Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Oct 24:17:1254154.
doi: 10.3389/fnins.2023.1254154. eCollection 2023.

Predictive analytics identifies key factors driving hyperalgesic priming of muscle sensory neurons

Sridevi Nagaraja  1   2 , Shivendra G Tewari  1   2 , Jaques Reifman  1
Affiliations

Predictive analytics identifies key factors driving hyperalgesic priming of muscle sensory neurons

Sridevi Nagaraja et al. Front Neurosci. .

Abstract

Hyperalgesic priming, a form of neuroplasticity induced by inflammatory mediators, in peripheral nociceptors enhances the magnitude and duration of action potential (AP) firing to future inflammatory events and can potentially lead to pain chronification. The mechanisms underlying the development of hyperalgesic priming are not well understood, limiting the identification of novel therapeutic strategies to combat chronic pain. In this study, we used a computational model to identify key proteins whose modifications caused priming of muscle nociceptors and made them hyperexcitable to a subsequent inflammatory event. First, we extended a previously validated model of mouse muscle nociceptor sensitization to incorporate Epac-mediated interaction between two G protein-coupled receptor signaling pathways commonly activated by inflammatory mediators. Next, we calibrated and validated the model simulations of the nociceptor's AP response to both innocuous and noxious levels of mechanical force after two subsequent inflammatory events using literature data. Then, by performing global sensitivity analyses that simulated thousands of nociceptor-priming scenarios, we identified five ion channels and two molecular processes (from the 18 modeled transmembrane proteins and 29 intracellular signaling components) as potential regulators of the increase in AP firing in response to mechanical forces. Finally, when we simulated specific neuroplastic modifications in Kv1.1 and Nav1.7 alone as well as with simultaneous modifications in Nav1.7, Nav1.8, TRPA1, and Kv7.2, we observed a considerable increase in the fold change in the number of triggered APs in primed nociceptors. These results suggest that altering the expression of Kv1.1 and Nav1.7 might regulate the neuronal hyperexcitability in primed mechanosensitive muscle nociceptors.

Keywords: action potential; computational analysis; hyperalgesic priming; inflammation; ion channels; musculoskeletal pain; nociceptor.

PubMed Disclaimer

Conflict of interest statement

SN and ST were employed by The Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc. The remaining author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

