Limits...
Spatio-temporal Dynamics and Mechanisms of Stress Granule Assembly.

Ohshima D, Arimoto-Matsuzaki K, Tomida T, Takekawa M, Ichikawa K - PLoS Comput. Biol. (2015)

Bottom Line: SGs were assembled as a result of applying arsenite to HeLa cells.This same distribution was also found in our experimental data suggesting the existence of multiple fusion steps in the SG assembly.Furthermore, we found that the initial steps in the SG assembly process and microtubules were critical to the dynamics.

View Article: PubMed Central - PubMed

Affiliation: Division of Mathematical Oncology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan.

ABSTRACT
Stress granules (SGs) are non-membranous cytoplasmic aggregates of mRNAs and related proteins, assembled in response to environmental stresses such as heat shock, hypoxia, endoplasmic reticulum (ER) stress, chemicals (e.g. arsenite), and viral infections. SGs are hypothesized as a loci of mRNA triage and/or maintenance of proper translation capacity ratio to the pool of mRNAs. In brain ischemia, hippocampal CA3 neurons, which are resilient to ischemia, assemble SGs. In contrast, CA1 neurons, which are vulnerable to ischemia, do not assemble SGs. These results suggest a critical role SG plays in regards to cell fate decisions. Thus SG assembly along with its dynamics should determine the cell fate. However, the process that exactly determines the SG assembly dynamics is largely unknown. In this paper, analyses of experimental data and computer simulations were used to approach this problem. SGs were assembled as a result of applying arsenite to HeLa cells. The number of SGs increased after a short latent period, reached a maximum, then decreased during the application of arsenite. At the same time, the size of SGs grew larger and became localized at the perinuclear region. A minimal mathematical model was constructed, and stochastic simulations were run to test the modeling. Since SGs are discrete entities as there are only several tens of them in a cell, commonly used deterministic simulations could not be employed. The stochastic simulations replicated observed dynamics of SG assembly. In addition, these stochastic simulations predicted a gamma distribution relative to the size of SGs. This same distribution was also found in our experimental data suggesting the existence of multiple fusion steps in the SG assembly. Furthermore, we found that the initial steps in the SG assembly process and microtubules were critical to the dynamics. Thus our experiments and stochastic simulations presented a possible mechanism regulating SG assembly.

No MeSH data available.


Related in: MedlinePlus

Gamma distribution of SG size both in SS and experiments, and the sensitivity of the dynamics of SG assembly to parameters.(A) The distribution of SG size at 55 min resembled gamma distribution both in SSs (top and middle panels) and experiments (bottom panel). Continuous lines are curves fitted by gamma distribution. The middle panel was drawn by eliminating SGs that contained smaller number of TIA-1 than 139 in the top panel mimicking limited detection of small SGs in our fluorescence microscopy. (B) Sensitivity of the dynamics of SG assembly. The numbers of SG at the peak (black bars) and at 60 min (gray bars) were sensitive to k1-1, k1-2, kb2, and kb3. They were virtually insensitive or only weakly sensitive to other parameters (top panel). The latency and the time to peak were sensitive to k1-1, k1-2, k2-2, kb2, kb3, and kb4. But they were almost insensitive or only weakly sensitive to other parameters (bottom panel). Rate constant affecting the SG number and time courses are shown in red and blue rectangles, respectively in the model of SG assembly (middle panel).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4482703&req=5

pcbi.1004326.g007: Gamma distribution of SG size both in SS and experiments, and the sensitivity of the dynamics of SG assembly to parameters.(A) The distribution of SG size at 55 min resembled gamma distribution both in SSs (top and middle panels) and experiments (bottom panel). Continuous lines are curves fitted by gamma distribution. The middle panel was drawn by eliminating SGs that contained smaller number of TIA-1 than 139 in the top panel mimicking limited detection of small SGs in our fluorescence microscopy. (B) Sensitivity of the dynamics of SG assembly. The numbers of SG at the peak (black bars) and at 60 min (gray bars) were sensitive to k1-1, k1-2, kb2, and kb3. They were virtually insensitive or only weakly sensitive to other parameters (top panel). The latency and the time to peak were sensitive to k1-1, k1-2, k2-2, kb2, kb3, and kb4. But they were almost insensitive or only weakly sensitive to other parameters (bottom panel). Rate constant affecting the SG number and time courses are shown in red and blue rectangles, respectively in the model of SG assembly (middle panel).

Mentions: We were interested in the SG size distribution, because it might provide us with a better perspective on the dynamics of SG assembly. Bars in the top panel of Fig 7A indicate SG size distribution at 55 min as a result of the simulation, which was fitted by a gamma distribution with a shape and scale parameters of 4.50 and 0.073, respectively. The distribution in experiments at 55 min was also fitted by a gamma distribution with a shape and scale parameters of 2.54 and 0.093, respectively (lower panel of Fig 7A). There were significant differences between the simulation and the experiment. To explore the reason, we hypothesized that small SGs were not detected in our fluorescence measurement, and estimated that the number of GFP-TIA1 molecules in the smallest SG was 139 (Materials and Methods). If we eliminated SGs smaller than this in our simulated data, the distribution was much similar to that by the experiment with a shape and scale parameters of 1.70 and 0.14, respectively (middle panel of Fig 7A). In addition, the shape and scale parameters were almost unchanged by the change in N from 5 to 100 (S9 Fig), suggesting that the SG size distribution was fitted reasonably well with a gamma distribution in our simulation. These results suggest that small SGs were not detected in our fluorescence measurement.


