Flagellar Adhesion between Mating Type Plus and Mating Type Minus Gametes Activates a Flagellar Protein-tyrosine Kinase during Fertilization in Chlamydomonas*

When Chlamydomonas gametes of opposite mating type are mixed together, flagellar adhesion through sex-specific adhesion molecules triggers a transient elevation of intracellular cAMP, leading to gamete activation in preparation for cell-cell fusion and zygote formation. Here, we have identified a protein-tyrosine kinase (PTK) activity that is stimulated by flagellar adhesion. We determined that the protein-tyrosine kinase inhibitor genistein inhibited fertilization, and that fertilization was rescued by dibutyryl cAMP, indicating that the genistein-sensitive step was upstream of the increase in cAMP. Incubation with ATP of flagella isolated from non-adhering and adhering gametes followed by SDS-PAGE and immunoblotting with anti-phosphotyrosine antibodies showed that adhesion activated a flagellar PTK that phosphorylated a 105-kDa flagellar protein. Assays using an exogenous protein-tyrosine kinase substrate confirmed that the activated PTK could be detected only in flagella isolated from adhering gametes. Our results indicate that stimulation of the PTK is a very early event during fertilization. Activation of the PTK was blocked when gametes underwent flagellar adhesion in the presence of the protein kinase inhibitor staurosporine, but not in the presence of the cyclic nucleotide-dependent protein kinase inhibitor, H8, which (unlike staurosporine) does not block the increases in cAMP. In addition, incubation of gametes of a single mating type in dibutyryl cAMP failed to activate the PTK. Finally, flagella adhesion between plus and minus fla10-1 gametes, which have a temperature-sensitive lesion in the microtubule motor protein kinesin-II, failed to activate the PTK at elevated temperatures. Our results show that kinesin-II is essential for coupling flagellar adhesion to activation of a flagellar PTK and cAMP generation during fertilization in Chlamydomonas.

each other via sex-specific adhesion molecules, the mtϩ and mtϪ agglutinins, on their flagella surfaces and then undergo a series of events composing gamete activation that prepares the gametes for cell fusion. Gamete activation includes cell wall loss, movement of proteins (including agglutinins) from the cell body to the flagella, modification of the flagellar tip, phosphorylation of a gamete-specific homeodomain protein involved in activation of zygote-specific genes (1), activation of mating structures, and redistribution of the adhesion/fusion protein Fus1 onto the surface of the mtϩ mating structure (Ref. 2, reviewed in Ref. 3).
Several years ago it was shown that one of the earliest biochemical events triggered by flagellar adhesion is a 10 -20fold increase in intracellular cAMP (4,5) brought about by adhesion-induced activation of a flagellar adenylyl cyclase (6 -9). Related experiments showing that addition of dibutyryl cAMP to gametes of a single mating type induced all of the events known to occur during flagellar adhesion has confirmed that cAMP is a key second messenger in Chlamydomonas fertilization (3,5). Consistent with a central role for cAMP in gamete activation, H8, an inhibitor of cAMP-dependent protein kinase A does not interfere with flagellar adhesion (5) or the adhesion-induced rise in intracellular cAMP (9), but all of the downstream events that normally accompany the increase in cAMP are blocked by H8.
In previous studies we have established that the protein kinase inhibitor staurosporine blocked flagellar adhesion-induced activation of the gamete-specific flagellar adenylyl cyclase (7)(8)(9). More recently, we determined that the microtubule motor protein, kinesin-II, also is essential for the flagellar adhesion-induced increase in cAMP (10). We found that fla10-1 gametes, which have a mutation in the kinesin-II gene (11)(12)(13), exhibited wild type levels of flagellar adhesion, but failed to undergo gamete activation (10,14). Our studies showed that the fla10 gametes were blocked at the step that couples agglutinin interactions to activation of adenylyl cyclase.
