Obituary: Professor Gerald Elliott
The early years
He (JT) and H.R. Boot had revolutionised microwave radar by strapping the anodes of the cavity magnetron so that it operated only in the fundamental mode. This created the technology that JT regularly like to tell us had won the war.
In the light of the 1954 papers, the overwhelming question became how the two sets of filaments move past one another and generate tension during contraction. This would require an answer compatible with physics and chemistry, though subtleties could surely have arisen in more than 3 billion years of biological evolution. Attention quickly focussed on the cross-bridges between the two sets of filaments, first observed in electron microscope thin sections by Hugh Huxley (Huxley 1957), who had also done pioneering one-dimensional low-angle X-ray (diffraction) studies of living muscle (Huxley 1953).
I have often pondered where it (her research) might have led her had she lived. In my opinion—though of course I am biased—Jean had the most flexible mind of those four sliding filament pioneers. I sometimes wonder if she could perhaps have made great breakthroughs in a muscle field that has—at least in my view—sadly become rather stagnant.
This remark expresses Gerald’s general feeling about the muscle field to which he made such major contributions. Later in life he became known as something of a ‘disgruntled maverick’ in the field. How did this arise? Sadly the seeds for this were sown at quite an early stage. Gerald’s successful PhD work was on the X-ray diffraction and electron microscopy of molluscan smooth muscles (the ‘muscles of mussels’) which have large paramyosin-containing filaments rather than the myosin filaments in vertebrate muscles. Among many other important observations he found from the electron microscopy of cross-sections of the Portuguese Oyster (Gryphaea angulata) opaque adductor muscle that the thick paramyosin filaments sometimes showed parallel stripes of staining as though the filaments contained layers of paramyosin molecules. Since it was thought that these filaments also contained myosin, probably as a roughly cylindrical layer on the whole surface of the paramyosin core (Squire 1971), the presence of a layered core structure rather than one with circular symmetry in cross-section was surprising; the myosin rods would make many different contacts with the underlying paramyosin layer structure, sometimes on a front face with the so-called ‘Bear–Selby net’ structure, sometimes on an edge, and sometimes on a ‘Bear–Selby net’ at the back with opposite hand (Bear and Selby 1956).
When Gerald had completed his PhD thesis in 1960 he suggested to JT that Hugh Huxley would be the best choice for an external examiner. However, JT in his wisdom, instead of asking Hugh Huxley, decided to ask the unrelated Andrew Huxley (Cambridge Physiology, soon to move to University College London, UCL). Andrew, the other main sliding filament hero, was known as the most feared ‘inquisitor’ of his generation, as anyone who gave a seminar at UCL Physiology in the 60 s or 70 s could testify. Thus Gerald turned up to be faced with one of the greatest scientific brains of the twentieth century, an ‘about to be’ Nobel Prize winner, got hauled over the coals line by line for hours and hours, and had to do a major re-write. In particular, Andrew, commenting on the thesis work, reckoned that the layers seen in paramyosin filament cross-sections ought to be visible in some longitudinal sections, but were apparently never seen. He didn’t like Gerald’s conclusion, but, with some caveats, in 1960 the thesis was finally finished (Elliott 1960).
Gerald used to say, a bit wistfully “Years later it turned out I was right after all… but no-one noticed”. Given that Andrew Huxley was probably the key scientific ‘antagonist’ of Gerald’s career, you could see the viva as rather setting the tone…
Pioneering X-ray diffraction studies of muscle
The axial periodicities in the muscles did not change as the muscle‘s length was changed (as required by the sliding filament model).
The volume of the muscle lattice stayed more or less constant as the sarcomere length was changed (confirming an earlier suggestion by Hugh Huxley).
The intensity of part of the equatorial X-ray pattern, namely the (1,1) peak, gradually reduced as the sarcomere length increased, consistent with the actin filaments being gradually withdrawn from the hexagonal A-band lattice in vertebrate muscles, once again as required in the sliding filament model.
Parts of the diffraction patterns ascribed to actin and myosin filaments had different axial spacings in patterns from vertebrate skeletal muscles, and also from molluscan muscles, but the axial spacings of the myosin and actin filaments appeared similar or related in patterns from insect flight muscle.
