Culturing Hippocampal and Cortical Neurons

Peter J. Meberg and Matthew W. Miller

Department of Biology University of North Dakota Grand Forks, North Dakota 58202

I. Introduction

II. Acquisition of Hippocampal and Cortical Neurons

A. Dissections

B. Cell Dissociation and Storage

III. Short-Term Culture Methods

A. Plating of Neurons

B. Culture Characteristics

IV. Long-Term Culture Methods

A. Preparation of Glia Stocks

B. Long-Term Maintenance of Neurons Using Glia-Conditioned Medium

C. Culture Characteristics V. Summary


Primary cultures of rat and mouse hippocampus and cerebral cortex are widely used to study neuronal properties such as axonal extension, synaptic transmission, and excitotoxicity. Short-term culturing of these neurons can be very straightforward and is perhaps easier than culturing cell lines once dissections are made and cell stocks are frozen. Long-term cultures of relatively pure neuronal populations require slightly more effort, but protocols are described that are less complicated than most published protocols. These include simpler ways to clean and coat coverslips, as well as using glia-conditioned medium to eliminate the need to make individual cocultures of neurons and glia. These methods consistently yield hippocampal and cortical cultures expressing dendritic spines and synapses that survive over 3 weeks in culture. For investigators employing biochemical

METHODS IN CELL BIOLOGY, VOL. 71 Copyright 2003, Elsevier Science (USA). All rights reserved. 0091-679X/03 $35.00

assays where a fairly large amount of protein is necessary, cortical neurons may be especially attractive to use as large amounts of tissue are obtained and available for culture.

I. Introduction

Cultured neurons from rodent hippocampus and cerebral cortex have been popular for studying a variety of cell functions. Short-term cultures of these neurons are useful for studying neurite initiation and extension, as well as the establishment of cell polarity. The ability of these neurons to survive for weeks in culture and form synapses allows for studies of synaptogenesis and synaptic transmission. At all stages these neurons have been used for studies of mRNA and protein trafficking and targeting, signal transduction, and excitotoxicity. Part of the popularity of these cultures stems from their physiological relevance. Both the cortex and the hippocampus are primary sites for acquisition and storage of memories, and are regions significantly affected by neurodegenerative diseases such as Alzheimer's disease, epilepsy, and stroke. The corticospinal tract is also damaged after spinal cord injury, for example. These properties make these cultures excellent and accessible models for studies of normal neuronal development, as well as neurodegeneration.

Hippocampal and cortical neurons are also good models because the synapses formed in culture are not completely unnatural. In the intact hippocampus, pyramidal cells send out recurrent collaterals, which form synapses on other pyramidal cells. In the intact cerebral cortex, a large proportion of cortical neurons also normally synapse on cortical neurons in other layers of the cortex.

The majority of cells in hippocampal cultures obtained from embryonic rats or mice are pyramidal neurons having a highly characteristic morphology, plus a smaller component of interneurons and glia. A drawback of these cultures is the relatively small amount of tissue obtained from hippocampal dissections. A much greater amount of tissue is available from cortical cultures, which is important if doing biochemical assays or Western blots. A drawback of cortical cultures is the greater cell heterogeneity, but there still remain a large fraction of neurons having classical pyramidal neuron morphology.

In general, short-term cultures (<5 days) do well grown in a variety of media, either with serum or in serum-free medium with growth supplements. However, because longer term cultures lose viability, give inconsistent results, or become overrun with glia if grown in this manner, studies requiring long-term cultures and synapse formation have used more complex culture methods that involve the presence of differentiated glia. The most commonly used method is the classic "Banker method'' (Goslin et al., 1998), which involves growing hippocampal neurons on coverslips placed in close apposition to a bed of glia. The Banker method can give beautiful cultures and allows for the harvest of fairly pure neuronal cultures, but is a relatively complicated method. Another commonly used method for long-term cultures is to grow a confluent culture of glia and then plate neurons directly on the glial bed. This method has been useful for electro-physiological studies (e.g., Bekkers and Stevens, 1989) and studying individual neurons. However, glia contamination prevents biochemical analyses of neurons alone.

Our goal is to describe methods for culturing hippocampal and cortical neurons that are relatively simple, but which give consistent, healthy cultures. The first method described is useful for cultures grown for a few days. The second method, using glia-conditioned medium, is useful for longer term cultures grown more than 10 days, when synapses form. These methods are modifications of culture methods described previously by several laboratories (Mattson and Kater, 1988; Brewer et al., 1993; Goslin et al., 1998). We have used the short-term culture methods for rat hippocampal and cortical neurons to study signal transduction/phosphoryl-ation (Meberg et al., 1998), effects of adenovirus-mediated overexpression of ADF/cofilin on neurite outgrowth (Meberg and Bamburg, 2000), and effects of neurodegenerative stimuli on actin inclusion body formation (Minamide et al., 2000). We are currently using the long-term cultures to characterize the function of ADF/cofilin at synaptic sites. Protocols are described in detail for rat neurons, but these methods should work equally well for mice, as many laboratories have published studies using similar methods for mouse hippocampal and cortical neurons (e.g., Xiang et al., 1996; Ferreira et al., 2000; Hasbani et al., 2001).

