Microtransplantation of Synaptic Membranes to Reactivate Human Synaptic Receptors for Functional Studies

Neurodegenerative disorders affect a large percentage of the population. Although their devastating consequences are well known, the link between the functional alterations of neurotransmitter receptors, which are critical for brain function, and their symptomatology is still poorly understood. Inter-individual variability, chronic nature of the disease, and insidious onset of symptoms are just some of the reasons that have delayed the understanding of the many brain disorders where chemical imbalances are well documented^1,2. Animal models have generated invaluable information and expanded our knowledge about the mechanisms underlying physiology and pathophysiology in evolutionary conserved systems; however, several interspecies differences between rodents and humans preclude the direct extrapolation of receptor function from animal models to the human brain³. Thus, initial efforts to study native human receptors were developed by Ricardo Miledi's lab using surgically removed tissue and frozen samples. These initial experiments used whole membranes that include neuronal synaptic and extra synaptic receptors as well as non-neuronal neurotransmitter receptors, and although they provide important information about diseased states, there is a concern that the mix of receptors complicates the interpretation of data^4,5,6,7. Importantly, synapses are the major target in many neurodegenerative disorders^8,9; therefore, assays to test the functional properties of affected synapses are fundamental to obtain information about disease-relevant changes affecting synaptic communication. Here, a modification of the original method is described: microtransplantation of synaptic membranes (MSM), which focuses on the physiological characterization of enriched synaptic protein preparations and has been successfully applied to study rat and human synaptosomes^{10,11,12,13,14,15}. With this methodology, it is possible to transplant synaptic receptors that were once working in the human brain, embedded in their own native lipids and with their own cohort of associated proteins. Moreover, because MSM data is quantitative, it is possible to use this data to integrate with large proteomic or sequencing datasets¹⁰.

It is important to note that many pharmacological and biophysical analyses of synaptic receptors are done on recombinant proteins^16,17. While this approach provides better insight into the structure-function relationships of receptors, it cannot provide information about complex multimeric receptor complexes found in neurons and their changes in disease. Therefore, a combination of native and recombinant proteins should provide a more comprehensive analysis of synaptic receptors.

There are many methods to prepare synaptosomes^{10,11,12,13,14,15} which can be adjusted for the requirements of a lab. The protocol begins with the assumption that synaptosomal enriched preparations were isolated and are ready to be processed for microtransplantation experiments. In the lab, the Syn-Per method is used following the manufacturer instructions. This is done because of high reproducibility in electrophysiological experiments^10,11. There is also abundant literature explaining how to isolate Xenopus oocytes^18,19, which can also be purchased ready for injection²⁰.

All research is performed in compliance with institutional guidelines and approved by the institutional Animal Care and Use Committee of the University of California Irvine (IACUC-1998-1388) and the University of Texas Medical Branch (IACUC-1803024). Temporal cortex from a non-Alzheimer's disease (AD) brain (female, 74 years old, postmortem interval 2.8 h) and an AD-brain (female, 74 years old, postmortem interval 4.5 h) were provided by the University of California Irvine Alzheimer's disease research center (UCI-ADRC). Informed consent for brain donation was obtained by UCI-ADRC.

NOTE: Unfixed human brain tissue should be treated as a source of blood borne pathogens (BBP). Accordingly, BBP training is needed prior to start experiments. This protocol is performed in a biosafety level 2 (BSL2) laboratory under BSL2 requirements. Guidelines and precautions within the laboratory include: no food or drinks allowed in the laboratory, good laboratory practices must be followed, personal protection equipment (gloves, gown, no open-toed shoes) is required, and the door must be closed at all times.

