FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2079504259> ?p ?o ?g. }

• W2079504259 endingPage "32869" @default.
• W2079504259 startingPage "32861" @default.
• W2079504259 abstract "The enzymatic cleavage of the nicotinamide-glycosidic bond on nicotinamide adenine dinucleotide (NAD+) has been proposed to go through an oxocarbenium ion-like transition state. Because of the instability of the ionic intermediate, there has been no structural report on such a transient reactive species. Human CD38 is an ectoenzyme that can use NAD+ to synthesize two calcium-mobilizing molecules. By using NAD+ and a surrogate substrate, NGD+, we captured and determined crystal structures of the enzyme complexed with an intermediate, a substrate, and a product along the reaction pathway. Our results showed that the intermediate is stabilized by polar interactions with the catalytic residue Glu226 rather than by a covalent linkage. The polar interactions between Glu226 and the substrate 2′,3′-OH groups are essential for initiating catalysis. Ser193 was demonstrated to have a regulative role during catalysis and is likely to be involved in intermediate stabilization. In addition, a product inhibition effect by ADP-ribose (through the reorientation of the product) or GDP-ribose (through the formation of a covalently linked GDP-ribose dimer) was observed. These structural data provide insights into the understanding of multiple catalysis and clues for drug design. The enzymatic cleavage of the nicotinamide-glycosidic bond on nicotinamide adenine dinucleotide (NAD+) has been proposed to go through an oxocarbenium ion-like transition state. Because of the instability of the ionic intermediate, there has been no structural report on such a transient reactive species. Human CD38 is an ectoenzyme that can use NAD+ to synthesize two calcium-mobilizing molecules. By using NAD+ and a surrogate substrate, NGD+, we captured and determined crystal structures of the enzyme complexed with an intermediate, a substrate, and a product along the reaction pathway. Our results showed that the intermediate is stabilized by polar interactions with the catalytic residue Glu226 rather than by a covalent linkage. The polar interactions between Glu226 and the substrate 2′,3′-OH groups are essential for initiating catalysis. Ser193 was demonstrated to have a regulative role during catalysis and is likely to be involved in intermediate stabilization. In addition, a product inhibition effect by ADP-ribose (through the reorientation of the product) or GDP-ribose (through the formation of a covalently linked GDP-ribose dimer) was observed. These structural data provide insights into the understanding of multiple catalysis and clues for drug design. Human CD38 is a type II transmembrane glycoprotein first identified as a surface antigen of lymphocytes (1Jackson D.G. Bell J.I. J. Immunol. 1990; 144: 2811-2815PubMed Google Scholar). It has since been shown to possess many cellular functions (2Deaglio S. Capobianco A. Bergui L. Durig J. Morabito F. Duhrsen U. Malavasi F. Blood. 2003; 102: 2146-2155Crossref PubMed Scopus (164) Google Scholar) and is directly involved in some human diseases such as B cell chronic lymphocytic leukemia (3Thornton P.D. Fernandez C. Giustolisi G.M. Morilla R. Atkinson S. A'Hern R.P. Matutes E. Catovsky D. Hemato. J. 2004; 5: 145-151Crossref PubMed Scopus (50) Google Scholar), AIDS (4Roussanov B.V. Taylor J.M. Giorgi J.V. AIDS. 2000; 14: 2715-2722Crossref PubMed Scopus (14) Google Scholar), and diabetes (5Antonelli A. Ferrannini E. J. Endocrinol. Investig. 2004; 27: 695-707Crossref PubMed Scopus (17) Google Scholar). In addition, the extramembrane domain of CD38 is found to be a multifunctional enzyme. As a member of the ADP-ribosyl cyclase family, CD38 not only can synthesize cyclic ADP-ribose (cADPR) 4The abbreviations used are: cADPR, cyclic ADP-ribose; ADPR, ADP-ribose; NAADP+, nicotinic acid adenine dinucleotide phosphate; NGD+, nicotinamide guanine dinucleotide; GDPR, GDP-ribose; cGDPR, cyclic GDP-ribose; GDPRI, GDP-ribose intermediate; MES, 4-morpholinoethanesulfonic acid. 4The abbreviations used are: cADPR, cyclic ADP-ribose; ADPR, ADP-ribose; NAADP+, nicotinic acid adenine dinucleotide phosphate; NGD+, nicotinamide guanine dinucleotide; GDPR, GDP-ribose; cGDPR, cyclic GDP-ribose; GDPRI, GDP-ribose intermediate; MES, 4-morpholinoethanesulfonic acid. from NAD+ but also can hydrolyze NAD+ and cADPR to produce ADPR (6Lee H.C. Zocchi E. Guida L. Franco L. Benatti U. De Flora A. Biochem. Biophys. Res. Commun. 1993; 191: 639-645Crossref PubMed Scopus (102) Google Scholar, 7Howard M. Grimaldi J.C. Bazan J.F. Lund F.E. Santos-Argumedo L. Parkhouse R.M. Walseth T.F. Lee H.C. Science. 