Science like sport has rules and procedures. One of the most important procedures is that if you want to claim something in science you have to devise an experiment and test the proposition. You then publish your procedures and your results so everyone can see what you did.
If people are happy with what you have done your claim is accepted. If they don’t like it they can repeat the work to check you haven’t made a mistake – or cheated – and if they come up with the same results your work and your claim is validated and accepted.
These rules have the great merit of simplicity in theory and are supposed to ensure that no good work is lost because the procedure for verifying is so simple.
The story of Sasi and his side by side (SBS) alternative to the Watson Crick double helix (W-C) model of DNA is an example of these simple rules being broken and potentially critically important work being ignored.
Sasi and his supporters are free to make this accusation quite simply because his work was initially criticised and then dismissed/forgotten even though no one has ever bothered to repeat it in over 40 years.
If Sasi had been a junior researcher, if this work had been published in some obscure journals or had been examining some unimportant field it would perhaps matter less. But as you will see (this website Stokes Papers 14, p 224 para 3; and Stokes Papers 15b Ch II p 31-34) Sasi was at the time one of the very few people in the world qualified to undertake this work, the DNA molecule is one of the most important molecules known to science and his work was published in leading journals including Current Science, Nature, Proceedings of the National Academy of Science (PNAS) and Nucleic Acids Research.
It is effectively impossible to watch DNA in action. To view it we have to extract it from the cell. To do this we have to follow the procedures like those outlined on the DNA extraction procedure page on this website (Johnson Material 18). This is not subtle and by definition it strips away any and all associated material and leaves a cold and naked molecule.
DNA is not and never has been a free entity in vivo but in order to understand it and work out its nature you have to strip it down and identify its range of movement in the same way as you might a human skeleton.
Stripped down DNA is too small for us to observe directly. Our assessment of it is by various techniques including x ray analysis, diffraction patterns etc which are almost by definition being done on a still/cystallised molecule.
If you are at a dance/disco and it is almost completely dark you know people are dancing but can’t say for sure exactly what moves they are making. Once the strobe light starts you get snapshots of their moves and from those you can infer the full arcs of their movement. This how individual film/cartoon frames work.
This is effectively what Sasi and his team did. They generated a series of crystals of each element of the molecule and analysed those through different ranges of movement. It was from these that they generated the overall picture of the molecule.
This part of the team’s approach was not unique but what was unique was that they prepared much more complex crystals than had been used previously and they did not limit their investigation by only examining right-handed options (supportive of the W-C) as most other groups previously had.
This aspect of their approach is what drew the attention of the philosophy of science student Terry Stokes and he did his PhD thesis on the role of logic in scientific innovation. He used this group’s work as his worked example and in the process has left us a remarkable diary/analysis/character study of the key players in DNA research of the time. This thesis is reproduced here in full (Stokes Papers 15 PhD thesis) together with a shorter dissertation (Stokes Papers 14 Dissertation) both with Prof Stokes’ kind permission.
THE BACKBONE
Up to Sasi’s research analysis of the secondary structure of the molecule had used mono nucleosides – very simple crystal preparations of the backbone elements or of the bases themselves. What Sasi had his team do was prepare and analyse di, tri and tetra nucleotide crystals. This meant that their analysis of the sugar phosphate ‘backbone’ included the effects of base stacking in creating the conformations. This allowed them to demonstrate the great range of viable options around the phosphodiester bonds in the sugar phosphate chains. (See Sasisekharan Papers 8)
They showed that the deoxy ribonucleic ‘backbone’ of DNA has 4 low energy or stable positions with relatively low energy pathways between them on 3 sides. In effect the molecule can twist from position 1 to 2 to 3 to 4 and back again but it cannot go from 1 to 4 ie it can’t twist full circle, a bit like a human backbone.
By contrast the ribonucleic backbone of RNA has only 2 stable positions because the extra hydroxyl group on C2 of the sugar restricts the rotation options of the phosphodiester bond (Sasisekharan Papers 8 p 188).
Non specialists may be surprised to learn that DNA is far more flexible than RNA which is simply not what the double helix model suggests or even what we learn about tRNA weaving its way through the cell.
We know that DNA emerged from the pre-existing ribonuclear protein world of which RNA, ribosomes and a few other molecules are retained functional remnants. The loss of its hydroxyl group made DNA capable of doing its job far more effectively than its ribonuclear precursor.
