DNA persistence length revisited

DNA restriction fragments ranging from 79 to 789 base pairs in length have been characterized by transient electric birefringence (TEB) measurements at various temperatures between 4 and 43 degrees C. The DNA fragments do not contain runs of four or more adenine residues in a row and migrate with normal electrophoretic mobilities in polyacrylamide gels, indicating that they are not intrinsically curved or bent. The low ionic strength buffers used for the measurements contained 1 mM Tris Cl, pH 8.0, EDTA, and variable concentrations of Na(+) or Mg(2+) ions. The rotational relaxation times were obtained by fitting the TEB field-free decay signals with a nonlinear least-squared fitting program; the decay of the birefringence was monoexponential for fragments < or = 241 base pair (bp) in length and multiexponential for larger fragments. The terminal relaxation times, characteristic of the end-over-end rotation of the DNA molecules, were then used to determine the persistence length (p) and hydrodynamic radius (r) of DNA as a function of temperature and ionic strength, using several different hydrodynamic models. The specific values obtained for p and r are model dependent. The wormlike chain model of P. J. Hagerman and B. H. Zimm (Biopolymers 1981, Vol. 20, pp. 1481-1502) combined with the revised Broersma equation (J. Newman et al., Journal of Mol Biol 1997, Vol. 116, pp. 593-606) appears to be the most suitable for describing the flexibility of DNA in low ionic strength solutions. The values of p and r obtained from the global least squares fitting of this equation are independent of DNA length, and the deviations of the individual values from the average are reasonably small. The consensus r value calculated for DNA in various low ionic strength solutions containing 1 mM Tris buffer is 14.7 +/- 0.4 A at 20 degrees C. The consensus p values decrease from 814 approximately 564 A in solutions containing 1 mM Tris buffer plus 0.2-1 mM NaCl and decrease still further to 440 A in solutions containing 0.2 mM Mg(2+) ions. The persistence length exhibits a shallow maximum at 20 degrees C and decreases slowly upon either increasing or decreasing the temperature, regardless of the model used to fit the data. By contrast, the consensus values of the hydrodynamic radius are independent of temperature. The calculated persistence lengths and hydrodynamic radii are compared with other data in the literature.

Do orientation effects contribute to the molecular weight dependence of the free solution mobility of DNA?

The free solution mobility of DNA increases with increasing molecular weight and then levels off and becomes constant at molecular weights above approximately 400 bp (Stellwagen, N. C., Gelfi, C., Righetti, P. G., Biopolymers 1997,42, 687-703). To investigate whether the increase in mobility could be attributed to an increased orientation of the larger DNA molecules in the electric field, the free solution mobility of DNA was measured by capillary electrophoresis as a function of electric field strength. Mixtures containing 20-, 118- and 422-bp DNA molecules, and 20-, 422- and 2116-bp DNAs, were studied. If the larger DNA molecules in each mixture were oriented by the electric field, their mobilities should increase with electric field strength faster than the mobility of the 20-bp oligomer, which is too small to be oriented by the electric fields used in this study. Instead, the ratios of the mobilities of the 118-, 422- and 2116-bp fragments to the mobility of the 20-bp oligomer were independent of electric field strength. Hence, orientation effects are not important for DNA molecules up to 2 kbp in size, in electric fields up to 500 V/cm in amplitude. An explanation is suggested.

Measuring the translational diffusion coefficients of small DNA molecules by capillary electrophoresis

The apparent translational diffusion coefficients of four 20 base pair (bp) DNA oligonucleotides with different sequences have been measured by capillary electrophoresis, using the stopped migration method. The diffusion coefficients of the four oligomers were equal within experimental error, and averaged (120 +/- 10) x 10(-8) cm(2) s(-1) in 40 mM Tris-acetate-EDTA buffer at 25 degrees C. Since this value is nearly identical to the translational diffusion coefficient determined for a different 20-bp oligomer using other methods, the stopped migration method can accurately measure the diffusion coefficients of small DNA oligomers. The apparent diffusion coefficient of a 118-bp DNA restriction fragment was also measured by the stopped migration method. However, the observed value was approximately 25% larger than expected from other measurements, possibly because the diffusion coefficients of larger DNA molecules are somewhat dependent on the ionic strength of the solution.

Preferential counterion binding to A-tract DNA oligomers

The free solution mobility of four 20 bp DNA oligomers, with and without A-tracts, has been measured by capillary electrophoresis in Tris-acetate buffer, to test the hypothesis that site-specific binding of monovalent counterions can occur in the narrow minor groove of A-tract DNAs. Preferential counterion binding has been proposed to cause A-tract bending because of asymmetric charge neutralization and collapse of the helix backbone toward the minor groove. Preferential counterion binding in A-tract DNAs should be manifested by a decrease in the electrophoretic mobility observed in free solution, compared to that of non-A-tract DNAs of the same size. Of the four sequences studied here, the slowest absolute mobility, indicative of the greatest counterion binding, was observed for a 20 bp oligomer containing two runs of A3T3 in phase with the helix repeat. A 20-mer containing phased CACA sequences migrated with the fastest mobility; 20-mers containing phased A5 tracts or phased runs of T3A3 migrated with intermediate mobilities. Very similar mobility differences were observed when 1-20 mM NaCl was added to the buffer. The results suggest that preferential counterion binding occurs in A-tract DNAs, especially those containing the AnTn sequence motif.

DNA and buffers: are there any noninteracting, neutral pH buffers?

The interaction of DNA with various neutral pH, amine-based buffers has been analyzed by free solution capillary electrophoresis, using a mixture of a plasmid-sized DNA molecule and a small DNA oligonucleotide as the reporter system. The two DNAs migrate as separate, nearly Gaussian-shaped peaks in 20-80 mM TAE (TAE, Tris-acetate-EDTA; Tris, tris[hydroxymethyl]aminomethane) buffer. The separation between the peaks gradually increases with increasing TAE buffer concentration because of differences in solvent friction between large and small DNA molecules. The two DNAs form complexes with the borate ions in TBE (Tris-borate-EDTA) buffer, with mobilities that depend on the DNA/borate ratio. In 45 mM TBE buffer, the two DNAs comigrate as a single sharp peak, with a mobility that is faster than either of the constituent DNAs in the same buffer. Hence, the mixed DNA-borate complex is stabilized by the binding of additional borate ions, possibly forming bridges between the different DNAs. The mixed DNA-borate complex is gradually dissociated into its component DNAs by increasing the TBE concentration, possibly because the borate binding sites become saturated at high buffer concentrations. Other neutral pH, amine-based buffers, such as Mops (3-[N-morpholino]propanesulfonic acid), Hepes (N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid]), Bes (N,N-bis[2-hydroxyethyl]-2-aminoethanesulfonic acid), Tes (N-tris[hydroxymethyl]methyl-2-aminoethanesulfonic acid), and tricine (N-tris[hydroxymethyl]methylglycine) also form complexes with DNA, giving distorted peaks in the electropherograms. The combined results indicate that borate buffers and most neutral pH, amine-based buffers interact with DNA.

