Synthesis and antimicrobial activity of chloramphenicol-polyamine conjugates

Accepted Manuscript

George E. Magoulas, Ourania N. Kostopoulou, Thomas Garnelis, Constantinos

M. Athanassopoulos, Georgia G. Kournoutou, Michael Leotsinidis, George P.

Dinos, Dionissios Papaioannou, Dimitrios L.Kalpaxis

PII:

S0968-0896(15)00377-6

DOI:

http://dx.doi.org/10.1016/j.bmc.2015.04.069

BMC 12281

Reference:

Bioorganic & Medicinal Chemistry

To appear in:

Received Date:

Revised Date:

Accepted Date:

10 March 2015

23 April 2015

24 April 2015

Please cite this article as: Magoulas, G.E., Kostopoulou, O.N., Garnelis, T., Athanassopoulos, C.M., Kournoutou,

G.G., Leotsinidis, M., Dinos, G.P., Papaioannou, D., L.Kalpaxis, D., Synthesis and antimicrobial activity of

chloramphenicol-polyamine conjugates, Bioorganic & Medicinal Chemistry (2015), doi: http://dx.doi.org/10.1016/

j.bmc.2015.04.069

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers

we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and

review of the resulting proof before it is published in its final form. Please note that during the production process

errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Synthesis and antimicrobial activity of chloramphenicol-polyamine

conjugates

George E. Magoulas^a, Ourania N. Kostopoulou^b,c, Thomas Garnelis^a, Constantinos

M. Athanassopoulos^a, Georgia G. Kournoutou^b, Michael Leotsinidis^d, George P.

Dinos^b, Dionissios Papaioannou^a,* and Dimitrios L.Kalpaxis^b,*

^aLaboratory of Synthetic Organic Chemistry, Department of Chemistry, University of

Patras,GR-26504 Patras, Greece

^bDepartment of Biochemistry, School of Medicine, University of Patras, GR-26504

Patras, Greece

^cPresent address: Center for Molecular Medicine, L8, Karolinska

Universitetssjukhuset, Solna, 17176 Stockholm, Sweden

^dDepartment of Public Health, School of Medicine, University of Patras, GR-26504

Patras, Greece

* Corresponding authors. Tel.: +30 2610962954, fax: + 30 2610962956 (D.P.); Tel.:

+30 2610996124, fax: + 30 2610969167 (D.L.K.)

E-mail addresses: dapapaio@upatras.gr (D. Papaioannou), dimkal@med.upatras.gr

(D.L. Kalpaxis)

ABSTRACT

A series of chloramphenicol (CAM) amides with polyamines (PAs), suitable for

structure-activity relationship studies, were synthesized either by direct attachment of

the PA chain on the 2-aminopropane-1,3-diol backbone of CAM, previously oxidized

selectively at its primary hydroxyl group, or from chloramphenicol base (CLB)

through acylation with succinic or phthalic anhydride and finally coupling with a PA.

Conjugates 4 and 5, in which the CLB moiety was attached on N4 and N1 positions,

respectively, of the N⁸,N⁸-dibenzylated spermidine through the succinate linker, were

the most potent antibacterial agents. Both conjugates were internalized into E. coli

cells by using the spermidine-preferential uptake system and caused decrease in

protein and polyamine content of the cells. Noteworthy, conjugate 4 displayed

comparable activity to CAM in MRSA or wild-type strains of Staphylococcus aureus

and Escherichia coli, but superior activity in E. coli strains possessing ribosomal

mutations or expressing the CAM acetyltransferase (cat) gene. Lead compounds, and

in particular conjugate 4, have been therefore discovered during the course of the

present work with clinical potential.

1

Keywords:

Chloramphenicol

Polyamines

Polyamine conjugates

Polyamine uptake system

Antibacterial activity

1. Introduction

Chloramphenicol (CAM) was isolated in 1947 from Streptomyces venezuele and

two years after was introduced in clinical practice [1]. Its molecule consists of a p-

nitrobenzene moiety, an 2-amino-1,3-propanediol moiety, and a dichloroacetyl tail

(Fig. 1). CAM is active against a wide range of bacteria, usually behaving as a

bacteriostatic drug, although it exhibits bactericidal activity against the three most

common causes of meningitis, Haemophilus influenza, Streptococcus pneumonia, and

Neisseria meningitides [2]. CAM inhibits protein synthesis by binding to the peptidyl

transferase (PTase) center of the bacterial ribosome and abrogating essential

ribosomal functions, like peptide-bond formation [3], termination of translation [4],

and translational accuracy [5].

Resistance or decreased sensitivity to CAM is frequently mediated by

78

mutations/modifications in residues of the ribosomal binding site [6-9], as well as by

decreased influx [10] or/and over-expression of efflux pumps [11], and acetylation of

hydroxyl groups in the 2-amino-1,3-propanediol moiety of CAM [12]. However the

most important concerns, that limit the utility of this antibiotic, relate with its adverse

effects, like neurotoxicity [13], bone marrow depression, and in some cases severe

aplastic anaemia [14]. Intensive research on the adverse effects of CAM has led to the

conclusion that the toxicity of CAM may have a mitochondrial basis [15], particularly

in cases in which underlying mitochondrial defects accentuate CAM binding to

mitochondrial ribosomes [16].

Elucidation of the X-ray structures of CAM bound to bacterial 50S ribosomal

subunits have revealed all the ribosomal residues implicated in the drug binding [17-

18

19]. Most of them are consistent with findings from mutational studies [6-9] and

footprinting analysis [20-2122], and agree with kinetic evidence suggesting that CAM

primarily binds adjacently to the A-site of the PTase center, blocking the

accommodation of the 3΄-aminoacyl end of tRNA within the catalytic region [3]. Two

2

nucleotides, A2058 and A2059 (Escherichia coli numbering is followed throughout

the text), placed outside the PTase center, seem to influence the binding of the drug

by allosteric effects [22] or constitute a secondary binding site of low affinity [23].

Interestingly, crystallographic study on 50S ribosomal subunits from Deinococcus

radiodurans in complex with CAM revealed two putative cationic centers within the

PTase center, involved in the stabilization of the ribosome-CAM complex [17]. One

of these centers, regarded as a Mg²⁺ion, was found to coordinate the interaction

between the 3-OH group of CAM and the carbonyl oxygen at position 4 of U2506. At

the resolution level of this study, however, the possibility of cationic center being

either a monovalent ion or a polyamine (PA) could not be precluded. In fact, kinetic

analysis using a cell-free system derived from E. coli has indicated that PAs enhance

the CAM potency [3]. On the other hand, photoaffinity labeling experiments have

revealed that the binding site of CAM on the ribosome is one of the preferred cross-

linking sites of PAs [24].

The above information prompted us to design and synthesize a series of CAM-PA

conjugates (CAM-PACs), expecting that the PA-conjugated portion could strengthen

the binding of CAM to the ribosome. Moreover, decorating CAM with PAs could

facilitate the transport of CAM into the bacterial cells by exploiting their capacity of

destabilizing the liposaccharide layer of the outer bacterial membrane [10] or/and

acting as a vector system by taking advantage of the bacterial uptake system for PAs

[25]. Previous studies from our group and other investigators have indicated that the

uptake system for PAs in bacteria consists of two members belonging to the ATP

binding cassette (ABC) transporters, characterized by substrate-specificity but also by

26

high tolerance for the import of non-native PAs [25-27].

The synthesized series of CAM-PACs 1-9 is depicted in Figure 1. In these

conjugates, the PA chain is either attached directly on the 2-aminopropane-1,3-diol

backbone of CAM, by selectively oxidizing the primary hydroxyl group of CAM, or

using a dicarboxylic acid linker (succinic or phthalic acid) replacing the

dichloroacetyl part of CAM. With these particular conjugates, we wanted to examine

the effect of (a) the site of attachment of the PA chain on the CAM antibacterial

activity (e.g. compounds 2 and 7) and at the same time to verify if a PA attached to

the 3-position of the 2-aminopropane-1,3-diol chain could mediate interaction of the

CAM-PAC with the U2506 nucleotide of the PTase center, (b) the size and the

number of the free amino functions of the PA chain (compounds 1-3), (c) the site of

3

attachment of CAM on the PA chain (compounds 4 and 5), (d) increasing the

lipophilicity of the PA chain by incorporating benzyl groups on N atoms (compounds

3 and 4), and (e) the nature and the flexibility of the linker (compounds 1 and 6). We

also included in this study two analogues of CAM in which the dichloroacetyl part of

the molecule has been replaced by the 1,2,4-triazole-3-carboxylate unit, which is

either directly connected to the 2-amino group (amide 8) or through a β-alanine spacer

(bisamide 9). With these compounds we wished to examine the effect of replacing the

two chlorine atoms of CAM by N atoms and further evaluate whether removing this

replacement away from the 2-aminopropane-1,3-diol main chain would have any

amplifying effect on the antibacterial activity of the parent compound.

Recently, we published on the effects of CAM-PACs 1-9 on peptide bond

formation catalyzed by the ribosome in a cell-free system derived from E. coli, the

growth of Gram- negative and Gram-positive bacteria, as well as on the viability of

human, healthy and cancer cells [28]. We found that compounds 4 and 5 exhibited

higher activity than CAM in inhibiting the in vitro peptide-bond formation,

compounds 1-3 and 8 displayed activity comparable to those of CAM, while the rest

were essentially inactive. Although compounds 4 and 5 displayed better activity than

CAM against CAM-resistant E. coli strains, they were less active against wild-type

Streptococcus aureus and E. coli cells. Compounds 2, 3 and 9 showed a little activity

against S. aureus, while the rest were quite inactive against both bacteria. This fact

revealed that penetration of the cellular envelope or/and efflux mechanisms may be

significant obstacles to the effectiveness of PA-CAM conjugates acting as

antibacterials. However, compound 4 compared to CAM possessed stronger toxicity

against human leukemic cells, and lower toxicity against healthy human leukocytes,

thus highlighting the potentiality of compound 4 as a safe antibacterial and anticancer

agent. In the present study, we focus on the synthesis of CAM-PACs 1-9, the

mechanisms by which they penetrate the bacterial outer and inner cellular membranes,

and the effects of efflux pumps and the CAM acetyltransferase activity on their

efficacy. Unexpectedly, we found that compounds 4 and 5, once administrated to E

coli cells, cause additionally severe depletion in intracellular polyamine pools, a

finding suggesting that these compounds might have additional targets for their

inhibitory effects beyond their impact on peptide-bond formation.

