Abstract: It is know from in vitro experiments that contractile forces of nonmuscle myosin-II (MyoII) in the cytoplasm affect the function of the nucleus. Furthermore, perinuclear pools of MyoII have been localized among several types of cultured cells. However, beyond cell culture experiments there is no evidence that cytoplasmic MyoII associates with the nucleus. The aim of the current experiments is to determine whether or not MyoII associates with the nucleus of cells in metazoan tissue. The giant cells within salivary gland organs from 3rd instar Drosophila melanogaster larvae were evaluated in living and fixed preparations. A UAS-Gal4 conditional expression system was used to drive gene expression of MyoII specifically within salivary gland organs. A GFP-MyoII protein trap line which uses the endogenous MyoII promoter to control expression of full-length GFP-MyoII was also employed. Additionally, antibody immunoreactivity was used to localize endogenous MyoII proteins. The results revealed a perinuclear localization pattern for the MyoII molecule. The molecule formed oligomerized (filament-like) conformations on the cytoplasmic side of the nuclear lamin. Furthermore, the MyoII α-helical coiled-coil tail was shown to be necessary for perinuclear localization and oligomerization. These experiments provide direct evidence for a nuclear association of MyoII within metazoan tissue.
INTRODUCTION
The nucleus can function as a mechanosensor that detects mechanical forces in the cytoplasm (Dahl et al., 2008). KASH (Klarsicht, ANC-1, Syne homology) proteins on the outer nuclear membrane and SUN (Sad1/UNC-84) proteins attached to nuclear lamins directly transfer force generated in the cytoplasm to the nucleus (Starr, 2009). It is known that the nucleus is sensitive to cytoplasmic forces generated by the nonmuscle myosin-II (hereafter referred to as MyoII) molecular motor. Among naïve mesenchymal stem cells, MyoII contraction affects transcriptional profiles that determine cell lineage (Engler et al., 2007; McBeath et al., 2004). Additionally, MyoII facilitates the translocation of the nucleus through the cytoplasm during interkinetic nuclear migration in the retina and nuclear positioning in the cytoplasm among migrating NIH 3T3 cells (Gomes et al., 2005; Norden et al., 2009). Furthermore, several types of cultured cells, stained with MyoII antibodies have revealed a prominent cytoplasmic pool of MyoII around the nucleus (Hirano et al., 1999; Kolega, 1998; Maupin et al., 1994). The existence of perinuclear pools of MyoII suggest localized force generation at the nucleus. However, perinuclear pools of MyoII are also believed to serve as solatable reservoirs that are mobilized when needed by the cell. For instance, migrating bovine microcapillary endothelial cells sequester perinuclear MyoII to re-enforce a 10 μm wide region just behind their leading lamellipodia (Kolega, 1997). Additionally, during cellular locomotion a significant proportion of perinuclear MyoII from Swiss 3T3 murine fibroblast become diffusible (Kolega and Taylor, 1993). Up to 79% of the perinuclear pool of MyoII is highly diffusible while 21% exhibit a low mobility of 4.1x10-9 cm2 sec-1 (37°C) (DeBiasio et al., 1988). These observations from cell culture experiments suggest that perinuclear pools of MyoII may generate force onto the nucleus and/or serve as a diffusible reservoir that reinforce tension generating loci in the cell (Engler et al., 2007; Kolega, 1997). However, high resolution microscopic evidence of such perinuclear reservoirs within living or fixed preparations of metazoan tissue has not been reported. Therefore, it is not clear whether the functions of perinuclear MyoII actually generalize beyond cultured cells (Norden et al., 2009).
In the current study living and fixed whole-mount salivary gland organs from larval Drosophila melanogaster were used to localize perinuclear pools of zip/MyoII (the Drosophila homolog of MyoII). These perinuclear pools exhibited an oligomerized conformation and co-distributed with filamentous actin. Furthermore, the tail domain of zip/MyoII was necessary for the assembly of perinuclear oligomers.
MATERIALS AND METHODS
GFP-zip/MyoII transgenic strains: The cloning and sequencing of GFP-zip/MyoII constructs into the pUAST vector and the generation of transgenic animal strains harboring UAS-GFP-zip/MyoII full-length or fragment have been described previously (Franke et al., 2005a, b). The genotypes of these and other animals used in the present study (12/06-10/09) are listed in Table 1. The UAS-Gal4 gene expression system was used to drive the expression of full-length UAS-GFP-zip/MyoII and UAS-GFP-zip/MyoII head+neck, neck+tail and tail domains in salivary glands of living Drosophila wandering 3rd instar larvae (review: Phelps and Brand, 1998). For instance, animals harboring full-length or domain specific UAS-GFP-zip/MyoII constructs were crossed with a transgenic line that harbored the salivary gland Gal4 driver (Cherbas et al., 2003). The GFP-zip/MyoII protein trap line (CC01626) which uses the endogenous zip/MyoII promoter to control expression of full-length GFP-zip/MyoII was also used in this study (Morin et al., 2001).
