As mentioned before, the semi-synthetic antibiotics were among the first industrial products to experience the benefits of biocatalysis. The historic background in fermentation in
CEPHALEXIN (>10 steps)
Figure 19.5 Integrated, semi-synthetic route for Cephalexin (DSM); note stoichiometric chemistry in all steps except sugar-based Penicillin G fermentation many companies together with the drivers presented in the previous section laid the basis for this early involvement. The relatively simple hydrolytic steps in which the natural side chains were removed from the P-lactam nucleus were the first to be done enzymatically, employing well-known amide acylases such as derived from E. Coli. As early as 1985 almost all 6-APA, the P-lactam nucleus for the semi-synthetic penicillins Ampicillin and Amoxicillin, was produced from Pen. G with biocatalysis. A decade was needed to bring the reverse reaction, coupling of a P-lactam nucleus such as 6-APA or 7-ADCA with a non-natural side chain, i. e. phenylglycine, to industrial practice. Cephalexin was not by accident the first representative. Product prices, market volumes and the commercial expectations at that time, 1990-2000, all just came together in these process innovations. Other cephalosporins were too small to cover the development costs whereas the large penicillins like Amoxicillin and Ampicillin were too large to take the entrepreneurial risk. A broad collaboration with the Dutch universities, well stimulated by the Dutch government, eventually provided a strong technological base for successful enzymatic coupling of 7-ADCA with D-phenylglycine amide at a few hundreds ton scale at the DSM site in Barcelona in 1997. Presently, industrially reliable enzyme catalyzed coupling procedures are available for all P-lactam nuclei with all commercially relevant synthetic side chains. Biocatalysis also made its way into the synthesis of the required side chains. Figure 19.6 summarizes the present (bio)synthesis route for Cephalexin, including the biosynthesis of
7-ADCA to be discussed below. After a few years of practice the biocatalytic process met all expectations in terms of economic and environmental advantages (discussed in more detail in Section 19.3.3). Moreover, product quality improved with respect to the level and number of impurities, crystal form and reproducibility.
A more deep-seated process change was underway in the meantime: the biosynthesis of the P-lactam nucleus 7-ADCA. The very elegant multi-step process developed by Gist Brocades in the 1970s lost competitive power because of its completely stoichiometric character demanding costly and laborious recycling of solvents and reagents or unacceptable environmental consequences. Genetic engineering of P. Chrysogenum, the well – known penicillin-producing mould, allowed ring expansion of the five-membered thiazoline ring in the penicillin skeleton to the six-membered ring of the desired cephalosporin moiety inside the cells of the micro-organism. The process performs best when the traditional side chain precursor, phenylacetic acid, is replaced by adipic acid. The latter can readily be removed by enzymatic hydrolysis employing a glutaryl-related enzyme. Product purification and isolation proved to be quite different from the experiences in the chemical process, eventually resulting in similar advantages in economy, ecology and quality as found with the enzymatically prepared Cephalexin. Industrial production of 7-ADCA through this bio-route was started successfully in 2000 at the DSM Gist site in Delft and is now running at full scale.
The remarkably short period of ca. 10 years needed to replace the traditional chemical routes by bio-based processes poses the question of the next process changes. Present state of the art in metabolic pathway engineering allows for bio-based synthesis of all intermediates, nuclei and side chains alike, of all large-scale commercialized lactam antibiotics. Even full biosyntheses of end products like Ampicillin or Amoxicillin are no longer utopistic. Whether these processes will see industrial reality is quite a different question. Product life cycles and market potentials must be large enough to earn back the substantial development costs, which is not an easy task given the already very low prices of the bulk end products. In fact, only full biosynthesis of complete end products meeting the efficiencies of the well-proven biosynthesis of natural Pen. G would promise sufficient economic and environmental advantages to allow for all development costs and investments. A more likely development could be a further integration of process steps on a single production site. Today, the process shown in Figure 19.6 is operated at three different sites: Delft for the 7-ADCA nucleus, Geleen (NL) for the side chain and Barcelona for the enzymatic coupling.