonase Activity

SMP30 has the ability to hydrolyze various aldonolactones. Kondo et al. (21) tested the enzyme’s gluconolactonase ability after noticing sequence similarity to several bacterial gluconolactonases. Using rat SMP30 purified from E. coli, the specific activity towards D- and L-glucono-δ-lactone, D- and L-gulono-γ-lactone, and D- and L-galactono-γ-lactone was determined in the presence of 75 μM Zn²⁺. In each case, it was determined that D-aldonolactones were the more reactive stereoisomeric substrates for the reaction shown in Figure 3. The specific activity in vitro of rat SMP30 with D-glucono-δ-lactone was 18- to 22-fold higher than with D-gulono- and D-galactono-γ-lactones, respectively (21). Glucono-δ-lactone is a six-membered ring, whereas gulono-γ-lactone and galactono-γ-lactone are five-membered rings. The closing of L-gulonate to L-gulono-γ-lactone is the penultimate step of the ascorbate biosynthetic pathway (Figure 2).

Figure 3.

SMP30 catalyzes the interconversion of D-glucono-δ-lactone and D-gluconic acid in the presence of divalent cations including Mn²⁺, Mg²⁺, Ca²⁺, and Zn²⁺. Opening of the lactone ring releases a proton. Of all the aldonolactone subtrates tested, SMP30 showed the greatest activity with D-glucono-δ-lactone.

The activity of human SMP30 with D-glucono-δ-lactone and its dependence on various divalent cations has also been examined (13). In the presence of 75 μM Zn²⁺, the K_M, k_cat, and k_cat/K_M values were determined to be 2.7 mM, 341 s⁻¹, and 126 mM⁻¹s⁻¹. The K_d values for Zn²⁺, Mn²⁺, Mg²⁺, and Ca²⁺ were determined in the presence of D-glucono-δ-lactone. Although SMP30 had the highest gluconolactonase activity in the presence of Zn²⁺, this is unlikely the metal of physiological relevance because the K_d for Zn²⁺ is above the free metal ion cellular concentration range. In light of the fact that SMP30 was thought to be a Ca²⁺-binding protein that played a significant role in Ca²⁺ regulation, it is interesting to note that SMP30 had the lowest affinity for Ca²⁺, and the free concentration of Ca²⁺ in a resting cell is far below the determined K_d value. It was speculated (13) that the lactonase activity of SMP30 may depend on Mn²⁺ or Mg²⁺ most of the time, switching to a Ca²⁺-dependent activity only in times of stress when Ca²⁺ levels are known to increase.

SMP30 is a Six-Bladed β-propeller

The crystal structure of human SMP30 was recently solved to a resolution of 1.4 Å (13). The structure reveals that SMP30 is a six-bladed β-propeller with 24 β-strands arranged in six β-sheets that form a torus around a central tunnel as shown in Figure 4. A single metal binding site is found inside the solvent-filled central tunnel and crystal structures bound to Zn²⁺ and Ca⁺ have been solved. In structural homologs of SMP30, such as a drug resistant protein from S. aureus named Drp35, squid DFPase, paraoxonase 1 (PON1), and a protein from X. campestris, there are four conserved residues that bind a catalytic Ca²⁺ atom. In SMP30, these residues correspond to E18, N154, D204, and N103; however, based on structures solved with Ca²⁺ or Zn²⁺ bound, only three of these residues actually coordinate with the metal ion in SMP30 (13). Although it is in close proximity, the side chain of N103 is rotated relative to its position in the structural homologs and does not appear to bind to the metal ion. The putative active site consists of four residues in the central tunnel just above the metal binding site (R101, M118, I34, D104). A flexible loop made up of six residues is located at the top of the central tunnel, forming a lid-like structure above the active site that could possibly play a role in substrate specificity or as a gatekeeper. The C-terminal tail runs across the opposite side of the central tunnel (13).

Figure 4.

