Reactivation at low ATP distinguishes among classes of paralyzed flagella mutants.

Paralyzed flagella (pf) mutants of Chlamydomonas have been distinguished by the inability of the intact cells to move. Demembranated flagella from these mutants are also immotile when reactivated under standard conditions, with millimolar ATP concentrations. Three of these pf mutants were previously found to be motile when reactivated under 3 alternate reactivation conditions: low ATP concentration (< or =50 microM); 0.1 mM ATP combined with >0.5 mM ADP; or 0.1 mM ATP combined with non-reactivating ATP analogs anthraniloyl ATP or methylanthraniloyl ATP. We have now surveyed all pf mutants in the Chlamydomonas Culture Collection and discovered that a great majority of these mutants can move under these alternate nucleotide conditions. Only pf22 and pf23, mutants missing multiple subsets of dynein arms, did not reactivate under those conditions. This suggests that the paralysis observed in most pf mutants is the result of inhibition by physiological ATP. Except for pf12, which has an abnormally symmetric bending pattern, all other pf mutants exhibit asymmetric bending patterns similar to wild-type. Previously, motility that was restored by the presence of suppressor mutations was found to lack the normal asymmetry of wild-type flagella or the suppressor by itself. The waveform of pf mutants at alternate reactivation conditions in the absence of suppressor shows that pf mutants with radial-spoke or central-pair defects are capable of asymmetric bending similar to wild-type. A complete radial-spoke/central-pair complex is not essential for the production of asymmetric bending patterns. Furthermore, this suggests that the symmetric waveform observed previously in suppressed pf mutants is due to the interaction between the pf and suppressor mutations.

Ability of paralyzed flagella mutants of Chlamydomonas to move.

Chlamydomonas mutants missing the central pair or radial spokes are paralyzed despite the fact that they have the full wild-type complement of functional dynein ATPases. We show here that these mutants can move under conditions of low ATP concentration, a combination of ATP and ADP, and a combination of ATP and ribose-modified ATP analogs. These conditions suggest an inhibitory role of ATP and that this inhibition can be relieved by ADP or analogs. The function of the central-pair/radial spoke complex may be to release this ATP inhibition in a controlled manner.

Characterization of a minus end-directed kinesin-like motor protein from cultured mammalian cells.

Using the CHO2 monoclonal antibody raised against CHO spindles (Sellitto, C., M. Kimble, and R. Kuriyama. 1992. Cell Motil. Cytoskeleton. 22:7-24) we identified a 66-kD protein located at the interphase centrosome and mitotic spindle. Isolated cDNAs for the antigen encode a 622-amino acid polypeptide. Sequence analysis revealed the presence of 340-amino acid residues in the COOH terminus, which is homologous to the motor domain conserved among other members of the kinesin superfamily. The protein is composed of a central alpha-helical portion with globular domains at both NH2 and COOH termini, and the epitope to the monoclonal antibody resides in the central alpha-helical stalk. A series of deletion constructs were created for in vitro analysis of microtubule interactions. While the microtubule binding and bundling activities require both the presence of the COOH terminus and the alpha-helical domain, the NH2-terminal half of the antigen lacked the ability to interact with microtubules. The full-length as well as deleted proteins consisting of the COOH-terminal motor and the central alpha-helical stalk supported microtubule gliding, with velocity ranging from 1.0 to 8.4 microns/minute. The speed of microtubule movement decreased with decreasing lengths of the central stalk attached to the COOH-terminal motor. The microtubules moved with their plus end leading, indicating that the antigen is a minus end-directed motor. The CHO2 sequence shows 86% identify to HSET, a gene located at the centromeric end of the human MHC region in chromosome 6 (Ando, A., Y. Y. Kikuti, H. Kawata, N. Okamoto, T. Imai, T. Eki, K. Yokoyama, E. Soeda, T. Ikemura, K. Abe, and H. Inoko. 1994. Immunogenetics. 39:194-200), indicating that HSET might represent a human homologue of the CHO2 antigen.

Structural and functional hierarchy of eukaryotic cilia and flagella.

There are now a variety of methods to investigate the morphofunctional aspects of eukaryotic cilia and cilia. These methods are useful for investigating the basic mechanism of eukaryotic axonemal mechanochemical function and understanding the function and interaction of its components. It is clear that the complex structure of eukaryotic axoneme requires the combination of all these techniques to unravel its mystery. The compositionally simple in vitro microtubule assays are crucial in investigating the functions of different dyneins within an axoneme. However, because such assays do not include other components of the axoneme and the important mechanical feedback present in a beating axoneme, reactivation of the entire structure will continue to play a basic role in the morphofunctional study of eukaryotic axonemes.