Because of the absence of the extra hydroxyl group compared to RNA it doesn’t make cross connections and so it maintains its overall shape no matter its sequence or length. DNA synthesis goes 50x quicker than RNA synthesis and protein synthesis.
UNWINDING: A MAJOR PROBLEM WITH THE WATSON CRICK DOUBLE HELIX
It is most unusual in nature for a molecule to be wrapped around itself like the W-C. This twisting makes it extremely difficult for it to disentangle itself and disentangling either partially for template formation or fully for replication lie at the heart of DNA’s most critical functions.
The other thing that this makes very difficult is the binding of the histones. They need to find their binding sites by travelling round the molecule to the major groove. They then have to be out of the way during replication and relocate on daughter molecules.
This very important problem was addressed head on by Crick and Watson in their follow up paper in 1953 and a very good history of this can be read in (Stokes Papers 15b Ch III ‘The Unwinding Problem’ p 48).
Following similar work by other authors in 1975 the Polish biologist Gorski outlined the problem (Stokes Papers 14 p217 para 4 and Stokes Papers 15b p51 para 4). The observed time it takes for the gut bacterium E.coli to replicate is between 30 and 41 minutes. With a single DNA strand of nearly 300,000 base pairs (2.82 x 105) which given the helical turn h of 10 for the molecule means the whole molecule would have to rotate at 6,900 – 9,000 revs/min or 115-150 revs/sec.
At the same time the associated histones etc would have to be delivered to the daughter strands which would themselves be connected to the still spinning mother molecule and would be rotating themselves as they took up their double helix form. All of this in an aqueous environment with highly polarised components attracting water molecules, creating drag etc.
In higher organisms there would be several DNA molecules performing this act of madness in close quarters simultaneously.
THE CHALLENGE IS TAKEN UP
This was the state of knowledge of DNA replication up until the mid 1970’s. Two decades after Watson and Crick identified this major problem there was no solution to it and yet no one had seriously examined an alternative model.
This is where Sasi and another, more theoretical group in New Zealand came in proposing the side by side model. (Stokes Papers 14 p 222 para 2 and p 224 para 3 and Stokes Papers 15b Ch I NZ and Ch II Indian). The point of the SBS model being that the unwinding is simply not an issue. There is no unwinding to be done because the two chains lie side by side and are simply pulled apart.
In addition the histone bonding is very straightforward on such an open structure as is histone release during template formation and replication.
In 1970-71 Sasi was Visiting Professor at Princeton working alongside Rob Langridge who had done the first study of base stacking in 1960. This and subsequent papers by Pullman in 1962 and De Voe in 1968 all assessed right-handed stacking only, the others following Langridge’s lead.
In 1970 the Matsui research group observed from both X-ray and circular dichroism (CD) data that the synthetic D-DNA was left-handed. They couldn’t convert it into A-DNA or B-DNA, regarded as the most similar to naturally occurring forms.
By doing more careful and detailed work than the Matsui group Sasi showed that D-DNA could be converted to A-DNA and B-DNA which were evidently right-handed by both X-ray and CD data. Because he was able to do this by simply changing the humidity and such a simple change could not be responsible for something as profound as a change in helical sense he concluded that all 3 forms were right handed.
However the anomaly of the CD data indicating that D-DNA was left-handed remained. Sasi would take up this issue with Langridge, who had been one of the Matsui researchers. Sasi describes Langridge and being ‘evasive’ on the subject. (Stokes Papers 14 p225 and Stokes Papers 15b ChII p32 para 39).
This may be because of the implied criticism of the Matsui group’s methods or of Langridge’s original base stacking work not having examined left-handed stacking or else as Sasi acknowledges because Langridge was distracted by plans to set up his new lab in San Francisco.
The observation of the anomaly and of the imprecision of some of the work lead Sasi to a take a significant decision. He left Princeton to take up his position as head of the Molecular Biophysics Unit at the Indian Institute of Science in Bangalore. As head of department he would be able to initiate his own research and he decided to examine all the component elements of the DNA molecule in fine and accurate detail.
The big problem was that he could recruit no molecular biologists or biochemists as doctoral research assistants. This was because he was proposing to examine left-handed possibilities and they viewed this as too unorthodox and career threatening to get involved with. (Stokes Papers 14 p 226 para 2 and Stokes Papers 15b Ch II p 35)
Finally after 2 years he recruited N. Pattabiraman and G. Gupta, a nuclear physicist and particle physicist respectively, one of whom had never heard of DNA. This ignorance was an advantage. They had no preconceptions and were very good mathematicians.
