Live imaging of cytoskeletal dynamics in embryonic Xenopus laevis growth cones and neural crest cells

Abstract

The cytoskeleton is a dynamic, fundamental network that not only provides mechanical strength to maintain a cell’s shape, but also controls critical events like cell division, polarity, and movement. Thus, how the cytoskeleton is organized and dynamically regulated is critical to our understanding of countless processes. Live imaging of fluorophore tagged cytoskeletal proteins allows us to monitor the dynamic nature of cytoskeleton components in embryonic cells. Here, we describe a protocol to monitor and analyze cytoskeletal dynamics in primary embryonic neuronal growth cones and neural crest cells obtained from Xenopus laevis embryos.

MATERIALS

Reagents

Crystalline collagenase (for neural tube dissections only)

Dissolved at 2 mg/mL in Steinberg’s media

Culture dish coating reagents^*

Culture reagents depend on the type of culture. Here, we use Poly-L-lysine (PLL) and laminin-coated plates for neural tube, and gelatin and fibronectin-coated plates for neural crest cells.

Culture media base^*

mRNA encoding fluorophore tagged protein of interest

Capped mRNAs are synthesized in vitro using the mMessage mMachine kit (Invitrogen).

Plating Culture Media for Neural Tube Explants/Neural Crest Cells^*

Ringer’s Solution^*

Steinberg’s Solution^*

Xenopus laevis embryos, staged appropriately for neural crest (stage 18) or neural tube (stage 20–23) isolation.

Equipment

Dissecting stereoscope

Dumont #5 forceps or equivalent

For neural tube dissections:

Electrolytically-sharpened tungsten needles

Agarose or sylgard-coated petri dish

For neural crest cells dissections:

Eyelash knife (or insect pins)

Plasticine clay-coated petri dish

Culture dish/imaging chamber (eg. as 35 mm MatTek glass bottom (No. 1.5) culture dish)

Imaging chambers will be coated with appropriate coating reagents depending on the type of culture (see “Culture dish coating protocol” in “Recipes” section)

Imaging microscope

Spinning disk confocal microscope (SDCM), total internal reflection fluorescence (TIRF) microscope, or widefield epifluorescence scope, equipped with an appropriate objective. Our SDCM rig uses a 63× Plan Apo 1.4 NA objective, while our TIRF uses a 60x, 1.49 NA TIRF objective. For detection, a growing number of modern CCD and scMOS cameras will be suitable; ensure that the chip size, sensitivity, and speed of your camera are appropriate for your needs. Our SDCM uses an Orca Flash 4 CCD; our TIRF imaging is typically acquired with an Andor ELYRA scMOS.

Software - Image J

To process and analyze images collected, we routinely use ImageJ, which is an open-source image processing program that runs on various operating systems.

Software - plusTipTracker

While other single particle tracking softwares are currently available for analyzing cytoskeletal dynamics, we use a MatLab based open-source software package developed by Danuser lab [3]. A detailed protocol for how to use the program to measure microtubule dynamics in growth cones is described by Stout et. al. [4]

METHOD

Fluorescent proteins used to visualize the cytoskeleton

1. In order to quantify changes in microtubule dynamics, we typically inject in vitro transcribed mRNA of fluorophore tagged EB1 or MACF-43 (a truncated protein that contains a minimal EB1-binding domain [5]), at 300 pg/embryo, or 75–100 pg/embryo, respectively (See Table 1). These mRNAs can be injected alongside overexpression or knockdown strategies for your protein of interest.

   Note: Injections are often made at 2–4 cell stages. However, tissue-targeted injections can be done based on fate maps provided on Xenbase.

2. Changes in actin filament dynamics can be monitored with fluorophore-tagged actin binding domains such as LifeAct, Utrophin, F-Tractin, or actin monomer. However, it has been reported that these markers may impact actin dynamics and architecture in a dose-dependent manner [6–8]. Therefore, care should be taken to use the probe that will least-impact the dynamic behavior of interest. We traditionally start with LifeAct or Utrophin mRNA concentrations at 100–300pg mRNA/embryo.

3. For cell–extracellular matrix adhesion dynamics, focal adhesions can be labeled with 50–75 pg/embryo of fluorophore-tagged Paxillin, Zyxin, or PAK2, in addition to many others [9, 10]. Note that focal adhesion proteins have notoriously-high cytoplasmic fluorescence, and TIRF microscopy greatly facilitates crisp adhesion segmentation in later analyses. Advanced imaging techniques such as fluorescence recovery after photobleaching (FRAP) can also be utilized to monitor the adhesion dynamics involved in cell migration and adhesion [11].

