Development of techniques for primary culture of C. elegans embryonic neurons

Development of techniques for primary culture of C. elegans embryonic neurons

 

Laird Bloom

MIT

 

from Ph.D. thesis, Massachusetts Institute of Technology, 1993

 

Introduction

 

One of the major limitations of the study of axonal outgrowth in C.

elegans is that direct manipulation of the environments of specific

neurons is not possible, either in partially-dissected preparations or

in culture. In experimental systems in which culture is available,

detailed studies of the interactions of growing axons with their

substrates are possible, including antibody perturbation studies of

cell-surface molecules, direct observation of growth cone behavior, and

pharmacological manipulation of growing axons' electrical activity,

second messengers, or cytoskeleton. In Drosophila, primary culture of

neurons from mutant strains has enabled study of membrane cycling in

shibire mutants and the electrophysiological defects in nap mutants (Wu

et al., 1983) . The availability of a technique for the culture of C.

elegans neurons would provide a new method for the analysis of mutants

defective in neuronal development and function. In particular, several

genes required for normal axonal outgrowth are believed to encode

molecules that affect the cytoskeleton, membrane structure, interaction

with the extracellular matrix, and signal transduction (Leung-Hagesteijn

et al., 1992; T. Otsuka et al., in preparation; Oshima, R. Steven, A.

Ruiz, J. Mancillas, and J. Culotti, personal communication). A

technique for studying these processes in mutant cells lacking specific

molecules in a defined environment might provide information applicable

to the study of axonal outgrowth in a variety of species..

No technique for culturing C. elegans neurons has been published.

Hedgecock et al. (1987) cited unpublished observations that embryonic

cells plated on an adhesive substratum send out single, unbranched

processes, but gave no indication of the conditions used. Conditions

that allow normal cell division to occur after embryos are permeabilized

or partially dissociated and reassembled have been reported. L. Edgar

(personal communication) has developed a technique for removing the

eggshells of embryos as young as the 1-cell stage by a combination of

enzymatic digestion and pipetting through a narrow aperture. These

embryos, which remain surrounded by a membrane, continue to divide to

produce up to 500 cells, and normal differentiation of the major

lineages occurs (as judged by markers for gut, muscle, and germline).

Occasional neuronal processes have been observed in these permeabilized

embryos (L. Edgar, personal communication). Blastomeres that are

separated, reassociated, and cultured under these conditions continue to

divide and differentiate normally (Goldstein, 1992).

The experiments described below were designed to extend these

techniques to the growth of large numbers of dissociated embryonic

cells. The goal was to define media, substrates, and cell isolation

techniques that would permit neuronal differentiation and axonal

outgrowth in short-term cultures. Because long-term culture was not

anticipated, efforts to maintain sterility were made only in preparation

of media. Bacterial and fungal contamination was sometimes evident in

three-day cultures, but not before this. Cells were kept on ice or at

room temperature during initial experiments with cell isolation

procedures. As this appeared to make little difference in the health

of the cells, experiments with different substrata and media were

conducted with cells kept at room temperature throughout the procedure.

 

Experimental Procedures

 

Details of the procedures used are discussed in the text.

Cultures were observed under Nomarski optics using a 100x Planapo

objective lens on a Zeiss Axiovert 10 inverted microscope or a Zeiss

Axiophot microscope. For antibody staining, cultures were fixed for 30

min at room temperature in 4% paraformaldehyde in PBS, followed by three

washes in PBS pH 7.2 containing 1% Triton X-100 and 1% BSA. Cultures

were blocked for 30 min. with 10% BSA in PBS, followed by overnight

incubation in primary antibody at 4'C, three room-temperature washes in

PBS, and a 1 hr incubation in secondary antibody at 37'C. Following

three washes in PBS, the cultures were mounted in Mowiol containing 1

mg/ml p-phenylenediamine and observed.

Ascites fluid from monoclonal antibody 611B1 (G. Pipierno) was used at

a 1:10 dilution, the anti-tubulin monoclonal antibody YL1/2 was used at

a 1:50 dilution. Rhodamine-conjugated goat-anti-mouse (Cappell) and

fluorescein-conjugated goat-anti-rat secondary antibodies (Jackson

ImmunoResearch) were used at a 1:400 dilution. All antibodies were

diluted in PBS.

