diameters in length) after an overnight incubation, but the proportion

was generally low. A few rare cultures contained many cells with

processes as long as 10 cell diameters. These processes appeared well

attached to the surface, and processes arising from neighboring cell

bodies were sometimes bundled together. Well-attached cells with

processes were observed in one- and two-day cultures, while three-day

cultures usually contained a high proportion of cells that appeared to

have detached from the surface. However, most of the neurons in the

intact worm nervous system complete their outgrowth within 24 hours of

fertilization, and the observation that neurites appeared to grow from

some cells in some cultures and remained healthy for at least this long

suggested that successful primary C. elegans cell culture was possible.

Alternative substrates and methods of cell dissociation were tested in

conjunction with a modified version of the medium designed by Edgar for

use with permeabilized embryos. Because adhesion and axonal outgrowth

were dramatically improved by the use of TESPA and modified Edgar's

medium, Wu's medium and PLL were not used in further experiments.

Edgar's medium is similar to Wu's medium, except that inulin and

polyvinylpyrrolidone (PVP) are included to raise the osmotic pressure of

the medium, and chicken egg yolk is added as a source of cholesterol.

The salts and buffer in Edgar's optimal medium are slightly different

from those in the modified Schneider saline used by Wu and co-workers

but are similar to those in several commercially-available media. Edgar

(personal communication) indicated that commercial L-15 medium was

nearly as good for permeabilized embryos as her optimized salt/buffer

solution, so bulk embryonic cell culture experiments were done with

prepared media obtained commercially. The modified Edgar's medium used

for initial studies consisted of 61% L-15, 20% fetal bovine serum, 15%

inulin solution (7.5 mg/ml in egg salts; see protocol below), 1% cell

culture base mix, 0.25% egg yolk, 5 mg/ml polyvinylpyrrolidone, and 1%

penicillin-streptomycin mix. The egg yolk was omitted after the initial

experiments without apparent ill effect.

Cells prepared either by Dounce homogenization or by

chitinase/chymotrypsin treatment and trituration were tested with this

medium on PLL-coated coverslips. Because attachment was poor in an

initial experiment, cells were allowed to attach to the coverslips for

two hours in a serum-free version of this medium (in which serum was

replaced by an equal volume of egg buffer) and then placed in

serum-containing medium. This led to a clear improvement in cell

attachment, probably because serum proteins can compete with the cells

for attachment to adhesive sites on the coverslip, and the period

without serum probably allowed the relatively poorly adhesive cells to

resynthesize surface and extracellular matrix adhesion molecules lost

during proteolytic dissociation. This improvement was not due simply to

the absence of serum, as cells kept in serum-free medium were

considerably less healthy than cells returned to serum-containing

medium.

Several other commercially-available media, many of which are used for

the culture of other invertebrate neurons, were compared with L-15 in

modified Edgar's medium. Cells were initially dissociated and plated on

activated TESPA in L-15-based media and then switched to

serum-containing media based on Schneider's Drosophila medium, Medium

199, Grace's insect cell culture medium, or CO2-independent medium

(Gibco; see protocol below). Most media were identical to the L-15

based medium except for the different commercial medium component.

Grace's insect cell culture medium has a higher osmolality than the

other media, and so the inulin and PVP were reduced by 10% in this

medium. A qualitative assessment of the survival and differentiation of

cells in these media suggested that some media promoted rapid neurite

outgrowth but poor survival while others promoted survival at the

expense of neurite outgrowth. For example, cells grown in Medium 199

exhibited striking neurite outgrowth after 20 hours, with some processes

as long as 20 cell diameters, but a high proportion of cells appeared to

be degenerating by 48 hours. Cultures that had had cells with long

processes at 20 hours now had cells with thick stumps about a cell

diameter in length. By contrast, cells in Grace's medium showed much

shorter neurites after 20 hours but also much less detachment and

degeneration after three days in culture. CO2-independent medium gave

an intermediate level of cell survival and neurite outgrowth in 20-hour

cultures, but had a high proportion of long, well-attached neurites

after three days. L-15 medium was generally similar to Medium 199, with

somewhat shorter neurites, and Schneider's Drosophila medium produced

both intermediate short-term neurite outgrowth and intermediate

longer-term survival. By four days in culture, cells in all media were

detaching and degenerating.