Ca²⁺-bound crystal structure of human SMP30. A) A top view of SMP30 shows the six-bladed β-propeller fold. The C-terminal tail traverses the central tunnel of the propeller and the bound Ca²⁺ ion is shown in purple. B) A side view of SMP30 shows several loops that extend above the propeller near the active site which is located just above the bound Ca²⁺ ion. C) Overlay of the crystal structures of human SMP30 (gold) and PON1 (blue). Although both proteins have a six-bladed β-propeller fold, PON1 binds two Ca²⁺ ions whereas SMP30 only binds one. Furthermore, PON1 is specific for Ca²⁺, while SMP30 can bind several different divalent cations. D) Side view of SMP30 with the metal-binding and putative active site residues highlighted. E18, D204, and N154 are bound to the metal ion. N103 and D104 are not close enough to bind the metal. Other putative active site residues, which are located just above the metal binding site include, I34, R101, and M118.

SMP30 as an Organophosphate Hydrolase

OP compounds, first synthesized in Germany in the 1930s, are acetylcholinesterase inhibitors that are found in insecticides and used as nerve agents in chemical warfare (63). Billecke et al. (9) identified an enzyme in rat liver capable of hydrolyzing DFP, soman, sarin, and tabun that Kondo et al. later confirmed to be SMP30 (12). The activity of SMP30 was tested with substrates of human PON1 (another mammalian enzyme capable of hydrolyzing OPs) to determine how closely related the two enzymes were. The enzymes are each six-bladed β-propeller enzymes (Figure 4C). SMP30 was unable to hydrolyze any of the putative physiological PON1 substrates such as phenyl acetate and napthylacetates, and 7-acetoxy-N-methylquinolinium (9). Furthermore, the DFPase hydrolytic activity of SMP30 is dependent on divalent cations such as Mg²⁺, Mn²⁺, and Co²⁺, with the highest activity observed in the presence of Mg²⁺ (9). Unlike PON1, which requires Ca²⁺ for activity, SMP30 was unable to hydrolyze DFP with Ca²⁺ (12). Although the ability of an enzyme to hydrolyze OPs is of potential interest in the engineering of bioscavengers, it tells us little about the enzymes true physiological purpose, since OPs are synthetic substrates that have been around for less than a century.

There are several enzymes that are structural and potential functional homologs of SMP30. SMP30 is one of a select few enzymes that have been shown to hydrolyze both lactones and OPs, including PON1, and bacterial phophotriesterases (64). Some six-bladed β-propellers, such as Drp35 from S. aureus and XC5397 from X. campestris hydrolyze lactones but not OPs, while still others, such as DFPase from the squid L. vulgaris, hydolyze OPs but not lactonases (65). As is characteristic for β-propeller proteins, there is very low sequence identity between any of the above mentioned proteins (13, 64, 65). It is known that for PON1, different residues in the active site catalyze the OP hydrolase and lactonase mechanisms (64, 65). The relationship and involvement of specific active site residues between OP hydrolase and lactonase activity in SMP30 is unknown.

Physiological Effects

High sequence conservation of SMP30 among vertebrates is suggestive of a significant, conserved physiological function; however, no such conserved function is known for this protein. Although SMP30 is a key enzyme in the synthesis of ascorbate in most non-primate mammals, there is no direct metabolic evidence of an essential biological function in primates and other mammals that lack the capacity to synthesize their own ascorbate. Furthermore, in SMP30 knockout mice, most differences in phenotype, such as brittle bones and low body weight, are corrected by a diet supplemented with vitamin C. Recent studies have tied SMP30 to a number of physiological effects of medical relevance that are independent of a link to vitamin C. Connections between SMP30 and osteoporosis, fatty liver diseases, glucose intolerance and cancer will be discussed next.

