Expression of Conventional and Unconventional Actins in Chlamydomonas reinhardtii upon Deflagellation and Sexual Adhesion

Abstract

Chlamydomonas has two actin genes, one coding for a conventional actin and the other coding for a highly divergent actin. The divergent actin NAP (for “novel actin-like protein”) is expressed only negligibly in wild-type cells but abundantly in a null mutant of conventional actin, the ida5 mutant. The presence of the dormant NAP gene suggests that NAP may also have its own function in wild-type cells under some conditions. However, no specific functions have been suggested. In this study, we examined the expression of actin and NAP in wild-type and ida5 cells under conditions where actin function has been shown to be important. We found that deflagellation induces the expression of NAP as well as that of actin in wild-type cells. The expressed NAP becomes localized to the regrown flagella, apparently without being associated with dynein. Mating of gametes also increased the expression of actin in wild-type cells and that of NAP in ida5 cells, resulting in accumulation of these proteins in flagella (in both wild-type and ida5 cells) and the fertilization tubule (only in wild-type cells). However, it did not induce significant NAP expression in wild-type cells. These and other observations suggest that the expression of actin and NAP mRNAs is controlled by two discrete mechanisms and that NAP plays a role in flagellar formation in wild-type cells.

Actin in most eukaryotes has highly conserved amino acid sequences and almost identical biochemical properties (23). Chlamydomonas reinhardtii has a single gene that codes for an actin 89% identical to rabbit skeletal muscle actin (24). However, in addition to the gene coding for this conventional actin, Chlamydomonas has a gene coding for a divergent actin (NAP, for “novel actin-like protein”) with sequence identity as low as 64% to the conventional actin (11, 15). Phylogenetic analysis indicates that this protein is a divergent member of the actin family rather than one of the actin-related proteins that have functions distinct from actin. Proteins similar to NAP have been found only in Volvocalean green algae (T. Kato-Minoura et al., unpublished data). Although some protozoa have multiple unconventional actins (4), Chlamydomonas is the only known organism that has genes for both conventional and unconventional actins, each present in a single copy. Naturally, it is an interesting question to ask how these two actins divide their roles in the cell and how the expression of their genes is controlled.

In wild-type Chlamydomonas, conventional actin is localized to the furrow region in dividing cells (8) and the fertilization tubule, i.e., a process formed in mating-type-plus (mt⁺) gametes during the mating reaction (2, 3). In addition, it is contained in some flagellar inner-arm dyneins as a subunit (19). Among those actin-containing structures, only the fertilization tubule can be stained with fluorescently labeled phalloidin, a probe for F-actin, suggesting that filamentous actin, if any, may be present in very small amounts within the cellular cytoplasm.

The unconventional actin NAP was first identified in an ida5 mutant, with a null mutation for conventional actin. This mutant lacks a subset of actin-containing inner-arm dyneins and is deficient in production of the fertilization tubule, but its growth rate and manner of cytokinesis are perfectly normal. This is probably because NAP is abundantly expressed in this mutant and functions as a substitute for actin. Strikingly, NAP is expressed in wild-type cells only in negligibly small amounts; both Western and Northern blot analyses using specific probes showed only extremely faint signals for NAP (11). We speculate that NAP expression in wild-type cells is suppressed by the presence of conventional actin. This is because, in an experiment using ida5 mutants transformed with the conventional actin gene, the amounts of NAP and actin in the transformed cells showed an inverse relationship (17). Hence, NAP may be expressed only when actin is not fully available in the cell. In other words, NAP might well be regarded as a spare actin recruited only when conventional actin is lost by mutation. However, from an evolutionary viewpoint, it seems highly unlikely that such a protein is preserved only for backing up possible actin loss. We may therefore reasonably expect that NAP plays an important role of its own in wild-type cells as well.

In this study, we wanted to obtain clues to the function of NAP in wild-type cells by finding conditions under which NAP is expressed. This approach was taken because we cannot directly examine the NAP function in Chlamydomonas, since no methods for targeted disruption or complete inactivation of a particular gene have been developed for this organism. We show here that deflagellation induces the expression of NAP in wild-type cells as well as in ida5 cells. Deflagellation of wild-type cells also induces actin expression, although it is less marked than the induction of NAP expression. In contrast, the mating reaction promoted actin expression in wild-type cells and NAP expression in ida5 cells but only negligibly induced NAP expression in wild-type cells. Therefore, deflagellation and the mating reaction appear to control the expression of actin and NAP through different pathways. Indirect-immunofluorescence observations indicated that NAP can be transported into flagella apparently independently of axonemal dyneins. These results suggest that NAP plays a role in the flagellar formation in wild type as well as ida5 cells distinct from its function as a dynein subunit.

