Many medical research teams continue to study the wide range of human conditions influenced by the structure and function of Z-DNA, for example, the differentiation of fibre cells in the adult lens1, a pox virus which complexes with Z-DNA2, the inhibition of transcriptional activation3, a conserved family of Z-DNA binding proteins4, the rôle of Z-DNA in the activity of the hippocampus in the Alzheimer Brain5, an RNA deaminase has a binding site for Z-DNA6, 7, and a human gene codes for the protein dlm-1 which has a Z-DNA binding domain8.  Z-DNA-forming sites have been reported within the human genome9.  New Z-DNA duplexes are being currently reported10. While the Z form has been assigned a left-handed helical sense11, a central difficulty about the crystallographic identification of Z‑form oligodeoxyribonucleotides and Z-DNA as being uniformly left‑handed is the creation of a crucial paradox. Highly regarded crystallographers, for example Sasisekharan & Brahmachari12 , and Leslie et al.13, record the complete conversion of the B to the Z form inside semi-crystalline fibres under the very mild conditions of humidity change alone.  Likewise, Mahendrasingham et al.14 studied the transition of the B to the D form under similarly mild conditions using time-lapse data recording. 

These crystallographers argue strongly from the direct experimental evidence of a clear change in well-defined diffraction patterns that the B, D and Z forms must have the same helical handedness to effect such changes from one polymorph to another inside semi-crystalline fibres with high conversions, since only small structural changes are conceivable in these circumstances. Much earlier, Franklin & Gosling15 had reported the complete interconversion of the A to and from the B form, and likewise Marvin et al.16 had reported the conversion of form B to C inside fibres with humidity change alone.  Therefore, if they have a uniform helical handedness along their whole molecular lengths, A, B, C, D and Z‑DNA would all have the same uniform helical handedness because these polymorphs are all variously interconvertible inside solid fibres. 

Now, Dickerson's group17,18 claims, from a study of a true oligodeoxyribonucleotide crystal, that B‑DNA is right‑handed.  Rich's group11 have studied a crystalline fragment, designated Z‑DNA, which they claim is left‑handed.  If the DNA polymorphs are claimed to each have a uniform helical handedness along their whole molecular lengths, and if the Z form fibre in vivo is claimed to have the same handedness as the crystalline Z-DNA fragment (a proposition which has never been directly established), it would seem that at least one team of crystallographers must be wrong.  Either DNA fibres of the forms A, B, C, D & Z have the same helical handedness as suggested by their various, ready interconversions inside fibres, or the B and Z polymorphs have opposite handedness as deduced from the structural solutions derived from X-ray diffraction from true crystals of their respective oligodeoxyribonucleotides which are then extended to the structure of the high polymer.

Sasisekharan & coworkers^{19, 20, 21, 22}, and Rodley & his coworkers ^{23, 24, 25, 26} have elaborated models of duplex DNA which consist of alternating stretches of right- and of left-handed helical winding.  Were such models to exist inside DNA fibres, any problem about a change of helical handedness among the A, B, C, D & Z forms would not arise.  In addition, if no change of helical handedness were to be contemplated to resolve the paradox, a further, right-handed paranemic, i.e., side-by-side (SBS), model^{27, 28, 29, 30} could equally well be considered (Figure 1).  Lee et al.³¹ have produced a very clear STM image of B-DNA as right-handed and paranemic in the conditions with which they worked (their Figure 3d).  Beebe et al.³², Barton's group^33,34 and Iwamoto & Hsu³⁵ have also used non‑crystallographic methods purporting to show that B‑DNA is uniformly right‑handed, at least under their experimental conditions.

To establish a mere change in DNA polymorph in semi-crystalline fibres it is only necessary to record a directly observable change in the X-ray diffraction pattern from the fibre as the predominant polymorph changes.  Using X-ray diffraction, to establish an actual atomic structure for a DNA polymorph, or for a true crystal of an oligodeoxyribonucleotide, however, it is necessary to proceed through many algorithms and computational processes. Thus an actual atomic structure is an indirect outcome deducible from a diffraction pattern only after much computation, whereas a change in polymorph is directly observable merely from changes in a diffraction pattern. It is therefore essential to scrutinise with extreme care the indirect outcome of the identification of the polymeric molecular structure of Z‑DNA and high‑salt fibres of poly (dG‑dC)₂ as uniformly left‑handed when deduced using X-ray diffraction from crystals of short lengths of oligodeoxyribonucleotides.

The Z form has been identified by vacuum ultraviolet³⁶ and Raman spectroscopy³⁷, for example, as well as through the use of monoclonal antibodies^38,39, and by using electron microscopy⁴⁰. These techniques do not themselves establish the helical handedness of the Z polymorph.  The Z polymorph, composed of particular sequences such as polydeoxyguanosine-cytidine, poly (dG.dC)₂, is considered to be a uniformly left-handed double helix in high salt media from the crystallography of short crystallised fragments known as oligodeoxyribonucleotides and to have 12 Watson-Crick base pairs per helical turn of pitch some 4.45 nm⁴¹. 

FIGURE 1

A DIAGRAMMATIC REPRESENTATION OF PARANEMIC RIGHT HANDED DUPLEX Z-DNA LOOKING

 TOWARDS THE LEADING EDGE OF THE BASE PAIRS (THE FRONT FACE IN FIGURE 2)


 

SPECTRAL INVERSIONS AS AN INDICATION OF CHANGES IN HELICAL HANDEDNESS

The earliest attempt to assign a change of handedness, however tentatively, was that of Pohl & Jovin⁴² in 1972 using the inversion of circular dichroism (CD) spectra from the synthetic polydeoxynucleotide poly (dG.dC)_n.  They identified the normal B form CD spectrum for the aqueous polydeoxynucleotide at low ionic strength, to which they assigned the designation R, reflecting the prevailing view at the time, and today too, that the B form helix was right-handed. At high ionic strengths they observed that the CD spectrum was substantially inverted in the spectral range accessible to them. To this DNA polymorph they assigned the designation L.

The possible existence of left‑handed DNA had been inferred earlier by Pohl from topological considerations regarding the chemistry of DNA⁴³ Unfortunately, there are several difficulties with the Pohl & Jovin paper⁴².  First, the designations R and L are not neutral and, further, they imply an acceptance of the notion that the reported CD inversion might reasonably be correlated with a change of helical handedness, though they remark:

"Although inversion of the circular dichroism spectra is suggestive of a change of the helix sense, it cannot prove it."

This view has been expressed elsewhere in the literature⁴⁰. Only rather later, did Pohl⁴⁴ understand in 1976 that similar inverted spectra could be obtained by calculation for right‑handed helices. Also, in 1976 the SBS models^{19,20,21,22,23} challenged another, less obvious assumption of Pohl & Jovin, namely, that the B form helices are of uniform hand along their whole lengths.

