Antibody Uptake Assay for Tracking Notch/Delta Endocytosis During the Asymmetric Division of Zebrafish Radial Glia Progenitors

Summary

This work develops an antibody uptake assay for imaging intra-lineage Notch/DeltaD signaling in dividing radial glia progenitors of the embryonic zebrafish brain.

Abstract

Asymmetric cell division (ACD), which produces two daughter cells of different fates, is fundamental for generating cellular diversity. In the developing organs of both invertebrates and vertebrates, asymmetrically dividing progenitors generates a Notch^hi self-renewing and a Notch^lo differentiating daughter. In the embryonic zebrafish brain, radial glia progenitors (RGPs)-the principal vertebrate neural stem cells-mostly undergo ACD to give birth to one RGP and one differentiating neuron. The optical clarity and easy accessibility of zebrafish embryos make them ideal for in vivo time-lapse imaging to directly visualize how and when the asymmetry of Notch signaling is established during ACD. Recent studies have shown that dynamic endocytosis of the Notch ligand DeltaD plays a crucial role in cell fate determination during ACD, and the process is regulated by the evolutionarily conserved polarity regulator Par-3 (also known as Pard3) and the dynein motor complex. To visualize the in vivo trafficking patterns of Notch signaling endosomes in mitotic RGPs, we have developed this antibody uptake assay. Using the assay, we have uncovered the dynamicity of DeltaD-containing endosomes during RGP division.

Introduction

Notch signaling controls cell fate decision and patterning during development in metazoans¹, and recent studies have shown that Notch signaling in stem cell division mainly depends on endocytic trafficking^2,3. Endocytosed Notch/Delta can activate Notch signaling in the nucleus and enhance the transcription of Notch target genes^4,5,6. Directional Notch/Delta endosomal trafficking was first observed in Drosophila sensory organ precursor (SOP) cells during its asymmetric cell division (ACD), resulting in a higher Notch signaling activity in pIIa than in pIIb^7,8. Antibody uptake assays with anti-Delta and anti-Notch antibodies have been applied to monitor the endocytic process in mitotic SOP cells. Notch/DeltaD endosomes move along with a kinesin motor protein to the central spindle during cytokinesis, and are asymmetrically translocated into the pIIa cell due to the antiparallel array of the asymmetric central spindle at the last moment of cell division^3,8. These studies have shed light on the molecular mechanisms regulating asymmetric division in Drosophila SOP cells, but it is unclear whether similar endocytic processes occur in vertebrate radial glia progenitors (RGPs).

Moreover, the molecular mechanisms that regulate asymmetric Notch/DeltaD signaling during vertebrate RGP division are not well understood. In zebrafish, it has been reported that the interaction of Notch and Delta facilitates the endocytosis of the DeltaD ligand⁹. It is unknown whether DeltaD endocytosis can impact the cell fate choice of daughter cells in the developing vertebrate brain. Recent studies show injecting fluorescently conjugated anti-DeltaD antibodies into the neural tube could label Sara endosomes specifically in neuroepithelial cells, and anti-DeltaD containing Sara endosomes preferentially segregate into proliferating daughter cells¹⁰. It has been suggested that Notch signaling from the endosomes could regulate daughter cell fate. Previous results have shown that most zebrafish RGP cells in the developing forebrain undergo ACD, and the daughter cell fate determination is dependent on intralineage Notch/DeltaD signaling¹¹. In order to elucidate the nature of the intralineage Notch/DeltaD signaling in zebrafish RGPs, we have developed the anti-DeltaD antibody uptake assay in the zebrafish developing brain. Using this protocol, we have successfully performed live labeling and imaging of DeltaD endocytic trafficking in mitotic RGPs.

