THE BIOTRANSFORMATION OF ENDOBIOTICS BY SULFONATION
Hirotoshi Fuda, PhD, Staff Scientist
Osamu Hanyu, MD, PhD, Postdoctoral Fellow
Motohiro Endo, MD, PhD, Visiting Fellow
Hanako Nakae, MD, PhD, Postdoctoral Fellow
Norman B. Javitt, MD, PhD, Adjunct Investigator
![Photo of Charles Strott, M . D .](graphics/strott.jpg)
We investigate the molecular mechanisms and biologic implications of modifying substances by sulfonation, a fundamental process in the biotransformation of endobiotics and of drugs and xenobiotics. Sulfonation, the transfer of an SO3-1 group from the universal sulfonate donor molecule 3¢-phosphoadenosine 5¢-phosphosulfate (PAPS) to an acceptor compound, is essential for normal growth and development as well as for maintenance of the internal milieu. We investigate enzymes that transfer the sulfonate moiety, i.e., sulfotransferases as well as the bifunctional PAPS synthetases producing the sulfonate donor. There are two classes of sulfotransferases: one class is tightly associated with membranes, especially the Golgi complex, while the other is composed of a superfamily of soluble or cytosolic enzymes (designated SULT). The former class sulfonates macromolecules such as glycosaminoglycans and proteoglycans, whereas the SULT class sulfonates low molecular weight compounds, such as hormones and neurotransmitters. We study the SULT2 family that sulfonates steroids and sterols. By modulating the availability of biologically active substances, sulfonation can influence biologic activity regardless of whether compounds act (1) in their unconjugated or sulfoconjugated form or (2) via a genomic or nongenomic mechanism. Thus, sulfotransferases play an essential role in specific physiologic systems and their associated disorders.
Cholesterol sulfate regulation of profilaggrin mRNA expression: studies involving siRNA for SULT2B1b and RORalpha
Cholesterol sulfate, which is produced by the action of the sterol/steroid sulfotransferase SULT2B1b, is a ubiquitous sulfolipid that has emerged as a multifaceted molecule with regulatory capabilities. Cholesterol sulfate is present in the epidermis, where it accumulates in the stratum granulosum; in addition, epidermal keratinocytes amass cholesterol sulfate during squamous differentiation. It is notable that expression of SULT2B1b localizes to the granular layer of the epidermis and thus coincides with the epidermal area containing the highest content of cholesterol sulfate. Furthermore, both SULT2B1b and the barrier protein filaggrin are localized in the epidermis; filaggrin is an acknowledged late marker of keratinocyte terminal differentiation and, in contrast to involucrin, an early marker of differentiation, which is expressed throughout the suprabasal region.
Interestingly, others have reported that, based on crystallographic data, cholesterol sulfate binds with high affinity to the retinoic acid receptor–related orphan receptor alpha (RORalpha); moreover, the high-affinity binding of cholesterol sulfate was found to translate into increased transcriptional activity of RORalpha. RORalpha mRNA is widely expressed in the mouse, although RORalpha protein is expressed in the periphery at detectable levels only in skin and testis. While RORalpha’s expression in mouse skin provides the basis for assuming similar expression in human skin, researchers had not yet conclusively demonstrated expression in human skin. The observation that, in the human epidermis, the granular layer contains the highest concentration of cholesterol sulfate, coupled with the observation that SULT2B1b and filaggrin co-localize in the granular layer, suggests an important functional association. The association is strongly supported by the finding that mouse keratinocytes in primary culture respond to administration of cholesterol sulfate with a dose-dependent increase in filaggrin formation. The possibility of association is even more intriguing when the above findings are considered in terms of the discovery that cholesterol sulfate binds to the nuclear receptor RORalpha with high affinity and that the latter is similarly expressed in the epidermis. Therefore, we raised several questions with respect to the physiological significance of SULT2B1b expression in the epidermal granular layer vis-à-vis filaggrin expression: whether cholesterol sulfate has a direct influence on the expression of filaggrin; the location of RORalpha expression in the skin of humans (if indeed RORalpha is expressed in human skin), particularly in relation to the expression of SULT2B1b and filaggrin; which of RORalpha’s four subtypes (a1–a4) are expressed in human skin; and whether RORalpha participates in the molecular action of cholesterol sulfate in the human epidermis.
We have now demonstrated that RORalpha is in fact expressed in human skin and localizes to the outer reaches of the epidermis. When we stained sections of normal human skin with an antibody against RORalpha, we observed intense staining of nuclei located primarily in the granular layer of the epidermis. Furthermore, it appears that only the a1 and a4 subspecies are expressed in human skin; a2 and a3 were not detectable. The pattern of RORalpha subspecies expression in human skin was similar to that found in the human cerebellum. In contrast to human skin and cerebellum, however, primary cultures of normal human epidermal keratinocytes (NHEK) clearly expressed the a4 species, whereas the a1 species was not demonstrable.
