Article

Andrey G. Tere4

0022-2836/© 2018 Elsevi

Binding and Action of Amino Acid Analogs ofChloramphenicol upon the BacterialRibosome

shchenkov1, Malgorzata D

obosz-Bartoszek2, Ilya A. Osterman1, 3,James Marks , 8, Vasilina A. Sergeeva1, 9, Pavel Kasatsky5,Ekaterina S. Komarova (Andreyanova) 3, 7, Andrey N. Stavrianidi 1, Igor A. Rodin1,Andrey L. Konevega5, 6, Petr V. Sergiev1, 3, Natalia V. Sumbatyan1,Alexander S. Mankin4, 8, Alexey A. Bogdanov1 and Yury S. Polikanov2, 8,

1 - Department of Chemistry and A.N. Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University,Moscow 119992, Russia2 - Department of Biological Sciences, University of Illinois at Chicago, Chicago, IL 60607, USA3 - Skolkovo Institute of Science and Technology, Skolkovo, Moscow region 143025, Russia4 - Center for Biomolecular Sciences, University of Illinois, Chicago, IL 60607, USA5 - Petersburg Nuclear Physics Institute, NRC “Kurchatov Institute”, Gatchina 188300, Russia6 - Peter the Great St. Petersburg Polytechnic University, Saint Petersburg 195251, Russia7 - Department of Bioengineering and Bioinformatics, Lomonosov Moscow State University, Moscow 119992, Russia8 - Department of Medicinal Chemistry and Pharmacognosy, University of Illinois at Chicago, Chicago, IL 60607, USA

Correspondence to Alexey A. Bogdanov and Yury S. Polikanov: Department of Biological Sciences, University ofIllinois at Chicago, Chicago, IL 60607, USA. bogdanov@belozersky.msu.ru; yuryp@uic.eduhttps://doi.org/10.1016/j.jmb.2018.01.016Edited by Sarah A. Woodson

Abstract

Antibiotic chloramphenicol (CHL) binds with a moderate affinity at the peptidyl transferase center of thebacterial ribosome and inhibits peptide bond formation. As an approach for modifying and potentiallyimproving properties of this inhibitor, we explored ribosome binding and inhibitory activity of a number ofamino acid analogs of CHL. The L-histidyl analog binds to the ribosome with the affinity exceeding that of CHLby 10 fold. Several of the newly synthesized analogs were able to inhibit protein synthesis and exhibited themode of action that was distinct from the action of CHL. However, the inhibitory properties of thesemi-synthetic CHL analogs did not correlate with their affinity and in general, the amino acid analogs of CHLwere less active inhibitors of translation in comparison with the original antibiotic. The X-ray crystal structuresof the Thermus thermophilus 70S ribosome in complex with three semi-synthetic analogs showed that CHLderivatives bind at the peptidyl transferase center, where the aminoacyl moiety of the tested compoundsestablished idiosyncratic interactions with rRNA. Although still fairly inefficient inhibitors of translation, thesynthesized compounds represent promising chemical scaffolds that target the peptidyl transferase center ofthe ribosome and potentially are suitable for further exploration.

© 2018 Elsevier Ltd. All rights reserved.

Introduction

Manyantibiotics stop growth of pathogenic bacteria,and thereby cure infections, by selectively bindingto the bacterial ribosomes and inhibiting proteinsynthesis. Antibiotics can interfere with translationby interacting with various functional centers of theribosome and either locking of a particular conforma-

er Ltd. All rights reserved.

tion of the ribosome or hindering the binding of itsligands. The peptidyl transferase center (PTC),located in the large ribosomal subunit, is targeted bya particularly broad array of inhibitors belongingto several distinct chemical classes, such as pheni-cols, lincosamides, oxazolidinones, pleuromutilins,streptogramins A, and others [1]. From the oldestchloramphenicol (CHL) to the newest Food and Drug

J Mol Biol (2018) 430, 842–852

843Amino Acid Analogs of Chloramphenicol

Administration-approved retapamulin, PTC-targetingdrugs are known as excellent antibacterial agents.Most of the PTC-targeting compounds inhibit

protein synthesis by competing with the positioningof the amino acid side chain of the incomingaminoacyl-tRNA (aa-tRNA) in the A site [2–4]. CHL,a typical example from this family, inhibits translationin a wide range of Gram-positive and Gram-negativebacteria [5]. It binds to theA site of thePTC in a creviceformed by the bases of the conserved nucleotidesU2504, A2451, and C2452 of the 23S rRNA and itsnitrobenzyl ring forms aπ-stacking interaction with thebase of C2452 [6,7]. The aromatic ring of theribosome-bound CHL overlaps with the placement ofthe side chains of the incoming aa-tRNAs, thusefficiently preventing the aminoacyl moiety ofaa-tRNA from properly accommodating into the PTCactive site. CHL was originally viewed as a universalinhibitor of peptide bond formation [8]. However, thismodel has been recently revised because the newerdata revealed CHL as a context-specific inhibitor oftranslation, whose activity depends on the nature ofspecific amino acids in the nascent chain and theidentity of the residue entering the A site [9].While CHL does not act upon the eukaryotic

cytoplasmic ribosome, it readily binds to the ribo-somes of the mammalian mitochondria [10–14]. Theinterference with mitochondrial translation, which isthe cause of the major side effects of CHL, signifi-cantly curbed the medical use of this drug in manycountries [5,15,16]. Amending the CHL structure withthe additional chemical entities that would formidiosyncratic interactions specifically with the bacterialribosome could be one approach for the developmentof more selective inhibitors. In addition, the rapidspread of antibiotic resistance has significantly limitedthe medical utility of many available antibiotics,including the PTC-targeting drugs. One of the recentlydiscovered but rapidly spreading resistance mecha-nisms operates via modification of the 23S rRNAnucleotide A2503, which is located in the binding siteof several inhibitors. The rRNA methyltransferase Cfradds a methyl group to the C8 position of the A2503and renders bacteria resistant to a wide range ofantibiotics targeting the catalytic center of the ribo-some [17,18]. The development of newer derivativesof the existing antibiotic platform, which could avoid aclash with the C8 methyl group of the Cfr-modifiedA2503, has proved to be a viable approach forovercoming such resistance [19].Previously, attempts were made to prepare pep-

