We demonstrate a cell culture protocol for the direct study of neuronal and glial components of the enteric nervous system. A neuron/glia mixed culture on coverslips is prepared from the myenteric plexus of adult mouse providing the ability to examine individual neuron and glia function by electrophysiology, immunohistochemical, etc.
The enteric nervous system is a vast network of neurons and glia running the length of the gastrointestinal tract that functionally controls gastrointestinal motility. A procedure for the isolation and culture of a mixed population of neurons and glia from the myenteric plexus is described. The primary cultures can be maintained for over 7 days, with connections developing among the neurons and glia. The longitudinal muscle strip with the attached myenteric plexus is stripped from the underlying circular muscle of the mouse ileum or colon and subjected to enzymatic digestion. In sterile conditions, the isolated neuronal and glia population are preserved within the pellet following centrifugation and plated on coverslips. Within 24-48 hr, neurite outgrowth occurs and neurons can be identified by pan-neuronal markers. After two days in culture, isolated neurons fire action potentials as observed by patch clamp studies. Furthermore, enteric glia can also be identified by GFAP staining. A network of neurons and glia in close apposition forms within 5 – 7 days. Enteric neurons can be individually and directly studied using methods such as immunohistochemistry, electrophysiology, calcium imaging, and single-cell PCR. Furthermore, this procedure can be performed in genetically modified animals. This methodology is simple to perform and inexpensive. Overall, this protocol exposes the components of the enteric nervous system in an easily manipulated manner so that we may better discover the functionality of the ENS in normal and disease states.
The enteric nervous system (ENS) is vast network of nerves and glia that runs the entire length of the gastrointestinal (GI) tract. The ENS functionally controls all aspects of digestion, including peristalsis, fluid absorption/secretion, sensation of stimuli, etc (for review see 1). It contains over 500 million neurons, more than found in the spinal cord, and contains every neurotransmitter class found in the brain. Furthermore, the ENS is unique in that it can function reflexively without input from the central nervous system 2. Understanding of the ENS is crucial, not only to understand its normal physiological role, but to understand its involvement in a variety of neuropathies which can be congenital (Hirschsprung’s disease), acquired (Chagas), secondary to disease states (diabetic gastroparesis), drug-induced (Opioid bowel syndrome), or due to injury (postoperative ileus) 1. In addition, enteric neurons can be a reservoir for viral infection (varicella zoster)3. Because of its similarities to the brain and the high levels of serotonin in the gut, medications aimed at treating central nervous system defects often have unwanted side effects on the ENS 2. It is also noteworthy that many neuropathies such as Alzheimer’s disease and Parkinson’s disease show similar cellular changes in the enteric neurons long before their appearance in central neurons, making the ENS an accessible model to study the pathogenesis of these diseases 4. Therefore, a thorough understanding of the ENS is a necessity in understanding disease states and preventing/predicting pharmacological side effects.
The neurons of the ENS have been traditionally studied in the guinea pig using wholemount preparations 5-7 or cultured neurons 8. Despite the ease at which neurons can be studied in this large animal, this model has many limitations including lack of genetically modified strains, lack of reagents specific to this species, and the high cost associated with ordering and housing these subjects. The development of a murine enteric nervous system model has the unique advantage of various knock out systems, a vast array of other established methodologies that can be used in conjunction with the cell culture technique, and the ability to provide a validation for the guinea pig model.
The ENS is comprised of three plexi that run the length of the gastrointestinal tract: the outer myenteric plexus (between the longitudinal and circular muscle) which is mainly responsible for the peristaltic actions of the gut, as well as the submucosal and mucosal plexi, (found under and within the mucosa, respectively) which largely controls fluid absorption/secretion and the detection of stimuli 1. This method begins with the isolation of the longitudinal muscle/ myenteric plexus (LMMP) preparation by peeling off the outer muscle layer of the GI tract. This dramatically cuts down on contamination issues that arise when the mucosal layer is involved in the isolation. As a result, this process is ideal for the study of neuronal control of motility rather than secretory actions of the ENS.
The method presented here results in a mixed culture of enteric neurons and glia. At least two different types of neurons are present based on previous electrophysiological and immunocytochemical observations 9. The presence of glia is highly advantageous, as they are not only an important cell type to study in their own right, but they contribute to the survival of the enteric neurons 10 and maintain native receptor expression on the neuronal cell surface 11. Furthermore, deficiencies of enteric glia may lead to abnormal gastrointestinal motility disease states, coined ‘neuro-gliopathies’ 12. Therefore, the ENS culture presented here results in several cell types that are ripe for investigation.
The advantages to this methodology are ease of isolation, inexpensive tool requirements, and a short time to master the technique by experienced lab personnel. Limitations of the methodology include low overall cell yield from high tissue volumes and the exclusion of ENS neurons from mucosal and submucosal plexi. This procedure will be highly advantageous to scholars specializing in electrophysiology, immunohistochemistry, single-cell PCR, and other methodologies.
