Inhibition of LuxS by S-ribosylhomocysteine analogues containing a [4-aza]ribose ring Venkata L.A. Malladi a , Adam J. Sobczak a,, Tiffany M. Meyer b , Dehua Pei b , Stanislaw F. Wnuk a,⇑ a Department of Chemistry and Biochemistry, Florida International University, Miami, FL 33199, USA b Department of Chemistry and Ohio State Biochemistry program, The Ohio State University, 100 West 18th Avenue, Columbus, OH 43210, USA article info Article history: Available online 28 July 2011 Keywords: Azahemiacetals Azasugars Homocysteine LuxS S-Ribosylhomocysteinase abstract LuxS (S-ribosylhomocysteinase) catalyzes the cleavage of the thioether linkage of S-ribosylhomocysteine (SRH) to produce homocysteine and 4,5-dihydroxy-2,3-pentanedione (DPD), the precursor to a small sig- naling molecule that mediates interspecies bacterial communication called autoinducer 2 (AI-2). Inhibi- tors of LuxS should interfere with bacterial interspecies communication and potentially provide a novel class of antibacterial agents. In this work, SRH analogues containing substitution of a nitrogen atom for the endocyclic oxygen as well as various deoxyriboses were synthesized and evaluated for LuxS inhibi- tion. Two of the [4-aza]SRH analogues showed modest competitive inhibition (K I 40 lM), while most of the others were inactive. One compound that contains a hemiaminal moiety exhibited time-dependent inhibition, consistent with enzyme-catalyzed ring opening and conversion into a more potent species (K I ⁄ = 3.5 lM). The structure–activity relationship of the designed inhibitors highlights the importance of both the homocysteine and ribose moieties for high-affinity binding to LuxS active site. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction Quorum sensing (QS) is a type of bacterial cell-to-cell commu- nication mediated through the production, release and detection of the small signaling molecules called autoinducers (AIs). 1–3 Such communication allows bacterial control of crucial functions in uni- ted communities for enhancement of symbiosis, virulence, antibi- otic production, biofilm formation, and many other processes. 4,5 Hence, there have been great interests in the synthesis of small molecules that can modulate QS pathways. 6–8 S-Ribosylhomocy- steinase (LuxS) is a key enzyme in the biosynthetic pathway of type II autoinducer, which mediates the interspecies quorum sens- ing among both Gram-positive and Gram-negative bacteria. The biosynthesis of AI-2 starts with the dual substrate-specific microbial enzyme 5 0 -methylthioadenosine/AdoHcy nucleosidase (MTAN), which catalyzes the depurination of S-adenosyl-L-homo- cysteine (SAH), a byproduct of many S-adenosyl-L-methionine- dependent methyltransferases reactions, to form S-ribosyl-L-homo- cysteine (SRH, Fig. 1). 9,10 SRH is subsequently converted to L-homocysteine and 4,5-dihydroxy-2,3-pentadione (DPD) by the LuxS enzyme. 11 DPD undergoes spontaneous cyclization to 1e and complexation with borate to form a furanosyl borate diester, which acts as the AI-2 in some bacteria. 11,12 Chemical synthesis of the unstable DPD has been accomplished recently by the groups of Jan- da 13 and Semmelhack, 14 which allowed the vital complexation properties of DPD with borate 15 to be studied and provided DPD as a reliable standard for investigation of AI-2 regulated QS processes. LuxS is a small metalloenzyme (157 amino acids in the Bacillus subtilis enzyme) containing Fe 2+ coordinated by His-54, His-58, Cys-126, and a water molecule. The native enzyme is unstable un- der aerobic conditions, but substitution of Co 2+ for Fe 2+ gives a highly stable variant with essentially wild-type catalytic activ- ity. 16–18 In the proposed catalytic mechanism, LuxS catalyzes con- secutive aldose–ketose (1a ? 1b) and ketose–ketose (1b ? 1c) isomerization steps and then b-elimination of Hcy from a 3-keto intermediate (1c ? 1d) to form DPD. 11,19 LuxS-catalyzed cleavage of the C5–S thioether bond in SRH is analogous to that of SAH hydrolase, which effects cleavage of an equivalent thioether bond in SAH by first oxidizing the C3 0 secondary alcohol into a ketone with an NAD + cofactor. 20,21 Zhou and co-workers designed and synthesized two LuxS sub- strate analogues, the S-(anhydroribosyl)-L-homocysteine (2) and S-(homoribosyl)-L-cysteine compounds, which blocked initial and final mechanistic steps, respectively (Fig. 2). 22 Pei and co-workers have prepared a series of stable analogues of the putative enedio- late intermediate, some of which showed submicromolar inhibi- tion of the enzyme (e.g., K I = 0.72 lM for isostere 3). 23 Zhang et al. found that the brominated furanones 4 covalently modify and inactivate LuxS. 24 Recognizing structural similarities between substrates of mammalian AdoHcy hydrolase and bacterial 0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.bmc.2011.07.043 ⇑ Corresponding author. Tel.: +1 305 348 6195; fax: +1 305 348 3772. E-mail address: wnuk@fiu.edu (S.F. Wnuk). On a faculty leave from University of Life Sciences, Department of Chemistry, Poznan, Poland Bioorganic & Medicinal Chemistry 19 (2011) 5507–5519 Contents lists available at SciVerse ScienceDirect Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc
13
Embed
Bioorganic & Medicinal Chemistryfaculty.fiu.edu/~wnuk/Publications-s/113.pdfOn a faculty leave from University of Life Sciences, Department of Chemistry, Poznan, Poland Bioorganic
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier .com/locate /bmc
Inhibition of LuxS by S-ribosylhomocysteine analogues containinga [4-aza]ribose ring
Venkata L.A. Malladi a, Adam J. Sobczak a,�, Tiffany M. Meyer b, Dehua Pei b, Stanislaw F. Wnuk a,⇑a Department of Chemistry and Biochemistry, Florida International University, Miami, FL 33199, USAb Department of Chemistry and Ohio State Biochemistry program, The Ohio State University, 100 West 18th Avenue, Columbus, OH 43210, USA
� On a faculty leave from University of Life SciencPoznan, Poland
a b s t r a c t
LuxS (S-ribosylhomocysteinase) catalyzes the cleavage of the thioether linkage of S-ribosylhomocysteine(SRH) to produce homocysteine and 4,5-dihydroxy-2,3-pentanedione (DPD), the precursor to a small sig-naling molecule that mediates interspecies bacterial communication called autoinducer 2 (AI-2). Inhibi-tors of LuxS should interfere with bacterial interspecies communication and potentially provide a novelclass of antibacterial agents. In this work, SRH analogues containing substitution of a nitrogen atom forthe endocyclic oxygen as well as various deoxyriboses were synthesized and evaluated for LuxS inhibi-tion. Two of the [4-aza]SRH analogues showed modest competitive inhibition (KI �40 lM), while mostof the others were inactive. One compound that contains a hemiaminal moiety exhibited time-dependentinhibition, consistent with enzyme-catalyzed ring opening and conversion into a more potent species(KI⁄ = 3.5 lM). The structure–activity relationship of the designed inhibitors highlights the importance
of both the homocysteine and ribose moieties for high-affinity binding to LuxS active site.� 2011 Elsevier Ltd. All rights reserved.