Table 1.

Useful cytoskeletal markers to visualize MTs, actin filaments, or focal adhesions.

Marker	Labels	Purpose	Concentration range per embryo	Imaging technique	Analysis software
EB1–3 [12]	MT plus-end	MT growth dynamics	100pg – 300pg mRNA	Spinning disk confocal microscopy (SDCM)	plusTipTracker [3, 4]
MACF-43 [5]	MT plus-end	MT growth dynamics	100pg – 300pg mRNA	SDCM	plusTipTracker [3, 4]
Fluorophore tagged tubulin	MTs	Uniform labelling to visualize MTs	900pg mRNA	SDCM	Fiji (ImageJ)[13]
		Speckle labelling to assess MT flow rates	300pg mRNA	Total internal reflection microscopy (TIRF)	QFSM software [14, 15]
Actin monomer (G-actin)	Actin	Speckle labelling to assess actin dynamics	300pg mRNA	TIRF	QFSM software [14, 15]
LifeAct -GFP [16] Utrophin [17] F-Tractin [18]	F-actin	Actin dynamics	LifeAct/Utrophin may affect dynamics dependent on the concentration. Optimization may be required. 60 −300 pg can be used to begin.	SDCM TIRF	Fiji (ImageJ)[13]
Paxillin [19], PAK2 [20], Focal Adhesion kinase [21], Zyxin, Vinculin, Talin	Focal adhesion	Focal adhesion dynamics	50–75 pg DNA, 250 – 500 pg mRNA	TIRF	FAAS [22]

Preparation of Neural Tube and Neural Crest Cell Cultures

Neural Tube Culture

Isolation and preparation of neural tube cultures have also been previously described in Lowery et al, 2012 [1].

1. Prior to performing the dissections, prepare the culture dishes (see Culture dish coating protocol in Recipes below). Fill the culture dishes with the neural tube explant culture media (see Recipes below). Additionally, due to variability in expression of injected mRNA, embryos may exhibit a mosaic of fluorescence. Thus, before dissections, it is useful to screen embryos with a fluorescence dissecting scope and identify which express fluorescence in the neural tube. Place fluorescent embryos at stage 20–23 into an agarose-coated plastic dish filled with Steinberg’s media. Dissection of embryos beyond stage 23 is possible, but it is more challenging as tissues adhere more tightly to each other and thus require longer treatment with collagenase.

2. Isolate neural tubes. Under a dissecting scope, remove the vitelline membrane with fine forceps, and then isolate the entire dorsal portion of the embryo, by making a series of incisions. While holding the embryo in place with one forceps, use the second to make an incision on the side of the embryo to expose the hollow interior. Then, use both forceps to pinch along the tissue between the dorsal and ventral halves of the embryo, thereby cutting the embryo in half to isolate the dorsal portion containing the neural tube. Place the dorsal explant in a small tube or dish with 2 mg/ml collagenase in Steinberg’s media for 15–20 min on a rotator, to loosen tissues, and then pipette dorsal explant into a plastic agarose-coated dish with fresh Steinberg’s media. Using a pair of forceps, gently dissect the neural tube from the dorsal epidermis and the ventral notochord. Slide the tip of the forceps between the epidermis and underlying tissue and slowly pull back the epidermis, revealing the neural tube beneath. Then, use one forceps to hold the tissue and another to slide the tip between the neural tube and notochord. Finally, use forceps to remove the somites on either side of the tube.

3. Transfer the neural tube to an agarose-coated dish filled with culture media. After collection of several neural tubes, transect each of them into approximately 20 thin slice explants using electrolytically-sharpened tungsten wires. Transfer explants to the prepared culture dishes, spreading out the explants evenly in rows.

4. After plating, do not move the dishes because this would disturb the attaching cells. Allow explants to adhere for 12–18 hours, ideally around 20–22 °C. We typically leave the dish of cells on the bench at room temperature overnight. Axons will extend radially, allowing various live imaging of growth cone behavior and cytoskeletal dynamics. Xenopus laevis neural explants typically send out neurites in a very robust manner by 24 hr after plating on the laminin/poly-lysine substrate if conditions are appropriate. With this substrate, growth cones are highly motile and can achieve axon lengths of up to 1 mm, extending in all directions outward from the explant, although typical lengths are 100 μm or more.

