Info

imr^J

L^ W ^

Fig. 1 Comparison of GFP expression in embryos at times after DNA and RNA injection. DNA (100 pg) or RNA (1 yg) was injected into a single blastomere of 4 cell stage embryos. Confocal images of GFP fluorescence were taken periodically over the first 24 h of development. One hour after RNA injection, GFP expression is already detected in a 16 cell stage embryo, whereas after 5 h of development, GFP fluorescence was still not detectable in DNA-injected embryos. After 10 h of development, DNA expression is detected; however, it is much less widespread as compared to RNA. After 24 h of development, GFP expression is highly mosaic in DNA-injected embryos, but is widespread and uniformly high in RNA-injected embryos. Scale: 500 ym.

Fig. 1 Comparison of GFP expression in embryos at times after DNA and RNA injection. DNA (100 pg) or RNA (1 yg) was injected into a single blastomere of 4 cell stage embryos. Confocal images of GFP fluorescence were taken periodically over the first 24 h of development. One hour after RNA injection, GFP expression is already detected in a 16 cell stage embryo, whereas after 5 h of development, GFP fluorescence was still not detectable in DNA-injected embryos. After 10 h of development, DNA expression is detected; however, it is much less widespread as compared to RNA. After 24 h of development, GFP expression is highly mosaic in DNA-injected embryos, but is widespread and uniformly high in RNA-injected embryos. Scale: 500 ym.

1. Equipment, Materials, and Common Solutions

Stereo microscope (Leica), micromanipulator (WPI), pressure injection system (Parker), Flaming Brown P-97 micropipette puller (Sutter Instrument Co.)

Blunt forceps (Dumont #5 standard tip) to handle and move embryos and a fine forceps (Dumont #55 biology tip) to break micropipette tips

Homemade embryo holding dish made from a 60-mm culture dish with 1-mm nylon mesh (Small Parts Inc.) cut to size and secured to the bottom of the dish with plastcine

Thin-walled borosilicate capillary tubing (with omega dot fiber, 1.0 mm O.D. x 0.75 mm I.D.; FHC) pulled to a fine tip using an electrode puller

Modified Ringers solution (MR in mM), 100 NaCl, 1.8 KCl, 2.0 CaCl2, 1.0 MgCl2, 5.0 HEPES, pH 7.6, using NaOH and filter sterilize or autoclave

2% L-cysteine (Sigma) made fresh in 10% MR. pH 8.0-8.1 using 10 M NaOH

5% Ficoll PM400 (Amersham) in 10% MR and autoclave

One cell to 64 cell stage embyros

2. Procedure a. Dejelly Embryos

Freshly fertilized embryos can be dejellied chemically after cortical rotation (0.5 hpf), although survival seems to improve if performed after the first cell division at 1.5 hpf. Transfer embryos into freshly prepared 2% cysteine and swirl for 3-5 min until jelly coats, but not vitelline membranes, have been removed. Wash embryos extensively in 10% MR.

b. Prepare Pipette Tips

Pull capillary tubes to a tip diameter of —1 ^m and then break manually using fine forceps leaving a tip diameter of —5 ^m. Backload micropipettes by placing small drop (0.5-1 ^l) of solution to the back end.

c. Calibrate Injection Volume

Using a transfer pipette, transfer dejellied embryos into the holding dish containing 5% Ficoll solution. Mount loaded micropipette onto the capillary holding device attached to a micromanipulator and pressure injection system. House air is usually sufficient to drive the pressure injection system (output set —20psi with ejection duration —1 s). Position micropipette tip above the medium in the center of the image field. Eject a volume of solution, and at a known magnification factor, measure the diameter of the suspended drop. If the drop does not stay on the tip of the micropipette, it is sometimes helpful to penetrate an embryo several times. Adjust the duration of the pressure ejection to achieve the desired drop size (volume).

d. Injection

Embryos with the most regular cleavage patterns should be arranged in the holding dish to allow rapid injection in succession. If specific blastomeres are being targeted, it is helpful to arrange four cell stage embryos with the most ideal pigmentation pattern in the same dorsal-ventral orientation. Position the tip of the micropipette just above the blastomere being injected and then lower it rapidly into the blastomere and then out slightly to relieve dimpling of the cell membrane. Inject a known volume of solution into a blastomere and then remove the tip rapidly. Repeat injections for at least two times the desired number of embryos to account for embryo loss. After 5-10 injections, remove tip from solution and reexamine drop diameter as clogging can reduce drop size. Rebreak tip and recalibrate if necessary. Transfer injected embryos into a fresh dish containing 5% Ficoll and allow them to recover several hours to overnight. After recovery, transfer all embryos into 10% MR without Ficoll and allow them to develop to the desired stage.

C. Lipofection

One of the principal shortcomings of expression by blastomere injection is that this leads to early and widespread expression in embryos (Fig. 1). This problem can be reduced with DNA injections; however, the expression of ectopic protein still begins well before neurogenesis. Therefore, defects that result from these early manipulations could be secondary effects of earlier abnormal events. An ideal manipulation for experiments testing the cell autonomous function of a particular gene product later in development is to express the protein in a limited number of cells at the time the event is being studied. For example, in studies examining axonal pathfinding, ectopic genes would ideally be expressed only in the specific neurons at times of particular guidance behaviors. One approach that has allowed late expression in a limited number of cells in the developing retina is lipofection (Holt et al., 1990; Ruchhoeft et al., 1999; Naka-gawa et al., 2000).