Having carried out these successful studies of resting or rigor muscles, Gerald then wanted to study actively contracting muscle in a time-resolved way and was very ingenious in using the available X-ray generator to best advantage. X-rays are produced in conventional X-ray generators by bombarding a metal target (the anode) with a beam of high-speed electrons from a heated filament, the whole system being encased in an evacuated chamber with Beryllium windows through which the X-rays can pass. The problem with this method is that most of the energy of the electrons appears as heat in the water-cooled target and if the electron beam current is too high the anode will melt. How to get a more intense X-ray beam without melting the anode? Gerald remembered a story he had heard from a much-decorated REME instructor “who had told me how he had driven a damaged truck without a cooling system out of the firing line in Normandy. ‘It meant I had to replace the engine’ the instructor said’ but at least I could do it without the b******s shooting at me.’ What Gerald did was to run the generator at 3 mA, rather than the normal 400 μA, but only for a short time. The anode would melt if the system was run at 3 mA for more than 2 s, so he ran the experiment for a second or so and then turned the electron beam off. After a while this could be repeated in a ‘stroboscopic’ experiment so that enough diffracted intensity from an active muscle could be built up over time. To do these experiments in the early 1960s, after Roy Worthington had departed for a job in America, Gerald collaborated with the physiologist Jack Lowy (the one who with Jean Hanson also visualised the actin filament helix for the first time in around (Hanson and Lowy 1963). Gerald was later joined by Barry Millman, a Canadian who had finished his PhD on the physiology of molluscan muscles with Jack, and together Gerald and Barry worked on molluscan and vertebrate muscles in the contracting state. So by the mid-1960s Gerald and his collaborators were ready to publish their findings on active vertebrate striated muscle. Meanwhile Hugh Huxley in Cambridge had been following up his own earlier work on the X-ray diffraction of vertebrate muscle with similar time-resolved studies of the active state. After some negotiations between King’s and Cambridge it was agreed that the two papers would be published back to back in Nature (see detailed comments on the history of this in Elliott 2007a, b).
The two 1965 Nature papers (Elliott et al. 1965; Huxley et al. 1965) were both milestones, being the first X-ray diffraction papers on active vertebrate striated muscle. They showed clearly that: (i) the filament axial spacings did not change significantly between relaxation and full activity, (ii) that the equatorial (1,1) reflection became much stronger in patterns from active muscle relative to the (1,0) reflection, consistent with increase of mass at the actin filament positions (Gerald’s team only), (iii) that the layer-lines from the myosin filaments, apart from those with intensity on the meridian, reduced in intensity when the muscle was active, indicating the movement of myosin cross-bridges and (iv) that the lattice spacing between myosin filaments changed with sarcomere length in the same sort of way in both resting and active muscle (as required for constant volume), although the active spacing was slightly smaller than the relaxed spacing (Gerald’s team only).
During this time in London Gerald was not just working on muscle. He was a School Governor at Dulwich College where he gave talks to encourage the boys to go into science. He also had a life-long interest in socialist and social democratic politics and stood for Parliament as a Labour candidate for Croydon North East in 1966 and 1970, narrowly losing to Bernard Weatherill, later speaker in the House of Commons and to become Lord Weatherill.
Off to America
In 1967 Gerald obtained a position at the Carnegie-Mellon University in Pittsburgh, Pennsylvania where he was appointed Professor of Chemistry and Biology. Here Gerald continued his X-ray studies of muscle as well as electron microscopy and laser diffraction, all in the interest of investigating the constant volume behaviour of muscle (April et al. 1971, 1972). He and his colleagues showed first that muscles of the walking legs of the crayfish (Orconectes) displayed constant volume behaviour, even after the sarcolemma had been removed, so that this appeared to be an intrinsic property of the myofilament lattice. They also showed that with skinned fibres the lattice spacing could be altered by changes in ionic strength and pH in a manner that was consistent with these changes altering the electrostatic forces between the filaments as in a liquid-crystalline solution. It is notable that Gerald’s first post-Doc at Pittsburgh was Ada Yonath who was later (2009) to win the Nobel Prize in Chemistry for her work on the structure of the ribosome.
During this period in the US Gerald and his family also spent three idyllic summers at the Marine Biology Institute at Woods Hole, a favourite summer meeting place for US muscle researchers and others.