Here, we have investigated the very early events that occur following flagellar adhesion to identify a flagellar protein kinase activity stimulated by adhesion, and we examined the role of kinesin-II in protein kinase activation. We show that incubation of gametes in the protein-tyrosine kinase inhibitor, genistein, does not interfere with flagellar adhesion, but inhibits the gamete activation that follows as assessed by gamete fusion. The block can be overcome by incubation of the gametes in dibutyryl cAMP, implicating a protein-tyrosine kinase in the pathway that connects flagellar adhesion to activation of adenylyl cyclase. Biochemical experiments demonstrate that immediately after flagellar adhesion is initiated, a flagellar protein-tyrosine kinase (PTK) is activated whose substrate is a soluble, ϳ105-kDa flagellar protein. Our data indicate that flagellar adhesion per se, and not the downstream events, in-cluding protein kinase A activation and gamete fusion that normally accompany flagellar adhesion, is sufficient to activate the PTK. Moreover, microtubule-based motility and intraflagellar transport are connected to regulation of the PTK, because the PTK is not activated during flagellar adhesion between fla10-1 mtϩ and mtϪ gametes, which have a lesion in the microtubule motor protein, kinesin-II.
Cells and Cell Culture-Chlamydomonas reinhardtii strains 21gr (mtϩ) (CC-1690), 6145c (mtϪ) (CC-1691), fus1-1 (mtϩ) (CC-1158) (all available from the Chlamydomonas Genetics Center, Duke University), and strains fla10-1 (4930 3-4, mtϩ) and fla10-1 (4930 6 -2, mtϪ) (provided by Dr. Susan Dutcher, Washington University, St. Louis) were cultured vegetatively in liquid culture with aeration and gametic cells were obtained as described previously (15). To activate gametes of a single mating type, cells were incubated with 15 mM Bt 2 -cAMP and 150 M papaverine in N-free medium for ϳ30 min. Assessments of gamete activation as determined by assaying cell wall loss and zygote formation as determined by formation of quadriflagellated cells were carried out as described previously (10). For the experiments at 32°C, wild-type gametes and fla10-1 mutant gametes were incubated with illumination in a water bath with constant aeration.
Isolation of Flagella-Flagella were isolated by use of the pH shock method of Witman et al. (16) with minor modifications as follows. Gametes (typically ϳ6 liters, at 5 ϫ 10 6 cells per ml in N-free medium) were harvested by centrifugation at 3,000 ϫ g for 3 min and resuspended in 5% sucrose in 20 mM HEPES, pH 7.2. The cells were placed on ice and the flagella were immediately detached by rapidly reducing the pH to 4.5 using 0.5 M acetic acid with constant stirring. Deflagellation, which usually occurred within 15 s, was confirmed using phasecontrast microscopy and the cells were neutralized with 0.5 M KOH. Cell bodies were removed by centrifugation through a 15-ml cushion of 25% sucrose in 20 mM HEPES, pH 7.2, in 50-ml conical tubes at 3,000 ϫ g for 9 min. The remaining cell bodies were removed from the top layer by centrifugation through a second 25% sucrose, 20 mM HEPES layer. Flagella were harvested from the overlying supernatant by centrifugation in an HB-4 rotor at 8,000 rpm for 15 min. The resulting pellet was thoroughly resuspended to a concentration of ϳ3 mg/ml protein in 5% sucrose, 20 mM HEPES, pH 7.2, containing a 1/100 dilution of the protease inhibitor mixture for plant cells from Sigma (catalogue number P9599). The suspension was divided into small portions, frozen in liquid nitrogen, and stored at Ϫ70°C.
For fractionation, flagella (ϳ300 g of protein in 100 l 5% sucrose, 20 mM HEPES, pH 7.2) isolated from mixed plus and minus gametes and frozen as described above were thawed on ice and centrifuged at 13,000 rpm (Sigma 1-15K refrigerated table top centrifuge, rotor number 12132) for 20 min to yield the freeze/thaw supernatant fraction and the disrupted, sedimented flagella. The sedimented flagella were resuspended in 100 l of buffer A that contained 0.1% Nonidet P-40, 20 mM HEPES, pH 7.2, 50 mM KCl, 10 mM MgCl 2 , 1 mM EDTA, and 2 mM dithiothreitol, kept on ice for 10 min, and centrifuged again as above to yield the 0.1% Nonidet P-40-soluble fraction. The sedimented material was resuspended by pipetting and sonication in 100 l of buffer A. Protein concentration was determined with the Coomassie Blue protein assay reagent of Pierce using crystalline bovine serum albumin as standard.