tide derivatives of CHL containing peptides of up to10 amino acids long aiming to identify interactionsbetween the peptide chain and the elements of thenascent peptide exit tunnel (NPET) [20,21]. In thecurrent study, with the goal of paving the way fordevelopment newer derivatives of CHL that wouldtarget the PTC, we have prepared severalsemi-synthetic analogs carrying various individual

amino acids attached to the drug. Using competi-tion-binding experiments, we investigated the affin-ities of the newly prepared compounds for theirtarget and found that some of them show improvedbinding to the bacterial ribosome. We also showedthat the mode of action of the newer analogs differsfrom the site-specific action of CHL. By solving theX-ray crystal structures of the Thermus thermophilus(Tth) 70S ribosome in complex with several CHLanalogs, we observe specific interactions of theamino acid moiety with rRNA, thereby rationalizingthe improved binding. Although the higher affinity ofthe derivatives to the vacant Escherichia coliribosome does not directly translate into strongerinhibition of protein synthesis, the compoundsdescribed here could open new directions forimproving the medical utility of amphenicol class ofthe ribosome inhibitors.

Results and Discussion

Synthesis of CAM derivatives

Chemical synthesis of CHL analogs carrying aminoacid residues instead of dichloroacetic moiety isbased on acylation of CHL amine (CAM), an inactiveCHL derivative, with activated amino acids (Figs. 1and S1) [22]. The overall synthesis scheme includesthree steps: (i) acid hydrolysis of CHL to yield CAM[23]; (ii) acylation of CAM by succinimide esters ofamino acids with protected α-amino group andside-chain groups; and (iii) de-protection of theobtained CAM-derivatives to yield aminoacyl-CAM(AA-CAM) (Figs. 1 and S1). Using this approach, wehave prepared CHL analogs aminoacylated withdifferent amino acids, including the N-protected ones(Table 1). Molecular weights and chemical structuresof all synthesized CAM derivatives were confirmed bymass spectrometry and 1H- and 13C-NMR. Althoughfew individual amino acid analogs of CHL havebeen studied previously [22,24,25], this work repre-sents the first systematic approach to synthesis andfunctional assessment of more than three dozens ofvarious AA-CAMs.

AA-CAM compounds bind tightly to the bacterialribosome

We used the competition-binding assay, exploitingBODIPY-labeled erythromycin (BODIPY-ERY) toassess the affinity of the synthesized compounds forthe ribosome [26,27]. While CHL binds in the A site ofthe PTC [6,28], ERY binds in the upper part of theNPET [7] but their binding sites sufficiently overlap sothat even unmodified CHL and ERY compete for theirbinding to the ribosome [29] (Fig. S2). The overlap ofthe AA-CAMs with ERY is expected to be even

Fig. 1. Schematic diagram of chemical synthesis of the histidine analog of CHL.

844 Amino Acid Analogs of Chloramphenicol

more pronounced. The apparent dissociation con-stant KDapp of CHL obtained using competition withBODIPY-ERY (2.8 ± 0.5 μM) is consistent with the

Table 1. Apparent dissociation constants (KDapp) ofamino-acid CAM derivatives with the E. coli 70Sribosomes

N-unprotected Compounds N-Protected Compounds

Compound KDapp (μM) Compound KDapp (μM)

CHL 2.8 ± 0.5 Ac-Pro-CAM (7a) 10 ± 1.5His-CAM (1) 0.24 ± 0.06 Ac-Arg-CAM (3a) 26 ± 10Lys-CAM (2) 1.6 ± 0.5 Ac-His-CAM (1a) 34 ± 5Arg-CAM (3) 3.2 ± 1.4 Ac-Lys-CAM (2a) 43 ± 12Ala-CAM (4) 5.3 ± 0.9 Ac-Ala-CAM (4a) 55 ± 13Gly-CAM (5) 5.7 ± 0.9 Ac-Gly-CAM (5a) 69 ± 14Gln-CAM (6) 6.3 ± 5.1 Ac-Phe-CAM (11a) 92 ± 22Pro-CAM (7) 7.5 ± 1.6 Ac-Trp-CAM (10a) 106 ± 36Tyr-CAM (8) 8.2 ± 1.5 Ac-Tyr-CAM (8a) 140 ± 50Ser-CAM (9) 8.4 ± 2.4 Boc-His-CAM (1d) 19 ± 5Trp-CAM (10) 12 ± 2 Boc-Gly-CAM (5d) 45 ± 8Phe-CAM (11) 19 ± 8 Boc-Phe-CAM (11d) 160 ± 120Val-CAM (12) 34 ± 8 Boc-Ala-CAM (4d) 170 ± 60Leu-CAM (13) 36 ± 6 Boc-Pro-CAM (7d) 440 ± 230Ile-CAM (14) 36 ± 10 f-Gly-CAM (5b) 45 ± 6Asn-CAM (15) 40 ± 26D-His-CAM (1b) 2.9 ± 0.8D-Ala-CAM (4b) 7.9 ± 1.4D-Pro-CAM (7b) 13 ± 2β-Ala-CAM (16) 6.3 ± 1.3

Numbers in theparenthesiscorrespond to theparticular compound fromthe synthesis scheme shown in Fig. S1.

previously published data determined by direct[14C]-CHL binding (2.3 μM [30]). Using this approach,we found thatmanyof the synthesizedCAMderivativesexhibited considerable affinity for the ribosome (KDappin the lowmicromolar range) (Table 1). Interestingly, allof the AA-CAM derivatives carrying free α-aminogroups bind to the ribosome with higher affinities thanthe corresponding compounds in which the α-aminogroup was modified by acetylation or formylation orprotected by the Boc group (Fig. S3; Table 1). Thisresult suggests that the positive charge, the small sizeof the α-amino group, or both contribute to the efficientribosome binding of AA-CAMs. Importantly, oneAA-CAM variant, His-CAM, binds to the ribosomewith a more than 10-fold higher affinity than CHL(His-CAM, KDapp = 0.24 ± 0.06 μM) (Fig. 2a; Table 1).