1993; 262: 1056-1059Crossref PubMed Scopus (666) Google Scholar, 8Takasawa S. Tohgo A. Noguchi N. Koguma T. Nata K. Sugimoto T. Yonekura H. Okamoto H. J. Biol. Chem. 1993; 268: 26052-26054Abstract Full Text PDF PubMed Google Scholar). Furthermore, at acidic conditions and in the presence of nicotinic acid, CD38 can catalyze a nicotinamide exchange reaction to produce NAADP+ from NADP+ (9Aarhus R. Graeff R.M. Dickey D.M. Walseth T.F. Lee H.C. J. Biol. Chem. 1995; 270: 30327-30333Abstract Full Text Full Text PDF PubMed Scopus (338) Google Scholar). Recent studies indicate that CD38 can also hydrolyze NAADP+ to ADPRP and that this hydrolysis reaction, likewise, occurs only at acidic pH (10Graeff R. Liu Q. Kriksunov I.A. Hao Q. Lee H.C. J. Biol. Chem. 2006; 281: 28951-28957Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar). Biochemical and structural characterizations reveal that two acidic residues, Glu146 and Asp155, at the active site of CD38 are the critical determinants of this novel acidic dependence (10Graeff R. Liu Q. Kriksunov I.A. Hao Q. Lee H.C. J. Biol. Chem. 2006; 281: 28951-28957Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar). The enzymatic products of CD38, cADPR, NAADP+, and ADPR, are all calcium messengers targeting different calcium channels and stores, indicating the central role of the enzyme in calcium signaling. The enzymatic syntheses of these signal molecules first require the removal of the nicotinamide moiety from the substrate to form a reactive intermediate, which can be further catalyzed to produce divergent signals through divergent pathways (11Schuber F. Lund F.E. Curr. Mol. Med. 2004; 4: 249-261Crossref PubMed Scopus (105) Google Scholar). As the catalytic ability of an enzyme is dependent on the structure of the reaction intermediate, it is thus important to characterize the chemical nature of the intermediate. To describe the reaction mechanism, a covalent intermediate was proposed in which the remaining ribonucleotide is linked to the catalytic residue of the enzyme after the departure of the nicotinamide (12Sauve A.A. Deng H.T. Angeletti R.H. Schramm V.L. J. Am. Chem. Soc. 2000; 122: 7855-7859Crossref Scopus (43) Google Scholar). However, mounting evidence points instead to the formation of an ionic intermediate, proposed to be an oxocarbenium ion that is stabilized by ion pair interactions with the catalytic residues (13Oppenheimer N. Mol. Cell Biochem. 1994; 138: 245-251Crossref PubMed Scopus (61) Google Scholar, 14Muller-Steffner H.M. Augustin A. Schuber F. J. Biol. Chem. 1996; 271: 23967-23972Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar, 15Yamamoto-Katayama S. Ariyoshi M. Ishihara K. Hirano T. Jingami H. Morikawa K. J. Mol. Biol. 2002; 316: 711-723Crossref PubMed Scopus (83) Google Scholar). Similar intermediates have also been proposed to exist for other NAD+-utilizing enzymes such as Sir2 enzymes (16Avalos J.L. Boeke J.D. Wolberger C. Mol. Cell. 2004; 13: 639-648Abstract Full Text Full Text PDF PubMed Scopus (117) Google Scholar), ADP-ribosyltransferase (17Seman M. Adriouch S. Haag F. Koch-Nolte F. Curr. Med. Chem. 2004; 11: 857-872Crossref PubMed Scopus (102) Google Scholar), and ADP-ribosylating toxins (18Han S. Craig J.A. Putnam C.D. Carozzi N.B. Tainer J.A. Nat. Struct. Biol. 1999; 6: 932-936Crossref PubMed Scopus (206) Google Scholar). In solution, the lifetime of the transient existence of such a cation is ∼10-12 s, preventing it from structural detection. In the active site of an enzyme, the cation intermediate may be stabilized by the enzyme for a longer lifetime. For example, during NGD+ catalysis by human CD38, the intermediate can stay in the active site of the enzyme with a half-life of about 6 ms (12Sauve A.A. Deng H.T. Angeletti R.H. Schramm V.L. J. Am. Chem. Soc. 2000; 122: 7855-7859Crossref Scopus (43) Google Scholar, 19Sauve A.A. Munshi C. Lee H.C. Schramm V.L. Biochemistry. 1998; 37: 13239-13249Crossref PubMed Scopus (98) Google Scholar). To our knowledge, there has been no structural report for the intermediate during nucleotide catalysis. To structurally probe this transient intermediate and to understand the structural basis for the multiple catalysis, we combined molecular biology, low temperature soaking technique, and cryocrystallography to obtain CD38 complex structures along the reaction pathway. To obtain the Michaelis complex, enzymatically inactive mutant E226Q crystals were produced to form a complex with NAD+. To obtain the enzyme-product complex, wild-type CD38 crystals were incubated with NAD+ to form a complex with ADPR. To get the intermediate, the wild-type CD38 crystals were incubated with NGD+ instead of NAD+. The incubation of CD38 crystals with NGD+ at low temperature resulted in the capture of a trapped intermediate and a GDP-ribose (GDPR) dimer in the active sites of two CD38 molecules. The analysis of these structural data provides insights into the understanding of the intermediate and mechanisms for multiple catalysis. Protein Expression, Purification, and Crystallization—The extramembrane domain of the wild-type CD38 was produced in a yeast expression system using a fermentation reactor and purified as reported previously (20Munshi C. Aarhus R. Graeff R. Walseth T.F. Levitt D. Lee H.C. J. Biol. Chem. 2000; 275: 21566-21571Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 21Liu Q. Kriksunov I.A. Graeff R. Munshi C. Lee H.C. Hao Q. Structure (Camb.). 2005; 13: 1331-1339Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar, 22Mushi C. Fryxell K.B. Lee H.C. Branton W.D. Methods Enzymol. 