Portal de Periódicos da CAPES

Worm Sperm and Advances in Cell Locomotion

1996; Cell Press; Volume: 84; Issue: 1 Linguagem: Inglês

10.1016/s0092-8674(00)80068-9

1097-4172

Julie A. Theriot,

Biocrusts and Microbial Ecology

Across the animal kingdom, fertilization requires the encounter between a large stationary egg and small motile sperm. To maximize their likelihood of reaching the egg before their competition, sperm are extraordinarily specialized cells, generally consisting of little more than a haploid nucleus, mitochondria to generate energy, and a highly efficient movement engine. Almost all animal sperm are flagellated and seek the egg by swimming quickly through a liquid environment. Nematodes, however, produce sperm that move by crawling along solid substrates (2Foor W.E Biol. Reprod. (Suppl.). 1970; 2: 177-202Crossref PubMed Google Scholar). These roundworm sperm extend pseudopods that look and behave like the actin-rich pseudopods of a wide variety of cells ranging from free-living soil amoebae to human white blood cells. The crawling sperm appear by most criteria to be exploiting classic actin-based cell motility, with one important difference: the sperm contain practically no actin (11Nelson G.A Roberts T.M Ward S J. Cell Biol. 1982; 92: 121-131Crossref PubMed Scopus (105) Google Scholar). The pseudopods of amoeboid nematode sperm extend using a complex and dynamic cytoskeleton based on filamentous polymers of the 14 kDa major sperm protein (MSP), which bears no apparent similarity to actin. MSP is expressed only during spermatogenesis (8Klass M.R Hirsh D Dev. Biol. 1981; 84: 299-312Crossref PubMed Scopus (76) Google Scholar), and no homologs or analogs have yet been found in other cell types or in other animals. By all appearances, nematodes have evolved an entirely unique cytoskeleton used for this single purpose. The study of how these amoeboid sperm develop and move serves two independent goals. First, it addresses an interesting evolutionary question. Nematode fertilization takes place inside the ovary, densely crowded with eggs, so crawling sperm would clearly have an advantage over flagellated sperm in invading the ovary and maneuvering around inside it to reach the target eggs. But it is not at all clear why nematode sperm employ a novel crawling mechanism rather than using the well-established actin-based cytoskeleton that serves for all other forms of cell crawling during nematode development. Second, and perhaps more interestingly, nematode sperm demonstrate a form of crawling motility that is morphologically and dynamically analogous to actin-based motility, but with an entirely different cast of characters. The actin cytoskeleton, of course, is responsible for much more than just motility. It is also generally the primary determinant of cell shape, and it organizes and supports many cell type–specific specializations such as microvilli and stereocilia. Further, the actin cytoskeleton mediates cellular shape changes in response to external signals, and it rapidly rearranges itself at the end of every cell cycle to perform cytokinesis. Of the hundreds of actin-associated proteins in every cell, most are involved to some extent in more than one of these functions. It is a frequent topic of debate among motility researchers whether particular proteins or particular types of biochemical activity are absolutely required for actin-based motility per se or whether they are primarily involved in other actin-dependent processes. It has not yet been possible to reconstitute actin-based lamellipodial or pseudopodal protrusion in a purified system, so we cannot reliably define the minimum requirements for such movement, and the informed opinions of individual researchers vary widely. By comparing actin-based motility with MSP-based motility, though, we may be able to begin to understand the true basic mechanistic rules of crawling movement in biological systems. By analogy, bats and birds have evolved flight independently, and the striking structural and functional similarities between bat and bird wings delineate the rules of vertebrate flight. Thus, examining a form of whole-cell motility that does not involve actin at all may yield novel and critical insights into the long-standing problem of actin-based cell motility. Ascaris lumbricoides is a large, unpleasant, parasitic nematode that presently inhabits the intestines of approximately 1 billion people. The almost identical species Ascaris suum prefers to dwell in pigs. Both look very much like their smaller, useful cousin Caenorhabditis elegans, except that Ascaris individuals can be about 400 times larger. Sperm development and motility in C. elegans have been investigated by a number of groups, mostly using genetic