Whole-mount immunofluorescence: Wandering 3rd instar larvae were washed in dH2O to remove adhering yeast. Salivary glands were dissected in phosphate buffered saline (PBS: 1.86 mM of NaH2PO4, 8.41 mM of Na2HPO4, 175.0 mM of NaCl, pH 7.4). The glands were then fixed in either 4% formaldehyde in PBS for 20 min, 3.7% formaldehyde in PBS for 2 h or 10% formaldehyde in PBS + 0.2% Tween for 10 min. There were no significant differences in antibody staining pattern or intensity for these fixatives, additionally, the former two fixatives preserved endogenous GFP fluorescence.
| Table 1: | Animal genotypes |
Following fixation the salivary glands were washed 3x5 min with PT (PBS and 0.1% Triton X-100) then 30 min with PBT (PBS, 0.1% Triton X-100 and 0.1% bovine serum albumin). They were then incubated in PBT+5.0% NGS (normal goat serum) for 2 h. Primary antibodies were added to PBT+5.0% NGS at a concentration of 1: 200 and incubated over night at 4°C. The primary antibodies used in this study were anti-Drosophila nonmuscle MHC-656 (Kiehart and Feghali, 1986), anti-Drosophila nuclear lamin (Riemer et al., 1995), anti-Drosophila spectrin-243 (Thomas and Kiehart, 1994) and anti-Drosophila moesin (Edwards et al., 1997). Endogenous GFP signal was preserved throughout these procedures therefore no anti-GFP antibody was needed. Following incubation with an antibody, the salivary glands were washed 3x5 min with PBT then incubated with secondary antibody (mouse Cy3 and/or rabbit Alexa Flur 488) in PBT+NGS at a concentration of 1: 600 for 1 h at 22°C. The salivary glands were then washed 3x5 min with PBT then incubated for 1 h in a DAPI solution (10 mL of 1X PBS, 10 μL of Tween-20, 1 μL of a 20 mg mL-1 DAPI/H2O solution) and/or a solution of PBT+NGS containing rhodamine phalloidin (1:600 concentration). Afterwards, the salivary glands were washed 3x5 min with PBT then incubated in mounting media (10% 1 M Tris-Cl pH 8.0, 90% glycerol, 0.5% N-propylgallate) for 15 min then mounted on glass slides. For live whole-mount immunoflurescence, wandering 3rd instar larvae were washed in dH2O to remove adhering yeast. Salivary glands were then dissected in Schneider’s Drosophila medium (Invitrogen Corporation, Carlsbad, CA. USA), mounted on glass slides in Schneider’s Drosophila medium and immediately imaged.
Laser scanning microscopy (LSM): A Zeiss LSM 510 confocal on an Axio Observer microscope mounted on a motorized Marzhauser scan stage (DC 120x100 mm) was employed in these studies. The objectives used to examine salivary gland specimens were a Zeiss EC Plan-NeoFluar 40x/1.30 oil objective and a Zeiss Plan-Apochromat 100x/1.4 oil objective. The excitation wavelengths were from a 405 nm Diode laser, 488 nm Argon laser and a 561 nm Diode laser. Conventional fluorescence filters (LP420, LP505 and LP575) for DAPI, green and red were employed with pinholes ranging from 66-128 μm and optical zooms of 1.3 to 2X. Serial 3 μm Z slices were obtained throughout the salivary glands starting from the ventral most surface to the dorsum. Gain and offset were optimized for the brightest central planes of the stack. The Zeiss LSM 510 version 4.2 software was used for offline analysis of the images, such as producing Z-stacks, rotating and orienting images. Colors in a few images were sometimes modified to better convey meaning and/or contrast. For instance, DAPI-blue in some images was changed to red and green was changed to white (see caption in each image). Huygens Essential version 3.0 was used for Surface Rendering (volume visualization) in order to separate or demonstrate the association between different color volumes, such as blue-DNA, red-lamin and GFP-zip/MyoII.
