- Email: [email protected]
Genome analysis of Amphioxus and speculation as to the origin of contrasting vertebrate genome organization patterns
Comp. Biochem. PhysioL. Vol. 63B, pp. 455 to 458
0305-0491/79/0801-0455502.00/0
© Pergamon Press Ltd 1979. Printed in Great Britain
GENOME ANALYSIS OF A M P H I O X U S AND SPECULATION AS TO THE ORIGIN OF CONTRASTING VERTEBRATE GENOME ORGANIZATION PATTERNS J. SCHMIDTKE,J. T. EPPLEN and W. ENGEL Institut ffir Humangenetik der Universit~it,D-3400 G6ttingen, West Germany
(Received 19 December 1978) Abstract--1. The genome of Amphioxus was investigated by DNA reassociation techniques for the
amount of repetitive and non-repetitive sequences and its pattern of organization. 2. A comparison of the amount of non-repetitive DNA between Amphioxus and the tunicate Ciona intestinalis does not support the hypothesis that the Cephalochordates have arisen from the Tunicates by polyploidy. 3. In the Amphioxus genome repetitive and non-repetitive elements are predominantly arranged in a short period interspersion pattern. Conclusions are presented as to the evolution of contrasting genome organization patterns amnng vertebrates.
INTRODUCTION
In eukaryotic genomes at least two different patterns of arrangement of repeated and non-repeated DNAsequences exist. One, known as the "Xenopus-pattern" (Davidson et al., 1975) is characterized by an intimate interspersion of non-repeated sequences (mostly one kilobase (kb) in length) with short repeated ones (averaging 0.3 kb), whereas the "Drosophila pattern" (Manning et al., 1975) consists of much longer non-repeated sequences (> 13 kb) adjacent to long (about 5.6 kb) repeats. Until quite recently it was thought that the Drosophila pattern is restricted to a few insect species only (Drosophila melanogaster, Chironomus tentans (Wells et al., 1976) and Apis mellifera (Crain et al., 1976)). It was found, however, that the genome organization of several bird species, namely, the duck (Epplen et al., 1978), the chicken (Arthur & Straus, 1978; Epplen et al., 1978) and the pigeon (Epplen et al., in preparation) in many respects resembles the Drosophila pattern more closely than the Xenopus pattern. In view of the presence of contrasting patterns of genome organization in vertebrates it is of great interest to know the genome organization of the cephalochordate Amphioxus (Branchiostoma). This species is thought to be the immediate phylogenetic ancestor of the vertebrates, or at least shares a common ancestry with the vertebrates (Bone, 1972; Jollie, 1973). The composition of repeated and non-repeated DNA sequences of the Amphioxus genome is also of interest in comparison to the genome of Ciona intestinalis representing the third chordate subphylum, namely the tunicates. The genome size of Ciona is 0.45 pg, the size of the Amphioxus genome is 1.28 pg (Atkin & Ohno, 1967). Ohno (1974) has speculated that an Amphioxus-like genome arose from a Cionalike genome by way of polyploidization. If the polyploidization hypothesis is correct, and if the presentday genome composition resembles the situation during chordate evolution, the Amphioxus genome should be expected to contain double the amount of non-
repetitive DNA than Ciona, while the extra DNA, which cannot be explained by polyploidization would consist of repeated sequences added at a later stage of genomic evolution. We here report the results from the genome analysis of Amphioxus using DNA reassociation techniques. The DNA sequence composition of the Ciona genome was studied before (Lambert & Laird, 1971). We have reinvestigated Ciona DNA by reassociation analysis under our own standard working conditions, MATERIAL AND METHODS
Origin of animals Ciona intestinalis specimens were collected from the harbour of Helgoland (North Sea). Amphioxus specimens came from the "amphioxus grounds" near Helgoland. DN A preparation Sperm was collected into filtered seawater, and DNA was isolated from the lysed cells and purified essentially according to the method of Kirby (1968) as described previously (Epplen et al., 1978). Fragmentation and sizing of the DNA DNA was fragmented with a Branson sonifier B 12 cell disruptor. Fragment lengths of about 200(Ciona) and 250(Amphioxus) base pairs were obtained by sonication at intervals for 600sec at setting 4. Long fragments of Amphioxus DNA (about 1200 base pairs) were obtained after sonication for 7 sec. Sizing of the fragments was performed by alkaline sucrose gradient analysis in the Spinco L2 50B ultracentrifuge using an isotopically labelled marker of sheared E. coli DNA sized with the electron microscope (Epplen et al., 1978). DNA meltin 0 experiments In order to determine base composition DNA was melted in 0.1 x SSC (0.015 M sodium chloride~).0015 M sodium citrate, pH 7.0) in a Unicam SP 1800 spectrophotometer using thermo-jacketed cuvettes. The change in absorbance was registered with a Unicam AR 25 linear recorder while the temperature was monitored using a Lauda digital thermometer with a Pt 100 resistance tern-
455
J. SCHM1DTKE,J. T. EPPLEN and W. ENGEL
456