5. Neurites and growth cones can be observed and imaged at room temperature 12–24 hr after plating. In general, expression of a RNA or plasmid product will result in variable expression between cells, and thus there may be a range of fluorescence expression levels between growth cones. We have observed RNA expression of fluorescent proteins to persist beyond 48 hr after plating, although this depends on the particular construct.

Cranial Neural Crest Cell Culture

A very helpful and thorough guide to neural crest isolation is given in Milet et al 2014 [2], and we advise that this be heavily relied upon for tissue identification and dissection technique. We offer minor modifications here.

1. As with the neural tube dissection protocol, prior to performing the dissections, prepare the culture dishes (see Culture dish coating protocol in Recipes below). Fill the culture dishes with the neural crest cell culture media (see Recipes below). Sort embryos to identify those with fluorescence in the cranial neural crest region at stage 18.

2. Strip vitelline membranes of stage 18 embryos and embed them gently in plasticine-clay in Steinberg’s media with the anterior dorsal regions exposed. Neural crest cells emerge from the tissue just along the anterior neural fold, which is raised slightly from the surrounding tissue.

3. Remove the skin above the neural crest using an eyelash knife. Apply gentle pressure along the edge of the neural fold to allow neural crest (2–3 cell layers) to separate, and lift the explant with a lateral/ventral flicking motion.

4. Rinse explants several times by gently pipetting into fresh culture media, and then transfer them into fibronectin-coated imaging chambers. (Rinses and culture are performed in a modified DFA solution, prepared as previously described [2]).

5. Allow explants to adhere (15–30 minutes). Tissue will begin collective cell migration within an hour of plating, and subsequently delaminate to begin single-cell movement after eight or more hours; allowing for various measurements of cytoskeletal dynamics in both individual and collectively- migrating cells.

Imaging Cytoskeletal Dynamics

1. Once cells are adhered, position the culture plate on microscope stage. Live imaging of intracellular cytoskeletal dynamics in either neuronal growth cones or neural crest cells can be performed using spinning disk confocal microscopy, TIRF microscopy (for dynamics at or near the cell membrane), or even widefield fluorescence microscopy (for microtubule plus-end dynamics). Cytoskeletal dynamics will generally require a relatively high-magnification objective (60x or 100x); we use a 63× Plan Apo 1.4 NA lens for SDCM.

2. Perform time-lapse imaging. To analyze growth and pause events of microtubule plus-ends, which exhibit very rapid dynamics, images should be captured at least every 2 seconds for a duration of 1–2 min. Given the rapid dynamics of MT plus-ends, spinning-disk confocal microscopy allows the easiest acquisition of large ROIs (whole cells). To visualize actin retrograde flow, using low-density labeling of fluorescent actin monomers (followed by quantitative fluorescent microscopy or kymograph analysis), an initial test with 1–2 second imaging intervals will provide a good starting point, but flow rates will vary greatly based on structure stability. Focal adhesion dynamics of neural crest cells can be captured every 1–2 minutes, for manual tracking of general size and turnover, but more sensitive tracking measures will often require much greater sampling rates, again within the order 5–10 seconds, for extended imaging periods (20–30 minutes or longer). General surveys of cell morphology, perhaps to assess cell polarization or filopodial density/number following a genetic perturbation, can perhaps be reduced to every few minutes or, collected at only select time points over a number of hours.

   Note: It is incredibly important that you optimize the laser power and exposure time of your acquisitions to minimize light toxicity. Using short exposure times and low laser power with reasonably higher gain will help reduce phototoxicity and minimize photobleaching.

3. Two fluorophore-labeled structures in separate imaging channels will not be captured simultaneously in typical single-camera/filter systems. Thus, if comparing dynamics of two labeled proteins, it will be necessary to take into account the time that passes while one is acquired and the system switches to capture the other. This can be quite minimal with some setups (i.e., triggered acquisition using a small ROI) and dramatically delayed with others (i.e., averaged line scans on a large ROI). Make sure your microscope can be optimized for rapid multichannel acquisition if this is of utmost importance.

4. If cytoskeletal drugs are to be applied while imaging, treatment should be administered using a perfusion chamber (many pharmacological agents are not easily soluble in aqueous medias, and precipitation can occur if a concentrated solution is administered into a corner of the dish). These can be constructed in the bottom of typical inverted imaging chambers using vacuum grease and an additional coverslip.