We have modified the original Xenopus in vivo lipofection technique (Holt et al., 1990) to ectopically express proteins during and after neurogenesis in the developing neural tube. The delay between lipofection and the earliest axon outgrowth in the spinal cord can be very short. For this reason, we prefer RNA transfection to DNA as it results in more rapid expression and, unlike DNA, does not require mitotic cells. DNA can be substituted in this protocol with the possible advantage of providing longer lasting and higher levels of expression over time; however, the lipofection solution must be injected into a region of the dividing neuroepithelium. In vivo lipofection is similar to blastomere injection except that neural plate stage through neural tube stage embryos (stages 14 to 20; 16.5 to 21.75 hpf) are used and a liposomal carrier reagent is included with the

Fig. 2 Expression in the spinal cord by RNA lipofection. Five hundred micrograms of RNA encoding a myc-tagged protein along with DOTAP was injected into the caudal region of the neural folds on the right side of a stage 17 embryo. After an additional 7.5 h of development, this stage 24 embryo was fixed in 4% paraformaldeyde/4% sucrose overnight. This embryo was then washed extensively with CMF-PBS, dissected to expose the spinal cord, and double labeled for myc (green) and neural-specific tubulin (red). Lateral view of a spinal cord showing two myc-positive neurons that have extended axons (arrows), although many other neurons are labeled in the caudal spinal cord. Dorsal is up and caudal is left. Scale: 100 ^m. (See Color Insert.)

Fig. 2 Expression in the spinal cord by RNA lipofection. Five hundred micrograms of RNA encoding a myc-tagged protein along with DOTAP was injected into the caudal region of the neural folds on the right side of a stage 17 embryo. After an additional 7.5 h of development, this stage 24 embryo was fixed in 4% paraformaldeyde/4% sucrose overnight. This embryo was then washed extensively with CMF-PBS, dissected to expose the spinal cord, and double labeled for myc (green) and neural-specific tubulin (red). Lateral view of a spinal cord showing two myc-positive neurons that have extended axons (arrows), although many other neurons are labeled in the caudal spinal cord. Dorsal is up and caudal is left. Scale: 100 ^m. (See Color Insert.)

genetic material. We have successfully lipofected embryos from stages 14-18 with this approach and believe this range can be extended. However, it is most reliable to target the spinal neurons using older embryos (stages 16-18), and because RNA is expressed so rapidly, delayed lipofection may be necessary for studies involving axon growth and guidance, which does not begin until stage 21. Lipo-fection into the developing neural tube has the added advantage of allowing targeted expression into different anterior-posterior positions along the develop-mentally graded axis of the spinal cord (Fig. 2). Our technique utilizes the transfection reagent DOTAP, although other lipofection agents that have been developed more recently have not been tested. This technique for expression is best suited for studies done in vivo, as typically a small population of cells are transfected.

1. Materials

DOTAP liposomal transfection reagent (Roche). DOTAP is bottled under argon as it oxidizes rapidly; keep vial sealed at 4° C but do not freeze. Due to the instability of DOTAP, vials typically last for less than 6 months.

500 yg/ml Fast Green dye in DEPC water

1-ml syringe with a 27-gauge needle

Stage 15-20 embryos

2. Procedure a. Prepare RNA/DOTAP

Chill several small Eppendorf tubes on ice. Remove a small amount of DOTAP with a syringe and transfer to a chilled Eppendorf tube. The optimum DOTAP-to-DNA ratio was determined to be 3:1 by weight (Holt et al., 1990), and we assume that a comparable optimum ratio exists for RNA, although this has not been tested. Dilute RNA to 0.33mg/ml and transfer 1 yl of RNA to a clean and chilled Eppendorf tube. Add 1 yl DOTAP (1 mg/ml) to 1 yl RNA to achieve 3:1 and flick to mix. Keep solution on ice, as RNA and DNA will precipitate. If a white precipitate does form, new DOTAP may be required. Add 1 yl chilled Fast Green solution to 2 yl of RNA/DOTAP solution and flick to mix. Dilution of RNA with Fast Green helps visualize the injected solution in embryos and facilitates the backloading of micropipettes, which often load poorly with concentrated DOTAP/RNA solutions.

b. Prepare Pipette Tips

Pull, break tips, and load micropipettes as described previously. Calibrate drop volume to at least 5nl. Tips may need to be rebroken occasionally between injections as they clog more easily when loaded with DOTAP.

c. Injection

Place dejellied, stage 14-19 embryos (16.25 to 21 hpf at 23° C) in a holding dish with 5% Ficoll. Inject 5 or more nanoliters into the dorsal region of the developing neural plate, neural folds, or neural tube. Transfer injected embryos into a fresh dish containing 5% Ficoll and allow them to them to develop to desired stage.

D. Single Cell Injection

The delivery of molecules by a single cell injection represents an extreme example of delayed and targeted labeling or manipulation as this procedure is designed for individual neurons in culture. This technique is well suited for the injection of nucleic acids, proteins, or fluorescent dextrans into neurons that have extended processes. The main benefit of this approach is that the acute and cell autonomous effects of proteins on function can be assessed. Further, fluorescent tracers or physiological indicators that may alter aspects of normal cell development or function chronically can be introduced acutely. The primary drawback of this approach is that it is time-consuming and requires specialized and relatively expensive equipment.