and microscopic techniques (reviewed by9L'Hernault, S.W. (1996). In The Nematode C. elegans, Volume 2, D. Riddle, T. Blumenthal, B.J. Meyer, and J. Priess, eds. (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press), in press.Google Scholar). The large size and quantity of Ascaris sperm have made them even more amenable to study, and over the past 15 years a series of papers from a handful of laboratories have described the ultrastructure and behavior of these remarkable cells (reviewed by3Heath J.P Curr. Biol. 1992; 2: 301-303Abstract Full Text PDF PubMed Scopus (5) Google Scholar, 13Roberts T.M Stewart M Curr. Opin. Cell Biol. 1995; 7: 13-17Crossref PubMed Scopus (40) Google Scholar). Ascaris sperm are well-polarized cells about 20–25 μm across. The pseudopod is a broad, flat structure that extends about 15 μm from the cell body and is filled with a dense meshwork of filaments that excludes membranous organelles. The pseudopod protrudes at the front of the cell and forms close adhesive contacts with the solid substrate over which the sperm crawls, dragging the cell body containing the nucleus and mitochondria along behind it (10Nelson G.A Ward S Exp. Cell Res. 1981; 131: 149-160Crossref PubMed Scopus (34) Google Scholar, 15Sepsenwol S Ris H Roberts T.M J. Cell Biol. 1989; 108: 55-66Crossref PubMed Scopus (84) Google Scholar) (Figure 1a). This morphological description could also refer to any of a variety of cells that move using actin, except that here MSP and not actin is the primary component of the filamentous network that fills the pseudopod (15Sepsenwol S Ris H Roberts T.M J. Cell Biol. 1989; 108: 55-66Crossref PubMed Scopus (84) Google Scholar). MSP-containing amoeboid sperm move at rates of over 1 μm/s, as fast as the fastest animal cells that crawl with actin. MSP is the most abundant protein in the sperm, making up about 15% of total cell protein in both C. elegans (8Klass M.R Hirsh D Dev. Biol. 1981; 84: 299-312Crossref PubMed Scopus (76) Google Scholar) and Ascaris (10Nelson G.A Ward S Exp. Cell Res. 1981; 131: 149-160Crossref PubMed Scopus (34) Google Scholar). MSP is not homogeneous, but is a family of closely related 12–14 kDa basic polypeptides encoded by up to 60 genes (14Scott A.L Dinman J Sussman D.J Ward S Parasitology. 1989; 98: 471-478Crossref PubMed Scopus (29) Google Scholar). Purified MSP can be induced to polymerize in vitro in the presence of a high concentration of water-miscible alcohols, in a form of controlled aggregation (5King K.L Stewart M Roberts T.M Seavy M J. Cell Sci. 1992; 101: 847-857PubMed Google Scholar). Like actin, MSP monomers associate in a linear, head-to-tail fashion, forming 10 nm–wide filaments that consist of two subfilaments coiled tightly around each other (16Stewart M King K.L Roberts T.M J. Mol. Biol. 1994; 242: 60-71Crossref Scopus (26) Google Scholar). Unlike actin, there is no ATP-binding site on the MSP monomer, and no nucleotide is required for polymerization. Individual MSP filaments can spontaneously self-assemble into larger helical bundles, called "macrofibers." Similar macrofibers are seen within the pseudopods of intact sperm (7King K.L Stewart M Roberts T.M J. Cell Sci. 1994; 107: 2941-2949PubMed Google Scholar). Inside the sperm pseudopod, MSP assembly is spatially regulated. Filament assembly occurs at the leading edge of the pseudopod, where the nascent filaments also braid themselves into macrofibers. Individual branchings of the macrofiber complex can be observed in living cells, and these remain stationary with respect to the substrate as the sperm cell moves forward (15Sepsenwol S Ris H Roberts T.M J. Cell Biol. 1989; 108: 55-66Crossref PubMed Scopus (84) Google Scholar, 12Roberts T.M King K.L Cell Motil.Cytoskel. 1991; 20: 228-241Crossref PubMed Scopus (41) Google Scholar) (Figure 1b and Figure 1c). In other words, MSP filament polymerization occurs at the leading edge at the same rate as the sperm moves forward, and the speed of whole-cell motility is tightly coupled to the rate of filament assembly. The filaments disassemble close to the cell body at the rear of the pseudopod, also at a constant rate, so the cell maintains its overall shape as it moves. In the intact cell, the spatial regulation of MSP filament assembly and disassembly is partly due to the presence of a pH gradient over the breadth of the pseudopod. At the leading edge, the intracellular pH is about 0.2 U higher than it is at the rear. Apparently, the very basic MSP prefers to self-associate at higher pH and to disassemble under more acidic conditions (6King K.L Essig J Roberts T.M Moerland T.S Cell Motil.Cytoskel. 