1. Introduction
Quorum sensing (QS) is a type of bacterial cell-to-cell commu-nication mediated through the production, release and detectionof the small signaling molecules called autoinducers (AIs).1–3 Suchcommunication allows bacterial control of crucial functions in uni-ted communities for enhancement of symbiosis, virulence, antibi-otic production, biofilm formation, and many other processes.4,5
Hence, there have been great interests in the synthesis of smallmolecules that can modulate QS pathways.6–8 S-Ribosylhomocy-steinase (LuxS) is a key enzyme in the biosynthetic pathway oftype II autoinducer, which mediates the interspecies quorum sens-ing among both Gram-positive and Gram-negative bacteria.
The biosynthesis of AI-2 starts with the dual substrate-specificmicrobial enzyme 50-methylthioadenosine/AdoHcy nucleosidase(MTAN), which catalyzes the depurination of S-adenosyl-L-homo-cysteine (SAH), a byproduct of many S-adenosyl-L-methionine-dependent methyltransferases reactions, to form S-ribosyl-L-homo-cysteine (SRH, Fig. 1).9,10 SRH is subsequently converted toL-homocysteine and 4,5-dihydroxy-2,3-pentadione (DPD) by theLuxS enzyme.11 DPD undergoes spontaneous cyclization to 1e andcomplexation with borate to form a furanosyl borate diester, whichacts as the AI-2 in some bacteria.11,12 Chemical synthesis of the
ll rights reserved.
: +1 305 348 3772.
es, Department of Chemistry,
unstable DPD has been accomplished recently by the groups of Jan-da13 and Semmelhack,14 which allowed the vital complexationproperties of DPD with borate15 to be studied and provided DPDas a reliable standard for investigation of AI-2 regulated QSprocesses.
LuxS is a small metalloenzyme (157 amino acids in the Bacillussubtilis enzyme) containing Fe2+ coordinated by His-54, His-58,Cys-126, and a water molecule. The native enzyme is unstable un-der aerobic conditions, but substitution of Co2+ for Fe2+ gives ahighly stable variant with essentially wild-type catalytic activ-ity.16–18 In the proposed catalytic mechanism, LuxS catalyzes con-secutive aldose–ketose (1a ? 1b) and ketose–ketose (1b ? 1c)isomerization steps and then b-elimination of Hcy from a 3-ketointermediate (1c ? 1d) to form DPD.11,19 LuxS-catalyzed cleavageof the C5–S thioether bond in SRH is analogous to that of SAHhydrolase, which effects cleavage of an equivalent thioether bondin SAH by first oxidizing the C30 secondary alcohol into a ketonewith an NAD+ cofactor.20,21
Zhou and co-workers designed and synthesized two LuxS sub-strate analogues, the S-(anhydroribosyl)-L-homocysteine (2) andS-(homoribosyl)-L-cysteine compounds, which blocked initial andfinal mechanistic steps, respectively (Fig. 2).22 Pei and co-workershave prepared a series of stable analogues of the putative enedio-late intermediate, some of which showed submicromolar inhibi-tion of the enzyme (e.g., KI = 0.72 lM for isostere 3).23 Zhanget al. found that the brominated furanones 4 covalently modifyand inactivate LuxS.24 Recognizing structural similarities betweensubstrates of mammalian AdoHcy hydrolase and bacterial
Figure 1. Biosynthetic pathway to AI-2. Enzymatic conversion of SRH to DPD by LuxS.
O
OH
SHO
O
NH2
HO
2
SNH
O
OH
OH
OHO
HONH2
3
O
OH
SHO
O
NH2
X5 X = H, F, Br, OMe
OH
OO
Br
Br
4
Figure 2. LuxS inhibitors.