SMP30 and Osteoporosis

Brittle bones and scurvy-like symptoms were observed in SMP30 knockout mice as a result of vitamin C deficiency (21). Ascorbate is a known cofactor in the hydroxylation of lysine and proline, which are required for collagen formation. Since collagen synthesis is a precursor of bone mineralization, ascorbate is required for bone formation (66). In addition, ascorbate contributes to bone formation by suppressing the activity of bone-resorbing osteoclasts and promoting the formation of bone-forming osteoblasts (67). Ascorbate deficiency is associated with the development of osteoporosis in elderly humans (67). Curiously, rats overexpressing SMP30 also showed bone loss and osteoporosis, with females being more affected than males (22). In SMP30 transgenic rats, SMP30 levels were elevated to 110% and 150% of wild type levels in the femoral-diaphyseal tissues of male and female rats respectively, and to 136% and 205% of wild type levels in the femoral-metaphyseal tissues of male and female rats (23). Bone morphologic changes were observed in femoral-diaphyseal and femoral-metaphyseal tissues alike, with greater changes observed in female transgenic rats. SMP30 transgenic rats had less highly mineralized and calcified tissue than wild type rats. In addition, the polar strength strain index and cortical thickness were significantly decreased in female SMP30 transgenic rats (23). SMP30 was also enhanced in bone marrow cells of transgenic rats. Differentiation of marrow cells into bone resorbing osteoclasts is stimulated by bone-resorbing factors such as the receptor activator of NF-κB ligand, parathyroid hormone, 1α,25-dihydroxyvitamin D₃ and prostaglandin E₂. When marrow cells from wild type and SMP30 transgenic rats were cultured in the absence of bone-resorbing factors, increased osteoclast formation was seen in cultures from SMP30 transgenic rats (68, 69). These results implicate SMP30 in bone resorption and the development of osteoporosis.

Nonalcoholic Fatty Liver Disease and Liver Fibrosis in SMP30 Knockouts

The designation, nonalcoholic fatty liver disease (NAFLD) applies to a range of fatty liver diseases including simple hepatic steatosis (the abnormal accumulation of lipids in cells), steatohepatitis, and even cirrhosis that are caused by neither alcoholic abuse nor hepatitis infection (27, 70, 71). It was estimated that in 2006, NAFLD affected 16–23% of the US population (70). Hyperlipidemia, insulin resistance, oxidative stress, and especially obesity are believed to contribute to NAFLD development (27, 70, 71). Over 90 percent of people with a body mass index of 39 kg/m² or greater have NAFLD (70). Furthermore, NAFLD is one of the most common causes of fibrosis, the build-up of scar tissue (71).

Recent observations in humans and knockout mice indicate an association between SMP30 and NAFLD. Among the phenotypic characteristics of SMP30 knockout mice is an accumulation of lipids, phospholipids, and neutral lipids in the liver. Lipid droplets can be seen in SMP30-deficient hepatocytes and around the central vein in the liver, suggesting a role for SMP30 in lipid metabolism (26). Since hyperlipidemia is a putative contributing factor to NAFLD development, abnormal lipid metabolism and lipid accumulation in SMP30 knockout mice indicates that SMP30 deficiency may indirectly lead to NAFLD. Observations in human subjects further support this hypothesis. Park et al. (27) observed decreased SMP30 levels in liver and blood samples from human NAFLD patients. Patients with the most severe degree of NAFLD had the lowest hepatic SMP30 levels and the highest low-density lipoprotein (LDL) and very low-density lipoprotein (VLDL) levels. Interestingly, VLDL1 production in the liver is commonly observed during development of insulin resistance, which is another one of the key risk factors for NAFLD occurrence (27). An estimated 70% of people with type 2 diabetes also have NAFLD (72). Decreased hepatic SMP30 levels were also observed in mice with D-galactosamine/lipopolysaccharide-induced acute liver failure (ALF) (73). Interestingly, plasma SMP30 levels were elevated in these mice as well as in human ALF patients (over 3 times those of healthy humans) (73). While these results do not conclusively determine whether decreased hepatic SMP30 expression is a cause or a result of NAFLD and ALF, they indicate that SMP30 is involved in the pathogenesis of this increasingly prevalent disease.