The fluorescently labeled anti-DeltaD-antibody is efficiently internalized into the RGPs along the forebrain ventricle. It has greatly facilitated the discovery of directional trafficking of DeltaD endosomes in the asymmetrically dividing RGPs^12,13. Compared to previous antibody uptake protocols developed for Drosophila notum cultures and the zebrafish spinal cord¹⁰, this protocol has achieved long-lasting and highly efficient anti-DeltaD labeling in the brain ventricle cell layer, specifically with less than 10 nL of microinjected antibody mixture. The hindbrain ventricle injection is very convenient for applying the antibody uptake assay in the developing brain, as the hindbrain ventricle is well expanded in zebrafish embryos and filled up with cerebrospinal fluid at early development stage¹⁴. By injecting the antibody mixture into the hindbrain ventricle without injuring any crucial developing tissues, the protocol has minimized possible damage to the imaging zone in the forebrain as much as possible. The reduced dosage of the injected primary antibody has also avoided potential side-effects of interfering with endogenous Delta-Notch signaling in vivo. This protocol can be easily combined with other pharmacological or genetic perturbations, utilized at different developmental stages, and possibly adapted to the adult brain as well as human pluripotent stem cell-derived 2D/3D brain organoids. Taken together, the protocol has made it possible to understand how and when Notch signaling asymmetry is established during ACD. The main challenge for successful implementation of this protocol is to achieve precise delivery of appropriate concentrations of the antibody based on specific experimental conditions.

Protocol

We have used the AB wild type line and transgenic line Tg [ef1a:Myr-Tdtomato] for the study. All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of California, San Francisco, USA (Approval Number: AN179000).

1. Preparation of zebrafish embryos

1. Set up fish crossing tanks in the afternoon before 5:00 p.m., with one female wild-type fish and one male Tg [ef1a:Myr-Tdtomato] fish by using dividers to separate them in each tank.
2. Remove the dividers before 11:00 a.m. from all crossing tanks the next morning. Keep quiet and do not disturb the fish while they are mating. Spawning typically occurs within 30 min after removing the dividers. The fertilized eggs remain on the bottom of the tank.
   1. Collect fertilized eggs from the tanks with a mesh filter. Transfer the eggs by washing them off into a Petri dish full of egg water and examine the embryos under a dissecting microscope at a 10x to 20x magnification.
3. Transfer the fertilized embryos to a clean Petri dish containing approximately 20 mL of embryo medium (100 mL of 1,000x stock solution contains 29.4 g of NaCl, 1.27 g of KCl, 4.85 g of CaCl₂.2H₂O, and 8.13 g of MgSO₄.7H₂O) with large-bore glass Pasteur pipettes. Keep the embryos at 28 °C. Keep 50 fertilized embryos per Petri dish with 30 mL of embryo medium at 28.5 °C.
   ​NOTE: One pair of fish normally produces 100-300 embryos, and the fertilization and survival rate of healthy embryos is over 95% for each mating.
4. The next morning, take the embryos out of the incubator at 8:00 a.m., when the embryos have reached the developmental stage of ~18-20 h post-fertilization (hpf). Observe them under an epifluorescence microscope at a 20x magnification, using white light at first. At that time, the zebrafish embryos are at the 20 somite stage.
   1. Discard dead embryos that are cloudy or ruptured under the microscope using a glass pipette.
5. Turn on the fluorescent lamp and choose the RFP filter setting of the microscope. Then, select the embryos with strong red fluorescence and transfer them to a new dish with egg water.
   1. Dechorionate the embryos manually with two fine forceps (the tips of forceps should be sharp and undamaged) under white light (Table of Materials).
   2. Hold the chorion with one pair of forceps and make a tear in the chorion with the other forceps. Open the tear carefully using the forceps, and make it large enough for the embryo to pass through by gently pushing the embryo with the tips of the other forceps.
6. Transfer the dechorionated embryos to a new Petri dish with fresh embryonic medium for a quick rinse before microinjection.

2. Preparation of microinjection

1. Use the capillaries (1.2 mm OD, 0.9 mm ID, with filament) for pulling fine injection needles on a puller. Design and optimize the pulling program according to the handbook.
2. Open the tip of the needle with forceps under a stereo dissection microscope. Make the diameter of the tip around 10 µm, and the taper angle around 30°.
3. Conjugate the mouse monoclonal anti-Dld antibody with the anti-Mouse-IgG-Atto647N before injection.
   1. For each injection experiment, mix 0.5 µL of the anti-Dld antibody (0.5 mg/mL) with 2 µL of the anti-Mouse-IgG-Atto647N antibody (1 mg/mL) by pipetting 5-10 times. Then, incubate at room temperature for at least 30 min (or on ice for 2-3 h).
   2. After incubation, add 2.5 µL of blocking buffer (10 mg/mL mouse IgG) and 0.5 µL of 0.5% phenol red to the antibody mixture, and pipette 10x to mix thoroughly, to block any unconjugated antibodies remaining in the mixture.
      NOTE: For each trial, prepare one extra mixture without the anti-Dld antibody and use it as a control.
4. Prepare 1% low-melting point agarose in the embryo medium. Heat the agarose contained mixture at 70 °C till the mixture turns transparent.
   1. Aliquot the agarose solution in 2 mL microcentrifuge tubes. Keep the aliquots in a heat block at ~40 °C.
5. Rinse the embryo in the tube containing 1% low-melting point agarose for 3 s.
6. Place the embryo on an inverted plastic Petri dish lid together with individual drops of agarose (~30-40 µL) to mount each embryo separately on the lid as shown in Figure 1. Lay the embryos flat laterally in the agarose and keep this position until the agarose has solidified at room temperature.
7. Mount 12 embryos in three rows one by one as above. Cover all the mounted embryos in the agarose with egg medium.