5508 V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519
S-ribosylhomocysteine (SRH) hydrolase (LuxS enzyme), we de-signed and synthesized SRH analogues with 6-(fluoro)vinyl moietyin place of the C5 and sulfur atoms which acted as weak/moderateinhibitors of LuxS enzyme.25 The SRH analogues 5 lacking enoliz-able hydroxyl group at C3 were found to be competitive substrateof LuxS.26,27 The time dependence inhibition with C3 halogenatedsubstrates was caused by enzyme-catalyzed elimination of halideions.27
Here, we report synthesis of [4-aza]SRH mimics in which thefuranose ring oxygen has been replaced by a nitrogen atom. Theresulting hemiaminals should have different stabilities28 relativeto the O,O-hemiacetals present in SRH and as a result differentrates of metabolic alteration. The higher basicity of the aza ana-logues is expected to have different binding strengths and ratesfor productions of the open chain aldehyde form–necessary forthe first isomerization to occur. Also, the aminosugars can beprotonated at physiological pH and the corresponding positivecharge may have an effect on binding to the enzymatic active site.Azasugars29 have been found to be potent inhibitors of glycosi-dases and glycosyltransferases30,31 and have been targeted astransition-state models.32,33 The 40-azanucleosides34 function astransition-state inhibitors of MTAN at the femtomolar level.35,36
2. Results and discussion
2.1. Chemistry
Our first target was 1,4-dideoxy-[4-aza]SRH 12 lacking the hy-droxyl group at C1 (Scheme 1). Compound 12 cannot undergo ringopening (which will preclude the initial step of the LuxS-catalyzedreaction) and may act as a competitive inhibitor of LuxS. Synthesisof 12 started with the protected 1-amino-1,4-anhydro-1-deoxy-D-ribitol 6, which was readily prepared from the commerciallyavailable D-gulonic acid c-lactone.37 However, attempted mesyla-tion of the N-benzyl protected 6 resulted in the formation of piper-idine derivative 13 as a mixture of two diastereomers (�3:1).Presumably, the mesylated pyrrolidine underwent a rearrange-ment reaction into the piperidines through an aziridine intermedi-ates.38,39 We found that replacement of the benzyl protectinggroup at ring nitrogen of 6 with a Boc group suppressed the nucle-ophilicity of the nitrogen and prevented ring expansion, allowingthe formation of stable 5-O-mesyl derivatives. Thus, silylation of6 with TBDMSCl and subsequent hydrogenation (5% Pd/C) in thepresence of (Boc)2O39,40 yielded 8 (97% from 5). Desilylation of 8(70%) followed by mesylation gave 10 as a stable compound(96%). Displacement of the mesylate group with a thiolate, gener-ated by reduction of properly protected L-homocystine19 withwater soluble tris(2-carboxyethyl)phosphine hydrochloride,26 gavethioether 11 (86%). Treatment of 11 with TFA effected simulta-neous removal of the N-Boc, acetonide and t-butyl ester protectiongroups to give the desired [4-aza]SRH analogue 12 in good yields(66%).
The second target was c-lactam 21, which contains an amidecarbonyl at C1 and nitrogen as a replacement of the ring oxygen(Scheme 2). It is noteworthy that, as opposed to the [4-aza]SRHanalogue 12 (or 23), the lactam nitrogen cannot be protonated atphysiological pH. Selective oxidation of the 5-O-TBDMS-azasugar7 at C1 with RuO2/NaIO4 under EtOAc/H2O biphasic conditions41
produced N-benzyl lactam 14a (65%) and a small amount (18%)of the corresponding N-benzoylpyrrolidinone byproduct, resultedfrom oxidation of the benzylic carbon of the N-protecting group.Desilylation of 14a with TBAF, followed by mesylation anddisplacement of the mesylate group with protected Hcy gave
N
OO
ROBn
6 R = H7 R = TBDMS
bN
OO
ROBoc
8 R = TBDMS9 R = H
cN
OO
MsOBoc
10
HN
OHOH
SHO
NH2
O
12
e
fN
OO
Boc
St-BuO
NHBoc
O
11
N
OOCl
13
a d
Bn
c
Scheme 1. Reagents and conditions: (a) TBDMSCl/imidazole/DMAP/CH2Cl2/rt; (b) H2/Pd-C/(Boc)2O/Et3N/EtOH/rt; (c) MsCl/Et3N/CH2Cl2/rt; (d) TBAF/THF/rt; (e)BocNHCH(CH2CH2SH)CO2t-Bu/LDA/DMF; (f) (i) TFA, (ii) TFA/H2O.
V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519 5509
thioether 18. Treatment of 18 with TFA removed all of the acid-la-bile protection groups to yield N-benzyl protected [4-aza]SRH lac-tam 20 (48%). However, all attempts to remove the N-benzyl groupfrom 18 or 20 (to yield 21) were unsuccessful [e.g., H2/Pd-C orPd(OH)2-C, Na/NH3(liq.), BCl3]. Our attempt to mesylate the
N-Boc protected 15b (prepared by RuO2-catalyzed oxidation of 8and desilylation of the resulting 15a) failed to produce 17, yieldingonly the starting material 15b. Fortunately, oxidation of the 5-O-mesyl and N-Boc protected pyrrolidine 10 with RuO2/NaIO4
afforded 17 efficiently (95%). Coupling of 17 with homocysteinate
5510 V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519
afforded thioether 19a with concomitant loss of Boc group at ringnitrogen. Subsequent deprotection with TFA followed by TFA/H2Ogave [4-aza]SRH lactam 21 (58%).
Our next target was hemiaminal 23. Since only the open alde-hyde form of SRH is catalytically active, we were interested inthe effect of the nitrogen substitution on the ring opening. Theexistence of azahemiacetals in equilibrium with dehydrated form(imine) as well as with open aldehyde and dimeric forms was re-ported for 4-azapentofuranoses.42,43 It is noteworthy that sugarN,O-acetals were found to be stable enough to undergo couplingwith nucleoside bases,41 or transformation to proline.44 Althoughdirect reduction of lactam 21 (or 19a) with LiBEt3H failed to yieldhemiaminal 23 (or 22), the protection of the ring nitrogen with aBoc group facilitated the reduction reaction.44 Thus, treatment of19a with (Boc)2O/DMAP gave N-Boc protected lactam 19b (93%)which upon treatment with LiBEt3H produced hemiaminal 22(92%) as a mixture of two anomers. Deprotection of 22 with TFAfollowed by TFA/H2O gave desired [4-aza]SRH (N,O-acetal) ana-logue 23 (72%) as a mixture of a/b anomers. Interestingly, no freealdehyde or imine proton peaks were visible on 1H NMR spectra.Compound 23 is stable when kept at 4 �C but decomposes slowlyin solution at ambient temperature especially at basic pH.
To determine whether the cyclic [4-aza]SRH exists in equilib-rium with the open chain aldehyde form, we carried out a limitedmodel study. Thus, N-Boc protected lactam 15a44 was reducedwith LiBEt3H to afford protected hemiaminal 24a (Scheme 3). Desi-lylation with TBAF yielded 24b, which was treated with TFA to give
d N
O
HO
24a R =24b R =
a
NH2
OHOH
HO N
H
OBn
26
15a R = TBDMS
N
O
ROBo
N
OO
ROBoc
O
b
Scheme 3. Reagents and conditions: (a) LiEt3BH/THF/-78 �C; (b) TBA
deprotected hemiaminal 25a as a mixture of anomers susceptibleto dehydration at pH higher than 7 to form imine 25c.43 Subse-quent treatment of 25a with O-benzylhydroxylamine gave ex-pected oxime 26 as the only product. The formation of oxime 26indicates that azasugar 25a exists in equilibrium with the openaldehyde form (25b) and that the equilibrium could be shifted bysubsequent transformations.