1. Equipment and Materials

Cultured neurons (see later)

Phase-contrast microscope with 20 or 40 x long working distance objectives

Eppendorf transjector 5246 or Narishege IM-200 microinjector

K T Brown type micropipette beveler (Sutter Instrument Co.)

Nanosep microconcentrators (Pal Filtron)

Ultracentrifuge (Beckman TL100 rotor or equivalent)

2. Procedure a. Prepare Injection Solutions

Injection reagents are diluted or dialyzed and, if necessary, concentrated in microinjection buffer [5mM HEPES (pH 7.2), 72 mM KCl, 12 mM NaCl, 0.05 mM EGTA] using the appropriate molecular weight cutoff microconcentra-tors. An alternate microinjection buffer that works well for Xenopus spinal neurons is 1.34mMK2HPO4, 0.44mMNaH2PO4,2.0mMHEPES (pH 7.4), 110mMKCl, 10 mM NaCl, 2 mM glucose, 1 mM Na pyruvate, 1 mM ATP, 0.1 mM GTP. Injection solutions must be clarified by ultracentrifugation at >40,000 x g for 20 min before loading to minimize plugging of the microneedle. Fluorescent conjugates of various molecular weight dextrans (^400 yM) are added to all solutions to verify successful injections as well as identify and track injected cells by fluorescence microscopy.

b. Prepare Pipette Tips

Pull microneedles to a tip diameter of —0.2 ^m. It is important to make micro-needles from thin-walled capillaries with omega dot fiber to minimize plugging of the needle tip and to ensure that injection solutions reach the tip without air bubbles. Microneedles are loaded either by adding injection solutions directly to the back of the microneedle or by using a drawn-out Pasteur pipette as a back-loading micropipette. For injections of viscous solutions, such as concentrated protein (>10mg/ml), it is necessary to first bevel microneedles (tip diameter —0.3 ^M) using a micropipette beveler.

c. Injection

Microinjections of isolated Xenopus spinal neurons growing in culture (see later) are done at room temperature on a microscope stage, and cells are allowed to recover for 1-4 h prior to subsequent manipulations. Back pressure is maintained to the injection needle to prevent dilution of the injection solution with culture medium through capillary action and to help prevent clogging. Injection times (—0.2 s) and pressures (usually 150% of the back pressure) are adjusted to minimize cell damage during microinjection. Using the Eppendorf microinjector, we position the microneedle tip adjacent to the cell soma and in the same focal plane as the nucleus to set the z limit. The needle is then moved up a few micrometers and is positioned directly over the soma and to one side of the nucleus. This is the starting position. During automatic injection, the needle tip will travel at an angle to the same x and y coordinates and down to the preset z limit, which should be near the middle of the soma. This is the preferred method as injection time is constant and damage to the cell is minimized. Injections can also be done by moving the microneedle down manually to penetrate the cell body adjacent to the nucleus. This method must be used if using the Narishege IM-200 injector.

III. Culturing Xenopus Spinal Neurons

Xenopus spinal neurons as a culture system offer several advantages over other model systems. First, Xenopus neurons and nonneuronal cells survive for more than 24 h in culture at room temperature without additives in simple buffered physiological salt solutions. Second, spinal neuron cultures contain a mixture of functionally distinct neuronal types that extend axons over a variety of substrata. Third, Xenopus neurons form functional synapses between neurons and with nonneuronal cells within 24 h in culture. The extent of interactions among neurons and with nonneuronal cells is dependent on the culture technique used. Culture methods exist that generate (1) isolated neurons with few interactions with other neurons, (2) neurons in tissue explants with extensive synaptogenesis with other neurons within the explant, and (3) mixed cultures containing neurons that synapse upon muscle and/or skin cells. In addition to varying the density and composition of cells cultured, our work has benefited from the ability to grow

Xenopus spinal neurons on a variety of different extracellular matrix proteins (laminin, fibronectin, tenascin, etc.), cell adhesion molecules (L1, N-cadherin), and artificial substrata (tissue culture plastic, poly-D-lysine, glass). The following sections outline the basic technique for isolating Xenopus neural tissue from different age embryos, along with variations in culture composition mentioned earlier.

Toward the end of Xenopus gastrulation, induction of the dorsal ectoderm results in formation of the neural plate, a spatially distinct population of neuronal precursors. Around 16 hpf the neural plate begins to invaginate (NF14; Nieuw-koop and Faber, 1967), taking on a cylindrical conformation. Closure of the neural tube in the Xenopus embryo occurs through stage 20 (21.75 hpf) and coincides with pioneering axon outgrowth. After closure of the neural tube, the intact spinal cord can be isolated easily from the somites (lateral tissue, which is the source of muscle cells in mixed cultures), notochord, and skin. These tissues contribute the majority of nonneuronal cells in culture, including myocytes, fibroblasts, melanocytes, and unidentified round cells. Following mechanical isolation, the neural tube may be dissociated and streaked onto the culture substrate as isolated cells or plated as large explants containing dozens of cells. The neural tube at this stage is composed of postmitotic primary neurons, mito-tically active neuronal precursors that generate secondary neurons, and glial cells. For an in-depth description of neuronal proliferation and differentiation in the Xenopus spinal cord, see Hartenstein (1989, 1993).