AA-CAM compounds are less efficient inhibitorsof translation than CHL and do not show thesame context specificity

To check whether the binding strength of AA-CAMscorrelates with inhibition of translation, we tested theirability to interferewith in vitroprotein synthesis. Additionof 30 μM CHL to the cell-free transcription–translationsystem based on the E. coli cell extract resulted in anear-complete inhibition of synthesis of firefly luciferasereporter (Fig. 2b). Consistent with previously published

Fig. 2. Binding and inhibitory properties of AA-CAM-derivatives. (a) Competition-binding assay to test the inhibition ofBODIPY-ERY binding to the E. coli ribosomes in the presence of increasing concentrations of AA-CAM derivativesmeasured by fluorescence anisotropy. (b, c) Inhibition of in vitro synthesis of firefly luciferase by AA-CAM derivatives in theE. coli S30 cell extract (b) or in the PURE system (c). All the inhibitors were present in the reaction at 30 μM. All thereactions were performed in triplicates, and error bars represent confidence interval (α = 0.05). Inhibitory activity of AA-CAMcompounds with free α-amino groups are shown as light gray bars; N-protected AA-CAM compounds, dark gray; and positivecontrol CHL, white bars. (d) Primer extension inhibition (toe-printing) analysis of site specificity of action of CHL and AA-CAMs. Thesyntheticmini-genewas translated in the cell-free translation (“PURE”) systemand sites of antibiotic-induced translation arrestwereanalyzed by primer extension. The reactions loaded onto lanes 1–6 containedmupirocin, an inhibitor of isoleucyl-tRNA synthetase.The sample in lane 2 (labeled “NONE”) contained no other antibiotics besides mupirocin. The control antibiotic retapamulin (RET)inhibits translation initiation and arrests the ribosome at the start codon (black arrowheads). Bands corresponding to theCHL-induced translation arrest at the fifth codon are indicated by the green arrowheads. Stalling of ribosomes at the seventh codonof the ORF due to the presence of mupirocin that causes depletion of isoleucyl-tRNA (lanes 1–6) is indicated by the bluearrowheads. U- and A-specific sequencing lanes are indicated. The nucleotide sequence of the gene and the correspondingencodedaminoacidsare indicatedon the left.Note that the reverse transcriptasestops15–16nucleotidesdownstream from the firstnucleotide of the P-site codon as indicated by the dashed arrow.

845Amino Acid Analogs of Chloramphenicol

results, removal of the dichloroacetyl moiety from CHLresults in nearly complete loss of its biological activity(Fig. 2b, CHL versusCAM) [23], likely due to the loss ofinteractions of the dichloroacetic moiety with theribosome [6]. However, similar to the previous studiesemploying peptide derivatives of CHL [5,20,31], theinhibitory activity of CAM can be partially restored byincorporation of specific amino acids (Fig. 2b), indicat-ing that the contacts of the aminoacyl side chain ofAA-CAMs with the ribosome may compensate for theinteractions lost upon the removal of the dichloroaceticg r o u p . H owe v e r , mo s t o f t h e t e s t e dAA-CAM-derivatives were less potent inhibitors oftranslation than CHL (Fig. 2b). Even His-CAM, whichexhibited higher than CHL affinity for the ribosome,reduced the yield of the functional luciferase by only60%.Only the activity of β-Ala-CAMapproached that ofCHL, although this derivative demonstrated a twofoldweaker binding to the ribosome compared to theparental compound. The modification of the α-aminogroup of AA-CAMs (with formyl, acetyl, or Boc groups)leads to a nearly complete loss of their inhibitorypotency (Fig. 2b). The inability of most of the AA-CAMcompounds to efficiently inhibit translation could be

potentially explained by the susceptibility of AA-CAMsto the action of peptidases present in the bacterialextract-based cell-free translation system. Therefore,we re-tested the inhibitory activity of some of thecompounds in the PURE cell-free translation systemreconstituted from purified components and thuslacking peptidases [32]. All of the tested compoundsexhibited inhibitory activity in the PURE system(Fig. 2c), suggesting that the stability of our compoundsis unlikely to be the key issue. Although the overallinhibition of translation by AA-CAMderivatives in thePURE system has improved compared to theireffect upon translation in the cell extract, it still didnot reach the level of inhibition elicited by the CHL,suggesting that the equilibrium binding affinities ofthe tested inhibitors to a vacant ribosome might notadequately reflect the interactions of these com-pounds with the translating ribosome carryingmRNAs, tRNAs, and nascent peptide chain. Severalpeptide derivatives of CAM have been prepared andtested previously in the binding and inhibitionassays [20,31]. Similar to our findings, thosederivatives were found to be poor inhibitors oftranslation.

Fig. 3. Chemical structures and electron density mapsof CAM-derivatives. Chemical structures and differenceFourier maps of His-CAM (A), D-His-CAM (B), and Lys-CAM(C) in complexwith theTth 70S ribosome. The refinedmodelof each compound is displayed in its respective electrondensity before the refinement (green mesh). Carbon atomsare colored yellow for His-CAM, orange for D-His-CAM,magenta for Lys-CAM, blue for nitrogens, and red foroxygens.

846 Amino Acid Analogs of Chloramphenicol

The inhibition of protein synthesis by CHL de-pends on the amino acid sequence of the nascentpeptide and the identity of the aa-tRNA, with themost efficient arrests occurring after alanine (and toa lesser extent after serine or threonine) that appearat the penultimate position of the nascent chain [9].To test whether the specificity of action of AA-CAMcompounds matches that of CHL, we used primerextension inhibition assay (toe-printing) to checkwhether the derivatives that exhibited inhibitoryactivity in the PURE system (Fig. 2d) would arrestthe ribosome at the same codon(s) where translationis arrested by CHL. In these experiments, we used ashort synthetic open reading frame (ORF) encodingthe peptide MFKAFKNIIRTRTL, where CHL effi-ciently arrests the ribosome at the fifth codon afterthe MFKAF peptide has been synthesized (Fig. 2d,lanes 3 and 9, green arrowheads). Some translationreactions were additionally supplemented withmupirocin, an inhibitor of isoleucyl-tRNA synthetase.The depletion of isoleucyl-tRNA efficiently traps theribosomes at the seventh codon of the ORF if theywere not arrested at any of the prior codons by theaction of the ribosome inhibitor (Fig. 2d, lanes 1–6,blue arrowheads) helping to evaluate the efficiencyof the antibiotic-dependent arrest [33]. Only a smallfraction of ribosomes pre-incubated with CHL wasable to reach the “trap” codon (Fig. 2d, lane 3)because most of the translation complexes werearrested at the fifth codon. In contrast, the intensity ofthe trap-codon band was only slightly diminished insamples supplemented with AA-CAM analogs andno strong arrest was observed at the fifth codon(Fig. 2d, lanes 4–6). This was in spite of the fact thatthe concentrations of AA-CAM compounds in thetoe-printing experiments (100 μM) were several-foldhigher than those used in the cell-free translationexperiments (Fig. 2b, c), where the same derivativeswere able to inhibit expression of the luciferasereporter with a considerable activity. We noted,however, that the toe-print bands corresponding tothe initiating ribosome were slightly enhancedcompared to the no-drug (+/− mupirocin) control(Fig. 2d, lanes 2 and 8 versus lanes 4–6 and 10–12,respectively, black arrowheads), indicating thatAA-CAM analogs weakly interfered with translationinitiation, likely by inhibiting the formation of the firstpeptide bond. We concluded that AA-CAM deriva-tives do not show the same site specificity of actionas the parent compound CHL, apparently due toidiosyncratic interactions of the aminoacyl moieties ofAA-CAMs with the ribosome.