II. Acquisition of Hippocampal and Cortical Neurons

Both cortical and hippocampal neurons can be obtained from the dissection of a single pregnant rat. Preparation, dissections, cell dissociation, and freezing of cell stocks can be completed in less than 3 h. With this relatively minimal time investment you can have cells ready for experiments for several weeks to follow, especially if using cortical tissue. Dissections are not difficult after minimal practice, if dissection tools are of good quality and a decent dissecting microscope is available.

A. Dissections

Fetal rats are obtained from timed-pregnant Sprague-Dawley rats (Charles River Laboratories) at a gestational age of 18 days (E18). Charles River Laboratories considers day 1 as the morning a vaginal plug is present after introduction of the mate the previous day. At this age, pyramidal cells, not granule cells, are present in the hippocampus (Bayer, 1980) and astrocytes are small in number compared to neurons. Cortical neurons can be obtained from earlier fetal ages (E16-E18), which may reduce glia numbers, but at this time hippocampal dissections are too difficult. Because mouse development is more rapid, cortical neurons from mice should be obtained at E15-E16 (e.g., Hasbani et al., 2001) and hippocampal neurons at E16 (e.g., Ferreira et al., 2000).

We typically obtain 10-14 fetuses/pregnant rat and acquire both cortical and hippocampal neurons in the same dissection. The protocol for the dissections is described and shown in Table I and Fig. 1. We find the best illumination for dissections is from beneath the tissue; dark-field (indirect illumination with a dark background) illumination may work best for early parts of the dissection and then direct illumination through a diffuser for cutting out the hippocampus.

Table I

Dissections of Hippocampal and Cortical Neurons

Equipment and supplies

Halothane (anesthetic) and sealable jar Scissors, scalpel, straight forceps (medium size) 70% ethanol

Two 100-mm and 6-8 60-mm sterile petri dishes containing Hank's balanced salt solution (HBSS) Dissecting microscope in a laminar flow hood Two pair fine forceps (Dumont #5) Fine spring scissors (Fine Science Tools)

Solutions HBSS (500 ml)

50ml 10x HBSS without Ca++ or Mg2+ (Invitrogen/Life Technologies) 1 mM HEPES adjust pH to 7.3


1. Pour ~1ml halothane onto small Kimwipe in small plastic beaker; place beaker in jar.

2. Place pregnant rat into jar to euthanize. Leave rat exposed to halothane for 1 min after breathing stops. Rat should not respond to toe pinch.

3. Ensure death of rat by lung puncture (insert scalpel underneath ribs to puncture lungs).

4. Place rat on back and use squirt bottle to spray abdomen with 70% ethanol. All tools should also be rinsed with 70% ethanol.

5. Lift skin and cut across abdomen with scissors to expose muscle beneath. Again rinse scissors in 70% ethanol.

6. Hold abdominal muscle up with forceps and cut through muscle with scissors to expose uterine horns.

7. Grab uterus with forceps and gently pull upward while cutting away connective tissue and fat with scissors. Ideally the uterus will not touch any external part of the rat!

8. Place uterus in 100-mm dish containing cold HBSS and transfer to a laminar flow hood for brain dissections.

9. Dip fine forceps and spring scissors into ethanol to sterilize and then flame off ethanol.

10. To remove fetuses, tear uterus above each fetus by pinching with two pair of fine forceps close together and then pulling apart. When fetus is exposed, pull up on umbilical cord with forceps to sever the cord and transfer fetus to a new 100-mm dish with HBSS.

11. Quickly cut head from fetus with scissors and transfer head to a 60-mm dish with HBSS.

13. Transfer brains to a 60-mm dish (we usually place 3/dish), making sure that brain tissue is always completely covered with HBSS in subsequent steps.

14. Separate cortical hemispheres from brain (Fig. 1C).

15. Remove meninges from hemispheres (Fig. 1D).

16. Cut hippocampus away from the medial surface of the cortex with spring scissors and place in a 60-mm dish (Fig. 1E). Place cortices in a separate dish.

Hippocampus Mouse

Fig. 1 Dissection of the cerebral cortex and hippocampus. (A) The dorsal view of the head is shown before dissection. To remove the brain, first pinch skull over brain with two pairs of forceps positioned close together and then gently pull them apart to tear away skin/skull covering the brain. Continue this process, always pulling at the same time in diametrically opposed directions with the two forceps until the brain is exposed. (B) The exposed brain can be removed by holding forceps tips together and then sliding them under the brain to gently lift it up while holding the nose down with another forceps. (C) The ventral surface of the brain is shown here, with the dashed line indicating fissure between the diencephalon (d) and the cerebral cortex (c). To remove the cortical hemispheres, use spring scissors to gently cut/tease apart the cortices from the diencephalon while holding the diencephalon with forceps. Be careful not to damage the medial portion of the cortex, which is where the hippocampus is located. (D) The medial surface of the cortical hemispheres is shown after the meninges (the outer covering of the brain containing visible blood vessels) have been removed. The meninges are removed by gently pulling horizontally apart with two pairs of forceps to not damage the cortex. When slightly loose, pull upward on the meninges, out of the liquid, and they should pull free from the brain. Make sure none of the meninges remain! After the meninges are removed the hippocampus should be readily visible as a crescent on the medial surface (noted by arrows). (E) Use a small scissors to cut along the dorsolateral edge of the hippocampus to free the hippocampus. For cortical cultures the olfactory bulb (olf) should be removed, indicated by the dashed line in (D).

