The activity of single neurons from adult-aged mice can be studied by dissociating neurons from specific brain regions and using fluorescent membrane potential dye imaging. By testing responses to changes in glucose, this technique can be used to study the glucose sensitivity of adult ventromedial hypothalamic neurons.
Studies of neuronal activity are often performed using neurons from rodents less than 2 months of age due to the technical difficulties associated with increasing connective tissue and decreased neuronal viability that occur with age. Here, we describe a methodology for the dissociation of healthy hypothalamic neurons from adult-aged mice. The ability to study neurons from adult-aged mice allows the use of disease models that manifest at a later age and might be more developmentally accurate for certain studies. Fluorescence imaging of dissociated neurons can be used to study the activity of a population of neurons, as opposed to using electrophysiology to study a single neuron. This is particularly useful when studying a heterogeneous neuronal population in which the desired neuronal type is rare such as for hypothalamic glucose sensing neurons. We utilized membrane potential dye imaging of adult ventromedial hypothalamic neurons to study their responses to changes in extracellular glucose. Glucose sensing neurons are believed to play a role in central regulation of energy balance. The ability to study glucose sensing in adult rodents is particularly useful since the predominance of diseases related to dysfunctional energy balance (e.g. obesity) increase with age.
The brain regulates energy homeostasis through the neuroendocrine and autonomic nervous systems. The ventromedial hypothalamus (VMH), comprised of the ventromedial nucleus (VMN) and the arcuate nucleus (ARC), is important for the central regulation of energy homeostasis. Specialized glucose sensing neurons, within the VMH, link neuronal activity and peripheral glucose homeostasis1. There are two types of glucose sensing neurons; glucose excited (GE) neurons increase while glucose inhibited (GI) neurons decrease their activity as extracellular glucose increases. VMH glucose sensing neurons are generally studied using electrophysiology or calcium/membrane potential sensitive dye imaging.
The electrophysiological patch clamp technique is considered to be the gold standard in the study of ex vivo neuronal activity. In this technique, a glass micropipette electrode is attached to the cell membrane via a high resistance (GΩ) seal. Patch clamp electrodes allow real time recording of action potential frequency (current clamp) or ion conductance (voltage clamp) changes within a single neuron. While the patch clamp technique provides detailed information regarding changes in specific ion channel conductances, a major drawback is that only one neuron may observed at a time. It takes approximately 30-45 min of recording to verify that one is recording from a glucose sensing neuron before even beginning a specific experimental treatment. Moreover, GI and GE neurons comprise <20% of the total VMH neuronal population. Compounding this issue is the lack, in many cases, of an identifying cellular marker for these neurons. Thus, it is clear that despite providing valuable electrical information that other techniques cannot, patch clamp analysis is laborious, time consuming and low yield.
The use of fluorescence imaging of dissociated VMH neurons allows for the study of hundreds of neurons simultaneously. Calcium sensitive dyes can be used to measure intracellular calcium changes, which indirectly correlate to changes in neuronal activity. Membrane potential sensitive dyes are used to monitor membrane potential changes. Measuring cellular membrane potential is a more direct index of neuronal activity compared to changes in intracellular calcium levels. Furthermore, membrane potential dye (MPD) imaging potentially detects smaller changes in membrane potential where action potential firing is not altered and intracellular calcium levels might not change. Both of these fluorescence imaging techniques have been used to study VMH glucose sensing neurons from juvenile mice2-7. While results are less detailed than those obtained with patch clamp electrophysiology, the strength of imaging experiments is that they simultaneously evaluate a large population of cells which inevitably include a significant number of glucose sensing neurons. MPD imaging is particularly useful for studying GI neurons which are more uniformly localized throughout the entire VMH; thus providing an adequate population to study in the dissociated VMH (~15% GI). In contrast, while GE neurons are densely localized to the ventrolateral-VMN and cell poor region between the ARC and VMN, they do not represent a significant number of neurons within the VMH (<1% GE). Moreover, by studying isolated neurons, astrocytic and presynaptic effects are eliminated. This can be an advantage in studying first order neuron effects, as well as a disadvantage since physiological connections and processes are lost.