1997; 280: 318-330Crossref PubMed Scopus (65) Google Scholar). The E226Q mutant was constructed as described previously (20Munshi C. Aarhus R. Graeff R. Walseth T.F. Levitt D. Lee H.C. J. Biol. Chem. 2000; 275: 21566-21571Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 23Graeff R. Munshi C. Aarhus R. Johns M. Lee H.C. J. Biol. Chem. 2001; 276: 12169-12173Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar), and the mutant protein was produced and purified using the same method as described for the wild-type CD38. Usable wild-type CD38 crystals were obtained using the hanging-drop vapor diffusion technique with 1 μl of protein solution equilibrated with 1 μl of precipitant solution containing 100 mm MES, pH 6.5, and 12% polyethylene glycol 4000. The CD38 E226Q mutant was crystallized with a slightly modified condition in which the concentration of polyethylene glycol 4000 was diluted to 10%. Preparation of CD38 Complexes—All complexes were obtained by soaking either native CD38 crystals or E226Q mutant crystals in solutions containing NAD+ or NGD+. The 10 mm NAD+ soaking solution was prepared with NAD+, crystallization mother liquor, and 30% glycerol. Glycerol was required to prevent formation of ice at low temperature. The E226Q-NAD+ complex was obtained by soaking one E226Q crystal in 10 mm NAD+ solution at room temperature for 30 s. The soaked crystal was then directly mounted for data collection under the protection of a cold nitrogen stream. It should be noted that a soaking experiment in some cases might freeze out authentic conformational changes associated with substrate binding. For the wild type-GDPR dimer and wild type-GDP-ribose intermediate (GDPRI) complex, one wild-type CD38 crystal was incubated with 20 mm NGD+ soaking solution at 0 °C for 20 min. Soaking at 0 °C can greatly slow down the reaction rate of NGD+ and reduce the damage to the crystal. Data Collection, Reduction, and Structure Refinement—All x-ray diffraction data were collected at the Cornell High Energy Synchrotron Source (CHESS) A1 station under cryoprotection at 100 K with a fixed wavelength of 0.976 Å. A total of 360 images with an oscillation angle of 1° each were collected for each crystal using a Quantum Q-210 CCD detector. The complete data sets were processed using the program package DENZO/SCALEPACK (24Otwinowski Z. Minor W. Methods Enzymol. 1997; 276: 307-326Crossref Scopus (38252) Google Scholar). The crystallographic statistics are listed in Table 1. The apostructure of soluble human CD38 was used as the initial model for structure solution by the molecular replacement method. Subsequent crystallographic refinements were done with the program REFMAC (25Collaborative Computational Project, Number 4 (CCP4) Suite: Programs for Protein CrystallographyActa Crystallogr. Sect. D Biol. Crystallogr. 1994; 50: 760-763Crossref PubMed Scopus (19667) Google Scholar). All substrates and products were built using the program O (26Jones T.A. Zou J.Y. Cowan S.W. Kjeldgaard M. Acta Crystallogr. Sect. A. 1991; 47: 110-119Crossref PubMed Scopus (12999) Google Scholar).TABLE 1Crystallographic data and refinement statisticsWT-ADPRE226Q-NADWT-GDPRIData collectionCell dimensions41.89341.45141.71a, b, c (Å)51.15250.71552.812α, β, γ (°)65.639/109.35665.200/108.75865.424/105.9391.22492.00791.8097.38396.37795.22Space groupP1P1P1Resolution (Å)50-1.7 (1.76-1.7)30-1.9 (1.97-1.9)30-1.7 (1.76-1.7)Unique reflections52,17037,73256,343Multiplicity3.63.33.9I/sigma19.56 (3.34)17.72 (2.54)21.37 (2.95)Rmerge (%)aRmerge = ∑|I - 〈I 〉|/∑I, where I is the integrated intensity of a given reflection.6.2 (30.6)6.2 (39.8)5.1 (46)Completeness (%)94.9 (71.8)95.8 (83.7)97.2 (95.9)RefinementR factor (%)bR = ∑||Fobs| - |Fcalc||/∑|Fobs|. Rfree was calculated using 5% of the data excluded from refinement.17.9217.6318.25Rfree factor (%)23.1323.3923.57Protein atoms3,9914,0164,016Water molecules543381695Ligands (atoms)36 + 36 + 3644 + 4436 + 37 + 73r.m.s. deviationsBond lengths (Å)0.0140.0170.015Bond angles (°)1.5311.7671.702a Rmerge = ∑|I - 〈I 〉|/∑I, where I is the integrated intensity of a given reflection.b R = ∑||Fobs| - |Fcalc||/∑|Fobs|. Rfree was calculated using 5% of the data excluded from refinement. Open table in a new tab Mutagenesis and Biochemical Assay—The mutants of CD38 at the Ser193 position were prepared by site-directed mutagenesis as described (20Munshi C. Aarhus R. Graeff R. Walseth T.F. Levitt D. Lee H.C. J. Biol. Chem. 2000; 275: 21566-21571Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar, 23Graeff R. Munshi C. Aarhus R. Johns M. Lee H.C. J. Biol. Chem. 2001; 276: 12169-12173Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar) and were verified and confirmed by DNA sequencing. Following a 24-h growth phase in YPD medium (1% yeast extract, 2% peptone, 2% glucose), protein expression was induced by addition of 0.5% methanol. The supernatants were harvested 72 h after methanol induction. The mutant proteins were purified by phenyl-Sepharose chromatography as described previously for proteins expressed with this system (27Munshi C.B. Lee H.C. Lee H. C Cyclic ADP-ribose and NAADP: Structures, Metabolism, and Functions. Kluwer Academic Publishers, Dordrecht, the Netherlands2002: 81-99Crossref Google Scholar). The protein content was determined by the Bradford assay. The elution of protein corresponded to the elution of activity measured with NGD+ as substrate. The GDP-ribosyl cyclase activity was measured based on fluorescence changes as described previously (28Graeff R.M. Walseth T.F. Fryxell K. Branton W.D. Lee H.C. J. Biol. Chem. 1994; 269: 30260-30267Abstract Full Text PDF PubMed Google Scholar). NADase activity was determined for all Ser193 mutants and compared with the NADase activities of wild-type CD38 and the E226G mutant. The Ser193 and E226Q mutants were assayed at a protein concentration of 10 μg/ml and the wild-type CD38 at 1 μg/ml. The times of the assays were selected so that 10-20% of the substrate was utilized. The reactions were conducted in 25 mm Tris-HCl, pH 7, with 0.5 mm NAD+ in volumes of 100 μl for times ranging from 1 to 30 min and were stopped by rapid mixing with 100 mm HCl. Preliminary