DNA and buffers: the hidden danger of complex formation

The free solution electrophoretic mobility of DNA differs significantly in different buffers, suggesting that DNA-buffer interactions are present in certain buffer systems. Here, capillary and gel electrophoresis data are combined to show that the Tris ions in Tris-acetate-EDTA (TAE) buffers are associated with the DNA helix to approximately the same extent as sodium ions. The borate ions in Tris-borate-EDTA (TBE) buffers interact with DNA to form highly charged DNA-borate complexes, which are stable both in free solution and in polyacrylamide gels. DNA-borate complexes are not observed in agarose gels, because of the competition of the agarose gel fibers for the borate residues. The resulting agarose-borate complexes increase the negative charge of the agarose gel fibers, leading to an increased electroendosmotic flow of the solvent in agarose-TBE gels. The combined results indicate that the buffers in which DNA is studied cannot automatically be assumed to be innocuous.

Conformational isomers of curved DNA molecules can be observed by polyacrylamide gel electrophoresis

Dimers, trimers and higher multimers of two 147-base pair restriction fragments called 12 A and 12B, obtained from the MspI digest of plasmid pBR322, migrate as sharp bands in agarose and dilute polyacrylamide gels, indicating that they are homogeneous in molecular weight. However, the electrophoretic bands corresponding to multimers of the curved fragment 12A are split into sharp sub-bands in more concentrated polyacrylamide gels. The relative intensities and spacing of the sub-bands depend on the number of monomers in the multimer, the pH of the buffer, and the presence or absence of divalent cations in the solution. Since band splitting is not observed for the normal 12B multimers under any gel-running conditions, the sub-bands observed for multimers of the curved fragment 12A must be attributed to conformational isomers which are in slow exchange on the electrophoretic time scale. Band splitting is also observed for multimers of a curved DNA fragment containing the kinetoplast bending locus and for plasmid pUC19 linearized by digestion with certain restriction enzymes. Plasmid pUC19 contains two nearly equidistant regions of intrinsic curvature (Strutz, K., Stellwagen, N. C., Electrophoresis 1996, 17, 989-995). Hence, DNA molecules containing two or more regions of curvature exist as discrete subpopulations of conformational isomers which can be observed as separate bands migrating in polyacrylamide gels.

Free solution mobility of DNA molecules containing variable numbers of cationic phosphoramidate internucleoside linkages

The free solution electrophoretic mobility of an 118-base pair DNA fragment containing zero, three, six or nine cationic phosphoramidate internucleoside linkages has been measured by capillary electrophoresis. The electrophoretic mobility decreases with the increasing number of cationic phosphoramidate linkages, as expected because of the reduced negative charge on the DNA molecules. The decrease in mobility is approximately linear for DNA molecules containing three and six cationic phosphoramidate linkages, but begins to level off when nine cationic phosphoramidate linkages have been added. The mobility also varies somewhat depending on whether the modified phosphoramidate linkages are located at the 5'- or 3'-end of the DNA molecule.

DNA-histidine complex formation in isoelectric histidine buffers

The free solution electrophoretic mobility of two DNA molecules of different molecular masses, 18 base pairs and 2686 base pairs, has been measured in isoelectric histidine buffers with and without added low-molecular-mass electrolytes. Extensive DNA-histidine complex formation is observed in isoelectric histidine buffer, as evidenced by distortion and splitting of the peaks in the electropherograms. Peak distortion and splitting can be decreased or eliminated by adding low-molecular-mass neutral salts to the solution, suggesting that the DNA-histidine complexes are stabilized by electrostatic interactions. The ability of various neutral salts to disrupt the DNA-histidine complexes depends on the molecular mass of the DNA and the concentration and type of added salt.

Apparent pore size of polyacrylamide gels: comparison of gels cast and run in Tris-acetate-EDTA and Tris-borate-EDTA buffers

The electrophoretic mobilities of DNA molecules in three different molecular weight ladders were measured in polyacrylamide gels containing different acrylamide concentrations (%T) and cross-linker ratios (%C), cast and run in Trisacetate-EDTA (TAE) buffer. The apparent pore radius of each gel was estimated from Ferguson plots of the relative mobilities of each of the DNA molecules, using the mobility of the monomer fragment in each molecular weight ladder as the reference mobility. The effective size of each of the DNA molecules was estimated from its radius of gyration. The apparent gel pore radii calculated in this manner ranged from 21 nm in gels containing 10.5%T, 5%C to 200 nm in gels with 4.6%T, 2%C, similar to the values observed for polyacrylamide gels cast and run in Tris-borate-EDTA (TBE) buffer (Holmes and Stellwagen, Electrophoresis 1991, 12, 612-619). Hence, the effective pore size of polyacrylamide gels is essentially independent of whether the gels are cast and run in TAE or TBE buffer.

Capillary zone electrophoresis of ds-DNA in isoelectric buffers: effect of adding of competing, nonamphoteric ions

When separating ds-DNA in isoelectric His buffer (pH=pI=7.6), in the 50-250 mM concentration range, some unique phenomena were observed: improved resolution for smaller DNA fragments, up to ca. 150 bp, and a rapid deterioration of resolution above this critical length (which corresponds to the persistence length). Such phenomenon depended also on voltage and concentration of sieving liquid polymer. Direct binding of His to the DNA helix was hypothesized, with resultant stiffening and an increment of diameter of the DNA fragments, thus inducing an early onset of reptation at the applied voltage in the 100-300 V/cm range. In order to prove this hypothesis, "competing ions" (notably NaCl and KBr) were added to the His background electrolyte: a partial reversal of the His effect was already apparent at low concentrations of such ions (10 mM) and was complete at higher concentrations (30 and 50 mM). By molecular modeling, it was found that His could be docking on the negatively charged oxygen (bound to the phosphate) by offering both charged (primary and tertiary amino) groups to simultaneous binding, thus forming a salt and neutralizing the negative charge borne by the oxygen. The following characteristic bond distances were found: 0.34 nm between the N (imidazolic) and O; 0.32 nm between the primary N and O; 0.36 nm between the two nitrogens engaged in salt formation with the oxygen. In addition, for complexation to occur, the distance between the noncharged nitrogen in the imidazole ring and the nearest phosphate oxygen (engaged in the phosphodiester bridge) should be 0.44 nm. Under these conditions, the two rings present (a six-membered, ideal one, salt-linked with the oxygen and rather highly elongated, and the imidazole) will not be precisely coplanar, since the primary and tertiary nitrogens will be one slightly above and one slightly below the plane of the drawing. Upon extensive binding, occupying every available phosphate site, pi-pi interactions could occur among the stacks of bound His residues, thus further stabilizing the complex.