All animal care and experimental procedures were in accordance with and approved by the Institutional Animal Care and Use Committee at Virginia Commonwealth University.
1. Preparation of Sterile Poly-D-Lysine- and Laminin-Coated Glass Coverslips in 24-Well Plates
2. Advance Preparation of Neuron Isolation Solutions
3. Harvest Longitudinal Muscle/Myenteric Plexus (LMMP) Preparation from Mice
4. Digest LMMP
Immediately following isolation of LMMP-derived cells, neurons and other cell types will not be readily evident. Living, round cells of indistinct phenotype can be seen as well as tissue detritus from incompletely digested tissue fragments and connective tissue. This flotsam is of no concern and will be largely removed with the first media change in two days. Do not attempt to clean the slides before this as the healthy, viable cells will be removed as well.
After one day in culture, neurons will begin to show neurite outgrowth. Specific identification of neurons may still be indistinct at this time. Provided the cells are adherent enough to transfer to an experimental chamber, they will be ready for electrophysiological study after one day of culture. However, the cells are more adherent after two days in culture and ideal for function studies (electrophysiology, calcium imaging) from approximately days 2 – 5.
Morphological features of neuron become distinct after approximately a week in culture (Figure 1). Ideal immunocyctochemical features can be identified after about ten days, when neurons display long projections and grow in an almost ‘ganglionic’ like fashion interspersed with glia (Figure 2). This staining suggests it may be possible to study the synaptic interactions of these cells using this methodology.
After one day in culture, neurons are recognized by their sine que non or ‘defining feature’, the action potential (Figure 3). Electrophysiologically, they can be functionally classified into at least two neuronal types: neurons that contain an after-hyperpolarization and increased current/density relationships of sodium and potassium and neurons that do not have an after-hyperpolarization and significantly less sodium and potassium conductance (Figure 4). AHP positive and negative neurons appear to correlate with AH and S neurons seen in previous guinea work, respectively. AHP positive neurons (AH neurons) have multiple long projections originating from around the cell body, while the AHP negative neurons (S neurons) have one long projection which branches many times. This, in conjunction with the immunocytochemical coding in Figure 1, suggests that neuronal population is very heterogeneous.
Figure 1. Immunohistochemical characterization of enteric neurons and glia isolated from the mouse longitudinal muscle. Confocal microscopy revealed neuronal-specific β-III-tubulin (Abcam, rabbit, 1:1,000) staining in whole mount ileal longitudinal muscle (A) preparation from the mouse. Cells isolated from longitudinal muscle/myenteric plexus (LMMP) preparations contain neurons (B; β-III-tubulin, Abcam, rabbit, 1:1,000) that stain positively for calbindin (C) (Chemicon, rabbit, 1:1,000) and calretinin (D) (Swant, rabbit, 1:2,000). Glia cells (E) were visualized with the glia-specific marker GFAP (Chemicon, mouse, 1:500). Antibodies were visualized via appropriate goat secondary antibody Alexa 488 (green, Molecular Probes, 1:1,000)0. Nuclei were visualized using Hoescht 33342 (blue, C-G, 1 μg/ml). No staining was seen when primary antibody was omitted (F). Modified and reprinted from Smith, T.H., et al. Morphine Decreases Enteric Neuron Excitability via Inhibition of Sodium Channels. PLoS One., doi:10.1371/journal.pone.0045251.g001 (2012). Click here to view larger figure.
Figure 2. Neurons and glia isolated from the mouse longitudinal muscle grow in close proximity to one another. Confocal microscopy images indicate the neurons (green, β-III-tubulin, Abcam, rabbit, 1:1,000) and glia (red, GFAP, Chemicon, mouse, 1:500) readily grow adjacent to one another and appear to interact in vitro. Click here to view larger figure.
Figure 3. Electrophysiology of cultured enteric neurons and glia. In current clamp mode, all neurons (A) displayed action potentials upon current injection. Glia (B) do not have action potentials but do display large electrotonic potentials in response to current injection. Protocols in A & B start with a current injection of -0.01 nA and increase to 0.09 nA in eleven 0.01 nA steps. Click here to view larger figure.
Figure 4. Neurons cultured from the mouse ileum are an electrophysiologically heterogeneous population. In current clamp mode, a current injection of 0.09 nA into neurons results in action potentials. S neurons (A), immediately return to the resting membrane potential following stimulation. AH-type neurons (B) display an afterhyperpolarization (AHP) following stimulation in which the resting membrane potential falls below baseline before slowly returning to the initial value. Modified and reprinted from Smith, T.H., et al. Morphine Decreases Enteric Neuron Excitability via Inhibition of Sodium Channels. PLoS One., doi:10.1371/journal.pone.0045251.g001 (2012).
