NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Riddle DL, Blumenthal T, Meyer BJ, et al., editors. C. elegans II. 2nd edition. Cold Spring Harbor (NY): Cold Spring Harbor Laboratory Press; 1997.

Cover of C. elegans II

C. elegans II. 2nd edition.

Show details

Section IIMitosis

A. The Wild-type Karyotype

The wild-type metaphase karyotype of C. elegans consists of 12 chromosomes in hermaphrodites, five pairs of autosomes and two sex chromosomes, and 11 chromosomes in males, which have a single sex chromosome. Mitotic stages are most easily visualized in developing embryos, where the chromosomes display many of the characteristic features of holocentric chromosomes that distinguish them from monocentric chromosomes. At metaphase, each chromosome as a whole orients and lies within the spindle, parallel to the equator of the spindle, whereas anaphase figures suggest that the entire chromosome moves broadside on toward the spindle pole (Albertson and Thomson 1982).

Cytological preparations of metaphase chromosomes are made by squashing younger embryos, which contain a higher proportion of dividing cells. The chromosomes lack a primary constriction and are rods, a few microns in length. The length decreases with the developmental age of the embryo. Few distinguishing features are apparent on the chromosomes. Variably, regions of the chromosomes may show some differential fluorescence after staining with nucleic-acid-specific fluorochromes, but there is no consistent banding pattern as is seen with mammalian chromosomes. After staining with Hoechst 33258, a dark band has been observed on the right third of linkage group V, coincident with hybridization of a probe for the 5S gene cluster (Albertson 1984a) and may reflect a lower affinity of the dye for the DNA sequence composition at this locus.

B. Holokinetic Organization

In electron micrographs of sections taken longitudinally through embryonic metaphase chromosomes, the kinetochore appears as a trilaminar plaque covering the poleward face of the chromosomes. It resembles the monocentric kinetochore in structure, as it is composed of inner and outer layers, 0.02 μm in width, separated by a more electron lucent layer, 0.03 μm in width. Analysis of electron micrographs of serial sections taken through dividing embryonic cells revealed that from 0 to 8 microtubules were attached to the kinetochore. When several microtubules were present, they were distributed along the kinetochore, with the greater number of microtubules being recorded on the longer kinetochores (Albertson and Thomson 1982). However, the actual number of microtubules per kinetochore is likely to be larger, since microtubule stabilization buffers were not used in the preparation of the specimens. If the number were on the order of tens, then for C. elegans chromosomes, which vary in length from 12 to 20 Mb, there would be approximately one kinetochore microtubule for every 106 base pairs of DNA, similar to the microtubule density observed for monocentric chromosomes (Bloom 1993). A conservation of microtubule density for the two types of kinetochore structure may reflect similar requirements for movement and microtubule capture during mitosis in higher eukaryotes (Pluta et al. 1995). Furthermore, the presence of multiple microtubule attachment sites at both the holocentric and the monocentric kinetochore requires coordination of the structural and cell cycle control mechanisms (Bloom 1993). For example, in order to prevent nondisjunction, the microtubule-binding sites must be properly spaced so that they are oriented toward one spindle pole at metaphase, and cell cycle checkpoints must be able to sense orientation and number of bound microtubules prior to the onset of anaphase (Pluta et al. 1995).

The holocentric organization of C. elegans chromosomes has also been demonstrated by experimentally inducing chromosome breaks and observing the maintenance of the fragments through several mitotic cell divisions (Albertson and Thomson 1982). Such tests are considered diagnostic of holocentric chromosomes (White 1973), since when monocentric chromosomes are treated in the same way, acentric fragments are formed that are not mitotically stable. Breakage of holocentric chromosomes, on the other hand, should not produce acentric fragments. Similarly, irradiation of monocentric chromosomes can result in the formation of dicentric translocation chromosomes, which suffer breakage-fusion-bridge cycles. In organisms with holocentric chromosomes, dicentrics are not formed, even when two entire chromosomes are fused. In C. elegans, segregation of both experimentally induced chromosome fragments and genetically characterized chromosome fragments, called free duplications (see below), appears cytologically to be nearly normal. Some abnormalities have been observed, however, suggesting that the fragments or small chromosomes behave somewhat differently compared to normal chromosomes. The fragments are often found at the edges of the metaphase plate in squash preparations, and they show evidence of abnormal segregation, including the presence of the fragments in the cytoplasm adjacent to the nucleus at interphase and lagging at anaphase (Albertson and Thomson 1982).