MATERIALS AND METHODS

Strains and growth medium.

C. reinhardtii strains used in this study were wild type (strain 137c) and ida5t (10). All cells were grown in Tris-acetate-phosphate (TAP) medium with aeration (7). For preparation of gametes, cells grown on an agarose TAP plate for 4 days under continuous light were suspended in nitrogen-free medium (9). After incubation for 5 hr with gentle shaking, the mating ability of the cells was checked by mixing cells of opposite mating types.

Flagellar regeneration and isolation.

Vegetative cells grown in TAP medium to a concentration of 5 × 10⁶ to 1 × 10⁷ cells/ml or gametes prepared as above were deflagellated by pH shock (26). The deflagellated cells were kept gently stirred with a magnetic stirrer at room temperature. The motility of most cells was recovered by 80 to 90 min after deflagellation.

For protein analysis of newly formed flagella, the regenerated flagella were detached by treatment with dibucaine and harvested by centrifugation (27). The flagella suspended in buffer A (30 mM sodium HEPES, 5 mM MgCl₂, 1 mM dithiothreitol, 0.5 mM EGTA [pH 7.4]) were centrifuged at 1,500 × g for 10 min while underlaid with buffer A containing 25% sucrose (12). The flagella at the interface and the upper layer were recovered and centrifuged again at 15,000 × g for 5 min. The pellet was suspended in buffer A containing 0.5% Nonidet P-40 and centrifuged at 15,000 × g for 10 min. The resultant pellet was suspended in a volume of buffer A the same as that of the supernatant. These samples are referred to as membrane and matrix (M&M) and axoneme fractions. Flagella from gametic cells were obtained as above.

Protein determination.

To determine the ratio of the amounts of protein in the flagellum and the cell body, cells were fractionated into flagella and cell bodies and dissolved in 6 M urea. The protein concentration in each sample was then determined using a protein assay kit (Bio-Rad, Hercules, Calif.).

Activation of gametes.

Equal numbers of mt⁺ and mt⁻ gametes were mixed and allowed to mate at 24°C under illumination. The mating efficiency, checked 30 min after mixing, was greater than 90% in every experiment. To initiate the mating reaction with isolated flagella, gametic cells were mixed with the flagella from gametes of the opposite mating type (5). For activation of gametes by elevating the intracellular cyclic AMP (cAMP) level, gametic cells were incubated with 15 mM dibutyryl-cAMP (db-cAMP) and 1 mM 3-isobutyl-1-methylxanthine (IBMX) for 60 min (18).

Quantification of transcripts.

The amount of transcripts within the cell was estimated by Northern hybridization. Total RNA was isolated from cells with TRIzol Reagent (Life Technologies, Inc., Rockville, Md.) as specified by the manufacturer. A 3-μg portion of each RNA sample was denatured by mixing with 3% formaldehyde and loaded on a nylon membrane (Biodyne B; Nihon Pall, Ltd., Tokyo, Japan) using a slot blot apparatus (20). The probes used for detection of actin and NAP transcripts were cDNA fragments from 3′-untranslated regions of the respective genes (11, 24). Hybridization was performed using a³²P-labeled probe as described previously (11). Hybridized signals were detected and quantified with a BAS-2500 imaging analyzer (Fuji Photo Film Co., Tokyo, Japan). The actin/NAP transcripts were quantified against calibrated signals from known amounts of plasmid DNAs carrying these genes.

SDS-PAGE and Western blotting.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed by the method of Laemmli (13). The protein concentration in each sample was determined using a Bio-Rad protein assay kit. After SDS-PAGE, the proteins were transferred to a polyvinylidene difluoride membrane (Immobilon; Millipore, Bedford, Mass.) by the method of Towbin et al. (25) and reacted with antibodies. Antisera specific for the N-terminal sequences of conventional actin (24) or that of NAP (11) were affinity purified using fusion proteins expressed in Escherichia coli (17). Monoclonal antibody specific for p28 was a generous gift from G. Piperno (14).

Fluorescence microscopy.