Regulatory role of nucleotides in axonemal function.

Axonemal sliding involves both sliding velocity and the extent of sliding, that is how many doublets slide. It is clear that axonemes cannot beat if all doublets were to slide simultaneously, thus sliding extent is important. Using the turbidimetric assay of sliding disintegration of Tetrahymena axonemes, we examined the sliding extent and th effect of APD, ATP, and ATP analogs on the sliding extent. Of course, ATP is necessary to produce sliding disintegration, but ATP alone did not produce extensive sliding disintegration. The additions of higher ATP concentration even in the presence of ADP inhibited sliding disintegration. We also observed sliding disintegration using ribose-modified ATP analogs, anthraniloylATP, and methylanthraniloylATP. The extent of sliding disintegration was proportional to the analog concentration. Thus in contrast to ATP, higher analog concentration was not inhibitory. These results indicate that high ATP concentration acts to inhibit the extent of sliding disintegration and that ADP relieves this inhibition. We propose a model in which the affinity of of multiple cooperative active sites are regulated by the binding of ATP or ADP to a regulatory site. This model provides a mechanism by which nucleotides regulate the extent of sliding necessary for effective axonemal bending.

Functional multiplicity of motor molecules revealed by a simple kinetic analysis.

We present a simple analytical solution for a kinetic model of motor molecule function with multiple arms. This model has a rate of motion proportional to the probability that all arms in a complex are detached from the cytoskeleton and, therefore, we refer to it as obligate cooperativity. The model has the form: v = Vmax/(1 + q/S)n, where Vmax is the maximum velocity, the product nq is the effective Michaelis constant at high [ATP], and n is the number of arms. Values of n = 2 and n = 1 give good fits to the heavy meromyosin and myosin S1 sliding velocity data, respectively, consistent with the number of active sites. Despite the complexity of the eukaryotic axoneme, beat frequency data from Chlamydomonas wild-type and oda mutants are also fit by this model.

Axonemes paralyzed by the presence of dyneins unable to use ribose-modified ATP.

Substrate analogs are useful for studying the structures of active sites and for distinguishing between similar enzyme activities. Fluorescent ribose-modified ATP analogs were used to investigate the functional differences between dynein ATPases. These analogs reactivate (support the movement of) sea urchin sperm axonemes, yet they do not reactivate wild-type Chlamydomonas axonemes. Surprisingly, the analogs reactivate the axonemes of mutants completely missing the outer arm dyneins. Competition experiments using ATP and these analogs provide strong evidence that the analogs bind to all dynein active sites but fail to release a subset of dyneins from rigor. We suggest that this subset of Chlamydomonas outer arm dyneins unable to use the analogs remains in rigor in the presence of the analogs and paralyzes the axoneme.

Sea urchin axonemal motion supported by fluorescent, ribose-modified analogues of ATP.

The axonemal motion supported by fluorescent ribose-modified analogues, anthraniloyl ATP (Ant-ATP) and methylanthraniloyl ATP (Mant-ATP), was investigated. Ant-ATP and Mant-ATP supported good vigorous motion. A detailed study of the movement shows that the maximum beat frequencies (Vmax) were significantly lower with the analogues. However, Michaelis constants (Km) for beat frequency were also significantly lower than with ATP. Thus the net effect of changes in these two parameters, Vmax/Km, was similar for ATP and Ant-ATP and higher with Mant-ATP. Thus these fluorescent analogues are good substrates for axonemal movement. The consistently higher value of Vmax/Km, a measure of substrate selectivity, with Mant-ATP over Ant-ATP suggests a feature of the ribose binding site. Other significant differences in the movement with the fluorescent analogues are quantified in terms of kinetic measures of sliding velocity and bend propagation velocity.

Cooperativity in axonemal motion: analysis of a four-state, two-site kinetic model.

A kinetic model for axonemal motion based upon a four-state mechanochemical cycle of dynein with two active sites is described. Our model analysis determines the pseudo-steady-state concentrations of enzyme species for specified rate constants, most of which are experimentally determined, with given substrate and product concentrations. The proportion of enzyme species in which both active sites are detached from the microtubule (denoted as "both detached"), numerically calculated from the model, appears to be proportional to experimental observations of flagellar beat frequency. This correlation between beat frequency and the both-detached enzyme species is maintained over a wide range of substrate concentrations and exhibited an apparent positive cooperativity at low substrate concentrations, which we call "obligate cooperativity." The unusual obligate cooperativity exhibited by flagellar beat frequency parallels that seen in the calculated proportion of the both-detached enzyme species and is interpreted as a requirement for a molecule of substrate to bind to each active site in a multimeric dynein in order to produce oscillatory motion. Furthermore, the proportion of the both-detached enzyme species correlates with experimentally observed changes in beat frequency with a nucleotide analog and with product inhibition.

