A profusion of plant forms surround us in gardens, parks, farmers' fields and the wild. This diversity is all the more remarkable as we increasingly understand that plant growth and development is controlled by surprisingly few and simple rules.

Scientists trying to understand how development is controlled can adopt one of two approaches. The first is to observe closely what happens during development by tracking the metabolism, growth and movement of individual cells, tissues or whole organs. Researchers often compare normal plants with plants that have mutations in one or more of the genes that control a particular aspect of development. In this way they can work back from what happens in normal development (and when development is disrupted) to explain how development is controlled.

Alternatively they can construct, from first principles, a set of rules that allow them to 'mimic' what happens during plant development. These rules can then be used to explore the real process of growth and development and investigate whether their imagined 'rules' exist in reality. It is this latter approach that John Innes Centre scientists (working in collaboration with a group at the University of Calgary, Canada) have adopted in studying inflorescence (flower spike) development. They have found, not only that their imagined rules really exist, but that they have been able to identify some of the plant genes that control these rules.

Making an inflorescence
An inflorescence begins life as a shoot with a growing tip that includes a vegetative meristem; a small group of dividing cells that add cells to the growing shoot. The main shoot can produce side shoots, each with its own meristem, and these side shoots may in turn produce side shoots of their own.

For a plant to produce flowers vegetative meristems must be converted into floral meristems. When this happens the meristem stops making shoot cells, shoot growth stops and the floral meristem develops into a flower bud and eventually a flower. Vegetative meristems are switched from vegetative to floral growth by a 'flowering trigger', which may be an internal or environmental signal.

The rules of the game:
As vegetative meristems develop they approach a critical point where they become sensitive to the 'flowering trigger' that switches them to floral meristem development.

Meristems can be 'mature' or 'immature'. The rate at which mature and immature meristems approach the point where they become sensitive to the flowering trigger can be different, and these differences affect inflorescence shape. If mature meristems become sensitive to the flowering trigger more slowly than immature meristems then a long flowering spike with many flowering side shoots will result - a type of inflorescence called a raceme. The other extreme, where immature meristems are slower to convert to flowering than mature meristems, results in an inflorescence called a cyme. Here the main shoot terminates in a flower, and inflorescence growth continues from a side shoot, this quickly produces a terminal flower and inflorescence growth is continued by another side shoot, and so on.

Other factors do affect inflorescence shape and structure, for example,

* the number of new side shoots produced before the main meristem forms a flower and so stops further inflorescence development
* where new meristems are formed on the main shoot - only on one side, alternately on one side and then the other, in pairs, or spiralling around the shoot.

Conclusion:
Starting from first principles scientists have worked out the simplest set of rules needed to produce the variety of inflorescences we see around us.

By applying these rules to software models of inflorescence growth they can grow many different inflorescence shapes in the computer. Some of these resemble inflorescences we know from ornamental, crop and wild plants, while others look like genetic mutations already found in living plants. This gives them clues as to which rules are controlled by which genes and how to breed for particular inflorescence types.

This approach is being expanded to look at the rules controlling growth of other plant organs and to relate these rules to the genes that control development in those organs. Eventually we will be able to model particular plant shapes and structures in computers, before going into the glasshouse to use that knowledge to combine the necessary genes in new ornamental and crop plants.

The John Innes Centre (JIC), Norwich, UK is an independent, world-leading research centre in plant and microbial sciences. The JIC has over 800 staff and students. JIC carries out high quality fundamental, strategic and applied research to understand how plants and microbes work at the molecular, cellular and genetic levels. The JIC also trains scientists and students, collaborates with many other research laboratories and communicates its science to end-users and the general public. The JIC is grant-aided by the Biotechnology and Biological Sciences Research Council. You can learn more about the John Innes Centre and our exhibits at previous Chelsea Flower Shows on our website at http://www.jic.ac.uk/chelsea/index.htm

The John Innes Centre is very grateful to the BBSRC for sponsoring 'Growing plants in microchips' and to Elonex plc for the loan of computers and screens.