In addition to chromosomes and chromosomal fragments, extrachromosomal high-copy-number arrays of DNA injected into the germ line are transmitted through both meiosis and mitosis (Mello et al. 1991). The DNA arrays are propagated less efficiently than chromosomes and at different rates for different arrays formed from the same DNA. Rates of mitotic loss per cell division of 1−2 × 10−2 and 3 × 10−3 have been measured for two arrays (Herman et al. 1995; L. Miller and S.K. Kim, pers. comm.). In comparison, the reported rate of mitotic loss per cell division for several free duplications varies from 10−4 to 5 × 10−3 (Herman 1995 and see below). The structure of the extrachromosomal DNA arrays has not been fully characterized, but it seems to differ from that of chromosomes and free duplications. The arrays appear to lack telomeres, since a probe for the C. elegans telomere failed to hybridize to several different arrays, although the presence of telomeres on several free duplications could be demonstrated using this same probe (D.G. Albertson, unpubl.). These observations are consistent with other independent evidence suggesting that at least some extrachromosomal arrays are circular in structure (S.K. Kim, pers. comm.), and therefore arrays and chromosomes might be expected to differ in their segregational behavior. Other aspects of array structure remain to be determined. For example, it is not known if a kinetochore is assembled on the arrays or what proteins interact with the arrays to package the DNA.

The observation that any DNA can be propagated as an extrachromosomal array in C. elegans suggests that there may be no specialized sequences required for segregation of holocentric chromosomes. On the other hand, the fact that neither free duplications nor DNA arrays segregate with the fidelity of normal chromosomes suggests that arrays and free duplications lack certain features that promote mitotic stability of wild-type chromosomes. Whether some of these features will turn out to be cis-acting sequences with properties and functions similar to those of monocentric centromeres remains to be determined. Monocentric centromeres carry out several functions in mitosis. They mediate attachment of the chromosome to the mitotic apparatus via the kinetochore. The centromere is also the site of the last attachment of sister chromatids and therefore may be expected to contain proteins directing sister chromatid segregation, as well as proteins involved in cell cycle control (Pluta et al. 1995). The role of specific DNA sequences in these various aspects of centromere function is best understood for the small, point centromeres of some yeasts, whereas for the larger, regional centromeres, such as those found in higher eukaryotes, it is not clear which of the various DNA sequence classes located at the centromeric constriction perform the different centromeric functions. In monocentric organisms, certain aspects of mitotic segregation can also be accomplished in the absence of a centromere, including the nucleation of microtubules, and their assembly into bipolar arrays on chromatin in egg or early embryo extracts (Sawin and Mitchison 1991; A. Hyman, quoted in Raff and Allan 1996), and the transmission, albeit inefficient, of acentric fragments (Steiner and Clarke 1994; Brown and Tyler-Smith 1995; Murphy and Karpen 1995). Therefore, some mechanisms may exist that promote segregation in the absence of a normal centromere, and the fact that DNA arrays can be transmitted in C. elegans should not be interpreted as conclusive evidence that cis-acting factors are not required for holocentric chromosome segregation. Further study of both monocentric and holocentric chromosomes will be required to understand the role of cis-acting factors and different segregation mechanisms.

C. Mutant Karyotypes

Reciprocal translocations between many pairs of chromosomes have been described. They have been useful for a variety of genetic analyses, including the characterization of cis-acting features involved in pairing and crossing over (see below) and as genetic balancers (for review, see Edgely et al. 1995). In some cases, the position of the breakpoints generates translocation chromosomes that are morphologically distinct in the light microscope, and these have been used for cytogenetics (Albertson 1985). Two additional types of chromosome rearrangements can be stably propagated because of the holocentric organization of the chromosomes. They are small chromosome fragments, called free duplications, and translocations involving fusion of two entire chromosomes.

1. Chromosomal Rearrangements

The first fusion chromosome to be described in C. elegans was the mnT12(IV;X) chromosome, in which the right end of X is joined to the left end of IV (Sigurdson et al. 1986). Animals homozygous for mnT12(IV;X) are viable and fertile, indicating that all essential genes on X and IV have been retained on the translocation chromosome. However, telomeric sequences could not be observed at the breakpoint by fluorescent in situ hybridization with a telomere probe, and therefore have been eliminated or reduced below the sensitivity of the assay by the fusion event. The metaphase karyotype of mnT12(IV;X) homozygotes consists of ten chromosomes, with the mnT12(IV;X) chromosomes easily distinguished cytologically by their length (Fig. 1a).