Assays for Protein-tyrosine Kinase Activity-Flagellar PTK activity was assayed in vitro using both endogenous substrates and an exogenously added PTK substrate. For assays using endogenous flagellar proteins as substrates, samples (5 l) of whole flagella (ϳ3 g/l protein) or flagellar fractions (ϳ0.3-3.0 g/l protein) in 5% sucrose, 20 mM HEPES buffer were mixed with 5 l of 2ϫ PTK buffer (20 mM HEPES, pH 7.2, 10 mM MgCl 2 , 2 mM dithiothreitol, 1 mM EDTA, 50 mM KCl, 2 mM ATP, 0.2% Nonidet P-40, 0.4 mM orthovanadate, 20 mM ␤-glycerolphosphate, and 2% Sigma plant protease inhibitor mixture) at 23°C for the times indicated in the figure legends. Samples were prepared for SDS-PAGE by addition of 5 l of 4ϫ SDS-PAGE sample buffer containing 0.25 M Tris, pH 6.8, 40% glycerol, 8% SDS, 0.1% bromphenol blue, and 0.4 M dithiothreitol. After boiling for 5 min, the samples were cooled on ice and loaded onto 8% mini-slab gels. Electrophoresis was performed at 48 mA for 1.5 h as described previously (17). To assay for PTK activity using an exogenous substrate, we used the dot-blot method of Allis et al. (18) with the peptide poly((L-glutamic acid:L-tyrosine)EY4:1) (PGT) as substrate. The freeze/thaw supernatants of flagella (10 l in 20 mM HEPES, pH 7.2, buffer containing 0.3 g/l protein) isolated from mtϩ gametes alone, mtϪ gametes alone, and from adhering mtϩ and mtϪ gametes were incubated with 10 l of PTK buffer that con- FIG. 2. Flagellar adhesion activates a protein-tyrosine kinase whose substrate is a ϳ105-kDa protein. mtϩ and mtϪ gametes (1-5 ϫ 10 8 cells/ml in N-free medium) were mixed together in N-free medium and agitated by aeration to allow them to undergo flagellar adhesion, which typically was maximal as assessed by microscopy after 2-3 min. The adhering gametes, as well as non-mixed mtϩ and mtϪ gametes alone, were harvested by centrifugation and their flagella isolated as described under "Experimental Procedures." Flagella (5 l containing ϳ15 g of protein) isolated from mtϩ gametes alone, from mtϪ gametes alone (mtϩ, mtϪ), and from mtϩ and mtϪ gametes that had been mixed together for 3 min to allow flagellar adhesion to occur (mtϩ & mtϪ) were frozen and thawed, incubated with 5 l of 2ϫ PTK buffer for 10 min, and analyzed for phosphorylation of endogenous proteins by SDS-PAGE and immunoblotting with anti-Tyr(P) (left panel). The middle panel shows the immunoblot membrane stained with Coomassie Blue G-250. The right panel shows results of an assay of a freeze/thaw supernatant of flagella (ϳ1.5 g of protein) isolated from adhering mtϩ and mtϪ gametes in which the assay was carried out in the presence and absence of ATP. The arrow indicates the 105-kDa protein whose phosphorylation in flagella from adhering cells was consistently identified using this assay. the method described below for immunoblotting.
Immunoblot Analysis-After SDS-PAGE, protein was transferred to a polyvinylidene fluoride membrane (Immobilon-P, Millipore, Bedford, MA) in buffer containing 25 mM Tris, pH 8.3, and 192 mM glycine at 100 V for 1 h. The membrane was blocked with 3% bovine serum albumin in 1ϫ TBST (20 mM Tris, pH 7.6, 137 mM NaCl, 0.05% Tween 20) (19) for 1 h at room temperature and incubated with a 1:1000 dilution of anti-Tyr(P) 4G10 in blocking solution overnight at 4°C. At the end of the incubation, membranes were washed three times with 1ϫ TBST for about 20 min and incubated for 30 min with horseradish peroxidase in blocking solution. The membrane was washed and developed with ECL Western blotting reagents as described by the manufacturer (Amersham Biosciences). In some experiments, proteins were detected after immunoblotting by staining the polyvinylidene difluoride membrane with Coomassie Brilliant Blue R-250 for ϳ5 min, followed by destaining in methanol:acetic acid:water (5:1:5).