In their paper, Pohl & Jovin make no reference to the pre-existing published work on the inversion of CD spectra. In 1969, Gabbay⁴⁵ reported his observation of oppositely induced circular dichroism in complexes of small, reporter molecules with nucleic acids.  With a range of substituted, unsymmetrical 2,4‑dinitroanilines he showed that double-stranded (ds) RNA and ds DNA would induce a positive and a negative CD spectrum respectively, very nearly mirror images, with peak and trough both at about 360‑365 nm.  Now, it is not claimed or suggested that the random sequence RNA and DNA enjoyed opposite helical handedness.  Therefore it is possible for nucleic acids in complexes to give rise to CD spectra whose form does not depend on the handedness of the helix. 

Gabbay suggested that the small, unsymmetrical reporter molecule is complexed with the RNA in one orientation, and with DNA in a reversed orientation, a suggestion of great potential significance for the structure of B-DNA relative to Z‑DNA. Furthermore, in a later paper⁴⁶, Gabbay and coworkers developed the proposition that the inversion of the CD spectrum seen for DNA complexes with their reporter molecule 4, created by changing the nucleic acid : reporter ratios, arises from the existence of different intercalation sites which have a range of binding affinities for the reporter.  Again, inversion of the CD spectrum can be induced without postulating any change of handedness of the helix, it seems.  By manipulation of the sequence of the stacked bases, and using intercalation, an inverted CD spectrum can be induced.

In later work, Zarlenga, Halsall & Day⁴⁷ reported an extensive study of the effect on CD spectra of the complexing of polydeoxynucleotides with a polycationic aromatic intercalator.  Their Figure 2 shows a large inversion for poly (dA‑dT)_n, again apparently without a change of handedness of the helix. Tomasz et al.⁴⁸ studied the interaction of mitomycin with poly (dG‑dC)_n.  The complexes produced showed an inversion of the CD spectrum rather similar to that of Pohl & Jovin⁴², but again a change of handedness of the polynucleotide helix was not the explanation of the inversion.  They urge caution in the use of CD changes alone as an indication of the existence of a left‑handed double helix. Again, Taboury & Taillandier⁴⁹ were able to induce a negative CD band in poly (dA‑dC)_n. poly (dG‑dT)_n, at 275 nm in the presence of Cs⁺, which might have implied a conversion to a left‑handed helix since the curve resembled that of Pohl & Jovin⁴² (Figure 2), but they were able to show that the helix must still be right handed.

Vacuum ultraviolet (VUV) CD frequencies not accessible to Pohl & Jovin were used later^{36,50 – 55} and the spectral changes and inversions reported did not confirm that a change of helical handedness had occurred during the B to Z transition. In the VUV, spectral changes were found to be dependent upon changes in base orientation. The CD spectral inversions in UV and VUV frequencies discussed here can be explained by the inversion of the orientation of the purine (guanine) through 180^o changing from a Watson-Crick base pair to a Hoogsteen base pair (Figure 2) in (dG.dC)_n sequences. Such a transition has not previously been considered in the literature as an explanation of the CD inversion observed by Pohl & Jovin.  However, Hoogsteen base pairs in polynucleotide fibres have been reported⁵⁶.

FIGURE 2

Watson-Crick (below) and Hoogsteen Base Pairing  (above)

Leading Edge, or Front Face, Drawn To The Left)

The purines and pyrimidines are connected to the deoxyribose at C1'. The guanosine undergoes an inversion rotation through 180^o  from one pairing to the other. The N at GN⁷ in Hoogsteen Base Pairs carries a positive charge.

CRYSTALLOGRAPHIC DETERMINATION OF THE ABSOLUTE HANDEDNESS OF Z-DNA

Pohl⁵⁷ considered "the possibility of adopting a left‑handed double‑helical structure for DNA firmly established" by the work of Wang et al.¹¹.

The proposition that the analysis of the X‑ray diffraction patterns derived from true crystals of oligodeoxynucleotides, as conducted by skilled practitioners, yields absolute 3‑dimensional structures at, or near, atomic resolution which are always beyond challenge, and which can be extended to include the high polymer in vivo, constitutes the core of the currently received wisdom in this field (see, for example, 10, 58, 59).

However, there are difficulties with this proposition.  For example, when applied to Z form oligodeoxyribonucleotide crystals, supposedly uniformly left-handed, with the B form crystals uniformly right-handed, there is the difficulty already discussed in the Abstract concerning the same handedness of the A, B, C, D and Z forms as demonstrated by polymorph inter-conversions inside semi-crystalline fibres under mild conditions of humidity change alone.

A further difficulty extends to the actual methodology used to deduce the structures of crystalline fragments of Z-form oligodeoxynucleotides.  Even using heavy atoms, such as in brominated oligodeoxyribonucleotides, for example, the structural solutions are explicitly deduced from earlier solutions derived from X-ray diffraction from oligodeoxynucleotides which themselves did not contain heavy atoms. Thus Chevrier et al.⁶⁰ found the structure of ^5BrCG^5BrCG^5BrCG from that reported for ^5MeCG^5MeCG^5MeCG⁶¹, itself based on the choice of an even earlier model of d(CGCGCG), prior to computational structure refinement:                      

“…The structure was solved by placing the brominated hexamer in the methylated lattice.”  (60, page 708)

Wang et al.⁶² reported on the structures within four different crystals of d(CGCGCG) which they consider to be composed of uniformly left‑handed double helices.

In solving the structures of their crystals they generated a helically regular hexamer with the 6₅ axis and 4.458 nm repeat which they had identified in the crystal.  In Table 2 they list the coordinates of single chains of ZI and ZII DNA and remark:                                                                       

 "Additional residues and the complementary chain may be generated from these coordinates."  [My italics]

The clear implication of these remarks is that the idealised coordinates for chains ZI and ZII were generated by the investigators themselves and then coordinates for the complementary chains were calculated mathematically.

Such a procedure would seem likely to predispose a left‑handed structural solution. 

Again, Pohl ⁶³ draws upon the work of Crawford et al.⁶⁴, whose structural solution for crystals of d(CGCG) was claimed to be left‑handed.  These workers made it clear, however, on page 4016, that they also tried A‑, B‑, C‑ & D‑DNA as search molecules, but found the best fit for a trial model of left‑handed Z‑DNA:

"We tried to solve the structure by using the method of molecular replacement, and assuming a number of search molecules (A‑DNA, B‑DNA, C‑DNA, D‑DNA).  Finally we became convinced that the method was not working or that somehow the structure was quite different from what we had anticipated....  Solution of the hexamer's strikingly unusual structure and the clear similarity of the cell dimensions between the tetramer and hexamer suggested that we should again attempt the method of molecular replacement using a trial model of left‑handed Z‑DNA."

This might suggest that a fresh model, as yet untried, perhaps a side-by-side^20,26 or even a paranemic model^27,30 might give an even better fit and that their range of trial models limited their possible range of structural solutions.  Such reservations as these hardly establish d(CGCG)⁶⁴ or d(CGCGCG)⁶² as left‑handed beyond challenge.

There is a further difficulty with the Z-DNA crystallographic studies reported by Rich’s group and by Drew et al.  In 1981 Hopkins⁶⁵ showed that, even for a double helix, whether right- or left-handed, there would be two possible stereoisomers in each helical hand.  The Z-DNA researchers made no reference to this and did not discuss how they came to prefer one isomer rather than the other.  The alternative stereoisomer for Z-DNA was investigated using NMR much later by Uesugi et al.⁶⁶ in 1988.

