Patch Clamp Technique with Integration of Automatic Cell Identification for Ion Channel Activities Recording and Cell Microinjection

膜片鉗技術結合自動細胞識別功能技術以用於進行離子通道訊息採集和細胞微注射

Student thesis: Doctoral Thesis

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date1 Apr 2022

Abstract

Ion channels are specialised proteins in the plasma membrane that provides a passageway for charged ions, including sodium (Na+), potassium (K+), chloride (Cl-), and calcium (Ca+) ions; and ion channel activities are crucial related to the physiology of living creatures. Abnormal ion channel activities can lead to neurodegenerative diseases such as Amyotrophic lateral sclerosis (ALS) and Alzheimer's disease (AD). Thus, recording of the ion channels’ activities from time to time is regarded as valuable data for tracing how such conditions were developed, and hence it may be possible to assist the diagnosis of patients. The patch clamp technique is considered a gold method to record the ion channels activities. This technique can measure the ultra-low current going down to pico ampere across the cell membrane.

In this thesis, the patch clamp technique is used to develop cytoplasmic microinjection with current feedback for small and non-spherical cells. Cell microinjection is commonly used for intracellular delivery of various substances into cells using micropipettes. It plays a crucial role in drug deliveries, cell transfection, and gene therapy using human cells. However, cell microinjection always requires a skilled operator to manipulate a micropipette carefully, so it is inefficient to carry out the cell microinjection, especially for cells that are small in micron scale. It is quite challenging to observe and determine whether the target materials are injected into the cell spontaneously with the naked eye of humans. To reduce the operational failure due to human errors, other research groups have introduced different feedback approaches such as visual and voltage feedback. Using the features of high current sensitivity of the patch clamp technique, we have introduced the patch clamp technique to monitor the microinjection by observing the electrical current passing through the living cell. The electrical current response not only can indicate the materials injected into the cell during the microinjection but also can verify the cell viability after the microinjection. In our studies, an electrical equivalent model for cytoplasmic microinjection has been developed and cell viability verification with SH-SY5Y (human-derived neuroblastoma cell) and HEK-293 (human embryonic kidney cell) have been performed after the microinjection.

After building the electrical equivalent model, we have converted the model into the equivalent electric circuit to calculate the current response with four different injection volumes during the microinjection. During the experiment, a linear relationship is found between the injection volumes and the drop of the current signals. In the equivalent electric circuit, ISeal (current passing the sealing resistance) and IAccess (current passing the access resistance) have been calculated and we have found that ISeal is the significant factor of Im (cell membrane current) for monitoring the injection volume.

Nevertheless, the patch clamp technique has a low throughput due to operational challenges. Traditional operations are complex and labour-intensive, especially the first step to identify suitable cells which can be poked with a glass micropipette without destroying the cell membrane based on microscopic images. The subtle morphological differences between suitable and unsuitable cells are difficult to detect by inexperienced operators or using traditional machine learning approaches. It is desirable to develop an automatic and accurate detection method to select suitable cells for every operation. Hence, a novel signalling model called CELL-YOLO has been developed to detect suitable cells for electrophysiological recording based on the YOLO (You only look once) model [23]. To verify its effectiveness, HEK-293 cells have been chosen as target cells in this study. CELL-YOLO has been trained with three different data sizes of HEK-293 cell images in our self-built datasets and three resulting models named CELL-YOLO104, CELL-YOLO306 and CELL-YOLO612 have been generated. The learning ability of CELL-YOLO has been verified by comparison among these three trained models. To assess CELL-YOLO’s performance, the newly developed YOLO models that are YOLOV3 and YOLOV4 have also been trained with the largest dataset for comparison. To quantify the performance, the accuracy can be measured using mAP50 (mean average precision when intersection over union (IoU)=50), and mAP75 (mean average precision when IoU=75), whilst the processing speed can be measured using FPS (frames per second).

We have found that the mAP50, mAP75, and FPS of CELL-YOLO612 are 35.65, 21.26, and 30, respectively, which are higher than those parameters of YOLOV3 and YOLOV4. The mAP and FPS results indicate the superiority of our proposed model in this specific application in terms of accuracy and processing speed.

Since the patch clamp technique can measure the ultra-low current that goes down to pico ampere, it is a reliable electrophysiological technique to reveal and quantify neuronal activities. The developed method has been applied to electrophysiological signals in both voltage and current clamp to verify if the induced pluripotent stem (iPS) cells can be differentiated into neuronal stem cells (NSCs) on different biocompatible extracellular matrices (ECMs). After plating the stem cells on the glass slip as a control and two different ECMs that are L-NHs (Left-handed nano-helices subtract) and NZs (Nano-zigzags subtract), the whole-cell voltage-clamp has been used to monitor the total current flow across the entire membrane of the stem cell due to the total ion channel activity in response to voltage stimuli. The activities of the Na+ and K+ ion channels of the stem cells have been recorded and those ion channel activities can be ascribed to neuronal signalling. Using the current clamp, the NSCs can exhibit spontaneous firing with current stimuli, and that signal firing is one of the typical neuronal signalling activities. Thus, the experimental result can prove that iPSC can be differentiated into NSC successfully.