B. Cell Dissociation and Storage

Because the methods of dissociating hippocampal and cortical neurons are nearly identical, both are summarized in Table II. The difference is additional initial steps for corticals that involve first chopping the hemispheres into small pieces before trypsin treatment and the inclusion of DNase I in the trypsin solution to break down DNA released by damaged cells. An important aspect of the dissociation is the vigor and number of trituration strokes. Too little trituration will result in incomplete dissociation, which is not bad if there are simply a

Table II

Dissociation of Hippocampal and Cortical Neurons

Equipment and supplies

Razor blade and medium forceps (corticals only)

15-ml sterile tubes (Corning)

Sterile transfer pipettes (Fisher)

Cotton-plugged Pasteur pipettes, autoclaved

Water bath heated to 37° C

Dimethyl sulfoxide (DMSO)

Freezing tubes (Corning)

Cryo freezing container (Nalgene) filled with isopropanol

Solutions HBSS (see Table I)

10 x trypsin (2.5%; Invitrogen/Life Technologies) DNase solution (10 ml) 100 mg DNase (Sigma) 10ml HBSS

pH to 7.0-7.8, sterile filter, store aliquots at —20 C DMEM/10% FBS

90 ml DMEM (Invitrogen/Life Technologies) 10 ml defined fetal bovine serum (Hyclone)


1a. Corticol. Chop into small pieces (~ 1 mm) with a razor blade that has been ethanol/flame sterilized (hold razor blade with forceps). Transfer pieces to two 15-ml sterile tubes using plastic transfer pipettes. Bring volume to 8 ml in each tube and add 1 ml DNase solution and 1 ml 10 x trypsin to each tube.

1b. Hippocampal. Transfer hippocampi to a 15-ml tube using plastic transfer pipettes. Bring volume to 4.5ml with HBSS. Add 0.5ml 10x trypsin solution.

2. Incubate for 15min at 37 C.

3. Remove supernatant and rinse two to three times with DMEM/10% FBS (allow tissue to sink to bottom between rinses). After final rinse, and DMEM/10% FBS to a final volume <2 ml.

4. Triturate 6-10 strokes with the tip of a Pasteur pipette at the bottom of the tube until large chunks are diminished.

5. Triturate an additional 6-10 strokes with a fire-polished pipette (to fire polish, rotate tip of Pasteur pipette in gas flame a few seconds until tip opening is just slightly smaller than originally).

6. If freezing cells, dilute cells to the appropriate volume with DMEM/10% FBS and add DMSO to 8% of total volume. Aliquot and freeze slowly by placing vials in cryo freezing containers and leaving at —80 C overnight. Long-term storage should be in liquid nitrogen.

few clusters of <10 cells left sporadically, but too much trituration will increase cell death. Hippocampal cultures require less vigorous trituration than cortical cultures. Occasionally there will be some cortical tissue that is resistant to trituration, but it is not worth trying to overtriturate to break them down! Simply avoid/discard those pieces when plating or aliquoting neurons for freezing or pass the cells through a sterile cell strainer (Falcon) if needed.

Cells can be plated immediately after dissociation or can be frozen and stored for later use. We bring the dissociated hippocampal neurons from 10 to 12 rats up in 3 ml Dulbeccos modified Eagles medium (DMEM)/fetal bovine serum (FBS) before freezing 2 x 106 cells/ml) and freeze in 0.5-ml aliquots. Corticals are brought to 15 ml DMEM/FBS before freezing (-2 x 107 cells/ml). Cells are viable for a month when stored at —80° C and viable for years when stored in liquid nitrogen (or vapor phase). Typically, about two-thirds of frozen cells are viable after being thawed, as determined by trypan blue staining.

III. Short-Term Culture Methods

A. Plating of Neurons

For studies on neurons grown less than 10 days in culture, culture methods are simple and cells survive well by plating in Neurobasal/FBS and then switching the medium to neurobasal containing B27 supplements (Table III). Because we typically use these cultures within 4 days, no feeding is necessary. Otherwise they should be fed by replacing one-half of the medium every 4-7 days. Cells are plated

Table III

Plating of Neurons on Coverslips and Culture Dishes

Supplies and protocol


Poly-D-lysine-coated coverslips or culture dishes (Table V)


100 ml neurobasal medium (Invitrogen) 2 ml B27 serum-free supplement (50 x) (Invitrogen) 1 ml 200 mM glutamine stock (~ 2mM final) Neurobasal/FBS (add 10 ml FBS instead of 2 ml B27)

1. If using frozen cells, thaw briefly in a 37 C water bath.

2. Dilute cells in Neurobasal/FBS and perform cell counts. Further dilute cells in Neurobasal/FBS for the appropriate cell number.