1. Xenopus oocyte microinjection preparation

1. To make injection needles, pull 3.5 inches borosilicate tubes using a micropipette puller. Once the microinjection needles are pulled, use a microscope and a razor blade to cut off enough of the tip of the needle so that microinjection can be performed (usually between 2-3 mm).
   NOTE: The length of the needle's tip to cut can vary.
2. Prepare 1x Barth's solution by adding 200 mL of Barth's Stock Solution (5x) to 800 mL of distilled water. Fill a 24-well, flat bottom tissue culture plate with 18 °C 1x Barth's solution.
   1. Prepare 5x Barth's stock solution as follows. In 1 L of distilled water, add 25.71 g of NaCl (sodium chloride), 0.372 g of KCl (potassium chloride), 0.301 g of CaCl₂ (calcium dichloride), 0.389 g of Ca(NO₃)₂(4H₂O) (calcium nitrate tetrahydrate), 1.01 g of MgSO₄(7H₂O) (magnesium sulfate heptahydrate), 1.008 g of NaHCO₃ (sodium bicarbonate), and 11.91 g HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid).
3. Isolate Xenopus oocytes using the protocols described in^18,19 and using 2x magnified stereoscope, selecting healthy looking, stage V-VI oocytes. Place about 10 oocytes per well and use one of the wells to house non-injected oocytes to be used as control cells when performing recordings.
4. Retrieve a small Petri dish, 1 cm tall and 6 cm in diameter, and place Nylon mesh inside so that it covers the bottom of the dish. Then, pour 15 mL of 1x Barth's solution to be used for performing microinjections of the oocytes.
5. Place a nanoinjector along the midline of the microscope platform. Turn the magnet to the Off position and slowly move the nanoinjector to the desired position while supporting it carefully. When the nanoinjector is properly positioned, turn the magnet back to the fully On position and make sure that it is secure.
6. For a sample size of 50.6 nL, use the following settings on the side of the nanoinjector: 1 is D (down), 2 is D, 3 is U (up), 4 is D, and 5 is U. Position a piece of self-sealing thermoplastic, or microcentrifuge tube cap, on the microscope stand, aligning it with the trajectory of the nanoinjector.
   NOTE: 50 nL is near the maximum amount of injected material that the cytoplasm can withstand²¹.
7. Fill up an insulin syringe about half its volume with mineral oil (about 0.5 mL/cc). Using this syringe, fill the microinjection needle with mineral oil. Ensure that a visible bead of oil is seen at the tip of the needle.
8. Expose the nanoinjector plunger to a minimum of 1-2 cm. Do this by holding down the EMPTY button until the plunger is visible at desired length. Once the plunger is exposed, slowly and carefully place the mineral oil filled microinjection glass needle onto the nanoinjector plunger, and make sure that it fits well through the black resistance O-ring and stops at the white resistance ring. Ensure not to bend the nanoinjector plunger in the process.
9. Once the microinjection glass needle is secure and in place, press EMPTY to void any potential air bubbles. Gently, with a paper tissue, clean off the excess mineral oil.

2. Loading the sample

1. Prepare synaptosomal enriched preparations (samples) from the human brain region of interest as described in^{10,11,12,13,14,15}, and store them in aliquots of about 5 μL at -80 °C until the moment of injection.
2. Before injection, transfer the aliquot to wet ice and keep it on ice at all times except for sonication and retrieval. The number and types of samples will be determined by the experimental design.
3. Sonicate the selected samples 3x in a bath with floating wet ice to avoid warming up of the sample. Use 5 s cycles each time, waiting 1 min on wet ice in between cycles.
4. Place 1 μL of sample onto the thermoplastic. Make an indention in the surface, if needed, before sample placement to prevent movement of the sample.
5. Use the knobs of the manipulator holding the nanoinjector to position the needle and move it toward the sample to be filled. Once the needle tip is in the sample, press and hold the FILL button until enough sample is taken up. About 0.5 μL is required for 10 oocytes. Using the knobs, move the nanoinjector's needle back to the highest position.

3. Injecting the oocytes

NOTE: Standardized rat cortex synaptosomal membranes are also injected in all experiments into a set of oocytes to measure changes in fusion capacity between different batches of oocytes.

