Limits...
Determining host metabolic limitations on viral replication via integrated modeling and experimental perturbation.

Birch EW, Ruggero NA, Covert MW - PLoS Comput. Biol. (2012)

Bottom Line: The level of detail of our computational predictions facilitates exploration of the dynamic changes in host metabolic fluxes that result from viral resource consumption, as well as analysis of the limiting processes dictating maximum viral progeny production.For example, although it is commonly assumed that viral infection dynamics are predominantly limited by the amount of protein synthesis machinery in the host, our results suggest that in many cases metabolic limitation is at least as strict.Taken together, these results emphasize the importance of considering viral infections in the context of host metabolism.

View Article: PubMed Central - PubMed

Affiliation: Chemical Engineering, Stanford University, Stanford, CA, USA.

ABSTRACT
Viral replication relies on host metabolic machinery and precursors to produce large numbers of progeny - often very rapidly. A fundamental example is the infection of Escherichia coli by bacteriophage T7. The resource draw imposed by viral replication represents a significant and complex perturbation to the extensive and interconnected network of host metabolic pathways. To better understand this system, we have integrated a set of structured ordinary differential equations quantifying T7 replication and an E. coli flux balance analysis metabolic model. Further, we present here an integrated simulation algorithm enforcing mutual constraint by the models across the entire duration of phage replication. This method enables quantitative dynamic prediction of virion production given only specification of host nutritional environment, and predictions compare favorably to experimental measurements of phage replication in multiple environments. The level of detail of our computational predictions facilitates exploration of the dynamic changes in host metabolic fluxes that result from viral resource consumption, as well as analysis of the limiting processes dictating maximum viral progeny production. For example, although it is commonly assumed that viral infection dynamics are predominantly limited by the amount of protein synthesis machinery in the host, our results suggest that in many cases metabolic limitation is at least as strict. Taken together, these results emphasize the importance of considering viral infections in the context of host metabolism.

Show MeSH

Related in: MedlinePlus

Measured and simulated phage production.Shown per infected host, across time, experiment compared to model predictions for integrated model system, and the T7 ODEs alone, on M9 minimal media with glucose, succinate, or acetate as carbon source (growth rates for T7 ODEs alone are , respectively). Error bars are standard deviation of n = 3. For glucose and succinate media the T7 ODEs time course is not visible because it falls directly beneath the integrated simulation line. The lower right panel quantifies the goodness of fit of the integrated simulation and the T7 ODEs alone to experimental observations using normalized mean squared error.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3475664&req=5

pcbi-1002746-g005: Measured and simulated phage production.Shown per infected host, across time, experiment compared to model predictions for integrated model system, and the T7 ODEs alone, on M9 minimal media with glucose, succinate, or acetate as carbon source (growth rates for T7 ODEs alone are , respectively). Error bars are standard deviation of n = 3. For glucose and succinate media the T7 ODEs time course is not visible because it falls directly beneath the integrated simulation line. The lower right panel quantifies the goodness of fit of the integrated simulation and the T7 ODEs alone to experimental observations using normalized mean squared error.

Mentions: Unlike either individual model, the integrated model is capable of predicting the viral infection dynamics for many different culture conditions. We tested model predictions for three previously unmodeled conditions: glucose, succinate, and acetate minimal media. In each case, we measured the phage production over time (Figure 5, bottom left and top panels). For glucose and succinate media, the models produced dynamics nearly identical to each other as well as similar to the experimental data. However, for infections on acetate minimal media, the integrated model was more accurate than the T7 ODEs alone. The two predicted time courses differ because the integrated model accounts for the slow growth and nutritional limitation of E. coli on acetate (roughly half of the growth rate on succinate). In particular, small decreases in gene product synthesis result in delayed achievement of the thresholds necessary for phage genome replication initiation. Furthermore, all of the simulations, from both the integrated model and the ODEs alone, deviate from the typical one-step-growth phage production trajectory. This is due to the rigid description of host DNA degradation and incorporation into viral genomes in the ODEs, which was originally characterized under a single environmental condition. Quantitative comparison of our observations to the model predictions verified that tryptone simulations were the most indicative of experiment, and that the tryptone and acetate integrated model simulations outperformed those of the ODEs alone (Figure 5, bottom right panel).