experiments showed that this was an effective method for stopping the reactions. The acid-containing mixture was filtered with Immobilon filters and neutralized with 100 mm Tris base. These neutral extracts were diluted to 1:10 in 50 mm Tris-HCl, pH 8, and assayed for NAD+ concentration. Standards for NAD+ were treated similar to assay mixtures. The final NAD+ content was determined on a fluorescence plate reader by mixing equal volumes of sample with a reagent containing 50 mm Tris-HCl, pH 8, 1% ethanol, and alcohol dehydrogenase (6 units/ml). The samples were quantified by measuring the fluorescence of the NADH product (excitation 340 nm, emission 440 nm). Enzyme-NAD+ Complex—To obtain a stable enzyme-NAD+ complex for crystal structure determination, the enzymatic activity of human CD38 needs to be totally blocked. Previous site-directed mutagenesis studies have shown that Glu226 is critical for the enzymatic activities of CD38 (20Munshi C. Aarhus R. Graeff R. Walseth T.F. Levitt D. Lee H.C. J. Biol. Chem. 2000; 275: 21566-21571Abstract Full Text Full Text PDF PubMed Scopus (102) Google Scholar). We therefore prepared the inactive mutant of CD38, E226Q, and obtained crystals of it. By incubating preformed CD38 E226Q crystals with NAD+, we found that NAD+ can easily diffuse into the active site and form the Michaelis complex shown in Fig. 1A. The linear NAD+ is constrained by the enzyme and is stabilized in the active site by extensive polar interactions involving residues Asp155, Glu146, Gln226, Trp125, Ser126, Arg127, and Thr221 and a structural water molecule (Fig. 1B). The electron densities are clear for most parts of NAD+, except for the adenine ring and the ribose moiety attached to it (Fig. 1B). The adenine terminus of NAD+ is out of the active site and thus not expected to be stabilized by the enzyme. The disordering of the adenine terminus reflects the flexibility of this portion of NAD+ in the active site, consistent with its ability to refold toward the nicotinamide end of the molecule so that the two ends can be brought together to form cADPR. The nicotinamide end of NAD+ penetrates deep into the active site (Fig. 1A), which is also consistent with cleavage of the nicotinamide-glycosidic bond being the first step of catalysis. The nicotinamide group of the bound NAD+ interacts with Glu146 and Asp155 through two hydrogen bonds to its amide nitrogen (Fig. 1B) and is additionally positioned by the parallel displaced π-π interactions between its pyridine ring and the Trp189 indole ring. These concerted interactions not only provide precise recognition and positioning of the nicotinamide moiety in the active site but also exert certain strain on the nicotinamide moiety, resulting in a rotation of the plane of the carboxamide group by 22° from the plane of the pyridine ring (Fig. 1C). This distortion of the substrate nicotinamide moiety may be important for its cleavage, as has been observed in the structure of SIR2-NAD+ complex (16Avalos J.L. Boeke J.D. Wolberger C. Mol. Cell. 2004; 13: 639-648Abstract Full Text Full Text PDF PubMed Scopus (117) Google Scholar). Two other parts of NAD+ that have strong interactions with the active site are the nicotinamide ribose and the diphosphate. The ribose is fixed by polar interactions with Gln226, mainchain Trp125 nitrogen, and a structural water molecule (Fig. 1B). The diphosphate forms four hydrogen bonds with residues, Ser126, Arg127, and Thr221. Overall, the effect of interactions between substrate and enzyme is an activated Michaelis complex. In this activated complex, the labile nicotinamide-ribosyl bond is stretched from the ideal value bond length of 1.50 Å (29Guillot B. Jelsch C. Lecomte C. Acta Crystallogr. Sect. C. 2000; 56: 726-728Crossref PubMed Scopus (12) Google Scholar) to 1.65 Å, promoting the dissociation of the nicotinamide moiety from the substrate. Reaction Intermediate Captured during NGD+ Catalysis—Nicotinamide guanine dinucleotide (NGD+) is a surrogate substrate of NAD+ and differs from NAD+ by replacing the NAD+ adenine group with a guanine group. NGD+ was used to obtain an intermediate because the reaction rate of NGD+ is much slower than that of NAD+ (by a factor of four (23Graeff R. Munshi C. Aarhus R. Johns M. Lee H.C. J. Biol. Chem. 2001; 276: 12169-12173Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar)), suggesting a longer lived intermediate during the NGD+ catalysis. There are two CD38 molecules in the asymmetric unit of preformed wild-type crystals (21Liu Q. Kriksunov I.A. Graeff R. Munshi C. Lee H.C. Hao Q. Structure (Camb.). 2005; 13: 1331-1339Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar). By soaking these wild-type crystals with NGD+, the active sites of both molecules were saturated with substrate NGD+, and reaction proceeded in the crystal. By analyzing the data collected from these crystals, we found that the active sites of the two molecules contain different nucleotides. In fact, as can be seen in Fig. 2, molecule B contains two nucleotides, a GDP-ribose intermediate and a hydrolyzed product, GDPR, whereas molecule A contains only one nucleotide. It is reasonable to observe different occupants in the active sites of two molecules. We reported previously that two molecules in the crystallographic asymmetric unit have slightly different structures because of the crystal constraints (21Liu Q. Kriksunov I.A. Graeff R. Munshi C. Lee H.C. Hao Q. Structure (Camb.). 