2,3,4-Trideoxy-[4-aza]SRH 38 lacking the enolizable hydroxylgroups at C2 and C3 was next prepared to examine the importanceof C2 and C3-OH groups for LuxS binding and catalysis. The keystarting material (S)-5-(bromomethyl)-2-pyrrolidone (27) wasconveniently prepared from L-pyroglutamic acid45 (Scheme 4). Dis-placement of the bromide in 27 with the L-homocysteinate affor-ded thioether 33 (79%), which was deprotected with TFAquantitatively to give 2,3-dideoxy-4-azaSRH analogue 34 as a tri-fluoroacetate. Displacement with the unprotected D/L-homocyste-ine produced racemic 36 (75%) as a sodium salt, which upontreatment with TFA was also converted to its trifluoroacetate salt.As expected, 1H NMR spectrum of 34 showed only one set of peakswhich are present in the spectrum of racemic 36. Treatment of 33with (Boc)2O/DMAP gave the N-Boc protected lactam 35, whichwas reduced with LiBEt3H to give hemiaminal 37. Subsequentdeprotection with TFA produced 38.
The 5-S-alkyl-2,3-dideoxy-[4-aza]SRH (e.g., 28/29) and the 5-S-alkyl-[4-aza]SRH analogues with different length of the alkylthiochain were also prepared.46 These cyclic azahemiacetals and theirancestor lactams were found to modulate Pseudomonas aeruginosa
H2
OHH
O
TBDMS H
H
O
c
OH c N
OHOH
HOH
OH
25a
25b
N
OHOH
HO
25c
(± H2O)
F/THF/rt; (c) (i) TFA/0 �C, (ii) TFA/H2O/0 �C; (d) BnONH2/pyr/rt.
RN
OSR''OOC
NHR'
33 R = H, R' = Boc, R'' = t-Bu34 R = R' = R'' = H35 R = R' = Boc, R'' = t-Bu36 R = R' = R'' = H (9R/S)
V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519 5511
QS.46 The alkylthiomethyl azahemiacetal 28/29 existed in solutionas an equilibrium mixture of anomers along with the open chainaldehydes [5–25%, 1H NMR (d 8.90), 13C NMR (d 180.8)] and thecorresponding imines 30/31 [3–30%; 1 H NMR (d 7.63), 13C NMR(d 167.0)].46 Treatment of 28 with O-benzylhydroxylamine alsoproduced the expected oxime 32,46 as observed for 25a.
To explore the possibility of the LuxS-mediated addition ofwater across carbon–nitrogen double bond, we synthesized animine-type analogue 43 (Scheme 5). The precursor 1-methylimino-cyclitol 39 was prepared by the Moriarty rearrangement47 of theexo-imino to endo-iminocyclitol, which involves inversion at C4of the L-lyxo sugar to give the D-ribo azasugar. The imine 3947
was mesylated at the primary alcohol to give 40 (85%), whichwas coupled with protected L-Hcy to give 41 (85%, Scheme 5).Treatment of 41 with TFA for a short time gave only isopropylideneprotected 42. We found that the protons at the C1-methyl groupare exchangeable with deuterium within few hours when com-pound 42 is dissolved in D2O. Treatment of 42 with aqueous TFA(9:1) yielded fully deprotected 43 in quantitative yield. Protonsat C1-methyl group of 43 were also exchangeable with deuterium.These exchange indicate that 1,4-ketimine-SRH analogue 43 mightbe expected to undergo enzyme-catalyzed hydrolysis to generate a[4-aza]SRH analogue with a methyl ketone rather than an aldehydeat C1. This change might affect the regioselectivity and rate of thefirst isomerization step in the LuxS-catalyzed reaction. We alsoproved that the methyl group protons in 39 are not susceptibleto exchange even if 39 was dissolved in D2O for several hours.Additionally, we noticed that observed low rate of exchange in39 (relatively to 42 and 43) can be enhanced exclusively in thepresence of acid or amino acid (TFA and glycine were used,respectively). Attempted, one-step deprotection of 41 with BCl3
led to a partial loss of chirality at C9 giving 43 as a mixture ofdiastereomers (2:3).
Our attempt to prepare the imine derivative of [4-aza]SRH wasunsuccessful. Thus, debenzylation of 7 and treatment of theaminoribitol 44 with N-chlorosuccinimide (NCS) followed bydehydrochlorination of the resulting N-chloroamine with lithiumtetramethylpiperidine gave unstable aldoimine of type 39 (Hinstead of CH3), as reported.48 However, couplings of suchaldoimine with Hcy to give the imine SRH analogue failed. Acid-catalyzed hydrolysis of such imine analogue could serve as analternative route to 4-azaSRH 23. Also, enzyme-mediated proton-ation of the imine nitrogen atom and the addition of water mightgenerate 23 and/or new species with an ‘amino group’ within theenzyme active site.
A nitrone analogue of SRH 49 was also targeted. Since nitronesare more electrophilic than imines such analogue might act as irre-versible inhibitors by forming a covalent adduct(s) with enzyme. Itis noteworthy that nitrones are overall neutral and cannot be pro-tonated at physiological pH. Thus, treatment of the aminoribitol 44with SeO2/H2O2 gave nitrone 4549 (74%; Scheme 6). Desilylationand subsequent mesylation gave 47 (56%). Coupling of 47 with
NRO NSROOC
NHR'
41 R = t-Bu, R' = Boc, R'',R'' = CMe242 R = R' = H, R'',R'' = CMe243 R = R' = R'' = H
protected Hcy afforded a nitrone-SRH derivative 48 (43%). Depro-tection of 48 with TFA produced unstable nitrone derivative 49(40%).