The use of L-15-based medium for initial cell preparation and plating

appeared to be important. Cells prepared and plated in Medium-199-based

media and then shifted to serum-containing Medium-199-based medium

consistently showed poorer adhesion and neurite outgrowth in overnight

cultures than did cells prepared and plated in L-15-based media and then

switched to Medium-199-based serum-containing medium. Cells cultured in

CO2-independent medium similarly showed better adhesion and neurite

outgrowth if they were prepared and plated in L-15-based media than they

were if prepared in plated in in Medium-199-based media. The reason for

the importance of the initial plating medium is not clear.

Nature of the cultured cells

Cells observed two hours after plating were usually flattened and

round. About 25 percent had extended short processes, usually less than

one cell diameter long. Almost none extended longer processes,

suggesting that the long processes observed after 16-20 hours in culture

had grown after plating.

Cultured cells with long processes after overnight growth resembled

cultured neurons from other organisms. They usually had relatively

flat, small cell bodies and a single major process, which often spread

out at the tip into a flattened, multiply-branched growth cone-like

structure (Fig. 4-9). This major process occasionally branched, but

most did not. Some cells were bipolar, with long neurites roughly equal

in caliber extending from opposite ends of the ovoid cell body.

Neurites often appeared to extend to another cell body and terminate

there. They rarely seemed to interact with one another, crossing

without changing trajectory when they intersected rather than forming

fascicles. Neurites arising from nearby cells in a group, however,

often grew out in fascicles, staying together for most of their length

but occasionally splitting into individual processes.

Evidence that the cells extending processes in culture were neurons was

provided by staining with a neuron-specific antibody. The monoclonal

antibody 611B1 was prepared against sea urchin axonemal microtubles

(Siddiqui et al., 1989) and stains only neurons in whole-mount C.

elegans preparations. The most intensely-staining cells are the

mechanosensory cells ALM, PLM, AVM, and PVM, but all neuronal processes

are reported to stain to some degree. Cultured cells were fixed,

permeabilized, and stained with 611B1 or a control anti-tubulin

monoclonal antibody, YL 1/2, that stains all C. elegans microtubules

(Hyman, 1989) . Most cells appeared to stain with the control tubulin

antibody (Fig. 4-10). Cells without processes had a ring of stain in

the cytoplasm but none in the nucleus. Cells with processes showed

intense staining along the length of the processes. By contrast, cells

without processes were not stained with 611B1, and a subset of the cells

with long neurites were stained (Fig. 4-10). This suggests that at

least some of the cells with long processes were neurons.

Optimized protocol

The optimized procedure for culturing cells determined by the

experiments described above is listed below. This procedure gave

reasonably reproducible cell survival and axonal outgrowth but the

variables have been by no means exhaustively explored.

Media:

L-15 Culture medium (L-15/CM)

24 ml L-15 (Gibco cat. 320-1415A; includes glutamine)

6 ml inulin (ICN cat. 102055; 7.5 mg/ml in egg salts,

autoclaved)

8 ml fetal bovine serum (heat inactivated; Gibco cat.

230-6140AG)

0.4 ml base mix

200 mg PVP

0.4 ml penicillin-streptomycin mix (Gibco cat. 600-5140AG)

Mix, let sit on ice 1-2 hr, filter-sterilize (0.22 um).

egg salts: 11.8 ml 1 M NaCl

4.8 ml 1 M KCl

83.6 ml H2O

base mix: 100 ml H2O

100 mg adenine

10 mg ATP

3 mg guanine

3 mg hypoxanthine

3 mg thymine

3 mg xanthine

3 mg uridine

5 mg ribose

5 mg deoxyribose

autoclave

PVP: polyvinylpyrrolidone MW 40,000

Dialyze against H2O 1-2 days and lyophilize; store

frozen.

L-15 Serum-free medium (L-15/SFM):

24 ml L-15

6 ml inulin

0.4 ml base mix

10 ml egg buffer

0.4 ml penicillin-streptomycin

100 mg PVP

Filter through 0.2 um filter and keep refrigerated.

Medium 199 Culture medium (M199/CM)

24 ml Medium 199 (Gibco cat. 320-1151AG; includes glutamine)

6 ml inulin (ICN cat. 102055; 7.5 mg/ml in egg salts,

autoclaved)

8 ml fetal bovine serum (heat inactivated; Gibco cat.

230-6140AG)

0.4 ml base mix

200 mg PVP

0.4 ml penicillin-streptomycin mix (Gibco cat. 600-5140AG)

Mix, let sit on ice 1-2 hr, filter-sterilize (0.22 um).

Egg buffer:

11.8 ml 1 M NaCl

4.8 ml 1 M KCl

0.34 ml 1 M CaCl2

0.34 ml 1 M MgCl2

2.0 ml 0.25 M HEPES (pH 7.4)

82.5 ml H2O

Filter-sterilize or autoclave before adding sterile HEPES.