By 48 hpf the Xenopus neural tube is composed of nonneuronal cells (glia and fibroblasts), mitotic secondary neurons, and fully differentiated postmitotic primary neurons. This neuronal population consists of eight distinct subtypes in vivo, including cholinergic motor neurons, substance P immunoreactive sensory neurons (Rohon-Beard cells), several classes of GABAergic and glycinergic inhibitory interneurons, and glutamatergic excitatory interneurons. The identity of neurons in the tadpole spinal cord and their functional circuitry have been reviewed by Roberts (2000). If development proceeds normally in culture, a 24-h-old culture from 24 hpf neural tissue should yield these eight types of neurons; however, it is likely that the differentiation of spinal neurons is altered in culture. Although it has been reported that many aspects of differentiation proceed normally in vitro, this has not been established for all neuronal subtypes. Undoubtedly cellular interactions that induce the differentiation of specific neuron types are lost during the culture process. One way to reduce and possibly help characterize these necessary and mostly undefined inductive cues is to culture spinal cords at later times in development.

Despite the uncertainties presented by a mixed neuronal population, cultured Xenopus spinal neurons offer many advantages for the study of cell biological mechanisms controlling motility, adhesion modulation, cytoskeletal dynamics, and the signaling mechanisms underlying these phenomena. These aspects of cell function appear to be well conserved among different neuronal types in culture. For example, our work examining the role of spontaneous and induced calcium transients and their effects on local adhesion and subsequent cytoskeletal rearrangements has not revealed subpopulations with divergent responses to calcium signals. The benefits of this system, which include the ability to easily introduce exogenous nucleic acids and proteins into developing neurons by blastomere injection and the rapid and robust growth of neurons at room temperature, greatly outweigh any limitations.

A. Dissociated Mixed Cultures

This type of culture has been used extensively in studies of the cellular and physiological mechanisms underlying synaptic development and plasticity. Using mixed cultures, electrophysiological recordings and imaging can be conducted easily due to the accessibility and accurate identification of presynaptic neuronal cell bodies and postsynaptic muscle cells. Mixed cultures also support studies of the development of the neuromuscular junction; however, it should be noted that neurons in mixed cultures, unlike neural-enriched cultures, are exposed to unidentified factors likely originating from myocytes that stimulate neurite outgrowth (Holliday and Spitzer, 1993). This approach has proven useful for examining the development of axons and myocytes, as well as synapses between them. The dissection and culture technique is used on embryos ranging from stage 15 to 24, although complete dissociation of spinal neurons past stage 20 is difficult and may require the addition of 5 mg/ml trypsin, as reported previously (Anderson and Cohen, 1977).

1. Materials

CMF-MR (same as MR, except CA2+/Mg2+ free with 1 mM EDTA

Stage 15-24 embryos

Untreated or coated culture dishes (with or without acid-washed glass coverslips)

Antibiotics (penicillin/streptomycin and gentamycin, Sigma)

2. Procedure a. Embryo Preparation

Transfer the required number of stage 15-22 embryos (Figs. 3A and 3B) into a 35-mm petri dish containing 100% MR for initial dissection. We normally use one dorsal section for each culture dish, although a higher or lower density can be achieved using more or less than one dorsal section per dish. All dissections are carried out at 10-50 x magnification using a standard dissection microscope equipped with a fiber optic external light source (Leica Optical). The jelly coat and vitelline membrane surrounding embryos are removed manually by pinning the jelly coat with blunt forceps while piercing and pulling away the surrounding membranes with a fine forceps.

Fig. 3 Schematic diagrams of neural plate and neural tube dissections. (A) Dorsal view of a stage 15 Xenopus embryo. (B) Lateral view of a stage 22 Xenopus embryo. Dorsal views of stage 15 (C) and stage 22 (D) embryos with dashed lines indicating the major incisions performed to separate the dorsal aspect of the embryo. (E) Schematic cross section of a neural plate indicating the location of the neuroectoderm (NE) relative to the paraxial mesoderm (PM), notochord (NC), and archenteron roof (AR). All germ layers can be cultured together in mixed cultures or NE can be isolated and plated separately as neural-enriched cultures. (F) Schematic cross section of a neural tube (NT) and associated tissues, including the somitic mesoderm (S), notochord (NC), and the surrounding dorsal pigmented epidermis (gray line).

Fig. 3 Schematic diagrams of neural plate and neural tube dissections. (A) Dorsal view of a stage 15 Xenopus embryo. (B) Lateral view of a stage 22 Xenopus embryo. Dorsal views of stage 15 (C) and stage 22 (D) embryos with dashed lines indicating the major incisions performed to separate the dorsal aspect of the embryo. (E) Schematic cross section of a neural plate indicating the location of the neuroectoderm (NE) relative to the paraxial mesoderm (PM), notochord (NC), and archenteron roof (AR). All germ layers can be cultured together in mixed cultures or NE can be isolated and plated separately as neural-enriched cultures. (F) Schematic cross section of a neural tube (NT) and associated tissues, including the somitic mesoderm (S), notochord (NC), and the surrounding dorsal pigmented epidermis (gray line).

