Vav3 is regulated during the cell cycle and effects cell division

Abstract

Vav3 is a member of the family of guanine nucleotide exchange factors implicated in the regulation of Rho GTPases. Although the exact in vivo function of Vav3 is unknown, evidence from several studies indicates a role distinct from Vav2 or Vav1. Here we report that the expression of Vav3 is regulated during the cell cycle. Strikingly, Vav3 was transiently up-regulated in HeLa cells during mitosis, whereas enforced expression of Vav3 perturbed cytokinesis and led to the appearance of multinucleated cells. These effects of Vav3 were RhoA-dependent, required phosphorylation of the regulatory tyrosine 173, but were not enhanced by N-terminal truncations. Thus, this report establishes that expression of Vav3 is strictly regulated in a cell cycle-dependent manner and implicates Vav3 in the control of cytokinesis.

Rho-family GTPases are implicated in the control of a variety of cellular functions by means of activation of the mitogen-activated protein (MAP) kinase and the c-Jun NH2-terminal kinase (JNK) cascades, and regulate actin cytoskeleton in response to extracellular stimuli (1). In particular, Rho proteins are also involved in the regulation of cytokinesis (2). As signal transducers, these proteins act as bimodal switches, active when GTP-bound and inactive when GDP-bound. Thus, control of Rho requires two different classes of proteins. Rho activators, known as guanosine nucleotide exchange factors, exchange GDP with GTP; on the other hand, GTPase-activating proteins reverse Rho to its inactive GDP-bound state (3).

Proteins with Rho-guanosine nucleotide exchange factor function contain the catalytic Dbl-homology (DH) domain. Among DH proteins, the Vav-family has been extensively characterized (4). All 3 Vav proteins are highly homologous (50–70% identity at the amino acid level) and share an array of structural motifs characteristic of proteins involved in signal transduction, including pleckstrin-homology, two Src-homology (SH) 3, and a single SH2 domain. Catalytic activity of Vav is regulated by tyrosine phosphorylation. Based on recent structural studies, a model was proposed in which tyrosine 174 of Vav1 stabilizes an N-terminal autoinhibitory loop blocking the catalytic site of DH domain; phosphorylation of this tyrosine leads to a steric clash with DH side chains and an unmasking of the catalytic site (5). Importantly, this mechanism is also consistent with earlier observations that Vav1 mutants with N-terminal truncations show constitutive guanosine nucleotide exchange factor activity no longer regulated by tyrosine phosphorylation (4).

Although Vav1 was originally identified as a truncated protein with transforming activity in fibroblasts (6), the endogenous Vav1 is expressed exclusively in hematopoietic cells; however, Vav2 and Vav3 show broad pattern of expression (7–10). Functional importance of Vav1 in several hematopoietic lineages was demonstrated by gene knock-out studies. Thus, Vav1 is essential in T lymphocytes (11–13), whereas B lymphocytes can use either Vav1 or Vav2 (14, 15). However, little is known at present about function(s) of the most recently identified Vav-family member, Vav3, although studies in nonlymphoid and lymphoid cell lines suggest that it has both common and distinct properties with other family members (7, 8, 10). In addition, effects of Vav3 on actin cytoskeleton and growth properties of NIH 3T3 cells implicated Vav3 as a potential activator of RhoA in vivo (8, 10).

In this report, we investigated a potential role for Vav3 in the regulation of cell division. We used the HeLa cell system to demonstrate that endogenous Vav3, but not Vav2, is stringently regulated during cell cycle, whereas its enforced expression disrupts cell division.

Materials and Methods

Plasmids.

Vav expression constructs were described (7). Expression constructs for RhoAN19, Rac1N17, and Cdc42N17 were generated by cloning cDNAs with reverse transcription–PCR, using Daudi cell RNA and subcloning into pCF1.HA or pCF1.GFP (HA, hemagglutinin; GFP, green fluorescent protein) expression vectors, as above. GFP- or HA-tagged N-terminal-deleted ΔN66 (amino acids 67–847), ΔN145 (amino acids 146–847), and ΔN184 (amino acids 185–847) deletions in Vav3 and RhoAN19, Rac1N17, and Cdc42N17 mutations were introduced by PCR-based site-directed mutagenesis and confirmed by sequencing. Primer sequences are available on request.

Abs.

