Thursday, 01 September 2016 00:17

Fitness landscapes for microbial pathogens in agricultural systems

Published in Microbial Sciences

Written by Lori R. Shapiro

E. tracheiphila infected Cucurbita pepoFigure 1. Cross section of Cucurbita pepo (squash) plant infected with Erwinia tracheiphila. The white ooze is pure E. tracheiphila culture blocking xylem vessels. Photo by: Scott Chimileski.

How is it that we are able to devote so little of our personal time and energy to producing or acquiring the healthy, safe food that we consume multiple times every day? A large part of the reason we seldom worry about agricultural output is that most of us benefit enormously from modernized, industrial food systems that provide consumers a secure, year-round supply of plentiful food choices at historically low prices. This was made possible by a series of technological advances in plant genetics, farm mechanization, synthetic fertilization, and irrigation that increase crop yield and are together referred to as the ‘Green Revolution’. While the Green Revolution addressed urgent needs to increase agricultural output and efficiency, it has created unforeseen problems.

One unintended side effect has been homogenization of our physical agricultural landscapes, and a reduction in the genetic diversity of domesticated crop plants (and animals) grown in these landscapes. Because of high genetic and physiological homogeneity that characterize modern industrial agro-ecosystems, microbial pathogens and insect pests have the evolutionary potential to quickly invade these populations. In response to this vulnerability of agricultural landscapes, a key development of the Green Revolution was the synthesis and deployment of chemical pesticides to suppress pest insect populations (both herbivores and plant pathogen disease vectors) in agro-ecosystems. However, this solution is imperfect. Several of the insect pest species that pose the greatest threat to crop yields have evolved resistance to multiple classes of chemical pesticides. These compounds also have adverse effects on non-target beneficial arthropods, such as pollinators and natural enemies. Further, these compounds do not protect against all biotic stress agents, and there are notably few tools to protect against plant pathogen infection other than pesticide applications to kill insect vectors.

One pathosystem that offers insight into rapid evolution of microbes in agricultural systems is the phytopathogenic bacterium Erwinia tracheiphila, its plant hosts, and herbivorous insect vectors. E. tracheiphila is a multi-host pathogen that is only known to infect two genera of Cucurbitaceae host plants: Cucurbita spp. (squash, pumpkin, zucchini, gourds), which are native to the Americas, and Cucumis spp. (cucumber, muskmelon) that are native to Eurasia but are now widely grown in temperate Eastern North America. Together, these are among the highest acreage vegetable crops grown worldwide. Yet, E. tracheiphila is only known from cultivated cucurbit agro-ecosystems in Northeastern and Midwestern North America. Unlike most bacterial phytopathogens, E. tracheiphila has no known environmental reservoirs. It is transmitted from infected to healthy plants by the striped cucumber beetle, Acalymma vittatum, which is a highly specialized cucurbit herbivore and one of the few insect species that can detoxify cucurbit anti-herbivore defensive compounds.

Acalymma vittatumFigure 2. The striped cucumber beetle vector of E. tracheiphilaAcalymma vittatum. Photo by: Nick Sloff.

A recent microbial sequencing project has begun to investigate the evolutionary and ecological dynamics of Erwinia tracheiphila. Surprisingly, the reference genome sequence of E. tracheiphila shows the canonical characteristics of a recent restriction to replication only in eukaryotic hosts. First, approximately 20% of the genes are pseudogenes. Pseudogenes, which are genes inactivated by frameshifts, premature stop codons, or deletions, are a very small percentage of most bacterial genomes. The presence of such a high number of pseudogenes in E. tracheiphila suggests that the functional versions of these pseudogenes may have contributed to the progenitor microbe's fitness in a previous niche, but do not contribute to fitness in the current niche. This ecological niche change was likely relatively recent since these pseudogenes are still present in the genome, indicating not enough evolutionary time has passed for negative selection to remove these non-functional genes. Approximately 20% of the genes were acquired through horizontal gene transfer (HGT). HGT is a mechanism by which microbes can rapidly gain novel functionality, and can facilitate emergence into a new niche. A number of horizontally acquired genes in E. tracheiphila are important for host interactions in other microbes, and are putatively important for virulence in E. tracheiphila as well. Finally, approximately 20% of the genome is mobile DNA (phage and insertion sequences), which can structurally reshape microbial genomes. E. tracheiphila does not have a CRISPR-Cas anti-mobile DNA defense system, which is common in other Erwinia spp. This lack of anti-mobile DNA defenses may be related to why the E. tracheiphila genome contains more than 20 prophage regions, and has been invaded by insertion sequences from more than two dozen families have invaded and proliferated. Bacteriophage often carry genes from distantly related microbes, and phage infection is a mechanism by which HGT may have occurred. Together, the structural changes in the E. tracheiphila reference genome are among the most dramatic yet documented in a microbial pathogen.

Often, a microbe that experiences such overwhelming structural genomic changes will quickly die – except, perhaps, if its physical environment is also rapidly changing, and by chance the novel variant happens to be evolutionarily fit in a new setting. Despite genomic characteristics indicating recent dramatic structural changes, six E. tracheiphila strains collected from Cucurbita pepo plants in the Northeastern and Midwestern US show high levels of genetic homogeneity, suggesting a recent population bottleneck and/or purifying selection acting on the population. Together, evidence of dramatic structural genomic changes combined with high genetic homogeneity in geographically distant populations suggest that E. tracheiphila has likely recently emerged into homogenized agricultural systems. These features make this pathosystem a model that can be used to investigate the ecological context of pathogen emergence, and the genomic signatures of rapid adaptation and host specialization on microbial genomes.

Cucurbita pepo ssp texana plantFigure 3. Characteristic wilting symptoms in a wild gourd plant infected with E. tracheiphila in the field. Photo by: Lori Shapiro.

To meet the nutritional needs of an exponentially increasing human population, more than 50% of the Earth’s surface has already been modified for agriculture, pastureland, or urbanization. In the United States, closer to 80% of the landscape has been converted. These ecological changes in landscape complexity affect the probability of pathogen emergence, and likely select for increased virulence of pathogens that do emerge. The first Green Revolution provided key innovations to increase agricultural output of staple crops. The next urgent challenge to food security is to protect productivity in agricultural systems from rapidly changing climatic and rainfall patterns, and to protect these ecologically simplified landscapes against invasion by destructive insect pests and phytopathogens. The solutions to these problems will be complex and multifaceted. One fundamental aspect of protecting future food security is understanding the ecological causes of phytopathogen and insect pest emergence in agro-ecosystems, and the evolutionary dynamics of these pathogens and pests when they do emerge. This knowledge will be crucial to proactively creating agricultural systems with durable resistance to pathogen invasion, and to managing the spread of novel pests and pathogens when they do emerge.


Lori Shapiro

Lori Shapiro is a post doctoral researcher co-advised in Roberto Kolter's laboratory at Harvard Medical School and Naomi Pierce's laboratory at Harvard University. Her research focuses on understanding the ecological context and molecular mechanisms of plant pathogen transmission, and the evolutionary signatures of rapid adaptation on microbial genomes.
Last modified on Tuesday, 14 March 2017 14:31