It seems too that Rich’s research group have never deposited their Z form structure factors with the Protein Data Bank, so it is not possible to scrutinise their diffraction data directly.

Yet another complication with oligodeoxyribonucleotide crystallography is that the structural refinement process frequently makes use of the computational software Nucleic Acid Least Squares, NUCLSQ.  For example, there is the report of Sadavisan et al.⁵⁸ that their refinement was carried out using the algorithm NUCLSQ, but this contains within it the sub-routine NAHELIX⁵⁹ which itself contains within its algorithm a model double helix:

“Z1- and Z2-model coordinates were calculated by means of the NAHELIX program … and the solution for the first molecule was obtained with the Z2-DNA model …”

Thus the double-helical, uniformly left-handed Z-DNA outcome reported by Sadasivan et al. using the umbrella program NUCLSQ was predetermined by the prior choice of computational algorithm.

THE APPLICATION OF NUCLEAR MAGNETIC RESONANCE (NMR) TO OLIGONUCLEOTIDE STRUCTURES

Nuclear magnetic resonance records the absorbance of specific radio-frequencies in an intense magnetic field by appropriate atomic nuclei, such as ¹H, whose nuclei are said to “resonate” at these frequencies.  These specific frequencies can be assigned to particular chemical environments by calibration against chemical standards.

Organic macromolecules have been characterised systematically by proton (¹H) NMR, a technique of such sophistication that many reports, for example, 67 & 68, dwell on the extension of 2D‑NMR into a third dimension.  Thus proton NMR has been employed to measure distances between resonating nuclei in proteins and in nucleic acids, though such distances must be smaller than 0.5 nm. At greater distances between resonating nuclei the signal strengths are much diminished.                 

However, the inter-proton distances in nucleic acids deduced using NMR are based upon prior knowledge of the standard distances found by x-ray diffraction (69, pages 47, 48, 50).

That the foundation of nucleic acid NMR studies of the distances between nuclei stands upon the literature values of the atomic coordinates of nucleic acids and oligodeoxyribonucleotides derived from X‑ray crystallography is also made very clear, for example, by Cheng et al.⁷⁰:                                                          

        "The through‑space magnetic field effect on the base protons in the helices due to the local environment can be calculated in terms of the ring‑current effect, the local atomic magnetic susceptibility effect, and the polarisation or electric field effect, when a set of coordinates for the helix is provided such as the coordinates from A‑, A'‑, B‑, or Z‑conformation, etc.  Since both d‑CGCG and d‑CGCGCG helices have been defined and characterised, they can therefore serve as models to calibrate existing nmr theory and procedures for calculation of the through‑space magnetic effects on the base protons and NH ‑ N protons..." (My italics)  

However, it is the reliability of the x-ray structures of d-CGCG and d-CGCGCG which are under challenge in this paper, and, consequently, inter-proton distances in the nucleic acids derived from NMR which are offered to support the double helix are compromised by being dependent upon the same x-ray studies as are being challenged.

The measurement of interatomic distances in the nucleic acids should reveal differences between those expected for the double helix, based on a B form diameter of some 2.0 nm, while the corresponding Z form helical diameters are some 1.8 nm for the double helix and SBS models, with 0.85 nm for each helix of the paranemic model.

In particular, based on the double helical model, if the Watson-Crick duplex DNA were not the actual structure present, some resonating nuclei would give rise to rates of cross‑relaxation far higher than would be expected from their relatively large distances apart on the Watson-Crick model because their true distances apart might be far smaller, for example, with the paranemic model. 

This is exactly what has been observed, as reported by Reid⁷¹:                                            

"However, one must exercise some caution in deriving distances (from the [NMR] spectrum) because more distant protons can … exaggerate their proximity via a process termed spin‑diffusion..."

“Spin diffusion” ⁷² has been discussed in protein NMR, and is widely invoked in nucleic acid studies to “explain” anomalous, large distance measurements.

Many NMR spectroscopists have reported difficulties over reconciling some rather large NMR signal intensities with the supposedly large distance apart of the resonating nuclei as deduced from structures defined by X‑ray crystallography.   For example, Chazin et al.⁷³ conducted a careful study of intensities deduced via "through‑bond" ¹H‑¹H connectivities rather than using "through‑space" connectivities.  They report:

"...readily measurable cross‑peak intensities prevail only between closely spaced hydrogen atoms.  These [NMR signals] (Figure 9) correspond to connectivities di(6,8;2'), di(CH3 ;6), ds(2";6,8) and ds(6,8; CH3), where the latter comes rather as a surprise in view of the relatively long distance of 0.38 nm ....Both Figures 9 & 10 clearly indicate that in the spectrum recorded with a mixing time of 100 ms, the distance information is already strongly distorted by spin diffusion."   (My italics)                                                                

Again, Wűthrich (74, page 238) records quite large NMR signal intensities for di(6,8;3') and ds(3';6,8), even though in the standard B form of DNA these two distances are around 0.45 nm and are therefore at the limit for direct detection.  With the paranemic model, and at points of change of helical handedness in SBS models, where the resonating ¹H nuclei are closer together than in double helical models, these distances will be much reduced and the large signal intensities are easy to understand without the need to invoke "spin diffusion". 

"Spin diffusion" is invoked to "account for" differences in distances which NMR spectroscopists have deduced with those different distances derived from X-ray crystallography. 

Anomalously high resonances are reported between ¹H nuclei in oligodeoxyribonucleotides thought to be some 0.4 – 0.5 nm apart in a double helix model.  It is not explained by NMR spectroscopists how "spin diffusion" can be observed between nuclei 0.4 – 0.5 nm apart, presumably involving magnetic interactions over smaller, intervening distances, yet "spin diffusion" has never been invoked as occurring between nuclei 0.2 – 0.3 nm apart.

With the possibility that the Watson-Crick double helix is not the DNA structure actually being studied, and that the sugar‑phosphates are wound on a helix of smaller diameter, it might not be necessary to invoke the intervention of spin‑diffusion⁷².

The need to invoke spin‑diffusion in certain oligodeoxyribonucleotides may arise because their secondary structures as determined by X‑ray crystallography are wrong, at least to some degree.