Aminoacyl moieties of AA-CAMsmediate specificinteractions with the ribosome

To unambiguously determine the mode of bindingof AA-CAM derivatives to the ribosome, we crystal-lized Tth 70S ribosomes in the presence of mRNA,

deacylated A-, P-, and E-site tRNAs, and His-CAMand solved the structure of the complex at 2.8-Åresolution. However, the other complexes preparedin this way did not diffract sufficiently well. Therefore,we amended our crystallization approach by usingribosome complexed with the protein Y (PY) as a toolto obtain structures of higher resolution [34,35]. Thisapproach is based on our recent finding that bindingof PY to a vacant 70S ribosome stabilizes it bylocking the head of the 30S subunit in an unrotatedstate, which leads to a better diffraction [34,35].Since PY binds in the mRNA channel on the 30Ssubunit far away from the CHL binding site in thePTC of the 50S subunit, it precludes the presence ofmRNA or A- and P-tRNAs [34]. However, this doesnot interfere with the binding of AA-CAMs. Using thisapproach, we were able to solve two additionalstructures of D-His-CAM and Lys-CAM in complexwith the PY-bound 70S ribosome at 2.7- and 2.6-Åresolutions, respectively (Table S1). An unbiased

847Amino Acid Analogs of Chloramphenicol

difference Fourier map, calculated using the ampli-tudes from the crystals and phases derived from amodel of the ribosome without the bound compound,revealed positive electron density resembling char-acteristic features of each of the compounds (Fig. 3).A single binding site for each of the AA-CAMs isobserved in the ribosome within the PTC of the largeribosomal subunit (Fig. 4).The binding position of the amphenicol parts of

AA-CAMs is identical to those observed previouslyfor parent CHL in the Tth [7] or Eco [28] ribosomes in

Fig. 4. Structure of His-CAM in complex with the 70S ribosbinding site (yellow) in the Tth 70S ribosome viewed from thecross-cut section through the ribosome (b). The 30S subunit ismRNA is in magenta, and the A- and P-site tRNAs are in greenclarity. (c, d) Close-up views of the His-CAM bound in the PTPotential H-bond interactions are indicated with dashed linesedge-to-face π-stacking with the nucleobase of U2506 of the

the absence of mRNA and tRNAs (Fig. S4). Theamino acid moieties of AA-CAMs are orientedt owa r d t h e NPET and exh i b i t u n i q uecompound-specific interactions at the PTC (Figs. 4and 5). Histidine side chain of His-CAM forms tiltededge-to-face π-stacking with the nucleobase ofU2506 (Figs. 4d and 5a). In addition, the α-aminogroup of His-CAM forms hydrogen bond (H-bond) viaa water molecule with the phosphate of nucleotideG2505 (Fig. 5a). These additional interactions of theaminoacyl moiety likely account for its increased

ome and A- and P-tRNAs. (a, b) Overview of the His-CAMPTC down the tunnel as indicated by the inset (a), or as ashown in light yellow, the 50S subunit is in light blue, theand dark blue, respectively. The E-site tRNA is omitted forC. The E. coli nucleotide numbering is used throughout.. Note that side chain of His-CAM compound forms tilted23S rRNA (shown as spheres).

Fig. 5. Side-chain-specific interactions of AA-CAMswith the ribosome. Compound-specific H-bond interactions of His-CAM(a), D-His-CAM (b), or Lys-CAM (c)with the nucleotides of the 23S rRNAare indicatedwith dashed lines. Stacking interactions ofHis-CAM are shown with the black arrow.

848 Amino Acid Analogs of Chloramphenicol

affinity to the ribosome compared to CHL. Contraryto His-CAM and consistent with its significantly loweraffinity, α-amino group of the D-His-CAM stereoiso-mer does not form the water-mediated H-bond withG2505 (Fig. 5b). Surprisingly, the orientation of theamino acid side chain of L-Lys-CAM is similar to thatof D-His-CAM (Fig. 5c). The lysine moiety of Lys-CAMextends toward the wall of the NPET and, similar toD-His-CAM, forms a single H-bond with the nucleo-base of A2059 explaining why their binding is less tightin comparison with His-CAM.In the structure of the ribosome/His-CAM complex,

the oxygens of the nitro group in the CAMmoiety formH-bonds with the A76 ribose hydroxyls of deacylatedA- and P-site tRNAs (Fig. 4c). It is unclear whethersuch interactions are possible in the translatingribosome when the P-site tRNA is attached to theaminoacyl or peptidyl groups. Nevertheless, ourstructures clearly demonstrate that amphenicol part

Fig. 6. Structural basis for resistance to CHL (a) and His-CA(red sphere) catalyzed by the Cfr-methyltransferase reveals a s(b). Note that the path of the His-CAM in the PTC is located furtthat should allow it to avoid a possible steric clash with this po

of the AA-CAMs is capable of anchoring the deriva-tives in the PTC active site directing the attachedamino acids in the direction corresponding to thegrowing peptide chain.Resistance to many peptidyl transferase inhibitors

(including CHL) can be conferred by Cfr, a methyl-transferase that methylates A2503 of the 23S rRNAat the C8 atom [18]. The methyl group added toA2503 by Cfr would invade the CHL binding pocket ifthe placement of the modified A2503 remainsunchanged (Fig. 6a). In silico modeling shows,however, that the ribosome-bound His-CAM whoseplacement in the PTC is somewhat shifted relative toCHL, would avoid the collision with C8-methyl groupof A2503 (Fig. 6b), suggesting that this compoundcould retain certain activity against the Cfr-modifiedribosome. Likewise, the resistance to CHL conferredby the mutation of A2503 to G [36], is likely to be lesspronounced in the case of His-CAM due to a more

M (b). Molecular modeling of the C8-methylation of A2503mall clash with CHL (a) but not with histidine analog of CHLher away from the C8 position of A2503 (compared to CHL)sition carrying methyl group in the Cfr-modified ribosome.