1. Place selected oocytes onto the small Petri dish that has the fitted nylon mesh and Barth's solution (about 10 oocytes). Arrange oocytes so that they are not on top of each other; this will be helpful during the injection process.
2. Using the knobs on the manipulator holding the nanoinjector, move the needle toward the oocytes. Once the needle is submerged in the Barth's solution, use the foot pedal for the nanoinjector to make sure that the microinjection needle is releasing sample. If the sample is being released it will be visible from the microscope view.
3. Once it is confidently established that the sample is being released, move on to microinjecting the oocyte. If sample is not being released, repeat the process of filling the needle.
4. Penetrate the oocyte just underneath the surface with the needle, no deeper, and use the foot pedal to inject the sample. The foot pedal will make a beep sound when the sample is released. Wait about 2-3 s. If the injection was successful, the cell will expand. Once this happens, use the knobs on the manipulator to exit the cell.
   NOTE: Fusion of membranes in the oocyte is a polarized process; therefore, the oocyte should be injected in the animal side of the cell, preferably above the equator, with the angle of the needle toward the animal side to maximize membrane fusion.
5. Move the Petri dish so that the next oocyte is aligned with the needle and repeat the process until all oocytes in the dish have been injected. Repeat this process until the desired number of oocytes has been injected. Make sure that one well of oocytes is left non-injected to be used as control.
6. Once all oocytes have been injected, retract the nanoinjector to its original position and remove the needle.