The Kinetics of Some Interactions of Z-DNA

There are many difficulties with X-ray crystallography as a technique for studying the structure of DNA and oligodeoxyribonucleotide fragments:

1    It is a derived technique that does not allow direct structural observations

2    It is an inaccessible technique that can allow several interpretations of the DNA and oligodeoxyribonucleotide diffraction patterns requiring highly specialised training to interpret the experimental results using many computational algorithms

3    Its substrates must be in an orderly, preferably highly crystalline state

4    Usually it is applied to stationary structures and progressive movement cannot be captured (though recent synchrotron studies (e.g., 75, 76, 77) are changing this limitation) so the kinetics of its behaviour in vivo are rarely reported with the crystallographic outcomes.  Generally, kinetic information does not inform the outcomes of crystallographic studies even where choices of interpretation of the diffraction pattern offer themselves

5    It can be applied with the highest resolution only to short nucleic acid sequences up to about 12 base pairs largely because longer ones do not crystallise well or easily

6    It cannot be applied to single molecules

7    High resolution studies cannot be expected when the substrate is immersed, dissolved or dispersed in a liquid.

In its biochemistry, in a sense, a main function of DNA depends upon its capacity to be active and dynamically engaged in complexes where it is often engaged as a single molecule interacting with other nucleic acids, proteins, methylating enzymes, etc.  It is rarely stationary, rarely, if ever, in an ordered array, in vivo, and it is always immersed and dispersed in an aqueous medium.

The chemistry and biology of Z‑DNA has been reviewed by Rich et al.⁷⁸.  Their Table 3 records that the B→Z transition can be induced not only by changes in ionic strength (used originally by Pohl & Jovin), but by methanol, ethanol and trifluoroethanol, and can be largely reversed by intercalators⁷⁹.  The pitches and diameters of B‑ and Z‑DNA are different^11,80.

The formation of Z‑DNA from B-DNA is enhanced by the following, for example:

1.  Methylation of deoxycytidine at ⁵dC ⁸¹ to form ^m5dC  

2.  Very low levels of certain divalent ions^82,83

3.  Spermine⁸⁴,

and

4.  A methylated nucleotide, ^m7G, in place of G in poly(dG‑dC)₂ ⁸⁰.

Recording the circumstances in which Z-DNA formation is enhanced invites explanations, but these have not been forthcoming in any reviews of Z-DNA, or its Chemistry or its Biology to date.  It remains highly unsatisfactory that bare experimental results can be reported with no accompanying effort to explain them.

In addition, and not recorded by Rich et al.⁷⁸, according to Miller and his colleagues⁸⁵, the B→Z transition of poly(d(G‑^m5C))₂ is facilitated by an order of magnitude when the B-DNA has been already wound onto nucleosomal particles as a superhelix, constrained by its close contact with the nucleosomal histones, compared with the rate for the B→Z transition for the polymer freely suspended in solution.  This experimental observation poses a severe test for the proposition that Z-DNA has a different helical sense from B-DNA.

If a change in helical sense is involved here, it is important that an explanation be offered as to how it could be an order of magnitude easier to form Z‑DNA from B‑DNA when the latter is wound around in a superhelix for 1.75 turns in direct contact with a solid histone surface compared to forming Z‑DNA from B‑DNA freely in solution.

An interesting result reported by Chen⁸⁶, among others, lies in the existence, with the methylated polynucleotide poly d(G.^5mC)₂, of a Z conformation at low‑salt concentrations and a Z conformation at high‑salt concentrations, both of which are interconvertible into a B conformation at intermediate salt concentrations, even when approached from either a low‑ or a high‑salt direction.  A monotonic increase, or decrease, in salt concentration leads to a Z→B→Z transition.  It seems an unpromising standpoint to maintain that Z conformations are uniformly left‑handed in the face of Z→B→Z conformational transitions when the cause, as distinct from the experimental conditions, of neither of the changes in helical handedness can be explained.

It is hard to understand why a left‑right‑left equilibrium in uniform helical handedness could be established with both transitions influenced by the same stimulus.  A rather similar salt‑induced Z→A→Z transition has also been claimed to represent a left‑right‑left uniform helical transformation⁸⁷.

In this paper is offered an explanation of the results cited here.  With the SBS and paranemic models, the 180⁰ inversion rotation of the nucleotide G from Watson-Crick pairing to Hoogsteen pairing, illustrated in Figure 1, could cause the CD inversion reported by Pohl & Jovin without postulating any change in helical handedness. 

In the paranemic model, the base pairs are stacked upon each other and, when wound onto nucleosomal histones, the leading edges of the base pairs, that is, those facing away from the deoxyribose sugar linked via C1’, face outwards towards a greater radius of rotation.  This could allow them to rotate through 180^o more rapidly from the Watson-Crick base pairing in B-DNA to Hoogsteen pairing in the nucleosomal Z-DNA than would be possible for a disordered polymer suspended and twisted in solution.  This could account for the report of Miller et al. that the B→Z transition was an order of magnitude faster in nucleosomal DNA.

It is possible in this way to explain the reported Z→B→Z and Z→A→Z transitions since there is now no change in helical handedness in either the SBS or paranemic models.  The different Z states could result from transitions between one or more of the many alternative base pairing schemes discussed in the literature (e.g., 41, page 120).

Anti-Z-DNA Antibodies

A further sensitivity to substitution at ⁵C is known to include a specific anti-Z-DNA antibody, clone 44 ³⁸.

The sensitivity of ^5MeC to certain anti-Z-DNA antibodies³⁸ might now be explicable since these antibodies could encounter the 5Me group as soon as they approach the leading edge of the base pairs in the Z-DNA in the paranemic model.  For the double helix, on the other hand, ⁵C is relatively inaccessible inside the helix where antibodies would have a reduced sensitivity to ^5MeC.

Another anti-Z-DNA antibody, found by Malfoy et al.⁸⁸, recognised the 2-NH₂ in guanosine. In Hoogsteen pairing with the paranemic model, in guanosine position 2, the atoms face outwards into the environment where they are readily accessible to antibodies.  For Watson-Crick pairing, position ²G is relatively inaccessible inside the double helix.  Thus the sensitivities of these antibodies to methylation at ⁵C and to the 2-NH₂ on guanosine lend support to a paranemic duplex with Hoogsteen pairing. 

DEFENCES OF THE WATSON-CRICK NUCLEIC ACID DOUBLE HELIX

Arnott⁸⁹, Arnott et al.⁹⁰ and Dover⁹¹ have formulated concise and demanding defences of the double helix based on X-ray diffraction studies of fibres.  These defences have been criticised in detail and decisively rejected. ^{27, Chapter 9}

Rodley & Bates have debated with Arnott the merits and problems of the Watson-Crick double helix and SBS models (the paranemic model excepted.) ⁹²

The classical crystallographic studies of nucleic acid fibres and oligodeoxynucleotide crystal fragments have been extensively reassessed ^{27, Chapters 7, 13 & 14}.

TOPOLOGICAL DEFENCES OF SUPERHELICAL WATSON‑CRICK DNA

Crick, Wang & Bauer⁹³ have considered in detail the topological implications of the models then known to them, and extend their discussion to include a then‑hypothetical model for which Lk = 0:

"Notice that for a true side‑by‑side (SBS) model (Lk = 0) a very easy experimental proof is possible.  One need only take such a piece of circular DNA and raise the temperature until the structure denatures.  The two chains should come apart into two distinct intact single‑stranded circles.... We think it is fair to say that the evidence in favour of the classical double helix is sufficently strong that a proposal for a true SBS structure is unlikely to be widely accepted unless a dramatic experimental demonstration is provided.... This hypothetical experiment is a good test for a true SBS model (Lk = 0).  Experiments have already shown that a circular DNA molecule, when denatured, does not separate into two parts but sediments as a single component....This proves that Lk ≠ 0 (at least for the great majority of those molecules)."      [My italics. CD.]