FIGURE 1
FIGURE 1
(A) Implementation of two parallel inflammation-induced G protein-coupled receptor (GPCR) signaling pathways in our nociceptive muscle neuron model. Shown are the four modeled neuronal transmembrane proteins, Nav1.8, Nav1.7, Kv1.1, and TRPA1, whose activation and inactivation kinetics are modified by protein kinases C (PKC) and A (PKA), which are activated by the two prostaglandin E2 (PGE2)-initiated intracellular signaling pathways in the model. The arrows of the ion channels indicate the direction of flow of the ions through the corresponding channel. In the first pathway, phosphorylation of the GPCR activates subunits Gαq, β, and γ of the receptor. The Gαq subunit activates membrane-bound phospholipase C (PLC) and phosphatidylinositol 4,5-bisphosphate (PIP2) to produce diacylglycerol (DAG), which in turn activates PKC. In the second pathway, phosphorylation of the GPCR activates subunits Gαs, β, and γ of the receptor. The Gαs subunit activates membrane-bound adenyl cyclase (AC) that activates cAMP, which in turn activates PKA. We also implemented the cAMP activation of Epac, which increases PKC activation (blue dashed arrow), thus providing feedback between the two inflammation-induced pathways. In addition, we added a mechanosensitive ion channel, TRPV4, previously not present in the model. Finally, we modeled the phosphorylation of Nav1.8, Nav1.7, Kv1.1, and TRPA1 by PKC and PKA, which modified their activation and inactivation kinetics. (B) Schematic showing the simulation inputs and outputs. In each simulation (one performed using the nominal parameter set and another using 50,000 distinct parameter sets generated for the global sensitivity analysis), we applied a step input of 5 or 100 mN mechanical force for 60 s at the simulation time point of 1 h. Next, at the 47- and 95-h simulation time points, we applied a step input of an inflammatory mediator (IM) at 100 nM for 30 min immediately followed by a 60-s step input of either 5 or 100 mN mechanical force. For each simulation that ran successfully, we used the time course of the membrane potential output to calculate the number of action potentials (APs) fired after the application of the mechanical force input once before (AP1) and twice after (AP2 and AP3) exposure to IM. Finally, we calculated the fold change in the number of APs fired after the first (FC1) and second (FC2) exposure to the IM, by dividing AP2 and AP3, respectively, by AP1. We classified the simulations where FC1 and FC2 were > 1 as primed neurons and those where FC1 and FC2 were ≤ 1 as non-primed neurons.
FIGURE 2
FIGURE 2
Model calibration. We calibrated the model’s predictions of the increase in action potential (AP) firing magnitude by fitting them to experimental data from rat spinal cord neurons in response to mechanical forces applied after the addition of an inflammatory mediator. Open bars (N = 6) show mean experimental data ± 1 standard error fold change in the magnitude of AP firing in response to innocuous (A) and noxious forces (B) after two subsequent exposures to inflammatory mediators (Bar et al., 2004). Solid bars show the results of model fitting to the experimental data. We used innocuous and noxious forces of 20 and 100 mN, respectively, to stimulate the neuron in the model. (C) Simulated time course trajectory of the membrane potential output in response to a 20 mN force applied before (at the 1-h time point) and after two subsequent inflammatory events (at the 47- and 95-h time points). The inset shows the shape and number of APs generated in each response.
FIGURE 3
FIGURE 3
Model validation. We validated the model by comparing the simulations of inflammation-induced action potential (AP) firing increase and mechanical threshold reduction with the corresponding experimental data. (A) Simulated time course trajectory of the membrane potential output in response to a 100 mN force applied before (at the 1-h time point) and after two subsequent inflammatory events (i.e., at the 47- and 95-h time points). The inset shows the number of APs generated and their magnitude. (B) Mean experimental data ± 1 standard error (SE) fold change in the number of APs fired in response to a noxious force after exposure to two subsequent inflammatory events in rat gastrocnemius muscle neurons (open bars, N = 10) (Hendrich et al., 2013). (C) Mean experimental data of the percentage reduction ± 1 SE in the mechanical threshold induced by two subsequent exposures to 100 nM PGE2 (open bars, N = 6). Solid bars in (B,C) show the corresponding model predictions. (D) Model predictions of the AP firing fold change in response to five different force stimulations after two inflammatory events. The fold changes after the first inflammatory event are referred to as FC1 and those after the second inflammatory event as FC2.
FIGURE 4
FIGURE 4
