FABRICACIÓN DEL ESPÉCIMEN 12

Having investigated the influence of plasmid-carriage on forward mutation to drug- resistance, the effects of plasmid-carriage on UV-induced mutation from histidine

auxotrophy to histidine independence (back mutation) were investigated for conçarison with previous workers.

Shaken overnight cultures grown in 25 ml of fully-supplemented DM medium were washed in 25 ml of DM salts solution by centrifugation and resuspended in 2.5 ml of DM salts solution. Reversion frequencies to histidine independence were determined by inoculating 0.1 ml of the ION concentrated cell suspensions into molten (46 ®C) 0.6% agar overlays containing all the required vitamin and amino acid supplements in excess, except for histidine which was limited. The overlays were carefully mixed and then poured onto pre-warmed DM (1.5%) agar base plates containing 0.28% glucose. Viability and post-UV survival was determined by adding 0.1 ml of diluted (10"^) cell suspensions to fully- supplemented overlays, which were also poured onto pre-warmed DM agar base plates. Plates determining viability were incubated immediately, together with those determining spontaneous reversion. Plates determining post-UV survival and UV-induced mutagenesis were irradiated before incubation at 37 ®C, using a UV fluence of 1 J m'^ s '\ Histidine revertant clones were scored after 2, 4 and 6 days incubation.

The frequency of spontaneous reversion to histidine independence was expressed as the average number of histidine-independent revertant colonies per viable cell plated. UV-induced mutation frequencies were expressed as the average number of revertant colonies detected on irradiated plates per surviving fraction of cells plated.

The limiting histidine concentration allowed several cell divisions before its exhaustion. From this point onwards only mutants which had become histidine independent continued to grow and form colonies against a background lawn of histidine- requiring cells which had stopped multiplying. Mutation frequencies were based on resistant clones recovered after four days incubation, since there was no significant increase in the number of revertants recovered after that period.

> 1 0 - ' ’ > 3 p K M IO I (/) R46 r* pYD O l R391 RP4 V) V o I 1 0 “ * L_ I 10-' 10 1-7 ■a o> u I > Z) a 0 20 4 0 60 80 UV d o s e

Figure 19. The influence o f plasmid-carriage on the frequency o f reversion to histidine independence in E. coli A B l 157. (Mean o f three experiments).

From Figure 19, four out of five plasmids tested increased UV-induced mutagenesis of strain ABl 157; two of these four plasmids, R46 and its derivative pKMlOl, also increased post-UV survival, whereas the remaining two plasmids, R391 and pYDOl, both increased the UV-sensitivity of strain ABl 157. In agreement with earlier workers (Molina et W-, 1979; Finney, 1980; Upton and Finney, 1983) plasmid RF4 conferred neither significant mutator activity, nor effected the post-UV survival of strain ABl 157 (Figure 19). After none of the UV doses tested (Table 20) did the presence of RF4 increase mutation frequency to 1.5 times that of the control, plasmid-free strain, which was the cut off point used by Finney (1980) to determine mutator phenotype. The scale of the mutator-effect demonstrated by the four plasmids was varied: pKMlOl produced the greatest effect (Figure 19), whilst also offering the highest level of

protection, whereas plasmids R391 and pYDOl, which both sensitised the host, produced the lowest mutator-effect (Figure 19). R46 which conferred lower levels of UV protection than pKMlOl (Figure 8), produced intermediate UV-induced mutation frequencies (Figure

19). Only three of the four mutator plasmids demonstrated a spontaneous mutator-effect. Plasmid R391 did not appear to demonstrate significant spontaneous mutator-activity (Figure 19, and Table 20). Whilst this conflicts with previous findings under conditions selective for the isolation of nalidixic acid-resistant clones (Table 19), these results are in agreement with the absence of spontaneous mutator-activity by R391 to reversion to tryptophan independence (Magadan, 1983). The scale of the mutator-effect determined by the different plasmids varied greatly (Table 20). Once again, as seen with UV-induced mutagenesis to nalidixic acid resistance, pKMlOl exhibited superior mutator-activity, whereas R391 and pYDOl demonstrated low mutator-activity, and R46 was intermediate (Figure 19). While the actual frequencies of mutation to nalidixic acid resistance and to histidine independence might differ, one would expect the plasmid-mediated mutation- enhancing effects of the two investigations to he similar. Indeed, the effects of plasmids RP4, R46, R391 and pYDOl on UV-induced histidine reversion compare well with those to UV-induced nalidixic acid resistance in the same strain (compare Tables 19 and 20). Despite the absence of any spontaneous mutator-activity by R391 to histidine reversion, the UV-induced reversion frequencies of this strain closely resemble those obtained with strain AB1157(pYD01). In parallel with earlier results (Figure 18 and Table 19) the unequivocal uniformity of this mutator-effect (Table 20), in spite of the decrease in post- UV survival of strains carrying R391 or pYDOl, supports the constitutive expression of the gene products responsible, and tends to argue against the inducibility of these plasmid- encoded systems by UV irradiation.

In contrast, the mutator-effect determined by plasmid R46 clearly increases with UV intensity and is therefore dose-dependent (Table 20). This confirms the earlier suggestion (Results 2.9) that gene expression in R46 is not only constitutive, but also inducible. Indeed, this inducible mutator-activity is magnified in pKMlOl, the deletion- derivative of R46, where the genes responsible for the negative regulation of muc gene expression in R46 are thought to have been deleted and thus, further supports these conclusions (Table 20).