4. Recording of ion currents using a two electrode voltage clamp

1. To make two electrode voltage clamp (TEVC) electrode needles, pull 15 cm long fire polished borosilicate glass tubes by using the micropipette puller. Once pulling is complete, fill with 3 M KCl using long capillary needles, or alternatively, submerge and boil the needles in 3 M KCl solution for 15 min under a continuous vacuum. The high temperature helps in the filling of the electrodes and removal of air bubbles.
   1. Prepare a 3 M KCl electrode fill solution by using KCl in the crystalline form and deionized water, and following these steps carefully so that reference potential of electrodes is not changed. Dry the KCl carefully in an oven for 2-3 h. Using an analytical balance, carefully weigh 223.68 g of KCl. Transfer the KCl to a 1 L glass bottle. Use this solution to chlorinate silver electrode wires.
      NOTE: For pulling glass electrodes, the pipette cookbook can be downloaded here: https://www.sutter.com/PDFs/cookbook.pdf.
2. Turn on all equipment used for the recording of ion currents: main systems, solution valves, microscope light, valve systems, oocyte clamp, and vacuum system. Turn on the desktop computer, log into WinEDR V3.9.1., and make sure that the program is running and that the setup values are correct as follows: record to disk, set recording duration, and clear out the simulator option.
   1. Provide the following initial setup values for the amplifier: for voltage electrode (Vm) section, turn off the negative capacity compensation (-C); for bath electrodes (Im) section, set the gain selector switch (range from 0.1 to 10) at 10, and the three-position toggle switch which selects the gain multiplier (x0.1, x1.0 and x10), at x1.0. The LED lights will indicate the gain multiplier selection; for clamp section, turn off the clamp mode selector, set the DC gain control to IN, and the gain control full-bandwidth open-loop (range 0 to 2000) to around 1200; for commands, set at 40mV and set the hold controls negative and scale multiplier to be in x2; for current electrode, use the Ve offset to establish a zero reference before impaling the oocyte.
   2. To record, open the WinEDR V3.9.1 software, go to the top menu, and select File > Open New File. Create your own folder and save the file.
   3. Next, in the main menu select Record > Record to Disk. When the two electrodes are inside the ooctyte, turn on the ringer in the VC-8 valve controller, and turn the clamp to Slow in the OC-725C amplifier. Then, go back to the WinEDR software and select Record at the top left of the screen.
   4. In the mark chart at the bottom left, write down your initial membrane potential and press Enter. As the experiment keeps going, you can write down the name of the drugs being used. Once the experiment is over, select Stop at the top left of the screen.
      NOTE: We use WinEDR or WinWCP software to perform TEVC because these are free and well suited for experiments, but any other software that is available can be used.
3. Make up all of the drug solutions needed to perform the experiment (e.g., GABA, Kainate) and confirm that the solution valves are working properly by switching them on and off and observing the displacement of the pinch valves.
4. Fill the tubes by pouring the corresponding solution on the syringe receptacle connected to the tube and label the valve controller to correlate with the solution tubes. Ensure that the flow of solution is steady, there are no bubbles and that the vacuum system is working properly.
5. Set up the microscope and recording chamber area. Place the agar bridges (see below) in the respective circular holes behind the recording chamber (distal from handling area), connecting the wells for the electrodes used as ground reference and the recording chamber. Add Ringer's working solution (1x) to the recording chamber and the ground reference electrode wells.
   1. Agar bridges are U-shaped borosilicate tubes, 2-3 cm long, filled with 3% agar in Ringer's solution. Make U-shaped tubes by cutting 15 cm long tubes, bending them on the open flame, and then filling them with hot 3% agar using an insulin syringe. Store filled agar bridges in Ringer's solution in the fridge until use.
   2. Make 20x Ringer's stock solution as follows: In 1 L of distilled water, add 134.4 g of NaCl, 2.982 g of KCl, 5.29 g of CaCl₂, and 23.83 g of HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid). For 1x Ringer's working solution, in a volumetric flask, add 200 mL of Ringer's stock solution (20x) to 3,800 mL of distilled water.
6. Prepare chlorinated silver electrodes prior to each experiment by electrolysis by placing silver wires in 3 M KCl solution and connecting the silver electrode to the positive terminal of a 9 V battery. After about 3 min a solid brown layer of AgCl is deposited into the electrode. Place the chlorinated silver electrode wires into the electrode housing.
   NOTE: Ground reference and recording chlorinated electrodes are silver/silver chloride (Ag/AgCl) electrodes.
7. Remove the borosilicate needles from the KCl solution. Use a syringe to extract a small amount of KCl solution from the open end of the electrode needle and replace the KCl solution with mineral oil; this will avoid water evaporation and changes in KCl concentration within the needle. Slide the glass needle over the chlorinated silver electrode wire and tighten them in place.
8. Using the right and left manipulators holding the electrodes, guide the electrode needles into the recording chamber that is filled with Ringer's Solution.
9. Check the resistance of each electrode by zeroing out the Vm and Ve offset knobs and then pressing the Electrode Test buttons. The resistance should be between 0.5-3 MΩ (directly read as 5-30 mV in the screen). If the resistance is out of the range, replace using new microelectrodes.
10. Fill the recording chamber with fresh Ringer's solution. Using a glass pipette, place a non-injected or a microtransplanted oocyte in the center of the recording chamber. Ensure that the oocyte is clearly visible under the microscope, with animal side up (see NOTE in step 3.4).
11. Guide the electrode into the Ringer's solution until they touch the oocyte membrane. Gently pierce the oocyte membrane in the animal side (see NOTE in step 3.4) with both electrodes and record the resting membrane potential. Turn on the Ringer's solution flow.
12. Using the oocyte clamp amplifier, change the mode to Voltage Clamp and set the holding voltage of -80 mV. The current on the monitor should be negative, usually between 0 and 0.4 microamperes.
13. Create a new file to save the recording as File > New > Save the Recording on the software. Start recording, press Record > Record to Disk. Add relevant information to the file by using the mark box dialog (test name, drugs used, etc.).
14. Apply agonists (e.g., GABA, glutamate, or kainate) for chemical stimulation by opening the valves and perfusing them into the recording chamber for 15 s. Usually, use a gain of 10x to plot the maximum number of points and reconstruct the response. If currents are very small, increase the amplification, or gain, in order for them to be evaluated and visible. Always make note of these changes.
    NOTE: The duration of the perfusion depends on the experimental paradigm. If the major goal is to measure maximum amplitude, 15 s will be enough to reach the peak for GABA or the plateau for kainate.
15. Always monitor solution levels. Ringer's solution will need to be refilled frequently as it is used between all solution uses. Once recording is completed, turn the voltage clamp to the OFF position and turn the Ringer's solution valve off. Remove the oocyte. Stop recording and save the file. Repeat these steps for all the injected oocytes.

5. Analyzing TEVC recordings

1. Save a copy of the recordings to a USB drive. From there, open the desired file to be analyzed. In the file that opens, the recording can be viewed, marked, and measured. Most common analysis includes measuring of maximum amplitude, activation time, desensitization, and rundown of repetitive applications.
2. Measure the AMPA maximum response and the GABA peak response. To acquire these measurements, first establish a zero level or baseline by finding the red line with the cursor and dragging it to the current generated by Ringer application or baseline.
3. Once the zero level is established, determine the response measurements by using the mouse to drag the green, vertical readout cursor to the desired part of the graph tracer. If wanted, export the traces to be analyzed in any other preferred software
   NOTE: If the zero level is unable to be defined or there is too much noise indicated on the tracer, export the WinEDR file to an offline analytical software, that will allow for modifications to provide a level baseline and the addition of filters in order to reduce noise.