In that same year, 1979, Stettler et al.⁹⁴ provided exactly the required dramatic experimental demonstration.  They describe the formation of a duplex with well‑defined physicochemical properties when complementary circular single strands of DNA associate under hybridisation conditions.

Crick et al.⁹³ had already identified correctly the reason why perhaps all, and certainly the majority of naturally occurring closed circular duplexes do not furnish single stranded closed circular DNA on denaturation.  Such duplexes contain superhelical turns and will not separate even based on the SBS or paranemic models, where Lk = 0 for a completely relaxed closed circular duplex (Form V).

Crick et al.⁹³ deduce that a direct experimental proof of the existence of a true SBS model (Lk = 0) would be possible by strand separation in a closed circular duplex having no supercoiling.  The work of Stettler et al.⁹⁴ furnishes the “dramatic” experimental demonstration that Crick et al. require.

In a similar paper to that of Crick et al.⁹³, Bauer, Crick & White⁹⁵ have reviewed the topological implications of superhelical DNA.  They consider the evidence that, in many specimens of DNA examined, the number of base pairs per double helical turn is 10.4 ± 0.1, and discuss briefly several models which have been proposed:

"...in which the two polynucleotide chains do not coil around each other to form a double helix but instead lie side by side over most of their length, having only a few helical turns.  Wang's results...show that these new models must be incorrect.  This topological argument is very powerful in that it eliminates all models of the side‑by‑side type, regardless of their molecular detail." [My italics. CD.]

As pointed out by Rodley⁹⁶, however, it is not a requirement of the SBS models that there be an exactly equal number of right and left turns, or parts thereof, and therefore the relaxed circular SBS duplexes need not have any net plectonaemic winding in vivo, and thus the topological objection of Crick et al. to SBS models fails.

The new, paranemic model for the structure of DNA is a side‑by‑side model (Lk = 0) which differs from earlier ones in that the relaxed linear, or relaxed circular duplex has identically zero double helical turns at any and all places along its length, including any length taken overall.  The diameter of each helix in the pair is determined by the base pair width, a necessary consequence of the geometry.

Thus it is not possible to uncritically endorse the conclusion of Bauer, Crick & White⁹⁵ that all side‑by‑side models are eliminated by the topology of supercoiling in covalently closed circular duplexes.

Moreover, in a lengthy review of the experimental literature, Wolffe & Kurumizaka⁹⁷ conclude that the theoretical results of Bauer, Crick & White⁹³ and of White & Bauer⁹⁸ are not supported by experiment.

A recent paper⁹⁹ sets out the present understanding of the cellular replication of chromosomal DNA but gives no consideration to those demonstrations of strand separation, in the absence of proteins, such as those reviewed in Current Science³⁰.

THE RESULTS OF SOME OTHER BIOCHEMICAL DISCIPLINES WHICH BEAR ON THE STRUCTURE OF DUPLEX DNA

Scanning Tunnelling Microscopy (STM) and Atomic Force Microscopy (AFM) have been used to
study solitary, individual biological macromolecules, often in an aqueous environment. In STM a stream of electrons is propelled into the target and the differential effects exerted by different
parts of the target are recorded. In AFM a very sensitive mechanical probe is dragged across a target and the changes in physical resistance are recorded. The results deducible from these techniques, as applied to DNA, have been reviewed27. In many ways, the results from AFM and STM complement the information accessible to X-ray diffraction which does not lend itself to application under the same conditions. Though some aspects of the results obtained from AFM and STM applied in a horizontal direction are prone to objection on the grounds that they might give
rise to artefacts, the application of these techniques in the vertical direction is more secure, especially in AFM, as the actual vertical displacements of the probe can be measured and are
self-consistent across many studies27.

As is well known, the measured heights of DNA do not accord with the expected values for a
double helix, 2.0 – 2.2 nm: "... the reduced height of DNA molecules [compared with those expected for double helical duplex DNA] observed with the atomic force microscope remains a "paradox" …100 “ and a “puzzle" 101. Nor is that the only puzzle. The scans variously show that single and double stranded nucleic acids mostly have the same heights, within experimental
error, of ~1 nm and 1.05 ± 0.16 nm102,103. These heights would seem to accord with the
inner helical diameters determined for single helices by X-ray diffraction from the helix of
principal diffractors, such as that of 0.772 nm for poly (ribo-cytidylic acid) (27, page 74; and
104) and 0.774 nm by Zimmerman et al.105 for poly (ribo-adenylic acid).

That single and double helices of DNA might actually have the same diameter lying on an AFM
plate would seem improbable, perhaps even impossible, for the double helix. However, with the paranemic model of duplex DNA27,30, for example, when lying flat on a plate, the only possible result is that single and duplex DNA strands would indeed have the same height.

TESTING THE HYPOTHESIS THAT Z-DNA MIGHT BE PARANEMIC AND RIGHT-HANDED AS IN FIGURE 1

1        To determine whether Z-DNA has a double-helical or a paranemic structure

A direct test might utilise AFM to distinguish between a double helical structure and the
paranemic model for the Z form.  It is known that brominated poly (dG.dC)_2, having 45% ^Br8G
and 20% ^Br5C, has been reported as adopting the Z conformation in all conditions of
temperature, humidity and salt concentration cited in the literature to date¹⁰⁶.

At low salt concentrations, and at low but not zero humidity, the unbrominated poly (dG.dC)₂
would adopt the B form and the brominated polymer would adopt the Z form.

Although the artefacts which arise with AFM have attracted much adverse comment, the height measurements made on AFM platens seem secure ^27,28 .  Then the B form polymer, if it were a double helix, would give rise to a height measurement under AFM of some 2.0 to 2.2 nm (a
result never reported to date), but would give rise to a value of some 1.3 to 1.4 nm if it were paranemic (a result reported many times ^27,28).  The Z form would give rise to a height value of some 1.8 nm if it were a double helix, and to a value of 0.9 to 1.0 nm if it were paranemic.

TABLE 1

Predicted Polymer Heights under AFM

These height differences are large enough to be distinguishable using AFM.

1        To test the hypothesis that the Z form is right handed

This topological experiment offers a decisive determination of the handedness of B-DNA compared with that of Z-DNA.  It is known that (dG.^m5dC)_n forms Z-DNA under more mild conditions than does a sequence such as (dG.dC)_n. A sample of (dG.dC)_n having a convenient, uniform length of, say, 100 bp, is fully methylated enzymatically at all ⁵C positions and kept in conditions such that it retains the B conformation. 

A suitable, natural covalently closed circular (ccc) duplex DNA of known length, say, 200 bp, and a site for each of two different restriction endonucleases,   is enzymatically demethylated at cytosine, cleaved with a restriction endonuclease at one site and given blunt ends.  The methylated sequence, itself with blunt ends and under conditions which maintain it in the B conformation, is then joined to one end of the opened natural DNA and the molecule circularly closed at as low a temperature as quickly as possible to introduce as few superhelical turns as practicable in the closure reaction.