These three were the core research team responsible for the initial set of key papers published in Current Science, Nature, PNAS and Nucleic Acids Research from 1976 to 1980. (Sasisekharan Papers 1-8).
THE CORE FINDINGS
Sasi’s team found that there was great flexibility in the sugar phosphate backbone. Like a regular chain it had strong covalently bonded links running along its length but that like a regular chain it had great flexibility between links (Sasisekharan Papers 1 and 2). They found that the bases bond covalently to the links and then bond by relatively weak hydrogen bonding to each other, holding the two chains together. This much is pretty much agreed between SBS and W-C. Sasi also agreed that the lowest energy stacking angle was 35o.
The departure for Sasi was that the lowest energy base stacking could be 35o for right or left-handed in most cases and that there are viable energy gradients between the two stacks again in most cases. Their two papers in Nucleic Acids Research (Sasisekharan Papers 3 and 4) shows these sequence specific stacking energy profiles.
What this implies is that the tertiary structure of the molecule is determined by the stacking between the bases and so the tertiary structure of the molecule is sequence specific. (Sasisekharan Papers 5-10).
INDEPENDENT CORROBORATION? Another left-handed form appears
The front cover of Nature Vol 282 Issue 5740 published on 13th December 1979 (Johnson Material 16, p 28 and online search) heralded the discovery of the ‘only’ left-handed DNA known to science. This was the famous Zig Zag which had been discovered by a group at MIT headed by Alex Rich. This model is still a double helix but clearly shows that left-handed elements are viable and significant.
Given that Sasi had published a paper outlining the parameters of left and right-handed stacking in Nature the year before (Sasisekharan Papers 6) with references to all his previous papers it is very strange from a science procedure/rules point of view that the MIT authors should claim that their model was the ‘only’ left handed example known to science. They do reference Sasi’s 1978 PNAS paper (Sasisekharan Papers 5) but still claim uniqueness. It is also strange that the referees and then the editor of Nature indulged their claim.
Sasi’s work is a comprehensive analysis of all the component elements of the molecule whereas the MIT work is a one off. You can see Sasi’s account of the background to the MIT announcement in the letter written to A Johnson in February 1982 (Sasisekharan Papers 11).
What is so interesting is that the computer the MIT team were using had no preconceptions about DNA structure either and despite their best efforts to ‘make’ it support the W-C the computer forced them to realise their mistake.
(Up to the Zig Zag one authority had even taken to saying that left-handed DNA couldn’t exist because the ribose sugars are in the dextro rather than laevo chiral form. This is like saying cars with the steering wheel on the left can’t make right hand turns. There is this sense of people trying much too hard.)
What is very surprising is that this independent support for such a significant part of Sasi’s work didn’t lead to his work being re-evaluated or repeated. Part of this has to do with a shift in science publishing which impacted science practice in profound and unhelpful ways from the mid 1970’s.
This shift is well described in this article by science journalist Stephen Buranyi in the UK’s Guardian newspaper on 27 June 2017 ‘Is the staggeringly profitable business of scientific publishing bad for science?’. This outlines that one of the changes in publishing/funding was that researchers were dis incentivised from repeating other people’s work. This disrupted the important verification/refutation process which is a core part of science rules and procedures. This is damaging for science in general and in this particular case it’s critical.
The ascription of the Zig Zag DNA as a ‘high salt’ anomaly and the ‘only’ left-handed version of the molecule became the received wisdom. It is deeply problematic and the failure to carefully examine this left-handed ‘anomaly’ and consider the implications of it is a glaring failure of science procedure/rules.
The way this fudge was so easily accepted makes about as much sense as when people say, ‘All xxxxx’s are a bit dumb but I have a friend who is an xxxxx but she/he is different.’ This is a standard way for human beings to have dodgy attitudes but maintain their human decency by celebrating their ability to make genuine friendships despite them. It’s a cake and eat it thing and has no place in the world as a whole and still less in science.
This loopy ‘high salt anomaly’ logic explanation of the ‘Zig Zag’ was for over a decade included as a footnote in the DNA section of the print edition of Encyclopaedia Britannica for instance. The author of that section on DNA was Alex Rich.
Interestingly the only person to actually look systematically into the effects of changing salt concentrations on DNA structure was Sasi. Working with new team members he produced papers in 1981 and 1984 which show that these changes are reversible (Sasisekharan Papers 9 and 10).