2005; 13: 1331-1339Abstract Full Text Full Text PDF PubMed Scopus (131) Google Scholar). In molecule B, the electron densities for the active site indicate the presence of an intermediate and a guanine ring from a hydrolysis product, GDPR (Fig. 3A). Compared with the electron density for NAD+ in Fig. 1B, the gap between the densities of GDPRI and the guanine ring clearly prove that the nicotinamide has been cleaved in the GDPRI complex. The electron density at the ribosyl C-1′ position clearly shows the absence of an OH group, or any linkage, attached to C-1′. This demonstrates that it is not a substrate or product but most likely an intermediate produced after the attached nicotinamide is released from the substrate, NGD+, but before the nucleophilic attack by water that would lead to the formation of GDPR. The observed structure is thus consistent with the predicted oxocarbenium ion intermediate or its glycal analogue (30Smar M. Short S.A. Wolfenden R. Biochemistry. 1991; 30: 7908-7912Crossref PubMed Scopus (33) Google Scholar). The guanine ring of the GDPR is very close to C-1′, with its guanine oxygen only 2.16 Å away (Fig. 3A). The close proximity of the guanine ring to the intermediate presumably blocks the access of water and greatly delays its nucleophilic attack on C-1′ of the intermediate, allowing it to accumulate sufficiently for visualization. In molecule A, the electron density for the nucleotide clearly shows that the nucleotide is neither a GDPRI (as there is no gap between densities of the ribose and guanine ring (Fig. 3B)) nor a rebound GDPR (as density attached to the far side of the ring extends out, whereas a rebound GDPR would end at guanine). Analysis of the electron density in molecule A points to a covalently linked GDPR dimer in the active site. That is, the newly formed oxocarbenium ion intermediate has been nucleophilically attacked by the guanine oxygen atom, resulting in a covalent C-O bond with a bond length of 1.67 Å. This bond length is similar to that observed for the nicotinamide-glycosidic bond in the Michaelis complex. The enzymatic formation of the GDPR dimer has not been reported previously, but an ADPR dimer has been detected during the enzymatic catalysis of NAD+ (31De Flora A. Guida L. Franco L. Zocchi E. Bruzzone S. Benatti U. Damonte G. Lee H.C. J. Biol. Chem. 1997; 272: 12945-12951Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar).FIGURE 3Stereo drawings of omit difference electron densities for the reaction intermediate and two reaction products. Only those parts within the active site are shown. All omit difference densities are represented by gray isomesh surfaces contoured at 2.5 σ. The omitted nucleotides are drawn as sticks, with the same color scheme as described for Fig. 1B, except that carbon atoms from the right-hand nucleotide are colored cyan. A, the oxocarbenium ion intermediate, GDPRI, and the guanine ring from a reaction product, GDPR. B, the GDPR dimer produced by nucleophilic attack of the guanine oxygen of GDPR on GDPRI, resulting in a covalent C-O linkage with a bond length of 1.67 Å. C, ADPR plus the adenine ring from a reaction product, ADP-ribose. Here the intermediate has been hydrolyzed by a water molecule, resulting in attachment of a hydroxyl group to the C-1′ carbon.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Enzyme-ADPR Complex Reveals Products with Divergent Conformations—To further validate the intermediate and its hydrolysis, we soaked wild-type CD38 crystals with NAD+ and determined their structure. In this way, we could observe the electron density for the final products, ADPRs, in the active site of both molecules (Fig. 4.). In molecule B, we observed density similar to that in Fig. 3A but with additional density present around the ribosyl C-1′ carbon (Fig. 3C). This density corresponds to an OH group added to the C-1′ carbon after the nucleophilic attack of the intermediate by a nearby water molecule, indicating the presence of ADPR, the fully hydrolyzed product from the substrate NAD+.It should be noted that the OH group is attached to the α-face of the ribosyl group (from the bottom), suggesting the existence of an inversion mechanism for NAD+ hydrolysis. Such an inverted anomic configuration is consistent with an ionic intermediate instead of a covalent intermediate linked to the C-1′ carbon from the α-face as reported previously (19Sauve A.A. Munshi C. Lee H.C. Schramm V.L. Biochemistry. 1998; 37: 13239-13249Crossref PubMed Scopus (98) Google Scholar). In the active site of molecule B, we also observed a somewhat noisy electron density corresponding to the adenine ring from another ADPR. Comparing the densities of the guanine and adenine rings (Fig. 3, A and C, respectively), it is clear that the guanine ring does have higher affinity for the active site, as judged by its well ordered presence in the crystal structure. In molecule A, we cannot find an ADPR dimer, although it has been proposed to exist (31De Flora A. Guida L. Franco L. Zocchi E. Bruzzone S. Benatti U. Damonte G. Lee H.C. J. Biol. Chem. 1997; 272: 12945-12951Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Unexpectedly, the electron density in the active site clearly shows that the bound species is a hydrolyzed product, ADPR, which has reentered in a reversed orientation (Fig. 4A). The electron density for this ADPR is excellent except at the terminal ribose (Fig. 4A). This is consistent with the ribose having a freely rotational single bond to the adjacent oxygen atom. Two phosphates of this ADPR