A limiting factor in both patch clamp electrophysiology and MPD/calcium dye imaging is the need to use younger animals (e.g. mice or rats <8 weeks of age). This is predominantly due to increased connective tissue in combination with decreased neuronal viability that occurs with age. In brain-slice electrophysiology studies, increased connective tissue makes it more difficult to visualize the neurons. Increased connective tissue also makes it harder to dissociate a large number of healthy neurons for imaging studies. Furthermore, neurons from younger animals survive longer during either patch clamp recording or imaging. However, the use of young mice can be a major limitation. Neuronal activity and/or responses to neurotransmitters or circulating nutrients change with age. For example, since energy balance is closely tied to reproductive status, the hypothalamic neurons regulating energy balance may respond differently in pre- vs postpubescent animals. Additionally, many diseases require long term treatment or do not manifest until adulthood. Prime examples of such diseases are dietary obesity or Type 2 Diabetes Mellitus. Since glucose sensing neurons are believed to play a role in these diseases we developed a methodology for successfully culturing healthy adult VMH neurons for use in MPD imaging experiments.
1. Animals
2. Preparation of Perfusion Solution, Coverslips, Glass Pipettes, and Media
3. Cardiac Perfusion
4. Brain Slicing and Dissection
5. Dissociation and Culture
6. Preparation for MPD Imaging
7. MPD Imaging
8. MPD Imaging Analysis
The precise dissection of the VMH away from other hypothalamic areas is important to obtain consistent results. The inclusion of other areas could dilute the VMH neuronal population, changing the % of depolarized neurons calculated. Furthermore, glucose sensing neurons have been identified in other hypothalamic regions, such as the lateral hypothalamus, which may differ functionally and mechanistically from VMH glucose sensing neurons. Figure 1 illustrates the correct anatomical locations for proper dissection. Following the protocol above, brain tissue containing the correct VMH region can be dissociated. Further dissection to precisely separate the VMN and ARC, while maintaining the entirety of each subpopulation, may not be possible. Figure 2 shows an example of healthy dissociated VMH neurons. Using immunocytochemistry, we confirmed that our preparation is >90% neuronal. Only healthy neurons should be used for data analysis and dishes with too many unhealthy neurons should be discarded. Neurons that are unhealthy often have very dark edges, take up the MPD to a greater extent, and have irregular shapes. Additionally, neurons should be plated at a density such that most neurons are not touching each other during recording. Neurons chosen for analysis should not be touching other cells or debris.
Data obtained from recordings of a GI neuron and a nonGI neuron are shown in Figure 3. After obtaining a baseline for 10 min, the extracellular glucose was decreased from 2.5-0.1 mM. The GI neuron shows a robust reversible response greater than the predetermined threshold of 10% change from baseline. The increase in fluorescence reflects depolarization. While GE neuron hyperpolarization responses are also detected, the frequency is too low (<1% neurons) for successful use of this technique on this subtype of glucose sensing neuron. A range of physiological glucose decreases were tested and the percentage of VMH neurons which reversibly depolarized was calculated for each dish. Figure 4 shows that a positive relationship exists between the magnitude of the glucose decrease and the percentage of depolarizing neurons. This demonstrates that successful primary culture of adult murine neurons can be used in conjunction with fluorescence imaging to allow for the study of adult VMH GI neurons.
Figure 1. Coronal section with markers for correct VMH dissection. The dissected tissue contains the VMN and ARC with minimal contamination from other hypothalamic areas. Diagonal cuts should be made from ~25-30% below the top of the third ventricle to ~25-30% between the point where the cortex meets the hypothalamic area (blue *) and the third ventricle (3V). Adapted from bregma -2.8 mm Paxinos and Watson 19989, corresponding to Mouse Brain bregma -1.7mm. Click here to view larger image.