RESULTS
To determine whether protein-tyrosine kinase activity was required during flagellar adhesion-induced gamete activation in Chlamydomonas, we used the PTK inhibitor, genistein. We pretreated mtϩ and mtϪ gametes in genistein for 30 min, mixed the cells together in the continued presence of the inhibitor, and assessed zygote formation as a measure of gamete activation. Phase-contrast and bright field microscopy (not shown) indicated that the inhibitor had little effect on flagellar adhesion, with 80 -90% of the cells forming the large, swirling clumps of adhering cells that typify cell-cell adhesion in this highly motile organism. On the other hand, the inhibitor had a significant effect on gamete fusion, an event that only activated gametes can undergo, blocking fusion by 30% at 10 g/ml and 87% at 20 g/ml (Fig. 1). To determine whether the genistein blocked cell fusion per se, or if the block was upstream of the adhesion-dependent increase in cAMP, gametes that had been incubated in genistein subsequently were incubated in dibutyryl cAMP and papaverine, a treatment known to induce gamete activation (5) and then mixed together. As shown in Fig. 1, the Bt 2 -cAMP treatment overcame the genistein inhibition of fusion. The reversal of inhibition by incubation of the gametes in dibutyryl cAMP was consistent with the idea that the genistein-sensitive step in the signaling pathway was upstream of the increase in cAMP induced by flagellar adhesion. These results implicated a PTK activity early during flagellar adhesion-induced gamete activation.
A Protein-tyrosine Kinase That Phosphorylates a 105-kDa Protein Is Activated by Flagellar Adhesion-To test for activation of PTKs induced by flagellar adhesion, we isolated flagella from mtϩ gametes alone and mtϪ gametes alone, and from mtϩ and mtϪ gametes that were undergoing flagellar adhesion and gamete activation. Each of the flagellar samples was incubated for 10 min in a protein kinase assay buffer containing 1 mM ATP and 0.1% Nonidet P-40 followed by SDS-PAGE and immunoblotting with anti-Tyr(P). As shown in Fig. 2 (left  panel), a tyrosine-phosphorylated protein of ϳ105 kDa was readily and consistently detected in the adhering sample (mtϩ and mtϪ) and was not present in flagella isolated from mtϩ or mtϪ gametes alone (mtϩ, mtϪ). The middle panel in Fig. 2 shows the immunoblot stained with Coomassie Blue. The presence of approximately equal amounts of tubulin, which are the two strongly stained bands at ϳ50 kDa, documented the comparable loading of the lanes. The right panel in Fig. 2 shows results from a control experiment in which the detergent-soluble, matrix fraction (see below) of flagella isolated from adhering gametes was assayed for PTK in the absence and the presence of ATP. The phosphorylated 105-kDa protein was detected only when the assay was carried out in the presence of ATP. These results suggested that a flagellar PTK that phosphorylated a 105-kDa flagellar protein was activated soon after mtϩ and mtϪ gametes were mixed together and underwent flagellar adhesion. Fig. 3A shows a time course for the in vitro phosphorylation assay. Samples prepared from flagella isolated from mtϩ gametes alone (mtϩ), from mtϪ gametes alone (mtϪ), and from adhering mtϩ and mtϪ gametes that had been mixed together for 10 min (mtϩ & mtϪ) were incubated as above in the protein kinase assay buffer for the indicated times followed by SDS-PAGE and immunoblotting with anti-Tyr(P). Phosphorylated 105-kDa protein was not detected in any of the samples at 0 min and did not appear at any time in the flagellar samples from the mtϩ or mtϪ gametes alone. In the flagellar samples isolated from adhering mtϩ and mtϪ gametes (mtϩ & mtϪ), phosphorylated 105-kDa protein was detectable within 2 min after the incubation with ATP began and continued to increase during the 15-min assay.