Advanced stages of NAFLD are characterized by liver fibrosis which can lead to cirrhosis. Hepatic stellate cells (HSCs), usually found in a quiescent state, are the main cells implicated in collagen production during liver fibrosis. Activation of HSCs is stimulated by transforming growth factor beta, insulin, and leptin and inhibited by peroxisome proliferator-activated receptor gamma (PPAR-γ) and adiponectin (71, 74). In contrast to the previously mentioned human studies in which decreased levels of SMP30 are associated with more severe NAFLD cases, another study done in SMP30 knockout mice demonstrated that the lack of SMP30 protected the mice from carbon tetrachloride (CCl₄)-induced liver fibrosis (74). Fibrosis was restored to levels seen in CCl₄-treated wild type mice when SMP30 knockouts were given a vitamin C enriched diet, however the level of fibrosis in wild type mice did not change upon addition of vitamin C. Furthermore, PPAR-γ was up-regulated in SMP30 knockout mice. Park et al. (74) concluded that vitamin C deficiency in SMP30 knockout mice indirectly inhibits HSC activation by up-regulating PPAR-γ. These results appear to be in conflict with those of the previously mentioned studies. However, the protective effects of SMP30 deficiency observed here were the result of vitamin C deficiency, a nutrient that humans must obtain through dietary intake.

SMP30 and Insulin Resistance

Declining glucose tolerance is seen in normal aging. Type 2 diabetes generally occurs later in life (25). Paradoxically, both overexpression of SMP30 and decreasing levels of SMP30 have been implicated in the development of insulin resistance. Overexpression of SMP30 in rat hepatoma H4-II-E cells led to insulin resistance. Genes that were normally up-regulated in response to treatment with insulin or glucose were not affected by glucose or insulin in SMP30 transfectants (75). The mechanism by which SMP30 inhibits expression of insulin signaling-related proteins is still unclear. Furthermore, H4-II-E cells over-expressing SMP30 demonstrated increased lipid production and glucose utilization (75). An in vivo study using SMP30 knockout mice showed that blood glucose levels were significantly higher and insulin levels significantly lower than in wild type mice thirty minutes after glucose administration. In islet cells isolated from the knockout mice, treatment with glucose or KCl caused less insulin secretion than in islet cells from wild type mice, indicating impaired insulin secretion in the absence of SMP30 (25). Interestingly, insulin stimulates SMP30 expression in human hepatoma cells (Hep G2); treatment of Hep G2 cells with 0.1 μM insulin for twelve hours caused increased SMP30 mRNA and protein levels (76). These studies indicate that SMP30 may be involved in the development of glucose intolerance and type 2 diabetes, however, little is known about this putative involvement.

SMP30 and Cancer

SMP30 appears to be involved in the regulation of cell proliferation and has recently been linked to breast and prostate cancers. Following treatment with carbon tetrachloride (CCl₄), liver cells from SMP30 knockout mice showed increased mitosis compared to those from wild type mice. In addition, SMP30 expression was up-regulated in response to treatment with a mitogen, lead nitrate. From this data, Ishigami et al. (29) proposed that SMP30 has a suppressive effect on cell proliferation and that it is important in regulating liver regeneration following hepatic injury. Other studies have also shown that SMP30 has a suppressive effect on cell proliferation in rat liver and kidney cells (77, 78). In addition, over-expression of SMP30 alters the expression of tumor-related genes such as c-myc, Ha-ras, c-src, p53, and Rb (retinoblastoma protein). In rat liver cells with over-expressing SMP30, the expression of the oncogenes c-myc, Ha-ras, and c-src was suppressed, while expression of the tumor suppressor genes p53 and Rb were up-regulated compared to wild type cells (79, 80). This suggests a protective role of SMP30 against cancer development. Furthermore, a recent study showed that SMP30 expression was decreased in human breast and prostate cancer tissues and cell lines and that lower SMP30 levels were associated with further progressed tumors (30). Whether decreased SMP30 levels are a cause or effect of cancer progression remains unclear.