Chitinase/Chymotrypsin:

5 mg chitinase (Sigma C-6137)

5 mg alpha-chymotrypsin (ICN 152272)

1 ml egg buffer

Dissolve; place on ice 30 min.; add 10 ml penicillin-streptomycin mix.

Spin if cloudy. Better after a day or so on ice. Can be frozen.

Alternative culture media:

In the recipe for M199/CM above, substitute an equal volume of

CO2-independent medium (Gibco cat. 320-8045AG), Grace's insect cell

culture medium (Gibco cat. 350-1590AG), or Schneider's Drosophila medium

(Gibco 350-1720AG). CO2-independent medium does not come with

glutamine, so add L-gln to a final concentration of 4 mM. Reduce the

amounts of inulin and PVP by 10% when using Grace's insect cell culture

medium.

Culture Procedure:

1. Grow N2 hermaphrodites on 9 cm plates until the plates are crowded

with gravid adults (but not starved). Plates can be left at 15'C

overnight before use. The yield of cells from the procedure below is

often not very high, so plan to use one plate per 2-3 culture samples.

2. Prepare culture media a day in advance.

3. Prepare cover slips with TESPA:

--Wash cover slips in 70% ethanol, 1% HCl.

--Rinse in several changes of H2O over an hour or so.

--Air dry.

--Dip cover slips in 0.5%-4% TESPA (3-aminopropyl-triethoxysilane;

Sigma) in acetone for 5 min. at room temperature.

--Wash in 2 changes of acetone.

--Air dry.

To add reactive aldehyde groups:

--Soak in 4% paraformaldehyde in PBS 30 min. at room temperature.

--Wash in several changes of H2O.

--Air dry.

Use within 1-3 days.

4. Pour M9 onto plates and scrape worms and eggs off with a rubber

policeman. Transfer to a 15 ml tube.

5. Pellet worms in clinical centrifuge and resuspend in M9. Repeat

until supernatant is clear (usually 2-3 washes).

6. Add 5 ml hypochlorite mixture and mix gently at room temperature

until almost all worms are dissolved (about 5 min.).

Hypochlorite mixture:

1 ml sodium hypochlorite (should be fairly fresh)

2.5 ml 1 M NaOH

1.5 ml H2O

7. Wash 3 times with M9.

8. Wash once with Egg Buffer.

9. Add 0.5 ml chitinase/chymotrypsin mixture. Mix gently at room

temperature until embryos round up, single cells start appearing, and

the outlines of individual cells at the edges of embryos are more

distinct than the eggshell (about 5 min.) It helps to monitor a small

drop under the dissecting microscope. Too much digestion almost

certainly damages the cells.

10. Add 5 ml L-15/CM and spin for 1 min at 2000 rpm in clinical

centrifuge.

11. Wash twice in 2 ml L15/CM.

12. Resuspend cells in 1 ml L-15/SFM.

13. Dissociate cells by sucking up and down about 30 times in a pasteur

pipet (slightly drawn out to narrow tip). Gentle dissociation is

critical for good attachment and survival, so the pipet tip should not

be too narrow or the pipetting too vigorous. Allow whole embryos and

large clumps of cells to settle 3-5 min.

14. Remove supernatant to a sterile eppendorf tube. Add 1 ml L-15/SFM

to pellet and repeat dissociation, again allowing 3-5 min for settling.

Repeat once or twice more, until there are no cells in the supernatant

following the settling step. If anything settles out of the

supernatants once they have been removed from the first tube, this

material can added back to the first tube for further dissociation.

15. Filter supernatants through 2 layers of fine nylon mesh (as small

as possible--test to be sure that embryos don't pass through). A 1 cm

square can be attached to the end of a plastic syringe (3cc) by cutting

the tip off the cap that comes with the syringe and screwing the

proximal half of the cap over the nylon.

16. Spin 10 min at 750 rpm (clinical centrifuge) or 1100 rpm (swinging

bucket microcentrifuge). Resuspend in 1-2 ml L-15/SFM and repeat.

17. Resuspend cells in L-15/SFM, assuming 25-30 ul per cover slip.

18. Plate 25-30 ul on each cover slip and incubate in a moist chamber at

20'C in the dark.

19. After 2 hr, remove most of the drop of medium with a drawn-out

pipet and replace with the same volume of M199/CM for best neurite

growth. Repeat. Incubate in a moist chamber in the dark at 20'C (up to

3 days).

Return to C. elegans Protocols
Return to Main Index