SCOPE 44 - Introduction of Genetically Modified Organisms into the Environment

5.1 INTRODUCTION

Two recent observations from our laboratory are relevant to the subject of this volume. They both concern the modulation of replication of the mitochondrial genome from cytoplasmic petite mutants of yeast. In the first case, modulation is due to non-coding sequences flanking the ori sequences, in the second to an environmental factor, temperature.

The underlying questions are of a very fundamental nature. In the first case, the problem is that of the influence of sequences flanking a given DNA segment on the functional properties of the latter. Obvious effects can be expected, and are found, owing to the presence or absence of regulatory signals in the flanking sequences. A more interesting situation arises when this is not the case and the effects are presumably just due to changes in DNA structure around the sequence under consideration. This is exactly the case investigated by us, by looking at the influence of non-coding sequences that flank the ori sequences of the mitochondrial genome of yeast.

The second question concerns the effect of the environment on the genome. In evolution, the environment mainly acts as a selection agent; organisms adapt to diverse environments by being selected for. The environment can also cause, in a direct or an indirect way, chemical alterations in the genome, as exemplified by the actions of mutagens. Another way in which the environment can affect the genome is by directly inducing changes in its higher-order structure, as shown in the results to be described here.

5.2 THE MITOCHONDRIAL GENOME OF YEAST

The mitochondrial genome of wild-type, respiratory competent Saccharomyces cerevisiae is made up of 25-50 genome units (in haploid and diploid cells, respectively), which segregate into the buds at cell replication. Each one of the identical genome units (Figure 5.1) comprises genes encoding ribosomal RNAs and transfer RNAs, as well as genes encoding protein subunits of enzyme complexes associated with respiration and energy production. The latter genes contain, in a number of strains, intervening sequences, some of which contain open reading frames. All coding sequences only represent one third of the mitochondrial genome, the rest being formed by non-coding, mostly intergenic, sequences. The latter are formed by long AT spacers, only 5% in GC, and by about 200 GC clusters, about 60% in GC (de Zamaroczy and Bernardi, 1985, 1986a, 1986b, 1987).

As intergenic sequences are highly repetitive, they confer an extreme instability to the mitochondrial genome of yeast, which is dispensable (yeast being able to survive through fermentation). Indeed, excisions of DNA segments from the mitochondrial genome occur very frequently at pairs of direct repeats essentially located in intergenic sequences (de Zamaroczy et al., 1983). The excised segment is amplified in tandem to become the repeat unit of the defective genome of the cytoplasmic respiratory deficient petite mutants. These defective genomes segregate into the buds and may undergo secondary excisions (Figure 5.2). Replication of spontaneous petite mutants is usually insured by the presence in its repeat units of one of the three to four functional ori sequences (de Zamaroczy et al., 1984; Baldacci et al., 1984) from the wild-type genome (Figure 5.3). Such ori sequences are about 300 bp long and comprise several features : (1) two GC clusters, A and B, which can form with the intervening AT sequences an A-B fold (see Figure 5.5); (2) a long AT sequence l; and (3) a GC cluster, C, which is followed by a region r in which mitochondrial RNA polymerase starts the synthesis of the RNA primer of one nascent DNA chain; the RNA primer for the other nascent DNA chain is started at the other side of cluster C (Figure 5.3).

5.3 THE EFFECT OF DNA REGIONS FLANKING THE ORI SEQUENCE ON REPLICATION EFFICIENCY

If one tests the replicative efficiency of ori+ petites (namely of petites carrying an intact ori sequence in their mitochondrial genome) by crossing with either a wild-type strain (suppressivity test) or another petite, the general rule is that the mitochondrial genome made up of shorter repeat units (and containing therefore more ori sequences) will show a higher replicative efficiency, as shown by its predominance in the progeny. We have found, however, exceptions to this rule (Figure 5.4). Indeed, some petite genomes characterized by longer repeat units may replicate better than other ones carrying shorter repeat units. Since comparisons were made among petite genomes carrying the same ori sequences, the only difference was the presence or absence of a few dozens of nucleotides in the regions flanking the ori sequences. Such regions are made up by AT spacers and GC clusters, may be located upstream or downstream of ori sequences, are different in length and primary structure, and therefore unlikely to carry any specific sequence signal.