Cholesterol sulfate added to NHEK cultures induced expression of profilaggrin mRNA (profilaggrin is a large precursor protein that contains 10 to 12 identical copies of the mature filaggrin protein arranged in tandem) in a dose-dependent manner, suggesting a new regulatory role for this sulfolipid. That the nuclear receptor RORalpha has been shown to bind tightly to cholesterol sulfate further suggested that induction of profilaggrin by cholesterol sulfate might occur via an interaction with RORalpha. RORalpha expression in normal human skin is similar to that of filaggrin and SULT2B1b. We carried out knockdown experiments with siRNA directed against mRNAs for SULT2B1b and RORalpha, with NHEK undergoing Ca2+-induced differentiation. Concomitant with siRNA-induced reduction in SULT2B1b and RORalpha expression, we observed a decrease in profilaggrin mRNA expression of up to 80 percent. Furthermore, supplementing cultures with cholesterol sulfate during knockdown of SULT2B1b resulted in partial recovery of profilaggrin expression, whereas supplementing cultures with cholesterol sulfate during RORalpha knockdown did not. Our studies thus suggest that cholesterol sulfate may regulate profilaggrin expression via the receptor’s ability to interact with the nuclear receptor RORalpha.
Direct interaction of cholesterol sulfate and RORalpha in the regulation of expression of the profilaggrin gene
To obtain evidence for a direct interaction of cholesterol sulfate with RORalpha, we cotransfected HeLa cells, which express neither cholesterol sulfotransferase (SULT2B1b) nor RORalpha, with RORalpha and the 2.1 kb 5¢-flanking region of the profilaggrin promoter, which contains an RORalpha response element fused to a reporter gene. Interestingly, we noted a significant induction of profilaggrin gene expression in the presence of either cholesterol sulfate or RORalpha while the combination of cholesterol sulfate and RORalpha led to a striking enhancement of induction. Cholesterol acetate, on the other hand, had no effect on promoter activity in the presence or absence of RORalpha. In addition, either mutation or deletion of the RORalpha response element resulted in loss of induction by RORalpha with or without cholesterol sulfate. These studies strongly suggest that cholesterol sulfate can modulate gene expression by directly interacting with a nuclear receptor (RORa lpha), an action reminiscent of a classical hormone.
RORalpha and cholesterol sulfate binding parameters
Recent structural and functional studies have led to the hypothesis that either cholesterol or a cholesterol derivative is the natural ligand of RORalpha, an orphan member of the subfamily of nuclear hormone receptors. Following a ligand exchange experiment, we solved the X-ray crystal structure of the ligand-binding domain of RORalpha in complex with cholesterol sulfate. In contrast to the 3-hydroxyl of cholesterol, the 3-O-sulfate group makes additional direct hydrogen bonds with three residues of the RORalpha ligand–binding domain. Compared with the complex with cholesterol, seven well-ordered water molecules are displaced, and the ligand is slightly shifted toward the hydrophilic part of the ligand-binding pocket, which is ideally suited for interactions with a sulfate group. The additional ligand-protein interactions result in a higher affinity of cholesterol sulfate than of cholesterol, as shown by mass spectrometry analysis under native conditions and by differential scanning calorimetry. Moreover, mutational studies have shown that cholesterol sulfate’s higher binding affinity translates into increased transcriptional activity of RORalpha. These findings suggest that, at least in part, the molecular action ascribed to cholesterol sulfate involves interaction with the RORalpha nuclear receptor. Nevertheless, we have yet to determine binding parameters for this important interaction. Thus, in an effort to generate the parameters, we attempted to overexpress the RORalpha protein in bacteria; however, success remains elusive owing to an inability to obtain sufficient amounts of soluble protein, undoubtedly because of its high content of cystine, particularly in its N-terminal region. We have now shifted to the baculovirus expression system and are in the process of expressing the sterol-binding domain of RORalpha, which is located in the C-terminal portion of the protein. With a sufficient amount of pure protein and by using Biocore T100 technology, we hope to determine RORalpha–cholesterol sulfate binding parameters.