2.2. Inhibition of LuxS
Compounds 12, 20, 21, 23, 28, 36, 38, and 43 were evaluated aspotential inhibitors of Co(II)-substituted B. subtilis LuxS. Compound12 inhibited LuxS in a concentration-dependent manner that isconsistent with competitive inhibition (Fig. 3a), with a KI value of48 lM (Table 1). Similarly, lactam 21 also behaved as a competitiveinhibitor with KI value of 37 lM. As expected, the lactam 20, whichcontains a bulky benzyl group at the ring nitrogen, was found to beinactive, likely due to steric reasons. Compounds 36 and 38 wereboth inactive toward LuxS, highlighting the importance of the ri-bose hydroxyl groups for enzyme binding. The proposed mecha-nism predicts that the C2 and C3 hydroxyl groups directlycoordinate with the catalytic metal ion during different catalyticsteps (Fig. 1). The lack of activity of compound 43, which containsa methyl group instead of a hydroxyl group at the C1 position, maybe caused by both loss of favorable interactions with the OH groupand the bulky size of the methyl group. Collectively, these resultssuggest that proper interactions between the ribose ring and theenzyme active site critically contribute to the formation of a pro-ductive E–S complex and subsequent catalysis.
Unlike the other analogues described above, inhibition of LuxSby the hemiaminal-containing analogue 23 was time dependent(Fig. 3b). Its inhibition kinetics can be described by the slow-bind-
ing equationE + I
K I
kK I*
E I
E I*
where KI is the equilibrium constant
for the formation of the initial E�I complex, k is the rate constantfor the conversion of the E�I complex to the tighter E�I⁄ complex,and KI
⁄ represents the dissociation constant of the E�I⁄ complex.To assess its potency, different concentrations of compound 23were preincubated with LuxS for 30 min at 4 �C and the residualenzymatic activity was measured. Plot of the residual activityagainst the inhibitor concentration resulted in an IC50 value of60 lM, from which a KI
⁄ value of 3.5 lM was estimated (Table 1).Unfortunately, the complex inhibition kinetics precluded an accu-rate determination of the KI value. While further work is clearly
5512 V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519
necessary to determine the exact mechanism of inhibition by 23,we propose a working hypothesis to explain the observed timedependence (Fig. 4). Since compounds 12 and 21, which are struc-turally similar to 23, did not exhibit time-dependent inhibition andour model study shows that the hemiaminal 25a exists in equilib-
Table 1Inhibition constants of [4-aza]SRH analogues against B. subtilis LuxS
Compound KI or KI⁄ (lM)
12 4821 3723 3.5
Figure 3. Inhibition of LuxS by compounds 12 and 23. (A) Reaction progress curvesin the presence of increasing concentrations of inhibitor 12 (0, 200, 400, 800, 1600,and 3200 lM). The last two curves were control reactions in the absence of LuxS.Inset, plot of remaining LuxS activity as a function of inhibitor 12 concentration. (B)Reaction progress curves of LuxS in the presence of increasing concentrations ofinhibitor 23 (0, 20, 40, and 50 lM) (without preincubation). Inset, plot of remainingLuxS activity as a function of inhibitor 12 concentration (after 30 minpreincubation).
HN
RS
OHOH OH
NH2
RS
OOH OH
M2+
H
M2+
23 23a
Figure 4. Proposed mechanism for the time-dependen
rium with the free aldehyde form (25b), we propose that hemiami-nal 23 may undergo ring opening to form aldehyde 23a. Due to itsstructural similarity to catalytic intermediate 1a (Fig. 1), 23a mayundergo the aldose-ketose isomerization reaction to form 2-ketone23b, which presumably binds to the LuxS active site with higheraffinity than the ribose analogue 23. This behavior is very similarto that of a class of halogenated SRH analogues (e.g., [3-F]SRHand [3-Br]SRH), which have been shown to undergo LuxS-catalyzed ring opening to form open-chain species that are morepotent LuxS inhibitors than the initial ribose analogues.27
The remaining compound 28 and its ancestor lactam showed nosignificant inhibition of LuxS.
3. Conclusions
We have synthesized [4-aza] S-ribosylhomocysteine analoguesin which the furanose ring oxygen has been substituted by anitrogen atom having also the additional modifications atanomeric carbon. Coupling of the protected 4-amino-5-O-meth-anesulfonyl-4-deoxy-D-ribono-1,4-lactam with homocysteinateand subsequent deprotection with TFA gave [4-aza]SRH with anamide carbonyl at anomeric carbon. Reduction of the N-Bocprotected lactam with LiBEt3H and acid catalyzed deprotectionproduced [4-aza]SRH hemiaminal analogue. The [4-aza]SRHanalogue lacking the hydroxyl group at C1 and the correspondinglactam derivative showed modest competitive inhibition (KI
�40–50 lM) of LuxS. The hemiaminal analogue exhibited time-dependent inhibition (KI
⁄ = 3.5 lM), consistent with the enzyme-catalyzed ring opening and generation of 2- and/or 3-ketoneintermediates, which presumably bind to the LuxS active site withhigher affinity than the ribose natural substrate.
4. Experimental procedure
The 1H (400 or 600 MHz) and 13C (100 MHz) NMR spectra weredetermined with solutions in CDCl3 unless otherwise noted. Massspectra (MS) were obtained with atmospheric pressure chemicalionization (APCI) technique and HRMS in AP-ESI or TOF-ESI mode.TLC was performed with Merck kieselgel 60-F254 sheets productswere detected with 254 nm light or by visualization withCe(SO4)2/(NH4)6Mo7O24�4H2O/H2SO4/H2O reagent. Merck kieselgel60 (230–400 mesh) was used for column chromatography. HPLCpurifications were performed using XTerra� preparative RP18
OBDTM column (5 lm 19 � 150 mm) with gradient program usingCH3CN/H2O as a mobile phase. Reagent grade chemicals were used,and solvents were dried by reflux over and distillation from CaH2
To a stirred solution of 637 (150 mg, 0.57 mmol) in anhydrousCH2Cl2 (5 mL) at rt under Ar atmosphere were added DMAP(7 mg, 0.05 mmol) and imidazole (93 mg, 1.36 mmol) followed byTBDMSCl (103 mg, 0.68 mmol). The mixture was then stirred for
NH2
RSHO
OH O
NH2RS
HO
M2+23b 23c
O OH
M2+
?
t inhibition of LuxS by [4-aza]SRH hemiaminal 23.
