Network motifs: theory and experimental approaches

Abstract

Transcription regulation networks seem to be built of a few regulatory patterns called network motifs. Each network motif can carry out defined information-processing functions. These functions have been experimentally studied in selected systems, mostly Escherichia coli. Negative autoregulation can speed up responses and reduce fluctuations, whereas positive autoregulation slows responses and increases variations. Coherent feedforward loops can show persistence detection, whereas incoherent feedforward loops show pulse generation and response acceleration. Single-input modules can generate temporal programmes of expression. Dense overlapping regulons can act as arrays of gates for combinatorial decision making. Developmental networks display these network motifs, and additional motifs, such as two-point positive-feedforward loops for decision making and memory, and cascades for regulating slow multi-step processes. Network motifs in systems that have been studied experimentally so far seem to be wired together in a 'modular' way that allows us to understand the dynamics of each individual motif, even when it is connected to the rest of the network. Evolution seems to have converged on the same motifs in different systems and different organisms, suggesting that they are selected for again and again on the basis of their biological functions. Other biological networks, such as signalling and neuronal networks, also show network motifs, some of which are similar to the motifs that are found in transcription networks.