The Open University
After Pittsburgh, Gerald moved back to the UK where he was appointed the founding Professor of Physics in the Open University (OU). He was at the OU from 1969 until his formal retirement in 1996. The OU was then relatively new and had no research facilities for Gerald, so through the good offices of Jean Hanson he spent a good bit of his time back at the Drury Lane Biophysics laboratory. There he had fruitful collaborations with Barry Millman, Carl Moos and the Japanese physiologist Ichiro Matsubara. Gerald and Ichiro found that, unlike the crayfish, intact frog semi-tendinosus muscles showed constant volume behaviour, but skinned fibres did not; the sarcolemma is necessary to maintain the constant volume (Elliott and Matsubara 1972). This was explained in terms of Donnan and osmotic equilibria across the sarcolemma. He also continued his work on active muscle (Millman and Elliott 1972). During this time Gerald also contributed to a number of OU undergraduate courses.
Although Gerald’s first love was the mechanism of muscle contraction, in the late 1950s he became convinced that electrical charges also play an important role in maintaining the lateral packing of collagen fibrils in the cornea, which is crucial to its optical transparency. So it was at this Oxford Research Unit that Gerald started in earnest on his research into the structure of the transparent tissues of the eye, particularly the cornea and the lens, a topic which he found he could fund more easily than his work on muscle, despite applying similar physical insights and similar experimental techniques in both research fields. It was a puzzle to him why this was so, and it caused a good deal of frustration. In fact Gerald was the first to correctly interpret the X-ray diffraction pattern from the cornea. This work led to two successive NIH grants, and several ground-breaking publications. Gerald’s ideas about the polyelectrolytic nature of the cornea and maintenance of transparency have long been accepted, and are seen as mainstream within the eye field (Elliott et al. 1980; Sayers et al. 1982).
At the Oxford Research Unit he also recruited Keith Meek, now Professor and Head of the Biophysics Research Group in the School of Optometry and Vision Science at Cardiff University. After Gerald retired from the OU, from 1997 to 2005 he was appointed as Distinguished Research Fellow in Vision Sciences in this group at Cardiff where he continued his successful eye research with Keith Meek and others, including Justyn Regini, one of Gerald’s later OU PhD students, who became a kind of protégé. Gerald retained the title of Emeritus Professor of Physics at the OU, was an Associate Fellow at Green Templeton College and an Honorary Research Associate in the Nuffield Laboratory of Ophthalmology, Oxford.
Gerald’s last muscle theories
Two roads diverged in a wood, and I, I took the one less travelled by, and that has made all the difference
So Gerald and Roy had different ideas from the ‘conventional’ view of muscle. They thought that this caused them to lose out on muscle grant funding and they found it hard to publish their ideas in the mainstream muscle literature. But do they have a point? Maybe now is not the time or place to make a judgement. But there are one or two things highlighted in these papers that perhaps should make us stop and think, whatever the ultimate value of their theory of contraction. For example they cite the paper by Lionne et al. (1996) which shows well-documented results that appear to contradict the conventional theory of ‘independent force generators throughout the A-band’. The conventional view of muscle is that the myosin heads act as independent force generators, so that, for example, the total isometric force that a muscle can generate reduces linearly as the amount of filament overlap reduces when the sarcomere length increases beyond rest length at about 2.1 or 2.2 μm (Gordon et al. 1966). Related experiments on ATP usage as a function of sarcomere length also show a drop as the filament overlap reduces (He et al. 1997). One might expect that ATP usage during unloaded muscle shortening might also depend on sarcomere length in the same way. In other words, the total ATP usage in going from a long length to a fixed shorter length (ATPc) might be expected to be higher if the shortening starts from a longer sarcomere length rather than a shorter one. This does not appear to be the case in the experiments of Lionne et al. (1996) as pointed out by Gerald and Roy. Figure 2 in Lionne et al. (reproduced in Elliott and Worthington 2012) illustrates their experiment where ATP usage was monitored in myofibrils that were allowed to undergo unloaded shortening from two different sarcomere lengths, 2.8 and 3.5 μm. In both cases there was a very fast early phosphate burst, followed by a rapid phase during the shortening itself which then changes abruptly to a slower phase when the shortening had stopped or was much reduced. Lionne et al. studied the size of the phosphate burst and of ATPc for fibrils shortening from a range of different starting sarcomere lengths down to a common shorter length. Apparently in contradiction to the conventional wisdom, the size of the phosphate burst and of ATPc were more or less constant for starting sarcomere lengths over the range from 2.7 to 3.5 μm. As Lionne et al. themselves say: “From a purely enzymatic point of view the situation is difficult to understand; we have made several attempts at fitting it to classical ATPase schemes, but without success.”