Figure 1. Mitotic chromosome preparations with chromosome rearrangements.

Figure 1

Mitotic chromosome preparations with chromosome rearrangements. The holocentric structure of C. elegans chromosomes allows the stable propagation of chromosome fragments and (more...)

Another fusion chromosome, mnT13(I;X), has been characterized cytologically and consists of the left end of X joined to the right end of I (Albertson and Thomson 1993). Two observations suggest that some essential genes have been lost from linkage group I in the formation of the translocation. First, animals homozygous for mnT13(I;X) also carry an additional copy of linkage group I. Second, the ribosomal gene locus (which maps to the right end of linkage group I) appears to be partially deleted on the translocation chromosome, since hybridization of a probe for the ribosomal genes results in a fluorescent hybridization signal of reduced intensity on the mnT13 chromosome (Albertson and Thomson 1993; D.G. Albertson, unpubl.). The possible loss of other linkage group I gene sequences from mnT13 has not been investigated.

Recently, Hodgkin and Albertson (1995) described the generation of attached X chromosomes (X^X) in C. elegans. These chromosomes resulted from rare recombination events involving attached inverted duplications of the X chromosome, and they should therefore be composed of two X chromosomes attached at their left ends (see Fig. 6 in Hodgkin and Albertson 1995). Although animals with X^X chromosomes were viable and the X^X could be propagated through both mitosis and meiosis, the chromosome broke down frequently. This behavior differs from the behavior of attached chromosomes studied in other organisms (White 1973) and may be due to breakage occurring as a result of the kinetic activity of the chromosome ends during meiotic segregation of these holocentric chromosomes (see below). Similar behavior has been reported for the szT1(X) translocation chromosome (McKim et al. 1988).

Free duplications appear cytologically as small chromosomes or chromosome fragments (Fig. 1b) (Herman et al. 1976). Approximately half of the genetic map is covered by different duplications (Herman 1995). The free duplications are usually present in one copy per cell, and they are transmitted through both somatic and germ-line divisions with lower fidelity than a normal chromosome. This segregational behavior has been exploited for the generation of genetic mosaics (for review, see Herman 1995). Three general observations about duplication behavior have been made from the genetic mosaic studies. First, larger duplications tend to be mitotically more stable than smaller duplications. This behavior can be seen by comparison of the mitotic stability of duplications and their derivatives that have been made either smaller by deletion or larger by fusion with other duplications. The deletion derivatives show decreased mitotic stability, whereas the fusion derivatives show increased mitotic stability relative to the progenitor duplication(s). There are exceptions to this rule, however (McKim and Rose 1990; Hedgecock and Herman 1995). For example, the mitotic stability of qDp3(III, f) is about 20 times greater than that of sDp3(III, f), although the physical size of the duplicated region of linkage group III included in qDp3 appears to be equal to or less than that of sDp3 (Hedgecock and Herman 1995; Herman 1995; D.G. Albertson, unpubl.). Second, it has been observed that the germ-line transmission frequency and mitotic stability of the duplications are correlated, although the germ-line transmission of the duplication, ctDp2, is lower than expected from its mitotic stability (Hunter and Wood 1990). Third, free duplications have been observed to undergo spontaneous rearrangement (deletion) in the germ line at a high frequency compared to normal chromosomes (Herman 1984; McKim and Rose 1990, 1994; Villeneuve and Meyer 1990a).

Two general explanations have been offered as to why free duplications should be less stable than the normal chromosomes. Cytological observations on chromosome fragments led to the suggestion that the probability of microtubule capture may depend on chromosome length (Albertson and Thomson 1982). Therefore, the somatic loss of duplications, which has been observed to occur either by simple loss or by nondisjunction (Hedgecock and Herman 1995; Herman 1995), might be explained by failure of the duplication to attach to the mitotic apparatus or by mal-orientation on the mitotic spindle. It has also been suggested that the different mitotic stabilities of free duplications might be due to the fact that they are ring chromosomes, rather than linear chromosomes (McKim and Rose 1990, 1994; C.P. Hunter and W.B. Wood, pers. comm.). Whether free duplications are linear or ring chromosomes is at present not known, since the small size of the duplications makes it impossible to distinguish between these two structures by light microscopy. However, the demonstration of telomere sequences on two duplications of linkage group III, qDp3 and eDp6 (D.G. Albertson, unpubl.), suggests that at least some free duplications are linear. It is also possible that loss of free duplications may occur because of malfunctions occurring in interphase, including, for example, incorrect inclusion in the interphase nuclear organization or aberrant DNA replication.