lytical genome size determination (Atkin & Ohno, 1967). E. coil D N A used as a kinetic standard had a COTt/2 value of 2.59 msec. If the sequence composition of the Amphioxus genome is compared to that of Ciona (Table 1, Fig. 2), it becomes evident that from these results the polyploidization hypothesis forwarded by O h n o (see Introduction) cannot be supported. In accordance with previous findings (Lambert & Laird, 1971), we found the Ciona genome to consist of 3 2 ~ repeated and of 57~o non-repeated D N A (the COT1/2 value of the slow fraction agrees well with the value expected from the analytical genome size determination). If the genome fraction sizes are corrected for the amount of non-reacting, apparently degraded material, which amounts to 11.6~ in the case of Ciona and to 13.8~o in Amphioxus, it can be seen that the Amphioxus genome has almost four times more single copy D N A than Ciona, whereas only double the amount would have been expected to have arisen by polyploidization. Although it is conceivable that large part of the single-copy D N A of Amphioxus had originated from degenerated repeats, added later during genome evolution, we do not feel that the D N A reassociation results are altogether a sound basis for further speculation on polyploid evolution of Amphioxus. It should be noted also that isozyme analysis of the two species does not confirm the polyploidization hypothesis (Schmidtke et al., 1977, Schmidtke et al., 1978).
perature meter. Guanine + cytosine (GC)-contents were calculated according to Mandel & Marmur (1968): GC~o = (T,, - 53.9) 2.44, where T,, denotes the temperature at 50~o hyperchromicity. The duplex content of long reassociated DNA fragments was estimated in the following way. DNA was reassociated to the desired COT (COT is the product of the DNA concentration (mols of nucleotides per litre) and time (sec)), the fraction bound to hydroxyapatite was eluted with 0.4M PB, adjusted to 0.12M PB and melted in the spectrophotometer as described above. The hyperchromicity observed was corrected for collapse hyperchromicity (2.7%). The T,, of the reassociated DNA was compared to the T,, of native DNA in 0.12 M PB. Assuming that 1 deg. C difference in T,, is equivalent to 1~o base mismatch lowering the hyperchromicity (H), the expected hyperchromicity of reassociated fragments paired at full length but carrying mispaired regions, was calculated (H native corrected). The mean proportion of DNA duplex (d) is therefore estimated as d=
(H observed - 0.027) (H native corr. - 0.027)'
Reassociation oJ the DNA DNA samples were heat-denatured in a boiling water bath for 5-10 min and afterwards transferred to a water bath at 60°C (DNA in 0.04M PB) and at 65°C (DNA in 0.55 M PB). After renaturation to the desired COT value, separation of single-stranded molecules from molecules bearing duplex regions was performed by chromatographing the samples over hydroxyapatite. Single-stranded molecules were eluted with 0.12M PB and doublestranded molecules with 0.4 M P B at 60°C. The kinetic parameters of the reassociation were calculated using the "BMDX 85 non-linear least squares" computer program (UCLA, 1970).
Genome organtzation of Amphioxus One commonly employed technique to demonstrate interspersion of single copy and repetitive sequences is by analysis of the reassociation kinetics of long D N A fragments (Britten & Smith, 1970; Davidson et al., 1973; Graham et al., 1974; Goidberg et al., 1975; Crain et al., 1976; Zimmerman & Goldberg, 1977; Epplen et al., 1978). Moreover, without exception this technique has proved to be indicative of the type of sequence interspersion. Figure 1, lower curve, shows the reassociation kinetics of long (about 1200 base pairs) fragments of Amphioxus DNA. It is seen that at a C O T of 10, at which point all repetitive sequences but less than only 4 ~ of the single copy sequences have reacted, hydroxyapatite binding is increased from 2 0 ~ (short D N A ) to 6 5 ~ (long DNA). Therefore, long fragments bound at this C O T must carry non-reacted single-copy-sequences adjacent to the reacted repeats. This result demonstrates directly
RESULTS AND DISCUSSION
Sequence composition of the Amphioxus and Ciona 9enomes The renaturation of short Amphioxus D N A fragments exhibited a complex kinetic behaviour (Fig. 1, upper curve). The best estimate of the composition of repeated and non-repeated sequences is presented in Table 1. According to this analysis a repeated sequence fraction, which may also contain palindromic sequences, with a COT~/2 > 10 -3, can be distinguished from a fraction containing moderately repeated sequences, repetitive frequency about 1000). The slow fraction consists mainly of non-repeated sequences as judged from a comparison of the kinetic rate constant with the value expected from the ana-
Table 1. Sequence composition of Amphioxus* and Cionat DNA Very fast fraction K COT1/2~
Amphioxus Ciona
0.098
> l03
< l0 -3
~o 0.078 0.318
Fast fraction K COT1/2~ 4.16 19.33
0.241 0.052
~o
Slow fraction K COT1/2~
0.686 0.00446 0.566 0.0105
224 95
* The hyperchromicity of native Amphioxus DNA was 0.267. The T,, of native DNA in 0.1 SSC was 70.9°C. From this a GC (guanine + cytosine) content of 41.5~o was calculated. This value is in accordance with previous determinations (40.8-42.6~o) using different methods (Russell & Subak-Sharpe, 1977). t The hyperchromicity of native Ciona DNA was 0.264. The T,, of native DNA in 0.1 SSC was 69.5°C. This corresponds to a GC content of 38.1~o, and agrees well with the previous determination of Lambert & Laird (1971) of 38.3~. ++The kinetic parameters are standardized for 0.12 M PB (Britten et al., 1974).