 

Results

 

Dissociation methods

 

Dissociation of cells for primary culture from other organisms is

usually achieved by a combination of dissection of neurons away from

other tissue, mild proteolytic digestion of basement membranes and other

connective tissues, and physical separation of cells. The small size of

C. elegans embryos makes dissection impossible, and so an additional

necessary step is the removal of the eggshell. Krasnow et al. (1991)

dissociated whole Drosophila embryos simply by gentle Dounce

homogenization. Edgar's technique for removal of the eggshell from

single embryos involved a brief digestion of intact embryos in a mixture

of chitinase and chymotrypsin to begin breakdown of the eggshell

followed by passage of the embryo through a drawn-out micropipet with a

diameter slightly smaller than that of an embryo.

Both Dounce homogenization and enzymatic digestion were used to free C.

elegans embryonic cells from the eggshell. In all cases, populations of

mixed-stage embryos were obtained from mixed-stage C. elegans

populations by washing in M9 followed by treatment with 20% sodium

hypochlorite solution in 0.5 M NaOH until all larvae and adults were

dissolved. Embryos were washed several times in M9 to remove

hypochlorite and then resuspended in egg buffer for

chitinase/chymotrypsin digestion or Ca/Mg-free medium for Dounce

homogenization (8 g/l NaCl, 200 mg/l KCl, 50 mg/l Na H2PO4.H2O, 1 g/l

NaHCO3, 1 g/l glucose; Wu et al., 1983). Dounce homogenization was

performed on a 1 ml cell suspension in a 15 ml glass homogenizer. A

drop of the supernatant was inspected under the dissecting microscope

every 10-20 strokes, and homogenization was stopped when a large number

of individual cells were visible (60-100) strokes. The cells and

embryos were then incubated in a cocktail of collagenase type IA, IV,

and VII (Sigma; 0.1 mg/ml each in Ca/Mg-free medium) for 60 min at room

temperature. In some preparations, collagenase-digested embryos were

sucked up and down repeatedly (triturated) in a drawn-out pasteur pipet

to separate the cells mechanically. Yields from the Dounce procedure

were usually low, regardless of the number of Dounce strokes, the

collagenase mixture used, and the inclusion of a trituration step.

Embryos to be dissociated by enzymatic digestion were prepared by

hypochlorite treatment as above. They were then incubated at room

temperature in a mixture of 5-10 mg/ml each chitinase (Sigma) and

alpha-chymotrypsin (ICN) with gentle agitation until the embryos in a sample

observed under the dissecting microscope began to round up and the

outlines of individual cells at the edges of the embryos began to become

more distinct (usually 5-6 minutes). The reaction was stopped by

several washes in culture medium (see below) containing fetal bovine

serum, which contains protease inhibitors. Treatment with two washes of

soybean trypsin inhibitor before the serum washes did not improve the

apparent health of the cells. Cells were then mechanically dissociated

by trituration in a pasteur pipet with a slightly drawn-out tip,

followed by a period of several minutes in which whole embryos and large

clumps of cells were allowed to settle out of the suspension. The

mechanical dissociation and settling steps were repeated with material

that settled out of suspension until few intact embryos remained. High

yields of cells could be obtained with this technique if the

dissociation was sufficiently gentle, a condition aided by keeping the

tip of the pipet only slightly smaller than the normal pasteur pipet tip

and by keeping the amount of pipetting to a minimum. In addition,

overdigestion with the chitinase/chymotrypsin appeared detrimental to

the health of the cells.

Following dissociation, cell suspensions were filtered through two

layers of fine nylon mesh stretched over the end of a 3 ml plastic

syringe. This effectively removed all of the whole embryos and most of

the L1 larvae that were released from their eggshells during the

dissociation procedure, but it allowed large clumps of cells to pass

through. Filtrates were then subjected to two rounds of low-speed

centrifugation (750 x g) to separate intact cells from particulate

material produced during dissociation. Cells were resuspended in a

volume of medium equivalent to 50 ml per sample to be plated

(approximately three samples per 9 cm plate of worms). This yielded a

drop of cells that was confluent in the center but allowed observation

of individual cells at the edges.

 

Substrates

 