The composition of most chromosome rearrangements has been determined by inspection of specific markers, usually by genetic mapping (Edgley et al. 1995), or in a few cases by mapping relative to the physical map by cytological or molecular methods (see, e.g., Kramer et al. 1988; Albertson 1993). The rearrangements have therefore been mapped with respect to only a small portion of the genome, and the possibility remains that the free duplications, as well as other chromosome rearrangements, are highly complex rearrangements. They may contain other, as yet undetected regions of the genome important for proper segregation and stability. For example, addition of telomeres to qDp3, a duplication of the central portion of linkage group III, may have occurred de novo or may have been formed by translocation of other chromosome ends. Determination of the composition of chromosome rearrangements with respect to the entire genome by chromosome painting (Chuang et al. 1994) or comparative genomic hybridization (Kallioniemi et al. 1992) may reveal some surprises, as well as information on how rearrangements are formed and the requirements for their stable meiotic and mitotic segregation.

2. Aneuploidy

Both whole chromosome and segmental aneuploidies are observed in C. elegans. Nondisjunction of the X chromosome gives rise to viable hermaphrodites that are trisomic for the X, but otherwise diploid (Hodgkin et al. 1979). These animals may be distinguished from wild type by their dumpy morphology. Animals trisomic for linkage group IV are also viable (Sigurdson et al. 1986). They have small brood sizes but are indistinguishable morphologically from wild-type diploids. Tetrasomics in C. elegans appear to be lethal, since progeny tetrasomic for the X, for example, are not recovered from appropriate crosses (Hodgkin et al. 1979). In contrast, four copies of some chromosomal regions have been observed, for example, in animals homozygous for a translocated duplication (Rogalski and Riddle 1988) or a single copy of a free duplication that itself contains two copies of a chromosomal region (e.g., +/+/eDp27).

D. Telomeres and Telomerase

The C. elegans chromosomes terminate in 4−9-kb blocks of the tandemly repeated sequence TTAGGC (Wicky et al. 1996), which is closely related to the telomeric repeats found in other, highly diverged eukaryotes and is most similar to the TTAGGG repeat characteristic of vertebrate and trypanosome telomeres (Zakian 1989). The TTAGGC sequence appears to be a general feature of nematode telomeres, since it is also found at the telomeres of the parasitic nematodes Ascaris and Parascaris (Müller et al. 1991; Zetka and Müller 1996). In addition to the telomeric blocks of TTAGGC repeats, the C. elegans genome contains numerous dispersed internal blocks of perfect (Cangiano and La Volpe 1993) and degenerate (D.G. Albertson, unpubl.) TTAGGC repeats concentrated in the terminal 30% of the chromosomes.

In C. elegans, the terminal TTAGGC repeats alone appear to be sufficient for the general chromosome capping functions attributed to telomeres. Of the 12 telomeres, 11 have been cloned, and sequence analysis of the subtelomeric regions reveals that the 11 telomeres do not have any sequences in common apart from the TTAGGC repeats (Wicky et al. 1996). Moreover, TTAGGC repeats are joined immediately adjacent to a nearly complete ribosomal DNA repeat unit at the telomere corresponding to the right end of chromosome I. The fact that each of the C. elegans subtelomeric regions is unique suggests that no specific DNA sequence apart from the TTAGGC repeats is required for basic telomere function. The absence of similarity among the C. elegans subtelomeric regions contrasts with the subtelomeric regions of other species, which often contain homologous and repeated sequences (Biessmann and Mason 1992; Kirk and Blackburn 1995). Although the C. elegans subtelomeric regions do not have sequences in common with each other, many do contain repetitive DNA. Three contain satellite-like tandem repeats, and five contain repeated sequences that cross-hybridize with internal genomic DNA fragments (Wicky et al. 1996).

The cloned telomeres represent an important resource for completion of the C. elegans physical map (Waterston et al., this volume), since most of the few remaining gaps in the map are located at the chromosome ends. Three of the cloned telomeres have been assigned to their chromosome ends of origin (Wicky et al. 1996; A. Rose et al., pers. comm.), and mapping of the remaining telomeric clones is in progress.