Genome analysis of Amphioxus and genome organization patterns
2L
q
457
- [O][] ~
g
[]
~6 o ~8
[]
-2
-I
÷1
0
~2
+3
[]
+4
Log cot
Fig. 1. Reassociation of Amphioxus DNA of fragment lengths of about 250 base pairs (13) and of about 1200 base pairs (N), The samples were renatured in 0.04 M and 0.55 M phosphate buffer. Fraction bound refers to the fraction reassociated (bound to hydroxyapatite). COT values are expressed equivalent to 0.12 M phosphate buffer. The lines are computer-plotted least squares fits to the data points. that in the Amphioxus genome repetitive and singlecopy sequences are intimately interspersed. The fact that even at the lowest COT (10 -3 ) binding of long fragments is most markedly increased (when compared with short fragments) shows that very highly repeated sequences and/or palindromic structures are interspersed throughout the genome. In order to determine the duplex content of long reassociated fragments the hyperchromicity of molecules bound to hydroxyapatite at two different values of COT was determined. After correction for thermal stability and collapse hyperchromicity (see Material and Methods) a duplex content of 6% at COT 0.045, and of 40% at COT 30 was calculated. At COT 0.045 all of the highly repeated sequences but only about 15% of the intermediate repeats and none of the single-copy sequences have reassociated, at COT 30 all of the intermediate repeats but only about 12% of the single-copy sequences are reannealed as judged from the 250 base pairs reassociation curve. The duplex contents, as determined, therefore show that most of the fragments carrying reassociated repetitive regions bear flanking sequences of lower repetitivity and stretches of single-copy DNA.
On the origin of contrasting vertebrate gexnome organization patterns It is concluded that the genome of Amphioxus is organized in a "short period interspersion" pattern (Davidson et al., 1973). Since Amphioxus may represent the immediate phylogenetic ancestor of the vertebrates, there is reason to believe that the "original" vertebrate genome organization pattern is in fact Xenopus-like. This pattern obviously has been main-
tained during the evolution of most of the vertebrates. The long period interspersion pattern observed in bird genomes seems to be restricted to this vertebrate class. This assumption is further supported by COT analysis of reptile DNA performed in our laboratory (Epplen et al., 1978; Epplen et al., unpublished). Reptiles and birds are often combined to the superclass Sauropsida. According to Romer (1971) Ayes and Crocodylia have arisen from the Archosauria, whereas the Chelonia have diverged earlier as part of the anapsidian line of reptile evolution. Up to now, one representative species of each of these two reptile orders has been investigated with regard to the genomic organisation of repetitive and non repetitive sequences, namely Caiman crocodylus (Crocodylia) and Terrapene carolina triungius (Chelonia). It was found that the genomes of these species are organized in a short period interspersion pattern. It should be emphasized that birds in general have smaller genome sizes than the other sauropsidians. Therefore, analogous to the development of the Drosophila pattern during insect evolution, as proposed by Crain et al. (1976), we speculate that the long single and repetitive elements in bird genomes have arisen by deletion of interspersed single-copy and repetitive regions. It is expected that the analysis of further suitable groups of related vertebrate species with varying genome size (e.g. fish) will help to elucidate a possible generality of this aspect of genome evolution. This appears to be especially important in view of the functional significance of genome organization patterns (Britten & Davidson, 1971). Does genome size reduction lead to specialization through loss of regulatory versatility?
za
.g 6 ,,,- 8 I -2
L -I
L
I
I
0
+1
+2
L +3
Log cot
Fig. 2. Reassociation of Ciona DNA at a fragment length of about 200 base pairs. For further details see legend to Fig. 1.