The fusion protein of glutathione S-transferase and Vav3 region surrounding DH and pleckstrin-homology domains (amino acids 357–525) expressed in Escherichia coli was purified and used for immunization of rabbits. Specificity of these Abs and no crossreactivity with other family members was confirmed by Western blotting cell lysates of Vav1, -2, or -3-GFP transfectants, as well as by staining cells deficient in Vav3 (W.S., K.F., and F.W.A., unpublished observations). HA-probe, anti-GFP, -RhoA, -Cdc42, -Rac1 (Santa Cruz Biotechnology), -tubulin (Sigma), -actin (Molecular Probes), and -PY (Biomol, Plymouth Meeting, PA) were used.

Cell Culture, Transfections, and Immunoblotting.

HeLa cells were cultured in DMEM with 10% FCS. For immunochemistry cells were plated in cover slips at 5 × 10⁴ cells per well in 24-well plates. For immunoprecipitation and immunoblotting cells were plated on a 100-mm dish at the density of 1.8 × 10⁶ cells per dish, and cells were transfected by using SuperFect (Qiagen, Chatsworth, CA) according to manufacturer's instructions. At 36 h after transfection cells were washed with ice-cold PBS and then lysed in cold Nonidet P-40 lysis buffer (50 mM Tris-Cl, pH 7.6/150 mM NaCl/10 mM NaF/1 mM Na3VO4/10% glycerol/1% Nonidet P-40/1 mM phenylmethylsulfonyl fluoride/1 μg/ml each of leupeptin, aprotinin, and pepstatin) for 10 min at 4°C. Plates were then scraped, and crude lysates were cleared by centrifugation at 14,000 × g for 10 min at 4°C. For immunoprecipitation, total lysates were incubated with 2 μg of Abs and 50:50 slurry of protein A-conjugated Sepharose beads (Amersham Biosciences) with rotation for 3 h at 4°C. Immunoprecipitates were washed 4 times with cold Nonidet P-40 lysis buffer, resuspended in 2× SDS sample buffer, and analyzed by Western blotting following standard procedures. Primary Abs were developed with horseradish peroxidase (HRP)-conjugated secondary Abs (Bio-Rad) and visualized by chemiluminescence (Amersham Biosciences).

Cell Cycle Synchronization.

Synchronization of HeLa cells was done by double thymidine block as described (16), with some modifications. Briefly, HeLa cells in the exponential growth phase were treated with 5 mM thymidine (Sigma) in DMEM containing 10% FCS for 12 h and washed twice with PBS and then cultured in fresh DMEM/10% FCS and then treated again with DMEM/10% FCS/5 mM thymidine for 12 h. After washing cells with PBS, the block was released by replacing the medium with fresh DMEM/10% FCS, and cells were harvested at 0, 1.5, 3, 6, 9, 9.5, 10, 11, and 12 h and analyzed. Flow cytometry was carried out by using a FACS Calibur (Becton Dickinson), and DNA was stained with propidium iodide (Molecular Probes). Calibration of FACS Calibur was performed with DNA QC Particles (Becton Dickinson). To determine the dominant stage of mitosis, confocal imaging (Laser Scanning Microscope 410, Zeiss) was used with samples showing rounding up of cells and chromosomal condensation; cells were fixed and stained with 4′,6-diamidino-2-phenylindole (DAPI) (Molecular Probes) according to manufacturer's instructions.

Immunofluorescence Staining.

HeLa cells were plated on glass coverslips in 24-well plates at 0.8 × 10⁴ cells per well and incubated overnight. The next day, the cells were washed with PBS and fixed with 4% paraformaldehyde in PBS for 20 min at room temperature, washed two times with PBS, and then permeabilized with PBS with 0.1% Nonidet P-40, followed by a wash with PBS. Primary Abs were diluted in PBS with 10% normal goat serum and incubated on the coverslips at room temperature for 2 h, followed by wash with PBS 0.1% Tween 20 (PBST) and PBS. Detection of primary Abs to Vav3 was carried out with Tyramide Signal Amplification Fluorescence Systems Cyanine3 (NEN), mouse Abs were developed with AlexaFluor 488-labeled goat anti-mouse IgG Ab (Molecular Probes), and the HA probe was stained with AlexaFluor 546-labeled goat anti-rabbit IgG Ab. DAPI (Molecular Probes) was used for counterstaining of nuclei at 1 μg/ml for 30 min at room temperature. After the final wash with PBS, coverslips were mounted with glycerin with paraphenylenediamine at 10 mg/ml. All images were obtained by using a confocal laser scanning microscope (Laser Scanning Microscope 410, Zeiss).