For immunofluorescence microscopy, cells were fixed either by the cold methanol method as described previously (10), or, for detergent-extracted cells, by the following method. The cell suspension was mixed with an equal volume of Nucleus buffer (10 mM MgCl₂, 2.5 mM KCl, 3.7 mM EGTA, 6.7 mM Tris-HCl [pH 7.5]) (28) containing 0.2% NP-40 and 12% hexylene glycol. At 3 min after mixing, the cells were fixed by adding 3 volumes of Nucleus buffer containing 8% paraformaldehyde and 0.2% glutaraldehyde. Indirect immunofluorescence microcopy was performed by the method of Sanders and Salisbury (21).

To detect F-actin signals, cells were fixed with 3.7% formaldehyde-0.1% glutaraldehyde or demembranated and fixed as described above. The cells were then stained with BODIPY-phallacidin (Molecular Probes, Inc., Eugene, Oreg.).

RESULTS

NAP gene expression in wild-type cells after deflagellation.

We examined by Northern hybridization the change in abundance of the NAP mRNA in wild-type cells under various conditions. Compared with the expression of actin, the expression of NAP mRNA was always extremely low irrespective of the phase of circadian cycle in vegetative cells grown under a 12-h light-12-h dark cycle. Nor did it increase when vegetative cells were differentiated into gametes (data not shown). However, we found that the amount of the NAP mRNA greatly increased when vegetative cells were deflagellated (Fig. 1B). The level of actin mRNA also increased after deflagellation, but it was less marked (Fig. 1A). In ida5 cells, in which the NAP mRNA level was originally higher than in the wild type, deflagellation induced a further threefold increase (Fig. 1B). In both wild-type and ida5 cells the mRNA expression decreased to the original level 80 to 100 min after deflagellation.

FIG. 1.

Expression of actin and NAP genes after deflagellation. Vegetative and gametic cells of both wild-type and ida5 strains were deflagellated by acid shock, and RNA was isolated from the cells at the indicated time after deflagellation. Actin and NAP mRNA levels were quantified by Northern hybridization as described in Materials and Methods. (A and C) Change in the amount of actin mRNA after deflagellation in wild-type cells. (B and D) Transient increase in the amount of NAP mRNA in wild-type and ida5 cells after deflagellation.

Transient expression of NAP mRNA was also observed on deflagellation of gametic cells, although the increase in wild-type gametes was modest (Fig. 1D). The slower accumulation of messengers in gametic cells (Fig. 1C and D) is consistent with the previous report that the flagellar regeneration and protein synthesis after deflagellation are slower in gametic cells than in vegetative cells (16).

Localization of NAP in regrown flagella.

The above results suggest that NAP might be a component of newly formed flagella. As shown in Fig. 2A, wild-type flagella regrown within 2 h after amputation contained a certain amount of NAP, indicating that NAP expressed upon deflagellation is transported into newly formed flagella. The amount of NAP in flagella was estimated to be ∼1% of the total amount of NAP expressed in the cell, based on the intensity of the NAP bands in SDS-PAGE and Western blot patterns (Fig. 2B). When the flagella were demembranated with 0.5% NP-40 and centrifuged at 15,000 × g for 10 min, most of the NAP remained in the precipitate (Fig. 2A). Hence, it is likely that NAP was contained in the axoneme. Interestingly, Western blot analysis indicated that the amount of the axonemal NAP decreased with time (Fig. 2C), suggesting that the axonemal NAP is gradually removed. Since the amount of actin did not show a concomitant increase (Fig. 2C), NAP is apparently lost without exchanging with actin.

FIG. 2.

Localization and lifetime of NAP expressed after deflagellation. (A) Wild-type cells were deflagellated by acid shock and allowed to produce flagella for 2 h. The flagella were then isolated and treated with the detergent NP-40. The soluble fraction (M&M) and the insoluble fraction (axoneme) were subjected to SDS-PAGE and transferred to a polyvinylidene difluoride membrane. The blot was probed with anti-NAP antibody. CBB, SDS-PAGE gel stained with Coomassie brilliant blue. (B) Protein samples of cell bodies and flagella were prepared 2 h after deflagellation and analyzed by Western blotting. Each lane contained the same amount of protein (5 μg). Quantification of the total protein in the cell body and flagellar fractions in this preparation gave a protein ratio for the flagellum to the cell body of 2.1%. Since the band intensity of the NAP in flagella with respect to that of the NAP in the cell body was roughly 50%, we estimated the amount of NAP transported into newly formed flagella to be about 1% of the total amount of NAP expressed in the cells. (C) NAP in flagella (top panel) isolated at various time after deflagellation was detected by Western blot analysis. 0.0 indicates flagella isolated before the deflagellation treatment. As a control experiment, actin (middle panel) and p28 (bottom panel) were also detected with antibodies specific to these proteins.