To determine whether the results were because of activation of a PTK or, for example, a change in accessibility of the substrate, we used an exogenous protein-tyrosine kinase substrate, polyglutamine tyrosine (4:1) (PGT) to assay for an adhesion-induced increase in PTK activity. Samples from flagella isolated from mtϩ gametes alone, mtϪ gametes alone, and from adhering mtϩ and mtϪ gametes were incubated with protein kinase buffer that included PGT. At the indicated times, samples were spotted onto nitrocellulose paper, and after several washes, the blot was processed to detect phosphotyrosine with the 4G10 antibody used for immunoblotting. As shown in Fig. 3B, there was substantial phosphorylation of the peptide substrate only in samples from adhering gametes (mtϩ & mtϪ; PGT). Control assays, which did not contain PGT, carried out with flagellar samples from adhering gametes displayed a low level of activity (mtϩ & mtϪ; no PGT), which presumably was because of phosphorylation of the endogenous 105-kDa substrate. The control, non-adhering mtϩ and mtϪ samples (mtϩ, mtϪ) showed only a low level of activity that was much less than that exhibited by the samples from the adhering cells. Thus, experiments using endogenous and exogenous substrates indicated that flagellar adhesion activated a flagellar PTK. We investigated the time during fertilization when the PTK was activated. To do this, mtϩ and mtϪ gametes were mixed together and at various times after mixing, the flagella were detached and isolated and samples were assayed for the presence of the active PTK. As shown in Fig. 4A while PTK activity was high in flagella isolated from gametes that had been mixed for 3 min or more, the PTK activity was detectable in flagella isolated from gametes that had been mixed for only 30 s. These results indicated that PTK activation took place very soon after gametes adhered to each other, most likely before gamete fusion, for example, which begins to take place 3-5 min after mixing. To further examine the properties of the PTK and its 105-kDa protein substrate, we identified the flagellar compartment in which the PTK and the substrate were localized. For fractionation, the flagella were disrupted by freezing and thawing and the sample was centrifuged to yield the freeze/thaw supernatant (matrix fraction). The sedimented flagella with their disrupted membranes (20,21) were further fractionated by detergent extraction and centrifugation and equivalent portions of each fraction were assayed for PTK activity. As shown in Fig. 4B, phosphorylation of the 105-kDa protein was detected only in the starting sample of whole flagella and in the initial freeze/thaw supernatant from the disrupted flagella. The results indicated that both the protein kinase and its substrate were released from flagella by freezing and thawing.

Activation of the Flagellar PTK Occurs at a Very Early
Step during Gamete Interactions-We used the protein kinase inhibitors staurosporine and H8 to examine the stage during flagellar adhesion and gamete activation that the PTK was activated. Previously, we showed that while flagellar adhesion was unaffected by these inhibitors, gamete activation was blocked, albeit at different steps by each inhibitor. The staurosporine-sensitive block is upstream of activation of adenylyl cyclase, as cAMP levels fail to increase when gametes undergo flagellar adhesion in the presence of staurosporine (9). We have also shown that H8, an inhibitor of cyclic nucleotide-dependent protein kinases, does not block the flagellar adhesion-induced increase in cAMP, but inhibits steps in gamete activation, including cell wall release and cell-cell fusion, that are downstream of activation of adenylyl cyclase (9). To test the effects of the inhibitors on activation of the PTK, we preincubated mtϩ and mtϪ gametes in each inhibitor and mixed the cells together in the continued presence of the inhibitors. Examination by light microscopy indicated that, as previously reported (9), flagellar adhesion occurred normally between mtϩ and mtϪ gametes in the presence of each of the inhibitors, but cell fusion was blocked (not shown). Flagella were isolated from mtϩ and mtϪ gametes that were mixed together in the presence and absence of each of the inhibitors as well as from control, nontreated, adhering cells and assayed for the PTK activity. As shown in Fig. 5A, the PTK became activated in flagella of control, adhering cells (C) and in flagella isolated from gametes adhering in the presence of H8, whereas activation of the PTK did not occur in flagella isolated from gametes that had undergone flagellar adhesion in the presence of staurosporine (St). These results indicated that activation of the PTK required a staurosporine-sensitive step and suggested that the PTK was activated at an early stage during gamete activation, downstream of agglutinin interactions and flagellar adhesion and upstream of the appearance of cAMP.
In an independent test of the results indicating that activa-

FIG. 4. Timing of activation and location of the flagellar PTK.