3. Plate cells on a poly-D-lysine-coated cell culture dish or coverslip. For 35-mm dishes, plate 100,000-500,000 viable cells in 2ml Neurobasal/FBS. For each coverslip plate 10,000-40,000 viable cells in 0.5 ml Neurobasal/FBS (2000-8000 cells/cm2). Surface tension keeps the cell solution on the coverslip, but care is required to not bump the dish and send cells over the edge of the coverslip!

4. After 4 h remove Neurobasal/FBS and replace with 2 ml Neurobasal/B27 medium.

on cleaned coverslips (Table IV) or plastic dishes, which are then coated with poly-D-lysine (Table V).

The coverslip holder we use is made for silicon wafers, but holds twenty-four 22mm coverslips. For poly-D-lysine coating, the coverslips are either placed flat on a piece of filter paper or inside a 35-mm dish before adding the solution. Coverslips can be rinsed either by sequential dipping in beakers, or by adding ultrapure H2O directly to the surface. We typically use the coated coverslips or dishes the same day for cultures, but have seen no adverse effects on short-term cultures when they are stored dry for a few days. However, long-term hippocampal or high-density cortical cultures occasionally exhibit fasiculation of processes if plated on dried and then UV-irradiated poly-D-lysine. To prevent this, the poly-D-lysine coating protocol (Table V) can be modified. Coverslips are UV irradiated prior to the addition of a sterile-filtered poly-D-lysine solution. The dishes or coverslips are

Table IV

Cleaning of Coverslips

Equipment and supplies

Glass coverslips (Carolina Biological Supply)a Plate heater

Teflon holder for 24 coverslips (Fluoroware, Inc.) Pyrex crystallizing dish (Corning)


1. Warm the 2% micro wash solution in a Pyrex dish on a plate heater at setting #2 (near boiling).

2. Soak coverslips in a coverslip holder in the micro solution for ~20min.

3. Rinse in complete changes of ultrapure H2O at least 10 times.

4. Put coverslips/holder in a 160 C oven for ~30min to dry.

aWe normally use 22-mm coverslips, which fit nicely into 35-mm petri dishes. Table V

Poly-D-Lysine Coating of Culture Dishes or Glass Coverslips

Supplies and solutions

35-mm tissue culture dishes (Falcon) or cleaned glass coverslips (Table IV) in petri dishes (Falcon) Poly-D-lysine solution (100 ^g/ml)

400 ^l of 5mg/ml poly-D-lysine mol wt 70,000-150,000 (Sigma) frozen aliquot 20 ml borate buffer Borate buffer, pH 8.4 1.24 g boric acid

1.90 g sodium tetraborate (Borax) to 400 ml distilled H2O

1. Add the poly-D-lysine solution to a culture dish (1.5ml) or as a 500-^l bubble covering a cleaned 22-mm glass coverslip, and incubate for 30 min.

2. Rinse three times with ultrapure H2O.

3. UV 10-20 min (if using coverslips, place the coverslip in a petri dish and then UV)

then rinsed three times with sterile H2O and once with the plating medium. The medium is removed immediately before plating neurons.

A useful type of dish for the live imaging of fluorescence can be made by attaching a glass coverslip to the bottom of a petri dish with a hole in it. Simply drill a 16-mm hole in a 35-mm petri dish and then attach a cleaned 22-mm glass coverslip to the bottom using a thin bead of silicone aquarium sealant (Dow Corning). Because poly-D-lysine does not attach well to glass under acidic conditions, it is important to use borate buffer during poly-D-lysine coating to buffer the acetic acid that is released from the drying aquarium sealant.

For studies involving oxidative damage to neurons, the B27 supplement used in the neurobasal medium may be inappropriate to use as it contains high levels of antioxidants. For this reason, B27 without antioxidants (Invitrogen) is available. Alternatively, cultures can be grown in N2 (Invitrogen) or N2.1 supplements. A recipe for the N2.1 supplement is described in Chapter 5. DMEM/N2.1 or neurobasal/N2 media support the growth of hippocampal and cortical neurons fairly well, but there is a slightly lower level of cell survival, and these media are not recommended for cultures kept longer than 3 days.

B. Culture Characteristics

Both cortical and hippocampal cultures contain primarily neurons, as we typically find <15% nonneuronal cells 3 days after plating, as assayed by cell morphology together with glial fibrillary acidic protein (GFAP) immunoreactivity, an astrocyte marker. Therefore, these cortical cultures are excellent for biochemical assays as the vast majority of the cells are neurons and there is a large supply of them. For visualization of individual neurons, we typically plate 3000-5000 viable cells/cm2 on glass coverslips. At lower densities, neurons grow poorly, and at higher densities, neurites begin growing across each other relatively quickly. For protein isolation we typically plate ^500,000 viable cells in 35-mm culture dishes and get approximately 30 yg of total protein from 3-day cultures. Protein levels increase with time in these cultures as neurite extension and arborization continue.

Hippocampal cultures initially look cleaner than cortical cultures, with a higher percentage of neurons having classical pyramidal cell morphology. Pyramidal cells initially have a single, obviously longer neurite, the axon, and several much shorter, unbranched dendrites. After several days in culture, these dendrites will begin to grow and branch. Cortical cultures tend to have some cell debris present and slightly more neurons that are not pyramidal cells, but a large proportion will still be pyramidal (Fig. 2A). The differences we find in frozen cultures compared to fresh are minor, but include a somewhat delayed initiation of neurites and the initial presence of slightly more cell debris, but glia seem less numerous (or proliferate more slowly).