Do DNA gel electrophoretic mobilities extrapolate to the free-solution mobility of DNA at zero gel concentration?

The electrophoresis of small DNA fragments has been measured in dilute agarose and polyacrylamide gels cast and run in Tris-acetate-EDTA (TAE) and Tris-borate-EDTA (TBE) buffers. Ferguson plots were constructed to extrapolate the mobilities to zero gel concentration and estimate the free solution mobility of DNA. In polyacrylamide gels, in both TAE and TBE buffers, the extrapolated mobilities at zero gel concentration increased gradually with decreasing DNA molecular weight, went through a maximum at approximately 60 bp, and then decreased again. The increase in the extrapolated mobilities with decreasing molecular weight observed for DNA fragments > or = 60 bp can be attributed to transient interactions between the migrating DNA molecules and the polyacrylamide gel fibers. If such interactions are eliminated by extrapolating the mobilities to both zero gel concentration and zero DNA molecular weight, the apparent free solution mobility of DNA is found to be 3.1 x 10(-4) cm2 V(-1) s(-1) in TAE buffer and 4.2 x 10(-4) cm2 V(-1) s(-1) in TBE buffer at 20 degrees C, reasonably close to the actual free solution mobilities measured in the same two buffers by capillary electrophoresis (N. C. Stellwagen et al., Biopolymers 1997, 42, 687-703). The significantly larger electrophoretic mobility observed in TBE buffer is most likely due to the formation of nonspecific, highly charged deoxyribose-borate complexes in this buffer medium. For DNA molecules < or = 60 bp in size, the decrease in the extrapolated mobilities with decreasing molecular weight parallels the decrease in their free solution mobilities observed by capillary electrophoresis. In agarose gels, the extrapolated mobilities of small DNA molecules at zero gel concentration appear to be independent of molecular weight. The apparent free solution mobilities are found to be (3.0 +/- 0.1) x 10(-4) cm2 V(-1) s(-1) in TAE buffer and (3.2 +/- 0.1) x 10(-4) cm2 V(-1) s(-1) in TBE buffer. The very similar mobilities observed in the two buffer media suggest that the borate ions in TBE buffer primarily form complexes with the galactose residues in the agarose gel fibers, rather than with the migrating DNA molecules, because of mass action effects. The formation of borate-agarose complexes, increasing the net negative charge of the agarose gel fibers, appears to be responsible for the markedly increased electroendosmotic flow observed in agarose gels cast and run in TBE buffer (N. C. Stellwagen, Electrophoresis 1992, 13, 601-603).

The free solution mobility of DNA

The free solution mobility of DNA has been measured by capillary electrophoresis in the two buffers most commonly used for DNA gel electrophoresis, Tris-borate-EDTA (TBE) and Tris-acetate-EDTA (TAE). The capillaries were coated with polymers of either of two novel acrylamide monomers, N-acryloylaminoethoxyethanol or N-acryloylaminopropanol, both of which are stable at basic pH and effectively eliminate the electroendosmotic mobility due to the capillary walls. The free solution mobility of DNA in TAE buffer was found to be (3.75 +/- 0.04) x 10(-4) cm2 V-1 s-1 at 25 degrees C, independent of DNA concentration, sample size, electric field strength, and capillary coating, and in good agreement with other values in the literature. The free solution mobility was independent of DNA molecular weight from approximately 400 base pairs to 48.5 kilobase pairs, but decreased monotonically with decreasing molecular weight for smaller fragments. Surprisingly, the free solution mobility of DNA in TBE buffer was found to be (4.5 +/- 0.1) x 10(-4) cm2 V-1 s-1, about 20% larger than observed in TAE buffer, presumably because of the formation of nonspecific borate-deoxyribose complexes.

DNA mobility anomalies are determined primarily by polyacrylamide gel concentration, not gel pore size

The dependence of DNA mobility anomalies on gel pore size has been studied in polyacrylamide gels with a wide variety of compositions, using molecular weight ladders containing multiple copies of normal (12B) and anomalously slowly migrating (12A) 147-base pair restriction fragments from plasmid pBR322 as the migrating probe molecules. If the gel pore size is increased by decreasing the total acrylamide concentration (%T) at constant cross-linker ratio (%C), the usual method of increasing gel pore size, the mobility anomalies decrease with increasing gel pore radius as though the 12A multimers were retarded by a sieving mechanism. However, the decrease in the mobility anomalies is independent of whether the apparent gel pore radius is larger or smaller than the DNA radius of gyration, suggesting that gel pore size is not the controlling variable. If the acrylamide concentration is held constant and the gel pore size is increased by decreasing %C at constant %T, the mobility anomalies of the largest 12A multimers (6 mers and higher) decrease with increasing gel pore radius, because of sieving effects, until the effective gel pore radius becomes approximately equal to the DNA radius of gyration, after which the mobility anomalies level off and become independent of gel pore size. The mobility anomalies exhibited by 5-mers and smaller multimers of fragment 12A are independent of gel pore radius in all gels with constant %T. Similar results are observed with a molecular weight ladder containing phased A-tracts from the kinetoplast bending locus. Since the anomalous electrophoretic mobilities depend primarily on the total acrylamide concentration in the gel, and not on the apparent gel pore radius, increases in the magnitude of the mobility anomalies with increasing gel concentration (and decreasing gel pore radius) cannot be taken as evidence for DNA curvature.

Intrinsic curvature of plasmid DNAs analyzed by polyacrylamide gel electrophoresis

The electrophoretic mobility of two small DNA plasmids, pUC19 and Litmus 28, linerarized by digestion with a variety of single-cut restriction enzymes, has been studied. The permuted sequence isomers migrate with identical mobilities in agarose gels, as expected, but exhibit different mobilities in large-pore polyacrylamide gels, suggesting that the parent plasmids contain sequence-specific sites of curvature and/or anisotropic flexibility. Both plasmids contain apparent bend centers near their origin(s) of replication; pUC19 also has a major apparent bend center near the promoter of the ampicillin resistance gene. These apparent bend centers are observed under a variety of experimental conditions, suggesting that they correspond to sites of stable curvature in the parent plasmids. Both plasmids also contain minor bend centers that are observed under a sub-set of electrophoretic conditions and disappear when divalent cations are added to the solution, suggesting that these apparent bend centers may correspond to localized regions of variable flexibility.