This molecule, containing the methylated insert, is recovered from the other DNA species and divided into two aliquots.  Half of the molecules containing the methylated insert are then cleaved at one site outside the inserted sequence and the conditions changed so that the methylated sequence forms Z-DNA while the natural sequence retains the B conformation, for example, at a raised level of monovalent ion and with the addition of divalent ions (107), and/or stabilising the Z-DNA with anti-Z-DNA antibodies (108).  The molecules with the stretch of Z-DNA are circularly closed again, under the exact conditions of temperature and enzyme concentration employed in the earlier ring closure, and returned to conditions such the methylated sequence again adopts the B conformation. 

According to the proposition that Z-DNA has a handedness opposite to that of B-DNA, the two aliquots will now have very different numbers of superhelical turns from each other.  According to the new thesis in this paper, however, they will have rather similar numbers of superhelical turns because the Z-DNA stretch will have retained the same handedness as B-DNA all the time.                            

It is important to ascertain by another, independent method whether the Z polymorph was present in the aliquot immediately after ring closure, and before it is returned to the B conformation, by using Z-DNA antibodies (for example, 109, 110, 111) or spectroscopic techniques (for example, 112, 113, 114). 

The binding of anti-Z-DNA antibodies, and the confirmed spectroscopic correlations for the presence of Z-DNA only demonstrate that Z-DNA is present - they do not demonstrate that left-handed DNA is present.  The binding of antibodies and spectroscopic resonant frequencies are not securely identified with absolute configurations in space but are only correlated with the Z conformation as such, whose helical handedness is here in dispute.                                                            

Summary

The proposition that Z-DNA is a uniformly left-handed double helix has been probed by challenging the meaning of the original CD spectral data, the X-ray evidence deduced from oligodeoxyribonucleotide crystals, explaining a previously unexplained paradox in AFM molecular heights, and challenging the measurement of inter-proton distances found by NMR when based upon X-ray diffraction studies.  The kinetics and mechanisms of the chemical and biochemical behaviour of Z-DNA offer further difficulties to a double-helical, uniformly left-handed paradigm.

Variously obviating some of these difficulties are the SBS models of Sasisekharan and Rodley and their coworkers^{19 – 22, 23 – 26}, while the paranemic model of duplex DNA^27,30 would seem to have the potential to resolve all the difficulties raised in this paper.

References

1          Claude E. Gagna, W. Clark Lambert, Hon-Reen Kuo, and Patricia N. Farnsworth; Localization of B-DNA and Z-DNA in Terminally Differentiating Fiber Cells in the Adult Lens; J Histochem Cytochem, 1997, 45,1511-1521

2             Sung Chul Ha, Neratur K., Lokanath, Dong Van Quyen, Chun Ai Wu, Ky Lowenhaupt , Alexander Rich , Yang-Gyun Kim, and Kyeong Kyu Kim;A poxvirus protein forms a complex with left-handed Z-DNA: Crystal structure of a Yatapoxvirus Z bound to DNA; Proc Nat Acad Sci (USA); 2004, 101, 14367 - 14372

3             Steven D. Sheridan, Craig J. Benham, and G. Wesley Hatfield; Inhibition of DNA Supercoiling-dependent Transcriptional Activation by a Distant B-DNA to Z-DNA Transition, J Biol Chem, 1999, 274, 8169-8174

4             Thomas Schwartz, Joachim Behlke, Ky Lowenhaupt, Udo Heinemann & Alexander Rich; Structure of the DLM-1−Z-DNA complex reveals a conserved family of Z-DNA-binding proteins; Nature Structural Biology 8, 761 - 765 (2001)

5             Anitha Suram, Jagannatha, K., Rao, S., Latha K. S. Viswamitra M. A.; First Evidence to Show the Topological Change of DNA from B-DNA to Z-DNA Conformation in the Hippocampus of Alzheimer's Brain; NeuroMolecular Medicine, 2002, 2, 289-298

6             Schade, M., Turner, C. J., Kühne, R., Schmieder, P., Lowenhaupt, K., Herbert, A., Rich, A. & Oschkinat, H., The solution structure of the Zα domain of the human RNA editing enzyme ADAR1 reveals a prepositioned binding surface for Z-DNA, Proc. Natl. Acad. Sci. USA, 1999, 96, 12465 – 12470

7             Schade, M., Turner, C. J., Lowenhaupt, K., Rich, A. & Herbert, A., Structure-function analysis of the Z-DNA-binding domain Zα of dsRNA adenosine deaminase type 1 reveals similarity to the (α + β) family of helix-turn-helix proteins, The EMBO J., 1999, 18, 470 – 479

8             Rothenburg, S., Schwartz, T., Koch-Nolte, F. & Haag, F., Complex regulation of the human gene for the Z-DNA binding protein DML-1, Nucl. Acids Res., 2002, 30, 93 – 1000

9             Műller, V., Takeya, M., Brendel, S., Wittig, B. & Rich, A., Z-DNA-forming sites within the human -globin gene cluster; Proc. Natl. Acad. Sci. USA, 1996, 93, 780 – 784

10           Pan, B. & Sundaralingam, M., Crystal structure of rGd(CGCGCG): a Z-DNA hexamer duplex with a 5’-(rG) overhang, Acta Cryst., 2003, D59, 433 – 437

11           Wang, A. H.-J., Quigley, G. J.,  Kolpak, F. J., Crawford, J. L., van Boom, J. H., van der Marel, G. & Rich, A., Molecular Structure of a Left-handed Double Helical DNA Fragment at Atomic Resolution, Nature, 1979, 282, 680 - 686

12           Sasisekharan, V. & Brahmachari, S. K., B To Z Transition in a DNA Fibre: The Question of Handedness of the Duplex, Current Science, 1981, 50, 10 -13

13           Leslie, A. G. W.,  Arnott, S., Chandrasekaran, R. and Ratliff, R. L., Polymorphism of DNA Double Helices, J. Mol. Biol., 1980, 143, 49-72

14           Mahendrasingham, A., Forsyth, V. T., Hussain, R., Greenall, R. J., Pigram, W. J. & Fuller, W., Time-resolved X-ray Diffraction Studies of the B to D Structural Transition in the DNA Double Helix, Science, 1986, 233, 195-197

15           Franklin, R. & Gosling, R. G., The Structure of Sodium Thymonucleate Fibres. I., The Influence of Water Content, Acta Cryst., 1953, 6, 673 – 677

16           Marvin, D. A., Spencer, M., Wilkins, M. H. F. & Hamilton, L. D., The Molecular Configuration of Deoxyribonucleic Acid. III. X-ray Diffraction Study of the C form of the Lithium Salt, J. Mol. Biol., 1961, 3, 547-565

17           Wing, R., Drew, H., Takano, T., Broka, C., Tanaka, S., Itakura, K. & Dickerson, R. E., Crystal Structure Analysis of a Complete Turn of B-DNA, Nature, 1980, 287, 755-758