2. Results and Discussion

4

2.1. Chemistry

The synthesis of CAM-PACs 1-5 is depicted in Scheme 1. It involves a one-pot

acylation of the commercially available chloramphenicol base (CLB) with succinic

anhydride followed by coupling with the appropriate selectively protected, with the

amino protecting groups triphenylmethyl (Trt) or butoxycarbonyl (Boc) or

trifluoroacetyl (Tfa), PA derivatives 10-13, in the presence of the coupling agent O-

(benzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium hexafluorophosphate (HBTU) and

ethyldiisopropylamine. With the exception of PA derivative 14, which was

synthesised from 13 through N⁴-trifluoroacetylation and trifluoroacetic acid (TFA)-

mediated N¹-detritylation, the preparation of the other required PA derivatives 10

[29], 11 [30], 12 [29] and 13 [31] have been described in the literature. From the thus

obtained bisamides, the protecting groups were routinely removed by TFA-mediated

acidolysis or saponification to provide the anticipated compounds 1-5 (see

experimental section). Similarly, acylation of CLB with phthalic anhydride, followed

by coupling with PA derivative 10 gave the anticipated bisamide, which upon TFA-

mediated deprotection provided conjugate 6. On the other hand, conjugate 8 was

readily obtained from coupling of CLB with the commercially available 1,2,4-

triazole-3-carboxylic acid (TCA). Finally, coupling of CLB with ‘active’ ester Trt-β-

Αla-OSu [32] followed by TFA-mediated deprotection and further coupling with TCA

gave conjugate 9.

The synthesis of conjugate 7 is depicted in Scheme 2. It involved selective

protection of the primary hydroxyl function of CAM with the bulky tert-

butyldimethylsilyl (TBDMS) group, followed by protection of the secondary hydroxyl

group with the acetyl (Ac) group to afford the fully protected compound 15. Selective

removal of the TBDMS group gave compound 16, which upon oxidation of the

primary hydroxyl function, provided the corresponding stable carboxylic acid 17 in

55% total yield. It should be noted that direct selective oxidation of the primary

hydroxyl group of CAM or N-Boc [33, 34] protected CLB has been reported but leads

to the corresponding acids in low to moderate yields. In particular the acid from CAM

could be only purified and isolated as its corresponding methyl ester [33]. Acid 17

was then coupled with SPM derivative 11 using the 1-ethyl-3-(3-

dimethylaminopropyl) carbodiimide hydrochloride (EDCI)/ 1-hydroxybenzotriazole

(HOBt) methodology to give the fully protected intermediate 18. From this

5

compound, CAM-PAC 7 was finally obtained upon sequential deprotection, first of

the secondary hydroxyl group, and then of the amino functions of the PA chain.

2.2. Biological evaluation

The antibacterial activities of PA-CAMCs against a panel of CAM-sensitive and

CAM-resistant bacteria are shown in Table 1. Compounds 4 and 5 were the most

active against S. aureus and E. coli strains of wild-type, followed by compounds 3,

and 2, however only against the S. aureus strain. It is evident that replacement of

hydrogen atoms by benzyl groups in the N8-amino function of the spermidine (SPD)

tail of compounds 3, a modification that enhances the lipophilicity of the PA chain

and possibly its potency to adopt stacking interactions with nucleotide bases of

ribosomal RNA, has prominent influence on its antibacterial activity and recognition

by the PA-transporters (compare conjugates 3 and 4). Indeed, footprinting analysis

and Molecular Dynamics (MD) simulations have revealed that incorporation of

benzyl rings at the polyamine end of compound 4 provides additional stability on the

binding to the ribosome through π-stacking interactions of the benzyl rings with

U2585 and U2586 of 23S rRNA [28]. It is yet known that a periplasmic component of

the SPD/spermine (SPM)-preferential uptake system, the PotD protein, specifically

recognizes PAs analogues with intact aminopropane portion(s) [25, 26], while benzyl

substitution at the N4 and N9 amino groups of SPM contributes favorably to the

affinity of acyl-PA analogues for the transporter [27]. Therefore, it is tempting to

assume that CAM-PACs 4 and 5 receive better recognition by the PotD protein, than

the non-benzylated ones 2, 3, and 7. Also, compound 4 is more active than 5,

probably because it possesses an intact aminopropane portion. The position of

polyamine tail and the CAM scaffold, through which the two molecules are coupled,

appears to be of paramount importance. Thus, attachment of SPM at position 3 of the

2-aminopropane-1,3-diol moiety of CAM, following oxidation of the hydroxymethyl

to the carboxyl group, instead of position 2 (the amino function functionalized by a

succinate linker), led to diminution of activity (compare compounds 2 and 7).

Interestingly, the length of the PA chain and the number of the free amino functions

that PA chain carries seem to play critical roles for activity. Thus, comparison of the

IC₅₀values for conjugates 1-5 reveals that the SPD chain offers better opportunities

for activity, followed by the SPM chain, whereas PUT chain is associated with less

activity.

6

The design of compound 7 was based on X-ray crystallographic studies in crystals

of Deinococcus radiodurans 50S ribosomal subunit complexed with CAM, in which

the p-nitrobenzene moiety of CAM is directed away from the catalytic center, while

two putative Mg²⁺ions, Mg-C1 and Mg-C2, mediate interactions between the 3OH

and one of the oxygens of the p-NO₂of CAM with U2506 and U2500 of 23S rRNA,

respectively, holding the drug in place [17]. Unfortunately, the activity of compound

7 in our study was much lower than expected (Table 1). In fact, a recent

crystallographic study conducted in E. coli ribosomes is in controversy with the

previous one [18]; such Mg²⁺ions are not detected, while the orientation of the drug is

rotated approximately by 180^o, thus placing the p-nitrobenzene ring in a position

stacking on C2452 and orienting 3OH away from U2506. Therefore, our results are

more consistent with the second model as well as with previous SAR-studies

indicating that the conformation and integrity of 2-aminopropane-1,3-diol moiety is of

great importance for the activity of CAM [1].

The loss of activity observed for compound 6 denotes the influence of the nature

and the flexibility of the linker on the activity of compound (compare compounds 1

and 6); less flexible and more lipophilic linker results to complete loss of activity.

Finally, replacement of the dichloroacetyl moiety of CAM by the 1,2,4-triazole-3-

carboxylate unit resulted in a product (compound 8) exhibiting worse activity, when

compared to the parent compound. Removing this replacement away from the 2-

aminopropane-1,3-diol moiety resulted in compound 9 that displayed mostly weak

activity, but still remaining lower than those of CAM.

In a closed system, like a cell-free system, an important contributor to the

efficacy of a drug is its affinity for the target. Drug and target are at equilibrium, and

thus affinity constants precisely quantify the concentration of the encounter complex,

provided that drug-inactivating and/or target-modifying enzymes are absent or

inactive. However, the cell is an open system in which drug concentration and the

target itself fluctuate over time. Assuming that the function and concentration of

target remain unalterable, the drug concentration depends on the relative activities of

uptake and efflux systems. The cytoplasm also must not contain drug-modifying

enzymes. For instance, E. coli possesses more than seven efflux transporters that

remove a broad range of antibiotics and other toxic compounds from the intracellular

space. One well-studied efflux pump is the AcrA-AcrB-TolC tripartite pump system

which is assembled by the outer membrane protein TolC, the periplasmic protein

7

AcrA, and the inner membrane protein AcrB [35]. Deletions in acrAB and/or tolC

genes result in increased susceptibility of E. coli to a panel of compounds, including

CAM [36]. To investigate the effect of this efflux system on our compounds, we

determined the IC₅₀values against an E. coli strain BL21 DE3 lacking the tolC gene

that codes the TolC protein. As shown in Table 1, this efflux system does not seem to

affect the antibacterial activity of CAM-PACs, as the IC₅₀values regarding this strain

were similar to those of wild-type E. coli. However, E. coli BL21 DE3 (ΔtolC) strain

was approximately 3-fold more sensitive to CAM. Nevertheless, the probability that

CAM-PACs are extruded through another set of efflux systems of RND family [37],

cannot be excluded with absolute certainty.

Regarding now the internalization of a drug into the bacterial cell, the first barrier

that should be overcome is the outer membrane. This requirement explains our

finding that the IC₅₀values for CAM and CAM-PACs measured in S. aureus are

smaller than those measured in E. coli (Table 1). S. aureus, as a representative of

Gram-positive bacteria, possessing a simpler and more permeable outer membrane

than those existing in E. coli, appears higher susceptibility to the tested compounds.

This prompted us to examine the activity of CAM-PACs against three multi-drug

resistant (MDR) MRSA isolates, GRE2272, GRE2372, and GRE2691. The resistance

of these isolates against a variety of antibiotics, in addition to methicillin, is reported

in Table S2. Interestingly, all of them showed similar or better susceptibility to

compound 4, than to CAM. Given that compound 4 is better tolerated than CAM by

human leukocytes [28], this conjugate seems to be a well promising lead candidate in

designing efficacious drugs against MDR Gram-positive bacteria.

CAM utilizes channel proteins, like pore-forming porins, for access to the

periplasm in Gram-negative bacteria, like E. coli [10]. Conjugation of CAM with PAs

should be excluded from the passage, since PA portion(s) increase the size of the

drug. Apart from steric hindrance, PA portions, due to their polycationic nature, may

bind to internal, negatively charged regions of porins and block penetration [38]. On

the other hand, compounds possessing polycationic components can destabilize the

liposaccharide layer of the outer membrane and cross the outer-membrane barrier, by

using a self promoted uptake pathway [37]. Moreover, the PA portion(s) may endow

CAM-PACs with the ability to penetrate the bacterial inner membrane, by utilizing

the polyamine uptake system, a group of polyamine carriers which pertain to the

ATP-binding cassette transporter family [25]. To test this hypothesis, S. aureus or E

8

coli cells grown in M9 medium were exposed to an IC₅₀dose of each CAM-PAC, at

which 50% of growth inhibition was observed. Interestingly, 10×IC₅₀exogenously

added polyamines were able to significantly salvage normal growth of bacterial cells

exposed to an IC₅₀dose of compounds 2-5 (Fig. 2). Ιn contrast, when exogenous

polyamines were added to S. aureus or E coli cells exposed to an IC₅₀dose of CAM,

no rescue was observed. It is mentioned that the concentration used for each

polyamine was much lower than the maximum one tolerated by S. aureus or E coli

cells [39]. We rationalized the competition observed as being due to the relative

propensity of compounds 2-5 to use the polyamine transport system for cell entry (see

also next paragraph). Except for compound 2, all the rest CAM-PACs of this

subgroup competed more efficiently with putrescine and spermidine for the

polyamine uptake system, than with spermine. This finding tempted us to assume that

CAM-PACs are mainly internalized into the cell through the transporter encoded by

the pPT104 operon, which constitutes the spermidine-preferential uptake system [25].

The molecular demands for the selective delivery of PA conjugates into eukaryotic

cells are slightly different than those required by the prokaryotic PA-uptake system

[40]. Nevertheless, the most effective member of CAM-PACs in bacteria, compound

4, was also able to penetrate eukaryotic cells. As detected previously [28], compound

4 compared to CAM possesses enhanced toxicity against human cancer cells, and

lower toxicity against healthy human cells. This was attributed to the highly active

PA-transporters (PAT) existing in cancer cells [41].