Membrane hyperpolarization activates trout sperm without an increase in intracellular pH.

Sperm from trout, like other sperm, are immotile in the seminal tract and initiate motility upon dilution into an appropriate fertilizing environment. Trout sperm motility is inhibited by high extracellular [K+] and can be activated by dilution of extracellular [K+]. Activation of trout sperm by the dilution of extracellular [K+] suggests regulation by membrane potential. Using the membrane potential-sensitive fluorescent dye 3,3'-dipropylthiocarbocyanine iodide (diS-C3-(5)) we directly measured the K+ contribution to the membrane potential. Manipulating the membrane potential with Cs+ and the ionophore valinomycin can override K+ regulation. We show that trout sperm can also be activated in the presence of inhibitory [K+] by the addition of divalent cations. Activation by divalent cations is explained by the cations' ability to mask membrane surface potential and thus alter the potential sensed by membrane voltage sensors. Using the surface potential-sensitive dye, 1-anilino-8-naphthosulfonate (ANS), we directly measure the divalent cations' ability to mask surface potential. We propose a model where membrane hyperpolarization is the trigger that initiates the cascade of events leading to trout sperm activation. An increase in intracellular pH has been suggested to be a conserved step in the activation of sperm motility. We show that increasing intracellular pH by procedures that activate sea urchin and mammalian sperm does not activate trout sperm. In contrast, there is a decrease in intracellular pH upon activation of trout sperm motility. Artificially decreasing intracellular pH is not sufficient for activation of motility in trout sperm in an inhibitory [K+]. Thus, unlike some other sperm, changes in intracellular pH do not regulate trout sperm motility.

ATP analogs substituted on the 2-position as substrates for dynein ATPase activity.

In contrast to previously studied ATP analogs, the two-substituted ATP analogs, 2-N3 ATP and 2-Cl ATP were good substrates for dynein ATPase. The Vmax for hydrolysis of both analogs was significantly higher than for ATP and the Km for both analogs was comparable to ATP. The higher hydrolytic rate for the analogs might be explained by a faster dissociation rate of the diphosphate product. This interpretation is supported by measurements of the dissociation rate of the inhibitor, vanadate. The estimate dissociation rate of vanadate with the analogs as substrate is approx. 2-fold higher than with ATP as substrate. These data together with previous studies on a variety of ATP analogs suggest that the 6-amino group on adenine is important for recognition by dynein and that the anti-conformation of the adenine, favored by 2-substituents, is the favored conformation of the nucleotide.

Mechanochemical coupling in eukaryotic flagella.

Quantitative analyses of ATP hydrolysis coupled to movement of eukaryotic flagella is important for understanding the relationship between ATP hydrolysis and movement. The difference in ATPase activity between intact motile axonemes (that is the cytoskeletal core of flagella) and homogenized or immotile axonemes has been assumed to be coupled to movement. However, recent findings on rates of steps in the dynein ATPase cycle and the effect of interaction with microtubules on those steps call for reassessment of movement-coupled ATPase. From these studies, it is clear that dynein ATPase activity is not as tightly coupled to interaction with microtubules as myosin ATPase activity is coupled to interaction with actin. The method by which axonemal movement is inhibited will critically affect the interpretation of difference in ATPase activity. If the homogenization or similar methods uncouple dynein, the difference in ATPase activity is not a useful measurement. Greater understanding of the relationship between dynein kinetics and axonemal movement may be obtained by use of conditions and substrates with known effects at specific steps in the dynein mechanochemical cycle and quantitating their effects on movement.

2-Chloro adenosine triphosphate as substrate for sea urchin axonemal movement.

The 2-substituted ATP analog 2-Chloro ATP was tested for its capacity to support axonemal movement. The movement of sea urchin axonemes reactivated with 2-Cl ATP appeared very similar to that with ATP. Detailed waveform analysis indicated that bend angle and shear amplitude were not significantly different for ATP and 2-Cl ATP. Although wavelength differs at particular nucleotide concentrations, if normalized to the beat frequency, it is similar for ATP and 2-Cl ATP. The main difference in the movement with the two analogs was seen in beat frequency and sliding velocity. The Vmax for beat frequency and mean sliding velocity was lower for 2-Cl ATP. The apparent Km for beat frequency and sliding velocity was much lower for 2-Cl ATP. The ratio of these two effects, that is, (Vmax/Km) is higher for 2-Cl ATP. Thus 2-Cl ATP is a good substrate for axonemal movement. The significantly lower Km of 2-Cl ATP was also demonstrated by its ability to support oscillatory motion at concentrations below that for ATP. The observations identify the structures and conformation of substrate necessary to support axonemal movement.