RESULTS AND DISCUSSION
Perinuclear localization of zip/MyoII in salivary glands: Figure 1 reveals the localization of zip/MyoII in whole-mount salivary gland cells using three independent methods. In one method endogenous zip/MyoII was localized through immunolabeling with Drosophila zip/MyoII antibody (Kiehart and Feghali, 1986). Panels A to A" shows DAPI stained DNA (panel A), epitope immunoreactivity for Drosophila zip/MyoII (panel A') and a merger of the two panels (panel A"). The results reveal a prominent perinuclear pool of zip/MyoII. This pattern of prominent perinuclear localization is consistent with the staining pattern observed among cultured cells (Hirano et al., 1999; Kolega, 1998; Maupin et al., 1994). In the second method, GPF-zip/MyoII protein trap (Morin et al., 2001) also revealed perinuclear localization (note that in all GFP studies, thendogenous GFP signal was preserved therefore, no staining was needed).
| Fig. 1: | Perinuclear localization of endogenous and transgenic zip/MyoII in single cells from whole-mount salivary gland organs. Endogenous (antibody stained) zip/MyoII (panels A-A") exhibit a prominent perinuclear localization. However, GFP-zip/MyoII protein trap (panels B-B") exhibit a modest perinuclear localization. Over-expression of GFP-zip/MyoII with the salivary gland specific Gal4 driver (panels C-C") result in prominent perinuclear localization. Arrows point to the perinuclear clusters of zip/MyoII |
Panels B to B" shows DAPI stained DNA (panel B), GPF-zip/MyoII fluorescence (panel B') and a merger of the two panels (panel B"). Unlike the immunolabeling method, the protein trap revealed a modest perinuclear accumulation. The third method, utilized the UAS-Gal4 expression system to over-express GFP-zip/MyoII (Brand and Perrimon, 1993; Franke et al., 2005a, b). Panels C to C" shows DAPI stained DNA (panel C), GPF-zip/MyoII fluorescence (panel C') and a merger of the two panels (panel C"). Similar to the immunolabeling method, the UAS-GAL4 system revealed a prominent perinuclear pool of GFP-zip/MyoII. This pattern of prominent perinuclear localization is consistent with the localization of GFP-MyoII among cultured cells. Furthermore, experiments where fluorescent dye labeled myosin-II is microinjected into cultured cells also reveal prominent perinuclear localization (DeBiasio et al., 1988; Kolega, 1998; Kolega and Taylor, 1993). Since, UAS-Gal4 expression of GPF-zip/MyoII (Fig. 1, panels C-C") is qualitatively similar to immunolabeling of endogenous zip/MyoII (Fig. 1, panels A-A"), the UAS-Gal4 expression system (Brand and Perrimon, 1993; Franke et al., 2005a,b) was used to further evaluate perinuclear conformations of GFP-zip/MyoII. Figure 2 reveals the ubiquitous nature of perinuclear pools of GPF-zip/MyoII by showing oligomerized conformations among each nuclei of a section of the salivary gland organ. This Fig. 2 also reveals the special distribution of GPF-zip/MyoII within the salivary gland.
| Fig. 2: | Oligomerized perinuclear pools zip/MyoII in a whole-mount salivary gland organ. In all panels green is GFP-zip/MyoII, red or blue is DNA and the cell junction protein, moesin is in white. Panel A is a Z-section through a whole-mount salivary gland. Panels B-F are stacked Z-sections. Panels B-C provides a 3D view of zip/MyoII and DNA. Panels D-F reveal the spacial orientation of DNA and/or GFP-zip/MyoII within the salivary gland in general and specific cells in particular. Panels G-G'' are 0-6 μm sections through the nucleus outlined in panel A |
Furthermore, Fig. 2 provides high resolution photomicrographs of GPF-zip/MyoII relative to DNA. GPF-zip/MyoII does not seem to significantly co-localize with DNA but assemblies into oligomerized structures that spread around DNA. Furthermore, Fig. 3 shows that these oligomers are localized adjacent to the nuclear lamin on the side that is opposite to DNA (cytoplasmic side). These findings suggest that zip/MyoII may not directly interact with particular DNA fragments or intranuclear compartments. Instead, zip/MyoII may exhibit a more global influence on the nucleus. For instance, nuclear mechanosensing and transduction might be driven in part, by zip/MyoII mediated contractile forces in the cytosol (Engler et al., 2007; McBeath et al., 2004).