The rescue in growth observed when CAM-PACs and exogenous polyamines are

simultaneously added in the medium of bacterial culture may arise either due to a

competition between them for the polyamine uptake system or to an increase of the

protein synthesis rate, owing to the enrichment of the intracellular environment in

polyamines. To differentiate between these two possibilities and to confirm that

CAM-PACs act on their target after entry into the cells, E. coli cells at an A₅₆₀of

0.020 were grown at 37^oC in M9 medium supplemented with compound 4 or 5 at

concentrations equal to their respective IC₅₀. In parallel, cells were grown in the

absence of CAM-PACs (control cultures). The cells were harvested when the control

cultures reached an A₅₆₀of 0.800, washed and then homogenized in 0.6 M HClO₄.

After centrifugation, the supernatant was analyzed by RP-HPLC, while the pellet was

extracted in 0.1 M NaOH and analyzed for protein content. Using this analytical

approach, we were able to detect micromolar quantities of compounds 4 and 5 in the

9

cellular extracts of E. coli (Table 2, Fig. S1). Although much less than the exogenous

doses, the intracellular concentrations of CAM-PACs detected were adequate to

substantially inhibit protein synthesis, given that the inhibition constant (K_i*) values

in the puromycin reaction, previously calculated for these compounds [28], are less

than 1 μΜ (Table S1). Consistently, inhibition of cell growth in the presence of

compounds 4 or 5 was accompanied by remarkable decrease in the protein content of

cells (Table 2). Surprisingly, compounds 4 and 5 once administrated to E. coli cells

interfered with polyamine metabolism, as detected by changes measured in the

intracellular polyamine content. Putrescine and spermidine intracellular pools were

depleted by around 75% and 80%, respectively, at the time of cell harvesting, i.e. 3 h

after the addition of compound 4, whereas the corresponding depletions when

compound 5 used as the challenging agent reached the values 90% and 95%,

respectively (Table 2). In contrast to some studies reporting liberation of the appended

polyamine from the drug scaffold once administrated within cells [42], we failed to

detect N⁸,N⁸-dibenzylspermidine in the cellular extracts by RP-HPLC (Fig. S2), and

therefore we suggest that conjugates 4 and 5 are not decomposed in the cellular

environment to any significant extent. Nevertheless, polyamines are essential for cell

growth and protein synthesis [43], and therefore the revealed effects on polyamine

content suggest that CAM-PACs may directly exhibit an additional mode of toxicity.

The origin of this additional effect of CAM-PACs is not clear at present.

It should be noted that, conjugate 4 was more efficient than CAM in inhibiting

the growth of A2058G and A2503C mutants of E. coli, while both mutations failed to

provide any growth advantage to E. coli cells against compound 5, similar to that

conferred against CAM (Table 1) [7,44,45]. It should be mentioned that nucleoside

A2058 along with A2059 participate in the formation of a hydrophobic crevice at the

entrance to the exit tunnel. As previously revealed [22,46], even minor changes in the

structure or conformation of these nucleosides can affect functions of the PTase center

and response to antibiotics, via allosteric effects transmitted through a signal

exchange network including A2503 and A2062. Therefore, it is perceived why

mutations A2058G and A2503G confer resistance to E. coli against CAM. However,

when compound 5 is used as antibacterial, the signal exchange network is interrupted

at the level of nucleoside A2062 (Fig. S3A), i.e. no hydrogen bond is formed between

A2062 and compound 5, a fact that impairs the transmission of signals to the PTase

center. On the other hand, compound 4 preserves this hydrogen bond and also

10

displays π-π interaction with U2585 (Fig. S3B), thus limiting the flexibility of this

nucleoside. As proposed by others, U2585 is implicated in the binding of peptidyl-

tRNA to the P-site [47,48]. In addition, Agmon et al suggested that loss of the

flexibility of U2585 may be detrimental for peptide-bond formation, given that U2585

along with A2602 control the spiral rotation of the acceptor 3΄-acceptor end of a new-

bound aminoacyl-tRNA (A/T site) as seeking out its functional orientation towards

the P-site bound peptidyl-tRNA [49].

Last but not least, compounds 4 and 5 exhibited higher activity than CAM in

inhibiting the growth of Rosetta(DE3)pLysS E. coli cells expressing the cat gene

(Table 1). Production of CAM acetyltransferase, encoded by the cat gene, is a major

mechanism by which bacteria become resistant to CAM. This enzyme catalyzes

transfer of the acetyl moiety from acetyl coenzyme A to CAM [50]. The O-acetoxy

derivatives of CAM are devoid of antibiotic activity, because they do not bind to

bacterial ribosomes [50]. The sensitivity of Rosetta(DE3)pLysS E. coli cells against 4

and 5 suggests that these compounds are not so good substrates for CAM

acetyltransferase compared to CAM. This was confirmed by calculating the

specificity of CAM and compounds 4 and 5 for CAM acetyltransferase, using the

ratio V_max/K_mto benchmark the capacity of each drug to be used as acceptor of acetyl

groups. As shown in Table 3 and Fig. S4, compound 4 behaves less efficiently as

substrate of CAM acetyltransferase, compared to CAM or compound 5, and therefore

is more capable of combating this type of CAM resistance.

3. Conclusions

A series of CAM-PACs (1-9) were readily synthesized from either CAM, through

indirect oxidation of its primary hydroxyl function, or from CLB through N-acylation

with succinic or phthalic anhydride, followed by coupling with suitably protected

polyamines and finally, routine O- and/or N-deprotection. The study of their

antibacterial activity against a variety of bacterial strains and in comparison to CAM

revealed the structural characteristics which are necessary for optimum activity,

namely the presence of the hydroxymethyl unit in the 2-amino-1,3-diol CAM subunit,

the preference for a more flexible aliphatic dicarboxylic acid linker and the N⁸,N⁸-

dibenzyl-SPD as the PA chain, preferably acylated at position N4. The most active

conjugate 4, incorporating these characteristics, presented comparable antibacterial

activity to CAM against MRSA or wild-type S. aureus and E. coli cells, and superior

activity against E. coli mutants that exhibit resistance against CAM. Given that

11

antibiotic resistance is considered to be one of the major global public health threats,

compounds that are capable to combat resistant strains acquire great importance.

Although oligo-amines, like polymyxins and aminoglycosides, are relatively

neurotoxic and nephrotoxic because of their polycationic nature at physiological pH

(7.4) [51,52], natural polyamines are ubiquitous and essential components of cells.

Due to their cellular origin, numerous derivatives and analogues of polyamines have

been synthesized and some of them are presently in early phases of clinical trials [53].

Nevertheless, high extracellular polyamine levels can lead to toxicity and

neurotoxicity. Despite that direct cytotoxic actions are observed only at mM

concentrations [54], any potential toxic effects of polyamine conjugates against

healthy human cells must be avoided. Encouragingly, a previous study of CAM-PACs

has indicated that these compounds are well tolerated by human leukocytes [28].

Certainly, further work has to be performed to refute the suspicion that toxic

properties may be shared by CAM-PACs.

Lead compounds, and in particular conjugate 4, have been therefore discovered

during the course of the present work with clinical potential. We are currently in the

process of exploiting the mechanisms through which these conjugates exert their

antibacterial effects and designing better compounds, taking into consideration results

from related molecular modeling studies and using conjugate 4 as the lead compound.

4. Experimental

4.1. Materials and analytical methods

Melting points were determined with a Buchi SMP-20 apparatus and are

uncorrected. IR spectra were recorded as KBr pellets, where necessary, on a Perkin

1

Elmer 16PC FT-IR spectrophotometer. H NMR spectra were obtained at 400.13

MHz and ¹³C NMR spectra at 100.62 MHz on a Bruker DPX spectrometer. Chemical

shifts are reported in δ units, parts per million (ppm) downfield from TMS. Electron-

spray ionization (ESI) mass spectra were recorded at 30V, on a Micromass-Platform

LC spectrometer using MeOH as solvent. The MALDI high-resolution MS (HRMS)

experiments were carried out in an Applied Biosystems MALDI-TOF/TOF A/B 4700

Proteomics Analyzer instrument using dithranol as matrix. Microanalyses were

performed on a Carlo Erba EA 1108 CHNS elemental analyzer in the Laboratory of

Instrumental Analysis of the University of Patras. Flash column chromatography

(FCC) was performed on Merck silica gel 60 (230-400 mesh) and TLC on 60 Merck

12

60F₂₅₄films (0.2 mm) precoated on aluminium foil. Spots were visualized with UV

light at 254 nm and ninhydrine solution or charring agents. All solvents were dried

and/or purified according to standard procedures prior to use. All reagents employed

in the present work were purchased from either Aldrich or Alfa-Aesar. The synthesis

and characterization of compounds 13 and 14 are described in the supplementary data

section.

4.2. General procedure for the synthesis of CAM-PACs 1-6

Succinic (0.1 g, 1.0 mmol) or phthalic (0.15 g, 1.0 mmol) anhydride was added to a

solution of CLB (0.21 g, 1.0 mmol) in DMF (3.0 mL). The reaction mixture was

stirred at ambient temperature for 1.5h and the consumption of CLB was checked by

TLC using as eluant CHCl₃/MeOH (1:1). Following completion of the reaction, the

o

resulting solution was cooled to 0 C, the appropriate PA derivative (1.1 mmol) was

i

added followed by HBTU (0.46 g, 1.2 mmol) and finally Pr₂NEt (0.52 mL, 3.0

mmol). The reaction mixture was stirred at ambient temperature for the indicated time

and diluted with EtOAc. The organic phase was washed twice with a 5% aq. solution

of NaHCO₃and twice with H₂O, dried over Na₂SO₄and evaporated to dryness. The

protected CAM-PACs were obtained in pure form following FCC purification.

4.2.1. N¹-(4-Aminobutyl)-N⁴-((1R,2R)-1,3-dihydroxy-1-(4-nitrophenyl)-propan-2-

yl)succinamide (1)

The N-protected compound 1 was synthesized according to the above general

procedure using 10 (0.36 g, 1.1 mmol) as the PA derivative. Reaction time: 1h; Yield:

0.47 g (75%); White foam; R_f(CHCl₃/MeOH 95:5): 0.16; IR (KBr, cm^-1): 3381, 3350,

2934, 1647, 1519, 1351, 753, 712; MS (ESI, 30eV): m/z 663.30 [M+K], 647.43

[M+Na], 625.39 [M+H], 243.16 [Trt]; HRMS (m/z): [M+H]⁺calcd for C₃₆H₄₁N₄O₆,

625.3026. Found, 625.3046; ¹H NMR (CDCl₃): δ 8.10 (d, J = 8.8 Hz, 2H), 7.51 (d, J =

8.8 Hz, 2H), 7.43-7.38, 7.27-7.21 and 7.18-7.13 (three m, 15H), 7.09 (d, J = 8.8 Hz,

2H), 6.28 (t, J = 5.6 Hz, 1Η), 5.10 (d, J = 3.2 Hz, 1H), 4.09 (unresolved dddd, 1H),

3.76 (dd, J = 5.2 and 11.6 Hz, 1H), 3.66 (dd, J = 4.4 Hz and 11.6 Hz, 1H), 3.08

(unresolved q, 2H), 2.47-2.36 (m, 2H), 2.35-2.24 (m, 2H), 2.12 (unresolved t, 2H),

13

1.46-1.39 (m, 4H); C NMR (CDCl₃): δ 173.3, 172.5, 149.1, 147.2, 145.9 (three C),

128.6 (six C), 127.8 (six C), 126.9 (two C), 126.3 (three C), 123.4 (two C), 72.6, 70.9,

62.8, 56.4, 43.3, 39.7, 31.3 (two C), 28.0, 27.2.