A, mtϩ and mtϪ gametes (10 ml, 1 ϫ 10 8 cells/ml) were mixed together and at the indicated times the pH of the suspension was lowered to 4.5 by addition of 0.5 M HAc to the culture medium. ϳ1 min later, 25% sucrose, 20 mM HEPES was added to a final concentration of 5% sucrose and the pH was adjusted to pH 7.2 with 0.5 M KOH. The detached flagella were isolated as described under "Experimental Procedures" and the whole flagellar samples were assayed for PTK activity using the endogenous substrate in the in vitro assay as described above. B, flagella (100 l containing ϳ300 g of protein in 5% sucrose, 10 mM HEPES, pH 7.2) isolated from adhering gametes were frozen and thawed and centrifuged to obtain the freeze/thaw soluble fraction. The sedimented flagella were resuspended in 100 l of buffer A, which contains 0.1% Nonidet P-40, for 10 min on ice, and centrifuged again to yield the 0.1% Nonidet P-40-soluble and -insoluble fractions. The Nonidet P-40 insoluble fraction was resuspended in buffer A, and equal volumes (5 l) of the starting material and each fraction were incubated for 10 min in PTK buffer and assayed for PTK activity as above.

FIG. 5. Activation of the PTK is a very early event after flagellar adhesion.
A, mtϩ and mtϪ gametes (15 ml each, 1ϫ 10 8 cells/ml) were separately preincubated in N-free medium for 30 min with 100 nM staurosporine or 50 M H8. The preincubated samples were mixed together for 3 min in the continued presence of the inhibitors, and flagella were isolated and assayed for PTK activity as above. C, nontreated control; St, 100 nM staurosporine; H8, 50 M H8. B, wild type mtϪ gametes were mixed with non-fusing fus1-1 mtϩ gametes. After 3 min their flagella were isolated and assayed for PTK activity as above. C, left panel, mtϩ and mtϪ gametes (1 ϫ 10 8 cells/ml each in 15 ml of N-free medium) were incubated separately with dibutyryl cAMP and papaverine for 30 min to induce gamete activation and their flagella were isolated and assayed for PTK activity as above. Similar analyses of flagella isolated from control mtϩ and mtϪ gametes alone (mtϩ, mtϪ), and from control, adhering mtϩ and mtϪ gametes that had been mixed together for 3 min (mtϩ & mtϪ) also are shown. Right panel, flagella isolated from mtϩ gametes alone and from mtϪ gametes alone were analyzed for activation of PTK in PTK assays that included 100 M cAMP. tion of the PTK was upstream of cyclic nucleotide-activated processes, including gamete fusion, we used fus1-1 mtϩ gametes, which are capable of undergoing flagellar adhesion and gamete activation, but are unable to fuse (22,23). fus1-1 mtϩ gametes were mixed with wild type mtϪ gametes and, 3 min after mixing, flagella were isolated and subjected to the in vitro PTK assay. As expected, the gametes underwent flagellar adhesion and gamete activation as evidenced by cell wall loss, but they did not fuse to form quadriflagellated zygotes (not shown). On the other hand, the in vitro assay showed that the PTK was activated (Fig. 5B). Thus, consistent with the H8 experiment above, flagellar adhesion in the absence of gamete fusion was sufficient to activate the PTK (Fig. 5B).
It was also possible that activation of the PTK did not require flagellar adhesion per se, but was downstream of the increase in cAMP and involved a cAMP-induced event that was not sensitive to H8. We tested this idea by assaying for the PTK activity in flagella isolated from gametes of a single mating type that were activated by incubation in dibutyryl cAMP and papaverine. The requirement for flagellar adhesion to initiate many of the cellular events that prepare gametes for cell fusion can be bypassed by incubating gametes of a single mating type in the cell-permeable cAMP analogue, Bt 2 -cAMP and the phosphodiesterase inhibitor, papaverine (5). We incubated gametes of a single mating type in Bt 2 -cAMP for 30 min, confirmed that they had become activated as determined by an assay for cell wall loss (not shown), and then assessed activation of the PTK. As shown in Fig. 5C, flagellar samples from adhering gametes exhibited the expected phosphorylation of the 105-kDa protein, whereas flagella isolated from mtϩ gametes or mtϪ gametes alone activated by Bt 2 -cAMP failed to show active PTK in the in vitro assay. These results indicated that activation of the PTK required flagellar adhesion and was upstream of the appearance of cAMP. As a further test for a possible role for cAMP in activating the PTK, we assessed whether inclusion of cAMP in the PTK assay would stimulate the PTK activity in flagellar samples isolated from mtϩ or mtϪ gametes alone. As shown in Fig. 5C, inclusion of cAMP in the assay of flagella isolated from gametes of a single mating type did not lead to activation of the PTK.