Outlook

On a molecular level, little is known about how SMP30 affects cancer development, liver fibrosis, glucose intolerance, and bone loss. More research is required to determine the significance of the associations of SMP30 with these physiological effects. The recently solved crystal structure of SMP30 will allow putative physiological functions to be explored on the molecular level. The combined future exploration of structure, enzyme activity, metabolomics and in vivo studies will undoubtedly reveal which previously assigned SMP30 functions are physiologically relevant and significant.

Figure 5.

Organophosphates hydrolyzed by SMP30 (9). A) Structures of tabun, DFP, sarin, and soman. B) Hydrolysis of sarin by SMP30 and Mg²⁺.

Highlights.

• SMP30 is highly conserved (70–90%) in vertebrates irrespective of whether the species is capable of synthesizing ascorbate. This indicates an additional important physiological function of SMP30.

• SMP30, predominantly found in the liver and kidneys, has also been detected in the lungs, ovaries, testes, bone, epidermis, stomach and brain. Ascorbate biosynthesis occurs only in the liver in capable species. The metabolic and molecular role of SMP30 in other tissues is currently unknown.

• Homologs have been identified in insects, bacteria and fungi; however sequence identity is less than 50%.

• Cells over-expressing SMP30 have enhanced Ca²⁺-pumping activity. SMP30 appears to regulate Ca²⁺-pumping activity of plasma and microsomal membranes indirectly via interactions with calmodulin.

• SMP30 protects cells from apoptosis. This protective effect is likely the result of the protein’s role in regulating Ca²⁺ homeostasis.

• SMP30 is involved in the biosynthesis of ascorbate, a powerful antioxidant molecule. A major protective effect of SMP30 against oxidative stress is the result of ascorbate synthesis.

• SMP30 is a lactonase capable of hydrolyzing various aldonolactones such as D- and L-gulono-γ-lactone, D- and L-glucono-δ-lactone, and D- and L-galactono-γ-lactone. Greatest activity was seen with D-glucono-δ-lactone in the presence of Zn²⁺, a six-membered lactone.

• The crystal structure of human SMP30 reveals a six-bladed β-propeller. The active site is located in the central tunnel of the propeller, just above the single metal-binding site.

• Unlike structural homologs, such as PON1, that bind two Ca²⁺ atoms, the slightly altered geometry of the SMP30 metal-binding site allows it to bind a single metal, yet with a selectivity for a wider variety of divalent cations (Ca²⁺, Zn²⁺, Mg²⁺ and Mn²⁺).

• SMP30 hydrolyzes DFP and other OPs in the presence of Mg²⁺ and Mn²⁺. OPs are not naturally occurring compounds so this activity is a non-physiological, yet fortuitous function of SMP30, and this function sheds little light on the physiological function.

• Over-expression of SMP30 induces bone loss and osteoporosis in rats.

• SMP30 deficiency causes accumulation of lipids in the liver and is associated with NAFLD and liver fibrosis.

• SMP30 expression has been associated with decreasing glucose tolerance and the development of insulin resistance, suggesting a role of SMP30 in the development of type 2 diabetes.

• SMP30 suppresses cell proliferation and modulates expression of oncogenes and tumor suppressor genes.

• Expression of SMP30 in human breast and prostate cancers is significantly lower than in non-cancerous breast and prostate cells.

• Little is known about how SMP30 affects bone development, hepatic lipid accumulation and fibrosis, insulin resistance, and cancer development on the molecular level. Future studies are necessary to determine the significance of these associations.