Dissociated cells from a variety of species generally attach to glass

or tissue culture plastic coated with nonspecific charged molecules such

as poly-L-lysine or polyornithine, relatively nonspecific adhesive

proteins such as the lectin concanavalin A (Chiquet and Acklin, 1986) ,

or species-specific extracellular matrix molecules such as laminin,

fibronectin, or collagen (Banker and Goslin, 1991) . Because nearly

all neurons are reported to show some adhesion and axonal outgrowth on

polylysine, initial experiments were done with glass coated with 0.01-1

mg/ml polylysine. C. elegans embryonic cells from some preparations

adhered well to PLL-coated glass cover slips, but often they failed to

remain adhered, or when they sent out axons, the axons seemed very

loosely attached and appeared to float in the medium. Because a more

adhesive substrate appeared to be necessary, the silane derivative TESPA

(3-aminopropyl-triethoxysilane; Sigma), often used for attaching tissue

sections to slides, was tested for its ability to support C. elegans

cell attachment and axonal outgrowth. Initial experiments (using cells

prepared by chitinase/collagenase treatment and trituration and grown in

modified Edgar's medium; see below) showed that TESPA-coated cover slips

allowed more extensive axonal outgrowth than did PLL-coated cover slips,

but this, too, was variable. Reactive aldehyde groups can be added to

TESPA by brief treatment with paraformaldehyde; cover slips covered with

1% TESPA and activated with 4% paraformaldehyde produced the most

consistent axon outgrowth (Fig. 4-9). Cover slips prepared less than

two days before use appeared to be more reliable than older cover slips.

Cells grown on uncoated clean glass failed to adhere.

Poly-L-lysine applied to activated TESPA-coated cover slips appeared to

be no better than activated TESPA alone. Initial experiments with the

vertebrate extracellular matrix proteins laminin, fibronectin,

thrombospondin, collagen (types I, III, and IV) applied to PLL-coated

cover slips showed no obvious improvement in cell attachment or axonal

outgrowth over that observed with PLL alone.

The apparent advantage of paraformaldehyde-activated TESPA over other,

less adhesive, substrates suggested that cells prepared by

chitinase/chymotrypsin and trituration were not particularly adhesive.

This might be caused by loss of cell surface adhesion molecules through

excessive protease treatment. The adhesivity of the culture substratum

has been shown in other organisms to affect the amount of neurite

outgrowth and cell spreading (Bray and Chapman, 1985) . It is possible

that less adhesive substrata would have promoted different behavior of

C. elegans cells, such as cell division rather than differentiation.

 

Media

 

Cell culture media typically contain salts, a buffering agent,

vitamins, precursors for amino acid and nucleic acid biosynthesis,

antibiotics, and a source of growth factors. While some invertebrate

cell culture systems use invertebrate tissues as a source of growth

factors (e.g., Aplysia or Helisoma hemolymph), many invertebrate cell

types have been successfully cultured in fetal bovine serum (Beadle et

al., 1988) . Because invertebrate hemolymph is not commercially

available, fetal bovine serum was used to develop media for C. elegans

cell culture. (Coelomic fluid isolated from earthworms (Arlington Bait

and Tackle, Arlington, MA) showed considerable toxicity to C. elegans

cells in an initial experiment and was not tested further.)

Most invertebrate cells are cultured in air incubators rather than in

the environment of CO2 in air used for most mammalian cells. Several

media designed for use in air incubators were tested for their ability

to support C. elegans cell adhesion, differentiation, and survival. A

medium similar to that used by Wu and co-workers for the culture of

Drosophila larval neurons (Wu et al., 1983) was tested in initial

experiments with cells isolated by Dounce homogenization and plated on

PLL-coated cover slips, and subsequent experiments were conducted with

modifications of the medium designed by Edgar for use with permeabilized

C. elegans embryos. In all experiments, a drop of dissociated cells

(approximately 50 ul) was placed in the center of a 22 x 22 mm or 24 x

50 mm glass coverslip recently coated with PLL or TESPA. In some

experiments, the drop was held in place with by a thick line made with a

grease pencil, but this could be omitted without serious difficulty.

Cover slips were placed cell-side up onto parafilm-covered microscope

slides in a moisture chamber made from a plastic box lined with

water-soaked Whatman filter paper. The chambers were covered in

aluminum foil to protect the cells from light and placed in the same

20'C incubator used for growing worms. Cells were viewed with Nomarski

optics and a 100x oil-immersion objective lens. 24 x 50 mm coverslips

could be placed directly onto the stage of an inverted microscope for

observation without disturbing the drop of medium. Smaller coverslips

were viewed with a conventional microscope after being inverted onto

viewing chambers made from microscope slides to which two coverslips had

been glued, separated by a 5 mm gap. This configuration allowed the

cells to remain in culture medium during observation without being

squashed, but was prone to drying. Cultures that were to be observed

multiple times were kept on larger coverslips.

The first medium tested was similar to that used by Wu and co-workers

(1983) , which contained 25% L-15 medium (Gibco), 66% modified Schneider

saline, and 9% fetal calf serum, heat-treated to inactivate complement.

After preparation by Dounce homogenization, C. elegans cells were washed

in this medium and plated on freshly-prepared PLL-coated cover slips