The isolation of analogue-resistant mutants has already been discussed in the field of primary metabolism, where the rationale was that a mutant resistant to the inhibitory effects of an analogue, which mimics the control characteristics of the natural metabolite, may overproduce the natural metabolite. This approach has been adopted, or may be adopted, in a number of guises in the field of secondary metabolism.
(i) Mutants may be isolated which are resistant to the analogues of primary metabolic precursors of the secondary metabolite, thus increasing the availability of the precursor.
(ii) Mutants may be isolated which are resistant to the feedback effects of the secondary metabolite.
(iii) Mutants may be selected which are resistant to the toxic effects of the secondary metabolite when added to the trophophase of the producing organism.
(iv) Mutants may be isolated which are resistant to the toxic effects of a compound due to the production of the secondary metabolite.
(0 Mutants resistant to the analogues of primary metabolic precursors of the secondary metabolite
This approach has been adopted by Elander et al. (1971) in the isolation of mutants of Pseudomonas aureofaciens overproducing the antibiotic pyrrolnitrin. Tryptophan is a precursor of pyrrolnitrin and although it is stimulatory to production it is uneconomic to use as an additive in an industrial process. Thus, Elander et al (1971) isolated mutants resistant to tryptophan analogues using the gradient plate technique described previously. A strain was eventually isolated which produced two to three times more antibiotic than the parent and was resistant to feedback inhibition by tryptophan. Addition of tryptophan to the improved strain would not result in higher pyrrolnitrin synthesis, indicating that tryptophan supply was no longer the limiting factor and the organism was producing sufficient endogenous tryptophan for antibiotic synthesis.
The role of tryptophan in the control of the candi-cidin fermentation has been investigated by Martin and co-workers (Martin, 1978). Candicidin is a polyene macrolide antibiotic containing an aromatic p-aminoacetophenone moiety derived from chorismic acid, a precursor of the aromatic amino acids, so that tryptophan and candicidin may be considered to be end products of a branched biosynthetic pathway. Tryptophan has been demonstrated to inhibit the biosynthesis of the antibiotic which led Martin et al. (1979) to isolate mutants resistant to tryptophan analogues. Such mutants produced more candicidin than the parent strain, apparently due to the removal of tryptophan control
Godfrey (1973) isolated mutants of Streptomyces lip-manii resistant to the valine analogue, trifluoroleucine. These mutants produced higher levels of cephamycin than the parent strain, and appeared to be deregulated for the isoleucine, leucine, valine biosynthetic pathway, indicating that valine may have been a rate-limiting step in cephamycin synthesis.
Methionine has been demonstrated to stimulate the biosynthesis of cephalosporin by Acremonium chryso-genum and superior producers have been isolated in the form of methionine analogue-resistant mutants (Chang and Elander, 1979). Lysine analogue resistant mutants have yielded a greater frequency of superior /3-lactam antibiotic producers compared with random selection (Elander, 1989).
(ii) Mutants resistant to the feedback effects of the secondary metabolite
There are many cases in the literature of a secondary metabolite preventing its own synthesis. Martin
(1978) cited the following examples: chloramphenicol, aurodox, cycloheximide, staphylomycin, ristomycin, puromycin, fungicidin, candihexin, mycophenolic acid and penicillin. The precise mechanisms of these controls is not clear but they appear to be specific against the synthesis of the secondary metabolite and not against the general metabolism of the cells. The mechanism of the control of its own synthesis by chloramphenicol appears to be the repression of arylamine synthetase (the first enzyme in the pathway from chorismic acid to chloramphenicol) by chloramphenicol. Jones and Vining (1976) demonstrated that arylamine synthetase was fully repressed by 100 mg dm"3 chloramphenicol, a level of antibiotic which neither affected cell growth nor the activities of the other enzymes of the chloramphenicol pathway.
Martin (1978) quoted several examples of correlations between the level of secondary metabolite accumulation and the level which causes 'feedback' control, which may imply that the factor limiting the yields of some secondary metabolites is the feedback inhibition by the end product. The selection of mutants resistant to feedback inhibition by a secondary metabolite is a far more difficult task than the isolation of strains resistant to primary metabolic control. It is extremely unlikely that a toxic analogue of the secondary metabolite could be found where the toxicity lay in the mimicking of the feedback control of the secondary metabolite, a compound which would not be necessary for growth. However, the detection of mutants resistant to feedback inhibition by antibiotics may be achieved by the use of solidified media screening techniques, similar to the miniaturized screening techniques previously described. The technique would involve culturing the survivors of a mutation treatment on solidified medium containing hitherto repressing levels of the antibiotic and detecting improved producers by overlaying the colonies with an indicator organism. The difficulty inherent in this technique is that the incorporated antibiotic, itself, will inhibit the development of the indicator organism. This problem may be overcome by adjusting the depth of the overlay or the concentration of the indicator such that an inhibition zone could be produced only by a level of antibiotic greater than that incorporated in the original medium. However, it is unlikely that this would be a satisfactory solution for the selection of a high-producing commercial strain. Another approach would be to utilize an analogue of the antibiotic which mimicked the feedback control by the natural product but which did not have antimicrobial properties. Inhibition of antibiotic synthesis by analogues has been demonstrated in the cases of aurodox (Liu et al., 1972) and penicillin (Gordee and Day, 1972). An alternative may be to use a mutant indicator bacterium for the overlay which is only sensitive to levels of the antibiotic in excess of that originally incorporated into the medium. In cases where it has been demonstrated that feedback control by the end product plays an important role in limiting productivity it would probably be worthwhile to design such procedures.