Figure 5.1 Physical map of a long mitochondrial genome unit of wild-type S. cerevisiae. Black areas correspond to mitochondrial genes or their exons (the last exon block of oxi3 comprises exons A 6-8 and introns a1 7,8); dotted areas to intervening sequences and intergenic open reading frames (ORF1-ORF5; Colin et al., 1985); radial lines indicate tRNA genes (the thr1 gene is the only one to have an anticlockwise orientation). White areas correspond to long AT spacers with embedded short GC clusters. Among mitochondrial genes, oxi1, 2, and 3 encode subunits II, III, and I, respectively, of cytochrome c oxidase; cob, cytochrome b; aap1, oli2, and oli1, subunits 8,6, and 9 of ATPase; var1, a protein associated with the small mitochondrial ribosome subunit; 9S corresponds to the central part of tRNA synthesis locus; 15S and 21S are the genes for the small and large ribosomal RNAs, respectively. Triangles indicate the location of ori sequences 1-8; they point in the direction cluster C to cluster A.

The localization of the repeat units from the mitochondrial genomes of petite mutants used in the present work on the physical map of the mitochondrial genome of wt cells is shown. When several petite genomes correspond to the same localization, they are of about equal sizes. In the case of petite b, the dotted line corresponds to a central deletion; in that of petite hp5, excision sequences were not determined. Some restriction sites are indicated.

(From Rayko et al., 1988.)

Figure 5.2 Scheme of the mechanism of formation of the defective mitochondrial genomes from spontaneous petite mutants. A mitochondrial genome unit from wildtype cells undergoes an excision. The excised segment is tandemly amplified to form a mitochondrial genome unit of the petite mutant. This genome replicates and segregates into the buds. It may also undergo secondary excisions (from Bernardi, 1983).

The conclusion of these investigations (Rayko et al., 1988) is that noncoding sequences flanking the ori sequences may affect the efficiency of replication in all likelihood by altering the higher-order structures of the mitochondrial genome. This conclusion appears to be of general interest, since it does not seem to be limited to mitochondrial ori sequences of yeast. Indeed, the deletions in a 500 bp region of plasmid PT181 (a 4.4 Kb multicopy plasmid of Staphylococcus aureus) external to the minimal replicon decreased the ability of the plasmid to compete with a coexisting incompatible plasmid, whereas in the homoplasmic state the deletions affected neither copy number nor plasmid stability (Gennaro and Novick, 1986). These deletions appeared to affect the interaction of RepC, a transacting initiator protein which is rate limiting for replication, and the plasmid origin of replication.

Figure 5.3 The repeat unit of the mitochondrial genome of primary petite a1/1R/1 (see Gaillard et al., 1980) is 884 bp in size and contains an ori sequence, ori1. This comprises three GC clusters, A, B and C (heavy overlines). Cluster B is separated from cluster A by two short AT sequences P and s, and from cluster C by a long AT sequence l (broken overline). Cluster C is followed by sequence r (light overline) which comprises, on the strand shown, a nonanucleotide (heavy underline) where transcription of the RNA primer of one nascent DNA chain starts (arrow pointing to the left); transcription of the RNA primer for the other nascent DNA chain starts before cluster C (arrow pointing to the right) (Baldacci et al., 1984). Sequence r is followed by a sequence ra which might pair with sequence la (within the l sequence) to form a hypothetical tertiary structure (de Zamaroczy et al., 1984). In the case of ori1, sequence r is preceded by a duplication, sequence r'. Excision sequences () of the secondary mitochondrial genomes from petites 14, 26, and Z1 (see Gaillard et al., 1980; de Zamaroczy et al., 1984) are indicated by boxes (From Goursot et al., 1988).