form four hydrogen bonds with active site residues Ser126, Arg127, and Phe222. Ser193 and one water molecule have direct polar interactions with the two hydroxyl groups of the ADPR adenine ribose. Differing from residue Glu226 in GDPRI or GDPR dimer complexes, Glu226 has only indirect interactions to ADPR via a water molecule (Fig. 4A). The reentered ADPR complex can also be obtained by cocrystallizing NAD+ with native CD38 (data not shown). During the cocrystallization, CD38 hydrolyzes NAD+ to ADPR, which can reenter the active site of the enzyme and form a complex. Product Inhibition and Identification of CD38 Inhibitors Useful for Drug Design—Because the active site of CD38 has affinity for ADPR and GDPR, it has been suggested that ADPR and GDPR might exert product inhibition. Therefore, we measured the effect of ADPR and GDPR on the cyclization activity of CD38 using NGD+ as a substrate, forming cyclic GDP-ribose (cGDPR). As shown in Fig. 4B, inhibition by ADPR and GDPR was indeed observed in the mm range. The biochemical data confirm that the rebinding of ADPR to the active site of the enzyme can compete with other substrates and thus inhibit the enzymatic activities of CD38. ATP has long been identified as a CD38 inhibitor and is implicated in insulin secretion and diabetes (32Okamoto H. Ave. Exp. Med. Biol. 2003; 527: 243-252Crossref PubMed Scopus (13) Google Scholar). As ATP also has an adenine ring and guanine ring has a higher affinity to the active site over the adenine ring, it might be the case that guanine-containing nucleotides also inhibit CD38 activity. We therefore measured the inhibitory effect of GTP, GDP, and cGMP on CD38 cyclization activities and found that indeed all of them produce obvious inhibition of the cyclization reaction of NGD+ (Fig. 4B). For example, 0.1 mm GTP can inhibit cGDPR formation by 50%, revealing that GTP is an inhibitor. The affinity of the CD38 active site for guanine and adenine rings suggests the inclusion of the guanine ring moiety in the design of effective human CD38 inhibitors. Based on complex structures we determined an analogue of the GDPR dimer might be one of the human CD38 inhibitors. As the catalytic activity of CD38 has been confirmed to relate directly to some diseases such as diabetes (32Okamoto H. Ave. Exp. Med. Biol. 2003; 527: 243-252Crossref PubMed Scopus (13) Google Scholar), asthma (33Deshpande D.A. White T.A. Dogan S. Walseth T.F. Panettieri R.A. Kannan M.S. Am. J. Physiol. 2005; 288: L773-L788Crossref PubMed Scopus (101) Google Scholar), and inflammation (34Partida-Sanchez S. Goodrich S. Kusser K. Oppenheimer N. Randall T.D. Lund F.E. Immunity. 2004; 20: 279-291Abstract Full Text Full Text PDF PubMed Scopus (175) Google Scholar), the effective inhibitors of human CD38 activity might be good drug candidates for treating these diseases. Structural Features of the Intermediate—After the cleavage and release of nicotinamide, the newly formed oxocarbenium ion intermediate is stabilized by the formation of two hydrogen bonds between its two ribosyl 2′-,3′-OH groups and the carboxyl group of the catalytic residue Glu226. Structurally conserved water may also contribute to stabilization by forming an additional hydrogen bond with ribosyl 2′-OH (Fig. 5A). The guanine ring from a product GDPR is tightly associated within the active site by both hydrophilic interactions with Glu146 and Asp155 and hydrophobic interactions with Trp189. These interactions place the guanine ring parallel to the Trp189 ring, with the guanine oxygen pointing to the ribosyl C-1′ carbon at a distance of 2.16 Å. It should be noted that the C-1′ of the intermediate (GDPRI) is 2.95 Å away from the catalytically important residue Glu226, too far to form a covalent linkage with the enzyme. In addition, the two oxygen atoms on the Glu226 carboxyl group form two hydrogen bonds with two ribosyl OH groups, preventing it forming covalent linkage to C-1′ carbon. The geometry of the GDPRI structure demonstrates that at least when NGD+ is used as a substrate, the catalysis of CD38 does not involve a covalent intermediate. Structural Basis for Nicotinamide Cleavage—To examine the structural changes that occur during the cleavage of nicotinamide from NAD+ and the formation of the intermediate, the Michaelis complex and the intermediate complex were aligned based on least-squares minimization of all main-chain atoms on active site residues Trp189, Glu146, Asp155, Trp125, Ser193, and Glu226 (Fig. 5B). In the aligned model, the nicotinamide ring is coplanar with the guanine ring, with its amide nitrogen fixed by interactions to Glu146, Asp155, and Trp189 such that the anomeric C-1′ of the nicotinamide ribose is 3.22 Å from the catalytic residue Glu226 and 3.19 Å from the Ser193 OH group. In wild-type CD38, a glutamate is at position 226. Its carboxyl group has higher nucleophilicity than the carboxamide group in the E226Q mutant and favors the formation of an extra hydrogen bond with the 2′-OH of the nicotinamide ribose. This extra interaction appears to be critical for the enzymatic activity of CD38, as the E226Q mutant has greatly reduced activity. This is consistent with observations in the crystal structure of CD157 complexed with nicotinamide mononucleotide (NMN+). In that case, the hydroxyl groups of the NMN+-ri-bose were also observed to form hydrogen bonds with a glutamate (15Yamamoto-Katayama S. Ariyoshi M. Ishihara K. Hirano T. Jingami H. Morikawa K. J. Mol. Biol. 2002; 316: 711-723Crossref