Within a few hours after injection, the synaptic membranes, carrying their neurotransmitter receptors and ion channels, begin to fuse with the oocyte plasma membrane. Figure 1 shows recordings of AMPA and GABA_A receptors microtransplanted into Xenopus oocytes. For most of the analysis, the responses from two or three oocytes per sample were measured, using two or three batches of oocytes from different frogs, for a total of six to nine oocytes per sample. This is done for a large cohort of human subjects to observe group differences. The analysis is straightforward and measures amplitude of responses. It is important to note that the fusion of the transplanted membranes is a polarized process, thus selection of the injection site is very important. Injection into the vegetal hemisphere or into the equator gives consistent results, with all injected oocytes successfully inserting receptors and generating ion currents that follow a unimodal distribution²². Interestingly, when the membranes were initially injected into the animal pole, the oocyte responses followed a bimodal distribution: one group of oocytes had none or very low responses while the other group had large responses, even larger than those from oocytes injected in the vegetal pole or in the equator. To determine if some of the oocytes injected in the animal pole had low responses because the human membranes were being injected and trapped within the nucleus of the oocyte, a mix of membranes isolated from the electrical organ of Torpedo and cDNA coding for the GABAρ1 subunit was injected into the pole of the animal side. The nicotinic receptors from the Torpedo membranes showed a fast activation and fast desensitization during perfusion of acetylcholine²³, while the ρ1 subunit forms homomeric GABAρ1 receptors that do not desensitize during continuous perfusion of GABA^24,25,26,27. If the co-injected sample was accidentally being delivered into the nucleus and not the cytoplasm, then the cDNA would be transcribed into RNA and translated into functional GABAρ1 receptors. Figure 2 shows that when the injection was aimed at the nucleus, oocytes with high responses to acetylcholine did not express GABAρ1 receptors; conversely, oocytes with zero or low response to acetylcholine expressed GABA receptors. Results of this experiment indicate, first, that in low-response oocytes, membranes were accidentally deposited into the nucleus of the oocyte; second, the injection of torpedo membranes into the nucleus does not interfere greatly with the transcription of ρ1 cDNA. A few co-injected oocytes had large responses to both acetylcholine and GABA, suggesting rupture of the nucleus during the injection. The enhanced insertion of transplanted membranes into the animal hemisphere of the oocyte mirrors the polarized localization of the heterologously-expressed neurotransmitter receptors ^26,28. Therefore, by injecting into the animal hemisphere without targeting the nucleus, it is possible to obtain larger responses. This is important to study specimens with low density of receptors, for example from Alzheimer's disease (AD) tissue with low numbers of receptors (Figure 3).

Figure 1: Representative ion currents of oocytes. Ion currents from oocytes injected with membranes from human synaptic receptors were recorded. AMPA receptors were activated with 100 mM Kainate, and GABA_A receptors with 1 mM GABA. V_H = -80 mV. Please click here to view a larger version of this figure.

Figure 2: Co-injection of Torpedo and GABAρ1. Co-injection of filtered membranes (0.1 µm) from the electric organ of Torpedo, rich in acetylcholine receptors, and cDNA coding for the GABAρ1 receptor into the animal pole produced mainly two groups of oocytes based on their responses: one group of oocytes (A) had large responses to acetylcholine (Ach; 1 mM) but no responses to 1 µM GABA (voltage clamped to -80 mV), and oocytes in group (B) had null or low responses to ACh but large responses to GABA. (C) Graph shows mean ± standard error of mean (SEM) of peak current in group 1 (n = 5 oocytes) and group 2 (n = 11 oocytes). One oocyte in group 2 had large GABA and ACh responses suggesting rupture of the nucleus; consequently the distribution of the responses was skewed to low values, as noted by the difference between mean and median of the distribution of membrane current (22 nA vs 6 nA). This result indicates that one of the causes of low-response oocytes is that the membranes are being injected and trapped into the nucleus of the oocyte. Please click here to view a larger version of this figure.