18           Drew, H. R., Wing, R. M., Takano, T., Broka, C., Tanaka, S., Itakura, K. & Dickerson, R. E., Structure of a B-DNA Dodecamer: Conformation & Dynamics, Proc. Natl. Acad. Sci. USA 1981, 78, 2179-2183

19           Sasisekharan, V. & Pattabiraman, N., Double Stranded Polynucleotides: Two Typical Alternative Conformations for the Nucleic Acids, Current Science, 1976, 45, 779 – 783

20           Sasisekharan, V., Pattabiraman, N. & Gupta, G., Some Implications of an Alternative Structure for DNA, Proc. Natl. Acad. Sci. USA, 1978, 75, 4092 – 4096

21           Gupta, G., Bansal, M. & Sasisekharan, V., Conformational Flexibility of DNA: Polymorphism & Handedness, Proc. Natl. Acad. Sci. USA, 1980, 77, 6486-6490

22           Gupta, G, Bansal, M. & Sasisekharan, V., Polymorphism & Conformational Flexibility: Right & Left Handed Duplexes, Int. J. Biol. Macromol., 1980, 2, 368-380

23           Rodley, G. A., Scobie, R. S., Bates, R. H. T. & Lewitt, R. M., A Possible Conformation for Double-Stranded Polynucleotides, Proc. Natl. Acad. Sci. USA, 1976, 73, 2959 -2963

24           Rodley,G. A. & Bates, R. H. T.; Is DNA really a double helix ?, Trends in Biochemical Science, 1984, 5, 231 - 234

25           Rodley, G.A., Millane, R.P., McKinnon, G.A., & Bates, R.H.T.,  Use of Axial Patterson in Assessment of Compatibility of Alternative B-DNA Structures with Fibre X-ray Data, J Roy Soc N Z, 1984, 14, 223 - 231

26           Rodley, G. A., Reconsideration of some results for linear and circular DNA; J. Biosc., 1995, 20, 245 - 257

27           Delmonte, C. S., ‘A New Structural Molecular Biology’, 1991, 1997 & 2003; ISBN 0-9512276-0-2; fully and freely downloadable copy of the whole book is available from: http://www.notahelix.com/delmonte/new_struct_mol_biol.pdf 

28           Delmonte, C., ‘Advances in AFM & STM Applied to the Nucleic Acids’, available in public libraries or via the author,1997, ISBN 0-9512276-2-9

29           Delmonte, C., ‘DNA Structure Puzzles’, available in public libraries or via the author, 2000, ISBN 0-9512276-3-7;

30           Delmonte, C. S. & Mann, L. R. B., Variety in DNA Secondary Structure, Current. Science. 2003, 85, 1564 – 1570, freely downloadable from:

                http://www.ias.ac.in/currsci/dec102003/1564.pdf

                with subsequent correspondence from:

http://www.ias.ac.in/currsci/may252004/1352.pdf

31           Lee, G., Arscott, P. G., Bloomfield, V. A. & Fennell Evans, D., Scanning Tunnelling Microscopy of Nucleic Acids, Science 1989, 244, 475‑477                        

32           Beebe, T. P., Wilson, T. E., Ogletree, D. F., Katz, J. E., Balhorn, R., Salmeron, M. B. & Siekhaus, W. J., Direct Observation of Native DNA Structures with the Scanning Tunneling Microscope, Science 1989, 243, 370-372

33           Barton, J. K., Kumar, C. V. & Turro, N. J., DNA-Mediated Photoelectron Transfer Reactions, J. Am. Chem. Soc., 1986, 108, 6391-6393

34           Barton, J. K., Metals and DNA: Molecular Left-Handed Complements, Science, 1986, 233, 727-734

35           Iwamoto, S. & Hsu, M., Determination of twist & handedness of a 39-base pair segment of DNA in solution, Nature, 1983, 305, 70-72

36           Sprecher, C. A. & Johnson, W. C., Change in Conformation of Various DNAs on Melting as Followed by Circular Dichroism, Biopol., 1982, 21, 321-329

37           Thamann, T. J., Lord, R. C., Wang, A. H.-J. & Rich, A., The High Salt Form of Poly(dG-dC).poly(dG-dC) is Left-handed Z-DNA: Raman Spectra of Crystals & Solutions, Nucl. Acids Res., 1981, 9, 5443-5457

38           Moeller, A., Gabriels, J. E., Lafer, E. M., Nordheim, A., Rich, A. & Stollar, B. D., Monoclonal Antibodies Recognise Different Parts of Z-DNA, Proc. Natl. Acad. Sci. USA, 1982, 257, 12081-12085

39           Nordheim, A., Pardue, M. L., Weiner, L. M., Lowenhaupt, K, Scholten, P., Moller, A., Rich, A. & Stollar, B. D., Analysis of Z-DNA in Fixed Polytene Chromosomes with Monoclonal Antibodies That Show Base Sequence-dependent Selectivity in Reactions with Supercoiled Plasmids & Polynucleotides, J. Biol. Chem., 1986, 261, 468-476

40           Castleman, H. & Erlanger, B. F., Electron Microscopy of “Z-DNA”, Cold Spring Harbor Symposia Quantitative Biology, 1982, 47, 133 – 142

41           Saenger, W., Principles of Nucleic Acid Structure, Springer-Verlag, 1984, ISBN 3 – 540 – 90761 – 0, page 284

42           Pohl, F. M. & Jovin, T., Salt-induced Cooperative Conformational Change of a Synthetic DNA: Equilibrium & Kinetic Studies with Poly(dG.dC), J. Mol. Biol. 1972, 67, 375-397

43           Pohl, F. M., Ein Modell der DNS-Struktur, Naturwissenschaften, 1967, 54, 616

44           Pohl, F. M., Polymorphism of a Synthetic DNA in Solution, Nature 1976, 260, 365-366

45           Gabbay, E. J., Topography of Nucleic Acid Helices in Solutions. XII. The Origin of the Oppositely Induced Circular Dichroism of Reporter Molecules Bound to Ribo- & Deoxyribonucleic Acid, J. Am. Chem. Soc., 1969, 91, 5136-5150

46           Gabbay, E. J., DeStefano, R. & Sanford, K., Topography of Nucleic Acid Helices in Solutions. XXIV. Intercalation Specificities of DNA & Their Possible Role in the Recognition Process, Biochem. Biophys. Res. Comm., 1972, 46, 155-161

47           Zarlenga, D. S., Halsall, H. B. & Day, R. A., A  Polycationic Amine that Induces Unique Conformational Changes in Poly (dA.dT) in Low Salt, Nucl. Acids Res., 1984, 12, 6325-6335

48           Tomasz, M., Barton, J. K., Magliozzo, C. C., Tucker, D., Lafer, E. M. & Stollar, B. D., Lack of Z-DNA Conformation in Mitomycin-modified Polynucleotides Having Inverted Circular Dichroism, Proc. Natl. Acad. Sci. USA, 1983, 80, 2874-2878

49           Taboury, J. A. & Taillandier, E., Right-handed & Left-handed Helices of Poly(dA-dC).(dG-dT), Nucl. Acids Res., 1985, 13, 4469-4483