Actin and NAP expression upon fertilization.

Since the above experiments suggest that NAP mRNA is expressed in wild-type cells under conditions in which actin expression is enhanced, we next examined NAP expression during fertilization, an event in which actin is known to play an important role. When wild-type mt⁺ and mt⁻ gametes were mixed and allowed to mate, the amount of actin mRNA rapidly increased ∼10-fold (Fig. 3A). Similarly, mating between the ida5 gametes resulted in a dramatic increase in the level of NAP mRNA (Fig. 3B). However, the expression of NAP mRNA in wild-type cells, which was originally weak, only slightly increased upon fertilization (Fig. 3B).

FIG. 3.

Expression of actin and NAP genes during mating. Gametic cells of wild-type and ida5 strains were mated (A and B) or mixed with flagella isolated from the gametes of opposite mating type (C and D). RNA was isolated from the gametic cells at various times after mixing. Actin mRNA (A and C) and NAP mRNA (B and D) were quantified as described in Materials and methods.

Various changes during the mating reaction are triggered by the adhesion between flagella of opposite mating types, which induces an increase in the intraflagellar cAMP concentration (6, 18). When gametic cells were activated by addition of flagella isolated from gametes of opposite mating types, the level of actin mRNA increased in wild-type cells and the level of NAP mRNA increased in ida5 cells (Fig. 3C and D). Again, the expression of NAP mRNA increased only slightly in wild-type cells. Except for some differences in the time course and magnitude, these changes in the actin and NAP abundance are similar to those observed in real mating reactions (Fig. 3A and B). These results confirm that the change in actin and NAP mRNA expression is triggered by flagellar adhesion. In addition, they indicate that enhancement of actin and NAP expression of the same magnitude occurs in gametes of both mating types.

The change in actin and NAP mRNA expression in mating gametes may be caused by an elevation of the intracellular cAMP level, as in various other events occurring during fertilization (18). We therefore next examined whether treatment of gametes with db-cAMP and IBMX increased actin and NAP mRNA levels. As shown in Fig. 4, the actin mRNA level in wild-type cells and the NAP mRNA level in ida5 cells increased after this treatment, although more slowly than when the gametes were stimulated by isolated flagella. In contrast, the level of NAP mRNA in wild-type cells did not show a noticeable change. The increased levels of actin mRNA in wild-type cells and that of NAP mRNA in ida5 gametes treated with db-cAMP/IBMX persisted for >80 min. In ida5 cells, the NAP mRNA level became abnormally high during the drug treatment. This is probably because the NAP expression in this mutant is free from suppression by actin (see Discussion).

FIG. 4.

Induction of actin and NAP gene expression in gametic cells upon treatment with db-cAMP and IBMX. Actin (A) and NAP (B) mRNAs were quantified as described for Fig. 3.

Localization of NAP in gametic cells treated with db-cAMP and IBMX.

The increased expression of actin and NAP mRNAs upon induction of the mating reaction may result in localization of these proteins to specific sites. To examine this possibility, we performed indirect fluorescence observations using specific antibodies. As observed previously, wild-type mt⁺ gametes treated with db-cAMP and IBMX produced fertilization tubules that contained actin (18). Indirect immunofluorescence microscopy with anti-NAP antibody revealed localization of NAP to the fertilization tubule and its basal region in the cell body of these cells (Fig. 5A and B). After the cell was extracted with NP-40, the NAP signal at the fertilization tubule was considerably weakened but not eliminated while the basal fluorescence completely disappeared (Fig. 5E and F). This observation suggests that while a large portion of NAP is concentrated in the cellular apical region in a soluble form, some of the NAP is contained in the fertilization tubule core. In contrast to the mt⁺ gametes, mt⁻ gametes did not display localized staining with the NAP antibody (data not shown).

FIG. 5.