V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519 5513
A solution of 7 (145 mg, 0.38 mmol), triethylamine (0.105 mL,0.76 mmol), di-tert-butyldicarbonate (126 mg, 0.57 mmol) andPd/C (5%, 300 mg) in ethanol (6 mL) was stirred under an atmo-sphere of hydrogen at room temperature for 6 h. The reaction mix-ture was filtered through Celite to remove the catalyst. The Celitewas washed with ethanol (5 mL) and washings and the filtratewere combined and evaporated. The residue was partitioned(EtOAc//�NaHCO3/H2O). The organic layer was washed (brine),dried (MgSO4) and evaporated. The residue was column chromato-graphed (20 ? 30% EtOAc/hexane) to give 8 (147 mg, 99%) withspectral properties as reported.44
TBAF (1 M/THF; 0.25 mL, 0.25 mmol) was added to a stirredsolution of 8 (66 mg, 0.17 mmol) in THF (5 mL) at ambient temper-ature. After stirring for 30 min, the reaction mixture was parti-tioned (EtOAc//�NaHCO3/H2O). The organic layer was washed(brine), dried (MgSO4) and evaporated. The residue was columnchromatographed (50 ? 60% EtOAc/hexane) to give 9 (32 mg,70%) with spectral properties as reported.51
TBAF (1 M/THF; 0.18 mL, 0.18 mmol) was added dropwise to astirred solution of 14a (49 mg, 0.12 mmol) in THF (10 mL) at 0 �C.After 5 min, the ice-bath was removed and reaction mixture wasallowed to stir at ambient temperature for 2 h. The reaction mix-ture was quenched with water and volatiles were evaporated.The residue was partitioned (EtOAc//�NaHCO3/H2O) and the organ-ic layer was washed (brine), dried (MgSO4) and evaporated. Theresidue was column chromatographed (80 ? 90% EtOAc/hexane)to give 14b52as a white solid (33 mg, 97%): 1H NMR d 1.37 (s, 3,CH3), 1.47 (s, 3, CH3), 3.35 (s, 1, OH), 3.54 (‘t’, J = 1.8 Hz, 1, H4),3.64 (d of m, J = 11.8 Hz, 1, H5), 3.85 (br d, J = 12.0 Hz, 1, H5’),
Oxidation of 8 (90 mg, 0.23 mmol) with NaIO4 (126 mg,0.7 mmol) and RuO2�xH2O (8 mg, 0.05 mmol) by Procedure D [col-umn chromatography (50 ? 60% EtOAc/hexane)] gave 15a (56 mg,60%) as a colorless oil with data as reported.44
Reduction of 15a44 (68 mg, 0.17 mmol) with LiEt3BH (1 M/THF;0.43 mL, 0.43 mmol) in anhydrous THF (2 mL) at -78 �C by the pro-cedure F gave 24a44 (68 mg, 100%) as a colorless oil with data asreported.
5516 V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519
A solution of the crude 25a (13 mg. 0.09 mmol) and O-ben-zylhydroxylamine hydrochloride (43 mg, 0.27 mmol) in anhydrouspyridine (4 mL) was stirred under an atmosphere of nitrogen atroom temperature for 12 h. Pyridine was evaporated to afford 26of sufficient purity (�95%) for spectroscopic characterization to-gether with the excess of BnONH2 used: 1H NMR (MeOH-d4) d3.52 (‘dt’, J = 4.1, 8.4 Hz, 1, H4), 3.79 (dd, J = 3.8, 11.5 Hz, 1, H5),3.85 (dd, J = 4.4, 8.7 Hz, 1, H3), 3.94 (dd, J = 8.4, 11.5 Hz, 1, H5’),4.13 (dd, J = 6.8, 8.7 Hz, 1, H2), 4.92–5.16 (2H, Bn; signals for ben-zylic protons were within the envelope of the solvent peak butcross peaks between them were observed in COSY), 7.41 (1, H1;signal for H1 was within the envelope of protons from benzylgroup but cross peaks of H1 to H2 were observed in COSY), 7.35–7.47 (m, 5H, Bn); 13C NMR d 56.2 (C4), 58.8 (C5), 71.2 (C3), 71.6(C2), 77.0 (Bn, confirmed by HETCOR), 128.9, 129.3, 129.5, 139.1(Bn), 151.9 (C1); MS (ESI) m/z 255 (100, MH+).
Treatment of 2745 (5S; 18 mg, 0.1 mmol) with protected L-Hcy(35 mg, 0.12 mmol) by Procedure B (step a and b, 48 h) gave50.5 mg of the yellow oil. This material was column chromato-graphed (EtOAc/MeOH, 19:1) to give 33 (31 mg, 79%) as a colorlessoil: 1H NMR d 1.45 (s, 9), 1.48 (s, 9), 1.77–1.92 (m, 2, H3,8), 2.02–2.14 (m, 1, H8’), 2.27–2.47 (m, 3, H2,2’,3’), 2.51–2.57 (dd, J = 8.1,
Stirring sodium salt of 36 (10 mg) in TFA (1 mL) for 1 h at ambi-ent temperature and evaporation of volatiles gave trifluoroacetateof 36 as a mixture of diastereomers (9R/S, �1:1): 1H NMR (D2O) d1.79–1.89 (m, 1, H3), 2.09–2.19 (m, 1, H8), 2.21–2.28 (m, 1, H8’),2.28–2.33 (m, 1, H3’,), 2.34–2.48 (m, 2, H2,2’), 2.678 (dd, J = 6.5,13.6 Hz, 0.5, H5, 9R), 2.683 (dd, J = 6.5, 13.6 Hz, 0.5, H5, 9S), 2.72(t, J = 7.6 Hz, 2, H7,7’), 2.776 (dd, J = 5.4, 13.6 Hz, 0.5, H5’, 9R),2.781 (dd, J = 5.4, 13.6 Hz, 0.5, H5’, 9S), 3.94 (‘quin’, J = 6.3 Hz, 1,H4), 4.181 (t, J = 6.4 Hz, 1, H9R), 4.186 (t, J = 6.4 Hz, 1, H9S).
