Identification of hypoglycemia-specific neural signals by decoding murine vagus nerve activity

Abstract

Background: Glucose is a crucial energy source. In humans, it is the primary sugar for high energy demanding cells in brain, muscle and peripheral neurons. Deviations of blood glucose levels from normal levels for an extended period of time is dangerous or even fatal, so regulation of blood glucose levels is a biological imperative. The vagus nerve, comprised of sensory and motor fibres, provides a major anatomical substrate for regulating metabolism. While prior studies have implicated the vagus nerve in the neurometabolic interface, its specific role in either the afferent or efferent arc of this reflex remains elusive.

Methods: Here we use recently developed methods to isolate and decode specific neural signals acquired from the surface of the vagus nerve in BALB/c wild type mice to identify those that respond robustly to hypoglycemia. We also attempted to decode neural signals related to hyperglycemia. In addition to wild type mice, we analyzed the responses to acute hypo- and hyperglycemia in transient receptor potential cation channel subfamily V member 1 (TRPV1) cell depleted mice. The decoding algorithm uses neural signals as input and reconstructs blood glucose levels.

Results: Our algorithm was able to reconstruct the blood glucose levels with high accuracy (median error 18.6 mg/dl). Hyperglycemia did not induce robust vagus nerve responses, and deletion of TRPV1 nociceptors attenuated the hypoglycemia-dependent vagus nerve signals.

Conclusion: These results provide insight to the sensory vagal signaling that encodes hypoglycemic states and suggest a method to measure blood glucose levels by decoding nerve signals.

Trial registration: Not applicable.

Keywords: Bioelectronic medicine; Decoding; Glucose; Hypoglycemia; Insulin; Vagus nerve.

Fig. 1

Recording Neural Events in Acute Hyper and Hypoglycemia. a Mice are anesthetized with isoflurane, the nerve is fitted with a bipolar, cuff electrode from Cortec and activity is recorded using a Plexon acquisition system. After 30 min’ baseline recording, mice received 6 mg/kg insulin, 2 g/kg glucose, or saline, and were recorded for an additional 30 min. b Blood glucose was measured via tail nick with glucometer throughout the recording. Blood glucose dropped significantly in mice that received insulin within 10 min. Blood glucose rose significantly in mice that received glucose

Fig. 2

Indicative neural responses to different injections. Each colored trace represents the response rate over time of a different CAP – solid lines correspond to lower firing rate CAPs (maximum of 10 CAPs/sec) and dotted lines corresponds to high firing rate CAPs (maximum of 30–45 CAPs/sec). The injection is indicated by a vertical black line occurring at 30 min. The thresholds encapsulating the baseline statistics of a particular CAP are indicated by horizontal lines of the same color. An example of a insulin VN response curves that include a CAP that increases its firing rate, along with their respective CAP waveforms and b an example of neural responses to the saline injections control, where there are no discernible responses to the injection, along with the respective CAP waveforms.

Fig. 3

Decoding Algorithm and illustrative example. Schematic Diagram of the decoder used to regress the CAP event rates to the blood glucose concentration. The means of 20 s, non-overlapping sliding windows over the event rates for the previous 4 min were used as features to the lasso regression model with a regularization factor of 1. The inputs were regressed against a non-time-shifted version of the smoothing spline fit to the measurements of the blood glucose concentration. Leave-one-out cross-validation with 12 folds was used to train 12 separate regression models, and the out-of-sample data was evaluated for each model to reconstruct a piecewise response

Fig. 4

Indicative wild type regression model performance from an insulin injection. The black line indicates the smoothed blood glucose measurements, with blue bars indicating interpolation error. The red line indicates the estimated blood glucose levels as an output of the regression model. The regression closely traces the smoothed blood glucose measurements, and the average error over the N = 6 subjects has an interquartile range between 14 and 22 mg/dl

Fig. 5

Indicative neural responses to different injections from TRPV1+ Cell-depleted mice. Each colored trace represents the response rate against time of a different CAP – solid lines correspond to lower firing rate CAPs (maximum of 10 CAPs/sec) and dotted lines corresponds to high firing rate CAPs (maximum of 30–50 CAPs/sec). The injection is indicated by a vertical black line occurring at 30 min. The thresholds encapsulating the baseline statistics of a particular CAP are indicated by horizontal lines of the same color. An example of the a insulin VN response curves, along with their respective CAP waveforms, and b an example of the neural response to the saline injection control, where there is no discernible response to the injection, along with the respective CAP waveform. An example of TRPV1+ Cell-depleted mice c regression model performance from an insulin injection. The black line indicates the smoothed blood glucose measurements, with blue bars indicating measurement error. The red line indicates the estimated blood glucose levels as an output of the regression model. The regression somewhat traces the smoothed blood glucose measurements, and the average error over the N = 7 subjects has an interquartile range between 29 and 37 mg/dl

Fig. 6

An example of neural responses to a glucose injection. Each colored trace represents the response rate against time of a different CAP – solid lines correspond to lower firing rate CAPs (maximum of 15 CAPs/sec) and dotted lines corresponds to high firing rate CAPs (maximum of 60 CAPs/sec). The injection is indicated by a vertical black line occurring at 30 min. The thresholds encapsulating the baseline statistics of a particular CAP are indicated by horizontal lines of the same color. The a glucose VN response curves include a CAP that decreases its firing rate are shown here along with their respective waveforms. An example of wild type b regression model performance from a glucose injection. The black line indicates the smoothed blood glucose measurements, with blue bars indicating interpolation error. The red line indicates the estimated blood glucose levels as an output of the regression model. The regression revolves around the smoothed blood glucose measurements, and the average error over the N = 6 subjects has an interquartile range between 25 and 66 mg/dl