Determining host metabolic limitations on viral replication via integrated modeling and experimental perturbation.

Birch EW, Ruggero NA, Covert MW - PLoS Comput. Biol. (2012)

Measured and simulated phage production.Shown per infected host, across time, experiment compared to model predictions for integrated model system, and the T7 ODEs alone, on M9 minimal media with glucose, succinate, or acetate as carbon source (growth rates for T7 ODEs alone are , respectively). Error bars are standard deviation of n = 3. For glucose and succinate media the T7 ODEs time course is not visible because it falls directly beneath the integrated simulation line. The lower right panel quantifies the goodness of fit of the integrated simulation and the T7 ODEs alone to experimental observations using normalized mean squared error.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3475664&req=5

pcbi-1002746-g005: Measured and simulated phage production.Shown per infected host, across time, experiment compared to model predictions for integrated model system, and the T7 ODEs alone, on M9 minimal media with glucose, succinate, or acetate as carbon source (growth rates for T7 ODEs alone are , respectively). Error bars are standard deviation of n = 3. For glucose and succinate media the T7 ODEs time course is not visible because it falls directly beneath the integrated simulation line. The lower right panel quantifies the goodness of fit of the integrated simulation and the T7 ODEs alone to experimental observations using normalized mean squared error.
Mentions: Unlike either individual model, the integrated model is capable of predicting the viral infection dynamics for many different culture conditions. We tested model predictions for three previously unmodeled conditions: glucose, succinate, and acetate minimal media. In each case, we measured the phage production over time (Figure 5, bottom left and top panels). For glucose and succinate media, the models produced dynamics nearly identical to each other as well as similar to the experimental data. However, for infections on acetate minimal media, the integrated model was more accurate than the T7 ODEs alone. The two predicted time courses differ because the integrated model accounts for the slow growth and nutritional limitation of E. coli on acetate (roughly half of the growth rate on succinate). In particular, small decreases in gene product synthesis result in delayed achievement of the thresholds necessary for phage genome replication initiation. Furthermore, all of the simulations, from both the integrated model and the ODEs alone, deviate from the typical one-step-growth phage production trajectory. This is due to the rigid description of host DNA degradation and incorporation into viral genomes in the ODEs, which was originally characterized under a single environmental condition. Quantitative comparison of our observations to the model predictions verified that tryptone simulations were the most indicative of experiment, and that the tryptone and acetate integrated model simulations outperformed those of the ODEs alone (Figure 5, bottom right panel).

Bottom Line: The level of detail of our computational predictions facilitates exploration of the dynamic changes in host metabolic fluxes that result from viral resource consumption, as well as analysis of the limiting processes dictating maximum viral progeny production.For example, although it is commonly assumed that viral infection dynamics are predominantly limited by the amount of protein synthesis machinery in the host, our results suggest that in many cases metabolic limitation is at least as strict.Taken together, these results emphasize the importance of considering viral infections in the context of host metabolism.

View Article: PubMed Central - PubMed

Affiliation: Chemical Engineering, Stanford University, Stanford, CA, USA.

ABSTRACT
Viral replication relies on host metabolic machinery and precursors to produce large numbers of progeny - often very rapidly. A fundamental example is the infection of Escherichia coli by bacteriophage T7. The resource draw imposed by viral replication represents a significant and complex perturbation to the extensive and interconnected network of host metabolic pathways. To better understand this system, we have integrated a set of structured ordinary differential equations quantifying T7 replication and an E. coli flux balance analysis metabolic model. Further, we present here an integrated simulation algorithm enforcing mutual constraint by the models across the entire duration of phage replication. This method enables quantitative dynamic prediction of virion production given only specification of host nutritional environment, and predictions compare favorably to experimental measurements of phage replication in multiple environments. The level of detail of our computational predictions facilitates exploration of the dynamic changes in host metabolic fluxes that result from viral resource consumption, as well as analysis of the limiting processes dictating maximum viral progeny production. For example, although it is commonly assumed that viral infection dynamics are predominantly limited by the amount of protein synthesis machinery in the host, our results suggest that in many cases metabolic limitation is at least as strict. Taken together, these results emphasize the importance of considering viral infections in the context of host metabolism.

Show MeSH
Related in: MedlinePlus