Spatio-temporal Dynamics and Mechanisms of Stress Granule Assembly.

Ohshima D, Arimoto-Matsuzaki K, Tomida T, Takekawa M, Ichikawa K - PLoS Comput. Biol. (2015)

Gamma distribution of SG size both in SS and experiments, and the sensitivity of the dynamics of SG assembly to parameters.(A) The distribution of SG size at 55 min resembled gamma distribution both in SSs (top and middle panels) and experiments (bottom panel). Continuous lines are curves fitted by gamma distribution. The middle panel was drawn by eliminating SGs that contained smaller number of TIA-1 than 139 in the top panel mimicking limited detection of small SGs in our fluorescence microscopy. (B) Sensitivity of the dynamics of SG assembly. The numbers of SG at the peak (black bars) and at 60 min (gray bars) were sensitive to k1-1, k1-2, kb2, and kb3. They were virtually insensitive or only weakly sensitive to other parameters (top panel). The latency and the time to peak were sensitive to k1-1, k1-2, k2-2, kb2, kb3, and kb4. But they were almost insensitive or only weakly sensitive to other parameters (bottom panel). Rate constant affecting the SG number and time courses are shown in red and blue rectangles, respectively in the model of SG assembly (middle panel).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4482703&req=5

pcbi.1004326.g007: Gamma distribution of SG size both in SS and experiments, and the sensitivity of the dynamics of SG assembly to parameters.(A) The distribution of SG size at 55 min resembled gamma distribution both in SSs (top and middle panels) and experiments (bottom panel). Continuous lines are curves fitted by gamma distribution. The middle panel was drawn by eliminating SGs that contained smaller number of TIA-1 than 139 in the top panel mimicking limited detection of small SGs in our fluorescence microscopy. (B) Sensitivity of the dynamics of SG assembly. The numbers of SG at the peak (black bars) and at 60 min (gray bars) were sensitive to k1-1, k1-2, kb2, and kb3. They were virtually insensitive or only weakly sensitive to other parameters (top panel). The latency and the time to peak were sensitive to k1-1, k1-2, k2-2, kb2, kb3, and kb4. But they were almost insensitive or only weakly sensitive to other parameters (bottom panel). Rate constant affecting the SG number and time courses are shown in red and blue rectangles, respectively in the model of SG assembly (middle panel).
Mentions: We were interested in the SG size distribution, because it might provide us with a better perspective on the dynamics of SG assembly. Bars in the top panel of Fig 7A indicate SG size distribution at 55 min as a result of the simulation, which was fitted by a gamma distribution with a shape and scale parameters of 4.50 and 0.073, respectively. The distribution in experiments at 55 min was also fitted by a gamma distribution with a shape and scale parameters of 2.54 and 0.093, respectively (lower panel of Fig 7A). There were significant differences between the simulation and the experiment. To explore the reason, we hypothesized that small SGs were not detected in our fluorescence measurement, and estimated that the number of GFP-TIA1 molecules in the smallest SG was 139 (Materials and Methods). If we eliminated SGs smaller than this in our simulated data, the distribution was much similar to that by the experiment with a shape and scale parameters of 1.70 and 0.14, respectively (middle panel of Fig 7A). In addition, the shape and scale parameters were almost unchanged by the change in N from 5 to 100 (S9 Fig), suggesting that the SG size distribution was fitted reasonably well with a gamma distribution in our simulation. These results suggest that small SGs were not detected in our fluorescence measurement.

Bottom Line: SGs were assembled as a result of applying arsenite to HeLa cells.This same distribution was also found in our experimental data suggesting the existence of multiple fusion steps in the SG assembly.Furthermore, we found that the initial steps in the SG assembly process and microtubules were critical to the dynamics.

View Article: PubMed Central - PubMed

Affiliation: Division of Mathematical Oncology, The Institute of Medical Science, The University of Tokyo, Tokyo, Japan.

ABSTRACT
Stress granules (SGs) are non-membranous cytoplasmic aggregates of mRNAs and related proteins, assembled in response to environmental stresses such as heat shock, hypoxia, endoplasmic reticulum (ER) stress, chemicals (e.g. arsenite), and viral infections. SGs are hypothesized as a loci of mRNA triage and/or maintenance of proper translation capacity ratio to the pool of mRNAs. In brain ischemia, hippocampal CA3 neurons, which are resilient to ischemia, assemble SGs. In contrast, CA1 neurons, which are vulnerable to ischemia, do not assemble SGs. These results suggest a critical role SG plays in regards to cell fate decisions. Thus SG assembly along with its dynamics should determine the cell fate. However, the process that exactly determines the SG assembly dynamics is largely unknown. In this paper, analyses of experimental data and computer simulations were used to approach this problem. SGs were assembled as a result of applying arsenite to HeLa cells. The number of SGs increased after a short latent period, reached a maximum, then decreased during the application of arsenite. At the same time, the size of SGs grew larger and became localized at the perinuclear region. A minimal mathematical model was constructed, and stochastic simulations were run to test the modeling. Since SGs are discrete entities as there are only several tens of them in a cell, commonly used deterministic simulations could not be employed. The stochastic simulations replicated observed dynamics of SG assembly. In addition, these stochastic simulations predicted a gamma distribution relative to the size of SGs. This same distribution was also found in our experimental data suggesting the existence of multiple fusion steps in the SG assembly. Furthermore, we found that the initial steps in the SG assembly process and microtubules were critical to the dynamics. Thus our experiments and stochastic simulations presented a possible mechanism regulating SG assembly.

No MeSH data available.


Related in: MedlinePlus