9.4—
Fractionation and Properties of DNA
9.4.1—
General Properties
Although single-stranded DNA does occur naturally most DNA molecules consist of two anti-parallel chains of nucleotides. Each nucleotide consists of deoxyribose, a phosphate group and a base. The four bases that occur in all types of DNA are the purines adenine and guanine and the pyrimidines thymine and cytosine. Adenine is capable of forming hydrogen bonds with thymine, and guanine with cytosine. Sometimes modified bases are present in small amounts and in plant nuclear DNA a substantial proportion of the cytosine residues are replaced by 5-methyl-cytosine. In double stranded DNA there is an equivalence between the amount of adenine and thymine and guanine and cytosine plus 5-methyl-cytosine. The two strands are joined together by hydrogen bonds between complementary pairs of bases to form a right handed helix. The bases are at right angles to the sugar residues, and project inwards, and the phosphate groups are on the outside. In the B form of DNA there are ten base pairs in one complete turn of the helix, which extend for 3.4 nm. The width of the helix is 1.8 nm. An alternative configuration is possible, the A form, where the bases are tilted and 11 pairs are accommodated per turn of the helix. Double stranded RNA and DNA-RNA hybrids adopt this latter configuration.
Many DNA molecules are known to be several millimetres in length. It is very difficult to determine the actual length of very large DNA molecules because they are extremely sensitive to shearing forces produced by various isolation procedures. It is known that the complete genome of E. coli, with a molecular
weight of 2.9 × 109 daltons, consists of a single circular DNA molecule of length approximately 1 mM . The total amount of DNA in a higher plant cell is equivalent to a length of approximately 1 m, but it is not known how many molecules this comprises.
The base composition of DNA is normally expressed as the per cent G + C content. This value, which is characteristic of a given species, varies widely in different organisms and in bacteria may be anything from 30% to 70%. The base compositions of a number of plant DNAs are given in Table 9.2.
9.4.2—
Buoyant Density Centrifugation
One of the most common methods of fractionating DNA is to exploit differences in density between different types of DNA in concentrated salt solutions. Caesium chloride is the most commonly used. The method involves prolonged high speed centrifugation of purified DNA in concentrated caesium chloride. The centrifugation produces a gradient of salt concentration in the centrifuge tube and the DNA migrates to occupy a position in the gradient which corresponds to its own buoyant density. A number of factors affect the buoyant density, such as the nature of the caesium salt, the presence of heavy metals or DNA-binding dyes, the pH and the temperature. Under constant conditions (usually 25°C in caesium chloride at neutral pH) the buoyant density of DNA is related to the GC content:
DNAs with different base compositions can therefore be separated by this method. Figure 9.4 shows the fractionation of Phaseolus aureus DNA (buoyant density = 1.695) and E. coli DNA (buoyant density = 1.710) in caesium chloride.
Most nuclear DNAs from higher plants have buoyant densities within the range 1.69–1.71 g cm–3 . However, the presence of 5-methyl-cytosine serves to reduce the density slightly, thereby giving rise to an under-estimate of the GC content. In general, 1% methylation decreases the buoyant density by 1 mg cm–3 . Certain sequences of bases may also distort the relationship between base composition and buoyant density. Furthermore, single-stranded DNA is denser than double-stranded DNA of similar base composition by approximately 0.015 g cm–3 and under alkaline conditions the density is increased by 0.06 g cm–3 . This is due partly to the fact that DNA becomes single-stranded under these conditions and also partly because the deprotonated adenine and thymine residues are neutralized by binding caesium ions.
9.4.3—
Satellite DNA
Figure 9.4 shows that DNA from a particular organism often forms an apparently homogenous band in caesium chloride. The width of the band depends
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Figure 9.4
The separation of a mixture of E. coli and P. aureus DNA in caesium chloride.