Electric birefringence of kilobase-sized DNA molecules

Transient electric birefringence has been used to characterize the rotational diffusion of linear, circularly permuted pBR322 and SV40 DNA molecules. The birefringence relaxation times vary with the site of linearization, suggesting that the circularly permuted DNAs have different conformations in solution. The longest relaxation times are observed for DNA sequence isomers linearized at the major bend centers identified by gel electrophoresis. SV40 sequence isomers linearized at other locations have faster, but approximately equal, terminal relaxation times, suggesting that their free solution conformations are relatively independent of the location of the bend center within the sequence. By contrast, the terminal relaxation times of the various pBR322 sequence isomers vary approximately in accord with their electrophoretic mobilities in large-pore polyacrylamide gels, suggesting that the different mobilities may reflect real conformational differences between the sequence isomers.

Effect of spike pulses on the orientation of the agarose gel matrix

The orientation of the agarose gel matrix in two-part, "stair-step" electric fields has been studied by transient electric birefringence. Stair-step electric fields are those in which a pulse of a given amplitude is immediately followed by a short, higher voltage "spike" pulse of the same polarity. A single stair-step pulse orients the agarose gel matrix as though the two portions of the pulse were individually applied to the gel. However, a series of consecutive stair-step pulses causes an anomalous increase in the amplitude of the birefringence, suggesting that increased numbers of agarose fiber bundles are orienting in the electric field. Spike pulses > or = 10 V/cm appear to cause junction zone breakdown, freeing large numbers of agarose fiber bundles and microgel domains from the constraints of the gel matrix. The implications of these results for pulsed field gel electrophoresis are discussed.

Use of polyacrylamide gel electrophoresis to detect structural variations in kilobase-sized DNAs

The electrophoresis of linear, kilobase-sized DNA molecules with permuted sequences has been studied in polyacrylamide and agarose gels. Plasmid pBR322, bacteriophage phi X174, and the SV40 minichromosome were each digested with a series of single-cut restriction enzymes. The linearized, permuted isomers of all three DNAs exhibit different mobilities in large-pore polyacrylamide gels, suggesting that all three DNAs contain sites of anisotropic, sequence-dependent curvature. Various experimental parameters such as acrylamide concentration, crosslinker ratio and buffer composition affect the magnitude of the observed differential mobilities. Band sharpness appears to be optimal in polyacrylamide gels containing 6.9-8.1%T and 0.5-1%C. Only small mobility differences are observed for the linearized, permuted sequence isomers in agarose gels.

Transient electric birefringence of agarose gels. II. Reversing electric fields and comparison with other polymer gels

The transient electric birefringence of low electroendosmosis (LE) agarose gels oriented by pulsed unidirectional electric fields was described in detail in Part I [J. Stellwagen and N. C. Stellwagen (1994), Biopolymers, Vol. 34, p. 187]. Here, the birefringence of LE agarose gels in rapidly reversing electric fields, similar in amplitude and duration to those used for field inversion gel electrophoresis, is reported. Symmetric reversing electric fields cause the sign of the birefringence of LE agarose gels, and hence the direction of orientation of the agarose fibers, to oscillate in phase with the applied electric field. Because of long-lasting memory effects, the alternating sign of the birefringence appears to be due to metastable changes in gel structure induced by the electric field. If the reversing field pulses are equal in amplitude but different in duration, the orientation behavior depends critically on the applied voltage. If E < 7 V/cm, the amplitude of the birefringence gradually decreases with increasing pulse number and becomes unmeasurably small. However, if E > 7 V/cm, the amplitude of the birefringence increases more than 10-fold after approximately 20 pulses have been applied to the gel, suggesting that a cooperative change in gel structure has occurred. Because there is no concomitant change in the relaxation times of the orienting particles, the large increase in the amplitude of the birefringence must be due to an increase in the number of agarose fibers and/or fiber bundles orienting in the electric field, which in turn indicates a cooperative breakdown of the noncovalent "junction zones" that cross-link the fibers into the gel matrix. The sign of the birefringence of LE agarose gels is always positive after extensive junction zone breakdown, indicating that the agarose fibers and fiber bundles preferentially orient parallel to the electric field when they are freed from the constraints of the gel matrix. Three other gel-forming polymers, high electroendosmosis (HEEO) agarose (a more highly charged agarose), beta-carrageenan (a stereoisomer of agarose), and polyacrylamide (a chemically cross-linked polymer) were also studied in unidirectional and rapidly reversing electric fields. The birefringence of HEEO agarose gels in reversing fields is very similar to that of LE agarose gels, suggesting that the orientation anomalies are not due to the occasional charged residues on the agarose backbone chain. The beta-carrageenan gels exhibit variable orientation behavior in reversing electric fields, suggesting that its internal gel structure is not as tightly interconnected as that of agarose gels.(ABSTRACT TRUNCATED AT 400 WORDS)

Transient electric birefringence of agarose gels. I. Unidirectional electric fields

The orientation of agarose gels in pulsed electric fields has been studied by the technique of transient electric birefringence. The unidirectional electric fields ranged from 2 to 20 V/cm in amplitude and 1 to 100 s in duration, values within the range typically used for pulsed field gel electrophoresis (PFGE). Agarose gels varying in concentration from 0.3 to 2.0% agarose were studied. The sign of the birefringence varied randomly from one gel to another, as described previously [J. Stellwagen & N.C. Stellwagen (1989), Nucleic Acids Research, Vol. 17, 1537-1548]. The sign and amplitude of the birefringence also varied randomly at different locations within each gel, indicating that agarose gels contain multiple subdomains that orient independently in the electric field. Three or four relaxation times of alternating sign were observed during the decay of the birefringence. The various relaxation times, which range from 1 to approximately 120 s, can be attributed to hierarchies of aggregates that orient in different directions in the applied electric field. The orienting domains range up to approximately 22 microns in size, depending on the pulsing conditions. The absolute amplitude of the birefringence of the agarose gels increased approximately as the square of the electric field strength. The measured Kerr constants are approximately 5 orders of magnitude larger than those observed when short, high-voltage pulses are applied to agarose gels. The increase in the Kerr constants in the low-voltage regime parallels the increase in the relaxation times in low-voltage electric fields. Birefringence saturation curves in both the low- and high-voltage regimes can be fitted by theoretical curves for permanent dipole orientation. The apparent permanent dipole moment increases approximately as the 1.6 power of fiber length, consistent with the presence of overlapping agarose helices in the large fiber bundles orienting in low-voltage electric fields. The optical factor is approximately independent of fiber length. Therefore, the marked increase in the Kerr constants observed in the low-voltage regime is due to the large increase in the electrical orientation factor, which is due in turn to the increased length of the fiber bundles and domains orienting in low-voltage electric fields. Since the size of the fiber bundles and domains approximates the size of the DNA molecules being separated by PFGE, the orientation of the agarose matrix in the applied electric field may facilitate the migration of large DNA molecules during PFGE.