b. Dissect Dorsal Region

Isolation of the dorsal aspect of the embryo is achieved through a series of precise incisions using fine forceps. An initial incision perpendicular to the longitudinal axis and immediately caudal to the hindbrain (Figs. 3C and 3D) serves to expose the hollow interior of the embryo. The entire dorsal region of the neural plate or tube can now be isolated by making two parallel incisions along either side. This is accomplished by inserting one forcep into the interior of the embryo adjacent to the neural tissue and pinching longitudinally in the posterior direction. A final cut near the posterior end of the embryo results in a rectangular, multi-layered piece of tissue containing the neuronal cells (Figs. 3E and 3F).

c. Dissociate Cells

Transfer all dorsal sections into a 60-mm petri dish containing 100% CMF-MR solution for a 30- to 60-min incubation period. Stage 15 neural plates should not require more than 30min to dissociate; however, longer incubation times and possibly additional mechanical disruption are necessary to dissociate older dorsal sections. Dorsal sections should be noticeably disassociated prior to plating.

d. Prepare Culture Dishes

Each dorsal section is plated onto culture substrata in 100% MR containing antibiotics (50 yg/ml penicillin/streptomycin and gentamycin). Culture substrata should be prepared ahead of time. Many cell types contained within dissociated dorsal sections will morphologically differentiate on a variety of substrata ranging from tissue culture plastic and glass left untreated to these same surfaces coated with ECM molecules. If ECM proteins are used, we typically coat them at ~10 yg/ml in 1 x PBS for at least 1 h prior to culturing.

e. Cell Plating

Unlike most primary cell cultures, dissociated Xenopus cells must be plated directly onto the culture substratum using a drawn-out glass pipette to adhere. Plating pipettes are made by flaming the neck of a 9-in. sterile Pasteur pipette over a Bunsen burner and pulling the softened glass about 3 to 6 in. to a less than ~1-mm shaft diameter. The tip is broken off repeatedly to achieve an even, beveled tip. Cells can now be plated by sucking up a dissociated dorsal section into the pulled glass pipette and carefully streaking the cells out onto the culture dish in parallel lines. Plating pipettes can also be used to triturate incompletely dissociated neural tubes. Repeat for the remaining dorsal sections onto separate culture dishes. Dishes should be maintained near 23° C and not moved for at least 30min after plating. Neurite outgrowth can be observed within 6 h of plating on laminin (LN), although this is influenced heavily by the tissue culture substrata, as discussed later.

B. Neural-Enriched Cultures

Isolation of the Xenopus neural tube in tissue culture provides a useful method for eliminating large numbers of nonneuronal cell types that can confound the analysis of cell autonomous functions. Dissection techniques described earlier and elsewhere (Tabti and Poo, 1990) provide a framework for this culture method. Examination of growth cone motility and other processes in vitro is often less complicated using isolated primary neurons free from potential postsynaptic targets (mainly myocytes) or other cell types that influence neurite outgrowth. This requires that neural tissue be isolated mechanically from other tissue, such as the somatic mesoderm, epidermis, and notochord.

Neural-enriched cultures can be prepared either as dissociated neurons or as spinal cord explants. Dissociated cultures allow for a more precise examination of properties such as axonal and dendritic length, intracellular protein localization, and neuronal morphology. In contrast, the cell bodies of growth cones extending from within explants are difficult to identify, and both cell bodies and neurites often interact with neighboring neurons. However, explant cultures are advantageous as they allow the simultaneous examination of many growth cones that typically extend out in close association. The ability to monitor aspects of cellular physiology, such as growth rate, calcium dynamics, protease activity, or membrane potential in multiple cells simultaneously, makes this system amenable to high-throughput approaches. In particular, this approach could aid in the analysis of the specific effects of various pharmacological agents, diffusible signals, or substratum molecules.

Neural tissue can be isolated as early as stage 15; however, at this time in developmental, the neural epithelium consists of a bilayer of pigmented superficial cells and nonpigmented cells that lay atop the deeper endoderm and paraxial mesoderm (Fig. 3E). Mechanical isolation of the neural tissue requires separation of these tissues with collagenase. At the other extreme, the dissection of embryos beyond stage 23 becomes increasingly difficult due to the tight association the spinal cord forms with the axial somites. At this stage, the spinal cord itself also seems to become incased in matrix molecules and embryos begin to have reflex movements. For our purposes, isolation of the neural tube between stages 21 and 23 is ideal, providing a self-contained, cylindrical neural tube that is separated easily from the surrounding epidermis, notochord, and somites (Fig. 3F).