849Amino Acid Analogs of Chloramphenicol

remote position of the amino acid side chain ofHis-CAM relative to the site of the mutation.Therefore, the structure of the ribosome with theAA-CAM derivatives suggests that some of theamino acid analogs of CHL could be less susceptibleto the action of specific resistance mechanisms.Further experiments would be required to checkwhether any of the synthesized CHL analogs areactive against ribosomes rendered resistant to CHL.The structure of PTC is highly conserved among

evolutionary distant ribosomes, including the ribo-somes of mitochondria, which are known to betargeted by CHL [10–14]. However, even minorvariations in the placement of the nucleotidesdirectly interacting with the drug, which could beinfluenced by the less conserved second-shellnucleotides, could account for the selectivity of theantibiotics that bind in the PTC active site as well asfor the differential effect of these drugs uponribosomes from different species. Because thecompounds examined in our study have fairly lowactivity against the bacterial ribosome, we have notpursued studies of their effects upon mitochondrialtranslation. However, we believe that adding anamino acid or its analog to the drug core couldpotentially influence the spectrum of action of thisclass of the ribosomal antibiotics opening a possi-bility of making them more selective inhibitors ofbacterial translation.

Conclusions

The goal of the current study was to develop aminoacid analogs of CHL as an approach for exploringthis line of derivatives as new inhibitors of transla-tion. We used organic synthesis to generate adiverse set of amino acid analogs of native CHLand examined their ribosome binding and inhibitoryproperties. We have demonstrated that His-CAM, ahistidine analog of CHL, exhibits 10 times higheraffinity for the bacterial ribosome as compared toCHL. We have noted, however, that the inhibitoryproperties of the semi-synthetic CHL analogs do notcorrelate with their binding affinities to the vacantribosomes and the compounds with high affinities,such as His-CAM, inhibit translation less efficientlythan CHL, while compounds with low affinities, suchas β-Ala-CAM, demonstrated inhibitory propertiessimilar to those of the parental drug. Our crystalstructures show that amino acid analogs of CHLestablish compound-specific interactions with thenucleotides of the 23S rRNA at the PTC and orientthe amino acid moiety in the direction of the upperpart of the peptide exit tunnel. The idiosyncraticinteractions with the ribosome of the amino acidresidue attached to the CHL core open the possibilityof increasing the selectivity of this class of antibioticsand possibly diminish their side effects mediated by

the action upon mitochondrial translation. Moreover,the possibility that some of the amino acid deriva-tives of CHL might act upon drug-resistant ribosomemakes them attractive compounds for further explo-ration by medicinal chemists, and we expect that ourfindings, including the structure of the AA-CAM–ribosome complexes described here, will serve as astarting point for such studies.

Materials and Methods

Reagents

Various amino acid derivatives used in chemicalsynthesis (Supplementary Methods) were fromFluka or Reanal; CHL was from Sigma; succinimideester of BODIPY was from Invitrogen; and N-hydro-xysuccinimide, N,N′-dicyclohexylcarbodiimide, andDMAP were from Merck. BODIPY-ERY was synthe-sized as described previously [37].

Chemical synthesis of AA-CAM derivatives

The general scheme for the synthesis of AA-CAMderivatives is shown in Fig. S1 and the details ofchemical synthesis are provided in the Supplemen-tary Methods section. CAM [(1R,2R)-2-amino-1-(4--nitrophenyl)propane-1,3-diol)] was prepared asdescribed previously [23]. Amino acids with pro-tected α- and side-chain amino groups wereactivated by reaction with N-hydroxysuccinimide inthe presence of N,N′-dicyclohexylcarbodiimide at0 °C. The resulting succinimide-reactive esters wereused for the acylation of CAM in the presence ofdiisopropylethylamine as a base at room tempera-ture. Subsequent deprotection was achieved bytreatment of the obtained amino-acid CAM deriva-tives with trifluoroacetic acid and appropriate scaven-gers. Synthesized AA-CAM derivatives were purifiedby column chromatography on silica gel using suitablesystems of solvents. For generating N-acetylatedvariants of AA-CAM, additional acetylation wasperformed by reacting the unprotected AA-CAMderivatives with the N-acetylsuccinimide. Purity andchemical structures of obtained compounds wereconfirmed by HPLC, LC–MS, and NMR spectroscopy(see Supplementary Methods).

In vitro binding assay

Binding affinities of CAM-derivatives to E. coliribosomes was analyzed by competition-bindingassay using fluorescently labeled BODIPY-ERY asdescribed before [26,27,37]. BODIPY-ERY (4 nM)was incubated with ribosomes (25 nM) for 30 min at25 °C in the buffer containing 20 mM Hepes–KOH(pH 7.5), 50 mM NH4Cl, 10 mMMg(CH3COO)2, and

850 Amino Acid Analogs of Chloramphenicol

0.05% Tween-20. The solution of CAM derivatives in arange of concentrations from 0.1 μM to 1 mM wasadded to the formed complex. The mixture wasincubated for 2 h until equilibrium was reached andthe values of fluorescencepolarizationweremeasured.

In vitro translation

Inhibition of firefly luciferase synthesis in cell-freetranslation systems by AA-CAMs was tested asdescribed previously [38]. Briefly, the in vitro tran-scribed firefly luciferase mRNA was translated usingeither PURExpress system (New England Biolabs) orE. coli S30 extract prepared according to Svetlov et al.[39]. Reactions were programmed with 100 ng mRNAand were carried out in 5 μL aliquots at 37 °C for30 min. The AA-CAMs were added to 30 μM finalconcentration. The activity of in vitro synthesizedluciferase was assessed using 5 μL of the substratefrom the Steady-Glo Luciferase Assay System(Promega).