Baseline inflammation-induced sensitization and hyperalgesic priming of muscle nociceptors. Shown are the number of action potentials (APs) fired in response to a mechanical force of (A) 5 mN and (C) 100 mN applied before (pink circles) and after the first (green circles) and second (blue circles) exposures to an inflammatory mediator. The insets show the mean (horizontal black lines) number of APs fired. (B,D) show the corresponding mean ± 1 standard error (SE) of the AP fold-change values after the first (FC1) and second (FC2) inflammatory events in response to 5 and 100 mN mechanical forces, respectively.
FIGURE 5
FIGURE 5
Partial rank correlation coefficient (PRCC) analysis identified key proteins and processes for action potential (AP) regulation. The bars show the PRCCs of the 141 model parameters with fold changes in the total number of APs generated after the first (FC1) and second (FC2) inflammatory events in response to a mechanical force of 5 mN (A,B) and 100 mN (C,D) in the primed neuron simulations. The PRCCs above their respective thresholds (dotted horizontal lines) that were statistically significant (i.e., p < 0.01; p-values depict the probability of seeing the observed correlation if no correlation exists) are indicated by solid black bars, and the labels of the bars show the ion channels/ion pumps or the rates of intracellular processes that these parameters describe in the model. We used the Spearman’s rank correlation method to compute the PRCC values, and obtained the p-values of the correlation using a Student’s t-test. Because the data were not normally distributed, we used a large-sample approximation while performing this test. The analyses are based on 2,438 and 1,605 simulations classified as primed neurons for applied mechanical forces of 5 and 100 mN, respectively.
FIGURE 6
FIGURE 6
Parameter distribution analysis identified key proteins and processes for action potential (AP) regulation. Shown are the distributions of parameter values across the simulations of primed (solid lines) and non-primed (dashed lines) neuron groups representing (A) Nav1.7 activation and (B) Kv1.1 activation with a 5 mN mechanical force used as input, and (C) TRPA1 activation and (D) Nav1.7 inactivation with a 100 mN mechanical force used as input. The x-axis indicates the normalized parameter values, and the y-axis represents the percentage of simulations in each neuron group in which the parameter values fell within a particular range (described in the “Materials and methods” section).
FIGURE 7
FIGURE 7
In silico analysis identified the relative contributions of specific neuroplastic changes in model-identified key ion channels to action potential (AP) generation in primed and non-primed neurons. We simulated an individual twofold expression increase in Nav1.7, Nav1.8, and TRPA1; an individual twofold expression decrease in Kv7. 2 and Kv1.1; as well as simultaneous modification of Kv7.2, Nav1.7, Nav1.8, and TRPA1 (gray bars) after the first exposure to an inflammatory event in primed and non-primed neuron sets identified during global sensitivity analysis. We compared the means and one standard error (SE) of the AP fold changes (FC1 and FC2) in response to 5 and 100 mN forces, after two subsequent inflammatory mediator exposures derived from simulations implementing the six modifications involving the individual and combined increase or decrease in expression of proteins with corresponding simulations with no modifications (i.e., the solid bars) in the (A,B) primed and (C,D) non-primed neuron groups. Solid bars in all panels indicate the means and one SE of the magnitude of AP fold change in simulations with no modification. Because the data from all the groups were not normal (which we established by performing a Kolmogorov–Smirnov test), we used a Wilcoxon rank sum test, which is typically used to compare the means of two independent samples when we cannot assume normality, to determine if any of the protein modification significantly changed AP firing. An asterisk (*) indicates that the mean AP fold change due to a particular modification was significantly different from the simulations with no modification, with p ≤ 0.01.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Aley K. O., Levine J. D. (1999). Role of protein kinase A in the maintenance of inflammatory pain. J. Neurosci. 19 2181–2186. 10.1523/JNEUROSCI.19-06-02181.1999 - DOI - PMC - PubMed
    1. Amir R., Devor M. (2003). Electrical excitability of the soma of sensory neurons is required for spike invasion of the soma, but not for through-conduction. Biophys. J. 84 2181–2191. 10.1016/S0006-3495(03)75024-3 - DOI - PMC - PubMed
    1. Araldi D., Ferrari L. F., Levine J. D. (2016a). Adenosine-A1 receptor agonist induced hyperalgesic priming type II. Pain 157 698–709. 10.1097/j.pain.0000000000000421 - DOI - PMC - PubMed
    1. Araldi D., Ferrari L. F., Levine J. D. (2016b). Gi-protein-coupled 5-HT1b/d receptor agonist sumatriptan induces type I hyperalgesic priming. Pain 157 1773–1782. 10.1097/j.pain.0000000000000581 - DOI - PMC - PubMed
    1. Baker M. D. (2005). Protein kinase C mediates up-regulation of tetrodotoxin-resistant, persistent Na+ current in rat and mouse sensory neurones. J. Physiol. 567(Pt. 3), 851–867. 10.1113/jphysiol.2005.089771 - DOI - PMC - PubMed