1. Materials

Electrolytically sharpened tungsten wire mounted on holding tool

Collagenase B (Boehringer Manheim)

2. Procedure a. Common Procedures

The isolation of dorsal sections can be carried out as in steps a and b of the protocol for mixed nerve-muscle cultures.

b. Collagenase Treatment

Using a Pasteur pipette, transfer dorsal sections to a 35-mm dish containing 1 mg/ml collagenase B in 100% MR and incubate for 5min. Subsequent dissection of neural tubes may be carried out in this collagenase solution; however, if a large number of neural tubes are being dissected, they should be done in batches, as prolonged collagenase exposure may be detrimental to cell survival.

c. Isolate Neural Tissue

Shortly after collagenase treatment, neural tissue will begin to separate from adjacent nonneuronal tissues. An electrolytically sharpened tungsten wire with a right angle bend is useful in facilitating tissue separation. When dissecting neural plate stage embryos, use the curved edge of the hooked wire to assist neural ectoderm division from underlying tissues. When dissecting neural tube stage embryos, first peel back the dorsal pigmented epithelia by pulling or hooking the edges that have begun to separate from the deeper tissue layers. Skin is removed most easily if peeled back intact. Upon removal of the epidermis, the neural tube and lateral somitic mesoderm are clearly visible (see Fig. 3F cross section). Somites can now be flaked away using a forceps or the curved edge of the tungsten wire, thus exposing the spinal cord and underlying notochord (small cylindrical structure ventral to the neural tube). Separate the notochord from the spinal cord by pushing the curved edge of the tungsten wire between these tissues, starting at sites where the collagenase has already loosened the attachment.

d. Plate Neural Tissue

For dissociated neural-enriched cultures, neural tissue should be transferred to 100% CMF-MR and plated as described earlier, taking extra caution when plating such a small amount of tissue. If neural tube explant cultures are being made, transfer all spinal cords back to 100% MR. Each neural tube must be cut into small pieces with a diameter of ^100 ^m prior to plating. A typical stage 22 spinal cord is approximately 1 mm in length and 100 ^m in diameter. Using a sharpened tungsten wire, cross section this tissue into at least 10 pieces followed by perpendicular cuts to yield ~20 individual explants. Transfer all explant pieces using a Pipetman onto prepared culture substrata submersed in medium as described previously. Explants should be spaced adequately and may be repositioned within the dish during plating; however, cultures should not be moved for several hours after plating to allow adhesion and preserve explant spacing. Outgrowth on laminin occurs as early as 4-6 h following culture and continues for more than 24 h after plating.

C. Cocultures

Although mixed cultures will often result in stochastic interactions between neurons and target cells (Stoop and Poo, 1995), synaptic connections are enhanced by coculturing neural tube explants near target tissue explants. Cocultures between neurons and targets, such as muscle and possibly skin, can be created in vitro (Fig. 4). Coculturing neurons with target cells proceeds essentially as described earlier, except that during the initial dissection, portions of skin or myo-tomal tissue are kept aside for coculturing. Once spinal cords have been isolated, transfer each tissue type separately into fresh 100% MR. Next, nonneuronal tissues are cut into small pieces as was done for neural tube explants. After all tissues have been cut into appropriate-sized pieces, transfer neural explants into a culture dish, taking care to place pieces in an orderly fashion. Finally, place target tissue explants in close proximity to spinal explants such that extending axons are likely to encounter target cells. Synaptically connected Xenopus spinal neurons in culture are also a very useful and often studied model system of synaptic physiology. Although neural tube explants contain functional synapses within the explant, these types of connections are studied most easily in dissociated cultures.

D. Culture Substrata

Xenopus neuronal cultures are well suited for investigations of cellular mechanisms regulating growth cone motility and guidance, as neurons extend axons on many biological and artificial tissue culture substrata that have different adhesive

Fig. 4 Cocultured spinal cord explant with surrounding myocytes. After 24 h in culture, cells were fixed and stained for neural-specific ^-tubulin (green) and actin (red; Alexa-546 phalloidin). Isolated growth cones (arrow) can be identified, as well as functional synaptic contacts between neurons and muscle cells (as suggested by a-bungarotoxin staining of clustered acetylcholine receptors, not shown). Scale: 50 ^m. (See Color Insert.)

Fig. 4 Cocultured spinal cord explant with surrounding myocytes. After 24 h in culture, cells were fixed and stained for neural-specific ^-tubulin (green) and actin (red; Alexa-546 phalloidin). Isolated growth cones (arrow) can be identified, as well as functional synaptic contacts between neurons and muscle cells (as suggested by a-bungarotoxin staining of clustered acetylcholine receptors, not shown). Scale: 50 ^m. (See Color Insert.)

and growth-promoting properties. We have tested a variety of tissue culture substrata, including laminin, fibronectin, tenascin-C, vitronectin, N-cadherin, L1, polylysine, tissue culture plastic, and untreated glass. Substrata-specific requirements exist for coating proteins onto culture dishes and glass coverslips, as well as limitations with respect to the types of cultures that will extend axons on a given substratum. For example, in order to bind cell adhesion molecules such as L1 and N-cadherin onto glass coverslips or plastic, these surfaces must be precoated with nitrocellulose as described previously (Payne et al., 1992).

Fig. 5 Xenopus spinal neuronal morphology varies depending on culture substrata. (A) Neurons plated on 10 ^g/ml tenascin-C have short axons and highly adherent growth cones with many filopodia (arrow). (B) Neurons plated on 250 ^g/ml poly-D-lysine have short axons and highly adherent growth cones with broad lamellipodial (arrow). (C) Neurons plated on 10 ^g/ml fibronectin extend moderate-length, unbranched axons with many filopodia along their length (arrowheads) and small, highly motile growth cones (arrow). (D) Neurons plated on 10 ^g/ml laminin extend very long, smooth axons that tend to be adherent only at their small, highly motile terminal growth cones (arrow). Scale: 10 ^m for A-C and 20 ^m for D.