Analyzing Cytoskeletal Dynamics

To quantify parameters of microtubule plus-end dynamics, we use plusTipTracker, a Matlab-based open source software package that automatically detects, tracks, and analyzes time-lapse movies of fluorophore tagged cytoskeletal proteins. For a detailed explanation of how to manage files to perform dynamics analysis refer to Stout et al., 2014 [4]. It should be noted that a currently supported version of this pipeline has been revamped to analyze multiple types of dynamic particles, and is now available under the name u-Track [3]. Actin network flow can be assessed using Quantitative Fluorescent Speckle Microscopy technique and software (QFSM) [15]. Other pipelines exist for focal adhesion dynamics (FAAS) [22], automated filopodia tracking (FiloQuant) [23], and single particle tracking (TrackMate) [24]. Additionally, customized analysis methods can be built and streamlined in Fiji, Imaris, or Matlab, as preferred.

TROUBLESHOOTING

Problem:

Cultured explants do not express the fluorophore tagged protein.

Solution:

One reason for low or absence of expression is that the needle tip could be clogged depending on the injected material. For example, morpholinos are highly prone to clogging needles. To ensure a successful injection, occasionally check if the solution is coming out from the needle. Another reason is that the amount of injected material may not be enough. Optimal concentration will vary based on transcript quality and molecular weight and should be titrated accordingly. Generally, the lowest possible concentration of a fluorophore-labeled protein is the most favorable one; low-level expression can achieve a much more favorable signal-to-noise ratio, and minimize the artifacts caused by exogenous protein expression. Heavily overexpressed plus-end tracking proteins can be especially hard to deal with, as the excess fluorescent protein will coat the entire microtubule lattice instead of selectively-labeling distal microtubule ends, rendering automated plus-end segmentation and tracking much more difficult.

Problem:

No neurite extension or cell adhesion is observed.

Solution:

Neurites might fail to grow if the cultured explants did not adhere well to the culture plate. To ensure cell adherence, use freshly made culture media and PLL and laminin coated dishes for neural tube explants, and Fibronectin coated dishes for cranial neural crest. Try not to disturb culture dish after explants are plated. Finally, it is possible that embryos from which the explants were obtained might be unhealthy, although if embryos can develop to the neural tube stage, this is less likely to be the issue. However, it is best to keep some whole intact embryos cultured in 0.1X MMR along with dissections to ensure that continued development occurs.

Problem:

Growth cones or cells die during imaging.

Solution:

Light toxicity is the major cause of the cell death. To avoid cell death, image samples at high gain rather than high laser power and keep exposure time short.

DISCUSSION

Xenopus laevis has been a powerful embryonic tool, not only for addressing questions regarding vertebrate development, but also for deciphering key cellular events. One advantage of Xenopus laevis as a model is that Xenopus embryos can tolerate extensive manipulation, and exogenous materials can be easily introduced via microinjections, such as fluorophore tagged cytoskeletal associated proteins. Compared to other systems, cell culturing methods are facile and do not require expensive culture equipment. Cultures can be generated in large quantities and maintained on the bench at room temperature.

The protocol described here takes advantage of this versatile model organism to examine the dynamics of the cytoskeletal components. Measures of cytoskeletal dynamics can be taken from any tissue of interest. To gather data from heterogeneous neuronal cultures, we use neural tube explants dissected from stage 20–23 embryos. Retinal cultures from embryos at stage 24 or beyond can also be used if a more homogenous neuronal population is desired. Cranial neural crest, a multipotent motile cell with well-established migration routes and fate determination cues, can be a valuable model for identifying effectors of collective cell migration, craniofacial development, and cancer metastasis.

The scope of the high-resolution live imaging of cytoskeletal dynamics in axonal growth cones and in cells can be expanded to study different aspects of the cytoskeletal behavior. By mimicking various physiological conditions, one can study how microtubule or actin dynamicity would change within the growth cone or cell in response to guidance signals introduced in culture media. Multiple cues can be introduced to the cells to test the spatiotemporal changes in cytoskeletal dynamics, which could mimic the pathfinding behavior of the growth cone in its native environment.