13

The thus obtained N-protected conjugate (0.4 g, 0.64 mmol) was treated at ambient

temperature with a 30% TFA solution in CH₂Cl₂(2 mL) for 1h. The resulting solution

was evaporated to dryness and the residue was triturated with Et₂O to afford

conjugate 1, as its corresponding trifluoroacetate salt. Yield: 0.29 g (90%); white

solid; IR (KBr, cm^-1): 3395, 3025, 3000-2700, 1676, 1542, 1445; MS (ESI, 30eV):

m/z 405.46 [M+Na], 383.41 [M+H], 365.47 [(M+H)-H₂O]; HRMS (m/z): [M+H]⁺

1

calcd for C₁₇H₂₇N₄O₆, 383.1931. Found, 383.1959; H NMR (d₄-MeOH): δ 8.17 (d, J

= 8.4 Hz, 2H), 7.62 (d, J = 8.4 Hz, 2H), 5.07 (d, J = 3.0 Hz, 1H), 4.12 (dt, J = 3.0 and

6.6 Hz, 1H), 3.74 (dd, J = 6.6 and 10.8 Hz, 1H), 3.54 (dd, J = 6.6 and 10.8 Hz, 1H),

3.23-3.08 (m, 2H), 2.91 (t, J = 7.2 Hz, 2H), 2.48-2.28 (m, 4H), 1.63 (quintet, J = 7.2

Hz, 2H), 1.54 (quintet, J = 7.2 Hz, 2H); ¹³C NMR (d₄-MeOD): δ 174.9 (two C),

152.2, 148.6, 128.6 (two C), 124.2 (two C), 72.1, 62.8, 58.0, 40.4, 39.5, 32.2, 32.0,

27.4, 25.8.

4.2.2.

N¹-(3-(4-(3-Aminopropylamino)butylamino)propyl)-N⁴-((1R,2R)-1,3

dihydroxy-1-(4-nitrophenyl)-propan-2-yl)succinamide (2)

Then N-protected compound 2 was synthesized according to the above general

procedure using 11 (0.55 g, 1.1 mmol) as the PA derivative. Reaction time: 2h; Yield:

0.66 g (83%); colorless oil; R_f(CHCl₃/MeOH 97:3): 0.19; IR (thin film, CHCl₃, cm^-1):

3480 3448, 2922, 1670, 1636, 1418, 1162, 728; MS (ESI, 30eV): m/z 835.37 [M+K],

819.31 [M+Na], 797.39 [M+H]; HRMS (m/z): [M+H]⁺calcd for C₃₈H₆₅N₆O₁₂,

1

797.4660. Found, 797.4697; H NMR (CDCl₃): δ 8.16 (d, J = 8.8 Hz, 2H), 7.58 (d, J

= 8.8 Hz, 2H), 7.05 (br.s, 1H), 6.93 (br.s., 1H), 5.17 (d, J = 4.0 Hz, 1Η), 4.07-4.00 (m,

1H), 3.86 (dd, J = 4.0 and 11.6 Hz, 1H), 3.76 (dd, J = 4.0 and 11.2 Hz, 1H), 3.27-3.18

(m, 4Η), 3.17-3.03 (m, 8H), 2.50-2.32 (m, 4H), 1.69-1.57 (m, 4H), 1.49-1.45 (m, 4H),

1.42 (s, 27H); ¹³C NMR (CDCl₃): δ 173.5, 172.4, 156.5, 156.1 (two C), 149.1, 147.3,

127.0 (two C), 123.4 (two C), 80.0 (two C), 79.2, 73.3, 63.1, 56.9, 46.9 (two C), 44.0,

43.6, 37.6, 36.1, 31.8, 31.6, 28.4 (nine C), 28.0, 27.7, 25.9 (two C).

The thus obtained N-protected conjugate (0.6 g, 0.75 mmol) was treated with a

50% TFA solution in CH₂Cl₂(4 mL) for 2h. The reaction mixture was evaporated to

dryness and the residue was triturated with Et₂O to afford conjugate 2, as its

corresponding tristrifluoroacetate salt. Yield: 0.58 g (92%); colorless oil; IR (thin

film, CHCl₃, cm^-1): 3396, 3030-2700, 1684, 1674, 1654, 1202, 1132, 722; MS (ESI,

30eV): m/z 519.45 [M+Na], 497.47 [M+H]; HRMS (m/z): [M+H]⁺calcd for

14

C₂₃H₄₁N₆O₆, 497.3088. Found, 497.3042; ¹H NMR (d₄-MeOH): δ 8.18 (d, J = 8.8 Hz,

2H), 7.79 (d, J = 9.2 Hz, 1H), 7.63 (d, J = 8.8 Hz, 2H), 5.01 (d, J = 2.8 Hz, 1H), 4.12

(unresolved ddt, 1H), 3.75 (dd, J = 6.4 and 10.8 Hz, 1H), 3.53 (dd, J = 6.4 Hz and

10.8 Hz, 1H), 3.13 (t, J = 7.6 Hz, 2H), 3.10-3.03 and 3.03- 2.95 (two m, 10H), 2.51-

2.30 (m, 4H), 2.08 (quintet, J = 7.6 Hz, 4H), 1.86-1.77 (m, 4H); ¹³C NMR (d₄-

MeOD): δ 176.6, 175.1, 152.7, 149.1, 129.0 (two C), 124.7 (two C), 72.5, 63.2, 58.4,

48.8, 48.6, 46.7, 46.4, 38.4, 37.2, 32.4, 32.3, 28.2, 25.9, 24.8, 24.7.

4.2.3.

N¹-(4-Aminobutyl)-N¹-(3-aminopropyl)-N⁴-((1R,2R)-1,3-dihydroxy-1-(4-

nitrophenyl)propan-2-yl)succinamide (3)

The N-protected compound 3 was synthesized according to the above general

procedure using 12 (0.69 g, 1.1 mmol) as the PA derivative. Reaction time: 2h; Yield:

0.67 g (72%); white foam; R_f(CHCl₃/MeOH 97:3): 0.27; IR (KBr, cm^-1): 3394, 3320,

3056, 3018, 2932, 2856, 1620, 1518, 1488, 1448, 1346, 1216, 750, 706; MS (ESI,

1

30eV): m/z 946.34 [M+Na], 924.42 [M+H], 682.45 [(M+H)-Trt], 243.47 [Trt]; H

NMR (CDCl₃): δ 8.14 (d, J = 8.8 Hz, 2H), 7.54 (d, J = 8.8 Hz, 2H), 7.48-7.41, 7.30-

7.22 and 7.21-7.14 (three m, 30H), 6.80 (br.s, 1H), 5.13 (d, J = 3.2 Hz, 1H), 4.02-3.96

(m, 1H), 3.82 (dd, J = 4.0 and 11.6 Hz, 1H), 3.71 (dd, J = 4.0 and 11.6 Hz, 1H), 3.41-

3.20 (m, 2H), 3.18-3.04 (m, 2H), 2.47-2.28 (m, 4H), 2.21-2.08 (m, 4H), 1.73-1.60 (m,

13

2H), 1.49-1.39 (m, 4H); C NMR (CDCl₃): δ 173.9, 172.0, 149.2, 147.2, 146.8 (six

C), 128.6 (twelve C), 127.8 (twelve C), 127.0 (six C), 126.5 (two C), 123.4 (two C),

73.2, 71.0 (two C), 63.0, 56.8, 48.0, 46.4, 44.3, 43.3, 31.6, 30.1, 28.8, 26.5, 25.8.

The thus obtained N-protected conjugate (0.6 g, 0.65 mmol) was treated with a

30% TFA solution in CH₂Cl₂(4 mL) for 2h. The reaction mixture was evaporated to

dryness and the residue was triturated with Et₂O to afford conjugate 3, as its

corresponding bistrifluoroacetate salt. Yield: 0.40 g (93%); white foam; IR (KBr, cm^-

¹): 3390, 3000-2700, 1681, 1642, 1531, 1461, 1347, 1202, 1150, 840, 722; MS (ESI,

30eV): m/z 462.39 [M+Na], 440.48 [M+H]; HRMS (m/z): [M+H]⁺calcd for

C₂₀H₃₄N₅O₆, 440.2509. Found, 440.2532; ¹H NMR (d₆-DMSO): δ 8.15 (d, J = 8.4 Hz,

2H), 7.89-7.70 (m, 6H), 7.58 (d, J = 8.4 Hz, 2H), 5.00 (unresolved d, 1Η), 3.95 (m,

1H), 3.54 (unresolved dd, 1H), 3.32-3.22 (m, 3H), 3.21-3.14 (m, 2H), 2.84-2.75 (m,

2H), 2.73-2.64 (m, 2H), 2.40-2.17 (m, 4H), 1.80-1.65 (m, 2H), 1.56-1.41 (m, 4H); ¹³C

NMR (d₆-DMSO): δ 172.4, 171.9, 152.4, 146.7, 127.9 (two C), 123.2 (two C), 69.9,

60.9, 56.4, 46.9, 44.4, 42.2, 39.0, 37.0, 30.7, 28.1, 25.9, 24.7.

15

4.2.4.

N¹-(3-Aminopropyl)-N¹-(4-dibenzylamino)butyl)-N⁴-((1R,2R)-1,3-

dihydroxy-1-(4-nitrophenyl)-propan-2-yl)succinamide (4)

The N-protected compound 4 was synthesized according to the above general

procedure using 13 (0.62 g, 1.1 mmol) as the PA derivative. Reaction time: 2h; Yield:

0.62 g (72%); white foam; R_f(PhMe/EtOAc 2:8): 0.27; IR (KBr, cm^-1): 3352, 2927,

1618, 1522, 1490, 1346, 1072, 748, 698; MS (ESI, 30eV): m/z 884.49 [M+Na],

862.32 [M+H], 620.41 [(M+H)-Trt], 243.41 [Trt]; HRMS (m/z): [M+H]⁺calcd for

1

C₅₃H₆₀N₅O₆, 862.4544. Found, 862.4576; H NMR (CDCl₃,): δ 8.14 (d, J = 8.8 Hz,

2H), 7.54 (d, J = 8.8 Hz, 2H), 7.46-7.14 (m, 25H), 6.70 (d, J = 6.4 Hz, 1H), 5.14 (d, J

= 4.4 Hz, 1Η), 3.98-3.89 (m, 1H), 3.84 (unresolved dd, 1H), 3.71 (unresolved dd, 1H),

3.55 (br. s, 4Η), 3.36-3.07 (m, 4H), 2.61-2.20 (m, 6H), 2.16-2.01 (m, 2H), 1.68-1.56

13

(m, 2H), 1.52-1.36 (m, 4H); C NMR (CDCl₃,): δ 174.1, 171.9, 149.0, 147.2, 146.8

(three C), 128.8 (two C), 128.6 (four C), 128.5 (six C), 127.9 (six C), 127.8 (four C),

126.9 (three C), 126.4 (two C), 126.3 (two C), 123.4 (two C), 73.5, 71.0, 63.2, 58.6

(two C), 57.0, 47.7, 46.3, 44.2, 41.2, 31.7, 29.7, 28.9, 26.2, 25.5.