RNase Protection Assay.

To generate anti-sense riboprobes, a 197-bp cDNA fragment of human Vav3 corresponding to nucleotides 676–873 and a 204-bp cDNA fragment of Vav2 corresponding to nucleotides 646–850 were generated by reverse transcription–PCR with Daudi B cell RNA. PCR fragments of Vav3 and Vav2 were subcloned into TA Vector (TEasy; Stratagene), sequenced, digested with EcoRI, and subcloned to pBluescript II SK(+) in anti-sense direction (with regard to T7 promoter). Anti-sense riboprobes were synthesized in vitro by T7 RNA polymerase with [³²P]UTP. Labeled riboprobes were gel-purified and checked for expected sizes. RNase protection assay was carried out as described (17). Briefly, the total RNA was hybridized with antisense RNA probe (about 10⁵ cpm) in 30 μl of 80% formamide/40 mM Pipes pH, 6.5/1 mM EDTA/0.4M NaCl at 45^oC overnight. RNA was then digested with RNase A, treated with proteinase K, phenol/chloroform extracted, and analyzed by 6% polyacrylamide gel electrophoresis in the presence of 8 M urea. The signal was detected by using x-ray film or PhosphorImager scanning, and data were analyzed with imagequant software (Molecular Dynamics). RNA was isolated by using Trizol reagent (Invitrogen).

Results

Expression of Vav3 Is Regulated During Cell Cycle.

Several recent experiments indicate a role for Vav3 in the regulation of Rho GTPases distinct from Vav1 or Vav2 (7, 8, 10). Notably, Vav3 was shown to be a specific activator for RhoA, at least in vitro (8). Because RhoA has a well established role in cytokinesis, we decided to investigate whether Vav3 was involved in this process. First, we analyzed the levels of expression and the intracellular distribution of Vav3 during stages of cell cycle. HeLa cells were used to identify distinct mitotic stages, and the expression of endogenous Vav3 protein was examined by immunofluorescence with Vav3-specific Abs (see Materials and Methods). Initial analyses performed with unsynchronized HeLa cell cultures revealed striking heterogeneity in the levels of Vav3 expression (Fig. 1A). Thus, interphase cells showed only weak reactivity with Vav3 Abs, whereas mitotic cells (distinguished by their condensed chromatin; Fig. 1A Left ), showed dramatically elevated expression of Vav3 protein (Fig. 1A Right). As predicted, staining of several other cellular proteins including tubulin, actin, Rac1, or Cdc42 did not reveal significant differences in relative levels of expression in interphase vs. mitotic cells (data not shown).

Figure 1.

Expression of the endogenous Vav3 protein increases in mitotic cells. (A) Immunofluorescence staining of unsynchronized HeLa cells counterstained with DAPI (blue only, Left) or rabbit polyclonal anti-Vav3 Abs (red) and DAPI (blue) (Right). Cells in G₂/M phase are distinguished from interphase cells because of the condensation of chromosomal DNA appearing as mitotic figures (white arrows). [Bar = 50 μm.] (B) HeLa cells at each cell cycle phase were stained with Vav3 Abs as in A (red, Left) and DAPI for DNA staining (blue) and transmitted light (gray) (Right). Mitotic stages were identified by using confocal UV-laser scanning microscopy; a, interphase; b, prometaphase; c, metaphase; d, anaphase; e, telophase; f, cytokinesis. [Bar = 25 μm.]

Next, we examined expression of Vav3 throughout the M phase in more detail and determined that the increase in Vav3 protein could first be detected after the disappearance of the nuclear membrane, during the transition from prophase to prometaphase (Fig. 1B). Vav3 expression reached a maximum in prometaphase and metaphase, and then gradually decreased throughout anaphase toward the onset of cytokinesis (Fig. 1B and data not shown). Taken together, these experiments show that in HeLa cells expression of Vav3 protein is regulated in the context of the cell cycle, rapidly increasing during mitosis.

Vav3 mRNA Levels Are Up-Regulated During Mitosis.

