The homeostatic framework has dominated our understanding of cellular physiology. We question whether homeostasis alone adequately explains microbial responses to environmental stimuli, and explore the capacity of intracellular networks for predictive behavior in a fashion similar to metazoan nervous systems. We show that in silico biochemical networks, evolving randomly under precisely defined complex habitats, capture the dynamical, multidimensional structure of diverse environments by forming internal representations that allow prediction of environmental change. We provide evidence for such anticipatory behavior by revealing striking correlations of Escherichia coli transcriptional responses to temperature and oxygen perturbations-precisely mirroring the covariation of these parameters upon transitions between the outside world and the mammalian gastrointestinal tract. We further show that these internal correlations reflect a true associative learning paradigm, because they show rapid decoupling upon exposure to novel environments.

No abstract

Chestnut blight, caused by the fungus Cryphonectria parasitica, has been effectively controlled with doublestranded RNA hypoviruses in Europe for over 40 years. The marked reduction in the virulence of C. parasitica by hypoviruses is a phenomenon known as hypovirulence. This virus-fungus pathosystem has become a model system for the study of biological control of fungi with viruses. We studied variation in tolerance to hypoviruses in fungal hosts and variation in virulence among virus isolates from a local population in Italy. Tolerance is defined as the relative fitness of a fungal individual when infected with hypoviruses (compared to being uninfected); virulence is defined for each hypovirus as the reduction in fitness of fungal hosts relative to virus-free hosts. Six hypovirus-infected isolates of C. parasitica were sampled from the population, and each hypovirus was transferred into six hypovirus-free recipient isolates. The resulting 36 hypovirus-fungus combinations were used to estimate genetic variation in tolerance to hypoviruses, in hypovirus virulence, and in virus-fungus interactions. Four phenotypes were evaluated for each virus-fungus combination to estimate relative fitness: (i) sporulation, i.e., the number of asexual spores (conidia) produced; (ii) canker area on field-inoculated chestnut trees, (iii) vertical transmission of hypoviruses into conidia, and (iv) conidial germination. Two-way analysis of variance (ANOVA) revealed significant interactions (P < 0.001) between viruses and fungal isolates for sporulation and canker area but not for conidial germination or transmission. One-way ANOVA among hypoviruses (within each fungal isolate) and among fungal isolates (within each hypovirus) revealed significant genetic variation (P < 0.01) in hypovirus virulence and fungal tolerance within several fungal isolates, and hypoviruses, respectively. These interactions and the significant genetic variation in several fitness characters indicate the potential for future evolution of these characters. However, biological control is unlikely to break down due to evolution of tolerance to hypoviruses in the fungus because the magnitudes of tolerance and interactions were relatively small.Despite widespread release over many years, the efficacy of biological control agents in populations of target pests has remained relatively stable (22,25). This stability contrasts markedly to the situation with chemical pesticides, where numerous target pests have evolved resistance to a wide range of pesticides (19). In a recent review, Holt and Hochberg (25) summarized some of the hypotheses that have been proposed to explain this dichotomy. These hypotheses include inadequate relevant genetic variation in target pests, genetic constraints (trade-offs) between resistance (or tolerance) to a biological control agent and fitness, weak selection, temporally varying selection, and coevolutionary dynamics. Few of these hypotheses have been addressed experimentally in biological control systems, although genetic variation in r...

The outcrossing rate in the chestnut blight fungus, Cryphonectria (Endothia) parasitica, was estimated in a natural population using data from six unlinked restriction fragment length polymorphism (RFLP) loci and DNA fingerprinting. This fungus was shown to have a mixed mating system with both self-fertilization and outcrossing in the same population. In a sample of 22 perithecia, all progeny segregated in 1:0, 1:1 or 0:1 ratios at RFLP loci; there was no evidence for multiple male parents in ascospore families from single perithecia. The multilocus estimate of the outcrossing rate was 0.74 for the RFLP data; the proportion of perithecia with progeny segregating for fingerprint fragments was 0.73. Vegetative compatibility was not a sensitive marker for detecting outcrossing as only 11 perithecia (50 per cent) showed segregation with this marker, compared with 16 (73 per cent) for DNA fingerprinting. We are not aware of any other report of a mixed mating system of a fungus in a natural population with an outcrossing rate significantly different from 1.0 or 0.0.