458
J. SCHMIDTKE, J. T. EPPLEN and W. ENGEL
Acknowledgements--Ms Sonja P~itzold is thanked for skilful technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft. REFERENCES ARTHUR R. R. ~ STRAUS N. A. (1978) DNA-sequence organisation in the genome of the domestic chicken (Gallus domesticus). Can. J. Biochem. 56, 257-263. ATKIN N. n. ~ OHNO S. (1967) DNA values of four primitive chordates. Chromosoma 23, 10-13. BONE Q. (1972) The Origin of Chordates. Oxford University Press, London. BRITTEN R. J. & SMITH J. (1970) A bovine genome. Carnegie Inst. Wash. Year Book 68, 378-386. BRITTEN R. J. & DAVIOSON E. H. (1971) Gene regulation for higher cells: A theory. Science N.Y. 165, 349-357. BRITTEN R. J., GRAHAM D. E. & NEUEELD B. R. (1974) Analysis of repeating DNA sequences by reassociation. In Methods in Enzymology (Edited by GROSSMAN L. & MOLDAVE K.L pp. 363-418. Academic Press, New York. CRAIN W. R., DAVIDSON E. H. & BRITTEN R. J. (1976) Contrasting patterns of DNA sequence arrangement in Apis mellifera (honeybee) and Musca domestica (housefly). Chromosoma 59, 1-12. DAV1DSON E. H., GALAU G. A., ANGERER R. C. & BRITTEN R. J. (1975) Comparative aspects of DNA organisation in metazoa. Chromosoma 51, 253-259. DAVIDSON E. H., HOUGH B. R., AMENSON C. S. ~g BRITTEN R. J. (1973) General interspersion of repetitive with nonrepetitive elements in the DNA of Xenopus. J. molec. Biol. 77, 1-23. EPPLEN J. TH., LEIPOLDT M., ENGEL W. & SCHMIDTKE J. (1978) Untersuchungen i,iber die Genomorganisation bei Sauropsiden. Arch. Genet. 51, 4. EPPLEN J. TH., LEIPOLDT M., ENGEL W. & SCHMIDTKE J. DNA sequence organisation in avian genomes. Chromosoma. 69, 307-321. GOLDBERG R. B., CRAIN W. R., RUDERMAN J. V., MOORE G. P., BARNETT T. R., H1GGINS R. C., GELFAND R. A., GALAU G. A., BRITTEN R. J. & DAVIDSON E. H. (1975) DNA sequence organisation in the genomes of five marine invertebrates. Chromosoma 51, 225-251.
GRAHAM D. E., NEUFELD B. R., DAVIDSONE. H. & BRITTEN R. J. (1974) Interspersion of repetitive and nonrepetitive DNA sequences in the sea urchin genome. Cell 1, 127-137. JOLLIE M. (1973) The origin of the chordates. Aeta 2oo1., Stockh. 54, 81 100. KIRBY K. S. (1968) Isolation of nucleic acids with phenolic solvents. In Methods in Enzymology, Vol. 12B (Edited by GROSSMANL. 8., MOLDAVE K.), pp. 87-I 12. Academic Press, New York. LAMBERT, Ca. C. ~ LAIRD, CH. D. (1971) Molecular properties of tunicate DNA. Biochim. biophys. Acta 240, 39-45. MANDEL M. & MARMURJ. (1968) Use of ultraviolet absorbance-temperature profile for determining the guanine plus cytosine content of DNA. In Methods in En2ymoIogy, Vol. 12B (Edited by GROSSMAN L. & MOLDAVE K.), pp. 195-206. Academic Press, New York. MANNING J. E., SCHMID C. W. • DAVIDSON N. 11975) Interspersion of repetitive and non-repetitive DNA sequences in the Drosophila melanogaster genome. Cell 4, 141-155. OHNO S. (1974) Animal Cytogenetics 4, Chordata 1, Protochordata, Cyclostomata, and Pisces, pp. 64-68. Gebriider Borntraeger, Berlin, Stuttgart. ROMER A. S. (1971) Veryleichende Anatomie der Wirbeltiere. Parey, Hamburg. RUSSELL G. J. & SUBAK-SHARPEJ. H. (1977) Similarity of the general designs of protochordates and invertebrates. Nature, Lond. 266, 533-535. SCHMIDTKE J., WEILER C., KUNZ B. & ENGEL W. (1977) Isozymes of a tunicate and a cephalochordate as a test of polyploidisation in chordate evolution. Nature, Lond. 266, 532-533. SCHMIDTKE J., KUNZ B. & ENGEL W. (1978) Gene numbers and genic variability in Ciona intestinalis and Branchiostoma lanceolatum. Can. J. Genet. Cytol. 20, 61-70. WELLS R., ROVER H.-D. & HOLLENBERG C. P. (1976) Non Xenopus-like DNA sequence organisation in the Chironomus tentans genome. Molec. Gen. Genet. 147, 45-51. ZIMMERMANJ. L. & GOLDBERG R. B. (1977) DNA sequence organisation in the genome of Nicotiana tabacum. Chromosoma 59, 227-252.