To determine whether changes in expression of Vav3 protein correlated with the levels of Vav3 mRNA, we carried out RNase protection assays with a riboprobe designed to protect a 197-bp fragment of hVav3 mRNA corresponding to nucleotides (676–873) of hVav3 cDNA (see Materials and Methods). In these experiments, HeLa cells were synchronized at the G₁/S transition by using the double thymidine-block method (16). Cells were subsequently released, cultured, and then harvested and analyzed at indicated times. Cell cycle progression was monitored by flow-cytometry (Fig. 2A) and UV-laser scanning microscopy (Fig. 2B and data not shown). Results of these experiments showed that the levels of Vav3 mRNA were very low at the G₁/S transition (Fig. 2C, time = 0 h) but rapidly increased in ≈9 h from the release of cell cycle block, reached a maximum in 10 h, and then quickly decreased to baseline levels (Fig. 2C). We conclude from these experiments that the expression of Vav3 mRNA reaches a maximum during prometaphase or metaphase or both, as cells of these two stages predominated in 10-h cultures (Fig. 2B Right).

Figure 2.

Expression levels of the Vav3 mRNA change in a cell cycle-dependent manner. Exponentially growing HeLa cells were synchronized at the G₁/S transition by double thymidine block. (A) Cells were harvested at several time points after release and analyzed by FACS, using propidium iodide staining. The results are presented as histograms of DNA content. (B) HeLa cells synchronized cultured on cover slips were fixed at 9.5 and 10 h after release, stained with DAPI, and examined under the confocal UV-laser scanning microscope. [Bar = 25 nm.] (C) RNase protection assay of synchronized HeLa cells probed with Vav3 anti-sense riboprobe. Equal amounts of total RNA (5 μg) were loaded. For loading control the same preparations of RNAs were hybridized with anti-sense probe for glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The last lane represents unsynchronized cultured HeLa cells (80% confluent). This experiment was carried out 3 times with similar results. (D) RNase protection assay of Vav2 mRNA in synchronized HeLa cells. RNase protection analysis was performed by using the same samples as used for Vav3 analysis (C) with anti-sense Vav2 riboprobe of the same length as Vav3 (see Materials and Methods). GAPDH was also used as a loading control. A representative experiment is shown (n = 4). Relative band intensity was quantified in some experiments by using a PhosphorImager (Molecular Dynamics).

Next, we examined whether Vav2, the only other Vav protein present in HeLa cells, was also up-regulated in mitosis. We used synchronized HeLa cell cultures, as described above, and examined Vav2 expression by RNase protection with Vav2-specific probes. Results of these experiments indicated that Vav2 mRNA levels did not show any overt changes within the limits of our assay within 12 h after a release from G₁/S block (Fig. 2D and Materials and Methods). We conclude that up-regulation of expression during mitosis may, therefore, distinguish Vav3 from other family members.

Deregulated Expression of Full-Length Vav3 Perturbs Cell Division.

Given strict regulation of Vav3 during mitosis, we examined whether enforced expression of Vav3 would impair cell division. To this end, we transiently transfected HeLa cells with cDNA expression constructs encoding full-length Vav3, -2, or -1 tagged with either HA or enhanced (E)GFP (Fig. 3A). Expression of all constructs was confirmed by Western blotting with either HA- or EGFP-specific Abs (data not shown). Strikingly, we observed that almost half of transfected HeLa cells (43.6 ± 4.2%) had a flattened shape and contained multiple nuclei at 48 h (Table 1, Fig. 3B). Although these phenotypic changes were not observed in transfections with vector plasmid alone, overexpression of either full-length Vav2 or Vav1 resulted in an increase in binucleated cells (12.7 ± 3.5% and 16.5 ± 2.9%, respectively; Table 1); however, in contrast to Vav3-expressing cells, these cells had typical morphology and hence most likely represented normal cell cycle intermediates (data not shown). Notably, approximately 10% of cells expressing Vav3 contained 3 or more nuclei, clearly a result of aberrant cell division, whereas such cells were virtually absent in Vav2-, Vav1-, or mock-transfected cultures (Table 1, Fig. 3B). In addition, only Vav3 transfections resulted in production of very large “giant” cells (≈1–2% of all transfected cells, Fig. 3B d, and data not shown). We conclude from these experiments that expression of full-length Vav3, but not Vav2 or Vav1, in HeLa cells disrupts cell division and leads to production of multinucleated cells.