A trace of E. coli DNA labelled with 3 H thymidine was mixed with P. aureus DNA
and centrifuged at 30,000 rpm at 25ºC in caesium chloride for 3.5 d using a fixed-angle
centrifuge rotor. Three-drop fractions were collected and the P. aureus DNA located
by absorbance at 260 nm (solid circles, continuous curve) and E. coli DNA detected by
measurement of radioactivity (open circles, dotted curve). (Unpublished result of D. Grierson.)
upon the slope of the density gradient, the molecular weight of the DNA and the difference in base composition between different regions of the DNA. In most plants certain regions of the genome are sufficiently different in base composition for them to occupy a position distinct from the main-band DNA. If they represent only a small percentage of the total DNA such sequences are difficult to detect but when present in large amounts they produce one or more peaks of satellite DNA. Melon DNA provides an example where the satellite is denser than the main-band (Fig. 9.5) but light satellites have also been observed, for example in flax. It has been assumed that the majority of satellite DNAs are nuclear in origin but recent studies with cucumber suggest that they may sometimes represent organelle DNA (Kadouri et al., 1975). Although they are widely distributed in plants no satellites have so far been found in monocotyledonous plants using caesium chloride fractionation (Table 9.3). However similar components can be detected using other techniques. For example, substances such as heavy metals and dyes often selectively bind to certain sequences within the DNA. This alters their buoyant density and thus generates satellites. One common method is to fractionate DNA in caesium sulphate gradients in the presence of silver ions. Using this procedure, Ingle (unpublished) has demonstrated the presence of a satellite DNA in Scilla (a monocotyledon) although in caesium chloride no such satellite is observed.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Figure 9.5
The separation of a mixture of E. coli and melon DNA in caesium chloride.
The experimental details were as described in the legend to Fig. 9.4. Melon
DNA was detected by absorbance at 260 nm (solid circles, continuous curve)
and E. coli DNA by measurement of radioactivity (open circles, dotted curve).
(Unpublished result of D. Grierson.)
9.4.4—
Dissociation and Reassociation of DNA
When double-stranded DNA in solution is heated above a certain temperature the hydrogen bonds between complementary base pairs are broken and the two strands dissociate. The process, which can also be brought about by treatment of DNA with alkali, is commonly referred to as melting or denaturation. Strand separation is accompanied by an increase of approximately 35% in the absorbance of ultra-violet light. This hyperchromic effect provides a means of monitoring the dissociation of double-stranded DNA by measuring the increase in absorbance of a DNA solution at 260 nm as the temperature is raised. The melting curves of DNA from a variety of sources are compared in Fig. 9.6. The temperature at which 50% of the DNA is melted, the Tm, depends upon a number of factors such as the salt concentration and pH. Under standard conditions, i.e. when the DNA is dissolved in standard saline citrate (S.S.C. being 0.15 M sodium chloride, 0.015 M trisodium citrate, pH 7) the base composition and the Tm are related in the following way:
In viruses, the majority of sequences within the genome are similar in base composition and for a given DNA a sharp melting profile is observed with a characteristic Tm. In contrast, higher organism DNA often consists of distinct subfractions, each with a slightly different Tm. This produces a broad melting
Figure 9.6
Melting curves of various DNA samples. (1) Musk melon main-
band and (4) satellite DNA. (3) a 2:1 mixture of DNA from bacterium
C and phage P58. (2) B. subtilis DNA. Melting was carried out in S.S.C.
(From Bendich & Anderson, 1974.)
profile which represents a number of overlapping melting curves. With DNA samples from organisms that contain prominent satellite bands in caesium chloride the melting curve often has two or more phases. Even where no large satellites are present heterogeneity within the DNA can readily be detected by plotting the rate of change of absorbance at 260 nm of a DNA solution as a function of temperature.
If, following denaturation of DNA by alkali or by heating, the pH of the solution is restored to neutrality or the temperature reduced below the Tm, complementary strands of DNA show a marked tendency to recombine to form double-stranded molecules. This process of duplex formation is termed renaturation or reassociation. The rate of the reaction is governed by the concentration of DNA sequences, pH, temperature and salt concentration. Renaturation is normally carried out at moderately high salt concentrations at a pH close to neutrality and at a temperature approximately 25°C below the Tm. Under these conditions, single-stranded DNA molecules collide at random. Most collisions occur in such a way that complementary sequences of bases are not in register and the strands do not reassociate. If, however, two strands approach in an anti-parallel configuration and potential regions of complementarity are in close proximity a nucleation event will take place and hydrogen bonds will form. Whether or not the strands remain together depends upon the nature of the neighbouring bases on either side of the nucleation site. If complementary sequences are arranged in the correct register along the length of the molecule a rapid zippering effect takes place to produce a stable double-stranded DNA molecule.