Toe-printing analysis of compound-inducedribosome stalling

Toe-printing experiments were performed in thePURExpress cell-free translation system (NewEngland Biolabs) following our published protocols[40,41]. The synthetic template encoding the aminoacid sequence MFKAFKNIIRTRTL was initiallygenerated by PCR reaction using a combination offive primers: 2 μM T7 (ATTAATACGACTCACTATAGGG), 2 μM NV1 (GGTTATAATGAATTTTGCTTATTAAC), and 0.2 μM of each of the following: Fv(AATACGACTCACTATAGGGCAACCTAAAACTTACACACGCCCCGGTAAGGAAATAAAAAT),inner (GCCCCGGTAAGGAAATAAAAATGTTCAAAGCATTCAAAAACATCATACGTACTCGTACTC), and Rev. (GGTTATAATGAATTTTGCTTATTAACCTTGCCTGCGCTTAAAGAGTACGAGTACGTATGATGT) as described in Ref. [42]. Theproduct was cloned into the pUC18 plasmid cut withthe SmaI restriction enzyme and the resultingpUCMFKAFK plasmid was verified by capillarysequencing. For the toe-printing reaction, the tem-plate was PCR-amplified from the pUCMFKAFKplasmid using T7andNV1primers.Whenneeded, theother antibiotics (retapamulin, CHL, and AA-CAManalogs) were present in the reaction at the finalconcentration of 100 μM. In addition, all thetoe-printing reactions were supplemented with theIle-RS inhibitor mupirocin (final concentration 50 μM),which caused translating ribosomes to pausewhenanisoleucine codon was encountered.

Crystallographic structure determination

Ribosome complexes with mRNA and tRNAs orwith PY were formed as described previously [35].

CAM derivatives were added to the pre-formedribosome complexes to a final concentration of250 μM prior to crystallization. All Tth 70S ribosomecomplexes were formed in the buffer containing5 mM Hepes–KOH (pH 7.6), 50 mM KCl, 10 mMNH4Cl, and 10 mM Mg(CH3COO)2, and then crys-tallized in the buffer containing 100 mM Tris–HCl(pH 7.6), 2.9% (w/v) polyethylene glycol 20K, 7%–12% (v/v) methyl-2,4-pentanediol, 100–200 mMarginine, 0.5 mM β-mercaptoethanol. Crystals weregrown by the vapor diffusion method in sitting dropsat 19 °C and stabilized as described previously [35],with the corresponding CAM-derivatives beingadded to the stabilization buffers (100 μM each).Diffraction data were collected using beamlines24ID-C and 24ID-E at the Advanced Photon Source(Argonne, IL). All crystals belonged to the primitiveorthorhombic space group P212121 with approxi-mate unit cell dimensions of 210 Å × 450 Å × 620 Åand contained two copies of the 70S ribosome perasymmetric unit. Each structure was solved bymolecular replacement using PHASER from theCCP4 program suite [43]. The search model wasgenerated from the previously published structuresof Tth 70S ribosome with bound mRNA and tRNAs(PDB entry 4Y4P from Ref. [35]) or with PY (PDBentry 4Y4O from [35]). The initial molecular replace-ment solutions were refined by rigid body refinementwith the ribosome split into multiple domains,followed by positional and individual B-factor refine-ment. The final models of the 70S ribosome incomplex with His-CAM and mRNA/tRNAs, or incomplex with D-His-CAM/Lys-CAM and PY weregenerated by multiple rounds of model building inCOOT [44], followed by refinement in PHENIX [45].The statistics of data collection and refinement arecompiled in Table S1.

Accession Numbers

Coordinates and structure factors were depositedin the RCSB Protein Data Bank with accession code6CFJ for the T. thermophilus 70S ribosome incomplex with His‐CAM, mRNA, A‐, P‐ and E‐sitetRNAs; 6CFK for the T. thermophilus 70S ribosomein complex with D‐His‐CAM and protein Y; and 6CFLfor the T. thermophilus 70S ribosome in complexwith Lys‐CAM and protein Y.

Acknowledgments

We thank V. N. Tashlitsky for help with LCMSanalysis, Y. K. Grishin and I. A. Godovikov for helpwith interpreting NMR spectra, E. N. Shapovalova forhelp with chiral chromatography analysis, M. S.Svetlov for help with translation inhibition assays, G. A.

851Amino Acid Analogs of Chloramphenicol

Korshunova for providing Boc derivatives of aminoacids, and А. А. Bogdanov Jr. for providing Acderivatives of amino acids. We are also thankful tothe members of the A.A.B, P.V.S., A.S.M., and Y.S.P.laboratories for discussions and critical feedback.This work is based upon research conducted at the

Northeastern Collaborative Access Team beamlines,which are funded by the National Institute of GeneralMedical Sciences from the National Institutes of Health(P41 GM103403). The Pilatus 6M detector on 24ID-Cbeamline is funded by a NIH-ORIP HEI grant (S10RR029205). The Eiger 16M detector on 24ID-Ebeamlineis funded by a NIH-ORIP HEI grant (S10OD021527). This research used resources of theAdvanced Photon Source, a U.S. Department ofEnergy (DOE) Office of Science User Facility operatedfor the DOE Office of Science by Argonne NationalLaboratory under Contract No. DE-AC02-06CH11357.This work was supported by Illinois State startup

funds (to Y.S.P.), a Russian Science Foundationgrant 14-24-00061-P (to Tatyana S. Oretskaya,“Study of antibiotics binding to bacterial ribosomes”),and Russian Foundation for Basic Research grants16-04-00709 (to N.V.S., “Synthesis of CHL ana-logues”) and 15-34-20139 (to I.A.O., “Translationinhibition analysis”), and the National Institutes ofHealth grants R01 AI125518 (to A.S.M.).Author Contributions: A.G.T. and V.A.S. per-

formed chemical synthesis of CAM-derivatives;N.V.S. purified CAM derivatives; A.G.T. performed70S ribosome binding assay; A.A.S. and I.A.R.performed mass spectrometry analysis; I.A.O.,E.S.K., and P.V.S. designed and performed in vitrotranslation inhibition assays; P.S.K. and A.L.K.prepared E. coli 70S ribosomes; J.M. and A.S.M.designed and performed toe-printing experiments;M.D.B. and Y.S.P. designed and performed X-raycrystallographic experiments. N.V.S., A.S.M., A.A.B.,and Y.S.P. supervised the experiments. All authorsinterpreted the results. A.G.T., N.V.S., A.S.M., A.A.B.,and Y.S.P. wrote the manuscript.