(iii) The isolation of mutants resistant to the toxic effects of the secondary metabolite in the trophophase
It has been demonstrated for many secondary metabolite producing organisms that the secondary metabolite is toxic to the producing cell when it is present in the trophophase (growth phase) (Demain, 1974). Thus, it appears that a 'switch' in the metabolism of the organism in the idiophase enables it to produce an otherwise 'autotoxic' product. Furthermore, it has been demonstrated that, in some cases, the higher the resistance to the secondary metabolite in the growth phase the higher is productivity in the production phase. Dolezilova et al. (1965) demonstrated that the level of production of nystatin by various strains of S. noursei was related to the resistance of the strain to the antibiotic in the growth phase; a non-producing mutant was inhibited by 20 units cm~3, the parent strain produced 6000 units cm~3 and was inhibited by 2000 units cmT3 in the growth phase and a mutant producing 15,000 units crrr 3 was found to be resistant to 20,000 units cm-3.
The possible relationship between antibiotic resistance and productivity may be used to advantage in the selection of high-producing mutants by culturing the survivors of a mutation treatment in the presence of a high level of the end product. Those strains capable of growth in the presence of a high level of the antibiotic may also be capable of high productivity in the idiophase. This approach has been used successfully for antifungal agesterols (Bu'Lock, 1980), streptomycin (Woodruff, 1966) and ristomycin (Trenina and Trut-neva, 1966) but without success for novobiocin (Hoeksema and Smith, 1961).
A similar rationale was used by McGuire et al. (1980) for the improvement of daunorubicin production, an anthracycline antitumour agent synthesized by the red pigmented streptomycete Streptomyces peuceti-cus. The red pigment is presumably anthracycline which shares its oligoketide origin with daunorubicin. The directed selection was based on the ability of the antibiotic cerulenin to suppress oligoketide synthesis. This was indicated in S. peuceticus by the lack of the red pigment on cerulenin agar. Thus, mutants wh were still capable of oligoketide synthesis in the pr ence of the inhibitor would remain red. Approximate" a third of the resistant mutants were superior daum)^ bicin producers. ru"
(iv) The isolation of mutants in which secondary metabolit synthesis gives resistance to toxic compounds '6
A potentially toxic compound may be rendered harmless by an organism converting it to a secondary metabolite or a secondary metabolite complexing with it. Ions of heavy metals such as Hg2+, Cu2+ and related organometallic ions are known to complex with /3-lactam antibiotics (Elander and Vournakis, 1986) and such agents have been used to select resistant mutants The logic of this selection process is that overproduction of a /3-lactam may be the mechanism for increased resistance to the heavy metal. Elander and Vournakis (1986) reported that the frequency of superior cephalosporin C producers was greater amongst mutants resistant to mercuric chloride than amongst random samples of the survivors of ultraviolet treatment.
The conversion of the penicillin precursor, phenyl-acetic acid, to penicillin is thought to be an example of detoxification by conversion to a secondary metabolite. It appears that strains capable of withstanding higher concentrations of the precursor may be able to synthesize higher levels of the end product. Polya and Nyiri (1966) applied this hypothesis in selecting phenylacetic acid-resistant mutants of P. chrysogenum and demonstrated that 7% of the isolates showed enhanced penicillin production. Barrios-Gonzalez et al. (1993) investigated the same phenomenon and developed methods for the enrichment of both spores and early idiophase mycelium resistant to phenyl acetic acid. Of the resistant spore population, 16.7% were superior penicillin producers as compared with 50% of the resistant idiophase population. This suggests that the selective force may be more 'directed' by using a population already committed to penicillin synthesis. Although the best mutants contained elevated levels of acyltransferase (the enzyme that directly detoxifies PAA) they also showed higher levels of isopenicillin N synthetase (cyclase) which the authors claimed to be the limiting step. Thus, it was claimed that the screening method was not specific to selecting for acyltransferase elevation but could select for strains having a faster carbon flow through the pathway enabling them to use PAA faster.
Ball (1978) stressed that the major difficulty in the selection of toxic precursor resistant mutants is that the site of resistance of a mutant may not result in the organism overproducing the end product — for example, the resistance may be due to an alteration in the permeability of the mutant or due to the mutants' ability to degrade the precursor to a harmless metabolite unrelated to the desired end product. Barrios-Gonzalez's approach of using mycelial fragments already committed to penicillin synthesis decreases the likelihood of such 'false selection' occurring.
Was this article helpful?