Compared to many other neuronal types, such as spinal motorneurons or dorsal root ganglia, axonal extension in cortical neurons is relatively slow. Axon lengths after 3 days of growth in culture typically range from 100 to 1000 ym. This slow

Primary Cortical Neurons Div

Fig. 2 Hippocampal and cortical cultures. Live cultures of cortical neurons are shown at 3 (A) and 21 (B) days after plating. (B) The long-term culture is a high-density culture, which yields plentiful protein for protein analysis. (C and D) Fixed cultures of 21-day hippocampal neurons are shown. Pyramidal cells are characterized by a primary apical dendrite with multiple branches along with several other less arborized dendrites (C). Some astrocytes (< 10% of cells) having long branched processes can be found in these cultures (D), especially if they are not treated with AraC a few days after plating. An astrocyte is shown near the center, with two smaller neuronal cell bodies adjacent to it. This is an obvious astrocyte, based on its large cell body and greater width at process branch points, but many others are more difficult to discern unless astrocyte markers such as GFAP are labeled. Scale bar: 100 ^m.

Fig. 2 Hippocampal and cortical cultures. Live cultures of cortical neurons are shown at 3 (A) and 21 (B) days after plating. (B) The long-term culture is a high-density culture, which yields plentiful protein for protein analysis. (C and D) Fixed cultures of 21-day hippocampal neurons are shown. Pyramidal cells are characterized by a primary apical dendrite with multiple branches along with several other less arborized dendrites (C). Some astrocytes (< 10% of cells) having long branched processes can be found in these cultures (D), especially if they are not treated with AraC a few days after plating. An astrocyte is shown near the center, with two smaller neuronal cell bodies adjacent to it. This is an obvious astrocyte, based on its large cell body and greater width at process branch points, but many others are more difficult to discern unless astrocyte markers such as GFAP are labeled. Scale bar: 100 ^m.

growth rate can be advantageous in certain circumstances. For example, if there is a lag time between introduction of a gene and sufficient levels of protein expression, axons will not be overgrown and impossible to follow in the culture before the expressed protein has an effect. However, the relatively slow growth makes it difficult to measure growth rates with live imaging unless it is done for hours. If a faster growth rate is needed, it does appear that the addition of glia-conditioned medium speeds growth rates (see Section IV,B).

Use of the serum-free neurobasal/B27 medium is not supposed to support cell division, but we find that glia do proliferate, albeit at a slow rate. Glia proliferation appears negligible for several days after plating, but even a slow rate of division becomes a problem if exponential growth occurs! Glial proliferation does become a problem after 7-14 days in culture, depending on the dissection, because neuronal protein content in these cultures becomes overwhelmed by contaminating glia. Also, even though neurons exhibit spines and synapse formation, it can be difficult to obtain clear images of individual neuronal arbors due to crowding astrocytes. To kill glia and other dividing cells, cultures can be treated with either 5-fluorodeoxyuridine/uridine (FDUR) or cytosine ^-D-arabinofuranoside (AraC), inhibitors of DNA synthesis. However, concentrations of AraC that prevent glia proliferation also kill neurons cultured in the described manner. FDUR seems less toxic to the neurons, but loss of the glia leads to either neurons that do not regularly form spines or unhealthy neurons that begin dying off after 10 days. We are able to avoid these problems and get more consistent long-term culture results by using the following methods.

IV. Long-Term Culture Methods

We wanted to have neurons in culture that consistently formed active synapses and spines without contaminating glia so that we could readily visualize dendritic arbors and do biochemical assays with neuronal protein, not glial protein. We also did not want the time commitment and complication involved in the many steps of the Banker method and wanted to be able to plate many neurons on 35- or 60-mm tissue culture dishes. To accomplish this, neurons are initially plated as described for short-term neuronal cultures. The only differences are that cultures are then fed with glia-conditioned medium and an AraC treatment is given that spares neurons and decreases the number of nonneuronal cells.

A. Preparation of Glia Stocks

As described for neuronal cultures, large stocks of glia cells can be prepared from a single dissection and then frozen for future use in many experiments. Glia are obtained from cerebral cortices in a manner similar to that described earlier for cortical neurons, but cortices are taken from postnatal animals. The dissociated cells are then grown in culture to allow glia to proliferate. The glia cultures are grown on uncoated culture dishes so that neurons do not attach well. After amplification of the glia number in culture, the cells are lifted and frozen in aliquots for future use as feeder cultures. This protocol for preparing glia cultures (Table VI) closely follows that described by Goslin et al. (1998). The protocol, described for three rat pups, provides sufficient cells for 20-30 frozen aliquots, and each aliquot is sufficient for feeding more than two dozen neuronal cultures for 3 weeks (see later).