Appendix A. Supplementary data

Supplementary data to this article can be foundonline at https://doi.org/10.1016/j.jmb.2018.01.016.

Received 29 August 2017;Received in revised form 22 January 2018;

Accepted 25 January 2018Available online 2 February 2018

Keywords:antibiotic;ribosome;

X-ray structure;protein synthesis;

peptidyl transferase center

Present address: V. A. Sergeeva, Research Centre forMedical Genetics, Russian Academy of Medical Sciences,

Moscow, 115,478, Russia.

Abbreviations used:CHL, chloramphenicol; AA-CAMs, aminoacyl-CAM; PTC,

peptidyl transferase center; aa-tRNA, aminoacyl-tRNA;NPET, nascent peptide exit tunnel; BODIPY-ERY,BODIPY-labeled erythromycin; ORF, open readingframe; PY, protein Y; H-bond, hydrogen bond; Tth,

Thermus thermophilus.

References

[1] N. Polacek, A.S. Mankin, The ribosomal peptidyl transferasecenter: structure, function, evolution, inhibition, Crit. Rev.Biochem. Mol. Biol. 40 (2005) 285–311.

[2] D.N. Wilson, Ribosome-targeting antibiotics and mechanismsof bacterial resistance, Nat. Rev. Microbiol. 12 (2014) 35–48.

[3] S. Schwarz, J. Shen, K. Kadlec, Y. Wang, G. BrennerMichael, A.T. Fessler, et al., Lincosamides, streptogramins,phenicols, and pleuromutilins:modeof action andmechanismsof resistance, Cold Spring Harb. Perspect. Med. 6 (2016).

[4] S.A. McKie, Antibiotics: where to throw the spanner in theribosomal machinery? Future Med. Chem. 8 (2016) 1981–2002.

[5] G.P. Dinos, C.M. Athanassopoulos, D.A. Missiri, P.C.Giannopoulou, I.A. Vlachogiannis, G.E. Papadopoulos,et al., Chloramphenicol derivatives as antibacterial andanticancer agents: historic problems and current solutions,Antibiotics 5 (2016).

[6] J.L. Hansen, P.B. Moore, T.A. Steitz, Structures of fiveantibiotics bound at the peptidyl transferase center of thelarge ribosomal subunit, J. Mol. Biol. 330 (2003)1061–1075.

[7] D. Bulkley, C.A. Innis, G. Blaha, T.A. Steitz, Revisiting thestructures of several antibiotics bound to the bacterialribosome, Proc. Natl. Acad. Sci. U. S. A. 107 (2010)17158–17163.

[8] S. Pestka, Chloramphenicol, in: J.W. Corcoran, F.E. Hahn,J.F. Snell, K.L. Arora (Eds.), Antibitoics: Mechanism of Actionof Antimicrobial and Antitumor Agents, Springer-Verlag,Berlin, Heidelberg, New York 1975, pp. 370–395.

[9] J. Marks, K. Kannan, E.J. Roncase, D. Klepacki, A. Kefi, C.Orelle, et al., Context-specific inhibition of translation byribosomal antibiotics targeting the peptidyl transferasecenter, Proc. Natl. Acad. Sci. U. S. A. 113 (2016)12150–12155.

[10] A.E. Barnhill, M.T. Brewer, S.A. Carlson, Adverse effects ofantimicrobials via predictable or idiosyncratic inhibition ofhost mitochondrial components, Antimicrob. Agents Che-mother. 56 (2012) 4046–4051.

[11] R. Singh, L. Sripada, R. Singh, Side effects of antibioticsduring bacterial infection: mitochondria, the main target inhost cell, Mitochondrion 16 (2014) 50–54.

[12] C.N. Jones, C. Miller, A. Tenenbaum, L.L. Spremulli, A.Saada, Antibiotic effects on mitochondrial translation and inpatients with mitochondrial translational defects, Mitochon-drion 9 (2009) 429–437.

[13] N.G. Ibrahim, J.P. Burke, D.S. Beattie, The sensitivity of ratliver and yeast mitochondrial ribosomes to inhibitors ofprotein synthesis, J. Biol. Chem. 249 (1974) 6806–6811.

852 Amino Acid Analogs of Chloramphenicol

[14] A.J. Lamb, G.D. Clark-Walker, A.W. Linnane, The biogenesisof mitochondria. 4. The differentiation of mitochondrial andcytoplasmic protein synthesizing systems in vitro by antibi-otics, Biochim. Biophys. Acta 161 (1968) 415–427.

[15] J.C. Hanekamp, A. Bast, Antibiotics exposure and healthrisks: chloramphenicol, Environ. Toxicol. Pharmacol. 39(2015) 213–220.

[16] B.H. Cohen, R.P. Saneto, Mitochondrial translational inhibitorsin the pharmacopeia, Biochim. Biophys. Acta 2012 (1819)1067–1074.

[17] L.K. Smith, A.S. Mankin, Transcriptional and translationalcontrol of the mlr operon, which confers resistance to sevenclasses of protein synthesis inhibitors, Antimicrob. AgentsChemother. 52 (2008) 1703–1712.

[18] K.S. Long, J. Poehlsgaard, C. Kehrenberg, S. Schwarz, B.Vester, The Cfr rRNA methyltransferase confers resistanceto Phenicols, Lincosamides, Oxazolidinones, Pleuromutilins,and Streptogramin A antibiotics, Antimicrob. Agents Chemother.50 (2006) 2500–2505.

[19] J.B. Locke, G.E. Zurenko, K.J. Shaw, K. Bartizal, Tedizolidfor the management of human infections: in vitro character-istics, Clin. Infect. Dis. 58 (Suppl. 1) (2014) S35–S42.

[20] P. Mamos, M.G. Krokidis, A. Papadas, P. Karahalios, A.L.Starosta, D.N. Wilson, et al., On the use of the antibioticchloramphenicol to target polypeptide chain mimics to theribosomal exit tunnel, Biochimie 95 (2013) 1765–1772.

