The nucleosome hypothesis represents one of the great paradigm shifts in our understanding of eukaryotic gene expression. Formulated by Roger Kornberg in 1974, its key concept is that eukaryotic DNA is tightly packaged around a core of structural proteins — histones — to generate a nucleosome array that is fundamental for controlling gene expression.

Throughout the 1960s, it was largely accepted that chromatin was a linear strand of DNA coated with a simple repeated arrangement of five histones that packaged it into so-called '100-Å fibres'. This view held that the DNA is encased within the histone protein, resulting in non-specific repression of transcription.

However, cracks in this view were beginning to surface. There was accumulating evidence to indicate that chromatin structures might not be so evenly distributed as originally thought. Evidence from X-ray diffraction, electron microscopy, and, in particular, a 1973 study by Hewish and Burgoyne in which chromatin was incubated with a nuclear deoxyribonuclease, led to the idea that histones might instead have a more periodic particulate structure. Another significant turning point was marked by the electron microscopy analyses of Olins and Olins, and Stanchfield and colleagues. They isolated a range of interphase nuclei and identified linear arrays of chromatin ∼70 Å in diameter. In some planes, the structures appeared connected by 15-Å strands and were said to resemble 'beads on a string'. At the time, however, these studies were met with considerable resistance — many suspected that the observed structures were merely artefacts of sample preparation.

It took the landmark analysis of Kornberg and Thomas to appreciate the significance of these observations, combine them with a detailed analysis of histone–histone interactions and arrive at the concept we now recognize as the nucleosome hypothesis. Using biochemistry, they found that the histone core consists of a tetramer of H3 and H4 subunits, and two copies each of H2A and H2B. The next crucial advance was the finding that an H3–H4 tetramer, two H2A–H2B oligomers and DNA were sufficient to recapitulate the known X-ray structure of chromatin. Kornberg noted that chromatin contains one copy of each subunit per 100 base pairs of DNA. With the observed ratios of one tetramer and two oligomers from the biochemical analysis, he reasoned that the repeating unit of chromatin must therefore contain 200 base pairs of DNA. This hypothesis also fitted neatly with the Burgoyne study. Kornberg initially reported his findings at a Ciba Foundation Symposium in London, in April 1974. By the time he presented his hypothesis at a Gordon Research Conference in August of the same year, there were few who disagreed.

Subsequent studies by Finch et al. in 1977, and by Luger et al. and Davey et al. in 1997, reported the X-ray crystal structure of the nucleosome complex. These studies were no mean feat, as the large variation in post-translational modifications between species makes crystallization particularly difficult. What is remarkable, however, is that almost all the predictions of the original theory have been borne out by these structural studies. Indeed, to this day, the nucleosome hypothesis remains a cornerstone of modern biology.