PubMed Scopus (83) Google Scholar). The aligned structures thus suggest that, after the release of the nicotinamide group, the ribose moiety is no longer restrained by steric hindrance from triplet Glu146-Asp155-Trp189 and rotates toward Glu226, with which it forms an additional hydrogen bond (Fig. 5B). The additional H-bond is seen in the structure of the intermediate. Intermediate Catalysis—The intermediate plays a central role in the multiple reactions catalyzed by human CD38. If NAD+ is used as the substrate, there are individual pathways from the intermediate to produce ADPR, cADPR, or an ADPR dimer. In the ADPR pathway, nucleophilic attack by a water molecule present in the active site forms ADPR (Fig. 5C). After hydrolysis of the intermediate, the newly formed ribosyl C-1′-OH group forms an additional hydrogen bond to Glu226, with a C-2′ ring pucker similar to that found in the Michaelis complex (Fig. 5C). The 2′-OH group, previously found to interact with Glu226, interacts with Ser193 OH through a hydrogen bond. In the cADPR pathway, the adenine end of the intermediate folds back into the active site, so that its N-1 atom can attack the C-1′ carbon of the intermediate to form cADPR (35Lee H.C. Aarhus R. Levitt D. Nat. Struct. Biol. 1994; 1: 143-144Crossref PubMed Scopus (138) Google Scholar). If the adenine N-1 of another ADPR attacks the intermediate, the dimerization reaction is favored to form an ADPR dimer. If NGD+ is used as the substrate, human CD38 can generate the corresponding products, GDPR, cGDPR (28Graeff R.M. Walseth T.F. Fryxell K. Branton W.D. Lee H.C. J. Biol. Chem. 1994; 269: 30260-30267Abstract Full Text PDF PubMed Google Scholar), and GDPR dimer. Regulative Role of Residue Ser193 during Catalysis—As described above, we observed the intermediate GDPRI in the active site of molecule B but a product GDPR-dimer in the other. Why do the occupants of the sites differ? In Fig. 6A, the two molecules have been aligned based on main-chain atoms of the conserved active site residues. According to the aligned structures, two guanine rings, one from the GDPR dimer and the other from the product GDPR, take the same orientation and position with their guanine oxygens close to the C-1′ carbons. Except for Ser193, all other active site residues from two molecules overlap quite well. As a hydrogen donor, the Ser193 OH group forms a hydrogen bond with the Trp189 main-chain oxygen (Fig. 6C), bringing the lone pair electrons of the oxygen in the -OH group to a suitable distance for a nucleophilic attack at the scissile bond in the NAD+ complex (Fig. 5B). Indeed, the importance of Ser193 in catalysis was verified by site-directed mutagenesis of Ser193, which resulted in greatly reduced enzymatic activity (Table 2). Sequence alignments of CD38 homologues show that Ser193 is strongly conserved among various species from invertebrate to human, supporting its functional role in the active site (Fig. 6B). As the nucleophilicity of the Glu226 carboxyl group is higher than that of the Ser193 hydroxyl group, the functional role of Glu226 is likely to be dominant in effecting the cleavage of the nicotinamide group. This is consistent with mutagenesis results showing that mutations at Glu226 are more effective in reducing the enzymatic activities than changes at Ser193. Nevertheless, the fact that both residues are close to C-1′ suggests that they may work in concert for efficient catalysis.TABLE 2Inhibition of NADase and GDP-ribosyl cyclase activities by mutation of Ser193CD38 and its mutantsNADaseGDP-ribosyl cyclasenmol/mg/minnmol/mg/minWild type115,366 ± 11,10714,812 ± 1,016S193T3,394 ± 6713,371 ± 349S193V647 ± 15776 ± 53 Open table in a new tab The side chain of Ser193 can adopt two alternative conformations. In the NAD+ and GDPR dimer complexes, Ser193 is close to the C-1′ carbon at distances of 3.19 and 3.39 Å, respectively. In the GDPRI intermediate, it shifts to the other conformation, with a longer distance of 4.76 Å to the C-1′ carbon (Fig. 6C). For both conformations of Ser193, its OH group forms a hydrogen bond with the Trp189 main-chain oxygen (Fig. 6C). It is thus likely that the long distance between the Ser193 OH group and C-1′ carbon reduces the stability of the intermediate and is responsible for the relatively weak density observed for C-1′ carbon seen in Fig. 3A. The unstable intermediate in molecule B could be the reason that the guanine oxygen does not attack the GDPRI intermediate nucleophilically even at a distance of 2.16 Å. In molecule A, Ser193 is only 3.39 Å from the C-1′ carbon, and the partial negative charge on its lone pair electrons can act to stabilize the cation intermediate. Hence the guanine oxygen can approach and attack the C-1′ carbon, forming the C-O bond seen in Figs. 3B and 6A. To summarize, the role of Glu226 in the activation of substrate NAD+, the formation of an ionic reaction intermediate after nicotinamide cleavage, the stabilization of the intermediate by residues Glu226 and Ser193, and the competitive pathways toward cADPR or ADPR are illustrated in Fig. 7. We thank Marliijn Hoogendoorn for expert help in producing and purifying the recombinant CD38, Tadhg Begley for useful discussion, and Marian Szebenyi for critical reading of the manuscript. The crystallographic data were collected at the Cornell High Energy Synchrotron Source (CHESS), which is supported by the National Science Foundation and National Institutes of Health NIGMS under Award DMR-0225180." @default.