0305-0491/79/0801-0455502.00/0
© Pergamon Press Ltd 1979. Printed in Great Britain
GENOME ANALYSIS OF A M P H I O X U S AND SPECULATION AS TO THE ORIGIN OF CONTRASTING VERTEBRATE GENOME ORGANIZATION PATTERNS J. SCHMIDTKE,J. T. EPPLEN and W. ENGEL Institut ffir Humangenetik der Universit~it,D-3400 G6ttingen, West Germany
(Received 19 December 1978) Abstract--1. The genome of Amphioxus was investigated by DNA reassociation techniques for the
amount of repetitive and non-repetitive sequences and its pattern of organization. 2. A comparison of the amount of non-repetitive DNA between Amphioxus and the tunicate Ciona intestinalis does not support the hypothesis that the Cephalochordates have arisen from the Tunicates by polyploidy. 3. In the Amphioxus genome repetitive and non-repetitive elements are predominantly arranged in a short period interspersion pattern. Conclusions are presented as to the evolution of contrasting genome organization patterns amnng vertebrates.
INTRODUCTION
In eukaryotic genomes at least two different patterns of arrangement of repeated and non-repeated DNAsequences exist. One, known as the "Xenopus-pattern" (Davidson et al., 1975) is characterized by an intimate interspersion of non-repeated sequences (mostly one kilobase (kb) in length) with short repeated ones (averaging 0.3 kb), whereas the "Drosophila pattern" (Manning et al., 1975) consists of much longer non-repeated sequences (> 13 kb) adjacent to long (about 5.6 kb) repeats. Until quite recently it was thought that the Drosophila pattern is restricted to a few insect species only (Drosophila melanogaster, Chironomus tentans (Wells et al., 1976) and Apis mellifera (Crain et al., 1976)). It was found, however, that the genome organization of several bird species, namely, the duck (Epplen et al., 1978), the chicken (Arthur & Straus, 1978; Epplen et al., 1978) and the pigeon (Epplen et al., in preparation) in many respects resembles the Drosophila pattern more closely than the Xenopus pattern. In view of the presence of contrasting patterns of genome organization in vertebrates it is of great interest to know the genome organization of the cephalochordate Amphioxus (Branchiostoma). This species is thought to be the immediate phylogenetic ancestor of the vertebrates, or at least shares a common ancestry with the vertebrates (Bone, 1972; Jollie, 1973). The composition of repeated and non-repeated DNA sequences of the Amphioxus genome is also of interest in comparison to the genome of Ciona intestinalis representing the third chordate subphylum, namely the tunicates. The genome size of Ciona is 0.45 pg, the size of the Amphioxus genome is 1.28 pg (Atkin & Ohno, 1967). Ohno (1974) has speculated that an Amphioxus-like genome arose from a Cionalike genome by way of polyploidization. If the polyploidization hypothesis is correct, and if the presentday genome composition resembles the situation during chordate evolution, the Amphioxus genome should be expected to contain double the amount of non-
repetitive DNA than Ciona, while the extra DNA, which cannot be explained by polyploidization would consist of repeated sequences added at a later stage of genomic evolution. We here report the results from the genome analysis of Amphioxus using DNA reassociation techniques. The DNA sequence composition of the Ciona genome was studied before (Lambert & Laird, 1971). We have reinvestigated Ciona DNA by reassociation analysis under our own standard working conditions, MATERIAL AND METHODS
Origin of animals Ciona intestinalis specimens were collected from the harbour of Helgoland (North Sea). Amphioxus specimens came from the "amphioxus grounds" near Helgoland. DN A preparation Sperm was collected into filtered seawater, and DNA was isolated from the lysed cells and purified essentially according to the method of Kirby (1968) as described previously (Epplen et al., 1978). Fragmentation and sizing of the DNA DNA was fragmented with a Branson sonifier B 12 cell disruptor. Fragment lengths of about 200(Ciona) and 250(Amphioxus) base pairs were obtained by sonication at intervals for 600sec at setting 4. Long fragments of Amphioxus DNA (about 1200 base pairs) were obtained after sonication for 7 sec. Sizing of the fragments was performed by alkaline sucrose gradient analysis in the Spinco L2 50B ultracentrifuge using an isotopically labelled marker of sheared E. coli DNA sized with the electron microscope (Epplen et al., 1978). DNA meltin 0 experiments In order to determine base composition DNA was melted in 0.1 x SSC (0.015 M sodium chloride~).0015 M sodium citrate, pH 7.0) in a Unicam SP 1800 spectrophotometer using thermo-jacketed cuvettes. The change in absorbance was registered with a Unicam AR 25 linear recorder while the temperature was monitored using a Lauda digital thermometer with a Pt 100 resistance tern-
455
J. SCHM1DTKE,J. T. EPPLEN and W. ENGEL
456
lytical genome size determination (Atkin & Ohno, 1967). E. coil D N A used as a kinetic standard had a COTt/2 value of 2.59 msec. If the sequence composition of the Amphioxus genome is compared to that of Ciona (Table 1, Fig. 2), it becomes evident that from these results the polyploidization hypothesis forwarded by O h n o (see Introduction) cannot be supported. In accordance with previous findings (Lambert & Laird, 1971), we found the Ciona genome to consist of 3 2 ~ repeated and of 57~o non-repeated D N A (the COT1/2 value of the slow fraction agrees well with the value expected from the analytical genome size determination). If the genome fraction sizes are corrected for the amount of non-reacting, apparently degraded material, which amounts to 11.6~ in the case of Ciona and to 13.8~o in Amphioxus, it can be seen that the Amphioxus genome has almost four times more single copy D N A than Ciona, whereas only double the amount would have been expected to have arisen by polyploidization. Although it is conceivable that large part of the single-copy D N A of Amphioxus had originated from degenerated repeats, added later during genome evolution, we do not feel that the D N A reassociation results are altogether a sound basis for further speculation on polyploid evolution of Amphioxus. It should be noted also that isozyme analysis of the two species does not confirm the polyploidization hypothesis (Schmidtke et al., 1977, Schmidtke et al., 1978).