Figure 3.

Expression of full-length Vav3 induce multinucleated cells. (A) Schematic of Vav3 constructs used in this study. Structural domains are indicated. (B) HeLa cells were transfected with GFP-tagged full-length Vav3 cDNA constructs and processed for confocal laser scanning microscopy after 30–72 h. Nuclei were stained with DAPI (blue). (B a–c) Exogenous Vav3-C expressed HeLa cells detected as multinucleated or binucleated cells 40 h after transfection. (B d) “Giant” cell, 72 h. [Bar = 25 μm.]

Table 1.

Induction of multinucleated cells by Vav proteins

Expression construct	Mononucleated, %		Binucleated, % (1)		Trinucleated, % (or more) (2)		Multinucleated, % (1) + (2)
	Mean	SD	Mean	SD	Mean	SD	Mean	SD
Vector	98.8	0.5	1.2	0.5	0	0	1.2	0.5
Vav3	56.5	4.2	34.4	3.3	9.2	3.8	43.6	4.2
Vav2	83.5	2.9	16.3	3.2	0.2	0.5	16.5	2.9
Vav1	87.3	3.5	12	3.2	0.7	0.5	12.7	3.5

GFP-tagged constructs were transfected into HeLa cells, as indicated. Cells were fixed 48 h after transfection, counterstained with DAPI, and 200 GFP-positive cells were counted in each sample under immunofluorescence microscope. Data is shown as percent of total GFP positive. Five independent experiments were carried out, and data represent the means. SDs are shown. 

Vav3 Effects Cytokinesis by a RhoA-Dependent Mechanism.

Because recent reports implicated RhoA as a potential downstream effector for Vav3, we decided to determine whether the ability of Vav3 to induce multinucleated cells depended on RhoA activity. To this end, we used an approach of transfecting HeLa cells with Vav3 either alone or together with one of the dominant-negative mutants of Rho-family members, including RhoA^N19, Rac1^N17, or Cdc42^N17 (Fig. 4). These experiments showed that only coexpression of RhoA^N19, but not Rac1^N17 or Cdc42^N17, attenuated the ability of Vav3 to induce multinucleated cells (Fig. 4). We obtained similar data by using either HA- or GFP-tagged Rho mutants; comparable levels of expression of RhoA^N19, Rac1^N17, and Cdc42^N17 were confirmed by Western blotting with anti-HA or anti-GFP Abs, as well as by analyses of GFP fluorescence (data not shown). We interpret these results as suggesting that the effects of Vav3 on cytokinesis maybe mediated by a RhoA-dependent pathway.

Figure 4.

Induction of multinucleated cells by Vav3 depends on RhoA activity, tyrosine 173, and N-terminal sequences. GFP-tagged wild-type-, mutated Vav3-, or DN GTPase-encoding constructs were transfected into HeLa cells. Cells were processed and analyzed as in Table 1. Five independent experiments were carried out, and data representing average and SDs are shown.

Induction of Multinucleated Cells by Vav3 Is Not Enhanced by Removal of N-terminal Sequences.

Previous studies demonstrated that N-terminal truncations of Vav1 and Vav2 induced guanosine nucleotide exchange factor activity (reviewed in ref. 4). To determine whether Vav3 was subjected to a similar regulatory mechanism, a series of Vav3 expression constructs was generated with progressively larger deletions of the N terminus (Fig. 3A): (i) the ΔN66Vav3, which lacks a part of CH domain; (ii) the ΔN146Vav3, which lacks the entire CH domain but retains part of the AC region with tyrosine 173; and (iii) the ΔN184Vav3, which loses both CH and AC domains with tyrosine 173. Surprisingly, transient transfection experiments showed that full-length Vav3 had the highest activity inducing multinucleated cells (Fig. 4). Thus, compared to full-length Vav3, the activity of ΔN184-Vav3 was slightly lower, further diminished in ΔN146-Vav3, and almost lacking for ΔN66-Vav3 (Fig. 4). We conclude from these experiments that the ability of wild-type Vav3 to induce multinucleated phenotype in HeLa cells does not depend on the removal of its N-terminal sequences.

Phosphorylation of Tyrosine 173 Is Required for Vav3 to Effect Cytokinesis.