Figure 5.4 Repeat units of petites harbouring mitichondrial genomes carrying identical on sequences and exhibiting suppressivities and/or competitive abilities in petite petite crosses which do not follow the repeat unit size rule (see Discussion). Repeat unit sizes and suppressivities are indicated. Open boxes indicate active ori sequences 2,3, and 5, blackened boxes indicate inactive ori sequences 4 and 7 (de Zamaroczy et al, 1984). The on sequences are all oriented with cluster C to the right of cluster A. Flanking genetic markers are indicated (see also Figure 5.1). Vertical bars on the left indicate the pairs of petite repeat units which are compared and discussed. (From Rayko et al., 1988.)

Concerning the mechanism by which the extra sequences favour replication, it should be pointed out that in different cases the extra sequences are different in primary structure, size, and position relative to the on sequences (Figure 5.4). This makes it unlikely that replication is favoured by specific 'replication enhancement' sequences, and rather suggests that the effect is due to differences in the DNA or nucleoid (see, for instance, Rickwood et al., 1981) structure in the different petite genomes under consideration. Such differences can be visualized as differences in DNA bending, DNA superhelicity, and/or DNA-protein interactions which may modify the secondary and tertiary structure of the ori sequences themselves (see de Zamaroczy et al.,1984) and, consequently, the initiation of DNA replication; obviously effects on the elongation rate of newly synthesized DNA strands are also conceivable.

Since the flanking sequences of ori sequences in the petite genomes under consideration are made up of AT spacers and GC clusters, our results provide evidence that intergenic non-coding sequences play a role in the modulation of an essential genome function, such as replication, and substantiate the general idea, supported by other lines of evidence (Bernardi, 1982, 1983; Bernardi and Bernardi, 1986; de Zamaroczy and Bernardi, 1986a, 1987), that non-coding sequences do play a physiological role in genome function.

As a final remark, it should be noted that effects similar to those described here might be responsible for at least some of the cases in which recombinant plasmids carrying particular inserts fail to replicate.

5.4 THE EFFECT OF TEMPERATURE ON REPLICATION EFFICIENCY

The second observation concerns the effects of temperature on the efficiency of replication in petite genomes that have a thermosensitive structure in their ori sequences. Two petites from our collection, a1/1R/14 and a1/1R/26 (called 14 and 26 henceforth), are made up of repeat units containing ori sequences which are partially deleted in their cluster A and also in the neighbouring sequences. As a consequence, these on sequences lack the A-B fold which requires an intact A cluster. They can form, however, replacement folds which are only made by AT base pairs (Figure 5.5). The lower thermal stability of the replacement fold compared to the A-B fold suggests that temperatures within a physiological range for yeast may affect its structure. We have, therefore, tested the suppressivity of these petites and compared them with that of a closely related petite which comprised, however, an intact ori sequence, a1/1R/Z1 (or Z1). As shown in Table 5.1, the petites containing the defective ori sequences showed a strong temperature dependence (in the 23-33°C range) of its replication efficiency as tested by suppressivity, whereas the control petite showed essentially no change under the same conditions.

The results presented (Goursot et al., 1988) indicate that an environmental factor, temperature, can reversibly affect the replicative ability of a genome by altering its secondary (and, possibly, its tertiary) structure. Indeed, (1) these changes cannot be ascribed to enzymes involved in DNA replication, as in temperature-sensitive mutants, since the petites discussed here are isonuclear and lack mitochondrial protein synthesis, like all petites; (2) the different effects of temperature on the replicative ability of petites Z1, 14, and 26 show an excellent correlation with those expected from the secondary structures of the postulated A-B fold and replacement folds (de Zamaroczy et al., 1984), an effect on tertiary structures being also possible. 