perature meter. Guanine + cytosine (GC)-contents were calculated according to Mandel & Marmur (1968): GC~o = (T,, - 53.9) 2.44, where T,, denotes the temperature at 50~o hyperchromicity. The duplex content of long reassociated DNA fragments was estimated in the following way. DNA was reassociated to the desired COT (COT is the product of the DNA concentration (mols of nucleotides per litre) and time (sec)), the fraction bound to hydroxyapatite was eluted with 0.4M PB, adjusted to 0.12M PB and melted in the spectrophotometer as described above. The hyperchromicity observed was corrected for collapse hyperchromicity (2.7%). The T,, of the reassociated DNA was compared to the T,, of native DNA in 0.12 M PB. Assuming that 1 deg. C difference in T,, is equivalent to 1~o base mismatch lowering the hyperchromicity (H), the expected hyperchromicity of reassociated fragments paired at full length but carrying mispaired regions, was calculated (H native corrected). The mean proportion of DNA duplex (d) is therefore estimated as d=
(H observed - 0.027) (H native corr. - 0.027)'
Reassociation oJ the DNA DNA samples were heat-denatured in a boiling water bath for 5-10 min and afterwards transferred to a water bath at 60°C (DNA in 0.04M PB) and at 65°C (DNA in 0.55 M PB). After renaturation to the desired COT value, separation of single-stranded molecules from molecules bearing duplex regions was performed by chromatographing the samples over hydroxyapatite. Single-stranded molecules were eluted with 0.12M PB and doublestranded molecules with 0.4 M P B at 60°C. The kinetic parameters of the reassociation were calculated using the "BMDX 85 non-linear least squares" computer program (UCLA, 1970).
Genome organtzation of Amphioxus One commonly employed technique to demonstrate interspersion of single copy and repetitive sequences is by analysis of the reassociation kinetics of long D N A fragments (Britten & Smith, 1970; Davidson et al., 1973; Graham et al., 1974; Goidberg et al., 1975; Crain et al., 1976; Zimmerman & Goldberg, 1977; Epplen et al., 1978). Moreover, without exception this technique has proved to be indicative of the type of sequence interspersion. Figure 1, lower curve, shows the reassociation kinetics of long (about 1200 base pairs) fragments of Amphioxus DNA. It is seen that at a C O T of 10, at which point all repetitive sequences but less than only 4 ~ of the single copy sequences have reacted, hydroxyapatite binding is increased from 2 0 ~ (short D N A ) to 6 5 ~ (long DNA). Therefore, long fragments bound at this C O T must carry non-reacted single-copy-sequences adjacent to the reacted repeats. This result demonstrates directly
RESULTS AND DISCUSSION
Sequence composition of the Amphioxus and Ciona 9enomes The renaturation of short Amphioxus D N A fragments exhibited a complex kinetic behaviour (Fig. 1, upper curve). The best estimate of the composition of repeated and non-repeated sequences is presented in Table 1. According to this analysis a repeated sequence fraction, which may also contain palindromic sequences, with a COT~/2 > 10 -3, can be distinguished from a fraction containing moderately repeated sequences, repetitive frequency about 1000). The slow fraction consists mainly of non-repeated sequences as judged from a comparison of the kinetic rate constant with the value expected from the ana-
Table 1. Sequence composition of Amphioxus* and Cionat DNA Very fast fraction K COT1/2~
Amphioxus Ciona
0.098
> l03
< l0 -3
~o 0.078 0.318
Fast fraction K COT1/2~ 4.16 19.33
0.241 0.052
~o
Slow fraction K COT1/2~
0.686 0.00446 0.566 0.0105
224 95
* The hyperchromicity of native Amphioxus DNA was 0.267. The T,, of native DNA in 0.1 SSC was 70.9°C. From this a GC (guanine + cytosine) content of 41.5~o was calculated. This value is in accordance with previous determinations (40.8-42.6~o) using different methods (Russell & Subak-Sharpe, 1977). t The hyperchromicity of native Ciona DNA was 0.264. The T,, of native DNA in 0.1 SSC was 69.5°C. This corresponds to a GC content of 38.1~o, and agrees well with the previous determination of Lambert & Laird (1971) of 38.3~. ++The kinetic parameters are standardized for 0.12 M PB (Britten et al., 1974).