V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519 5517
Chemical shifts observed for TFA salt of 36(9R/S) were differentfrom its sodium salt but parallel the signals for the trifluoroacetateof 34(9S) derived from Hcy.
Treatment of 37 (20 mg, 0.04 mmol) with an excess of TFA(1 mL) by Procedure C (step a, 2 h at ambient temperature) gavetrifluoroacetate of 38 (13 mg, 95%; light yellow oil) as a complexmixture of isomers accompanied (�10%) by the open aldehydeform [1H NMR d 8.89 (s)]: MS (APCI) m/z 217 (100, [M-17]+).
Treatment of 41 (62 mg, 0.136 mmol) with TFA by Procedure C(step a, ambient temperature) gave 42 (41 mg, 99%) as a colorlessoil: 1H NMR (D2O) d 1.38 (s, 3), 1.41 (s, 3), 2.10–2.20 (m, 1, H8),
A stirred solution of 44 (70 mg, 0.24 mmol) and SeO2
(0.01 mmol, 1.1 mg) in acetone (3 mL) was cooled to �4 �C underN2 atmosphere and H2O2 (25%) was added slowly (3–4 h) untilthe reaction was completed (as judged by TLC). Volatiles are evap-orated and the residue was partitioned (EtOAc//NaHCO3/H2O).Theorganic layer was collected, washed (brine) and dried (MgSO4).The resulting solid was chromatographed (50% EtOAc/hexane) togive 45 (54 mg, 73%) as a white solid with data as reported.49
Treatment of 48 (72 mg, 0.15 mmol) with TFA by Procedure C(step a, 5 h; step b, 6 h at 0 �C) gave crude 49. Purification on HPLC(5% CH3CN/H2O at 2.5 mL/min; tR = 10–14 min) afforded 49(16 mg, 40%) as a white solid: 1H NMR (D2O) d 2.01–2.17 (m, 2,H8,8’), 2.83–2.89 (m, 2, H7,7’), 2.97 (dd, J = 6.3, 14.4 Hz, 1,H5),3.06 (dd, J = 3.8, 14.4 Hz, 1, H5’), 3.73–3.77 (m, 1, H9), 4.08–4.22(m, 1, H4), 4.40 (dd, J = 3.2, 6.0 Hz, 1, H3), 4.89–4.96 (m, 1, H2),7.25 (s, 1, H1); 13C NMR d 27.6 (C7), 29.9 (C5), 30.5 (C8), 53.7(C9), 70.4 (C3), 78.3 (C4), 80.7 (C2), 141.8 (C1), 173.9 (C10); MS(APCI) m/z 265 (100, MH+); HRMS (TOF MS-ESI) m/z calcd forC9H16N2O5SNa [M+Na]+ 287.0672; found 287.0664.
4.41. LuxS Inhibition Assay
SRH was prepared by incubating SAH (typically 10 mM) withnucleosidase Pfs (2 lM) overnight at 4 �C and the completion ofthe reaction was monitored spectrophotometrically by the absorp-tion difference between SAH and adenine (De276 = �1.4 mM�1
cm�1). A typical LuxS reaction (total volume = 1.0 mL) contained50 mM HEPES (pH 7.0), 150 mM NaCl, 17.8 lM SRH, and 150 lM5,5’-dithiobis(2-nitrobenzoic acid) (DTNB). The reaction wasinitiated by the addition of LuxS (final concentration 0.8 lM) andmonitored continuously at 412 nm (e = 14 000 M�1 cm�1) in aPerkin-Elmer k20 UV–vis spectrophotometer at room temperature.
For compounds that showed time dependent inhibition, the inhib-itor and LuxS (1.6 lM) were preincubated for 30 min at 4 �C andthe reaction was then initiated by addition of SRH.
Acknowledgment
We thank NIH (SC1CA138176, AI62901, and DE019667) andFIU’s Doctoral Evidence Acquisition Fellowship (V.L.A.M) for theirsupport.
References and notes
1. Waters, C. M.; Bassler, B. L. Annu. Rev. Cell Dev. Biol. 2005, 21, 319.2. Bassler, B. L.; Losick, R. Cell 2006, 125, 237.3. Winans, S. C. Nature 2005, 437, 330.4. Camilli, A.; Bassler, B. L. Science 2006, 311, 1113.5. Lebeer, S.; De Keersmaecker, S. C. J.; Verhoeven, T. L. A.; Fadda, A. A.; Marchal,
K.; Vanderleyden, J. J. Bacteriol. 2007, 189, 860.6. Ni, N.; Li, M.; Wang, J.; Wang, B. Med. Res. Rev. 2009, 29, 65.7. Mattmann, M. E.; Blackwell, H. E. J. Org. Chem. 2010, 75, 6737.8. Galloway, W. R. J. D.; Hodgkinson, J. T.; Bowden, S. D.; Welch, M.; Spring, D. R.
Chem. Rev. 2011, 111, 28.9. Lee, J. E.; Cornell, K. A.; Riscoe, M. K.; Howell, P. L. Structure 2001, 9, 941.
10. Lee, J. E.; Cornell, K. A.; Riscoe, M. K.; Howell, P. L. J. Biol. Chem. 2003, 278,8761.
11. Pei, D.; Zhu, J. Curr. Opin. Chem. Biol. 2004, 8, 492.12. Chen, X.; Schauder, S.; Potier, N.; Van Dorsselaer, A.; Pelczer, I.; Bassler, B. L.;
Hughson, F. M. Nature 2002, 415, 545.13. Meijler, M. M.; Hom, L. G.; Kaufmann, G. F.; McKenzie, K. M.; Sun, C.; Moss, J. A.;
Matsushita, M.; Janda, K. D. Angew. Chem., Int. Ed. 2004, 43, 2106.14. Semmelhack, M. F.; Campagna, S. R.; Federle, M. J.; Bassler, B. L. Org. Lett. 2005,
7, 569.15. Semmelhack, M. F.; Campagna, S. R.; Hwa, C.; Federle, M. J.; Bassler, B. L. Org.
