|Home | About | Journals | Submit | Contact Us | Français|
The ability to differentially label single cells has important implications in developmental biology. For instance, determining how hematopoietic, lymphatic, and blood vessel lineages arise in developing embryos requires fate mapping and lineage tracing of undifferentiated precursor cells. Recently, photoactivatable proteins which include: Eos1, 2, PAmCherry3, Kaede4-7, pKindling8, and KikGR9, 10 have received wide interest as cell tracing probes. The fluorescence spectrum of these photosensitive proteins can be easily converted with UV excitation, allowing a population of cells to be distinguished from adjacent ones. However, the photoefficiency of the activated protein may limit long-term cell tracking11. As an alternative to photoactivatable proteins, caged fluorescein-dextran has been widely used in embryo model systems7, 12-14. Traditionally, to uncage fluorescein-dextran, UV excitation from a fluorescence lamp house or a single photon UV laser has been used; however, such sources limit the spatial resolution of photoactivation. Here we report a protocol to fate map, lineage trace, and detect single labeled cells. Single cells in embryos injected with caged fluorescein-dextran are photoactivated with near-infrared laser pulses produced from a titanium sapphire femtosecond laser. This laser is customary in all two-photon confocal microscopes such as the LSM 510 META NLO microscope used in this paper. Since biological tissue is transparent to near-infrared irradiation15, the laser pulses can be focused deep within the embryo without uncaging cells above or below the selected focal plane. Therefore, non-linear two-photon absorption is induced only at the geometric focus to uncage fluorescein-dextran in a single cell. To detect the cell containing uncaged fluorescein-dextran, we describe a simple immunohistochemistry protocol16 to rapidly visualize the activated cell. The activation and detection protocol presented in this paper is versatile and can be applied to any model system.
Note: The reagents used in this protocol can be found in the table appended at the end of the article.
* This procedure uses the zebrafish as the animal model system.
* This procedure assumes that the embryo expresses reporter GFP. The embryo is viewed using the argon ion (488 nm) laser source attached to the LSM 510 META NLO microscope. A single GFP positive cell is selected and photoactivated using the Mai Tai laser (femtosecond laser) coupled to the LSM 510. The procedure is the same for non-fluorescent embryos. Alternatively, white light can be used to find the cell to be activated, followed by activation using the Mai Tai laser source.
Figure 1. Photoactivation of caged fluorescein-dextran. (A,B) Brightfield and fluorescence image of a lateral view zebrafish embryo at the 13/14-somite stage before photoactivation. (C,D) The embryo was photoactivated at the13/14-somite stage in one of the somites (arrow) with a wavelength of 740 nm, scan time of 31 secs, and an average laser power of 47 mW. Post-activation, (d) fluorescence from uncaged fluorescein-dextran is observed. Orientation: Dorsal (right), Ventral (left). Scale bar 100 μm.
Figure 2. Immunodetection of caged fluorescein-dextran. Figure 2 depicts the immunostaining of uncaged fluorescein-dextran in an 18/19 hpf zebrafish embryo. At the 10-somite stage a small region of the embryo was activated in the lateral plate mesoderm using the same laser parameters stated in Figure 1. Using the immunodetection procedure, the arrow identifies the activated region with minimal background staining. Orientation: Dorsal (top), Ventral (bottom). Scale bar 100 μm.
This paper describes a versatile method for single cell fate mapping based on the photoactivation of caged fluorescein-dextran. Uncaging of caged fluorescein-dextran in single cells is achieved using two-photon absorption from a Mai Tai femtosecond laser coupled to the Zeiss LSM 510 META confocal microscope. Uncaged fluorescein-dextran in photoactivated cells is rapidly visualized using a simple immunohistochemistry procedure.
In this protocol the two-photon absorption wavelength for photoactivation of caged fluorescein-dextran is 740 nm. This wavelength was experimentally determined by photoactivating caged fluorescein-dextran injected wildtype embryos. The laser excitation wavelength was varied from 600 to 800 nm, and the presence or absence of fluorescein fluorescence post-activation was qualitatively determined. We found that 740 nm produced the strongest fluorescein signal following activation. Ideally, 740 nm should work for all model systems; however, we advise testing a range of wavelengths to determine the optimal two-photon excitation wavelength for your sample.
In caged fluorescein-dextran injected wildtype embryos, for each two-photon excitation wavelength tested we also varied the average laser power. For an excitation wavelength of 740 nm, an average laser power of 47 mW produced a stronger fluorescein signal at the activated region in comparison to lower average laser powers. Higher average laser powers did not produce stronger fluorescein signals, but induced ablation of the tissue. Since femtosecond laser pulses are efficient at ablating biological tissue15, we recommend determining the threshold average laser power for photoactivation that avoids tissue ablation.