[21] A. Bougas, I.A. Vlachogiannis, D. Gatos, S. Arenz, G.P.Dinos, Dual effect of chloramphenicol peptides on ribosomeinhibition, Amino Acids 49 (2017) 995–1004.

[22] R. Vince, R.G. Almquist, C.L. Ritter, S. Daluge, Chloramphen-icol binding site with analogues of chloramphenicol andpuromycin, Antimicrob. Agents Chemother. 8 (1975) 439–443.

[23] M.C. Rebstock, H.M. Crooks, J. Controulis, Q.R. Bartz,Chloramphenicol (chloromycetin). IV. Chemical studies, J.Am. Chem. Soc. 71 (1949) 2458–2462.

[24] D. Drainas, P. Mamos, C. Coutsogeorgopoulos, Aminoacylanalogs of chloramphenicol: examination of the kinetics ofinhibition of peptide bond formation, J. Med. Chem. 36 (1993)3542–3545.

[25] M. Michelinaki, P. Mamos, C. Coutsogeorgopoulos, D.L.Kalpaxis, Aminoacyl and peptidyl analogs of chloramphen-icol as slow-binding inhibitors of ribosomal peptidyltransfer-ase: a new approach for evaluating their potency, Mol.Pharmacol. 51 (1997) 139–146.

[26] K. Yan, E. Hunt, J. Berge, E. May, R.A. Copeland, R.R.Gontarek, Fluorescence polarization method to characterizemacrolide–ribosome interactions, Antimicrob. AgentsChemother. 49 (2005) 3367–3372.

[27] A.G. Tereshchenkov, A.V. Shishkina, V.V. Karpenko, V.A.Chertkov, A.L. Konevega, P.S. Kasatsky, et al., Newfluorescent macrolide derivatives for studying interactionsof antibiotics and their analogs with the ribosomal exit tunnel,Biochemistry (Mosc) 81 (2016) 1163–1172.

[28] J.A. Dunkle, L. Xiong, A.S. Mankin, J.H. Cate, Structures ofthe Escherichia coli ribosome with antibiotics bound near thepeptidyl transferase center explain spectra of drug action,Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 17152–17157.

[29] D. Vazquez, Binding of chloramphenicol to ribosomes. Theeffect of a number of antibiotics, Biochim. Biophys. Acta 114(1966) 277–288.

[30] J.L. Lessard, S. Pestka, Studies on the formation of transferribonucleic acid-ribosome complexes. 23. Chloramphenicol,aminoacyl-oligonucleotides, and Escherichia coli ribosomes,J. Biol. Chem. 247 (1972) 6909–6912.

[31] A.G. Tereshchenkov, A.V. Shishkina, V.N. Tashlitsky, G.A.Korshunova, A.A. Bogdanov, N.V. Sumbatyan, Interactionof chloramphenicol tripeptide analogs with ribosomes,Biochem. Mosc. 81 (2016) 392–400.

[32] Y. Shimizu, Y. Kuruma, T. Kanamori, T. Ueda, The PUREsystem for protein production, Methods Mol. Biol. 1118(2014) 275–284.

[33] N. Vazquez-Laslop, D. Klepacki, D.C. Mulhearn, H. Ramu, O.Krasnykh, S. Franzblau, et al., Role of antibiotic ligand innascent peptide-dependent ribosome stalling, Proc. Natl.Acad. Sci. U. S. A. 108 (2011) 10496–10501.

[34] Y.S. Polikanov, G.M. Blaha, T.A. Steitz, How hibernationfactors RMF, HPF, and YfiA turn off protein synthesis,Science 336 (2012) 915–918.

[35] Y.S. Polikanov, S.V. Melnikov, D. Soll, T.A. Steitz, Structuralinsights into the role of rRNA modifications in proteinsynthesis and ribosome assembly, Nat. Struct. Mol. Biol. 22(2015) 342–344.

[36] B.T. Porse, R.A. Garrett, Sites of interaction of streptograminA and B antibiotics in the peptidyl transferase loop of 23SrRNA and the synergism of their inhibitory mechanisms, J.Mol. Biol. 286 (1999) 375–387.

[37] A. Shishkina, G. Makarov, A. Tereshchenkov, G.Korshunova, N. Sumbatyan, A. Golovin, et al., Conjugatesof amino acids and peptides with 5-o-mycaminosyltylonolideand their interaction with the ribosomal exit tunnel, Bioconjug.Chem. 24 (2013) 1861–1869.

[38] Y.S. Polikanov, I.A. Osterman, T. Szal, V.N. Tashlitsky, M.V.Serebryakova, P. Kusochek, et al., Amicoumacin A inhibitstranslation by stabilizingmRNA interaction with the ribosome,Mol. Cell 56 (2014) 531–540.

[39] M.S. Svetlov, A. Kommer, V.A. Kolb, A.S. Spirin, Effectivecotranslational folding of firefly luciferase without chaperonesof the Hsp70 family, Protein Sci. 15 (2006) 242–247.

[40] N. Vazquez-Laslop, C. Thum, A.S. Mankin, Molecularmechanism of drug-dependent ribosome stalling, Mol. Cell30 (2008) 190–202.

[41] C. Orelle, T. Szal, D. Klepacki, K.J. Shaw, N. Vazquez-Laslop, A.S. Mankin, Identifying the targets of aminoacyl-tRNA synthetase inhibitors by primer extension inhibition,Nucleic Acids Res. 41 (2013), e144. .

[42] C. Orelle, S. Carlson, B. Kaushal, M.M. Almutairi, H. Liu, A.Ochabowicz, et al., Tools for characterizing bacterial proteinsynthesis inhibitors, Antimicrob. Agents Chemother. 57(2013) 5994–6004.

[43] A.J. McCoy, R.W. Grosse-Kunstleve, P.D. Adams, M.D.Winn, L.C. Storoni, R.J. Read, Phaser crystallographicsoftware, J. Appl. Crystallogr. 40 (2007) 658–674.

[44] P. Emsley, K. Cowtan, Coot: model-building tools formolecular graphics, Acta Crystallogr. D Biol. Crystallogr. 60(2004) 2126–2132.

[45] P.D. Adams, P.V. Afonine, G. Bunkoczi, V.B. Chen, I.W.Davis,N. Echols, et al., PHENIX: a comprehensive Python-basedsystem formacromolecular structure solution, ActaCrystallogr.D Biol. Crystallogr. 66 (2010) 213–221.