The thus obtained N-protected conjugate (0.55 g, 0.64 mmol) was treated with a

30% TFA solution in CH₂Cl₂(4 mL) for 1h. The reaction mixture was evaporated to

dryness and the residue was triturated with Et₂O to afford conjugate 4, as its

corresponding bistrifluoroacetate salt. Yield: 0.51 g (94%); white foam; IR (KBr, cm^-

¹): 3386, 3000-2700, 1676, 1641, 1552, 1458, 1348, 1202, 1141, 836, 722; MS (ESI,

30eV): m/z 642.41 [M+Na], 620.48 [M+H]; HRMS (m/z): [M+H]⁺calcd for

C₃₄H₄₆N₅O₆, 620.3448. Found, 620.3455; ¹H NMR (d₆-DMSO): δ 8.23 (d, J = 8.0 Hz,

1H), 8.12 (d, J = 8.8 Hz, 2H), 7.47 (d, J = 8.4 Hz, 2H), 7.89-7.73 (m, 3H), 7.59 (d, J =

8.8 Hz, 2H), 7.54-7.39 (m, 10H), 5.00 (unresolved d, 1Η), 4.41-4.13 (m, 5H), 3.98

(unresolved dd, 1H), 3.56 (unresolved dd, 1H), 3.35-3.10 (m, 4H), 2.96-2.85 (m, 2H),

2.75-2.65 (m, 2H), 2.39-2.23 (m, 4H), 1.81-1.66 (m, 2H), 1.46-1.26 (m, 4H); ¹³C

-

NMR (d₆-DMSO): δ 172.3, 171.9, 158.8 (q, J = 32.7 Hz, CF₃CO₂), 148.2, 146.7,

131.5 (two C), 130.0 (two C), 129.3 (four C), 128.2 (two C), 127.9 (two C), 127.8

-

(two C), 123.2 (two C), 117.3 (q, J = 295.8 Hz, CF₃CO₂), 69.9, 60.9, 56.7 (two C),

56.4, 46.6, 44.3, 42.1, 38.7, 37.0, 30.6, 27.9, 25.9, 25.7.

4.2.5. N¹-(3-(4-(Dibenzylamino)butylamino)propyl)-N⁴-((1R,2R)-1,3-dihydroxy-1-

(4-nitrophenyl)-propan-2-yl)succinamide (5)

16

The N-protected compound 5 was synthesized according to the above general

procedure using 14 (0.57 g, 1.1 mmol) as the PA derivative. Reaction time: 1h; Yield:

0.47 g (65%); colorless oil; R_f(CHCl₃/MeOH 9:1): 0.33; IR (thin film, CHCl₃, cm^-1):

3408, 3331, 3037, 2938, 2854, 1618, 1552, 1490, 1346, 1215, 1072, 748, 698; MS

(ESI, 30eV): m/z 754.37 [M+K], 738.41 [M+Na], 716.39 [M+H]; HRMS (m/z):

1

[M+H]⁺calcd for C₃₆H₄₅F₃N₅O₇, 716.3271. Found, 716.3297; H NMR (CDCl₃): δ

8.18 (d, J = 8.8 Hz, 2H), 7.58 (d, J = 8.8 Hz, 2H), 7.45-7.23 (m, 10H), 6.89 (br.s, 1H),

6.63 (br.s, 1H), 5.18 (d, J = 3.6 Hz, 1Η), 4.09-4.00 (m, 1H), 3.88 (dd, J = 4.0 Hz and

9.6 Hz, 1H), 3.76 (dd, J = 4.0 Hz and 9.6 Hz, 1H), 3.62 (br. s, 4H), 3.36 (t, J = 6.8 Hz,

2H), 3.24 (q, J = 6.4Hz, 2H), 3.18-3.07 (m, 2H), 2.61-2.35 (m, 6H), 1.68 (quintet, J =

6.4 Hz, 2H), 1.62-1.46 (m, 4H); ¹³C NMR (CDCl₃): δ 173.4, 172.6, 148.9, 147.3,

137.7 (two C), 129.0 (four C), 128.4 (four C), 126.9 (four C), 123.4 (two C), 73.3,

63.1, 58.5 (two C), 56.9, 52.5, 47.5, 44.0, 36.2, 31.7, 31.6, 28.7, 26.9, 26.3.

Finally, water (1 mL) and K₂CO₃(0.46 g, 3.36 mmol) were added to a solution of

the thus obtained N⁴-protected conjugate (0.4 g, 0.56 mmol) in MeOH (15 mL). The

o

resulting suspension was refluxed at 65 C for 1h. Upon completion of deprotection,

salts were filtered off and discarded, the filtrate was evaporated to dryness and the

residue was taken up in EtOAc. The organic phase was washed twice with brine, dried

over Na₂SO₄and evaporated under vacuo. Pure compound 5 was obtained following

FCC purification. Yield: 0.20 g (59%); colorless oil; R_f(CHCl₃/MeOH 8:2): 0.12; IR

(thin film, CHCl₃, cm^-1): 3296, 2934, 2808, 1644, 1520, 1449, 1346, 1076, 750, 700;

MS (ESI, 30eV): m/z 620.48 [M+H]; HRMS (m/z): [M+H]⁺calcd for C₃₄H₄₆N₅O₆,

1

620.3448. Found, 620.3483; H NMR (CDCl₃): δ 8.13 (d, J = 8.8 Hz, 2H), 7.54 (d, J

= 8.8 Hz, 2H), 7.36-7.24 and 7.23-7.18 (two m, 10H), 7.10 (t, J = 4.8 Hz, 1H), 7.02

(d, J = 8.4 Hz, 1H), 5.12 (d, J = 2.8 Hz, 1H), 4.06 (unresolved ddt, 1H), 3.81 (dd, J =

4.0 Hz and 9.6 Hz, 1H), 3.71 (dd, J = 4.0 Hz and 9.6 Hz, 1H), 3.52 (br. s, 4H), 3.37-

3.25 (m, 2H), 3.18-3.08 (m, 2H), 2.57 (t, J = 6.4 Hz, 2H), 2.48-2.26 (m, 6H), 1.65-

13

1.36 (m, 6H); C NMR (CDCl₃,): δ 173.1, 172.4, 149.4, 147.2, 139.8 (two C), 128.7

(four C), 128.1 (four C), 126.9 (two C), 126.8 (two C), 123.3 (two C), 73.0, 63.1,

58.3 (two C), 56.3, 53.0, 49.0, 47.4, 38.3, 31.7, 31.6, 28.2, 27.3, 24.8.

4.2.6. N¹-(4-Aminobutyl)-N²-((1R,2R)-1,3-dihydroxy-1-(4-nitrophenyl)-propan-2-

yl)phthalamide (6)

17

The N-protected CLB conjugate was synthesized according to the above general

procedure using 10 (0.36 g, 1.1 mmol) as the PA derivative. Reaction time: 2h; Yield:

0.57 g (85%); white foam; R_f(CHCl₃/MeOH 95:5): 0.26; IR (KBr, cm^-1): 3395, 3294,

2921, 1656, 1515, 1337, 757, 707; MS (ESI, 30eV): m/z 695.48 [M+Na], 673.50

[M+H]; HRMS (m/z): [M+H]⁺calcd for C₄₀H₄₁N₄O₆, 673.3026. Found, 673.3042; ¹H

NMR (CDCl₃): δ 8.14 (d, J = 8.8 Hz, 2H), 7.49 (d, J = 8.8 Hz 2H), 7.46-7.39, 7.34-

7.21 and 7.19-7.05 (three m, 20H), 6.82 (unresolved t, 1Η), 5.10 (d, J = 2.8 Hz, 1H),

4.09 (unresolved ddt, 1H), 3.89 (dd, J = 2.8 and 11.6 Hz, 1H), 3.72 (dd, J = 3.6 Hz

and 11.6 Hz, 1H), 3.42 (s, 1H), 3.27 (q, J = 5.6 Hz, 2H), 2.17 (t, J = 5.6 Hz, 2H),

1.60-1.47 (m, 4H); ¹³C NMR (CDCl₃,): δ 170.1, 169.2, 148.8, 147.3, 145.8 (three C),

135.5, 134.1, 130.6, 130.0, 129.0, 128.6 (six C), 128.2, 127.8 (six C), 126.8 (two C),

126.3 (three C), 123.4 (two C), 73.4, 71.0, 63.1, 57.2, 43.4, 40.3, 28.0, 27.2.

The thus obtained N-protected conjugate (0.5 g, 0.74 mmol) was treated with a

30% TFA solution in CH₂Cl₂(2 mL) for 1h. The reaction mixture was evaporated to

dryness and the residue was triturated with Et₂O to afford conjugate 6, as its

corresponding trifluoroacetate salt. Yield: 0.38 g (95%); white solid; IR (KBr, cm^-1):

3390, 3000-2700, 1651, 1520, 840, 730; MS (ESI, 30eV): m/z 453.32 [M+Na], 431.41

[M+H]; HRMS (m/z): [M+H]⁺calcd for C₂₁H₂₇N₄O₆, 431.1931. Found, 431.1967; ¹H

NMR (d₄-MeOH): δ 8.22 (d, J = 8.8 Hz, 2H), 7.72 (d, J = 8.8 Hz, 2H), 7.52-7.47 (m,

3H), 7.41-7.36 (m, 1H), 5.13 (d, J = 4.0 Hz, 1H), 4.31 (ddd, J = 4.0, 5.6 and 6.4 Hz,

1H), 3.81 (dd, J = 5.6 Hz and 11.2 Hz, 1H), 3.67 (dd, J = 6.4 Hz and 11.2 Hz, 1H),

3.35-3.35 (2H under solvent), 3.00-2.92 (m, 2H), 1.71 (quintet, J = 6.4 Hz, 2H), 1.64

13

(quintet, J = 6.4 Hz, 2H); C NMR (d₄-MeOH): δ 172.1, 171.7, 151.9, 148.9, 137.2,

136.8, 131.5, 131.4, 129.0, 128.9, 128.8 (two C), 124.4 (two C), 72.5, 62.9, 59.0,

40.4, 39.7, 27.3, 25.6.