The activity of Vav1 is regulated by phosphorylation of a conserved tyrosine 174 (5). To determine whether the phosphorylation of an analogous residue in Vav3 (position 173) was required for Vav3 to induce multinucleated cells, we mutated this amino acid to phenylalanine (Y173F substitution) and performed transient transfection experiments in HeLa cells. Strikingly, the ability of Vav3^Y173F to induce multinucleated cells was almost completely abolished (Fig. 4). To determine whether wild-type Vav3 was tyrosine-phosphorylated in HeLa cells, we performed immunoprecipitation and Western blotting experiments that showed that Vav3 protein was in fact constitutively tyrosine-phosphorylated, albeit at low level (data not shown). These experiments indicate that phosphorylation of tyrosine 173 in Vav3 is required for its ability to induce multinucleated cells.

Discussion

Vav3 Is Regulated During Cell Cycle.

In this report we show that in HeLa cells, the expression of Vav3 mRNA and protein changed during the cell cycle. Thus, Vav3 was expressed at low levels in interphase, but rapidly increased within hours of entry into M phase. We propose that these changes represent a novel mode of regulation of the Vav3 activity. Consistent with this view, enforced expression of wild-type Vav3 disturbed cell division and induced multinucleated cells, suggesting a block in cytokinesis. Moreover, regulated expression of Vav3 distinguished it from other family members, as the endogenous Vav2 did not show significant changes in expression throughout the cell cycle, and neither Vav2 nor Vav1 could induce multinucleated cell phenotype. We note, however, that the analyses of endogenous Vav2 expression were limited at this time to measuring steady-state levels of Vav2 mRNA, because of a lack of suitable Vav2 Abs.

Vav3 Effects on Cell Division Depend on RhoA.

In this report we show that production of multinucleated cells by Vav3 could be specifically suppressed by coexpression of dominant negative RhoA, but not Rac1 or Cdc42. These data strongly suggest that Vav3 effects cytokinesis via a mechanism that involves activation of the endogenous RhoA. In this context, deregulated RhoA activity has been implicated in uncoupling of nuclear division and cell division in a variety of experimental systems (2). In addition, recent reports showed that deregulating RhoA activity, for example by enforced expression of MgcRacGAP, disturbs cytokinesis (18).

N-terminal Truncations Do Not Enhance Production of Multinucleated Cells.

We were surprised to find that full-length Vav3 showed higher activity of inducing multinucleated cell phenotype than mutants lacking N-terminal sequences. It is possible that the loss of N terminus may lead to down-regulation of Vav3 activity in vivo, for example by changing its localization or sequestering it from RhoA. In this context, although one previous report using NIH 3T3 cells found no evidence for activation of Vav3 transforming potential by N-terminal deletions, a more recent study reached the opposite conclusion (8, 10); the exact reasons for these differences are not known at present. Because our studies presented here are limited to HeLa cells, it is possible that the effects of Vav3 on cytokinesis may be cell type-specific.

Vav3 Effects Cell Division via a Mechanism That Depends on Tyrosine 173.

Recent NMR studies implicated tyrosine 174 as a crucial determinant of “open” and “closed” confirmation of Vav1 with regard to its DH site accessibility (5). Because this tyrosine is evolutionarily conserved, we investigated whether the corresponding tyrosine in Vav3 (position 173) could also play a regulatory role. We mutated this tyrosine and showed that the Y173F substitution virtually abolished the ability of Vav3 to induce multinucleated phenotype. These results suggest that tyrosine 173 of Vav3 represents a key structural determinant of the N-terminal autoinhibitory loop. We also found that wild-type Vav3 but not Vav3^Y173F was tyrosine-phosphorylated, albeit weakly, in HeLa cells (K.F., W.S., F.W.A., unpublished observations). However, more detailed analyses will be required to distinguish whether this weak phosphorylation was in fact constitutive or whether it occurs only in a subset of cells, for example, at a particular stage of the cell cycle.

In summary, our studies presented here implicate Vav3 as an upstream regulator of RhoA during cytokinesis in HeLa cells. In this context, the up-regulation of Vav3 during mitosis represents a novel mode of regulation among Vav proteins.

Abbreviations

DH

Dbl-homology

HA

hemagglutinin

GFP

green fluorescent protein

DAPI

4′,6-diamidino-2-phenylindole