The PTK Is Not Activated during Flagellar Adhesion in fla10-1 Gametes, Which Contain a Lesion in the Microtubule Motor Protein Kinesin-II-Previously, we showed that coupling of flagellar adhesion to the increase in cAMP was disrupted in gametes of the fla10-1 mutant (10, 14), which express a kinesin-II with a temperature-sensitive defect in the 90-kDa motor subunit FLA10 because of a single amino acid substitution in the motor domain (13). The defective kinesin-II is expressed in fla10-1 gametes, but at much lower levels than in wild type cells. When fla10-1 cells are grown at room temperature they are fully flagellated and exhibit IFT. On the other hand when grown at elevated temperature (32°C), the cells do not have flagella (13,21,24). In cells grown at room temperature and shifted to 32°C, IFT particle movement ceases after about 40 min and within 1-2 h, the flagella begin to resorb (12,21,24). For our experiments, fla10-1 mtϩ and fla10-1 mtϪ gametes cultured at 21°C were shifted to 32°C for 40 min, mixed together, and their flagella were isolated 10 min after mixing. Flagella also were isolated from adhering mtϩ fla10-1 gametes and mtϪ fla10-1 gametes that were mixed together at 21°C, from wild type gametes mixed together at 21°C, and from wild gametes shifted to 32°C for 40 min before being mixed together for 3 min. As shown in Fig. 6, when analyzed by the in vitro assay, flagellar samples from wild type gametes mixed at 32°C showed a slight reduction in ability to phosphorylate the 105-kDa protein compared with flagellar samples isolated from wild type gametes mixed at 21°C. On the other hand, although phosphorylation of the 105-kDa protein was similar to wild type levels in assays of the 21°C fla10-1 samples, phosphorylated 105-kDa protein was not detected in assays of the 32°C fla10-1 samples. The bottom panel of Fig. 6, which is a loading control, shows the same samples probed with an antibody against a 48-kDa Chlamydomonas protein (25,26), and documents equivalent protein loading. Thus, coupled with their failure to undergo increases in cAMP during flagellar adhesion at elevated temperature (10), fla10-1 gametes also failed to show flagellar adhesion-induced activation of the PTK. DISCUSSION Previous studies on the molecular events regulated by flagellar adhesion during fertilization in Chlamydomonas indicated that flagellar protein kinases are activated at an early step after adhesion between mtϩ and mtϪ flagella, before the adhesion-induced increase in cAMP (7-9, 20, 26). In addition, we recently reported that in cells with a defect in the microtubule Agglutininmediated flagellar adhesion between mtϩ and mtϪ gametes leads to activation of a staurosporine-sensitive serine-threonine protein kinase(s), which in turn activates a genistein-sensitive PTK whose substrate is the 105-kDa protein. The kinesin-II sensitivity of PTK activation indicates that the motor protein functions very early in the pathway. Subsequent to phosphorylation of the 105-kDa protein, adenylyl cyclase is activated, and the consequent increase in cAMP leads to gamete activation and fusion to form a quadriflagellated zygote. motor protein kinesin-II, gamete activation was blocked at a step downstream of flagellar adhesion but upstream of the adhesion-induced increase in cAMP (10). Here, we identified a protein-tyrosine kinase activity that was regulated by flagellar adhesion per se, and not by the downstream events that accompany flagellar adhesion such as increased levels of cAMP, activation of cyclic nucleotide-dependent protein kinases, or cellcell fusion. Initial experiments with the PTK inhibitor genistein implicated PTK activity in fertilization downstream of flagellar adhesion and upstream of the increase in cAMP (Fig. 1). Use of in vitro PTK assays revealed a PTK activity that was present in flagella isolated from adhering gametes and that was nearly undetectable in flagella isolated from gametes of either mating type that had not been mixed with gametes of the opposite mating type (and therefore had not undergone flagellar adhesion). Importantly, flagellar adhesion was sufficient to activate the PTK, but not in gametes with a lesion in the microtubule motor protein, kinesin-II. A working model consistent with these data is that flagellar adhesion leads to a rapid stimulation of PTK activity in a pathway involving kinesin-II-dependent intraflagellar transport. PTK activation is followed, in turn, by activation of the adenylyl cyclase and gamete activation (Fig. 7).
