The hypothesis that crystals could have been primitive genetic materials has been put to the test by US scientists.

Nearly three decades ago, Graham Cairns-Smith proposed that the first genetic systems must have been more primitive than the sophisticated chemistries of DNA and RNA. He argued that crystals, particularly clay minerals, have the capacity to act as primitive genes.

His idea was that imperfect crystals can act as genes by transferring information from one crystal to another by means of their imperfections. Screw dislocations, for instance, are often replicated through crystal growth, so the arrangement of these dislocations in a crystal can be considered a store of information. This so-called 'crystals-as-genes' hypothesis has captivated the imagination of scientists for many years, said  Bart Kahr from University of Washington, Seattle, US, but it has never been put to the test, until now.

Kahr and colleagues have designed the first experiment to examine the idea that crystals can act as a source of transferable information, using crystals of potassium hydrogen phthalate.

'Our paper is really an experimental test of an idea whether or not that idea is relevant to life's origins,' said Kahr.

The researchers looked at the distribution of hillocks in the crystal system. 'Mother' crystals with these imperfections were cleaved with a razor and used as seeds to grow 'daughter' crystals from solution. The resulting daughter crystals were imaged using fluorescence microscopy to see whether the spatial distribution of the hillocks had been inherited. 

Kahr found that, as Cairns-Smith suggested, the distribution of the crystal defects can be transferred from one crystal to another. However, a large number of new hillocks, 'mutations', were also observed. For crystals to resemble genes there must be more inheritance than mutation in successive generations.

'While we determined that the dislocations in the crystal system that we studied were not faithful enough to store and transfer information form one generation to the next,' said Kahr, 'we did demonstrate how we can use luminescent molecules to identify the sequence in time and space of all growth active dislocations.'

Cairns-Smith himself is not deterred. 'I would say that the success of [the] idea that RNA preceded DNA has provided inadvertent support for crystal genes,' he said. 'The big thing missing...  is an account of how activated nucleotides might have appeared on the primitive Earth as feedstock for replicating RNA molecules. The kind of organic chemical competence required here could only have been the result of natural selection - based of course on some other genetic material.' 

Kahr, however, is more guarded. 'We wanted to try to bring one aspect of the multifaceted proposal of Cairns-Smith to the realm of repeatable experimental science.'

'I would hope that our experiment would encourage scientists to subject other aspects of the broad crystal-as-genes hypothesis to the scrutiny of experiment,' added Kahr.

While they may not have unravelled the mysteries of life's origins, the scientists have provided new tools for studying crystal growth mechanisms, an area of great interest in many aspects of materials science.

Caroline Moore