Animals Used
This protocol has been optimized for Swiss Webster mice. However, this method is easily adaptable to other small-sized mammals such as rats and to other strains of mice. We have successfully performed preliminary isolations with C57 mice and μ-opioid receptor knock-outs. However, it is also possible that other strains of mice may be problematic due to morphological variations in the GI tract. Furthermore, there are known differences between mouse strains (C57Bl/6 vs. Balb/c) in the neuronal circuitry of the LMMP preparation 13, which can affect the resulting mix of neurons in the final isolation. Additionally, age should be taken into consideration when using this method, as age-related neuronal losses occur in the myenteric plexus that are specific to cholinergic neurons and their corresponding glia, and can be seen as early as 12 months of age 14. Finally, when adapting this protocol to other species of animals, care should be taken to ensure that the myenteric plexus comes away with the longitudinal muscle rather than remaining attached to the circular muscle.
Tissue Used
Neurons and glia can be cultured from both the ileum and colon using this protocol. Ileal isolations are easier to perform due to a greater amount of available tissue and the ease at which the longitudinal muscle separates from the thick circular muscle. A minimum of two animals are needed to occupy 12 wells of the 24-well plate for applications such as electrophysiology or immunocytochemistry. The isolation of colonic material is more problematic. The thick colonic longitudinal muscle is more difficult to separate from the circular muscle. Furthermore, the underlying circular muscle is thinner in the colon, making it more susceptible to tears. Also, the colon is significantly shorter than the ileum so three animals, at minimum, are required to occupy 12 wells of a 24-well plate. Both the ileum and the colon can be harvested and plated separately from the same animal, but additional animals or the use of less cell plating area will be required for the colon isolation.
Duration of Cell Culture
Electrophysiological recordings are optimally performed in isolated enteric neurons/glia after 2 to 6 days in culture; at this time cells are rounded and pliant, ideal conditions for making a whole-cell seal. After a week in culture cells become flatter and more visually differentiated as the cells adhere to the plate. Immunocytochemical experiments are best done when cells are more firmly attached to the plate, around day 10, to prevent cell loss during numerous wash steps. Mouse enteric cells have been maintained in culture for up to three weeks. Patterns of receptor and neurotransmitter expression over time have yet to be studied in these cultures. However, previous studies on neuronal survival and biochemical/morphological differentiation in traditional guinea pig enteric neuron isolations have found that neurons survived up to 15 days and maintain a high degree of histochemical and morphological properties up to three weeks 8, which suggests that neurons derived from other animals sources may do the same.
Cell Yield
The overall yield of cells from this technique is 12 wells of a 24-well plate from two mice when isolating from the ileum. This is a relatively low number of neurons compared to tissue volume, as the myenteric plexus consists of less than 1% of the total gut wall. One drawback to this method is that to increase the number of cells obtained, one must increase the number of animals used. When animal numbers are increased, it is suggested that several lab members collaborate in the first step of the isolation of the LMMP. This reduces the amount of time the cells are in a disrupted state before final plating and this increases cell survival.
Cell Density
As written, this protocol uses two animals to yield an ileal isolation of low-density cell confluence (~10 – 40% measured after one day in culture). Low-density cell plating is ideal in this methodology to reduce the risk of contamination in the final culture. It is also useful when single cell electrophysiology is performed. Cell number can be increased by using more animals or reducing the final plating area, but risk of contamination increases. To off-set this problem when higher cell densities are required, either increased washing steps or the increased use of antibiotics is suggested.
Cell Heterogeneity
Neurons acquired using these protocols are comprised of at least two functionally distinct populations, as indicated by electrophysiological characterization (Figure 3). However, there are many known types of neurons with specific neurochemical coding patterns found in the guinea pig 15 and the same is probably true of the mouse. This cell heterogeneity is both an advantage and disadvantage to this methodology. Cell heterogeneity is beneficial when performing single-cell functional studies, observing cell-cell interactions or in immunocytochemistry, among others. Cell heterogeneity may be particularly advantageous when studying neuron/glia interactions in culture (Figure 3). However, cell heterogeneity is a hindrance to methodologies such as immunoblotting, where changes in protein expression, etc., cannot be contributed to any one cell type or neuron type. Examination of glia will be less problematic as only two subtypes are shown to exist in the LMMP preparation: intraganglionic and intramuscular enteric glia 16. Other cell types may also be present in this culture, such as interstitial cells of Cajal or fibroblasts.
