Development of a somatic 'gene readout' organelle, the macronucleus, in the ciliate Oxytricha trifallax
by Kevin Williams, Tom Doak, David Witherspoon & Glenn Herrick
| |
|
Ciliates and Oxytricha: Oxytricha is large single-celled microbe, representative of stichotrich [hypotrich] ciliates. It is covered with bundles ["cirri"] of cilia, it uses cirri to "walk" in freshwater on the substrate (Fig. 1).
| |
Figure 1. Scanning electron microscopy of Oxytricha trifallax, showing cortical features such as the oral apparatus at the far right and cirri (bundles of cilia). Photo by GW Grimes
| |
 |
|
Each ciliate carries two different nuclei in its single cell, a large "macronucleus" (MAC) and a small "micronucleus" (MIC). Cells clonally proliferate by binary fission (Fig. 2). The MAC is the "somatic" nucleus, a temporary specialized gene expression organelle comprised of nearly pure gene DNA. The MIC's genes are never expressed (transcribed) directly and the MIC functions solely as a germline nucleus to carry genes from one sexual generation to the next (Fig. 2).
| |
Figure 2. Nuclei in the ciliate life cycle. Legend: Vegetative cells conjugate (top), diploid MICs perform meiosis, haploid gametic nuclei are exchanged between mates, zygosis of gametic nuclei, and mitotic duplications of zygotic nuclei. Development of a new MAC (bottom), which begins transcription (green halo) and directs the proliferation of the exconjugant into a clone of vegetative cells (left side).
| |
 |
|
The MAC is replaced by a new one following cell mating, when the cell begins its genetic life with a new dipliod zygotic complement in a new MIC. The new MAC develops from a mitotic copy of the new MIC; once the new MAC is competent to support the cell, the old MAC is destroyed (Fig. 3 shows the nuclei of an exconjugant in the midst of this 3-day developmental progression), and finally the new MAC programs the newly emergent vegetative cell to proliferate into a vegetative clone (Fig. 2) of cells (such an exconjugant-founded clone is called a "karyonide.").
| |
Figure 3. An exconjugate, nuclei stained with DAPI. This cell had recently separated from its mating partner and was in the midst of generated its new MAC (MAC anlage, "AN"). The anlage is in mid-polyteny, the parental old MAC ("oM") is in two pynotic fragments, being destroyed apototically. The new MIC ("mi") rests in a pocket of the surface of the new MAC.
| |
 |
|
MAC development: (Fig. 4): The MAC precursor ("anlage," auf Deutsch) begins with a genome identical to the new MIC, but development strips it of all but 5% of its sequences. The remainder consists nearly completely of functioning genes. These genes are packaged mostly one per tiny MAC chromosome. These chromosomes are created by massive chromosome breakage and telomere formation. The mature MAC carries ~20,000 kinds of chromosomes, with an average length of 2400bp; most carry only one gene (some carry two or three genes).
| |
Figure 4. Macronuclear development. Mitotis of the zygotic nucleus generates the new MIC and the MAC anlage, in which the new MAC develops (Figure 1, bottom). Initial amplification (polytenization) is followed by elimination of nearly all (95%) of the germline sequences; chromatids are fragmented and telomeres are formed on each end, generating mature mini-chromosomes ("gene-sized"); this collection of ~20,000 kinds of MAC chromosomes is further amplified to a final ploidy, avg. ~1000.
| |
 |
Our specific interests in MAC development:
• Sequence elimination by DNA splicing (review: Klobutcher and Herrick 1997). On average each germline sequence segment destined for a particular MAC chromosome is interrupted by ~3 interruptions that often interrupt functional open reading frames. These interruptions -- Internal Eliminated Sequences, or IESs -- are precisely and efficiently spliced out, stitching together the uninterrupted MAC-destined sequence. We have studied families of active transposons ("TBEs") that massively infest the germline, but are completely eliminated from the MAC anlage, and we have made the case that all other IESs are probably ancient residua of old transposon invasions. We remain puzzled by our observation (Witherspoon et al. 1997) that the genes of the TBEs are functioning under a strong purifying selection for their function (we see a dearth of non-synonymous codon mutations).
• Alternative chromatid breakage and multi-gene MAC chromosomes. We have intensely studied the "81-MAC" locus (e.g., Herrick et al. 1997), which is processed to generate a nested set of three MAC chromosomes which share a 1.6kbp common region (Fig. 5). This region bears a gene, and comprises the smallest of the three chromosomes (plus telomeres); the two larger chromosomes have gene-bearing "arms" attached to their copies of the common region. These are the first known examples of hypotrich MAC chromosomes that do not consist solely on one gene, and contradict the catchy phrase, "gene sized." (Seegmiller et al. 1997)
| |
Figure 5. Display of multiple Telomere Addition Sites at the right end of the 81-MAC common region by the technique of Ligation-Mediated MAC-End PCR.
| |
 |
• Chromosome breakage and telomere addition. Having devised a new technique, Kevin Williams has discovered an important phenomenon operating at the ends of the 81-MAC common region. The new technique is an anchored PCR procedure, in which an oligomer anchor is ligated to the 5'P of MAC DNA chains; the data are generated by the extension of a radio-labeled primer; the extension products are displayed on a "sequencing" gel in parallel with sequencing lanes across the MIC sequence which is broken and which has the telomere addition sites, TASs. Figure 6 shows an example of a display lane for the TAS region at the "right" end of the common region, along with the sequencing of a cloned display product. Note the large number of TASs; such a "multiple-TAS" region is also seen at the "left" end of the common region, in contrast to the "single TAS" regions at the "far" ends of the two arms, which bear single TASs (not shown).