Kinetics of vanadate dissociation: estimation of the rate by inhibitor inactivation.

Vanadate (+5) is a potent inhibitor of a variety of ATPases including dynein ATPase. We describe a method useful for estimating the functional dissociation rate of vanadate from the active site which does not rely on classical physical separation techniques. The method involves spectrophotometrically monitoring the enzymatic activity as the inhibitor dissociates from the enzyme and is inactivated by norepinephrine. Norepinephrine effectively reverses vanadate inhibition by reducing vanadate (+5) to oxovanadium (+4). This reduction by norepinephrine is sufficiently fast for these purposes--addition of vanadate after norepinephrine shows no inhibition of ATPase activity. The mathematical estimation procedure is generally useful for estimation of dissociation rates of other reversible inhibitors which can be quickly inactivated after dissociation from the enzyme. The rate of dissociation of vanadate from dynein with ATP and 2-N3ATP as substrates using this method was estimated to be in the ranges 0.0023-0.0042 and 0.0057-0.0075 s-1, respectively. These rates permit estimation of the rates of vanadate association with dynein by using the reported dissociation constant for vanadate. The results are consistent with the very fast and potent inhibition of dynein ATPase activity observed.

Activation of the dynein adenosinetriphosphatase by microtubules.

Previous work has indicated that following the rapid adenosine 5'-triphosphate (ATP) induced dissociation of the microtubule-dynein complex, the rate-limiting step in the ATPase cycle is product release [Johnson, K. A. (1983) J. Biol. Chem. 258, 13825-13832], which occurs at a rate of approximately 2-6 s-1. In this report we complete the analysis of the ATPase cycle by examining the effect of microtubules on the rate of product release. For these studies we used repolymerized Tetrahymena axonemal microtubules and microtubule-associated protein (MAP) free bovine brain microtubules which were shown to be free of any measureable ATPase activity. Tetrahymena 22S dynein bound to these microtubules predominantly by the ATP-sensitive site and at a rate giving an apparent second-order rate constant of (0.2-1) X 10(6) M-1 s-1, which is 50-fold greater than the rate observed with brain microtubules containing MAPs. ATP induced the rapid dissociation of the microtubule-dynein complex with an apparent second-order rate constant vs. ATP concentration equal to 1.6 X 10(6) M-1 s-1; this value is only slightly lower than that observed in the presence of MAPs. After the ATP-induced dissociation, the dynein reassociated with the microtubules following a lag period due to the time required to hydrolyze the ATP. The duration of the lag time for reassociation decreased with increasing microtubule concentration, suggesting that microtubules increased the rate of ATP turnover. Direct measurements at steady state showed that the specific activity of the dynein increased with increasing microtubule concentration.(ABSTRACT TRUNCATED AT 250 WORDS)

Bending patterns of Chlamydomonas flagella: II. Calcium effects on reactivated Chlamydomonas flagella.

Ca2+ has profound effects on the movement of cilia and eukaryotic flagella, including those of Chlamydomonas. Two clear changes seen in Chlamydomonas flagella with changes in Ca2+ are beat frequency and symmetry. Photographic and computer assisted analysis of flagellar bending patterns on a uniflagellate mutant of Chlamydomonas have been used to examine details of the effects of Ca2+ on the movement of ATP-reactivated, demembranated flagella. In addition to the forward mode bending pattern seen at low Ca2+ concentrations (10(-9)M), which has a frequency of about 50 Hz and the reverse mode bending pattern seen at high Ca2+ concentrations (10(-4)M) with a frequency around 70 Hz, we carefully examined bending patterns in the intermediate Ca2+ concentration range of 1-6.5 X 10(-6)M. In this intermediate range, the bending patterns have significantly reduced asymmetry and slightly increased frequency, compared to the motility observed at low Ca2+ concentrations. These observations indicate that changes in these two parameters of motion do not occur in parallel and suggest that the effects of Ca2+ may be a multicomponent process. Physiologically, these changes in the beat pattern at intermediate Ca2+ may signal either 1) the beginning stages of transition to the symmetrical, high-frequency beating seen at high Ca2+, or 2) a more normal forward mode motility for the trans flagellum as suggested by Kamiya and Witman [1984]. No large amplitude bending patterns associated with transitions between forward and reverse mode beating in intact cells were seen at the intermediate Ca2+ concentrations.