Plating Cells
In this methodology, cells are plated in twelve wells. Occasionally, one or two of these wells will develop contamination in the first day after isolation. If this occurs, the slides from the contaminated wells are promptly discarded and the well is rinsed with 70% ethanol to kill the remaining contamination. This protocol can be modified to plate all of the isolated cells in fewer wells or in a single dish, and as a result, the risk of contamination will increase. Again, when the risks of contamination increase, extra wash steps or increased antibiotic use is suggested. The antibiotic/antimycotic liquid (Gibco) used in the protocol consists of Amphotericin B, Streptomycin, and Penicillin. Other antibiotic combinations or increased concentrations can be used to optimize the reduction of contamination while preparing the rinse and complete neuron media.
Attaching Cells
Cells are plated on coverslips coated with laminin and poly-D-lysine. Laminin is important in the survival of enteric neurons and glia, and encourages neurite outgrowth and neuronal development 17, and as such, is considered essential to this isolation. However, alternative attachment factors to poly-D-lysine, such as ornithine or Matrigel, may be tried as desired.
In conclusion, this isolation provides a mixed culture of enteric neurons and glia suitable for a wide range of study techniques. This methodology is easy to master, inexpensive, and time efficient. It is our hope that this cell model will contribute to the understanding of various pathologies associated with the ENS.
The authors have nothing to disclose.
National Institute of Health Grant DA024009, DK046367 & T32DA007027.
Reagents | |||
Fisherbrand Coverglass for Growth Cover Glasses (12 mm diameter) | Fisher Scientific | 12-545-82 | |
Poly-D-lysine | Sigma | P6407- 5 mg | |
24-well cell culture plate | CELLTREAT | 229124 | May use any brand |
Laminin | BD Biosciences | 354 232 | |
ddH2O | Can prepare in lab | ||
15 ml Sterile Centrifuge Tube | Greiner Bio-one | 188261 | May use any brand |
50 ml Sterile Centrifuge Tube | Greiner Bio-one | 227261 | May use any brand |
NaCl | Fisher BioReagents | BP358-212 | MW 58.44 |
KCl | Fisher BioReagents | BP366-500 | MW 74.55 |
NaH2PO4 .2H2O | Fisher Chemicals | S369-3 | MW137.99 |
MgSO4 | Sigma Aldrich | M7506-500G | MW 120.4 |
NaHCO3 | Sigma Aldrich | S6014-5KG | MW 84.01 |
glucose | Fisher Chemicals | D16-1 | MW 180.16 |
CaCl22H2O | Sigma Aldrich | C5080-500G | MW 147.02 |
F12 media | Gibco | 11330 | |
Fetal Bovine Serum | Quality Biological Inc. | 110-001-101HI | May use any brand |
Antibiotic/antimycotic 100x liquid | Gibco | 15240-062 | |
Neurobasal A media | Gibco | 10888 | |
200 mM L-glutamine | Gibco | 25030164 | |
Glial Derived Neurotrophic Factor (GDNF) | Neuromics | PR27022 | |
Sharp-Pointed Dissecting Scissors | Fisher Scientific | 8940 | May use any brand |
Dissecting Tissue Forceps | Fisher Scientific | 13-812-41 | May use any brand |
Cotton-Tipped Applicators | Fisher Scientific | 23-400-101 | May use any brand |
250 ml Graduated Glass Beaker | Fisher Scientific | FB-100-250 | May use any brand |
2 L Glass Erlenmyer flask | Fisher Scientific | FB-500-2000 | May use any brand |
Plastic rod (child's paint brush) | Crayola | 05 3516 | May use any brand |
Carbogen | Airgas | UN 3156 | 5% CO2 |
10 ml Leur-lock Syringe | Becton Dickinson | 309604 | May use any brand |
21 G x 1 1/2 in. Hypodermic Needle | Becton Dickinson | 305167 | May use any brand |
Collagenase type 2 | Worthington | LS004174 | |
Bovine Serum Albumin | American Bioanalytical | AB00440 | |
2 ml Microcentrifuge Eppendorf tubes | Fisher Scientific | 13-864-252 | May use any brand |
Nitrex Mesh 500 µM | Elko Filtering Co | 100560 | May use any brand |
Pipette Set | Fisher Scientific | 21-377-328 | May use any brand |
Sharpeining Stone | Fisher Scientific | NC9681212 | May use any brand |
Equipment | |||
LabGard ES 425 Biological Safety Cabinet (cell culture hood) | Nuaire | NU-425-400 | May use any brand |
10 L Shaking Waterbath | Edvotek | 5027 | May use any brand |
Microcentrifuge 5417R | Eppendorf | 5417R | May use a single larger centrifuge with size adapters |
Allegra 6 Series Centrifuge | Beckman Coulter | 366816 | May use any brand |
HuluMixer Sample Mixer | Invitrogen | 15920D | |
AutoFlow Water Jacket CO2 Incubator | Nuiare | NU-4750 | May use any brand |
Analytical Balance Scale | Mettler Toledo | XS104 | May use any brand |