DNA renaturation can be followed by measuring any property that distinguishes single-stranded from double-stranded DNA. The decrease in absorbance at 260 nm that accompanies duplex formation is often exploited for
this purpose but fractionation by hydroxyapatite is also commonly used. Under appropriate conditions only double-stranded DNA binds to hydroxyapatite (a hydrated form of calcium phosphate); single strands pass straight through a column and double-stranded material may subsequently be eluted either by increasing the salt concentration or by raising the temperature above the Tm. This procedure has the advantage that the properties of the single-stranded and renatured fractions may subsequently be studied separately.
9.4.5—
Repeated Sequence DNA
Information about the sequence content of DNA can be obtained by studying the renaturation rate. Double-stranded DNA is first sheared to lengths of a few hundred nucleotides by ultrasonication or by forcing through a narrow gauge needle. This is designed to liberate short sequences of DNA and to allow them subsequently to react independently of neighbouring sequences. The DNA is then denatured and allowed to renature at the desired concentration under controlled conditions. Renaturation is followed by monitoring the absorbance of the solution at 260 nm in a spectrophotometer cell maintained at the appropriate temperature, or alternatively by taking samples at intervals and fractionating them by hydroxyapatite chromatography. In some situations partially double-stranded segments of DNA are produced, which have single-stranded 'tails'. These can be removed by treating the reassociated DNA with SI nuclease, which is specific for single-stranded DNA, before hydroxyapatite fractionation.
The reassociation reaction follows second order kinetics, which means that it is governed by the concentration of the reacting sequences. This is determined by the DNA concentration and the sequence content. For example, synthetic poly-dA. dT renatures almost instantaneously, whereas at the sme concentration naturally occurring DNA takes much longer to renature. In general the renaturation rate is inversely proportional to the number of base pairs in the genome (providing that no repetitive sequences are present). In principle, therefore, the number of 'genes' or sequences may be estimated by comparing the renaturation rate of an unknown DNA with that from an organism with a well characterized genome. The results of renaturation experiments are often expressed as a 'Cot' plot. In this form the extent of DNA renaturation is plotted as a function of the initial concentration of the DNA (in moles of nucleotides per litre) multiplied by the renaturation time (in seconds) plotted on a log scale. One advantage of this is that different types of DNA can easily be compared by their Cot 1/2 values (the value of Cot at which 50% of the DNA becomes renatured) (Fig. 9.7a). For a variety of microorganisms and viruses the Cot 1/2 is related to the genome size. However, this is not strictly true for eukaryotic DNA which behaves in a more complex way.
In bacteria and viruses most DNA sequences are represented only once in the genome but in higher organisms many sequences occur in multiple copies. For this reason a plot of the renaturation of higher plant DNA deviates markedly
Figure 9.7
Reassociation kinetics of various DNA samples. (a) Musk melon satellite
(sat.), phage PS8 (PS8), Bacillus subtilis (B. sub. ), phage T4 (T4). The data
are corrected for differences in GC content and to take account of different
solvent concentrations. (b) Helianthus annuus DNA. The slowest reassociating
fraction behaves approximately as expected for sequences present in one copy per
haploid genome. (a) redrawn from Bendich and Anderson, 1974; (b) from Flavell et al., 1974.)
from a second order curve (Fig. 9.7b). A substantial proportion of the genome is so highly repetitive that it renatures even more rapidly than bacterial DNA. A second, intermediate, fraction contains families of repetitive sequences present in a lower frequency. Finally, there exists a unique fraction of DNA which essentially comprises sequences present in only one or a small number of copies in the genome. Flavell et al., (1974) have studied the reassociation kinetics of the DNA from 23 higher plant species, which vary in 2C nuclear DNA content between 1.7 and 98 × 10–12 g. The results, summarized in Table 9.4, show that in plants with relatively large genomes an average of 80% of the DNA consists of sequences present in 100 copies or more. For plants with smaller genomes
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
the average amount of repetitve DNA is 62%. These results indicate that part of the variation in nuclear DNA mass can be accounted for by variation in the amount of repeated sequence DNA.