Figure 5.5 Potential secondary structure of the 'A-B fold' of the on sequence present in the repeat units of mitochondrial genomes from petites a1/1R/1 and Z1 (ori1) and of the 'replacement folds' that can be formed in the oril sequence present in the mitochondrial genomes of petites 14 and 26. In the case of petite 14, the residual nucleotides from the partially deleted p stretch (p) can generate a hairpin structure within nucleotides from the preceding repeat unit, but the stem, only formed by 13 AT nucleotides, carries different terminal and side loops compared to the A-B fold. In the case of petite 26, the upper part of the stem and terminal loop are identical to those of the A-B fold, but the lower part is replaced by three A:T pairs (three nucleotides are derived from the preceding repeat unit). The sequences involved in the structure shown (GC clusters A and B, sequences p and s) are those indicated in Figure 5.1. (From Goursot et al., 1988.)

Table 5.1 Suppressivity of petites 14, 26, and Z1 at different temperatures

Interestingly, the conclusion that DNA secondary structure is required for ori activity in vivo has also been very recently reached for bacteriophage G4, where a strong temperature-dependent impairment of replication was found after introducing by site-directed mutagenesis point mutations that destabilize intrastrand base-pairing in the ori sequence (Lambert et al., 1987). In the latter case, hairpin formation concerns single-stranded and not double-stranded DNA. The possibility should be left open, therefore, that hairpin formation in the petite genomes discussed here may be affected in the single-stranded DNA made during DNA polymerase action. This would not change, however, the basic conclusion that temperature differentially and reversibly affects secondary DNA structures as present during the replication of the mitochondrial genomes of petites 14 and 26.

From a general viewpoint, these results indicate the existence of a novel type of environment-genome interaction, in which reversible changes in higher-order DNA structures are induced with profound consequences on a basic genome function, such as replication. These changes concern genome transconformations which, although non-inheritable, can be maintained for many generations in the presence of the appropriate environmental condition. Genome transconformations can provide, therefore, strong selective advantages or disadvantages and play an important role in evolution, independently of classical mutations, which involve changes in the primary structure of DNA.

It should also be noted that genome transconformations of the type just described are likely to be found in organelle genomes from other poikilothermic organisms and also in prokaryotic genomes, which have a similar nucleoprotein organization and can replicate at widely different temperatures. Moreover, similar phenomena might (1) be induced by other environmental factors; (2) affect other genome functions (e.g. transcription); and also (3) be operative in other organisms.

REFERENCES

Bernardi, G. (1982) Evolutionary origin and the biological function of non-coding sequences in the mitochondrial genome of yeast. In: Slonimski, P.P., Borst, P. and Attardi, G. (Eds.) Mitochondrial Genes, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, pp. 269-78.

Bernardi, G. and Bernardi, G. (1986) Compositional constraints and genome evolution. J. Mol. Biol. 24, 1-11.

de Zamaroczy, M. and Bernardi, G. (1985) Sequence organization of the mitochondrial genome of yeasta review. Gene 37, 1-17.

de Zamaroczy, M. and Bernardi, G. (1986b) The primary structure of the mitochondrial genome of Saccharomyces cerevisiaea review. Gene 47, 155-77.

de Zamaroczy, M., Faugeon-Fonty, G. and Bernardi, G. (1983) Excision sequences in the mitochondrial genome of yeast. Gene 21, 193-202.

Gaillard, C., Strauss, F. and Bernardi, G. (1980) Excision sequences in the mitochondrial genome of yeast. Nature 283, 218-20.

Goursot, R., Goursot, R. and Bernardi, G. (1988) Temperature can reversibly modify the structure and the functional efficiency of ori sequences of the mitochondrial genome from yeast. Gene, 69, 141-5.

Rayko, E., Goursot, R., Cherif-Zahar, B., Melis, R. and Bernardi, G. (1988) Regions flanking on sequènces affect the replication efficiency of the mitochondrial genome of ori+ petite mutants from yeast. Gene 63, 213-26.

Rickwood, D., Chambers, J.A.A. and Barat, M. (1981) Isolation and preliminary characterization of DNA-protein complexes from the mitochondria of Saccharomyces cerevisiae. Exp. Cell. Res. 133, 1-13.