1994; 27: 193-205Crossref PubMed Scopus (43) Google Scholar). Lowering the intracellular pH by the addition of weak acids results in the total disassembly of the MSP macrofiber web. Removal of the acid allows the mesh to reform, starting from the leading edge (12Roberts T.M King K.L Cell Motil.Cytoskel. 1991; 20: 228-241Crossref PubMed Scopus (41) Google Scholar). This result raises the interesting possibility that the leading edge membrane might contribute to the nucleation of new MSP filaments. One missing link between the simple in vitro polymerization of pure MSP and the complex whole-cell motility of living sperm has been provided by the reconstitution of membrane propulsion in sperm extracts (Italiano et al., 1996 [this issue of Cell]). By adding ATP to an extract prepared from lysed sperm, 4Italiano J.E Roberts T.M Stewart M Fontana C Cell. 1996; 84 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar are able to observe polymerization and bundling of MSP filaments. Most remarkably, every elongating macrofiber has a membrane vesicle at one end that appears to be pushed forward as the fiber grows. The macrofibers remain stationary with respect to the substrate, elongating by filament assembly at the end where the vesicle is attached (Figure 1d and Figure 1e). Using antibodies directed against plasma membrane proteins and against a phosphotyrosine epitope found at the leading edge, they are able to demonstrate that all of the vesicles moving at the ends of MSP macrofibers are derived from the leading edge plasma membrane, the exact site of MSP filament nucleation in intact sperm. Fiber formation and vesicle propulsion require MSP, membrane components, cytosolic components, and ATP. In the cytosolic extracts, MSP filaments are able to assemble at rates similar to the rate of whole-sperm movement, but the filaments do not disassemble, presumably because there is no pH gradient present in this experimental system. The MSP filament assembly and vesicle propulsion observed by 4Italiano J.E Roberts T.M Stewart M Fontana C Cell. 1996; 84 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar appear to represent a true reconstitution of events taking place at the leading edge of the locomoting sperm. The movement of vesicles at the end of a column of bundled MSP filaments bears strong similarity to some uncommon forms of actin-based motility. Intracellular bacterial pathogens, including Listeria monocytogenes and Shigella flexneri, move rapidly in the host cell cytoplasm by inducing the assembly of actin filaments into a long, cylindrical "comet tail" strikingly like the MSP macrofibers seen by Italiano et al. (reviewed by17Theriot J.A Annu. Rev. Cell Dev. Biol. 1995; 11: 213-239Crossref PubMed Scopus (104) Google Scholarreferences therein). Actin filament assembly occurs only at the very front of the comet tail, immediately adjacent to the bacterial surface. The actin filaments in the tail remain stationary in the cytoplasm as the bacteria are propelled forward. It has been argued that this type of bacterial motility is equivalent to a simplified form of lamellipodial protrusion, where the bacterium imitates a short fragment of the lamellipod's leading edge (18Theriot J.A Mitchison T.J Trends Cell Biol. 1992; 2: 219-222Abstract Full Text PDF PubMed Scopus (52) Google Scholar). This argument is substantially strengthened by the present observation, which directly demonstrates that short fragments of the leading edge can generate comet tail–like structures and movement in an unrelated system. One particularly fascinating issue raised by this observation is whether directed filament assembly, by itself, can produce propulsive force. For any directed movement to occur in a biological system, there must be some mechanism for transducing the chemical energy of nucleotide hydrolysis into mechanical energy. There are at least three possible sources of force for actin-based lamellipodial protrusion: motor proteins, such as myosin, that hydrolyze ATP to move along actin filaments; osmotic swelling of the actin gel, which takes advantage of the energy stored in the chemiosmotic transmembrane potential; and actin polymerization itself, with its associated ATP hydrolysis (reviewed by1Condeelis J Annu. Rev. Cell Biol. 1993; 9: 411-444Crossref PubMed Scopus (388) Google Scholar). Vesicles propelled by MSP filaments in cytosolic extracts cannot be using osmotic forces since there is no intact plasma membrane. There is no evidence for any MSP-based motor protein in nematode sperm, and it seems highly unlikely that one would have evolved for this sole purpose. It is most probable, therefore, that filament assembly itself is driving movement. But how can MSP assembly create force? For actin, the unidirectionality of assembly-driven movement can be guaranteed by the actin ATPase, but MSP does not bind or hydrolyze nucleotide. 