Figure 2. Brightfield image of healthy VMH neurons from an adult mouse. This image was taken 24 hr after dissociation and has been used for MPD imaging. A few examples of neurons that would (blue *) or would not (red x) be used for analysis are marked. Click here to view larger image.
Figure 3. Representative MPD fluorescence traces of a GI neuron (green line) and a nonGI neuron (orange line). Cells were perfused with 2.5 mM glucose (G) recording solution for 10 min, followed by a decrease to 0.1 mM G for 15 min and a return to 2.5 mM G for 15 min. In the GI neuron, the percent change from baseline increased to over 10% during the 0.1 mM G perfusion period (40% average from 20-25 min) and decreased to less than half of this increase in the 2.5 mM G reversal period (15% average from 35-40 min). Click here to view larger image.
Figure 4. The percentage of adult VMH neurons which reversibly depolarize in response to decreased glucose detected using fluorescent MPD imaging. A positive relationship exists between the magnitude of the glucose decrease and the percentage of depolarizing neurons. Click here to view larger image.
The key to being able to study activity of neurons from adult mice is the ability to dissociate healthy neurons. Dissociation of hypothalamic neurons from adult mice is more difficult at several key steps in the protocol compared to neurons from juvenile mice. We have overcome this problem in a number of ways. Making thick 500 µm brain slices minimizes mechanical damage to neurons compared to the usual 250-350 µm slices used for brain tissue from younger mice. However, thicker slices require greater attention to cardiac perfusion and papain digestion of brain tissue. If blood is seen on brain tissue, evaluate the precision and placement of the cut in the right atrium and the needle placed in the left ventricle. The flow rate and quantity of perfusion solution can also be adjusted. A thorough perfusion is imperative since blood is toxic to the neurons and will result in a lower yield of healthy neurons. Papain digestion must be efficient enough to allow for gentle trituration and easy release of single cells from brain tissue. An inefficient papain digestion is suspected when the VMH tissue sinks quickly at the end of digestion and when larger tissue pieces are present during the second trituration. With good digestion, tissue pieces are barely detected by the naked eye during the third trituration. If the trituration is not gentle, neuronal health will suffer. If an inefficient papain digestion is suspected, the concentration of papain can be increased by 5 U/ml or a new vial of papain can be used. However, the papain concentration should not be increased above 30 U/ml since this can impair neuronal health.
The exposure to BSA is another critical step in the protocol. After triturated cells have been centrifuged on the BSA gradient, both the amount of time neurons are exposed to BSA and the amount of BSA allowed to remain with the pellet should be minimized. The supernatant and the BSA gradient should be aspirated immediately following the centrifugation period. Vacuum suction can be used to remove the majority of supernatant, but the last ~100 µl should be removed manually with a pipette so that the cell pellet is not disrupted. Another important aspect of obtaining healthy neurons is to optimize cell density. Neurons plated too sparsely are not as healthy. Since so few neurons are obtained from each mouse brain, it is not practical to count cells before plating. During the time that would be spent counting, a significant number of cells would adhere to the plastic and would be lost. Using 4 dishes per mouse (~1 x 104 cells/ml) as a starting point, empirical determination of the number of dishes is the most practical approach to achieving optimal neuron density. The yield of VMH neurons from older mice can be expected to be about half of the yield from young mice. Further troubleshooting to improve neuron yield involves the quantification of neurons remaining at each step of the dissociation protocol to pinpoint steps of heavy neuronal attrition.
