Results and Discussion

Results and Discussion

Figure 1 illustrates the restriction patterns of N.tabacum ct-DNA generated by eight different enzymes.It is evident that EcoRI produces the largest number of fragments among the enzymes used and therefore has the highest resolving power to uncover differences.As represented in Figure 2（a），of the 40 fragments generated by the EcoRI enzyme，the first 10—13 in regions Ⅰ and Ⅱ are worth noting.In many species there are 6 fragments in each of these two regions.Fragments 1—3 and 6 in region I are stable and present in all 35 Nicotiana species examined so far.In contrast，fragments 4 and 5 are extremely variable.One or both can be altered；one or both may be absent.Unlike region I，the six fragments in region Ⅱ are stable with only a few exceptions.The constant fragments may indicate that a mutation in this region is not likely to survive because of the essential function of the fragment.Indeed，the fragments 6 and 9 are known to contain the 23s and 16s rRNA genes respectively（Sugiura and Kusuda，1979）.

In region Ⅲ only half of the fragments are constant.The variability increases in the remaining regions（Ⅳ—Ⅵ）.It is such a combination of diversity and similarity in fragment pattern that forms the basis of species specificity as well as the overall identity of Nicotiana ct-DNA.The EcoRI patterns in Figure 2（a）contain essential information reflecting many important evolutionary events.For example，the ct-DNA of N.noctiflora bears close resemblance to that of N.gossei because they are historically related［see Figure 2（a）lane c and d］.The same information exists among the numerous fragments generated by BamHI［see Figure 2（b）］and many other restriction enzymes.Among the seven Nicotiana species analyzed in Figure 2（b），ct-DNAs of N.longiflora，N.sylvestris and N.tabacum（lane c，e，and f）have an identical BamHI fragment pattern.Although N.longiflora and N.langsdorffii ct-DNAs（lanes c and g）bear a close resemblance they are clearly different.N.glauca ct-DNA（lane b）is quite similar to that of N.plumbaginifolia（not shown）but is distinguishable from that of both N.gossei and N.otphora（lane a and d）.The difference of ct-DNA between N.gossei and N.otophora is remarkable.They differ in 13 out of the 27 BamHI bands.This matches the extent of variation in their EcoRI fragments in which 18 of the 40 bands are different（see Figure 2（a）lanes a and d）.Therefore，both EcoRI and BamHI can be used to measure accurately the degree of variability of ct-DNA in Nicotiana species.

In the case of restriction enzymes which generate only a small number of fragments［see Figure 1（b）］SalI and SmaI were selected for this study because they have been frequently used to construct restriction maps of higher plant ct-DNA（Bedbrook and Bogorad，1976；Jurgenson and Bourque，1980；Herrmann et al.，1980）.

Figure 3（a）is a typical SalI restriction pattern of Nicotiana ct-DNA in which there are 9—11 bands.There is no difference in Sall pattern among many Nicotiana species and only in four species closely related to N.tomentosa are some small fragments combined to form a large one.In these species（N.tomentosa，N.tomentosiformis，N.knightiana and N.paniculata）fragments 7 and 10 are replaced by fragment 6 which is now represented as doublet［see Figure 3（a），lane b］.

In contrast，the SmaI fragment patterns are variable and provide much more revealing information.Figures 3（b）and（c）show the SmaI pattern of ct-DNA from several key Nicotiana species on the evolutionary scheme.They represent six unique types of SmaI ct-DNA fragment patterns.In the 22 species examined there are a combined total of 21 distinct SmaI bands.Eight are present in every species and thirteen are variable and scattered（see Table 1）.On the average there are 13—15 bands in each species.The fragments 17 and 20 are doublets in all species whereas fragment 15 exists in duplicate in most species but is missing in N.tomentosa，N.tomentosiformis，N.knightiana and N.paniculata.Only in N.tabacum，N.sylvestris and N.longiflora are there two copies of fragment 18（see Table 1）.Since the molecular weight of each fragment is known（see Table 1）it is therefore possible to trace the origin and conversion of each variable band in all six species.For example，fragment 3 is unique and found only in N.plumbaginifolia［see Figure 3（b），lane a］.It is replaced in N.langsdorffii by fragments 7 and 9［see Figure 3（b），lane b］.The combined molecular weight of fragment 7 and 9（12.0×106）is practically the same as that offragment 3（12.5×106）.Since all the other fragments are constant the conversion from fragment 3 of N.plumbaginifolia to 7 and 9 of N.langsdorffii is clearly demonstrated.It is thus deduced that ct-DNA of N.langsdorffii，a more advanced species，has evolved from N.plumbaginifolia by the introduction of a single SmaI site into fragment 3 of N.plumbaginifolia［see Figure 3（b），lanes a，b and Figure 4］.Likewise，a gain of a single SmaI site into fragment 6 of N.langsdorffii ct-DNA at different positions has generated the ct-DNAs of N.sylvestris and of N.otophora［see Figure 3（b），lanes c、d and Figure 4］.The two new SmaI fragments generated in each case are fragments 12 and 18 for N.sylvestris，13 and 16 for N.otophora.Their combined molecular weights are 6.85×106 and 6.7×106 which is essentially equal to the original size of 6.9×106 for fragment 6（see Table 1）.This is confirmed by nick translation and hybridization experiment（data not shown）.The alternative interpretation in this case is that initially both N.sylvestris and N.otophora could have identical Smal sites but subsequently an inversion occurred in one of them.