The zip/MyoII molecular motor is classically divided to three functionally distinct domains. There is the actin binding ATPase head domain, the light-chain regulatory neck domain and the tail domain (Franke et al., 2006). Individual MyoII molecules may efficiently assemble into filaments which represent the major force generating conformation (Craig and Woodhead, 2006; Liu et al., 2008). Filament assembly is regulated by the tail domain through hydrophobic and electrostatic interactions and phosphorylation induced mechanical folding (Hostetter et al., 2004; Lee et al., 1994; Turbedsky et al., 2005; Liu et al., 2008). It has been shown in vitro that the assembly of Drosophila zip/MyoII filaments is specifically determined by the tail domain via evenly distributed alternating charge repeats (Liu et al., 2008).
| Fig. 3: | zip/MyoII exhibits cytoplasmic perinuclear localization. Panels A-F are serial Z-zections through a single nucleus. Note that GFP-zip/MyoII is localized on the cytoplasmic side of the nuclear lamin (red). Panels G-G' are 3D reconstructions of panels B-F. Scale bar (10 μm) in panel C applies to all panels |
| Fig. 4: | zip/MyoII tail domain regulates the assembly of perinuclear oligomers. Panels A-C are photomicrographs of living whole-mount salivary gland organs. Note that perinuclear oligomers are observed with expression of the tail and neck+tail domains and not the head+neck domain |
Therefore, additional work was pursued to determine whether or not the tail domain was responsible for specifying the assembly of perinuclear oligomers. Figure 4 reveals the expression of the tail, neck-tail and head-neck domains in living salivary gland organs. Like the antibody staining studies and the GFP tagged full length zip/MyoII, the tail and neck-tail constructs exhibited perinuclear oligomers. The tail formed more highly condensed perinuclear oligomers than the neck-tail construct. In contrast, the head-neck construct localized at cell borders and failed to exhibit perinuclear oligomers.
| Fig. 5: | zip/MyoII co-localize with perinuclear actin. Panel A is phalloidin stained actin in whole mount salivary gland. The arrow reveals projections from both perinuclear and cell border actin pools. Panel B is the same salivary gland showing perinuclear localization of GFP-zip/MyoII. Panel C is the merger of panels A and B. Note that the yellow loci reveal strong perinuclear co-localization. Panel D is the Z-stack profile of the organ |
| Fig. 6: | Perinuclear zip/MyoII co-localize with some actin-binding proteins. Panel A shows that spectrin (red, antibody stain) is localized around the nucleus of salivary gland cells. Panel A1 shows the same salivary gland with co-localization of spectrin (red) and GFP-zip/MyoII. Note that the yellow loci are areas of strong co-localization. Panel B shows that moesin (red, antibody stain) is localized at the periphery of the salivary gland. Panel B1 shows no co-localization between GFP-zip/MyoII and moesin |
This lack of perinuclear localization and oligomerization has been demonstrated previously for myosin-II GFP-head constructs in COS-7 cells (Ikebe et al., 2001).
Zip/MyoII co-localize with perinuclear actin. Actin is believed to anchor the nucleus in the cytoplasm and both actin and MyoII interact to exert tension on the nucleus (Gomes et al., 2005; Starr and Han, 2003). Therefore, it was suspected that GFP-zip/MyoII may co-localize with actin at the nucleus. Figure 5 reveals that GFP-zip/MyoII co-localizes with polymeric (phalloidin stained) actin around the nucleus. This actin localization pattern is similar to phalloidin labeled actin shells around the nucleus of newly divided 3T3 cultured cells (Clubb and Locke, 1998). The actin binding protein spectrin, exhibited co-localization with GFP-zip/MyoII in perinuclear pools. However, another actin binding protein moesin, did not co-localize with perinuclear pools of GFP-zip/MyoII (Fig. 6).
CONCLUSION
The existence of perinuclear pools of MyoII is well established among several types of cultured cells. In the current experiments these cell culture observations were extended by revealing perinuclear pools of MyoII in living and fixed preparations of Drosophila salivary gland organs. MyoII in cooperation with filamentous actin is known to generate force at the nucleus during disperate conditions such as mitosis, cellular locomotion and cell lineage determination. Among the giant nuclei of salivary gland cells both zip/MyoII and filamentous actin were localized around the nucleus which indirectly implicates a force generating capacity at the nucleus.
ACKNOWLEDGMENTS
The author would like to thank Professor Dan Kiehart for helpful comments on an earlier version of the manuscript. The lamin Dm0(ADL67.10) monoclonal antibody developed by Paul A. Fisher was obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biology, Iowa City, IA 52242. The work was supported by the Hargitt Cell Biology Research Award.