Localization of NAP expressed in gametic cells treated with db-cAMP and IBMX. Cells were fixed with cold methanol (A, B, and I to L), fixed with formaldehyde-glutaraldehyde (C and D), or extracted with 0.1% NP-40 and fixed with aldehyde (E to H). The fixed cells were stained with anti-NAP antibody (A, B, E, F, and I to L) or BODIPY-phallacidin (C, D, G, and H). In ida5 gametic cells, as in vegetative cells (11), NAP was present in the midportion of the cell body before the treatment with db-cAMP and IBMX (I and J). After the treatment, it accumulated at the apical region of the cell body and in the flagella (K and L). (A, C, E, G, I, and K) Fluorescent image. (B, D, F, H, J, and L) differential interference contrast image. Bar, 10 μm.

Unlike the wild-type mt⁺ gametes, ida5 mt⁺ gametes treated with db-cAMP and IBMX did not form the fertilization tubule, despite the expression of an abnormally high level of NAP mRNA. Instead, they showed prominent NAP localization at the apical region of the cell body. In addition, a strong NAP signal was found along the length of the flagella (Fig. 5K and L). Such a strong flagellar localization of NAP was unexpected. The flagellar NAP signal indicates that a large amount of NAP is present without binding to dyneins attached to the outer doublets. In fact, the fluorescence greatly decreased upon detergent treatment of the sample, suggesting that the NAP is present in the membrane/matrix fraction (data not shown). To examine the manner of NAP accumulation in the flagella of ida5 gametes, flagella detached from the ida5 gametes before and after db-cAMP and IBMX treatment were extracted with NP-40 and separated into the M&M and axoneme fractions (see Materials and Methods). Consistent with the fluorescence microscope observation, Western analyses detected more abundant NAP in the flagella from db-cAMP- and IBMX-treated cells (Fig. 6). The amount of NAP increased in both M&M and axoneme fractions. To see whether the NAP in the axoneme and M&M fractions is associated with inner-arm dyneins, each fraction was probed with an antibody specific to p28, a subunit of NAP-containing inner-arm dyneins in ida5 cells (14). Figure 6 clearly shows that p28 was always absent from the M&M fraction and that its amount in the axoneme fraction did not change even when the amount of axonemal NAP increased. Thus, the excess NAP in the db-cAMP-treated gametes present in both M&M and axoneme fractions should not be associated with inner-arm dynein.

FIG. 6.

Transport of NAP into flagella independently of inner-arm dyneins. Flagella were isolated from ida5 gametes before and after treatment with db-cAMP and IBMX for 1 h and separated into the M&M and axoneme fractions. Each sample subjected to SDS-PAGE was transferred to polyvinylidene difluoride membranes. The blots were probed with anti-NAP antibody (middle panel) and anti-p28 antibody (right panel). CBB (left panel), SDS-PAGE gel stained with Coomassic brilliant blue. The amount of NAP in flagella, especially in the M&M fraction, increased after the db-cAMP treatment, whereas the amount of p28, a component of certain inner-arm dynein species, did not change with the treatment.

DISCUSSION

We have shown that NAP, a divergent actin, is transiently expressed upon deflagellation of wild-type Chlamydomonas. This is the first case where a particular cellular condition has been found to induce NAP mRNA in wild-type cells.

In wild-type cells, deflagellation enhances NAP mRNA and, less markedly, actin mRNA (Fig. 1). Most genes encoding flagellar proteins are up-regulated upon deflagellation. Typically, after flagellar amputation, the abundance of those mRNAs increases immediately, peaks in 40 to 50 min, and returns to the initial level by ∼90 min (22). Since actin is a subunit of several inner-arm dyneins (19), it is not surprising that its mRNA level transiently increases after deflagellation (Fig. 1A). However, the transient and marked expression of NAP mRNA upon deflagellation is rather unexpected since NAP has not been detected in wild-type flagella. The 5′-upstream region of the actin gene contains a special sequence called the “tub” box (24), which has been found in genes of many other flagellar proteins and is thought to be involved in the enhancement of the expression of these genes upon deflagellation (1). Intriguingly, the NAP gene also has this sequence motif (15). Thus, this characteristic alone suggests the involvement of NAP in some flagellar function.