SULT2B1b in oxysterol metabolism: role in cytotoxicity and cell signaling
In contrast to the many studies that have established the deleterious effects of oxysterols on biologic processes, few studies have focused on metabolic pathways for oxysterol disposal. Of particular interest are the high levels of oxysterols in atheromas, which are associated with instability and rupture, a prelude to myocardial infarction. A major oxysterol found in atheromas as well as in other tissues is 7-ketocholesterol (7-KC), which is known from cell culture studies to induce cell injury at concentrations present in vivo. For this reason, researchers are investigating metabolic pathways that can lead to a reduction in 7-KC’s toxicity. For example, it has been shown that 7-KC is a substrate for 27-hydroxylation, thus forming a more water-soluble triol that lowers the intracellular concentration of 7-KC in macrophages. We recently reported that, in addition to accelerating transport, 27-hydroxylation prevents loss in cell viability and, in co-culture, nullifies the toxicity of 7-KC, findings consonant with the differential effects of oxysterols when used in combination. Another potential metabolic pathway for the metabolism of oxysterols emerged with the discovery that one member of the SULT2 family of cytosolic sulfotransferases, SULT2B1b, has a particular affinity for cholesterol. Although it is well recognized that sulfonation of steroid hormones affects the hormones’ biologic activity and can influence their disposal, the concept of an analogous pathway for C27 sterols has received limited attention.
Oxysterols constitute a class of cholesterol derivatives that exhibit broad biological effects, ranging from cytotoxicity to regulation of cell signaling; for example, the liver X nuclear receptor (LXR) regulates expression of genes engaged in fatty acid and cholesterol metabolism. The role of oxysterols such as 7-KC in the development of retinal macular degeneration and atheromatous lesions is of particular interest, but little is known about their metabolic fate. We established that the steroid/sterol sulfotransferase SULT2B1b, which is known to sulfonate cholesterol efficiently, effectively sulfonates a variety of oxysterols, including 7-KC. We observed that the cytotoxic effect of 7-KC on 293T cells was attenuated when we transfected the cells, which do not express SULT2B1b, with SULT2B1b cDNA. Importantly, protection by transfection from 7-KC–induced loss of cell viability correlated with the synthesis of SULT2B1b protein and production of the 7-KC sulfoconjugate (7-KCS). Moreover, adding 7-KCS to the culture medium of 293T cells in amounts equimolar to 7-KC protected against loss of cell viability. In addition, MCF-7 cells, which highly express SULT2B1b, were significantly more resistant to the cytotoxic effect of 7-KC. We found that oxysterol substrates for SULT2B1b extend to 7alpha/7beta-hydroxycholesterol and 5alpha,6alpha/5beta,6beta-epoxycholesterol as well as the 7alpha-hydroperoxide derivative of cholesterol. Thus, by acting on a variety of oxysterols, SULT2B1b offers a potential pathway for modulating the injurious effects of these compounds in vivo. Employing LC-ESI/MS (liquid chromatography–electrospray ionization/mass spectrometry) analysis, a technique commonly used to identify a variety of biologically important substances, we also found that the sulfoconjugate of 7-KC does indeed occur in vivo as demonstrated for human atheromatous tissue. This is an important finding, given the high levels of oxysterols, particularly 7-KC in atheromas. In contrast to 7-KC’s toxic effects, as we report here, the sulfoconjugate of 7-KC is completely non-toxic to cells that are sensitive to the adverse effects 7-KC. Thus, SULT2B1b, which is expressed in human macrophages, offers an important in vivo metabolic route for the efficient detoxification of 7-KC and other cytotoxic oxysterols. In summary, it now appears that oxysterols have a broad range of biological effects and that SULT2B1b plays a significant role throughout the spectrum. Not only does SULT2B1b inactivate classes of oxysterols that are cytotoxic, but it also inactivates classes of oxysterols involved in cell signaling.
Fuda H, Javitt NB, Mitamura K, Ikegawa S, Strott CA. Oxysterols are substrates for cholesterol sulfotransferase (SULT2B1b). J Lipid Res 2007;48:1343-52.
Lee JW, Fuda H, Javitt NB, Strott CA, Rodriguez I. Expression and localization of sterol 27-hydroxylase (CYP27A1) in monkey retina. Exp Eye Res 2006;83:465-9.
COLLABORATOR
Inna Gorshkova, PhD, Division of Bioengineering and Physical Science, ORS, NIH, Bethesda, MD
Yuko Higashi, MD, PhD, Kagoshima University Graduate School of Medical and Dental Sciences, Kagoshima, Japan
Shigeo Ikegawa, PhD, Faculty of Pharmaceutical Sciences, Kinki University, Osaka, Japan
Ignacio R. Rodriguez, MD, PhD, Laboratory of Retinal Cell and Molecular Biology, NEI, Bethesda, MD
Peter Schuck, PhD, Division of Bioengineering and Physical Science, ORS, NIH, Bethesda, MD
For further information, contact chastro@mail.nih.gov.