3. Microinjection

1. Put the embedded embryos under the stereomicroscope and cover the agarose with egg water.
2. Set the air pressure injector together with micromanipulators, placing them close to the microscopes as shown in Figure 1. Use the steel gas cylinder containing gaseous nitrogen (N₂) under high pressure as the air resource.
   1. Open the gas valve only after the embryos have been mounted in the agarose. Then, front load the prepared glass needle with 2 µL of antibody mixture on the micromanipulator, as shown in Figure 1, when using the front fill module of the microinjector.
   2. Tune the input pressure to ~80-90 psi, and the injection pressure to ~20 psi.
3. Calibrate the injection volume by using a micrometer under the microscope, as described previously¹⁵. Set the tune time duration to be from 10 ms to 120 ms, according to the size of needle opening. Deliver each pulse of injection by tapping the paddle. Tune the injection volume of each delivery to ~4-5 nL.
4. Poke the tip of the microinjection needle through the dorsal roof plate of the hindbrain posterior to the r0/r1 hinge point, and inject about 10 nL (two or three pulses) of antibody mixture without hitting the brain tissue. Observe the flowing of red fluids in the brain ventricle.
   NOTE: The hindbrain ventricle is posterior to the midbrain hindbrain boundary. The injected phenol red-containing antibody mixture fills up the brain ventricle from the hindbrain to the forebrain immediately by diffusing with cerebrospinal fluid.
5. After injection, remove the tip of the needle from the embryo swiftly by rotating the knob of the micromanipulator. For a successful injection, the red dye of the injected mixture remains in the brain ventricle stably without leaking into the surrounded agarose.
6. Move the mounting plate under the microscope to locate another mounted embryo at a suitable position for repeats.
   1. After injecting six to eight embryos, peel the agarose with a microsurgical knife to release the embryos from the embedded agarose. Transfer the microinjected embryos to a fresh dish with 30 mL of embryo medium, and place them at room temperature for the next steps.

4. Mounting and time-lapse live imaging

1. After 30 min, transfer the selected embryos to 10 mL of embryo medium. Add 420 µL of tricaine stock (4 mg/mL) to 10 mL of embryo medium to anesthetize the embryos.
2. To mount the embryos, prepare 0.8% low melting point agarose containing the same concentration of tricaine in the tube. Keep the aliquots in a heat block at ~40 °C.
3. Use a glass pipette to immerse the injected embryos in the warm agarose for 3 s. Then, immediately remove the embryos from the agarose with the same glass pipette and place the embryos on the center of 35 mm glass bottom culture dishes with a drop of agarose from the tube. Place only one embryo per drop on the glass.
4. Orientate the embryos gently with a fiber probe or loading tip to keep the dorsal side of the embryonic brain as close to the glass bottom as possible. Tune the embryo position gently to extend the embryo without curling as the agarose solidifies gradually.
5. Afterward, check the embryo position by flipping the glass bottom dish over. Ensure that the whole dorsal forebrain with the correctly mounted embryos can be seen under the microscope.
6. Add 2-3 mL of 28.5 °C preheated embryonic medium containing tricaine to cover the embryo. Place the dish properly on the temperature-controlled stage of the confocal microscope, as shown in Figure 1. Adjust the temperature of the imaging chamber to be at 28.5 °C. The embryo is now ready for imaging.
7. Perform time-lapse live imaging with a fixed time interval using a 40x water immersion objective.
   1. Use the imaging and microscope control software (µMANAGER: https://micro-manager.org/)¹⁶.
   2. Select the imaging channels of 564 nm and 647 nm for imaging the membrane fluorescence of MyR-Tdtomato transgenic zebrafish and endosomal anti-Dld-Atto647N in the cell (Figure 1).
   3. Set the laser power at 30% for both channels. Use an exposure time per z-plane of 200 ms for each channel.
   4. For each mounted zebrafish embryo, use a scanning z-step of 1 µm for 20 z-planes in total, and a scanning cycle of 100 timeframes.
   5. Set the time interval between each scanning cycle at 20 s and the scanning mode as channel, slice.