Lett. 2004, 6, 2635.16. Zhu, J.; Dizin, E.; Hu, X.; Wavreille, A.-S.; Park, J.; Pei, D. Biochemistry 2003, 42,
4717.17. Zhu, J.; Patel, R.; Pei, D. Biochemistry 2004, 43, 10166.18. Rajan, R.; Zhu, J.; Hu, X.; Pei, D.; Bell, C. E. Biochemistry 2005, 44, 3745.19. Zhu, J.; Hu, X.; Dizin, E.; Pei, D. J. Am. Chem. Soc. 2003, 125, 13379.20. Turner, M. A.; Yang, X.; Yin, D.; Kuczera, K.; Borchardt, R. T.; Howell, P. L. Cell
Biochem. Biophys. 2000, 33, 101.21. Yuan, C.-S.; Liu, S.; Wnuk, S. F.; Robins, M. J.; Borchardt, R. T. Biochemistry 1994,
33, 3758.22. Alfaro, J. F.; Zhang, T.; Wynn, D. P.; Karschner, E. L.; Zhou, Z. S. Org. Lett. 2004, 6,
3043.23. Shen, G.; Rajan, R.; Zhu, J.; Bell, C. E.; Pei, D. J. Med. Chem. 2006, 49,
3003.24. Zhang, T.; Lee, B. W. K.; Cannon, L. M.; Ritter, K. A.; Dai, S.; Ren, D.; Wood, T. K.;
Zhou, Z. S. Bioorg. Med. Chem. Lett. 2009, 19, 6200.25. Wnuk, S. F.; Lalama, J.; Garmendia, C. A.; Robert, J.; Zhu, J.; Pei, D. Bioorg. Med.
Chem. 2008, 16, 5090.26. Wnuk, S. F.; Robert, J.; Sobczak, A. J.; Meyers, B. P.; Malladi, V. L. A.; Zhu, J.;
Gopishetty, B.; Pei, D. Bioorg. Med. Chem. 2009, 17, 6699.27. Gopishetty, B.; Zhu, J.; Rajan, R.; Sobczak, A. J.; Wnuk, S. F.; Bell, C. E.; Pei, D. J.
Am. Chem. Soc. 2009, 131, 1243.28. Jencks, W. P. Chem. Rev. 1972, 72, 705.29. Pearson, M. S. M.; Mathé-Allainmat, M.; Fargeas, V.; Lebreton, J. Eur. J. Org.
Chem. 2005, 2159.30. Kajimoto, T.; Liu, K. K. C.; Pederson, R. L.; Zhong, Z.; Ichikawa, Y.; Porco, J. A.;
Wong, C. H. J. Am. Chem. Soc. 1991, 113, 6187.31. Wong, C.-H.; Provencher, L.; Porco, J. A.; Jung, S.-H.; Wang, Y.-F.; Chen, L.;
Wang, R.; Steensma, D. H. J. Org. Chem. 1995, 60, 1492.32. Schramm, V. L. Acc. Chem. Res. 2003, 36, 588.33. Schramm, V. L. Arch. Biochem. Biophys. 2005, 433, 13.34. Yokoyama, M.; Momotake, A. Synthesis 1999, 1541.35. Lee, J. E.; Singh, V.; Evans, G. B.; Tyler, P. C.; Furneaux, R. H.; Cornell, K. A.;
Riscoe, M. K.; Schramm, V. L.; Howell, P. L. J. Biol. Chem. 2005, 280, 18274.36. Singh, V.; Evans, G. B.; Lenz, D. H.; Mason, J. M.; Clinch, K.; Mee, S.; Painter, G.
F.; Tyler, P. C.; Furneaux, R. H.; Lee, J. E.; Howell, P. L.; Schramm, V. L. J. Biol.Chem. 2005, 280, 18265.
37. Fleet, G. W. J.; Son, J. C. Tetrahedron 1988, 44, 2637.38. Calvez, O.; Chiaroni, A.; Langlois, N. Tetrahedron Lett. 1998, 39, 9447.39. Lee, J.; Hoang, T.; Lewis, S.; Weissman, S. A.; Askin, D.; Volante, R. P.; Reider, P. J.
Tetrahedron Lett. 2001, 42, 6223.40. Haidle, A. M.; Myers, A. G. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12048.41. Qiu, X. L.; Qing, F. L. J. Org. Chem. 2005, 70, 3826.42. Kim, Y. J.; Kitahara, T. Tetrahedron Lett. 1997, 38, 3423.43. Witte, J. F.; McClard, R. W. Tetrahedron Lett. 1991, 32, 3927.44. Zanardi, F.; Battistini, L.; Nespi, M.; Rassu, G.; Spanu, P.; Cornia, M.; Casiraghi,
G. Tetrahedron: Asymmetry 1996, 1167, 7.45. Otsuka, M.; Masuda, T.; Haupt, A.; Ohno, M.; Shiraki, T.; Sugiura, Y.; Maeda, K. J.
Am. Chem. Soc. 1990, 112, 838.
V. L. A. Malladi et al. / Bioorg. Med. Chem. 19 (2011) 5507–5519 5519
46. Malladi, V. L. A.; Sobczak, A. J.; Maricic, N.; Murugapiran, S. K.; Schneper, L.;Makemson, J.; Mathee, K.; Wnuk, S. F. Bioorg. Med. Chem. 2011, 19. doi:10.1016/j.bmc.2011.07.044.
47. Moriarty, R. M.; Mitan, C. I.; Branza-Nichita, N.; Phares, K. R.; Parrish, D. Org.Lett. 2006, 8, 3465.
48. Evans, G. B.; Furneaux, R. H.; Gainsford, G. J.; Schramm, V. L.; Tyler, P. C.Tetrahedron 2000, 56, 3053.
49. Evans, G. B.; Furneaux, R. H.; Hausler, H.; Larsen, J. S.; Tyler, P. C. J. Org. Chem.2004, 69, 2217.
50. Horenstein, B. A.; Zabinski, R. F.; Schramm, V. L. Tetrahedron Lett. 1993, 34,7213.
51. Murruzzu, C.; Riera, A. Tetrahedron: Asymmetry 2007, 18, 149.52. Moreaux, V.; Warren, H.; Williams, J. M. Tetrahedron Lett. 1997, 38, 4655.