50           Sutherland, J. C. & Griffin, K. P., Vacuum Ultraviolet Circular Dichroism of Poly(dI-dC).poly(dI-dC): No Evidence for a Left-handed Double Helix, Biopol., 1983, 22, 1445-1448

51           Uesugi, S., Ohkubo, M., Ohtsuka, E., Ikehara, M., Kobayashi, Y., Kyogoku, Y., Westerink, H. P., van der Marel, G. A., van Boom, J. H. & Haasnoot, C. A. G., Conformation of Ribooligonucleotide Duplexes Containing an Alternating C-G Sequence Which Show an Unusual Circular Dichroism Spectrum, J. Biol. Chem., 1984, 259, 1390-1393

52           Riazance, J. H., Baase, W. A. & Johnson, W. C., Evidence for a Z-form RNA by Vacuum UV Circular Dichroism, Nucl. Acids Res., 1985, 13, 4983-4989

53           Lewis, D. G. & Johnson, W. C.; Circular Dichroism of DNA in the Vacuum Ultraviolet, J. Mol. Biol., 1974, 86, 91-96

54           J.C. Sutherland, J. C., K.P. Griffin, K. P., P.C. Keck, P. C. & P.Z. Takacs, P. Z., Z-DNA: Vacuum Ultraviolet Circular Dichroism, Proc. Natl. Acad. Sci. USA, 1981, 78, 4801-4804

55           Behe, M. J., Vacuum UV CD of the Low-salt Z-forms of Poly(rG-dC).poly(rG-dC), and Poly(dG-m5dC).poly(dG-m5dC, Biopol., 1986, 25, 519-523

56           Ikehara, M., Hattori, M. & T. Fukui, T., Synthesis & Properties of Poly (2‑Methyladenylic Acid): Formation of a Poly(A). Poly(U) Complex with Hoogsteen‑Type Hydrogen Bonding, Eur. J. Biochem., 1972, 31, 329‑334

57           Pohl, F. M., Salt-induced Transition between Two Double-helical Forms of Oligo (dC.dG), Cold Spring Harbor Symposia on Quantitative Biology, 1983, 47, 113-117

58           Sadasivan, C., Karthe, P. & Gautham, N., Space-Group Degeneracy in the Packing of a Non-Selfcomplementary Z-DNA Hexamer, Acta Cryst., 1994, D50, 192 - 196

59           Malinina, L., Tereshko, V., Ivanova, E., Subirana, J. A., Zarytova, V. & Nekrasov, Y., Structural Variability and New Intermolecular Interactions of Z-DNA in Crystals of d(pCpGpCpGpCpG), Biophys. J., 1998, 74, 2482 – 2490

60           Chevrier, B., Dock, A. C., Hartmann, B., Leng, M., Moras, D., Thuong, M. T. & Westhof, E., Solvation of the Left-handed Hexamer (^5BrCG)₃ in Crystals Grown at Two Temperatures; J. Mol. Biol., 1986, 188, 707-719

61           S. Fujii, S., A.H.-J. Wang, A. H.-J., G. van der Marel, G., J.H. van Boom, J. H. & A. Rich, A., Molecular Structure of (^m5dC.dG)₃; The role of the methyl group on 5-methylcytosine in stabilising Z-DNA; Nucl. Acids Res., 1982, 10, 7879-7892

62           Wang, A. H.-J., Quigley, G. J., Kolpak, F. J., van der Marel, G., van Boom, J. H. & Rich, A., Left-handed Double Helical DNA: Variations in the Backbone Conformation, Science, 1981, 211, 171-176

63           Pohl, F. M., Salt-induced Transition between Two Double-helical Forms of Oligo (dC.dG), Cold Spring Harbor Symposia on Quantitative Biology, 1983, 47, 113-117

64           J.L. Crawford, J. L., F.J. Kolpak, F. J., A.H.-J. Wang, A. H.-J., G.J. Quigley, G. J., J. van Boom, J., G. van der Marel, G. & A. Rich, A., The Tetramer  d(CpGpCpGp) Crystallises as a Left-handed Double Helix, Proc. Natl. Acad. Sci. USA, 1980, 77, 4016-4020

65           Hopkins, R. C., Deoxyribonucleic Acid Structure: A New Model, Science, 1981, 211, 289 – 291

66           Uesugi, S., Lee, B. L., Ikehara, M., Fujii, S. & Tomita, K., Hybrid Hexanucleotide Duplex Containing Cyclonucleotides and Deoxynucleotides: The d(TA) Segment Can Adopt a High Anti Left-Handed Double-Helical Structure, Biochem., 1988, 27, 521 - 525

67           Ringe, D., Protein Structure: An Extra Dimension to NMR; Nature, 1988, 332, 303

68           Oschkinat, H., Griesinger, C., Kraulis, P. J., Sorenson, O. W., Ernst, R. R., Gronenborn, A. M. & Clore, G. M., Three-dimensional NMR Spectroscopy of a Protein in Solution, Nature, 1988, 332, 374-376

69           Shindo, H., Backbone Geometry of Oriented DNA Fibers as Revealed by ³¹P Chemical Shielding Anisotropy, Adv. Biophys., 1985, 20, 39-57

70           Cheng, D. M., Kan, L.-S., Frechet, D., T'so, P.O.P., Uesugi, S., Shida, T. & Ikehara, M.,¹H- and ³¹P-Nuclear Magnetic Resonance Studies on the Conformation of d-(CpGpCpG)₂ and d-(CpGpCpGpCpG)₂ Short Helices in B-Conformation; Biopol., 1984, 23, 775-795

71           Reid, B. R., Proton Assignment Strategies in Nucleic Acid NMR Studies, Chapter 3 in NMR in the Life Sciences; Eds. Bradbury, E. M. & Nicolini, C., Plenum Press, 1986, ISBN 0-306-42279-4

72           Kalk, A. & Berendsen, H. J. C., Proton Magnetic Relaxation & Spin Diffusion in Proteins, J. Mag. Res., 1976, 24, 343-366

73           Chazin, W. J., Wüthrich, K., Hyberts, S., Rance, M., Denny, W. A. & Leupin, W., ¹H Nuclear Magnetic Resonance Assignments for d-(GCATTAATGC)₂ using Experimental Refinements of Established Procedures, J. Mol. Biol., 1986, 190, 439-453

74           Wüthrich, K., NMR of Proteins & Nucleic Acids, Wiley-Interscience, 1986, ISBN 0-471-82893-9

75           Mahendrasingham, A., Forsyth, V. T., Hussain, R., Greenall, R. J., Pigram, W. J. & Fuller, W., Time-resolved X-ray Diffraction Studies of the B to D Structural Transition in the DNA Double Helix, Science, 1986, 233, 195-197

76           Forsyth, V. T., Greenall, R. J., Hussain, R., Mahendrasingham, A., Nave, C., Pigram, W. J. & Fuller, W., X-ray High-angle Fibre Diffraction Studies of Nucleic Acid Structure Using the Daresbury Synchrotron Radiation Source, Trans. Biochem. Soc. (London), 1986, 14, 553-557

77      &n