RECIPES

Plating Culture Media for Neural Tube Explants

Reagent	Amount to add	Final Concentration
Fetal Bovine Serum (Gibco, 10438018)	20 μl	1%
Pen / Strep (Sigma, P4333)	20 μl	1% (or 50 μg/ml)
BDNF (100μg/ml, Sigma, B3795)	0.5 μl	25ng/ml
NT3 (25μg/ml, Sigma, SRP3128)	2 μl	25ng/ml
Culture media base	up to 2 ml (per 35mm plate)

Prepare plating culture media fresh and keep at 4°C. While our protocol uses this media, it should be noted that this is an enriched media common for older Xenopus retinal ganglion cultures, where neurons have reduced energy reserved. Young spinal cord cultures will survive and grow for more than 24 hr in pure Ringers’s Media without additional growth factors, and this is indeed one of the benefits of using this system. Optional: add NT3 and BDNF (to final concentrations of 25 ng/ml each) to culture media to increase axon outgrowth. Axon outgrowth is greatly enhanced by addition of BDNF and NT3, but these may be omitted for guidance assays if cues will be added later.

Culture Media Base

Reagent	Amount to add	Final Concentration
L-15 (Sigma, L-1518)	98 ml	49%
1X Ringer’s Solution	100 ml	50%

Sterilize by vacuum filtering. Add 2 ml of Antibiotic -Antimycotic (100X, GIBCO, 15240062). Keep at 4°C.

Ringer’s Solution (10X)

Reagent	Amount to add	Final Concentration
NaCl	33.6 g	1.15 M
KCl	0.93 g	25 mM
CaCl2	1.11 g	20 mM
EDTA (0.5M pH 8.0)	5 ml	5 mM

Adjust the pH to 7.4 and complete the volume up to 500 ml with distilled water Sterilize by autoclaving.

Plating Culture Media for Neural Crest Cells

Modified Danilchik’s for Amy (DFA)

Reagent	Amount to add	Final Concentration
NaCl	1.55 g	53 mM
Na2CO3	.265 g	5 mM
Potassium Gluconate	.527 g	4.5 mM
Sodium Gluconate	3.82 g	35 mM
MgSO4	0.060 g	1 mM
CaCl2	0.055 g	1 mM

Add reagents to 400mL of DI water. Adjust pH to 8.0 with Bicine, and fill to final volume (500mL) with DI water. Sterilize by filtering. Can be prepared in bulk & frozen in small aliquots. Add 50 μg/mL Gentamycin Sulfate and 1 μg/mL Bovine Serum Albumin to culture media prior to use.

Steinberg’s Media

Reagent	Amount to add	Final Concentration
NaCl	3.39 g	58 mM
KCl	0.05 g	0.67 mM
Ca(NO3)2	0.07 g	0.44 mM
MgSO4	0.16 g	1.3 mM
Tris (1M, pH 7.8)	2.3 ml	4.6 mM

Add 250 ml of DI water and adjust the pH to 7.8 and fill to final volume (500ml) with DI water. Sterilize by autoclaving.

Culture dish coating protocol

Making neural tube culture plate

1. Coat 35mm MatTek glass bottom (No. 1.5) culture dish with 500 µl of 100µg/mL Poly-L-Lysine (Sigma P 4832).

2. Rinse off PLL by washing the dish with 1X sterile PBS three times.

3. Add 500 µl of 10 µg/ml laminin (Sigma, L 2020) into the center of the plate and incubate at 37°C for 1h.

4. Rinse off laminin by washing the dish with 1X sterile PBS three times and use immediately.

5. Coat coverslips with enough solution of 200 μg/ml poly-lysine in PBS to pool over the surface of the coverslip, of either Mattek or Lab-Tek culture dishes, and incubate for one hour (alternatively, one could use glass coverslips sitting loose in a Petri dish, for later placement on glass slides for imaging and immunocytochemistry). Aspirate and wash with an excess of PBS 3 times, and let dry. Then coat dishes with 10 μg/ml laminin in PBS (we usually use 500 μl per coverslip, enough to pool over the surface) for one hour at 37 degrees. Aspirate laminin solution and wash 3 times with an excess of PBS, being careful not to let laminin-coated surfaces become exposed to the air interface.

Making neural crest cell culture plate

1. Heat up 2mg/mL Gelatin Stock (Sigma, G1890) in the microwave allowing it to boil for 5–10 sec then allow it to cool down for 5 min.

2. Add 500 µl of cooled gelatin into the center of 35mm MatTek glass bottom (No. 1.5) culture dish.

3. Rotate gently for 25 min at room temperature.

4. Discard the gelatin and rinse the dish with 1X sterile PBS several times.

5. Add 500 µl of fibronectin (Sigma, F1141) at 20 µg/ml and incubate dishes at 4°C overnight.

6. Next day remove fibronectin, rinse the plate with 1X sterile PBS several times and use to culture cells.