What is the Research About?

• This research is focused on using machine learning (ML) and bioelectronics to control cells in real-time. This is achieved through bioelectronic devices that can sense and regulate certain biological processes, especially related to cell voltage (Vmem).
• Bioelectronics can interface electronic devices with biological systems, and in this study, they aim to control cell functions like differentiation and proliferation using bioelectronic devices.

What is Vmem (Cell Membrane Potential)?

• Vmem is the electrical charge difference across a cell’s membrane. It’s important because it affects cell functions like growth, movement, and communication with other cells.
• Changing Vmem can influence how cells behave, like differentiating into other types of cells or growing in certain ways. This makes it crucial for regenerative medicine and synthetic biology.

What Are Bioelectronic Devices?

• Bioelectronic devices connect biological systems with electronic signals to control biological processes. These devices can either sense or act on biological processes.
• One key challenge is translating biological signals (like ions and molecules) into electronic currents and vice versa. Iontronics solves this by controlling ions directly instead of using traditional semiconductors.

Machine Learning and Bioelectronics

• Machine learning (ML) can help bioelectronic devices adapt to changes in biological systems. In this study, ML is used to control bioelectronic devices that adjust cell membrane voltage (Vmem) by managing the pH of the surrounding environment.
• Using an adaptive ML algorithm, the researchers were able to control Vmem in human-induced pluripotent stem cells (hiPSCs) in real-time.

How Do Proton Pumps Control pH?

• Proton pumps are devices that add or remove protons (H⁺ ions) from a solution. These pumps are used to control the pH of the solution surrounding the cells.
• Changes in pH affect Vmem. Increasing protons (acidifying the solution) causes cell depolarization (lower Vmem), while decreasing protons (alkalizing the solution) causes cell hyperpolarization (higher Vmem).

Setting Up the Experiment

• The researchers used a proton pump array integrated with a machine learning controller to adjust the pH in the system and control the Vmem of hiPSCs.
• Each proton pump in the array controls the pH in a specific area. A fluorescent probe (SNARF dye) is used to measure pH changes in real-time.

Machine Learning-Based Control

• The ML algorithm uses real-time feedback from the fluorescent measurements to adjust the proton pump’s voltage. This process helps the system “learn” how to maintain the desired pH and Vmem level.
• The system adjusts the proton pump’s voltage automatically based on the current state and target goal, effectively “closing the loop” between sensing and actuation.

How Does the Algorithm Work?

• The algorithm doesn’t require prior training. It continuously adjusts and learns based on data collected during the experiment.
• It uses a radial basis function (RBF) neural network, which helps make quick adjustments in real-time based on the measured data.

Experiment Results

• The system successfully maintained control of Vmem in hiPSCs for up to 10 hours.
• Short-term Vmem control was achieved by using various waveforms like triangles, sine waves, and square waves to manipulate the proton pump and monitor cell response.
• For long-term control, the researchers alternated periods of pump activation and resting states, allowing for better regulation of Vmem without overstimulating the cells.

What is the Impact of This Research?

• This study is a proof-of-concept for controlling Vmem in non-excitable cells using bioelectronics and machine learning.
• It opens new possibilities for controlling cell functions like proliferation and differentiation, which can have applications in regenerative medicine and synthetic biology.
• By combining bioelectronics with real-time adaptive control, the system can be used for long-term manipulation of cell behavior.

Conclusion

• This research shows that it’s possible to control the membrane potential of non-excitable cells for extended periods using bioelectronic devices and machine learning.
• The results have significant implications for applications in synthetic biology, regenerative medicine, and bioelectronics, where controlling cell functions is crucial.

研究背景

• 本研究旨在使用机器学习（ML）和生物电子学实时控制细胞。这通过生物电子设备实现，能够感知并调节细胞电压（Vmem）等生物过程。
• 生物电子学可以将电子设备与生物系统连接，在这项研究中，他们试图通过生物电子设备控制细胞的分化和增殖等功能。

什么是Vmem（细胞膜电位）？

• Vmem是细胞膜两侧的电荷差。这对细胞的功能至关重要，如生长、运动和与其他细胞的交流。
• Vmem的变化会影响细胞的行为，例如分化成其他类型的细胞或以特定方式生长。这在再生医学和合成生物学中至关重要。

什么是生物电子设备？

• 生物电子设备将生物系统与电子信号连接，控制生物过程。这些设备可以感知或作用于生物过程。
• 其中一个关键挑战是将生物信号（如离子和分子）转化为电子电流，反之亦然。离子电子学通过直接控制离子而非传统的半导体来解决这一问题。

机器学习与生物电子学

• 机器学习（ML）有助于生物电子设备适应生物系统的变化。在这项研究中，机器学习被用来控制通过管理周围环境的pH来调节细胞膜电位（Vmem）的生物电子设备。
• 通过适应性机器学习算法，研究人员成功地实时控制了人诱导的多能干细胞（hiPSCs）中的Vmem。

质子泵如何控制pH？

• 质子泵是可以通过电压调节从溶液中添加或移除质子（H⁺离子）的设备。这些泵用于控制细胞周围溶液的pH。
• pH的变化会影响Vmem。增加质子（使溶液酸性）会导致细胞去极化（Vmem降低），而减少质子（使溶液碱性）则会导致细胞超极化（Vmem升高）。

实验设置

• 研究人员使用一个集成了机器学习控制器的质子泵阵列来调节系统中的pH并控制hiPSCs中的Vmem。
• 每个质子泵控制一个特定区域的pH。使用荧光探针（SNARF染料）实时测量pH变化。

机器学习控制

• 机器学习算法利用荧光测量的实时反馈来调节质子泵的电压。这个过程帮助系统“学习”如何保持所需的pH和Vmem水平。
• 系统根据当前状态和目标目标自动调整质子泵的电压，有效地“闭合”了感知与执行之间的控制回路。

算法如何工作？

• 算法不需要事先训练。它会根据实验过程中的数据持续调整和学习。
• 它使用一种径向基函数（RBF）神经网络，这有助于在实时基础上做出快速调整。

实验结果

• 系统成功地在hiPSCs中维持了长达10小时的Vmem控制。
• 短期Vmem控制是通过使用各种波形（如三角形、正弦波和方波）来操控质子泵并监测细胞响应。
• 对于长期控制，研究人员交替进行质子泵激活和休息状态，这使得Vmem的调节更加有效，而不会过度刺激细胞。

这一研究的影响是什么？

• 这项研究展示了使用生物电子设备和机器学习控制非兴奋性细胞膜电位的概念验证。
• 它为控制细胞功能，如增殖和分化，开辟了新的可能性，尤其在再生医学和合成生物学领域。
• 通过结合生物电子学与实时适应性控制，系统可以用于长时间操作细胞行为。

结论

• 这项研究表明，使用生物电子设备和机器学习可以在长时间内控制非兴奋性细胞的膜电位。
• 这些结果对合成生物学、再生医学和生物电子学等领域具有重要影响，其中控制细胞功能至关重要。