4.3.

(1R,2R)-3-((tert-butyldimethylsilyl)oxy)-2-(2,2-dichloroacetamido)-1-(4-

nitrophenyl)propyl acetate (15)

To a suspension of CAM (0.97 g, 3.0 mmol) and imidazole (0.25 g, 3.6 mmol) in

CH₂Cl₂(10 mL) TBDMSCl (0.54 g, 3.6 mmol) was added and the reaction mixture

was stirred at ambient temperature for 30 min. Then, Et₃N (0.84 mL, 6.0 mmol) and

Ac₂O (0.57 mL, 6.0 mmol) were added and the reaction mixture was stirred for

additional 12h. The reaction mixture was diluted with CH₂Cl₂and the organic phase

was washed once with 5% cold aq. solution NaHCO₃and twice with H₂O, dried over

18

Na₂SO₄and evaporated under vacuum to dryness. The residue was subjected to FCC

purification to give intermediate 15. Yield: 1.29 g (90%); colorless oil; R_f

(PhMe/EtOAc 95:5): 0.16; IR (thin film, CHCl₃, cm^-1): 3314, 2948, 2930, 2858, 1748,

1682, 1608, 1526, 1472, 1348, 1226, 1118, 840, 778; MS (ESI, 30eV): m/z 519.27

and 517.25 [M+K], 503.39 and 501.38 [M+Na], 481.44 and 479.40 [M+H] 421.33

and 419.44 [M-AcOH]; HRMS (m/z): [M+H]⁺calcd for C₁₉H₂₉Cl₂N₂O₆Si, 479.1172.

1

Found, 479.1161; H NMR (CDCl₃): δ 8.24 (d, J = 8.8 Hz, 2H), 7.58 (d, J = 8.8 Hz,

2H), 6.96 (d, J = 9.2 Hz, 1Η), 6.12 (d, J = 8.0 Hz, 1H), 5.91 (s, 1H), 4.34 (dddd, J =

2.4, 4.0, 8.0 and 9.2 Hz, 1H), 3.62 (dd, J = 4.0 and 10.4 Hz, 1H), 3.42 (dd, J = 2.4

13

and 10.4 Hz, 1H), 2.12 (s, 3H), 0.93 (s, 9H), 0.06 and 0.05 (two s, 6H); C NMR

(CDCl₃): δ 170.0, 164.0, 148.1, 144.2, 128.2 (two C), 123.9 (two C), 73.5, 66.3, 61.3,

55.0, 25.8 (three C), 20.9, 18.2, -5.6 and -5.5 (two C).

4.4.

(1R,2R)-2-(2,2-dichloroacetamido)-3-hydroxy-1-(4-nitrophenyl)propyl

acetate (16)

A solution of 15 (1.2 g, 2.5 mmol) in AcOH/H₂O (2:1, 12.5 mL) was stirred at

ambient temperature for 12h and then diluted with EtOAc. The organic phase was

washed once with 5% cold aq. solution NaHCO₃and twice with H₂O, dried over

Na₂SO₄and evaporated to dryness to give pure compound 16, following FCC

o

purification. Yield: 0.87 g (95%); white solid; m.p: 116-117 C; R_f(PhMe/EtOAc

1:1): 0.13; IR (KBr, cm^-1): 3461, 3272, 2916, 1743, 1684, 1520, 1348, 1237, 1059;

MS (ESI, 30eV): m/z 405.24 and 403.19 [M+K], 389.21 and 387.26 [M+Na], 307.33

and 305.32 [M-AcOH]; ¹H NMR (CDCl₃): δ 8.24 (d, J = 8.8 Hz, 2H), 7.60 (d, J = 8.8

Hz, 2H), 6.96 (d, J = 9.2 Hz, 1Η), 6.22 (d, J = 6.8 Hz, 1H), 5.90 (s, 1H), 4.37 (dddd, J

= 4.0, 4.8, 6.8 and 9.2, 1H), 3.70 (dd, J = 4.8 and 11.2 Hz, 1H), 3.55 (dd, J = 4.0 and

13

11.2 Hz, 1H), 2.17 (s, 3H); C NMR (CDCl₃): δ 170.2, 164.4, 148.0, 143.9, 127.9

(two C), 124.0 (two C), 73.0, 66.2, 61.1, 55.3, 21.0. Anal. (C₁₃H₁₄Cl₂N₂O₆) calcd: C,

42.76; H, 3.86; N, 7.67. Found: C, 42.97; H, 3.70; N, 7.49.

4.5.

(2S,3R)-3-Acetoxy-2-(2,2-dichloroacetamido)-3-(4-nitrophenyl)propanoic

acid (17)

o

To a round bottom-flask containing MeCN (16 mL) at 0 C, H₅IO₆(1.12 g, 4.93

mmol) was added and the mixture was stirred for 15 min. Then, the resulting cold

mixture was transferred to a flask containing compound 16 (0.8 g, 2.19 mmol) and

finally a catalytic amount of PCC (ca. 10 mg) was added. The resulting mixture was

19

o

stirred at 0 C for 30 min and then at ambient temperature for further 1h.

Subsequently, it was diluted with EtOAc and the organic phase was washed once

brine, once with saturated aqueous solution NaHSO₃and once again with brine, then

dried over Na₂SO₄and evaporated to dryness. Pure Acid 17 was obtained after FCC

o

purification. Yield: 0.53 g (64%); white solid, m.p. 104-106 C ; R_f(CHCl₃/MeOH

8:2): 0.14; IR (KBr, cm^-1): 3600-2800, 1732, 1694, 1606, 1520, 1376, 1350, 1234,

1046, 856, 812, 716; MS (ESI, 30eV): m/z 781.01 and 779.01 [2M+Na], 419.35 and

1

417.21 [M+K], 403.19 and 401.24 [M+Na], 359.42 and 357.28 [M-CO₂]; H NMR

(d₄-MeOD): δ 8.20 (d, J = 8.8 Hz, 2H), 7.63 (d, J = 8.8 Hz, 2H), 6.50 (d, J = 3.2 Hz,

1Η), 6.37 (s, 1H), 4.87 (d, J = 3.2 Hz, 1H), 2.18 (s, 3H); ¹³C NMR (d₄-MeOD): δ

172.0, 169.8, 164.8, 147.6, 144.9, 127.2 (two C), 123.1 (two C), 75.1, 65.9, 57.9,

19.4. Anal. (C₁₃H₁₂Cl₂N₂O₇) calcd: C, 41.18; H, 3.19; N, 7.39. Found: C, 41.25; H,

2.92; N, 7.12.

4.6. N¹-((2S,3R)-3-Acetoxy-2-(2,2-dichloroacetamido)-3-(4-nitrophenyl)

propano-yl)-N⁴,N⁹,N¹²-tri(tert-butoxycarbonyl)spermine (18)

To an ice-cold solution of acid 17 (0.38 g, 1.0 mmol), PA derivative 11 (0.60 g, 1.2

mmol), HOBt (0.16 g, 1.2 mmol) and Et₃N (0.15 mL, 1.1 mmol) in THF (1.5 mL),

EDCI (0.38 g, 2.0 mmol) was added portion-wise over 1.5 h. Upon addition, the

resulting mixture was stirred at ambient temperature for 12h. Then, it was evaporated

under vacuo and the residue was partitioned between EtOAc and a 5% aq. solution of

citric acid. The organic phase was washed with a 5% aq. solution NaHCO₃and brine,

dried over Na₂SO₄and evaporated to dryness. Compound 18 was obtained pure

following FCC purification. Yield: 0.52 g (60%); colorless oil; R_f(PhMe/EtOAc

35:65): 0.23; MS (ESI, 30eV): m/z 903.11 and 901.16 [M+K], 887.25 and 885.24

[M+Na], 865.31 and 863.27 [M+H], 765.28 and 763.27 [(M-Boc)+H]; HRMS (m/z):

[M+H]⁺calcd for C₃₈H₆₁Cl₂N₆O₁₂, 863.3725. Found, 863.3764; ¹H NMR (d₄-MeOH):

δ 8.29-8.24 (m, 1H), 8.22 (d, J = 8.4 Hz, 2H), 7.64 (d, J = 8.4 Hz, 2H), 7.41-7.37 (m,

1H), 7.32-7.26 (m, 1H), 6.34 (s, 1Η), 6.30 (unresolved d, 1H), 4.90 (unresolved dd,

1H), 3.26-3.06 (two m, 10H), 3.03 (t, J = 6.8 Hz, 2H), 2.13 (s, 3H), 1.68 (unresolved

quintet, 2H), 1.51-1.40 (m, 33H); ¹³C NMR (d₄-MeOH): δ 168.1, 166.2, 166.1, 155.5,

154.6, 154.5, 146.4, 143.0, 126.2 (two C), 121.7 (two C), 78.1, 78.0, 77.0, 72.5, 64.3,

64.0, 46.5 (under the solvent), 45.3, 44.4, 42.1, 36.4, 36.0, 25.9 (six C), 25.8 (three),

25.6, 24.3, 24.0, 23.6, 17.8.

20

4.7. (2S,3R)-N-(3-(4-(3-aminopropylamino)butylamino)propyl)-2-(2,2-dichloro-

acetamido)-3-hydroxy-3-(4-nitrophenyl)propanamide (7)

To a solution of intermediate 18 (0.43 g, 0.5 mmol) in MeOH/H₂O (5.0 mL, 9:1),

K₂CO₃(0.14 g, 1.0 mmol) was added. The reaction mixture was refluxed for 2 h and

then evaporated to a minimum volume, taken up in CHCl₃and washed twice with

H₂O. The organic layer was dried over Na₂SO₄and evaporated to dryness. The

deacylated intermediate was purified by FCC. Yield: 0.33 g (80%); colorless oil; R_f

(PhMe/EtOAc 35:65): 0.21; MS (ESI, 30eV): m/z 861.25 and 859.24 [M+K], 845.23

and 843.11 [M+Na], 823.28 and 821.28 [M+H]; HRMS (m/z): [M+H]⁺calcd for

C₃₆H₅₉Cl₂N₆O₁₁, 821.3619. Found, 821.3668.