Fig. 5 Xenopus spinal neuronal morphology varies depending on culture substrata. (A) Neurons plated on 10 ^g/ml tenascin-C have short axons and highly adherent growth cones with many filopodia (arrow). (B) Neurons plated on 250 ^g/ml poly-D-lysine have short axons and highly adherent growth cones with broad lamellipodial (arrow). (C) Neurons plated on 10 ^g/ml fibronectin extend moderate-length, unbranched axons with many filopodia along their length (arrowheads) and small, highly motile growth cones (arrow). (D) Neurons plated on 10 ^g/ml laminin extend very long, smooth axons that tend to be adherent only at their small, highly motile terminal growth cones (arrow). Scale: 10 ^m for A-C and 20 ^m for D.

Xenopus growth cones cultured on laminin, fibronectin, tenascin-C, and poly-lysine exhibit unique morphologies, and there is a marked difference in both the rate and the extent of axonal outgrowth (Fig. 5). In particular, spinal neurons extend short neurites on highly adhesive substrata such as tenascin-C, polylysine, and tissue culture plastic (on the order of 50-100 ym at 24 h in culture under control conditions) exhibiting highly filopodial and lamellipodial morphologies on tenascin and polylysine, respectively. In contrast, on laminin and other growth-permissive substrata, neuronal growth cones are highly motile and exhibit more balanced filopodial and lamellipodial morphologies, resulting in increased neurite lengths of up to 1 mm at 24 h in culture.

Growth-permissive substrata such as laminin and, to a lesser extent, fibronectin are suitable for neural tube explant cultures, as a large number of neurites project from the explant and form an extensive halo of radiating neurites. However, highly adhesive or inhibitory substrata, such as those mentioned earlier, should not be used for explant cultures. On surfaces that tend to be less growth permissive, explants rarely adhere and those that do stick often do not extend neurites out of the explant, probably due to the explant providing a more permissive environment for growth. Therefore, when working with less growth permissive substrata, mixed or enriched dissociated neuronal cultures should be used; however, even here increased axon fasciculation and growth upon nonneuronal cells are expected.

IV. Live Cell Imaging and Manipulations

One of the biggest advantages of working with Xenopus neurons in culture is the ability to examine cells expressing a particular mutant or marker protein less than 36 h after embryo fertilization with live cell imaging. This benefit has been advanced further with the explosion of fluorescent fusion protein technologies. As is clear from many studies, viewing the dynamic distribution of proteins in live cells can be much more informative and accurate as compared to viewing the same proteins distributed in fixed, static cells. With different color protein variants, it is now possible to view the dynamic distribution of several proteins simultaneously in living cells. The color variants of fluorescent proteins have also greatly expanded the utility of fusion proteins by taking advantage of the overlaps of fluorophore excitation and emission spectra and close intra- and intermolecular interactions of fusion proteins. Based on the properties of FRET, new expressable proteins are being designed daily that not only report the location of proteins in live cells, but also report protein function or some aspect of cell physiology (Pollok and Heim, 2000). This type of work is particularly useful when studying motile cells, as protein distribution and function, as well as cell physiology, can each be correlated with cellular behavior.

A. Imaging Chambers

Live cell imaging studies often require the rapid exchange of culture solutions. The perfusion of solutions containing pharmacological agents, function blocking antibodies, vital dyes, or fixatives allows the examination of behavioral and physiological changes during and after treatment. In addition, combining rapid solution changes with high temporal and spatial resolution imaging allows for time-resolved measurementzs of intracellular ion and fluorescent protein dynamics. To this end, we have developed a technique for mounting cells plated on loose glass coverslips onto glass slides that allows for the fast perfusion of culture medium in combination with confocal imaging on an upright microscope. Our technique results in a culture chamber volume ^120 yl, which allows for a complete solution exchange in less than 5 s and the use of limited quantities of reagents. Using thinner spacers or small openings will create even lower volume chambers. These chambers are useful for live cell imaging as well as rapid fixation followed by immmunocytochemistry.

1. Materials

Neuronal cultures on loose 22 x 22-mm glass coverslips

300- to 500-ym-thick plastic shims (Small Parts, Inc.)

Precleaned glass microscope slides (Fisher)

High vacuum grease (Dow Corning)

18-mm volume-reducing ring (Thomas Scientific)

2. Procedure a. Preassemble Chamber

Seal precut plastic shim (Fig. 6A) onto a glass slide using a small bead of high vacuum grease applied to all sides of the plastic spacer. Apply a second bead of high vacuum grease to the outer surface of the plastic spacer that will seal with a glass coverslip. Four small spacer squares can be sealed onto the glass slide with high vacuum grease in positions to meet the corners of the glass coverslip as an alternative to a complete precut-shaped plastic spacer.

b. Remove Submerged Coverslip

Glass coverslips containing cultured cells must be lifted from culture dish without exposing cells to air (Fig. 6B), which for Xenopus cells leads to an immediate disruption of membranes and cell death. To remove a submerged coverslip, we use an 18-mm glass ring with two spots of vacuum grease placed at either side to seal onto a coverslip surrounding cultured cells. Next, use a forceps to carefully pull the glass ring with the attached coverslip from the culture dish and place it onto a clean glass ring, which functions as a mounting stage.