4Italiano J.E Roberts T.M Stewart M Fontana C Cell. 1996; 84 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar find that ATP is indeed required for this form of in vitro motility, though they have not identified how it is used. It is possible that the energy requirement comes into play at the surface of the membrane vesicle. Pure MSP will not polymerize except in the presence of 30% ethanol, but in this system, filaments are forming under physiological conditions. The identity and the mechanism of action of the putative nucleating factor present at the leading edge plasma membrane may address this conundrum. One possible way of using ATP hydrolysis energy would be by transient phosphorylation of MSP at elongating filament ends, and the localization of the phosphotyrosine epitope at the leading edge described by 4Italiano J.E Roberts T.M Stewart M Fontana C Cell. 1996; 84 (this issue)Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar might be consistent with this. Beyond the fundamental mystery of how force is generated at the leading edge and the identity and activity of the nucleating factor, there are numerous other unanswered questions relating to the mechanism of nematode sperm crawling motility. Though C. elegans sperm will move on glass, they locomote most efficiently on their natural substrate (11Nelson G.A Roberts T.M Ward S J. Cell Biol. 1982; 92: 121-131Crossref PubMed Scopus (105) Google Scholar). This indicates that the cells form some specific adhesive contacts with the substrate. For actin, indirect connections of the cytoskeleton to other cells or to the extracellular matrix are mediated through specific adhesion molecules such as integrins or cadherins. There must presumably be some analogous transmembrane protein in nematode sperm. But how does it attach to the MSP macrofiber web? How does it coordinate adhesion with the dynamic movement of the pseudopod? Can it function in signal transduction as well as in the physical connection? How is the polarity of the pseudopod initially determined? To address these questions, it will be necessary to identify the other protein components of the sperm pseudopod besides MSP. Here, the power of C. elegans genetics can come into play. These worms exist primarily as self-fertilizing hermaphrodites. Mutant hermaphrodites that have a defect in spermatogenesis are self-sterile, but can produce viable offspring when mated with wild-type males making normal sperm. This convenient arrangement allows for a simple screen to identify genes that are required for sperm development. Of the more than 60 such genes identified so far, about a dozen are associated with specific defects in sperm pseudopod formation, polarity, structure, or motility (reviewed by9L'Hernault, S.W. (1996). In The Nematode C. elegans, Volume 2, D. Riddle, T. Blumenthal, B.J. Meyer, and J. Priess, eds. (Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press), in press.Google Scholar). We can expect that the combination of biochemical studies in Ascaris and genetic studies in C. elegans will succeed in answering many of the interesting questions about nematode sperm motility over the next several years. At this point, combining what is known about the actin-dependent locomotion of amoebae and animal cells and the MSP-dependent locomotion of nematode sperm, we can already list some of the fundamental mechanistic rules that govern cell crawling. Pseudopods are always a specialization formed at front of the cell that pull the cell body along rather than pushing it. The pseudopod structure is most efficient when it is broad and flat, allowing the cell to maximize the area of the contact with its substrate. The propulsive structure is filled with filaments made up of a directional polymer. The polymer assembles and elongates primarily at front of the pseudopod, and polymer assembly appears to be nucleated by factors concentrated at the leading edge membrane. Polymer disassembly is biased to occur mostly at the rear of the pseudopod. Lateral polymer interactions, supplied by filament cross-linking and bundling proteins for actin and by MSP filament homotypic interactions for MSP, are necessary to maintain the structural integrity of the pseudopod and allow it to extend. The crawling motility of nematode sperm is of great use to us as well as to the nematode. Besides representing an interesting biological system in its own right, it allows us to appreciate which features of the actin cytoskeleton are critical for motility and which are most likely to be involved in other actin-dependent cell behaviors. Furthermore, the nematode sperm has provided strong support for the once-controversial idea that cytoskeletal filament assembly can directly generate motile force.