B. Long-Term Maintenance of Neurons Using Glia-Conditioned Medium

Neurons are plated in Neurobasal/FBS as described for short-term cultures in Table III and are then fed with glia-conditioned medium on the day of plating (day 0) and again on days 4, 7, 14, and (if needed) 21 (Table VII). This feeding schedule has also been used successfully by other laboratorics for both rat and mouse hippocampal cultures (Morales et al., 2000). The glia-conditioned medium

Table VI

Preparation of Glia Stocks

Equipment and supplies

Scissors, blunt forceps, fine forceps (Dumont #5)

50-ml sterile centrifuge tubes

Five 10-ml sterile pipettes

Primaria 75-cm2 culture flasks (Falcon)

Cell strainer, 70 ^m (Falcon)

Solutions 70% ethanol HBSS (Table I) 10 x trypsin (2.5%; Invitrogen) 10 x trypsin-EDTA solution (Sigma) diluted to 1X DNase solution (Table II) PBS (without Ca2+ and Mg2+) Glial DMEM (GMEM) 500 ml DMEM

50 ml heat-inactivated horse serum

5 ml 100X penicillin/streptomycin (Invitrogen) GMEM/10%DMS0


1. Decapitate newborn or 1-day-old rat pups, letting heads fall into a sterile dissection dish.

2. Dissect out cortices as described for neuronal dissections (Table I, steps 11-15).

3. Chop tissue with a razor blade into small pieces.

4. Transfer tissue to a 50-ml sterile tube and bring volume to ~12ml with warm HBSS.

5. Add 1.5ml of DNase and 1.5ml of 10 x trypsin solutions.

6. Incubate tube for 10min in a 37 C water bath, swirling the tube continuously.

7. Triturate with a 10-ml pipette « 12 times.

8. Return to the water bath and swirl another 5 min.

9. Triturate with a 5-ml pipette 15 times.

10. Draw cell suspension up in a pipette and pour through a sterile screen into a 50-ml centrifuge tube containing 15ml GMEM.

11. Pellet cells at 200 x g for 3 min.

12. Remove supernatant, tap tube on bench to loosen cells, and add 2ml GMEM.

13. Stir gently with a pipette tip, add 13 ml GMEM, and pipette up and down a few times.

14. Plate approximately 7.5 x 106 cells in 15ml total GMEM in Primaria flasks.

15. The next day remove GMEM from four dishes and add 12ml fresh GMEM.

16. Feed every other day with GMEM. To do so, first shake/whack flasks before each feeding to remove loosely adhered cells, remove medium, and replace with fresh GMEM.

17. When nearly confluent (7-10 days), lift cells by rinsing once in warm PBS, adding 2ml

1 x trypsin-EDTA and incubating for 2-3 min at 37 C. Add 8 ml GMEM and pipette over the flask surface to dislodge cells. Transfer cells to a 15-ml centrifuge tube and pellet by centrifugation at 200 x g for 3 min.

18. Remove supernatant and resuspend cells in 3 ml GMEM/10% DMSO. Freeze 0.5-ml aliquots in cryotubes and place in a Nalgene freezing box containing isopropanol. Freeze at —80 C overnight and then transfer to liquid nitrogen.

Table VII

Long-Term Maintenance of Neurons Using Glia-Conditioned Medium

Reagents and solutions

Neurobasal/FBS (Table III) Neurobasal/B27 (Table III) Glia-conditioned medium (Table VIII) AraC (Sigma)


1. Plate neurons as described in Table III.

2. Replace Neurobasal/FBS with 50% neurobasal/B27 and 50% glia-conditioned medium (Table VIII) instead of simply Neurobasal/B27.

3. On day 4 postplating, replace one-half of the medium with glia-conditioned medium containing 2-4 ßM AraC (final concentration of 1-2 ßM).

4. On day 7, and every 7 days thereafter, replace one-half of the medium with fresh glia-conditioned medium.

Table VIII

Plating and Maintenance of Glia Feeder Cultures

Reagents and solutions GMEM (Table VI) Neurobasal/B27 (Table III) AraC (Sigma)

Primaria 75-cm2 culture flasks (Falcon) Protocol

1. Briefly thaw the cryovial containing 0.5 ml glia in a 37 C water bath.

2. Transfer cells to a 15-ml centrifuge tube containing 5 ml GMEM. Pellet cells at 200 x g for 3 min.

3. Remove supernatant and resuspend cells in 6 ml GMEM.

4. Add 3 ml of cell suspension to each of two Primaria flasks containing 9 ml of GMEM.a

5. Feed cultures with GMEM every other day until they reach confluence (~7 days).

6. When cultures reach confluence, replace GMEM with neurobasal/B27 containing 5 ^M Arac to glia cultures to halt cell proliferation.

7. Continue feeding approximately twice weekly with neurobasal/B27 only.

8. For glia-conditioned medium add fresh neurobasal/B27 and leave on cells 24 h. When removed, this "glia-conditioned medium'' should be added directly to neuronal cultures.

"The cells will look quite sparse initially. The 0.5-ml glia aliquots can also be split among four flasks, rather than two, and will just take longer (10-14 days) to reach confluence.

is prepared by adding fresh Neurobasal/B27 medium to glia cultures for 24 h and then removing that medium and adding it to the neurons (Table VIII). On day 4, AraC is added to the glia-conditioned medium right before addition to the neuronal cultures to kill dividing cells. For some unknown reason, the AraC addition to glia-conditioned medium does not harm neurons, but the addition to unconditioned Neurobasal/B27 is toxic to neurons. Because the effective concentration for AraC that we use is lower than commonly reported by other laboratories, it is possible that variations in lots or storage conditions may affect its efficacy, and this should be tested.