Representative Results

In Figure 2A, the embryos injected with Atto647N, without binding with the primary antibody, showed background fluorescence in the brain ventricle. Very few engulfed fluorescent particles can be observed in the cells. The anti-Dld-Atto647N injected zebrafish embryos showed large amounts of internalized fluorescent particles in most cells of the developing forebrain (Figure 2A, right panel). After zooming in to focus on mitotic RGPs, as shown in Figure 2B, we were able to record the dynamic movement of intracellular anti-Dld-Atto647N endosomes during cell division (Video 1). Most mitotic RGPs showed anterior or posterior asymmetric segregation of anti-Dld-Atto647N endosomes into two daughter cells¹². We also noticed that the asymmetry became stabilized toward the end of anaphase, resulting in asymmetric inheritance of Notch signaling endosomes by the daughter cells afterward¹².

Figure 1: Flowchart of experimental steps. Transgenic zebrafish expressing the MyR-Tdtomato reporter are outcrossed with AB wild-type on day 1. On day 2, red fluorescent embryos are selected for microinjection, followed by time-lapsing imaging. The whole experiment on the 2^nd day takes about 8 h. Please click here to view a larger version of this figure.

Figure 2: Time-lapse imaging of anti-Dld-Atto-647N antibody uptake in the developing zebrafish brain. (A) Anti-Dld-Atto-647N uptake after 30 min of microinjection in the forebrain of a 1 day post-fertilization (dpf) embryo. On the left panel is the embryo injected with Atto-647N only. On the right panel is the embryo injected with anti-Dld-Atto-647N. Dash lines indicate the apical layer along the ventricle. The rectangle zone is shown in high magnification in B. Scale bar = 40 µm. Abbreviations: Di = Diencephalon; Tel = Telencephalon; Ve = Ventricle. (B) Time-lapse imaging panels of anti-Dld-Atto-647N dynamics during mitosis. The eight timeframes in the montage covered the entire mitotic cycle, depicting the uptake and segregation of anti-Dld-Atto-647N into two daughter cells of a dividing RGP. Scale bar = 5 µm The cell membrane is outlined. In both A and B, MyR-Tdtomato labeling on the membrane is blue, and anti-Dld-Atto-647N is magenta. Please click here to view a larger version of this figure.

Video 1: Dynamics of internalized anti-Dld-Atto-647N in dividing radial glia cells throughout mitosis (40 timeframes and an interval time of 30 s between each frame; 20 min in total). Please click here to download this Video.

Discussion

We have developed an antibody uptake assay for labeling and imaging endosomal Notch/Delta trafficking in zebrafish radial glia progenitors with high efficiency. Compared to previous methods used for tracking labeled anti-DeltaD antibody in Drosophila SOP cells^7,8, our method used microinjection instead of incubation of samples in the conjugated antibody. Fluorescently conjugated anti-Dld antibodies were microinjected into the hindbrain ventricle; this technique does not cause injury, as evidenced by the normal development and full recovery of embryos after injection. The embryonic ventricles are continuous, allowing the injected antibody to freely disperse toward the forebrain after microinjection. The injected anti-Dld-atto647N lasts for over 24 h as a stable resource of continuous antibody uptake in the brain. Compared to neural tube injection, introduced in a previous report¹⁰, the zebrafish hindbrain ventricle injection does not cause tissue injury and enables selective anti-Dld-Atto647N uptake by mitotic RGPs lining the brain ventricles in vivo. Although we have not checked the antibody uptake in neuroepithelial cells in the spinal cord, the anti-Dld-Atto647N antibody uptake should be applicable there, just as in the brain. The conjugated antibodies injected into the hindbrain ventricle are expected to diffuse along the spinal cord as well.