In N.tomentosa，N.tomentosiformis，N.knightiana and N.paniculata elimination instead of gain of the Smal site seems to be the mechanism of the evolution of their ct-DNA.As illustrated in Figure 3（b）and（c），the unique fragment 11（5.1×106）of N.tomentosa appeared which has a combined molecular weight（5.1×106）of fragments 15 and 19 not found in this species（see Table 1）.The formation of fragment 14（4.0×106）resulted from the combination of the missing fragments 15（second copy，3.5×106）and 21（0.61×106）.Therefore，the elimination of SmaI sites between fragments 15 and 19，15 and 21，is clearly demonstrated（see Figure 4）.Since all other fragments are equal，the reduction of fragment size from 7 to 8 in N.paniculata［see Figure 3（c），lane b and Figure 4］may indicate a deletion.To date，no polymorphism for a restriction site has been observed in this study although such intraspecific variability was reported in N.debneyi（Scowcroft，1979）and in animal mitochondrial genomes（Castora et al.，1980）.

From the combination of gain and elimination of SmaI sites it is established that fragments 7 and 9，12 and 18，13 and 16，15 and 19 and 15 and 21 are linked.This is in agreement with the SmaI restriction map of N.tabacum ct-DNA（unpublished data）on which fragments 15，19，7，9，12 and 18 are linked together and arranged in this order.On the basis of this information a sequence of evolution of ct-DNA from several species in Nicotiana is proposed.Figure 4 is a diagramatic representation indicating the sequential addition of three SmaI sites and the deletion of SmaI and SalI sites during the course of ct-DNA evolution.Because all of these observed changes are clustered together in one region they may represent a“hot spot”as defined by Benzer（1961）.This region or at least the SmaI site seems to mutate much more frequently than others.(https://www.daowen.com)

Figure 5 is a phylogenetic tree of ct-DNA evolution within this genus.Goodspeed（1954）assummed that the ancestral stocks having 5 pairs of chromosomes had diverged into several branches.It appears the present-day species N.plumbaginifolia represents one of them at the 12-paired level.In this proposed scheme of evolution of the Nicotiana chloroplast genome，N.plumbaginifolia may have served as the ancestral stock from which all the Australian and some of the American species were derived.It is proposed here that ct-DNA of N.langsdorffii originated directly from N.plumbaginifolia.There is adequate cytogenetic evidence to support this claim（Goodspeed，1954）.N.langsdorffii may in turn have given rise to N.noctiflora，N.Iongiflora and N.alata.The evidence that N.noctiflora includes N.langsdorffii elements is available（Goodspeed，1954）.At this point of evolution the ancestor of N.noctiflora had migrated to Australia and formed the current Australian population.N.noctiflora ct-DNA has similar or identical restriction fragment patterns generated by SalI，SmaI，BamHI and EcoRI，to that of N.langsdorffii and many Australian species including N.suaveolens，N.gossei，N.excelsior，N.megalosiphon and N.fragrans.The evidence is therefore very convincing that the ancestor of N.noctiflora is indeed the link between these two populations of Nicotiana as often suggested（Goodspeed，1954；Chen et al.，1976）.The restriction enzyme analysis of N.africana ct-DNA also indicate that it may have been derived from an ancestor similar to N.noctiflora.

The proposed sequence of ct-DNA evolution from N.plumbaginifolia to N.iangsdorffii，N.iongiflora，N.sylvestris and N.tabacum deserves special treatment.These are very closely related species and all belong to the same section Alatae except N.tabacum（Goodspeed，1954）which is the progeny of a cross between N.sylvestris and N.tomentosiformis（Gray et al.，1974）.This proposal is based on the restriction analysis of their ct-DNAs but is strongly supported by evidence obtained from taxonomic and biochemical studies（Goodspeed，1954；Chen et al.，1976）.The recent biochemical evidence is particularly convincing.All five species in this sequence contain the identical isoelectric focusing pattern of the large subunit of RuBPCase，a product of the chloroplast genome（Chen et al.，1976）.Their differential relationship is clearly illustrated in Table 2 which strongly supports this proposed sequence.

Chloroplast DNAs from N.tomentosa，N.tomentosiformis，N.knightiana and N.paniculata are clearly separated from all the other species studied here.Their immediate ancestor is different from but probably parallel to N.plumbaginifolia.However，the close relationship of this group is highly supported by other evidence（Goodspeed，1954）.Furthermore，the evidence obtained from amino acid analysis and peptide mapping of the large subunit of RuBPCase（Kung et al.，1976）and the current analysis of restriction fragments of ct-DNA［see Figure 3（b）、（c）and Figure 4］further divide these species into two subgroups；N.tomentosa and N.tomentosiformis for one and N.knightiana and N.paniculata for the other（see Figure 5）.This is in agreement with the results obtained previously（Rhodes et al.，1981）.

It is interesting to note that in this proposed scheme of evolution（see Figure 5）there is a close correlation between the restriction pattern of ct-DNA and the isoelectric focusing pattern of the large subunit of RuBPCase.It is well established that the large subunit of RuBPCase is coded by ct-DNA（Kung，1976）.Whether such a correlation bears any functional relationship is currently under investigation.