The glia-conditioned medium is made by growing frozen stocks of glia in DMEM/horse serum for approximately a week until glia form a confluent layer and then the growth medium is switched to Neurobasal/B27 (Table VIII). At this time, glia are also treated with AraC to prevent the continued proliferation of unwanted cell types such as microglia right before using them for the production of glia-conditioned medium. These cultures should be comprised primarily of type 1 astrocytes (Goslin et al., 1998) and will look somewhat like confluent fibroblasts, being flattened, spread cells without significant extensions. These feeder cultures can be used for approximately 3 weeks after switching to neurobasal/B27, but should be discarded if bare areas begin appearing in the dish or smaller cells start appearing over the monolayer.

Our typical schedule is (i) start the glia 7-10 days before the neurons; (ii) replace the medium with 12 ml fresh neurobasal/B27 the day before plating neurons; (iii) completely remove the medium and add to neurons on the next day; and (iv) replace glia growth medium with fresh neurobasal/B27 until the day before the next feeding. The 12 ml of conditioned medium/flask is sufficient to feed 12 cultures. Therefore, one aliquot of frozen glia can be used to feed 24-48 neuronal culture dishes (see Table VIII). Glia can be grown in Primaria flasks, but we also use regular 100-mm tissue culture dishes. If not using Primaria-treated plastic, glia will just achieve confluence more slowly (~10 days versus 7).

C. Culture Characteristics

Cortical and hippocampal neurons treated with AraC and grown in the glia-conditioned medium have good long-term survivability (>21 days), develop extensive axonal and dendritic arbors (Figs. 2B and 2C), and exhibit hallmarks of mature excitatory synapses, such as dendritic spines and pre- and postsynaptic markers in close apposition (Fig. 3). In these cultures, <10% of the cells are nonneuronal, based on morphology together with GFAP labeling in 21-day cultures.

This low level of glia means that the glia contamination of neuronal protein is minimal in these cultures, especially considering the protein content of the extensive axonal and dendritic arbors. When 300,000 viable neurons are plated on 35-mm dishes, we typically harvest 75-100 ^g of total protein/dish. If additional amounts of protein are needed, larger dishes can be used, but the cell densities should not be increased. At higher neuronal densities, the fasciculation of processes can occur and adhesion to the substrate is reduced.

Factors released from glia may be essential for promoting synapse formation in certain neurons (Ullian et al., 2001). One of the factors that promotes synaptogen-esis is cholesterol (Mauch et al., 2001). Glia-conditioned medium also appears

Glia Cholesterol

Fig. 3 Dendritic spines and synapses in long-term cultures. (A and B) A portion of a dendritic arbor is shown for a hippocampal neuron labeled with DiI (Molecular-Probes). Numerous spines are evident throughout the dendrites. The boxed area is shown enlarged (B). (C) The presence of synapses is demonstrated along dendrites of another neuron by the presence of pre- and postsynaptic markers in close apposition. Synaptophysin (presynaptic) is shown in red and PSD-95 (postsynaptic) in green, with overlapping fluorescence yellow. Many synapses can be seen; three are indicated by arrows. For immunocytochemistry, cells were fixed in 4% paraformaldehyde and permeabilized with Triton X-100 before the addition of synaptophysin (Zymed, rabbit polyclonal, no dilution) and PSD-95 (Upstate Biotechnology clone K28/43, 1:100 dilution). Scale bars: 10^m; scale is the same in B and C. (See Color Insert.)

Fig. 3 Dendritic spines and synapses in long-term cultures. (A and B) A portion of a dendritic arbor is shown for a hippocampal neuron labeled with DiI (Molecular-Probes). Numerous spines are evident throughout the dendrites. The boxed area is shown enlarged (B). (C) The presence of synapses is demonstrated along dendrites of another neuron by the presence of pre- and postsynaptic markers in close apposition. Synaptophysin (presynaptic) is shown in red and PSD-95 (postsynaptic) in green, with overlapping fluorescence yellow. Many synapses can be seen; three are indicated by arrows. For immunocytochemistry, cells were fixed in 4% paraformaldehyde and permeabilized with Triton X-100 before the addition of synaptophysin (Zymed, rabbit polyclonal, no dilution) and PSD-95 (Upstate Biotechnology clone K28/43, 1:100 dilution). Scale bars: 10^m; scale is the same in B and C. (See Color Insert.)

to increase the rate of neurite extension, as the average lengths of axons and dendrites in the cortical cultures are increased by about 30% after 3 days in the presence of glia-conditioned medium. Therefore, when studying neurite outgrowth in cortical neurons, it may be advantageous to include glia-conditioned medium if a faster growth rate is desired. The glia-conditioned medium does not appear to increase neuronal survival, as neuron densities 3 days after plating are equivalent to that of neurons grown in neurobasal/B27 alone.

