Synthesis and antibacterial activity of new antibiotics arising from cephalosporin-monobactam coupling

New β-lactam antibiotics were obtained by coupling the cephalosporin Cefotaxime with monobactams in order to assess the possibility to enhance the cephalosporin activity through a synergistic dual-action mechanism. The activities were tested, in vitro, against a panel of selected bacteria. Preliminary results showed a light change in antibacterial activity when compared with that of the starting cephem counterpart.


Introduction
The dual-action mechanism exploited by cephalosporins coupled with other antibiotics has been described. 1,2It consists of a primary interaction between the β-lactam ring of the cephem counterpart with penicillin binding proteins or β-lactamases, that results in the release of the other antibiotic moiety in position 3, following a chemical displacement mechanism (Scheme 1).So far, with cephem-quinolones [3][4][5] (Figure 1), the dual-action mechanism has involved two active moieties inhibiting different molecular targets: i.e. cephalosporin exploiting a cell wall activity, 6,7 quinolone acting inside the cytoplasmic membrane at the DNA level. 8 To the best of our knowledge, no dual action products arising from the coupling between two β-lactam moieties have been reported, so far.In the present paper we describe the synthesis and the antibacterial activity of new compounds deriving from the above reported chemical linkage.A cephem and a monobactam moiety, both acting against bacterial cell wall targets, were used as parent compounds.
The expected dual action of the cephem-monobactam molecule and the monobactam, when released after enzymatic displacement, was based on the possible synergism resulting from a simultaneous action of these two moieties against different penicillin binding proteins (PBPs). 10he chosen β-lactam moieties have been Cefotaxime 11 (see Table 2), well known for its antibacterial activity, and monobactams 11, (E)-12 and (Z)-12.Monobactam (Z)-7 is the commercially available antibiotic drug Aztreonam ® , 12 especially used against Gram-negative aerobic organisms.The structures of unknown monobactams 11, (E)-12 and (Z)-12 were designed on the basis of the following considerations: (1) The 3-amidic and 4-methyl substituents in trans geometry mimic the structure of Aztreonam ® . 12(2) The tetrazole ring, successfully used as a β-lactam ring activating group, 13,14 was preferred over the sulfonic acid anion, present at the N-1 position of the Aztreonam ® .This neutral group, in fact, did not introduce further charges in the cephem-monobactam molecule, charges that could make difficult its penetration through the bacterial cell wall.(3) A pyridyl group at the 3-C substituent has been already used in N-(2H-tetrazol-5-yl)-azetidin-2-ones 14 obtaining monobactam with antibacterial activity.Furthermore, its introduction in the skeleton of monobactam has allowed linking the monobactam to the cephem as 3' quaternary ammonium salts.Compounds of this type have been shown to be a "third generation" antibacterial agents with excellent activity against a wide variety of Gram-positive and Gram-negative pathogens. 15
The preparation of the unknown β-lactams 5, (E)-6 and (Z)-6 , employed as 3'-cephalosporin substituents (see Table 1   The anti and the syn isomers of the methoxyiminopyridin-3-ylacetic acids (E)-10 and (Z)-10 were obtained by treating oxopyridin-3-ylacetic acid 14 (synthesized by oxidation of 3acetylpyridine 13 as reported in the literature 16 ) with O-methylhydroxylamine (Scheme 3).After esterification with diazomethane the two isomers were separated by flash chromatography and were subsequently hydrolyzed to the free acids (E)-10 and (Z)-10: The configuration was assigned on the basis of the relative rates of methyl ester hydrolysis.In fact, for a series of αalkoxyimino esters it has been demonstrated that the Z isomers (syn) hydrolyze much more slowly than the corresponding E forms (anti). 17 The coupling products (Z)-1, (Z,E)-2, (Z,Z)-2, (Z,Z)-3 and (Z)-4 are quaternary cephalosporin derivatives and were prepared by modification of the general method of Bonjouklian and Phillips (in Scheme 4 the synthesis of (Z,E)-2 is reported as a typical example). 15,19Silylation of Cefotaxime 15 in acetonitrile with N-methyl-N-(trimethylsilyl)trifluoroacetamide (MSTFA) was followed by in situ formation of the 3'-iodide derivative with trimethylsilyliodide (TMSI).Excess of trimethylsilyliodide was destroyed by addition of tetrahydrofuran.The 3'-iodide was then replaced by the pyridin-containing monobactam (E)-6 in acetonitrile.The silylated cephem was hydrolyzed by treatment with water, and concomitant detritylation of the tetrazole ring on the monobactam moiety occurred.
Cefotaxime-monobactam derivatives (Z)-1, (Z,E)-2, (Z,Z)-2, (Z,Z)-3 and (Z)-4 were tested for their in vitro activity in comparison with Cefotaxime, with compound (Z)-7 (Aztreonam) ® and with the free monobactam partners 11, (E)-12 and (Z)-12.Despite the lack of any in vitro activity of monobactams 11, (E)-12 and (Z)-12 (MIC>128µg/mL) the coupling compounds (Z)-1, (Z,E)-2, (Z,Z)-2 were tested against some selected bacteria.As a matter of fact, the inactivity in vitro of the monobactam counterpart, against the whole bacterial cell, might have been ascribed to difficulty of penetration through the bacterial cell wall.In contrast, the coupling compounds (Z)-1, (Z,E)-2, (Z,Z)-2 could have been able to exploit their potential intrinsic activity on their PBP targets, once entered the bacterial cell wall.Further studies are planned to understand whether the lack of activity of monobactam (E)-12 and (Z)-12 has to be ascribed to a lack of intrinsic activity or a difficult penetration through the Gram-negative outer membrane.Compounds (Z)-1, (Z,E)-2, (Z,Z)-2, (Z,Z)-3 showed better activity with respect to compound (Z)-4 , which is a poor antibacterial agent.However, although maintaining some antibacterial activity, compound (Z)-1 was significantly less active than Cefotaxime.By contrast, the antimicrobial activity of compounds (Z,E)-2, (Z,Z)-2 against most of bacterial strains matched that of Cefotaxime (Table 2), and compound (Z,Z)-2 was even somewhat more active than Cefotaxime against one strain of E. coli and one strain of K. pneumoniae.Since compounds (Z,E)-2, (Z,Z)-2 differ only in the geometry of the monobactam moiety's methoxy-imine, their chemophysical properties have to be very similar.Thus, if the activity of these compounds had to be ascribed only to the cephem nature without any concomitant dual action, it would be hard to understand the differences between the antibacterial activity of (Z,E)-2, (Z,Z)-2.Otherwise, stating the occurrence of the dual action mechanism, these differences can be ascribed to the release of an inactive monobactam moiety in the case of the compound (Z,E)-2 and of an active one in the case of the compound (Z,Z)-2.Since this dual-action mechanism should depend on the PBP's and/or βlactamases, it could happen that it is exploited to a different extent in different bacterial species and strains.In contrast, compound (Z,Z)-3, derived from coupling of Cefotaxime with Aztreonam (Z)-7, was less active than either Cefotaxime or Aztreonam alone.This poor activity could be caused by a difficult cell wall penetration due to the total negative charge.Furthermore, the comparison of its activity with that of Aztreonam, that is better against Gram-positive organisms but worst against Gram-negative, may suggest that compound (Z,Z)-3 maintains a cephalosporin-type spectrum of activity without significantly releasing the monobactam counterpart, as a consequence of a possible preferred primary interaction between the monobactam-ring and its target PBPs in Gram negative bacteria.These results strongly suggest that the expected dual-action mechanism is at least partially exploited.Work is in progress in order to assess the possibility of exploiting a dual action mechanism by coupling different β-lactam-antibiotics.