"Flip-flop" orientation of agarose gel fibers in pulsed alternating electric fields

The orientation of the agarose gel matrix in pulsed electric fields has been studied by transient electric birefringence. Two types of agarose with different degrees of charge were studied, in addition to agarose solutions and gels containing beta-carrageenan, a stereoisomer of agarose, and polyacrylamide. Agarose gels exhibit normal orientation behavior when short, high voltage pulses are applied to the gel. The sign of the birefringence is positive and the relaxation times are consistent with the orientation of dangling fiber ends parallel to the electric field. When long, low voltage pulses, of the amplitude and duration used for pulsed field gel electrophoresis, are applied to the gel, completely different orientation effects are observed. The amplitude of the birefringence (i.e., extent of orientation) is much larger than expected from the high field results, and the birefringence decay curves contain multiple components of opposite sign. The relaxation times are consistent with the orientation of long agarose chain bundles or fibers, as well as large three-dimensional domains. Chain bundles or fibers of the lengths observed in the agarose gels are also observed in agarose solutions, suggesting that the fibers that are free to orient in the gels had previously formed in the sol phase and are only weakly integrated into the matrix structure. In rapidly reversing low voltage electric fields, the sign of the birefringence of the agarose gels reverses from positive to negative in phase with the reversing electric field. This alternating change in the sign of the birefringence suggests that the agarose fibers "flip-flop" in orientation from parallel to perpendicular every time the electric field reverses its direction. Similar effects are observed for agarose gels with different charge densities. The flip-flop orientation and reorientation of agarose fibers within the matrix in reversing electric fields may decrease the microscopic viscosity of the gel, increasing the mobility of large DNA molecules migrating through the gel during electrophoresis. Polyacrylamide gels do not exhibit an anomalous reversal of the sign of the birefringence in reversing electric fields. Hence, the orienting fibers in these gels do not change their direction of orientation in reversing electric fields. Extensive orientation is observed in beta-carrageenan gels, similar to that observed in agarose gels. However, little orientation occurs in polyacrylamide gels, which are chemically crosslinked.(ABSTRACT TRUNCATED AT 400 WORDS)

Agarose gel pore radii are not dependent on the casting buffer

Previous studies have shown that the apparent pore size of agarose gels is dependent on the buffer in which the gel is cast and run (D.L. Holmes and N.C. Stellwagen, Electrophoresis 1990, 11, 5-15; N.C. Stellwagen and D.L. Holmes, Electrophoresis 1990, 11, 649-652). However, these studies, based on the mobility of DNA restriction fragments, neglected the effect of electroendosmosis. By measuring the mobility of vitamin B12 under various experimental conditions, it is shown here that electroendosmosis is highly buffer-dependent. When the observed mobilities of DNA are corrected for electroendomosis, the apparent pore radii of agarose gels are found to be independent of the casting buffer.

The effect of gel structure on matrix orientation

Four polymeric gels of different structure, low electroendosmosis (LE) agarose, highest electroendosmosis (HEEO) agarose, beta-carrageenan, and polyacrylamide, were studied by transient electric birefringence to determine the importance of various structural features on the orientation of the gels in an electric field. The two types of agarose, but not the polyacrylamide or beta-carrageenan, exhibited anomalous orientation effects. Both agarose and beta-carrageenan exhibited large birefringence signals, suggesting that the noncovalent hydrogen bonds joining the agarose fibers within the matrix allow the high degree of orientation of the gel. The spatial arrangement of the sugars of the agarose backbone is necessary for the anomalous orientation effects in reversing electric fields.

Orientation of the agarose matrix by pulsed electric fields

DNA molecules less than ~20 kilobase (kb) pairs in size are efficiently fractionated by unidirectional electrophoresis in agarose gels (1). The variation of DNA mobility with gel concentration can be described by the Ogston mechanism of pore size distribution, if the observed mobilities are first extrapolated to zero electric field strength (2). Therefore, sieving of the macromolecules by the matrix appears to be the dominant mechanism of mol wt separation in this size range.

Transient electric birefringence of two small DNA restriction fragments of the same molecular weight

The transient electric birefringence of two small DNA restriction fragments of the same molecular weight, one of which migrates anomalously slowly on polyacrylamide gels, has been investigated. Both fragments exhibit negative birefringence. The decay of the birefringence of the anomalously slowly migrating fragment is 8-9% faster than that of the normally migrating fragment. The faster birefringence decay of the anomalous fragment 12A persists under a variety of buffer conditions, suggesting that it is due primarily to static bending and/or curvature of fragment 12A. In reversing electric fields the absolute amplitude of the birefringence of fragments 12A and 12B decreased about 26% before returning to the steady state value. The minimum in the birefringence occurred faster than expected from the birefringence decay times and decreased with increasing electric field strength, suggesting that the minimum is due to a slow polarization of the ion atmosphere. For both fragments, the rise of the birefringence in the Kerr region is about 10% slower than the field-free decay. The buildup of the negative birefringence is preceded either by an interval when no birefringence is observed or by a small positively birefringent transient, suggesting that a small transverse ionic polarizability is also present. Both DNA fragments exhibit Kerr law behavior over most of the range of electric field strengths investigated. Analysis of the shapes of the saturation curves suggests that differences may exist in the polarization mechanisms of the two fragments.