To an ice-cold solution of the deacetylated compound (0.30 g, 0.37 mmol) in

CH₂Cl₂(1 mL), TFA (1 mL) was added. The reaction mixture was stirred at ambient

temperature for 1h, evaporated to dryness, triturated with Et₂O and refrigerated. The

supernatant solution was decanted to afford conjugate 7 as the corresponding tris-

trifluoroacate salt. Yield: 0.29 g (90%); colorless oil; MS (ESI, 30eV): m/z 545.41 and

543.32 [M+Na], 523.30 and 521.28 [M+H]; HRMS (m/z): [M+H]⁺calcd for

C₂₁H₃₅Cl₂N₆O₅, 521.2046. Found, 521.2089; ¹H NMR (d₆-DMSO): δ 8.86 (br. s, 2H),

8.74 (d, J = 8.8 Hz, 1H), 8.67 (br.s, 2H), 8.20 (d, J = 8.8 Hz, 2H), 7.99 (br.s, 3H), 7.65

(d, J = 8.8 Hz, 2H), 7.62 (t, J = 8.8 Hz, 1H), 6.60 (s, 1H), 5.23 (d, J = 3.0 Hz, 1H),

4.51 (dd, J = 3.0 and 8.8 Hz, 1H), 3.03-2.81 (two m, 12H), 1.90 (quintet, J = 8.0 Hz,

2H), 1.73 (unresolved quintet, 2H), 1.67-1.56 (m, 4H); ¹³C NMR (d₆-DMSO): δ

-

168.9, 164.1, 158.9 (q, J = 32.3 Hz, CF₃CO₂), 150.2, 147.1, 128.1 (two C), 123.4

-

(two C), 117.2 (q, J = 295.5 Hz, CF₃CO₂), 72.3, 66.8, 59.2, 46.6, 46.5, 45.0, 44.3,

36.6, 36.4, 26.3, 24.2, 23.1, 23.0.

4.8. N-((1R,2R)-1,3-Dihydroxy-1-(4-nitrophenyl)propan-2-yl)-1H-1,2,4-triazole-3-

carboxamide (8)

To an ice-cold solution of CLB (0.21 g, 1.0 mmol) in DMF (1.2 mL), TCA (0.14 g,

i

1.2 mmol), Pr₂NEt (0.52 mL, 3.0 mmol) and PyBrOP (0.67 g, 1.44 mmol) were

added sequentially. The reaction mixture was stirred at ambient temperature for 2 h

and then diluted with EtOAc. The organic phase was washed twice with a 5% aq.

solution of NaHCO₃and thrice with H₂O, dried over Na₂SO₄and evaporated to

dryness to afford an oily residue. Conjugate 8 was purified with FCC. Yield: 0.25 g

(80%); white foam; R_f(CHCl₃/MeOH 8:2): 0.26; IR (KBr, cm^-1): 3122, 3028, 2927,

21

2840, 1655, 1541, 1426, 1334, 1256, 1062, 807, 739,; MS (ESI, 30eV): m/z 637.19

[2M+Na], 346.33 [M+K], 330.39 [M+Na], 308.41 [M+H]; HRMS (m/z): [M+H]⁺

1

o

calcd for C₁₂H₁₄N₅O₅, 308.0995. Found, 308.0948; H NMR (d₆-DMSO, 40 C): δ

8.40 (s, 1H), 8.13 (d, J = 8.8 Hz, 2H), 7.68 (d, J = 9.2 Hz, 1H), 7.62 (d, J = 8.8 Hz,

2H), 7.61 (s, 1H), 7.00 and 6.99 (two s, 2H), 5.13 (d, J = 2.4 Hz, 1Η), 4.22-4.15 (ddt,

J = 2.4, 7.6 and 9.2 Hz, 1H), 3.66 (1H, dd, J = 7.6 and 10.8 Hz), 3.52 (1H, dd, J = 7.6

and 10.8 Hz); ¹³C NMR (d₆-DMSO, δ): 172.6, 158.1, 152.0, 146.9, 135.6, 127.6 (2C),

123.5 (2C), 69.4, 61.0, 56.7.

4.9.

N-(3-((1R,2R)-1,3-Dihydroxy-1-(4-nitrophenyl)propan-2-ylamino)-3-oxo-

propyl)-1H-1,2,4-triazole-3-carboxamide (9)

This compound was obtained through a three-step sequence, as described below:

4.9.1. N-Acylation of CLB using ‘active’ ester Trt-β-Ala-OSu

To an ice-cold solution of CLB (0.21 g, 1.0 mmol) in DMF (1.2 mL), Trt-β-Ala-

OSu (0.47 g, 1.1 mmol) and ⁱPr₂NEt (0.52 mL, 3.0 mmol) were added. The reaction

mixture was stirred at ambient temperature for 1h and then diluted with EtOAc. The

organic phase was washed twice with a 5% aq. solution of NaHCO₃and thrice with

H₂O, dried over Na₂SO₄and evaporated to dryness to afford an oily residue. The

acylated CLB was purified with FCC. Yield: 0.47 g (89%); white foam; R_f

(CHCl₃/MeOH 97:3): 0.1; IR (KBr, cm^-1): 3416, 3226, 3056, 2926, 2880, 1670, 1616,

1516, 1344, 1262, 1082, 746, 706; MS (ESI, 30eV): m/z 564.36 [M+K], 548.36

[M+Na], 526.38 [M+H], 243.41 [Trt]; ¹H NMR (CDCl₃): δ 8.10 (d, J = 8.6 Hz, 2H),

7.51 (d, J = 8.6 Hz, 2H), 7.42-7.36, 7.30-7.24 and 7.23-7.18 (three m, 15H), 5.18 (d, J

= 3.2 Hz, 1H), 4.17 (dt, J = 3.2 and 4.8 Hz, 1H), 3.89 (1H, dd, J = 4.6 and 11.2 Hz),

13

3.82 (1H, dd, J = 4.6 and 11.2 Hz), 2.53-2.26 (m, 4H); C NMR (CDCl₃): δ 173.2,

162.6, 148.9, 147.3 (three C), 128.6 (six C), 128.1 (six C), 127.0 (two C), 126.8 (three

C), 123.5 (two C), 73.3, 71.9, 63.5, 56.0, 40.1, 36.5.

4.9.2. Removal of the Trt protecting group

The thus obtained N-protected intermediate (0.4 g, 0.76 mmol) was treated with a

30% TFA solution in CH₂Cl₂(2 mL) for 1h. The reaction mixture was evaporated to

dryness and the residue triturated with Et₂O to afford the corresponding detritylated

compound in the form of its trifluoroacetate salt. Yield: 0.29 g (95%); white foam; IR

22

(KBr, cm^-1): 3390, 3000-2700, 1676, 1522, 1458, 1202, 1135; MS (ESI, 30eV): m/z

284.42 [M+H].

4.9.3. Coupling with TCA

To an ice-cold solution of the above obtained salt (0.25 g, 0.63 mmol) in DMF (0.8

mL), ⁱPr₂NEt (0.33 mL, 1.89 mmol) was added. The reaction mixture was stirred at 0

^oC for 10 min and then TCA (0.064 g, 0.57 mmol) and PyBrOP (0.32 g, 0.68 mmol)

were added sequentially. The mixture was stirred at ambient temperature for 1h and

then it was diluted with EtOAc. The organic phase was washed once with a 5% aq.

solution of NaHCO₃and thrice with H₂O, dried over Na₂SO₄and evaporated to

dryness. Compound 9 was purified with FCC. Yield: 0.16 g (75%); white foam; R_f

(CHCl₃/MeOH 9:1): 0.12; IR (KBr, cm^-1): 3117, 3041, 2931, 1648, 1540, 1060; MS

(ESI, 30eV): m/z 417.45 [M+K], 401.24 [M+Na], 379.28 [M+H]; HRMS (m/z):

1

[M+H]⁺calcd for C₁₅H₁₉N₆O₆, 379.1366. Found, 379.1398; H NMR (d₆-DMSO): δ

9.27 (t, J = 5.2 Hz, 1H), 8.30 (s, 1H), 8.14 (d, J = 8.8 Hz, 2H), 7.68 (d, J = 9.2 Hz,

1H), 7.57 (d, J = 8.8 Hz, 2H), 5.82 (br. s, 1Η), 5.00 (1H, unresolved d), 4.84 (br. s,

1H), 4.02-3.94 (m, 1H), 3.54 (unresolved dd, 1H), 3.29 (dd, J = 6.0 and 10.4 Hz, 1H),

13

3.20 (q, J = 6.8 Hz, 2H), 2.36-2.20 (m, 2H); C NMR (d₆-DMSO): δ 170.1 (two C),

156.9, 156.3, 152.4, 146.8, 127.8 (two C), 123.2 (two C), 69.8, 60.9, 56.4, 36.4, 34.3.

4.10. Biological evaluation

4.10.1. Antibacterial assays

The antibacterial activity of CAM-PACs was assessed against CAM-sensitive S.

aureus and E. coli strains, as well against two CAM-resistant strains of E. coli lacking

chromosomal rrn alleles, but containing pKK35 plasmids possessing mutated 23S

rRNA (A2058G or A2503C), kindly offered by Prof. A.S. Mankin (University of

Illinois). E. coli ΔtolC strain BL21 DE3 with impaired a proton-dependent MDR

efflux pump causing multidrug resistance, AcrAB-TolC, offered by Dr D.N. Wilson

(University of Munich), was included in our study to test if this mechanism of

resistance affects the efficacy of CAM-PACs. Three methicillin-resistant S. aureus

(MRSA) isolates belonging to the ST80 clone that carries the staphylococcal cassette

chromosome mec type A and exhibits multi drug resistance behavior (Table S2 in

Supplementary Data section), were kindly offered by Prof. I. Spiliopoulou-Sdougkou

(National Reference Laboratory for Staphylococci, School of Medicine, University of

23

Patras, Greece). E. coli strain, Rosetta(DE3)pLysS, containing rare tRNA genes and

a chloramphenicol resistant gene in the pLysS plasmid, was purchased from Novagen.

CAM was tested as a reference compound. The in vitro antibacterial activity was

determined as previously described [28], by calculating the IC₅₀from bacterial

cultures grown in LB (Luria Broth) medium in the absence or the presence of drugs at

different concentrations. Since the tested compounds differ in molecular mass, IC₅₀

values were expressed in μM. Molecular mass of each compound was unequivocally

determined by MS spectrometry. The activity of CAM acetyltransferase (CAT) using

CAM or CAM-PACs as acceptor substrates was assayed by using enzyme from E.

coli, acetyl coenzyme A as donor substrate, and 5,5΄-dithiobis(2-nitobenzoic acid)

purchased from Sigma-Aldrich, following the manufacturer’ protocols

(http://www.sigmaaldrich.com/technical-documents/protocols/biology/enzymatic-

assay-of-chloramphenicol-acetyltransferase.html ). The V_maxand K_mvalues of CAT

were determined by fitting the substrate concentrations [S] and the obtained

ΔΑ_412nm/min (V_o) values into Equation 1,

V_o= V_max[S]/(K_m+[S])

(Equation 1)

and then by dividing the V_maxvalue with 0.0136 to convert its units in μM·min^-1.