| |
Figure 6. Anatomy of the MAC chromosomes of the 81-MAC family, and the Telomere Addition Sites used to generate them. The polytene chromatids bearing the precursors of this family in the MAC anlage we believe are always broken, and precisely at the same TASs, at the outside positions (narrow arrows), but are alternatively broken at the central TAS regions flanking the common region (broad arrows). Only these three chromosomes is ever observed; we have never seen arm-alone or two-armed chromosomes in this family. Thicken horizontal arrows represent the exons of the three genes of the locus, separated by conventional spliceosomal introns.
| |
 |
|
We infer the existence of cis acting sequences that dictate each TAS region and each TAS. Displays across the multiple-TAS region in vegetative cells which carry MACs that were independently developed shows that the multiple-TAS pattern is fully reproducible (Fig. 7A). Also, the exact pattern is sensitive to the allele of the 81-locus that is processed (Fig. 7B). Thus, we envision three cis-acting sequences guiding the process of chromatid breakage and telomere addition: a binding site of the enzyme complex that performs these actions, a "cut" site for the endonuclease, and a site or set of linked sites to which telomerase adds telomere repeats. In other ciliates it is known that a chromosome break site (CBS) binding a complex of the "cutter" and apparently also of telomerase, non-covalently associated with the binder/cutter. We hypothesize a third type of cis-acting site, at which an exonuclease pauses, as it erodes cut sites which are not immediately tended to by telomerase, but at which telomerase ultimately acts, thus generating the reproducible mixed TAS patterns. Presumably, fixed TASs result from binding, cutting and telomerase action with no intervening exonuclease erosion.
| |
Figure 7. Evidence for genetic determinants of the multiple TAS pattern displayed from the right side of the 81-MAC family common region.
A. Independent episodes of MAC development generate the same pattern. Clones of exconjugants with the same genotype of the 81-MAC locus were established and display patterns were developed from the DNA of each clone; 3 O. trifallax alleles are used in these experiments: "3," "a," and "b." Note also that the patterns are subtly sensitive to the genotype.
B. Genotype dependence of the pattern. Displays of four different O. trifallax strains are show on the right; to the left the display pattern is shown for strain 3.5 of the genotypically different sibling species, O. fallax strain 3.5. The "verify" lane is the display from a PCR product generated from the product that generated the "3.5" display: the same subtelomeric primer was coupled with a telomeric-repeats primer, amplifying only products with telomeres; the sizes of the display products of bone fide products shift down the gel by a predicted size. The bottoms of these two lanes are displayed at the far left, illustrating the verification of a MAC chromosome that has telomeres added only 4 nt 5' of the transcription start site for the common-region gene: does this copy of the gene function?
| |
 |
Two hypotheses for the mechanism of generation of mixed TAS patterns.
• In these models we strive also to explain the correlation between the incidences of Mixed TAS regions and alternative processing Common Region boundaries (Fig. 5), regions which are not cut on each chromatid, generating the larger family chromosomes with an arm appended to one end or the other of the CR.
• We assume that a loose complex exists between Telomerase and a "CBS-binder / cutter" nuclease.
• We assume an exonuclease activity that erodes to-be-"telomered" ends when not yet capped by telomerase.
• The two offered hypotheses are kinetic, chain-of-event models that each invoke a poor cis-acting site in the mixed TAS region: TAS regions that sport either a weak CBS or cut site, or both, are "mixed TAS" or "multiple-TASs" regions, and that in these cases cutting occurs, but immediate action of telomerase is thwarted.
• The "hesitant-cutter" or "impatient-telomerase" Hypothesis: the case of a weak/hard-to-cut site. The consequence of a weak cut site is seen to be: while waiting for the "hesitant cutter" to act on the recalcitrant site, sometimes telomerase "impatiently" dissociates before the cut is made, and is absent to heal the broken ends when they are subsequently made. Consequently, erosion of the raw breaks ensues, until telomerase can finally act. Telomerase eventually acts on the eroded end, most often at recessed positions at which the eroder has paused. These TASs thus exist at or near these pause sites. The result is what we observe at multiple-TASs regions, where the internal TASs are dictated by the third kind of cis-acting sequence, the pause sites.
• The "frustrated telomerase" Hypothesis: the case of a weak CBS binding site. If the binding of the complex to the CBS is weak, the complex may bind transiently and cut, but then dissociate from the site before telomerase can act. That is, the telomerase is "dragged away," "thwarted" or "frustrated." Again, in its absence the eroder erodes and pauses, telomerase returns to cap the population of ends, generating what we observe at multi-TAS regions.
Herrick G, Hunter D, Williams K, Kotter K. 1987. Alternative processing during development of a macronuclear chromosome family in Oxytricha fallax. Genes Develop. 1:1047-1058.
Klobutcher LA, Herrick G. 1997. Developmental genome reorganization in ciliated protozoa: the transposon link. Prog. Nucleic Acid Res. and Mol. Biol. 56:1-62.
Witherspoon D, Doak TG, Williams KR, Seegmiller A, Seger J, Herrick, G. 1997. Selection on the protein-coding genes of the TBE1 family of transposable elements in the ciliates Oxytricha fallax and O. trifallax. Mol. Biol. Evol. 14:696-706.
Seegmiller A, Williams KR, Herrick G. 1997. Two two-gene macronuclear chromosomes of the hypotrichous ciliates Oxytricha fallax and O. trifallax generated by alternative processing of the 81 locus. Dev. Genet. 20: 348-357.
| |
|
|