UV dose (J m-')

Increase in F produced by plasmids [F(R^)/(R')F]

RP4 pKMlOl R46 R391 pYDOl 0 1.0 16.0 4.2 1.1 1.7 10 1.3 17.9 5.0 3.0 1.5 20 0.9 19.8 5.2 2.3 1.6 40 0.9 24.3 7.1 1.4 1.2 60 0.9 25.4 8.1 2.0 1.4 80 1.1 32.2 13.6 1.8 1.8

Table 20. Increase in UV-induced mutagenesis to histidine independence produced by plasmids present in E. coli strain AB1157.

Thus, in summary a clinically-isolated nalidixic acid-resistant dvsenteriae type 1 strain from Kashmir (Panhotra et d., 1985), which was multiply drug resistant was shown to carry four plasmids. The 40 mDa R plasmid designated pYDOl conferred trimethoprim resistance and was transferred at relatively high frequency into an E. coli strain ABl 157. Transfer of the three remaining plasmids was undetected. Although pYDOl itself was proven not to confer nalidixic acid resistance, it was shown to determine the mutator phenotype. This plasmid increased the frequencies of both spontaneous and UV- induced mutation to both nalidixic acid resistance and reversion to histidine independence in strain ABl 157. UV sensitivity studies of strains carrying pYDOl revealed that the plasmid belonged to the much smaller group of mutator plasmids that sensitise rather than protect their host strain to the lethal effects of UV irradiation. The post-UV survival curves demonstrated by strain ABl 157(pYD01) closely resembled those of strain ABl 157 carrying the clinically-isolated IncJ group plasmid R391. Furthermore, the UV-induced mutator-activities determined by pYDOl to both forward mutagenesis to nalidixic acid resistance and reversion to amino acid independence paralleled those determined by R391. The established mutator-activity of these plasmids appeared to be constitutive and apparently independent of UV dose. In contrast, the broad host-range plasmid R46, which confers the more common UV protection phenotype, facilitated higher mutator-activity that was both constitutive and UV dose-dependent.

bactericidal concentrations of nalidixic acid.

3.1. Introduction.

The selection of quinolone-resistant Gram negative organisms by passage in sub- inhibitory levels of nalidixic acid, cinoxacin or norfloxacin has been well documented (Lumish and Norden, 1975; Tenney et d ., 1983). Tenney et (1983) concluded that the level of nalidixic acid resistance induced was only to concentrations of drug close to that used for the selection of mutants; selection was rapid and resistant clones were stable.

Cairns et d . (1988) proposed in a highly controversial paper that ‘directed mutations’ could occur as a specific response to an environmental challenge. This paper provoked a storm of criticism (Scientific Correspondence in Nature 336, 525-528; 1988) for two reasons. Firstly, it challenged the orthodox view that mutational events were separate from their selection; and secondly, it suggested an unknown mechanism by which the environment could influence bacteria to generate or retain mutations. More recently. Hall (1990) has presented evidence that specific mutations occur in E. coli more frequently when advantageous than when neutral. Stahl (1990) believes that this work (Hall, 1990) has not only strengthened, but also expanded the work of Caims et d . (1988).

Both Cairns et (1988) and Hall (1990) used strains of E. coli that were unable to metabolise a particular biochemical substrate, because of a genetic defect. Cairns et (1988) studied the accumulation of mutations conferring the ability to ferment lactose in an initially lactose negative population, in the absence or presence of lactose. Hall (1990) investigated the distribution of mutations conferring the ability to metabolise tryptophan. These authors found that when E. coli was exposed to either substrate, cells appeared to preferentially mutate the defective genes back to their functional state.

Hall (1990) proposed that in nature bacteria are commonly under nutritional stress, rather than growing exponentially with DNA replication as the predominant cellular metabolic activity. Thus, if the mechanism of mutation under natural conditions is to be

understood, then experiments determining mutation frequencies should be performed using bacteria grown under conditions of nutritional stress (Hall, 1990), which is precisely what both Hall (1990) and Cairns et d . (1988) did.

In a non-dividing population, DNA synthesis is principally confined to DNA repair operations (Hall, 1990), which could induce the preferential mutation by a random mechanism involving error-prone DNA repair (Stahl, 1990). If such a mutation produces a protein that allows the cell to metabolise the limiting substrate, the cell would then be able to grow, shifting DNA synthesis towards DNA replication, which would fix the new mutation within the chromosome (Stahl, 1990). Thus, the environment would appear to be tailoring the mutations produced.

In the case of mutation to quinolone resistance, incubation in nutrient broth containing a bactericidal concentration of nalidixic acid will kill sensitive cells, but allow pre-existing mutants to grow through. Also, since nalidixic acid is a potent inducer of the SOS response (Phillips et ^ ., 1987), the proportion of mutants in the treated population might increase due to error-prone repair of nalidixic acid-damaged DNA. The presence of mutator plasmids in this system should increase both the frequency at which pre­ existing resistant cells occur, and the cellular capacity for SOS repair, and thereby increase the proportion of resistant mutants even further.

In order to establish a suitable concentration of nalidixic acid to use in these experiments, the effect of increasing concentrations of nalidixic acid on the growth of E. coli ABl 157 in nutrient broth was investigated.

3.2. Determination of sub-inhibitory and inhibitory concentrations of nalidixic acid