Recent experiments have provided molecular genetic evidence for telomerase activity in C. elegans. Wicky et al. (1996) demonstrated that a terminally deleted chromosome had acquired a new telomere and showed that this new telomere had arisen by de novo addition of telomeric repeats at a site that lacked preexisting TTAGGC repeats. The junction site contains three bases present in both the ancestral DNA sequence and the telomeric repeat; these three bases presumably acted as a primer for telomere addition by allowing limited pairing with the RNA template of the telomerase enzyme. This healing event is analogous to the new telomere formation that occurs during the developmentally regulated process of chromatin diminution in Ascaris (Tobler et al. 1992), where telomeric repeats are also added de novo at chromosomal sites that have one to four bases of overlap with the TTAGGC repeat (Müller et al. 1991; S. Jentsch and F. Müller, pers. comm.). Similar ambiguity at the junctions between subtelomeric satellite or ribosomal DNA repeat sequences and the TTAGGC telomeric repeats at several of the endogenous C. elegans telomeres further suggests that these telomeres may also have arisen by telomerase-mediated healing events and that de novo telomere formation by telomerase may play a part in genome evolution.

1. Mutations Affecting Chromosome Segregation

The mutation him-10(e1511ts) causes a temperature-sensitive defect in mitotic chromosome segregation (A.M. Villeneuve, unpubl.). In him-10 hermaphrodites shifted to 25°C during larval growth, oocyte nuclei contain widely varying numbers of chromosomes, some having many more chromosomes than normal and some having many fewer. Since C. elegans oocyte nuclei are paused in meiotic prophase, prior to the meiosis I division (Schedl, this volume), an aberrant number of chromosomes at this stage is clear evidence of defective chromosome segregation during the mitotic proliferation of the germ line. Further investigation will be required to determine the underlying malfunction (e.g., in assembly or function of the mitotic spindle) responsible for these errors in chromosome segregation.

It is likely that him-10(e1511) affects mitotic chromosome segregation in somatic cells as well as in the germ line. Although this has not yet been demonstrated for whole chromosomes, the him-10 mutation was found to cause a four- to sevenfold increase in the frequency of somatic mitotic loss of several free duplications, making it a useful tool for genetic mosaic analysis (Hedgecock and Herman 1995).

2. Mutations Affecting the Mitotic Cell Cycle

Mutations in several genes result in uncoupling of the chromosome and cell division cycles during postembryonic development. Newly hatched L1 larvae carrying these mutations are normal, perhaps because embryonic cell divisions are controlled by stored maternal products. Most postembryonic cell divisions are defective, resulting in a characteristic Sterile-Uncoordinated (Unc) phenotype since the development of the gonad and the later larval and adult nervous system requires cell division (Albertson et al. 1978; Sulston and Horvitz 1981). In one class of mutants, typified by lin-5 , nuclear division and cytokinesis fail but DNA synthesis continues, resulting in polyploid blast cells. In lin-5 mutants, many nuclei undergo abortive cycles of chromosome condensation and nuclear envelope breakdown, but metaphase and anaphase fail and the nuclear envelope reforms. In another class of mutants, typified by unc-59 and unc-85 , polyploid nuclei appear to arise both by failed nuclear division and by nuclear fusion following failed cytokineses. In lin-6 mutants, in contrast, cell divisions continue in the absence of DNA replication, producing smaller and smaller cells that eventually die.

Postembryonic cells of these Sterile-Unc mutants are unable to coordinate cell cycle events to ensure that they occur in the proper order; some aspects of the cell cycle continue unchecked despite the fact that events that normally precede them have failed. Cell division in C. elegans is governed by a classical checkpoint regulation program, however, as evidenced by the fact that pharmacological disruption of specific cell cycle events can elicit cell cycle arrest. For example, cell division in L1 larvae can be arrested prior to entry into mitosis (presumably in S phase) by treatment with the DNA synthesis inhibitor hydroxyurea (S. van den Heuvel, pers. comm.). Either the Sterile-Unc mutants cannot sense whether a prior event in the cycle has been completed or else their specific defects do not result in production of a signal capable of eliciting cell cycle arrest by a checkpoint mechanism.

In contrast to the Sterile-Unc genes, two genes have been identified that appear to be involved in regulating progression through the cell cycle: emb-29 , which is required for late embryonic mitoses (Hecht et al. 1987), and glp-3 , which is required for germ-line mitoses (L. Kaydyk et al., pers. comm.). Mutations in these genes cause an apparent arrest of the cell cycle prior to entry into mitosis, at the G2/M transition.

Copyright © 1997, Cold Spring Harbor Laboratory Press.
Bookshelf ID: NBK20136
PubReader format: click here to try


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...