Analyses of the actin and NAP mRNAs have provided evidence that their expression is regulated in different ways. Mating between mt⁺ and mt⁻ gametes or treatment of gametes with db-cAMP and IBMX greatly enhanced actin mRNA expression in wild-type cells and NAP mRNA expression in ida5 cells. Curiously, however, neither mating nor db-cAMP treatment significantly enhanced NAP mRNA expression in wild-type cells. Since the only difference between wild-type and ida5 cells is the presence of actin or its mRNA in the cytoplasm, we speculate that actin (or its mRNA) has a function in suppressing the expression of the NAP mRNA. This idea is in accordance with our previous proposal that the expression of the NAP mRNA is suppressed by the actin concentration within the cell, a proposal based on the observation that the extent of NAP expression is inversely proportional to the amount of actin in several ida5 cells transformed with actin genes (17) (see the introduction). Thus, wild-type cells express only an extremely small amount of NAP because actin is abundantly present. The results obtained in the present study suggest that the expression of NAP mRNA in mating wild-type gametes is also suppressed by the abundant actin. In contrast to the mating reaction of gametes, deflagellation of vegetative cells enhances the expression of both actin and NAP mRNAs in wild-type cells, suggesting that actin does not suppress NAP expression in this case. Thus, mating and deflagellation apparently stimulate the expression of NAP mRNA through different mechanisms. We propose that Chlamydomonas cells are equipped with two distinct regulatory systems controlling the expression of actin and NAP mRNAs, in which NAP mRNA expression is suppressed by the presence of actin in one system but not in the other (Fig. 7). The former system operates under normal conditions as well as during mating, whereas the latter system operates specifically during reflagellation. A complication is that NAP expression is enhanced only modestly when gametic cells are deflagellated, indicating that the controlling pathway differs somewhat between vegetative cells and gametes. However, it should be noted that the increase in NAP expression is still significant even in gametes.

FIG. 7.

Phenomenological model of expression control of the actin and NAP genes. The expression of actin and NAP are controlled by two distinct regulatory systems: one operating under normal conditions or during mating, and the other operating specifically during reflagellation. The former system tends to activate the expression of both actin and NAP, but the NAP expression in this case is suppressed by the expressed actin. Thus, wild-type cells contain only a negligible amount of NAP under normal conditions or during mating. In contrast, the latter system activates both actin and NAP expression without suppression by actin in vegetative cells or with only partial suppression in gametic cells (indicated by a broken line). Thus, both actin and NAP are transiently expressed upon deflagellation of wild-type cells.

An unexpected finding in the present study was that NAP became localized to flagella when its expression was enhanced by db-cAMP and IBMX treatment of the cells (Fig. 5). The NAP in the flagella was apparently not associated with inner-arm dyneins (Fig. 6). In a preliminary experiment, we observed that actin also becomes localized to the flagella when its expression is enhanced in wild-type cells (A. Ohara and M. Hirono, unpublished observation). Such dynein-independent flagellar localization of NAP or actin has not been detected in previous studies. Although it is not certain whether such flagellar NAP and/or actin has any specific function, we can at least say that flagella are equipped with a mechanism for transporting NAP and/or actin independently of inner arm dynein.

Wild-type mt⁺ gametes treated with db-cAMP and IBMX displayed NAP localization at the fertilization tubule and the basal region (Fig. 5A and B). Some NAP signal remained after the cell was extracted with NP-40 (Fig. 5E and F), indicating that at least part of NAP is incorporated into the core of the fertilization tubule. This observation raises the possibility that NAP forms copolymers with actin, although NAP has been thought to be unable to polymerize by itself because the mutant lacking actin lacks the fertilization tubule. However, further studies are necessary to determine the mechanism underlying the NAP accumulation at the fertilization tubule in mt⁺ wild-type gametes as well as that at the apical end in mt⁻ ida5 gametes. The appearance of NAP in the fertilization tubule did not apparently accompany transcription enhancement (Fig. 3). Whether it accompanies a change in translation level and whether the NAP in the fertilization tubule performs some specific function also await further studies.

In summary, the above findings show that (i) NAP is abundantly expressed after deflagellation of wild-type vegetative cells, (ii) expression of NAP and actin mRNAs is controlled in different ways such that only the NAP expression is deflagellation specific in wild-type cells, and (iii) NAP can be transported into the flagella without being associated with dynein. All these observations point to the possibility that NAP is involved in flagellar formation as a nondynein component. This may be the answer to the puzzling question of why the minor actin NAP, present in only negligible amounts under normal conditions in wild-type cells, is preserved in Chlamydomonas. Verification of this possibility awaits further studies, including isolation of mutants lacking NAP.