Experimental Section
General Procedures.All starting compounds, unless otherwise stated, were purchased.Reactions were run under an atmosphere of dry nitrogen or argon.FT-IR Spectra were recorded on a Perkin-Elmer infrared spectrometer, mass spectra at 70 eV, using the electron impact mode were obtained on Finnigan MAT GCQ instrument, NMR spectra on spectrometers Varian VXR 200, Varian Gemini 300, or Varian Mercury 400 MHz using the residual signal of the solvent as internal standard.HPLC analysis were carried out using a HP 1100 instrument, column Hibar Lichrospher 100 RP-18 (5µm), eluting with a gradient from KH 2 PO 4 buffer (0.01N, pH 3.2) to MeCN/ KH 2 PO 4 buffer 85/15.

(E) and (Z)-2-(Methoxyimino)-2-(3-pyrid-3-yl)acetic acid [(E)-10] and [(Z)-10]. O-Methyl-
hydroxylamine hydrochloride (8.3g, 98 mmol, 32.5 mL of a 25% solution in water) was added at room temperature and stirring to oxopyridin-3-ylacetic acid 14 (3.7g, 24 mmol).The pH 5 was adjusted by adding a saturated solution of NaHCO 3 .The resulting mixture was stirred over night affording a homogeneous solution (pH 5.2).1N HCl was added to adjust pH 4, and the water was removed in vacuo to afford a white solid.This product was dissolved in methanol and heated to 60 °C under stirring.The precipitate was separated by decantation and was a mixture of (E)-10 and (Z)-10, which was trearted diazomethane.Separation of the corresponding methyl esters by flash chromatography (cyclohexane/ethyl acetate 7:3) followed by hydrolysis with NaOH/MeOH, and subsequent addition of HCl (aq) to adjust pH 4.

4 Na
Products 1,2 are diasteromeric mixtures.b All products 1-4 gave 1 H-NMR and IR spectra consistent with the structure shown.

Table 1 .
Monobactam moieties and coupling products

Table 2 .
In Vitro antibacterial activity (MIC a , µg/mL) of the most representative compounds a MIC = Minimum inhibitory concentration.b gram-negatives.c gram-positives.