Genome analysis of Amphioxus and genome organization patterns
2L
q
457
- [O][] ~
g
[]
~6 o ~8
[]
-2
-I
÷1
0
~2
+3
[]
+4
Log cot
Fig. 1. Reassociation of Amphioxus DNA of fragment lengths of about 250 base pairs (13) and of about 1200 base pairs (N), The samples were renatured in 0.04 M and 0.55 M phosphate buffer. Fraction bound refers to the fraction reassociated (bound to hydroxyapatite). COT values are expressed equivalent to 0.12 M phosphate buffer. The lines are computer-plotted least squares fits to the data points. that in the Amphioxus genome repetitive and singlecopy sequences are intimately interspersed. The fact that even at the lowest COT (10 -3 ) binding of long fragments is most markedly increased (when compared with short fragments) shows that very highly repeated sequences and/or palindromic structures are interspersed throughout the genome. In order to determine the duplex content of long reassociated fragments the hyperchromicity of molecules bound to hydroxyapatite at two different values of COT was determined. After correction for thermal stability and collapse hyperchromicity (see Material and Methods) a duplex content of 6% at COT 0.045, and of 40% at COT 30 was calculated. At COT 0.045 all of the highly repeated sequences but only about 15% of the intermediate repeats and none of the single-copy sequences have reassociated, at COT 30 all of the intermediate repeats but only about 12% of the single-copy sequences are reannealed as judged from the 250 base pairs reassociation curve. The duplex contents, as determined, therefore show that most of the fragments carrying reassociated repetitive regions bear flanking sequences of lower repetitivity and stretches of single-copy DNA.
On the origin of contrasting vertebrate gexnome organization patterns It is concluded that the genome of Amphioxus is organized in a "short period interspersion" pattern (Davidson et al., 1973). Since Amphioxus may represent the immediate phylogenetic ancestor of the vertebrates, there is reason to believe that the "original" vertebrate genome organization pattern is in fact Xenopus-like. This pattern obviously has been main-
tained during the evolution of most of the vertebrates. The long period interspersion pattern observed in bird genomes seems to be restricted to this vertebrate class. This assumption is further supported by COT analysis of reptile DNA performed in our laboratory (Epplen et al., 1978; Epplen et al., unpublished). Reptiles and birds are often combined to the superclass Sauropsida. According to Romer (1971) Ayes and Crocodylia have arisen from the Archosauria, whereas the Chelonia have diverged earlier as part of the anapsidian line of reptile evolution. Up to now, one representative species of each of these two reptile orders has been investigated with regard to the genomic organisation of repetitive and non repetitive sequences, namely Caiman crocodylus (Crocodylia) and Terrapene carolina triungius (Chelonia). It was found that the genomes of these species are organized in a short period interspersion pattern. It should be emphasized that birds in general have smaller genome sizes than the other sauropsidians. Therefore, analogous to the development of the Drosophila pattern during insect evolution, as proposed by Crain et al. (1976), we speculate that the long single and repetitive elements in bird genomes have arisen by deletion of interspersed single-copy and repetitive regions. It is expected that the analysis of further suitable groups of related vertebrate species with varying genome size (e.g. fish) will help to elucidate a possible generality of this aspect of genome evolution. This appears to be especially important in view of the functional significance of genome organization patterns (Britten & Davidson, 1971). Does genome size reduction lead to specialization through loss of regulatory versatility?
za
.g 6 ,,,- 8 I -2
L -I
L
I
I
0
+1
+2
L +3
Log cot
Fig. 2. Reassociation of Ciona DNA at a fragment length of about 200 base pairs. For further details see legend to Fig. 1.