For detection of the activated PTK in these experiments, it was essential to incubate the flagellar samples with ATP before carrying out SDS-PAGE and immunoblotting. As shown in Fig.  2, when ATP was left out of the in vitro assay, phosphorylated 105-kDa protein was not detected; similarly, in the experiments to examine the relationship between the extent of phosphorylation and assay time, no phosphorylation of the 105-kDa protein was detected in the 0 time point sample. Previous experiments from our laboratory and others (7,9,26,27) on protein kinase activity in Chlamydomonas have shown that flagella contain high levels of protein phosphatase activity, and the omission of phosphatase inhibitors in the in vitro assay used in the current experiments led to much reduced levels of phosphorylation (not shown). Use of the exogenous PTK substrate PGT was important as an independent confirmation of PTK activation and also should be useful for future experiments to identify and further characterize this adhesion-regulated PTK.
Several independent approaches indicated that activation of the PTK was a very early event during gamete activation: activation required flagellar adhesion, occurred within seconds after gametes of opposite mating type were mixed together, and took place at the same time or earlier than the increases in levels of cAMP shown previously (5,10). The results strongly indicated that PTK activation was not dependent on increases in cAMP or cell fusion and was induced by flagellar adhesion per se. Thus, fus1-1 mutant gametes, which are capable of flagellar adhesion but not gamete fusion, showed adhesion-dependent activation of the PTK. Similarly, when gametes underwent flagellar adhesion in the presence of the inhibitor of cyclic nucleotide-dependent protein kinases, H8, gamete activation was blocked as assessed by the absence of cell wall loss and gamete fusion, but the PTK still was activated. And, while incubation of gametes of a single mating type in dibutyryl cAMP bypasses the requirement for flagellar adhesion and induces the cellular changes that normally accompany flagellar adhesion, the PTK was not activated in such artificially activated cells. In addition, inclusion of cAMP in the in vitro assay for PTK activity in flagella isolated from gametes of a single mating type did not lead to activation of the PTK. Whereas the evidence is compelling that PTK activation is not downstream of activation of adenylyl cyclase and that regulation of both activities is tightly coupled to flagellar adhesion, determining whether PTK activation is upstream of activation of the flagellar adenylyl cyclase or in a parallel signaling pathway remains to be achieved.
Activation of the PTK (Fig. 4) and increases in cAMP (5,9,10) are the earliest biochemical or cellular events that have been demonstrated after flagellar adhesion is initiated, occurring within seconds after cells are mixed together. Activation of the PTK (Fig. 5) and the increase in cAMP and gamete activation are blocked when cells undergo flagellar adhesion in staurosporine (9), suggesting that the activity of one or more protein kinases is essential for activation of the PTK and the adenylyl cyclase. Finally, adhesion-induced stimulation of the PTK (Fig. 6), increases in cAMP and gamete activation (10) fail in fla10-1 gametes, which have a lesion in the microtubule motor protein kinesin-II (13). The simplest model that explains the data is that a single signaling pathway is initiated by flagellar adhesion. Because cAMP-regulated events are downstream of the PTK, it is likely that the PTK is upstream of adenylyl cyclase. In addition, all of the cellular events characterized to date that compose gamete activation, including flagellar tip activation, movement of agglutinin molecules from the cell body to the flagella, cell wall loss, and activation of mating structures are downstream of activation of adenylyl cyclase. That is, all these cellular responses can be induced by incubation of gametes of a single mating type in dibutyryl cAMP. Therefore, if the PTK is part of a signaling pathway that is independent of cAMP, the events induced by that putative pathway have not yet been reported.
The observation that fla10-1 gametes at elevated temperature fail to activate the PTK (Fig. 6) is the second demonstration that kinesin-II plays a direct or indirect role in adhesioninduced signaling in Chlamydomonas (10,28). Microtubule motor proteins have been implicated in other signal transduction pathways in Chlamydomonas (10, 28) (reviewed in Refs. 3, 29, and 30) and it will be interesting to determine the role of kinesin-II function in coupling agglutinin interactions during flagellar adhesion to gamete activation. For example, kinesin-II-dependent movement of IFT particles might be required for maintaining proper levels of key signaling proteins in the flagella. Or, binding between mtϩ and mtϪ agglutinin molecules might trigger conformational changes in the interacting molecules that induce them to associate with kinesin-II or IFT particles as an early step in signal transduction. The results presented here should now make it possible to continue a detailed dissection of the molecular pathways that underlie cell-cell adhesion-induced signal transduction during fertilization in Chlamydomonas.