Estimation of polyacrylamide gel pore size from Ferguson plots of linear DNA fragments. II. Comparison of gels with different crosslinker concentrations, added agarose and added linear polyacrylamide

The mobilities of various DNA fragments in two normally migrating molecular weight ladders were studied in polyacrylamide gels containing different concentrations of the crosslinker N,N'-methylenebisacrylamide (Bis). The acrylamide concentration ranged from 2.5-10.5%T (w/v); the Bis concentration ranged from 0.5-10%C (w/w), with respect to total acrylamide. Ferguson plots were constructed for each of the DNA fragments in gels of each composition. The Ferguson plots of the different multimers in each molecular weight ladder were nearly parallel in gels containing 0.5-3%C, converged close to a common intercept at zero gel concentration in gels containing 4%C, and crossed at approximately 1.5%T in gels containing 5 and 10%C. If the mobilities observed for the different DNA fragments at zero gel concentration were also extrapolated to zero DNA molecular weight, a common limiting mobility was observed in gels of all crosslinker concentrations. This limiting mobility was approximately equal to the free solution mobility of DNA. The effective pore radius of each gel was estimated from Ferguson plots based on relative mobilities, using the mobility of the smallest DNA fragment in each molecular weight ladder as the reference mobility. The calculated gel pore radii ranged from 142 nm to 19 nm, respectively, for gels containing 4.6%T, 1.5%C, and 10.5%T, 5 or 10%C. These pore radii are an order of magnitude larger than previously accepted values, but are consistent with scanning electron microscope measurements (Rüchel, R., et al., J. Chromatogr. 1978, 42, 77-90).(ABSTRACT TRUNCATED AT 250 WORDS)

Estimation of polyacrylamide gel pore size from Ferguson plots of normal and anomalously migrating DNA fragments. I. Gels containing 3% N,N'-methylenebisacrylamide

The mobilities of normal and anomalously migrating DNA fragments were determined in polyacrylamide gels of different acrylamide concentrations, polymerized with 3% N,N'-methylenebisacrylamide as the crosslinker. The DNA samples were a commercially available 123-bp ladder and two molecular weight ladders containing multiple copies of two 147-base pair (bp) restriction fragments, obtained from the MspI digestion of plasmid pBR322. One of the 147 bp fragments is known to migrate anomalously slowly in polyacrylamide gels. Ferguson plots were constructed for all multimer ladders, using both absolute mobilities and relative mobilities with respect to the smallest DNA molecule in each data set. If the retardation coefficients were calculated from the relative mobilities, and the rms radius of gyration was used as the measure of DNA size, the Ogston equations were obeyed and the gel fiber parameters could be calculated. The effective pore sizes of the gels were estimated from the gel concentration at which the mobility of a given DNA molecule was reduced to one-half its mobility at zero gel concentration. The estimated pore radii ranged from approximately 130 nm for 3.5% gels to approximately 70 nm for 10.5% gels. These values are much larger than the pore sizes previously determined for the polyacrylamide matrix.

Electric birefringence of dilute agarose solutions

The technique of transient electric birefringence was used to investigate the orientation of agarose solutions in pulsed electric fields. If the agarose was dissolved in deionized water, the sign of the birefringence was positive when the electric field was small, indicating that the agarose molecules were orienting parallel to the electric field lines. The decay of the birefringence was rapid, consistent with the orientation of individual agarose helices. The amplitude of the birefringence, but not the birefringence decay times, increased as the agarose solution aged, suggesting that the helices formed slowly from the sol state. Increasing the amplitude or duration of the pulsed electric field caused additional negative, and then positive, birefringence signals to appear, characterized by much slower rise and decay times, consistent with the formation of aggregates. The slowest decay times ranged from 7.5-9.0 s, suggesting that the aggregates were several microns in size. When agarose was dissolved in dilute Tris buffer instead of deionized water, the fast positive birefringence signal was not observed, suggesting that individual helices were not present in solutions containing dilute buffer.

Resolution of a paradox in the electrophoresis of DNA in agarose gels

A paradox was observed in a previous study of the electrophoresis of linear DNA fragments in agarose gels (D. L. Holmes and N. C. Stellwagen, Electrophoresis 1990, 11, 5-15). The pore size of the agarose matrix was more accurately determined if the root-mean-square radius of gyration was used to measure DNA macromolecular size. However, the Ogston equations were obeyed and other gel parameters such as the apparent fiber radius and fiber volume appeared to be better described if the geometric mean radius was used to measure DNA size. This paradox can be resolved if relative mobilities (with respect to the smallest DNA molecule in the data set) are used to construct the Ferguson plots, instead of absolute mobilities. Using relative mobilities and the root-mean-square radius of gyration, the Ogston equations are obeyed and the pore size of the matrix is consistent with values determined by other methods.

Anomalously slow electrophoretic mobilities of DNA restriction fragments in polyacrylamide gels are not eliminated by increasing the gel pore size

The effect of gel pore size on the anomalous mobility of certain curved DNA restriction fragments in polyacrylamide gels was studied by comparing the electrophoretic mobilities of normal and anomalous fragments in gels of varying composition but constant acrylamide concentration. Molecular weight ladders were prepared from two 147 base pair G-C rich restriction fragments, called 12A (anomalous) and 12B (normal), obtained from the MspI digestion of plasmid pBR322. The electrophoretic mobilities of multimers of the two fragments increasingly diverged with increasing molecular weight. The anomalous mobility of fragment 12A was essentially independent of gel pore size, regardless of whether the pore size was varied by increasing the acrylamide concentration at constant cross-linker concentration or by increasing the cross-linker concentration at constant acrylamide concentration. The anomalous mobility of higher multimers of fragment 12A decreased with increasing gel pore size when the pore size was varied by changing the gel concentration. However, when the gel pore size was changed by varying the cross-linker concentration at constant acrylamide concentration, the anomalous mobility of higher multimers of fragment 12A went through a maximum and then decreased as the pore size was increased. Copolymerizing acrylamide with high molecular weight linear polyacrylamides had no effect on the anomalous mobility of the 12A multimer ladder, even though the apparent absolute mobilities of all fragments increased markedly. Only by incorporating charged residues into the gel matrix or by copolymerizing acrylamide with the intercalator ethidium bromide could the difference in mobility between the 12A and 12B multimer ladders be substantially reduced or eliminated. Similar results were observed with a molecular weight ladder containing 78 base pair repeats of the bending locus of kinetoplast DNA. These results suggest that pore size alone is not responsible for the anomalously slow migration of curved DNA molecules in polyacrylamide gels.

The electric field dependence of DNA mobilities in agarose gels: a reinvestigation

The electric field dependence of the electrophoretic mobility of linear DNA fragments in agarose gels was reinvestigated in order to correct the observed mobilities for the different temperatures actually present in the gel during electrophoresis in different electric field gradients. When corrected to a common temperature, the electrophoretic mobilities of DNA fragments less than or equal to 1 kilobase pairs (kbp) in size were independent of electric field strength at all field strengths from 0.6 to 4.6 V/cm if the gels contained less than or equal to 1.4% agarose. The mobilities of larger DNA fragments increased approximately linearly with electric field strength. If the agarose concentration was higher than 2%, the mobilities of all DNA fragments increased with increasing electric field strength. The electric field dependence of the mobility was larger in gels cast and run in Tris-borate buffer (TBE) than in gels cast and run in Tris-acetate buffer (TAE), and was more pronounced in gels without ethidium bromide incorporated in the matrix. Ferguson plots were constructed for the various DNA fragments, both with and without extrapolating the temperature-corrected mobilities to zero electric field strength. Linear Ferguson plots were obtained for all fragments less than or equal to 12 kbp in size in agarose gels less than or equal to 1.4% in concentration if the mobilities were first extrapolated to zero electric field strength. Concave upward curvature of the Ferguson plots was observed for DNA fragments greater than or equal to 2 kbp in size at finite electric field strengths. Convex downward curvature of the Ferguson plots was observed for DNA fragments greater than or equal to 1 kbp in size in agarose gels greater than or equal to 2% in concentration. The mobilities of the various DNA fragments, extrapolated to zero agarose concentration and zero electric field strength, decreased with increasing DNA molecular weight; extrapolating to zero molecular weight gave an "intrinsic" DNA mobility of 2.7 x 10(-4) cm2/Vs at 20 degrees C. The pore sizes of LE agarose gels cast and run in TAE and TBE buffers were estimated from the mobility of the DNA fragments.(ABSTRACT TRUNCATED AT 400 WORDS)

