Thermal Transients Excite Neurons through Universal Intramembrane Mechanoelectrical Effects

Modern advances in neurotechnology rely on effectively harnessing physical tools and insights towards remote neural control, thereby creating major new scientific and therapeutic opportunities. Specifically, rapid temperature pulses were shown to increase membrane capacitance, causing capacitive currents that explain neural excitation, but the underlying biophysics is not well understood. Here, we show that an intramembrane thermal-mechanical effect wherein the phospholipid bilayer undergoes axial narrowing and lateral expansion accurately predicts a potentially universal thermal capacitance increase rate of $\sim 0.3\%/°\mathrm{C}$. This capacitance increase and concurrent changes in the surface charge related fields lead to predictable exciting ionic displacement currents. The new MechanoElectrical Thermal Activation theory’s predictions provide an excellent agreement with multiple experimental results and indirect estimates of latent biophysical quantities. Our results further highlight the role of electro-mechanics in neural excitation; they may also help illuminate subthreshold and novel physical cellular effects, and could potentially lead to advanced new methods for neural control.

Popular Summary

Roughly a decade ago, physicists discovered that brief pulses of short-wave infrared light can excite neurons by inducing rapid temperature fluctuations in the surrounding tissue. The discovery of infrared neural stimulation (INS) motivated work towards potential clinical applications where lasers restore neural function, ranging from optical cochlear implants to optical heart pacers. Five years ago, researchers found that the underlying biophysics is mediated by a rapid thermal jump in the capacitance of the neuron’s membrane capacitor, leading to strong capacitive currents that cause the excitation. However, it was not clear how to tie together temperature rise and capacitance change and draw a clear biophysical connection between them. To fill this gap, we present a predictive biophysical explanation, which reveals that thermal transients alter the membrane’s capacitance through a mechanism that was completely unexpected—by changing the membrane’s dimensions.

As the temperature increases, the membrane undergoes minute structural changes in which it expands laterally while its thickness shrinks. Interestingly, both effects independently contribute to a net increase in capacitance. In fact, we use earlier physical measurements of these temperature-induced changes to predict and very accurately validate our model against capacitance rate-of-change measurements, which, as we show, repeatedly produce a universal value of about 0.3% per degree Celsius.

This new explanation can advance our understanding of thermal neuromodulation techniques and the design of advanced new methods, thus paving the way towards new neurotherapeutic protocols. It also draws interesting new relations between structural and functional biophysics, and it highlights emerging new connections between intramembrane mechanics and neural excitation processes.

Figure 1

Universal membrane capacitance thermal response rate, explained by membrane dimensional changes. (a) Membrane capacitance change rates measured in different studies and preparations (gray bars) versus predictions from the thermal dimensional response of POPC membrane bilayers (naive plate capacitor model), and from a detailed biophysical model accounting for spatial charge distribution. (b) Schematized membrane thinning and area expansion under temperature elevation. This observed process [30, 31, 32, 33] is thought to result from an increase in the phospholipid molecules’ fatty acid chains trans-gauche rotational isomerization, which shortens the tails’ effective length and increases the area per phospholipid molecule. Biophysically, these two phenomena contribute to a predictable increase in both the membrane hydrophobic core’s and total capacitance. Error bars for the direct capacitance measurements are $\pm \mathrm{s}.\mathrm{e}.\mathrm{m}$, and for the model predictions are $\pm$ chi-squares distribution-related uncertainty [31].