458
J. SCHMIDTKE, J. T. EPPLEN and W. ENGEL
Acknowledgements--Ms Sonja P~itzold is thanked for skilful technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft. REFERENCES ARTHUR R. R. ~ STRAUS N. A. (1978) DNA-sequence organisation in the genome of the domestic chicken (Gallus domesticus). Can. J. Biochem. 56, 257-263. ATKIN N. n. ~ OHNO S. (1967) DNA values of four primitive chordates. Chromosoma 23, 10-13. BONE Q. (1972) The Origin of Chordates. Oxford University Press, London. BRITTEN R. J. & SMITH J. (1970) A bovine genome. Carnegie Inst. Wash. Year Book 68, 378-386. BRITTEN R. J. & DAVIOSON E. H. (1971) Gene regulation for higher cells: A theory. Science N.Y. 165, 349-357. BRITTEN R. J., GRAHAM D. E. & NEUEELD B. R. (1974) Analysis of repeating DNA sequences by reassociation. In Methods in Enzymology (Edited by GROSSMAN L. & MOLDAVE K.L pp. 363-418. Academic Press, New York. CRAIN W. R., DAVIDSON E. H. & BRITTEN R. J. (1976) Contrasting patterns of DNA sequence arrangement in Apis mellifera (honeybee) and Musca domestica (housefly). Chromosoma 59, 1-12. DAV1DSON E. H., GALAU G. A., ANGERER R. C. & BRITTEN R. J. (1975) Comparative aspects of DNA organisation in metazoa. Chromosoma 51, 253-259. DAVIDSON E. H., HOUGH B. R., AMENSON C. S. ~g BRITTEN R. J. (1973) General interspersion of repetitive with nonrepetitive elements in the DNA of Xenopus. J. molec. Biol. 77, 1-23. EPPLEN J. TH., LEIPOLDT M., ENGEL W. & SCHMIDTKE J. (1978) Untersuchungen i,iber die Genomorganisation bei Sauropsiden. Arch. Genet. 51, 4. EPPLEN J. TH., LEIPOLDT M., ENGEL W. & SCHMIDTKE J. DNA sequence organisation in avian genomes. Chromosoma. 69, 307-321. GOLDBERG R. B., CRAIN W. R., RUDERMAN J. V., MOORE G. P., BARNETT T. R., H1GGINS R. C., GELFAND R. A., GALAU G. A., BRITTEN R. J. & DAVIDSON E. H. (1975) DNA sequence organisation in the genomes of five marine invertebrates. Chromosoma 51, 225-251.
GRAHAM D. E., NEUFELD B. R., DAVIDSONE. H. & BRITTEN R. J. (1974) Interspersion of repetitive and nonrepetitive DNA sequences in the sea urchin genome. Cell 1, 127-137. JOLLIE M. (1973) The origin of the chordates. Aeta 2oo1., Stockh. 54, 81 100. KIRBY K. S. (1968) Isolation of nucleic acids with phenolic solvents. In Methods in Enzymology, Vol. 12B (Edited by GROSSMANL. 8., MOLDAVE K.), pp. 87-I 12. Academic Press, New York. LAMBERT, Ca. C. ~ LAIRD, CH. D. (1971) Molecular properties of tunicate DNA. Biochim. biophys. Acta 240, 39-45. MANDEL M. & MARMURJ. (1968) Use of ultraviolet absorbance-temperature profile for determining the guanine plus cytosine content of DNA. In Methods in En2ymoIogy, Vol. 12B (Edited by GROSSMAN L. & MOLDAVE K.), pp. 195-206. Academic Press, New York. MANNING J. E., SCHMID C. W. • DAVIDSON N. 11975) Interspersion of repetitive and non-repetitive DNA sequences in the Drosophila melanogaster genome. Cell 4, 141-155. OHNO S. (1974) Animal Cytogenetics 4, Chordata 1, Protochordata, Cyclostomata, and Pisces, pp. 64-68. Gebriider Borntraeger, Berlin, Stuttgart. ROMER A. S. (1971) Veryleichende Anatomie der Wirbeltiere. Parey, Hamburg. RUSSELL G. J. & SUBAK-SHARPEJ. H. (1977) Similarity of the general designs of protochordates and invertebrates. Nature, Lond. 266, 533-535. SCHMIDTKE J., WEILER C., KUNZ B. & ENGEL W. (1977) Isozymes of a tunicate and a cephalochordate as a test of polyploidisation in chordate evolution. Nature, Lond. 266, 532-533. SCHMIDTKE J., KUNZ B. & ENGEL W. (1978) Gene numbers and genic variability in Ciona intestinalis and Branchiostoma lanceolatum. Can. J. Genet. Cytol. 20, 61-70. WELLS R., ROVER H.-D. & HOLLENBERG C. P. (1976) Non Xenopus-like DNA sequence organisation in the Chironomus tentans genome. Molec. Gen. Genet. 147, 45-51. ZIMMERMANJ. L. & GOLDBERG R. B. (1977) DNA sequence organisation in the genome of Nicotiana tabacum. Chromosoma 59, 227-252.