Electrophoresis of DNA in oriented agarose gels

Oriented agarose gels were prepared by applying an electric field to molten agarose while it was solidifying. Immediately afterwards, DNA samples were applied to the gel and electrophoresed in a constant unidirectional electric field. Regardless of whether the orienting field was applied parallel or perpendicular to the eventual direction of electrophoresis, the mobilities of linear and supercoiled DNA molecules were either faster (80% of the time) or slower (20% of the time) than observed in control, unoriented gels run simultaneously. The difference in mobility in the oriented gel (whether faster or slower) usually increased with increasing DNA molecular weight and increasing voltage applied to orient the agarose matrix. In perpendicularly oriented gels linear DNA fragments traveled in lanes skewed toward the side of the gel; supercoiled DNA molecules traveled in straight lanes. If the orienting voltage was applied parallel to the direction of electrophoresis, both linear and supercoiled DNA molecules migrated in straight lanes. These effects were observed in gels cast from different types of agarose, using various agarose concentrations and two different running buffers, and were observed both with and without ethidium bromide incorporated in the gel. Similar results were observed if the agarose was allowed to solidify first, and the orienting electric field was then applied to the gel for several hours before the DNA samples were added and electrophoresed. The results suggest that the agarose matrix can be oriented by electric fields applied to the gel before and probably during electrophoresis, and that orientation of the matrix affects the mobility and direction of migration of DNA molecules. The skewed lanes observed in the perpendicularly oriented gels suggest that pores or channels can be created in the matrix by application of an electric field. The oriented matrix becomes randomized with time, because DNA fragments in oriented and unoriented gels migrated in straight lanes with identical velocities 24 hours later.

Flexibility of smooth and skeletal tropomyosins

The rotational relaxation times of nonpolymerizable skeletal and smooth muscle tropomyosin were measured by analysis of the decay of the zero-field birefringence at different temperatures and salt concentrations. Skeletal tropomyosin in solution is equally well modeled as a rigid rod or as a semiflexible rod with a persistence length of 150 nm. Smooth muscle tropomyosin does not fit the rigid rod model but is well approximated by a semiflexible rod model with a persistence length of 55 nm. The results indicate that smooth muscle tropomyosin is either a more flexible molecule than skeletal muscle tropomyosin or is a curved structure with an end-to-end length shorter than the coiled-coil contour length. Smooth muscle tropomyosin controls the actomyosin ATPase differently from skeletal muscle tropomyosin and it had been suggested that the reason is because it is more rigid; clearly, another explanation must be sought.

Orientation of DNA and the agarose gel matrix in pulsed electric fields

Transient electric birefringence has been used as an analytical tool to study the orientation of DNA in agarose gels, and to study the orientation of the matrix alone. The sign of the birefringence of DNA oriented in an agarose gel is negative, as observed in free solution, indicating that the DNA molecules orient parallel to the direction of the electric field. If the median pore diameter of the gel is larger than the contour length of the DNA molecule, the DNA effectively does not see the matrix and the birefringence relaxation time is the same as observed in free solution. However, if the median pore diameter of the gel is smaller than the contour length of the DNA, the DNA molecule becomes stretched as well as oriented. For DNA molecules of moderate size (less than or equal to 4 kb), stretching in the gel causes the birefringence relaxation times to increase to the values expected for fully stretched molecules. Complete stretching is not observed for larger DNA molecules. The orientation and stretching of DNA molecules in the gel matrix indicates that end-on migration, or reptation, is a likely mechanism for DNA electrophoresis in agarose gels. When the electric field is rapidly reversed in polarity, very little change in the orientation of the DNA is observed if the DNA molecules were completely stretched and had reached their equilibrium orientation before the field was reversed in direction. Hence completely stretched, oriented DNA molecules are able to reverse their direction of migration in the electric field with little or no loss of orientation. However, if the DNA molecules were not completely stretched or if the equilibrium orientation had not been reached, substantial disorientation of the DNA molecules is observed at field reversal. The forced rate of disorientation in the reversing field is faster than the field-free rate of disorientation. Complicated patterns of reorientation can be observed after field reversal, depending on the degree of orientation in the original field direction. The effect of pulsed electric fields on the orientation of the agarose gel matrix itself was also investigated.(ABSTRACT TRUNCATED AT 400 WORDS)

Orientation of the agarose gel matrix in pulsed electric fields

The technique of transient electric birefringence was used to investigate the effect of pulsed electric fields on the orientation of the agarose gel matrix. Orientation of the gel was observed at all electric field strengths. Very slow, time-dependent effects were observed when pulses of 10-100 V/cm were applied to 1% gels for 0.5-2 seconds, indicating that domains of the matrix were being oriented by the electric field. The sign of the birefringence reversed when the direction of the applied electric field was reversed, indicating that the domains tend to orient in the perpendicular direction after field reversal. Theories of gel electrophoresis will need to incorporate the orientation of the matrix in order to provide a complete explanation of electrophoresis in agarose gels.