Figure 2

Detailed (Gouy-Chapman-Stern-based) membrane biophysical model. (a) Membrane subregions: hydrophobic core, containing the phospholipid molecules’ tails. The outer parts of this region contain intracellular (${\sigma }_i$) and extracellular ${\sigma }_o$ Unhandled Math Content: Z) surface charges. Intracellular and extracellular bulk regions containing mobile charges ($Q$ and $-Q$) and the surface charges’ fixed counterions ($-{\sigma }_i$ and $-{\sigma }_o$). Two Stern layers composed of the phospholipid polar heads and the closest ions to the membrane. $\mathrm{\Delta }{\mathrm{\Phi }}_j$ is the potential drop on each subregion (total, $V_m=\sum \mathrm{\Delta }{\mathrm{\Phi }}_j$). (b) Membrane series circuit where capacitors and sources approximate each subregion ($\mathrm{\Delta }{\mathrm{\Phi }}_j=Q/C_j+V_j$). $V_j\approx -{\sigma }_i/{\mathrm{C}}_j$ for $j=B_i$ or $S_i$ and $V_j\approx {\sigma }_o/{\mathrm{C}}_j$ for $j=B_o$ or $S_o$ is the potential resulting from respective surface charges and both $C_j$ and $V_j$ are temperature dependent. Overall, $Q=C_m(V_m-V_{\mathrm{S}.\mathrm{C}.})$, where $V_{\mathrm{S}.\mathrm{C}.}$ is the total surface charge-related potential. (c) Predicted thermal changes in partial and total membrane capacitance under alternative models: fixed geometry model predicts a net reduction while the varying-dimension model correctly predicts an increase [see also Fig. 1]. (d) Predicted thermal response of intracellular, extracellular (upper panel), and total (bottom panel) surface charge-related potentials.

Figure 3

Empirical results versus model-based predictions for artificial bilayer membrane INS experiments. (a) Predicted versus measured capacitance dynamics for 10 ms pulses. The capacitance change is proportional to the temperature transient (inset: temperature time derivative). (b),(c) Underlying membrane currents and potentials in response to the temperature transient in (a). The overall response is determined by the additive contribution of the membrane capacitance and surface charge-related potential thermal responses. (d)–(g) Predicted versus measured thermal membrane currents (d),(f) and induced charge (e),(g) for PE:PC membranes and for PE:PC:PS membranes. Addition of 14 mM ${\mathrm{MgCl}}_2$ (f),(g) or 1 mM ${\mathrm{GdCl}}_3$ (g) to the “extracellular” compartment shifts towards more depolarizing currents through membrane absorption of excess ions changing the extracellular surface charge density. The simulations use the experimental parameters of Shapiro et al. [16] (Table SI of Supplemental Material [40]). Error bars are $\pm \mathrm{s}.\mathrm{e}.\mathrm{m}$.

Figure 4

Empirical cellular neurostimulation results versus model-based predictions. (a),(b) Predicted versus measured membrane currents, potentials, and induced charge, for 10 ms, 7.3 mJ INS stimulation of Xenopus oocyte in ${\mathrm{H}}_2\mathrm{O}$- and ${\mathrm{D}}_2\mathrm{O}$-based extracellular media; illumination of 5% of the surface area assumed [16]. (c) Predicted somatic membrane potentials in a cortical pyramidal neuron stimulated by threshold and 90% subthreshold simulated 1 ms PAINTS transients. The membrane potential ($V_m$) thermal response is dictated by surface charge- and capacitance-related potentials ($V_{\mathrm{S}.\mathrm{C}\mathrm{.}}$ and $V_{Cm}$, respectively), where $V_{Cm}$ is also affected by the membrane’s charge dynamics. (d) Predicted versus empirical thresholds [Fig. 5(a) in Ref. 18]. The constant $\alpha$ determined from the best fit was used to calculate the surface charge difference in rat cortical neurons (solid blue line, $\mathrm{\Delta }\sigma ={\sigma }_i-{\sigma }_o=-0.107{\mathrm{C}/\mathrm{m}}^2$); predictions for an alternative model ($\mathrm{\Delta }\sigma =-0.05{\mathrm{C}/\mathrm{m}}^2$, dashed red line) are also shown for comparison. ${\sigma }_i\approx -0.12{\mathrm{C}/\mathrm{m}}^2$ and ${\sigma }_o\approx -0.013\mathrm{C}/{\mathrm{m}}^2$ for $V_{Cm}$ and ${\mathrm{V}}_{\mathrm{S}.\mathrm{C}\mathrm{.}}$ the calculation in (c). Error bars are $\pm \mathrm{s}.\mathrm{e}$.