c. Seal Chamber

Remove the glass ring from the loose coverslip, being careful not to tip the culture as medium will flow off and expose cells to air. Quickly place the aligned greased side of the preassembled glass slide with a spacer onto the glass coverslip. If the slide is tilted slightly when sealed onto the coverslip, air bubbles will be forced out of one open end. Note that the two remnants of grease on the glass coverslip should be placed at the sides rather than the ends of the chamber to prevent interference with medium flow. With the coverslip sealed in place, the

Fig. 6 Schematic diagram of perfusion chamber assembly. (A) A spacer cut from a 400-^m-thick plastic shim with the dimensions indicated yields a chamber volume of 120 ^l when sealed with a coverslip. First attach the spacer to a glass slide with high vacuum grease and then grease the outer surface of the spacer in preparation to receive a coverslip containing cultured cells. (B) Use a greased glass volume-reducing ring to remove the submerged coverslip. (C) Remove the volume-reducing ring from the coverslip and seal quickly with a glass slide, taking care to avoid introducing air bubbles. (D) Turn the slide right side up and press on all sides to tighten the seal. Mount the slide on a microscope stage using slide clips on coverslip corners (asterisks) to prevent coverslip movement during perfusion.

Fig. 6 Schematic diagram of perfusion chamber assembly. (A) A spacer cut from a 400-^m-thick plastic shim with the dimensions indicated yields a chamber volume of 120 ^l when sealed with a coverslip. First attach the spacer to a glass slide with high vacuum grease and then grease the outer surface of the spacer in preparation to receive a coverslip containing cultured cells. (B) Use a greased glass volume-reducing ring to remove the submerged coverslip. (C) Remove the volume-reducing ring from the coverslip and seal quickly with a glass slide, taking care to avoid introducing air bubbles. (D) Turn the slide right side up and press on all sides to tighten the seal. Mount the slide on a microscope stage using slide clips on coverslip corners (asterisks) to prevent coverslip movement during perfusion.

chamber may now be inverted, revealing the inlet and outlet ports on either side. To assure a complete seal, press all sides of the coverslip down with forceps. If small square spacers were used instead of complete chamber mold, all sides of coverslip, except inlet and output ports, must be sealed with vacuum grease. Test the chamber by adding medium to the inlet side and suctioning from the output side, taking care not to suction the chamber dry.

d. Mount Chamber on Stage

These chambers are designed for use with an upright microscope using oil, water, or air objectives; however, they could also be used with an inverted microscope using long working distance air or water objectives. Place the slide onto the microscope stage and clamp down all four corners of the glass coverslip using slide clamps. It is important to clamp directly onto the glass coverslip rather than the glass slide, as this will prevent movement of the coverslip during perfusion. We use specially designed slide clamps that precisely meet the four corners of the mounted glass slide. Finally, position a hypodermic needle or fine tubing with constant suction near the edge of coverslip within the output port. A properly positioned suction tubing will not suck a chamber dry when input stops, but will draw off fluid rapidly when flow begins.

B. Live Cell Imaging

For most live cell imaging we prefer to use a confocal microscope for the increased clarity, speed, sensitivity, and versatility offered by these systems. We use a three-channel Olympus Fluoview 500 laser-scanning confocal equipped with an argon ion gas laser (488 nm), a green helium neon laser (543 nm), and a red helium neon laser (633 nm). This laser combination allows us to excite GFP and RFP, as well as many commercially available fluorophores. In addition to scanning lasers, our system is equipped with a 100-W mercury light source focused along a separate light path for simultaneous imaging and UV uncaging of photosensitive compounds (see later).

The conditions used to collect confocal images vary depending on what is being examined. In general, when imaging fluorescence signals from vital dyes, such as Fluo-4 or CM DiI,we are less concerned about spatial resolution and more concerned about the viability of living cells, especially if we are collecting many images over a short period of time (high-temporal resolution). However, when imaging fluorescent fusion proteins with the intent of analyzing dynamic protein localization, high-spatial resolution is essential (^200 nm/pixel). To maximize the detected signal while minimizing laser power to the sample, we open the confocal aperture beyond the optimal Airy disc size. The photo multiplier tube (PMT) gain is also set high (50-75% of maximum), allowing laser power to be left below 5% of maximum. These values will vary for different confocal systems. For hightemporal, low-spatial resolution imaging, we also scan fast (2 ms/line) and reduce the scan region in the vertical dimension to fewer than 40 horizontal lines. Under these conditions, images can be captured at frequencies up to 8-12 Hz. We collect images at this rate to detect brief calcium signals that often persist for less than 200 ms. Even higher frequency imaging can be achieved by performing one-dimensional line scans along the length of individual filopodia. In this mode, a 2-ms temporal resolution is attained at the sacrifice of spatial resolution and signal to noise. In contrast, fluorescent proteins are typically imaged at a higher spatial and lower temporal resolution using a high NA oil objective, a near optimal pinhole diameter, and a lower PMT gain to increase the signal-to-noise level. Slower scan speeds (5 ms/line) are also used when imaging fluorescent proteins.

C. Rapid Fixation and Staining

Rapid fixation followed by immunocytochemical labeling after live cell imaging allows for the correlation of cellular behaviors, protein localization, or physiological activity with receptors and other key signaling molecules associated with cell motility. This is done by directly perfusing fixative through the imaging chamber on the microscope stage immediately following acquisition of a

Was this article helpful?

0 0

Post a comment