Do not make the mistake of cursorily looking at cultures at low magnification and assuming that you are seeing extensive networks of axons and dendrites when the apparent neurites are in fact astrocyte processes when observed more carefully! Differentiated type 2 astrocytes can look somewhat like neurons, having extensive and branched processes (Fig. 2D). Astrocytes can usually be discerned from neurons because the cells bodies are flatter, larger, or more irregular then the round, phase-bright neuronal cell bodies. Dendrites and axons also taper or retain their diameter as you move away from the cell body, whereas astrocytic processes often get wider, especially at branch points. Of course, the best way to distinguish astrocytes from neurons is by GFAP immunostaining of astrocytes or by using neuronal markers. To label astrocytes, fix cells in 4% paraformaldehyde, permea-bilize with Triton X-100, and label with a GFAP antibody (Chemicon mAB3402, 1:4 dilution). Astrocyte proliferation should not be a problem when treating the cultures with AraC, but should still be monitored in case frozen AraC aliquots begin to lose their efficacy.

V. Summary

We have described protocols for the short- and long-term culture of healthy hippocampal and cortical neurons. The protocols for cleaning and coating coverslips or culture dishes are faster and simpler than many commonly used protocols. The use of frozen stocks of neurons and glia also greatly reduces the frequency of required dissections. Finally, the use of glia-conditioned medium is (1) effective in producing long-lived neurons with spines and synapses and (2) easier and less complicated than coculturing coverslips of neurons over beds of glia. While short-term cultures do not require glia-conditioned medium, it appears essential that long-term cultures be exposed to factors released from glia. It is also important that glia in neuronal cultures be reduced by AraC treatments if immunocytochem-ical or biochemical assays are to be performed.


Thanks to Steve Schmidt and Kelsey Thibert for performing assays on glia number, neurite lengths, and cell survivability. This work was supported in part by the National Institutes of Health (NS40760) and a Basil O'Connor Starter Scholar Research Award (5-FY99-850) from the March of Dimes Birth Defects Foundation.


Bayer, S. A. (1980). Development of the hippocampal region in the rat. II. Morphogenesis during embryonic and early postnatal life. J. Comp. Neurol. 190, 115-134.

Bekkers, J. M., and Stevens, C. F. (1989). NMDA and non-NMDA receptors are co-localized at individual excitatory synapses in cultured rat hippocampus. Nature 341, 230-233.

Brewer, G. J., Torricelli, J. R., Evege, E. K., and Price, P. J. (1993). Optimized survival of hippocampal neurons in B27-supplemented Neurobasal, a new serum-free medium combination. J. Neurosci. Res. 35, 567-576.

Ferreira, A., Kao, H. T., Feng, J., Rapoport, M., and Greengard, P. (2000). Synapsin III: Developmental expression, subcellular localization, and role in axon formation. J. Neurosci. 20, 3736-3744.

Goslin, K., Asmussen, H., and Banker, G. (1998). Rat hippocampal neurons in low-density culture. In "Culturing Nerve Cells'' (G. Banker and K. Goslin, eds.), 2nd Ed., pp. 339-370. MIT Press, Cambridge, MA.

Hasbani, M. J., Schlief, M. L., Fisher, D. A., and Goldberg, M. P. (2001). Dendritic spines lost during glutamate receptor activation reemerge at original sites of synaptic contact. J. Neurosci. 21, 2393-2403.

Mattson, M. P., and Kater, S. B. (1988). Isolated hippocampal neurons in cryopreserved long-term cultures: Development of neuroarchitecture and sensitivity to NMDA. Int. J. Dev. Neurosci. 6, 439-452.

Mauch, D. H., Nagler, K., Schumacher, S., Goritz, C., Muller, E. C., Otto, A., and Pfrieger, F. W. (2001). CNS synaptogenesis promoted by glia-derived cholesterol. Science 294, 1354-1357.

Meberg, P. J., and Bamburg, J. R. (2000). Increase in neurite outgrowth mediated by overexpression of actin depolymerizing factor. J. Neurosci. 20, 2459-2469.

Meberg, P. J., Ono, S., Minamide, L. S., Takahashi, M., and Bamburg, J. R. (1998). Actin depolymerizing factor and cofilin phosphorylation dynamics: Response to signals that regulate neurite extension. Cell Motil. Cytoskel. 39, 172-190.

Minamide, L. S., Streigl, A. M., Boyle, J. A., Meberg, P. J., and Bamburg, J. R. (2000). Neurodegenerative stimuli induce persistent ADF/cofilin-actin rods that disrupt distal neurite function. Nature Cell Biol. 2, 628-636.

Morales, M., Colicos, M. A., and Goda, Y. (2000). Actin-dependent regulation of neurotransmitter release at central synapses. Neuron 27, 539-550.

Ullian, E. M., Sapperstein, S. K., Christopherson, K. S., and Barres, B. A. (2001). Control of synapse number by glia. Science 291, 657-661.

Xiang, H., Hochman, D. W., Saya, H., Fujiwara, T., Schwartzkroin, P. A., and Morrison, R. S. (1996). Evidence for p53-mediated modulation of neuronal viability. J. Neurosci. 16, 6753-6765.

This Page Intentionally Left Blank

+5 -1


  • Rufina
    How much fudr do you add to neuronal culture to kill dividing cells?
    5 months ago
  • hagos
    Can you grow primary rat cortical neurons in flasks?
    11 days ago

Post a comment