Figure 3: Polarized insertion of cell membranes in Xenopus oocytes. GABA and kainate responses of oocytes injected into the animal or vegetal poles, with unfiltered membranes obtained from a non-AD brain (female, 74 years old, postmortem interval 2.8 h) and an AD-brain (female, 74 years old, postmortem interval 4.5 h). Oocytes injected near the animal pole, without targeting the nucleus, gave larger responses than oocytes injected into the vegetal pole, thus allowing the study of tissue samples with very low numbers of receptors. Student's t-test between vegetal and animal poles: ** p < 0.01, *** p < 0.001, Bars indicate mean ± SEM of peak current, n = 5 oocytes each column. Please click here to view a larger version of this figure.

Analysis of native protein complexes from human brains is needed to understand homeostatic and pathological processes in brain disorders and develop therapeutic strategies to prevent or treat diseases. Thus, brain banks containing snap frozen samples are an invaluable source of a large and mostly untapped wealth of physiological information^29,30. An initial concern to use postmortem tissue is the clear possibility of mRNA or protein degradation that may confound the interpretation of data. For example, the pH of the brain consistently shows a positive correlation with the quantification of mRNA by quantitative real time-PCR, and this correlation is independent of the pathology³¹. Inter-subject differences of pH may lead to differences of mRNA quantification with large variances that may obscure the interpretation of results³². Interestingly, despite the lower quantification of mRNA transcripts as the pH acidifies, the covariances among the different transcripts are maintained³³, providing a framework for obtaining reliable transcriptome profiling of pathological states. Importantly, while mRNA is highly degradable, transmembrane proteins in the postmortem tissue are very resistant to degradation³⁴. Previous experiments in the lab have demonstrated that GABA and glutamate receptors of the human brain are functional after extremely large postmortem intervals³⁵. Moreover, the MSM method allows to directly measure the effects of potential confounding factors like pH at moment of death, agonal state, and mRNA and protein degradation directly on the function of receptors, thus providing ways to use this information in the analysis and interpretation of data.

Modifications and troubleshooting of the technique
As mentioned before, the MSM is a slight modification of the original transplantation of total membranes, which is a method that has been widely validated and confirmed by many labs in academics and the pharmaceutical industry^{12,36,37,38,39,40}. For example, microtransplantation of receptors was important in the development of Eli Lilly's LY3130481, which is a TARP gamma-8-selective antagonist of AMPA receptors which shows brain region selectivity¹². MSM follows the same microtransplantation principle using synaptosomal enriched preparations; therefore, having good synaptosomal preparations is important. We use the Syn-Per method because the results that are obtained with this procedure are very consistent and the amplitude of the ion currents correlate very well with the levels of synaptic proteins in proteomic experiments¹⁰. However, the microtransplantation or the MSM can be adapted to study receptors from many sources (e.g., insect membranes^38,39, neurolemma⁴⁰) with excellent results. There are some disorders where synapses are strongly affected like Alzheimer's disease, or are available in low quantities; in this case, injection of the membranes in the animal side helps in getting larger currents. Increasing the amount of protein injected also increases the fusion of membranes and the size of the responses. Although there is a negative correlation between the concentration of membranes injected and the survival of oocytes, it is possible to find a suitable concentration of membranes that allows the given experimental analysis.

Limitations of MSM
The MSM is a quantitative method that was developed for the study of human receptors or ion channels; therefore, it is well suited to study tissue with limited availability. Because Xenopus oocytes do not have endogenous glutamate or GABA receptors, any response to GABA or glutamate in microtransplanted oocytes comes from the injected receptors that fused with the membrane of the oocytes. However, an important limitation of MSM is the study of ion channels or transporters that are also endogenously expressed in oocytes, as the separation of ion currents from microtransplanted and endogenous channels/transporters is difficult. Future studies silencing endogenous genes may be helpful with this limitation.

The significance to existing methods is that MSM incorporates the native membrane with its corresponding proteins and receptors, and thus, provides a more physiological environment for the analysis. It has been found that the properties of the receptors in their native membrane are retained after transplantation to the oocytes, as observed in neurotransmitter receptors AMPA-type GluR1, α7-AcChoRs, and α4β2-AcChoRs from cultured cells⁴¹.

Future applications of the technique include the study of the electrophysiological properties of different proteins and receptors of interest in the membranes of cultured and non-cultured cells and tissues from healthy and diseased human brains.