• W2079504259 created "2016-06-24" @default.
• W2079504259 creator A5004956377 @default.
• W2079504259 creator A5005037936 @default.
• W2079504259 creator A5024966093 @default.
• W2079504259 creator A5039103109 @default.
• W2079504259 creator A5045363063 @default.
• W2079504259 creator A5058259419 @default.
• W2079504259 date "2006-10-01" @default.
• W2079504259 modified "2023-10-11" @default.
• W2079504259 title "Structural Basis for the Mechanistic Understanding of Human CD38-controlled Multiple Catalysis" @default.
• W2079504259 cites W1496899827 @default.
• W2079504259 cites W1539796472 @default.
• W2079504259 cites W1548497076 @default.
• W2079504259 cites W1559286577 @default.
• W2079504259 cites W1566643819 @default.
• W2079504259 cites W1834009592 @default.
• W2079504259 cites W1914120156 @default.
• W2079504259 cites W1966879697 @default.
• W2079504259 cites W1972403141 @default.
• W2079504259 cites W1976530842 @default.
• W2079504259 cites W1979845008 @default.
• W2079504259 cites W1981393301 @default.
• W2079504259 cites W1986505744 @default.
• W2079504259 cites W1987171087 @default.
• W2079504259 cites W1990637370 @default.
• W2079504259 cites W1991414624 @default.
• W2079504259 cites W2000232064 @default.
• W2079504259 cites W2000638509 @default.
• W2079504259 cites W2001641653 @default.
• W2079504259 cites W2001695130 @default.
• W2079504259 cites W2007480599 @default.
• W2079504259 cites W2008755938 @default.
• W2079504259 cites W2013083986 @default.
• W2079504259 cites W2015906906 @default.
• W2079504259 cites W2027723382 @default.
• W2079504259 cites W2039757972 @default.
• W2079504259 cites W2045266309 @default.
• W2079504259 cites W2064976108 @default.
• W2079504259 cites W2082340361 @default.
• W2079504259 cites W2095319998 @default.
• W2079504259 cites W2095837755 @default.
• W2079504259 cites W2138985573 @default.
• W2079504259 cites W2287444830 @default.
• W2079504259 doi "https://doi.org/10.1074/jbc.m606365200" @default.
• W2079504259 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/16951430" @default.
• W2079504259 hasPublicationYear "2006" @default.
• W2079504259 type Work @default.
• W2079504259 sameAs 2079504259 @default.
• W2079504259 citedByCount "56" @default.
• W2079504259 countsByYear W20795042592012 @default.
• W2079504259 countsByYear W20795042592013 @default.
• W2079504259 countsByYear W20795042592014 @default.
• W2079504259 countsByYear W20795042592015 @default.
• W2079504259 countsByYear W20795042592016 @default.
• W2079504259 countsByYear W20795042592017 @default.
• W2079504259 countsByYear W20795042592018 @default.
• W2079504259 countsByYear W20795042592019 @default.
• W2079504259 countsByYear W20795042592021 @default.
• W2079504259 countsByYear W20795042592022 @default.
• W2079504259 countsByYear W20795042592023 @default.
• W2079504259 crossrefType "journal-article" @default.
• W2079504259 hasAuthorship W2079504259A5004956377 @default.
• W2079504259 hasAuthorship W2079504259A5005037936 @default.
• W2079504259 hasAuthorship W2079504259A5024966093 @default.
• W2079504259 hasAuthorship W2079504259A5039103109 @default.
• W2079504259 hasAuthorship W2079504259A5045363063 @default.
• W2079504259 hasAuthorship W2079504259A5058259419 @default.
• W2079504259 hasBestOaLocation W20795042591 @default.
• W2079504259 hasConcept C12426560 @default.
• W2079504259 hasConcept C12554922 @default.
• W2079504259 hasConcept C127413603 @default.
• W2079504259 hasConcept C161790260 @default.
• W2079504259 hasConcept C183696295 @default.
• W2079504259 hasConcept C185592680 @default.
• W2079504259 hasConcept C2524010 @default.
• W2079504259 hasConcept C33923547 @default.
• W2079504259 hasConcept C55493867 @default.
• W2079504259 hasConcept C70721500 @default.
• W2079504259 hasConcept C86803240 @default.
• W2079504259 hasConceptScore W2079504259C12426560 @default.
• W2079504259 hasConceptScore W2079504259C12554922 @default.
• W2079504259 hasConceptScore W2079504259C127413603 @default.
• W2079504259 hasConceptScore W2079504259C161790260 @default.
• W2079504259 hasConceptScore W2079504259C183696295 @default.
• W2079504259 hasConceptScore W2079504259C185592680 @default.
• W2079504259 hasConceptScore W2079504259C2524010 @default.
• W2079504259 hasConceptScore W2079504259C33923547 @default.
• W2079504259 hasConceptScore W2079504259C55493867 @default.
• W2079504259 hasConceptScore W2079504259C70721500 @default.
• W2079504259 hasConceptScore W2079504259C86803240 @default.
• W2079504259 hasIssue "43" @default.
• W2079504259 hasLocation W20795042591 @default.
• W2079504259 hasOpenAccess W2079504259 @default.
• W2079504259 hasPrimaryLocation W20795042591 @default.
• W2079504259 hasRelatedWork W1967978480 @default.
• W2079504259 hasRelatedWork W1974837029 @default.
• W2079504259 hasRelatedWork W2066069824 @default.

• next