Effect of pulsed and reversing electric fields on the orientation of linear and supercoiled DNA molecules in agarose gels

When linear or supercoiled DNA molecules are imbedded in agarose gels and subjected to electric fields, they become oriented in the gel matrix and give rise to an electric birefringence signal. The sign of the birefringence is negative, indicating that the DNA molecules are oriented parallel to the electric field lines. If the DNA molecules are larger than about 1.5 kilobase pairs, a delay is observed before the birefringence signal appears. This time lag, which is roughly independent of DNA molecular weight, decreases with increasing electric field strength. The field-free decay of the birefringence is much slower for the DNA molecules imbedded in agarose gels than observed in free solution, indicating that orientation in the gel is accompanied by stretching. Both linear and supercoiled molecules become stretched, although the apparent change in conformation is much less pronounced for supercoiled molecules. When the electric field is rapidly reversed in polarity, very little change in the birefringence signal is observed for linear or supercoiled DNAs if the equilibrium orientation (i.e., birefringence) had been reached before field reversal. Apparently, completely stretched, oriented DNA molecules are able to reverse their direction of migration with little or no loss of orientation. If the steady-state birefringence had not been reached before the field reversal, complicated orientation patterns are observed after field reversal. Very large, partially stretched DNA molecules exhibit a rapid decrease in orientation at field reversal. The rate of decrease of the birefringence signal in the reversing field is faster than the field-free decay of the birefringence and is approximately equal to the rate of orientation in the field (after the lag period).(ABSTRACT TRUNCATED AT 250 WORDS)

Orientation of DNA molecules in agarose gels by pulsed electric fields

The electric birefringence of DNA restriction fragments of three different sizes, 622, 1426, and 2936 base pairs, imbedded in agarose gels of different concentrations, was measured. The birefringence relaxation times observed in the gels are equal to the values observed in free solution, if the median pore diameter of the gel is larger than the effective hydrodynamic length of the DNA molecule in solution. However, if the median pore diameter is smaller than the apparent hydrodynamic length, the birefringence relaxation times increase markedly, becoming equal to the values expected for the birefringence relaxation of fully stretched DNA molecules. This apparent elongation indicates that end-on migration, or reptation is a likely mechanism for the electrophoresis of large DNA molecules in agarose gels. The relaxation times of the stretched DNA molecules scale with molecular weight (or contour length) as N2.8, in reasonable agreement with reptation theories.

Circular dichroism and thermal melting of two small DNA restriction fragments of the same molecular weight

The thermal melting and circular dichroism of two 147 base pair restriction fragments of pBR322 have been studied. The fragment with the higher GC content, 12B, melts at a higher temperature than the other fragment, 12A, as expected. The melting temperatures are proportional to the logarithm of the concentration of NaCl or tris(hydroxymethyl)aminomethane (Tris) buffer, between 1 mM and 0.2 M added salt. In 1 mM Tris buffer, the melting temperatures of the two fragments are nearly equal. The circular dichroism spectra of fragments 12A and 12B in 0.2-10 mM Tris buffer are characteristic of B-form DNA. In 81% ethanol, the circular dichroism spectra of the two fragments are characteristic of A-form DNA. With 1 mM Tris buffer as the supporting electrolyte, fragment 12A exhibits a very sharp B----A transition, with a midpoint at 79% ethanol. However, a biphasic transition is observed for fragment 12B, with midpoints at 73% and 80% ethanol. This biphasic transition may represent the conversion of separate domains of fragment 12B from the B conformation to the A conformation; half of this fragment is much more GC rich than the other half. Methods are also described for preparing polymers of the 12A and 12B fragments.

Accurate molecular weight determinations of deoxyribonucleic acid restriction fragments on agarose gels

The electrophoresis of various DNA restriction fragments ranging in size from 47 to 6000 base pairs has been examined as a function of agarose concentration, electric field strength, and time of electrophoresis. A typical sigmoidal curve was obtained when the logarithm of the molecular weight was plotted as a function of mobility. The logarithms of the mobilities of all fragments were a linear function of gel concentration, if the mobilities of fragments greater than or equal to 1000 base pairs were first extrapolated to zero electric field strength. The slopes of the lines, called the retardation coefficients, were found to be linearly proportional to the effective hydrodynamic surface areas of the fragments, as predicted by the Ogston theory of pore size distribution. The logarithm of the mobility of native DNA fragments was inversely proportional to Mr0.8 over the entire molecular weight range, if the mobilities of fragments larger than 1000 base pairs were first extrapolated to zero electric field strength. The logarithm of the mobility of denatured, single-stranded DNA molecules was inversely proportional to the square root of molecular weight. The agreement of these results with the Ogston theory argues against a reptation mechanism for the movement of DNA molecules less than or equal to 6000 base pairs through agarose gels.

Anomalous electrophoresis of deoxyribonucleic acid restriction fragments on polyacrylamide gels

A detailed study has been made of the polyacrylamide gel electrophoresis of DNA restriction fragments obtained from two plasmids, pBR322 and p82-6B. Variables studied were molecular weight, gel concentration, temperature, and electric field strength. The retardation coefficients of the larger fragments (greater than 800 base pairs) were independent of molecular weight. The retardation coefficients of the smallest fragments (less than or equal to 300 base pairs) were proportional to Mr1/3, and therefore to the mean geometric radii of the fragments. The logarithm of the relative mobility of all fragments was also proportional to Mr1/3. The anomalous migration of certain fragments on polyacrylamide gels was found to be "transportable" into fragments generated by different restriction enzymes. Anomalous migration was enhanced at lower temperatures and disappeared upon increasing the temperature. A fragment which migrated anomalously slowly migrated even more anomalously when dimerized; dimerizing a normally migrating fragment resulted in the normal migration of the dimerized fragment. Anomalously migrating fragments were found to be localized in distinct regions of the pBR322 circle.

A small DNA restriction fragment with an apparent permanent dipole moment

The electric birefringence of two DNA restriction fragments, each containing 147 base-pairs, has been investigated. The decay of the birefringence was the same for the two fragments, with a relaxation time corresponding to the reorientation of fully extended, rod-like molecules. However, the birefringence saturation behavior of the two fragments was markedly different: one fragment oriented by the expected induced-dipole mechanism, while orientation of the other fragment followed the theoretical curve for permanent dipole orientation. This difference in behavior must be due to differences in the base-pair sequences of the two fragments.

Structural transition produced by electric fields in aqueous sodium deoxyribonucleate

It was found that the birefringence of aqueous solutions of sodium DNA is anomalous when electric fields of high intensity (>/=10(4) v/cm) are applied. The magnitude of the birefringence first rose upon application of the orienting pulse, then fell as the field was sustained above a critical value. The occurrence of the effect depended upon macromolecular and electrolyte concentrations. Upon removal of the field, the birefringence was rapidly restored and then it decayed with an increase of the reorientational relaxation times, relative to those observed below the critical field. It is proposed that the electric field may cause aggregation of the macromolecules and then produce a structural transition concomitant with the electric field orientation effect. This transition may correspond to the "B" right harpoon over left harpoon "A" structures identified in x-ray studies, or to a "B" right harpoon over left harpoon "V" structure change, where "V" is a postulated new helical form stabilized by cooperative interactions of base and dipoles in the electric field. Field induced transitions of this type would be of interest in connection with molecular mechanisms of transport through membranes, nerve impulse transmission, or information storage.