The take home message from Gerald is, I think, that we should be more open minded in the muscle research field, that we should not necessarily believe everything that we are told or read about in the published literature, that we should question, question, question the current ‘wisdom’ all the time, that there should be space in the field for those with alternative views, and that funding of grants should not depend on whether or not the applicant supports the ‘party’ line. How often in published muscle papers do people interpret their observations in terms of the current wisdom, the swinging crossbridge theory, without at the same time wondering how else the same data could be explained and then providing reasons to choose between the alternatives. We really need to do this! I once made this point to a senior muscle researcher who replied “But that’s really hard, John”. My unspoken response to this was: ‘But that is what scientists are supposed to do isn’t it! Is the muscle field somehow different?’
Building on Gerald’s convictions, let me give examples of what I mean. When I came into the muscle field in the late 1960s it was thought that the myosin filaments in vertebrate striated muscles and insect flight muscles were all 2-stranded (Huxley and Brown 1967; Reedy 1968). I looked at the data on which this was based and realised (Squire 1971) that the number of strands could be more, and could be different for different muscle types, which turned out to be the case; vertebrate striated muscle myosin filaments are 3-stranded and insect flight muscle myosin filaments are 4-stranded. Secondly, when David Parry and I did our work on the steric blocking model of thin filament regulation (Parry and Squire 1973), we didn’t just say that the X-ray data on this could be explained by a shift of tropomyosin in the grooves of the actin filament. We also said to ourselves: ‘If it is not a tropomyosin shift then what else could it be?’ We then tested all the alternatives that we could think of and found that the tropomyosin shift was really the only model that would work. It is exactly this kind of alternative thinking which Gerald was always pushing for. I could go on with many more examples, but suffice it to give one more. Recently the interference effects on the meridian of X-ray diffraction patterns from vertebrate striated muscles (e.g. Piazzesi et al. 2002; Reconditi et al. 2004; Huxley et al. 2006a, b and many more) were interpreted almost entirely in terms of movements of the lever arms of the myosin heads. It was not until others looked at alternative interpretations of these observations (Knupp et al. 2009) that it was shown these observed effects were almost entirely due to relative movements of the motor domains of the myosin heads, some attached to actin and some detached, and that sadly these intensities revealed little about the behaviour of the lever arms. It is this kind of testing of alternatives that we should be doing all the time if we don’t want to lead ourselves down blind alleys, and it is what we need to do to take the field forward. I am convinced, like Gerald, that several ideas that we all hold as central to our current concepts on the mechanism of muscle contraction will not stand up to detailed scrutiny. This critical approach was Gerald’s philosophy, and as he himself put it in one particular example: “I think that any graduate student who agrees with their supervisor on all points is probably in the wrong game!” This sums up his very healthy attitude to science and is one that we in the muscle field would be foolish not to follow. The field has been stuck in a rut for long enough.
Caroline: Gerald was a cheerful, gregarious character who enjoyed working with all sorts of people, especially women whose roles as serious scientists he greatly admired. He was always quick to acknowledge the diverse contributions of his many collaborators: he is sole author on only a handful of more than 100 publications. Although his concepts, particularly on crossbridges, failed to become mainstream, he never allowed his desire for personal popularity to attenuate his insistence upon the awkward implications of rigorously performed measurements.
Austin: Gerald’s interest in muscle research lasted to the end of his life: his last e-mail before his death was to Hugh Huxley, who he had known since the 1950s. Though disappointed that his final exposition of his concept of muscle contraction published early last year (Elliott and Worthington 2012) did not elicit more discussion, he remained philosophical. ‘Time will tell.’ He would say.
Various people who knew Gerald well have contributed to parts of this obituary, especially his wife Katalin and son Austin, and some of his collaborators and friends including Professor Belinda Bullard, Professor K.W. Ranatunga, Professor Caroline Pond, Professor Keith Meek and Dr. Justyn Regini. I am grateful to them all, as I am to have known and interacted with Gerald who enriched all our lives.
- Elliott GF (1960) Electron microscope and X-ray diffraction studies of invertebrate muscle fibres. PhD Thesis, University of LondonGoogle Scholar
- Elliott GF (2007a) X-rays, twitching muscles and burning anodes. Physiol News 67:6–10Google Scholar
- Elliott GF (2007b) Living history. Physiol News 69:42Google Scholar
- Huxley HE (2008) Memories of early work on muscle contraction and regulation in the 1950's and 1960's. Biochem Biophys Res Commun 369:34–42Google Scholar