Two-photon absorption occurs within a small interaction volume. To ensure proper activation of caged fluorescein-dextran, the cell must be brought into proper focus. In the above protocol, GFP positive cells were simultaneously imaged under white light to confirm cell focus. Therefore, white light acquisition in addition to other channels is advisable. After confirming cell focus, our protocol suggests magnifying the activated cell. A zoom factor of 50 to 70 is suggested, however, this value depends on the size of the chosen cell. Increasing the zoom isolates the activated cell from adjacent cells, and limits the scan area for two-photon activation. Ideally, the scan area should cover a large area of the cell.
Detection of single activated cells requires that the background staining be minimal. In the above immunodetection protocol it is important that the sample is blocked in blocking buffer for 5 to 6 hours (step 4.13). When developing with NBT/BCIP, uncaged fluorescein-dextran should become visible in 10 to 20 mins. If background staining is strong, we suggest a longer blocking time and additional washes to remove the antibody (step 4.17).
Figures 1 and 2 provide representative results of photoactivation and immunodetection. In Figure 1, early 1- to 2-cell stage wildtype embryos were injected with caged fluorescein-dextran. The embryos were raised to mid somitogenesis and mounted in low melting agarose in the lateral orientation. Figure 1(a,b) depict brightfield and fluorescent images of the embryo before photoactivation. Evidence that the embryo remained uncaged is demonstrated by the absence of fluorescein fluorescence in Figure 1(b). A small region within a somite was activated with a wavelength of 740 nm for 31 secs using an average laser power of 47 mW, Figure 1(c; arrow). Post-activation, two-photon uncaging of fluorescein-dextran occurred as shown by the fluorescence from the activated area, Figure 1(d; arrow). Activation was not accompanied by laser ablation, as the somitic tissue remained intact, Figure 1(c). Figure 2 demonstrates the immunodetection of uncaged fluorescein-dextran. A small region in the lateral plate mesoderm of a 10-somite stage embryo was photoactivated using the same laser parameters as mentioned in Figure 1. The embryo was raised and processed using steps 4.1 through 4.20. The arrow in Figure 2 locates the region activated, as indicated by the accumulation of blue precipitate. Only the activated location is clearly detected without interfering background staining.
1 X PBT (1 L) 100 mL 10 X PBS 2 g BSA 10 mL 20 % Tween
10 X PBS (1 L) 80 g NaCl 2 g KCl 6.1 g Na2HPO4 1.9 g KH2PO4 ddH2O to 1 L pH to 7.3
4 % Paraformaldehyde (160 mL) 32 % Paraformaldehyde (two 10 mL vials) 8 mL 20 X PBS 132 mL ddH2O
AP Buffer ( 50 mL) 1 mL 5 M NaCl 2.5 mL 1 M MgCl2 5 mL 1 M Tris pH 9.5 250 μL 20 % Tween 41.25 mL ddH2O
Antibody solution (for 1 sample) 5.5 μL Acetone powder 40 μL PBT 0.8 μL LS 0.1 μL Anti-fluorescein-AP antibody
2 % LS (360 μL for 1 sample) 7.2 μL 100 % LS 352.8 μL PBT
1 X Danieau media (1 L)‡ 11.6 mL 5 M NaCl 700 μL 1 M KCl 400 μL 1 M MgSO4 600 μL 1 M Ca(NO3)2 5 mL 1 M HEPES ddH2O to 1 L pH adjust to 7.6
‡ Danieau is an ideal salt solution for rearing dechorionated zebrafish embryos. Other media may be used, provided they are isotonic solutions with properly osmolarity (~ 300 mOsm for zebrafish)
NBT stock (1 mL) 50 mg Nitro Blue Tetrazolium 0.7 mL dimethyl formamide anhydride 0.3 mL ddH2O
BCIP stock (1 mL) 50 mg 5-bromo-4-chloro-3-indolyl phosphate 1 mL dimethyl formamide anhydride
NBT/BCIP developing solution 1 mL AP buffer 4.5 μL NBT stock 3.5 μL BCIP stock
No conflicts of interest declared.
We thank J. Chen for providing the caged fluorescein-dextran synthesis protocol. This research was supported by the March of Dimes award 5-FY09-78, the American Heart Association award 09BGIA2090075, and the NIH / NHLBI award R01 HL107369-01 to S.S. A special thank you to C. Closson for his microscopy assistance.