4.10.2. Quantification of the intracellular levels of polyamines and compounds 4

and 5

Polyamines and compounds 4 and 5 in E. coli cells were determined by reverse

phase high-performance liquid chromatography (RP-HPLC-UV). For these assays,

100 mg cells (wet weight), isolated from E. coli cultures grown at 37^oC in M9

medium (48 mM Na₂HPO₄, 22 mM KH₂PO₄, 9 mM NaCl, 19 mM NH₄Cl) enriched

with 0.03 mM FeCl₃, 0.1 mM CaCl₂, 1 mM MgSO₄, 0.01 mM vitamin B₁, 0.6%

glucose and 0.1% casamino acids and supplemented with compound 4 or 5 at

concentrations equal to their IC₅₀(9.4 μΜ and 35.5 μΜ, respectively), were washed

and then homogenized in 0.6 M HClO₄. Cultures grown in the absence of compounds

4 and 5 were used as a reference.The homogenate was incubated in ice for 1h, and

then centrifuged at 12,000×g for 20 min at 4^oC. The pellet was extracted twice with 1

mL 0.6 M HClO₄and recentrifuged. The three supernatants were pooled and used to

determine the levels of polyamines and compounds 4 and 5, while the pellet was

hydrolyzed o/n in 1 ml of 0.1 M NaOH and used for protein determination [55]. Each

24

supernatant, after dansylation and extraction of the dansyl derivatives by toluene [56],

was dried, redissolved in 10 mM phosphate buffer pH 4.4 and fractionated by RP-

HPLC (Waters 600 HPLC system equipped with a tunable UV absorbance detector

Waters 486), using a C₁₈-Symmetry, 3.5-μm, 75 mm×4.6 mm column obtained from

Waters, as previously described [57]. Polyamines and N⁸,N⁸-dibenzylspermidine, a

putative degradation product of compounds 4 and 5, were identified by comparing

retention times. 1,6-Diaminohexane was used as internal standard. Verification of the

identity of each peak was further done by U.V. and mass spectrometry. The detection

limits for PUT, SPD, and N⁸,N⁸-dibenzylspermidine were 52 pmol, 18 pmol, and 16

pmol at 258 nm, respectively. For determination of the intracellularly accumulated

compounds 4 and 5, the cellular extracts were directly fractionated by RP-HPLC,

using the same C₁₈-Symmetry as above, and the fractions were analyzed by optical

scanning at 275 nm. The detection limits for compounds 4 and 5 were 25 pmol and 30

pmol at 275 nm, respectively. The coefficient of variation (C.V.) of the intra-assays

ranged from 0.05 to 0.07, while those of the inter-assays ranged from 0.13 to 0.17.

Data processing was made, using the Millenium 32 software-Waters, and the results

were expressed as μmol /g wet weight.

4.10.3. Statistical analysis

All biological and biochemical experiments were carried out at least three times

with two replicates per time, and the data were expressed as means±standard

deviation. Significant differences between mean values were measured at p < 0.05 by

the F-Scheffe test (SPSS program 20.0 for Windows).

Footnotes

Abbreviations used: Boc, tert-butoxycarbonyl; CAM, chloramphenicol; CAM-PACs,

CAM-polyamine conjugates; CLB, chloramphenicol base; EDCI, 1-ethyl-3-(3-

dimethylaminopropyl)carbodiimide hydrochloride; HBTU, O-(benzotriazol-1-yl)-

N,N,N’,N’-tetramethyluronium hexafluorophosphate; HOBt, 1-hydroxybenzotriazole;

MRSA, methicillin resistant S. aureus; PA, polyamine; PTase, peptidyl transferase;

PUT, putrescine; rRNA, ribosomal RNA; SPD, spermidine; SPM, spermine; TBDMS,

tert-butyldimethylsilyl; TCA, 1,2,4-triazole-3-carboxylic acid; Tfa, trifluoroacetyl;

TFA, trifluoroacetic acid; Trt, triphenylmethyl; WT, wild-type

25

Conflict of interest

All authors declare no conflict of interest.

Supplementary data

Supplementary data associated with this article can be found in the online version,

at http://dx.doi.org/...........

Acknowledgements

We thank Prof. Giovanni Sindona and Prof. Anna Napoli from the Department of

Chemistry of the University of Calabria (Italy) for providing the HRMS analyses, and

Athina Fidaki from the Department of Public Health, School of Medicine, University

of Patras (Greece) for her technical assistance with RP-HPLC chromatography. We

also, thank Prof. A.S. Mankin, Prof. I. Spiliopoulou-Sdougkou, and Dr D.N. Wilson

for providing us with mutated strains of E. coli and MRSA isolates. We also thank the

University of Patras for funding the above work.

Legends to Figures and Schemes

Figure 1. Structures of CAM, natural PAs and synthesized CAM-PA conjugates.

Figure 2. Rescue of the growth of S. aureus and E. coli by exogenously added

polyamines, when bacteria were dosed at a IC₅₀concentration of each CAM-PACs 2-

5. S. aureus and E. coli cells at an A₅₆₀of 0.020 were grown in M9 medium in the

absence (white bars) or the presence of each CAM-PAC, added at a concentration

equal to their respective IC₅₀(white bars with lined pattern). CAM was used as a

reference compound. Similar cultures containing exogenous putrescine (light gray

bars), spermine (dark gray bars), or spermidine (black bars) at 10×IC₅₀were also

monitored until the optical density of the control culture (grown in the absence of

drug) reached the value 0.800 at 560 nm. Values are expressed as a percentage of

optical density measured in control cultures.

*Significant different from the value measured in cells grown in the presence of each

CAM-PAC at IC₅₀(p < 0.05).

26

Scheme 1. Synthesis of CAM-PA conjugates 1-9. Reagents and reaction conditions:

o

(i) succinic anhydride, DMF, 25 C, 1.5 h; (ii) 10 or 11 or 12 or 13 or 14, HBTU,

o

ⁱPr₂NEt, 0 C, 15 min then 25 C, 1-2 h, yields for the two steps: 65-83 %; (iii)

CF₃CO₂H-CH₂Cl₂(1:1), 0 ^oC, 5 min then 25 ^oC, 1 h, 90-95%; (iv) K₂CO₃, MeOH, H_2-

O, 65 ^oC, 1 h, 59% (for 5); (v) phthalic anhydride, DMF, 25 ^oC, 10 h; (vi) 10, HBTU,

o

ⁱPr₂NEt, 0 C, 15 min then 25 C, 1 h, 85% for the two steps; (vii) TCA, PyBrOP,

o

i

ⁱPr₂NEt, DMF, 25 C, 2 h 80% (for 8), 75% (for 9); (viii) Trt-βAla-OSu, Pr₂NEt,

DMF, 25 ^oC, 1 h, 89%.

Scheme 2. Synthesis of acid 17 and CAM-PA conjugate 7. Reagents and reaction

t

o

conditions: (i) BuMe₂SiCl, imidazole, CH₂Cl₂, 25 C, 30 min, then Ac₂O, Et₃N, 25

o

^oC, 12 h, 90%; (ii) AcOH-H₂O (2:1), 25 C, 12 h, 95%; (iii) H₅IO₆, PCC (cat.),

o

MeCN, 0 C, 30 min then 25 C, 1 h, 64%; (iv) 11, EDCI, HOBt, THF, 0 C, 1.5 h,

o

then 25 C, 12 h, 60%; (v) (a) K₂CO₃, MeOH-H₂O (9:1), reflux, 2 h, 80%, (b)

CF₃CO₂H-CH₂Cl₂(1:1), 0 ^oC, 5 min then 25 ^oC, 1 h, 90%.

27

Table 1. Determination of IC₅₀for CAM and CAM-PACs, indicating the concentration of each compound needed to inhibit the growth of wild-

type and mutant S. aureus and E. coli cells by half ^a

Compound

IC₅₀(μΜ)

E. coli^b

MRSA

S. aureus^b

E. coli^b

(A2058G)

E. coli^b

(A2503C)

E. coli

(ΔtolC)

E. coli

(GRE2272)

(GRE2372)

(GRE2691)

(WT)

Rosetta(DE3)pLysS

(Cam^R)

8.0 ± 0.9

>200

6.0 ± 0.8

>200

14.5 ± 1.2

>200

3.1 ± 0.3

>200

6.2 ± 0.5

>200

15.5 ± 1.3

>200

24.7 ± 2.3

>200

2.0 ± 0.5

>200

110.0 ± 5.7

CAM

>200

1

2

3

4

5

6

7

8

9

>200

45.3 ± 5.5

12.7 ± 1.0

4.7 ± 0.5

13.7 ± 1.2

>200

>100

>200

>150

>100

7.0 ± 0.9

>100

9.0 ± 0.8

>100

10.0 ± 0.9

>200

9.4 ± 0.8

35.5 ± 3.6

>200

9.4 ± 1.0

32.3 ± 3.0

>200

9.4 ± 0.9

37.1 ± 3.1

>200

9.0 ± 1.2

33.4 ± 3.8

>200

19.0 ± 2.1

42.5 ± 5.1

>200

>100

>300

>100

66.0 ± 4.6

>200

^aData represent the mean±SD values obtained from three independently performed experiments, with two replicates per experiment. E. coli

TA531 cells lacking chromosomal rrn alleles, but containing pKK35 plasmids possessing wild-type 23S rRNA displayed the same IC₅₀value for

each drug, like those of wild-type (WT) E. coli K12 cells. ^bData in these columns have been obtained from ref. 28 and are presented here for the

sake of comparison.

28

Table 2. Protein, polyamine, and CAM-PAC content in E. coli cells grown in the

presence or absence of compounds 4 and 5^a

Parameter assayed

Cells grown in

the absence of

CAM-PACs

Cells grown in the

presence of comp. 4 presence of comp. 5

Cells grown in the

Culture optical

density (560 nm)

at the time of

harvesting

0.800 ± 0.070

0.416 ± 0.030*

0.440 ± 0.035*

Protein

(mg/g w.w.)

57.00 ± 6.35

7.60 ± 1.20

1.63 ± 0.50

13.10 ± 1.92*

1.14 ± 0.15*

0.47 ± 0.05*

1.06 ± 0.12

21.8 ± 3.00*

0.46 ± 0.08*

0.35 ± 0.03*

Putrescine

(μmol/g mg/g w.w.)

Spermidine

(μmol/g mg/g w.w.)

Compound 4

(μΜ)^b

Compound 5

3.96 ± 0.47

(μΜ)^b

^aThe cells used for these assays were grown in polyamine deficient M9 medium in the

presence of compound 4 or 5 added at a dose equal to their individual IC₅₀, and

harvested when the control culture reached 0.800 at 560 nm. The cells were then

collected by centrifugation and extracted with 0.6 M HClO₄. The supernatants were

assayed for polyamines, compound 4, and compound 5 by RP-HPLC chromatography,

while the pellets were extracted in 0.1 M NaOH and analyzed for protein.

^bThe concentrations of compounds 4 and 5 were estimated in nmol/g wet weight and

then expressed in μΜ by assuming a cell density equal to 1 g/ml.

*Significant different from the value measured in cells grown in the absence of CAM-

PACs (p < 0.05).

29

Article Doi

Synthesis and antimicrobial activity of chloramphenicol-polyamine conjugates

DOI: 10.1016/j.bmc.2015.04.069

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.bmc.2015.04.069

Full text of DOI:10.1016/j.bmc.2015.04.069

Products guided by the article

R&D Labs maybe for 334513-28-9

Relevant to this article

Hot Product