For the successful application of the antibody uptake assay, the most important issue is to use a primary antibody that has high binding affinity and specificity for the extracellular domain of a membrane protein target. In addition to labeling Notch/Delta signaling, the protocol is applicable for labeling other membrane proteins or extracellular proteins by using similar strategies. Our protocol has demonstrated highly efficient anti-Dld antibody uptake in developing zebrafish embryos because of the high quality and high specificity of the anti-Dld primary antibody. Secondly, the conjugation of anti-Dld primary antibody with the fluorescent secondary antibody is critical for achieving a high level of antibody uptake efficiency in vivo. We have found Atto non-bleaching secondary antibodies to be more stable and much brighter than other fluorescent antibodies, such as Zenon and Alexa secondary antibodies used in Drosophila notum cultures^7,10. We chose Atto647N as the fluorescent label, as we normally would use it together with GFP or RFP transgenic zebrafish or some other fluorescent markers sharing similar excitation wavelengths as GFP or RFP. Other non-bleachable secondary antibodies with different excitation wavelengths should also work with the protocol. In our protocol, we shortened the incubation time of primary antibodies mixed with the Atto secondary antibody from 12 h to 30 min, and decreased the concentration of Anti-Dld antibodies to 0.05 mg/mL in the final microinjection mixtures. A higher concentration of anti-Dld antibody (0.25 mg/mL) did not increase the antibody uptake efficiency in zebrafish embryos; this may be the reason for needing long incubation times, in order to achieve sufficient labeling with fluorescent secondary antibodies¹⁰. We did not use anti-Dld antibody at a higher concentration as in previous studies¹⁰, because there is a possibility that injected anti-Dld-Atto647N antibodies compete with the endogenously expressed Notch/Delta when binding DeltaD ligand on the membrane of RGPs, thereby interfering with cis or trans Notch/Delta signaling in vivo. We also found that blocking buffer added after primary antibody conjugation was helpful for increasing the specificity of the antibody uptake assay in vivo. As shown in Figure 2, most RGPs in the forebrain internalize anti-Dld-Atto647N actively. Without adding the blocking buffer, most cells show very few internalized anti-Dld-Atto647N endosomes (data not shown).

The main limitation of the protocol is the availability of the antibodies that can be used for the uptake assay. Our protocol is applicable for labeling membrane expressed proteins, if a good antibody to the extracellular domain is available. For cytosolic and nuclear proteins, the antibody uptake assay is not applicable, as the conjugated antibodies cannot pass through the cell membrane to bind to their intracellular binding targets. Live imaging of intracellular proteins mainly depends on different types of fluorescent transgenic models. In recent decades, more fluorescent dyes have been developed for live-cell imaging, such as the Silicon Rhodamine-like (SiR) technology, which has significantly contributed to the live cell imaging of DNA, RNA, actin, and tubulin¹⁷. A combination of different live labeling strategies would be the best way to study complex signaling pathways in live cells.

Materials

Name	Company	Catalog Number	Comments
35mm glass bottom culture dish	MatTek corporation	P35GC-1.5-10-C
air pressure injector	Narishige	IM300
Anti-Mouse-IgG-Atto647N	Sigma-Aldrich	50185
CaCl2.2H2O	Sigma-Aldrich	C3306
Capillaries, 1.2 mm OD, 0.9 mm ID, with filament	World Precision Instruments	1B120F-6
CSU-W1 Spinning Disk/High Speed Widefield	Nikin	N/A	Nikon Ti inverted fluorescence microscope with CSU-W1 large field of view confocal.
Dumont Medical Tweezers Style 5	Thomas Scientific	72877-D
Flaming-Brown P897 puller	Sutter Instruments	N/A	https://www.sutter.com/manuals/P-97-INT_OpMan.pdf
KCl	Millipore	529552
MgSO4.7H2O	Sigma-Aldrich	M2773
micromanipulators	World Precision Instruments	WPI M3301R
Mouse anti-Dld	Abcam	AB_1268496
Mouse IgG blocking buffer from Zenon	Thermofisher Scientific	Z25008
NaCl	Sigma-Aldrich	S3014
Phenol red	Sigma-Aldrich	P0290
Stemi 2000	Zeiss	N/A
Tricaine	Sigma-Aldrich	E10521
UltraPureTM low melting point agarose	Invitrogen	16520050