TOPOISOMERASE AND MULTIPLE REPLICATION POINTS
Another significant development in the mid 1970’s that impacted the perception of Sasi’s work was the discovery of the topoisomerases. It was found that replication was initiated and speeded up by the inclusion of these enzymes. This was seized on by researchers as a solution to the problem of the unwinding and, as it used to say in the online encyclopedia entry ‘The advent of topoisomerases killed the side by side models’. The current entry under ‘Obsolete models of DNA structure’ is still pretty emphatic.
Except the discovery of topoisomerases doesn’t confirm the W-C at all. All it establishes is that whatever the structure of DNA may be it replicates faster with topoisomerase than without it. The leap to assuming that it confirms the W-C is again problematic from a science procedure/rules point of view.
Explanations for how topoisomerase unwinds the double helix could very easily be reworked to explain how it would work with SBS.
Other evidence is cited: ‘However, the structure of DNA was subsequently confirmed in solution via gel electrophoretic methods [26] and later via solution NMR[27] and AFM[28] indicating that the crystallography process did not distort it. The structure of DNA in complex with nucleosomes, helicases and numerous other DNA binding proteins also supported its biological relevance in vivo.[29]’
In addition the evidence showing multiple replication points in a single DNA molecule driven along by the work of topoisomerases is offered as confirming the Watson Crick, with the multiple points radically reducing or eliminating the need for any unwinding.
You should try this with a double twisted rope and as many fingers as you can to separate it. You’ll find that it doesn’t solve the problem at all.
The reality of individual and collective human psychology is that once people accept a particular conceptual framework or paradigm they will ‘arrange’ for all evidence to support it. It is just the way human beings are. It’s how our belief systems work.
If people believed in SBS as strongly as they believe in W-C these same arguments would be being made to support SBS. The multiple replication points would be explained as the SBS simply being pulled apart at multiple points.
The whole point is that science rules and procedures were established as a way of negating this well understood human psychological/perceptual problem and preventing subjectivity from wrecking science. It is only when Sasi’s work is repeated and either validated or refuted that a properly objective scientific judgement can be delivered about it.
The following speculations appear in outline in Johnson Material 16 p24 para 4
BIO EVOLUTIONARY PARSIMONY
The great problem with the solutions proposed for W-C function is the need for complex associated action from a number of different molecules/enzymes etc acting in sequence to perform core functions. This would require the ‘passive’ retention of a number of them until the evolution of the last of them allowed the whole process to become ‘live’. This doesn’t work very well from a bio evolutionary standpoint.
The SBS reverses the relationship between DNA and the associated molecules like histones. For W-C these associated molecules are enablers and facilitators of DNA function. With the SBS they are more simply stabilisers, with their removal enabling DNA’s inherent functionality to be expressed.
The SBS model suggests an inherent, sequence specific functionality which is very satisfying from an evolutionary biological point of view. It then sees the histones etc as subsequent embellishments to this basic functionality.
We know that DNA has many long repeat sequences with the important coding sequences between them. In day to day template formation with the SBS model one can envisage an overall increase in temperature causing the hydrogen bonding of the heterogeneous coding sequences to ‘break’ or open, allowing the relevant chain to offer its bases as a template while the homogeneous repeat sequences being more thermally stable remaining closed, like zip ends.
Now it is possible that key associated molecules like the histones and topoisomerases are also ribonuclear world relics that found a new association when DNA emerged. However you would imagine that similar problems around sequential operations and ‘passive’ retentions in that previous system would still apply. Who can say?
The really simple thing to do would be to repeat Sasi’s work. If it was found to be wrong that would be that for Sasi’s SBS – but if it isn’t wrong then a whole new world of research opens up.
THE DYNAMISM THING
Beyond Sasi’s basic assertion of sequence specific tertiary structure as we have seen his work shows that the bases can shift stack from one side to the other across viable energy gradients. Since living systems operate through a range of temperatures way above absolute zero the molecule must be thermally active. This is like a rope ladder with the centre of the treads being hinged and spring loaded. This means that the entire molecule is thermally dynamic.
You can also look at the very simplified animated artist’s impression of what the dynamic SBS might look like (Johnson Material 17). In reality it would be much more complex with each ring molecule of each component changing conformation/flexing as the molecule moved. The contribution of ring flexing to the overall flexibility of the molecule is central to Sasi’s approach.
This dynamic model might be the better representation of the molecule and it makes sense of the idea of associated material like histones having a primary function of limiting/moderating the molecule’s fundamental dynamism and inherent functionality.