Using our protocol, healthy dissociated neurons can survive in culture up to 2-3 days. However, the use of filtered astrocyte-conditioned growth media allows neurons to survive up to 2 weeks in culture. Growth media can be conditioned by incubating on confluent primary astrocytes overnight. Astrocytes from the same rodent strain, sex, and brain region should be utilized. While the use of astrocyte-conditioned media introduces another variable, this modification may be necessary for alternative experiments requiring more time in culture. Moreover, for reasons which are not clear, we have found that the ability to sense glucose is easily lost with suboptimal neuronal health, despite the maintenance of responses to other stimuli, such as glutamate. Thus, additional care must be taken. To observe neuronal responses to decreased glucose, it is imperative to avoid contamination by bacteria, mold, salts, soap, and glucose. All glassware and nonsterile plastics should be thoroughly washed with dH2O. Besides living contaminants, detergents and salts would affect neuronal integrity and activity. Small amounts of glucose could significantly change glucose concentrations. This would disrupt results since glucose sensing neurons are very sensitive to small changes in extracellular glucose within the nonhyperglycemic physiological range of 0.1-2.5 mM10,11.
When utilizing MPD imaging, consistency and side-by-side experimentation between treatment groups are vital for meaningful results. Different treatment groups should be tested on the same day if possible or at least on alternating days. This minimizes possible variations in culture preparations or MPD aliquots. Additionally, we found the least variation occurs when imaging is performed within 24 hr of neuron dissociation for reasons mentioned above. Since results are analyzed as a % of depolarizing neurons, accuracy and consistency in VMH dissection and neuronal preparation is also essential. The overall VMH neuronal population harvested must be constant and must not contain neurons from other regions. Thus, the protocol above should be followed closely and precisely. However, if alternative experiments are to be performed and population maintenance is not a concern, tissue punches can be made to isolate VMN or ARC neurons.
The methods described here establish how to harvest healthy VMH neurons from an adult mouse and how to use MPD imaging to study glucose sensing neurons. Instead, experiments could be designed to examine VMH neuronal responses to stimuli other than glucose changes. Alternatively, the dissociation protocol may be extrapolated to harvest neurons from other brain regions. Using neurons from adult mice, studies can now be performed using murine disease models that develop or progress with age. Moreover, healthy adult neurons can be further studied using techniques other than MPD imaging such as calcium imaging, electrophysiology, single-cell PCR, immunocytochemistry or a combination of these techniques.
The authors have nothing to disclose.
NIH R01 DK55619, NIH R21 CA139063
Neurobasal-A Medium (Custom) | Invitrogen | 0050128DJ | custom made glucose free |
Hibernate-A Medium (Custom) | BrainBits | custom made glucose free | |
Penicillin streptomycin (20,000 U/ml) | Invitrogen | 15140 | other vendors acceptable |
Stericup vacuum filter units (0.22 μm) | Millipore | other vendors acceptable | |
25 mm Glass coverslips | Warner | #1 25mm round | |
18 mm Glass coverslips | Warner | #1 18mm round | |
GlutaMAX | Invitrogen | 35050 | |
B27 minus insulin (50x) | Invitrogen | 0050129SA | |
Razor blade | VWR | 55411 | |
Vibratome & cooling chamber | Vibratome | Series 1000 Sectioning system | |
Vibratome blades | Polysciences | 22370 | injector or double edge blades from other vendors acceptable |
Papain, suspension | Worthington | LS003124 | |
BSA, suitable for cell culture | Sigma | other vendor acceptable | |
DNAse, for cell culture | Invitrogen | other vendor acceptable | |
cloning cylinders, 6 mm x 8 mm | Bellco Glass | 2090-00608 | |
Membrane Potential Dye (blue) | Molecular Devices | R8042 | |
In-line heater | Warner | SF-28 | |
Syringe pumps | WPI | sp100i | other vendor acceptable |
Closed chamber | Warner | RC-43C | |
Polyethylene tubing | Warner | PE-90 | |
Metamorph | Molecular Devices | alternate image analysis software acceptable | |
Microscope | Olympus | BX61 WI |
used with 10X objective |
Camera | Photometrics | Cool Snap HQ | |
Narrow Cy3 Filter Set | Chroma | 41007a | |
Illumination System | Sutter Instruments | Lambda DG-4 |