Alternative processing of scrambled genes generates protein diversity in the ciliate Chilodonella uncinata

Abstract

In ciliates, chromosomal rearrangements occur during the development of the somatic macronuclear genome from the germline micronuclear genome. These rearrangements are extensive in three ciliate classes-Armophorea, Spirotrichea, and Phyllopharyngea-generating a macronucleus with up to 20,000,000 gene-sized chromosomes. Earlier, we have shown that these three classes also share elevated rates of protein evolution relative to other ciliates. To assess the evolution of germline-limited sequences in the class Phyllopharyngea, we used a combination of traditional and walking PCR to analyze micronuclear copies of multiple genes from two lines of the morphospecies Chilodonella uncinata for which we had previously characterized macronuclear sequences. Analyses of the resulting data yield three main results: (1) conserved macronuclear (somatic) regions are found within rapidly evolving micronuclear (germline) regions; (2) gene scrambling exists within this ciliate lineage; and (3) alternative processing of micronuclear regions yields diverse macronuclear beta-tubulin paralogs. To our knowledge, this is the first study to demonstrate gene scrambling outside the nonsister class Spirotrichea, and to show that alternative processing of scrambled genes generates diversity in gene families. Intriguingly, the Spirotrichea and Phyllopharyngea are also united in having transient "giant" polytene chromosomes, gene-sized somatic chromosomes, and elevated rates of protein evolution. We hypothesize that this suite of characters enables these ciliates to enjoy the benefits of asexuality while still maintaining the ability to go through sexual cycles. The data presented here add to the growing evidence of the dynamic nature of eukaryotic genomes within diverse lineages across the tree of life.

Figure 1. Diagram of micronuclear structures of a) three scrambled genes and b) three non-scrambled genes

The graphs are sliding window analyses of pairwise divergence (π) between sequences from USA and Poland, calculated using DNAsp (Librado and Rozas, '09). The cartoons for each gene indicate the structure of the MDSs (boxes) and IESs (lines) that are shared between the lines. Numbers below the MDSs are position relative to coding domain. Scrambling is indicated by non-sequential numbering of MDSs (bold numbers with asterix), or in the case of actin, by the inversion of MDS numbering and arrow below boxes. Jagged edges reflect missing data and ‘?’ indicate areas where CDS/UTR boundaries cannot be determined.

Figure 2. MAC β-tubulin paralogs P1, P2 and SP1 share some identical regions

Graph is sliding window analyses of pairwise divergence (π) calculated using DNAsp (Librado and Rozas, '09). Top comparison is of macronuclear (MAC) P1 and P2; bottom comparison of P2 and SP1. Regions with the same color at identical positions correspond to shared sequences, and colors correspond to micronuclear loci in Figure 3.

Figure 3. Micronuclear sequences can be alternatively spliced together to make diverse β-tubulins

Each color refers to a micronuclear locus (MIC P1, MIC P2, and MIC SP1), boxes indicate MDSs, black lines are traditional IESs and thick colored lines are alternatively processed IESs. The large MDS used to generate SP1 (Figure 1A) is divided into four MDS (MIC SP1-alt1-4) that are then spliced to MDSs from the MIC P1 locus to generate the MAC paralog P1 (Fig 31). These same four alternatively processed MIC SP1 MDSs are combined with alternatively processed MDSs from MIC P1 plus MDSs from MIC P2 to generate MAC P2 (Fig 3B). The brackets and asterix above the MAC molecules correspond to the region of sequence data below each figure, shown to exemplify boundaries: * = MAC P1, ** =MAC P2. The nucleotides in boxes are the conserved pointer sequences as detailed in Table 1, with uppercase letters for MDSs and lowercase letters for IESs.

Figure 4. Conserved motif that may be a cis-acting signal for excision of both scrambled and non-scrambled IESs

Motif was generated using MEME (Bailey et al., '09) by searching among all IESs from the polish strain of C. uncinata with the constraint of identifying only one motif per sequence. The motif is significant as it has an E value of 4.8e+000 while analysis of the same sequences shuffled yields a much higher E value of = 1.2e+005. A similar motif was obtained when analyzing the IESs from the USA line, the scrambled IESs and the non-scrambled IESs, regardless as to whether we constrained the program to find just one motif per sequence or any number of motifs.

Figure 5. Evolution of genome characteristics among ciliates

A suite of four characters is found in the non-sister classes Phyllopharyngea and Spirotrichea: giant chromosomes, extensive fragmentation, elevated rates of protein evolution and gene scrambling. ‘+’ and ’−’ refer to presence and absence of characters, respectively. Symbols in parentheses indicate lineages for which there are only few data while no entry means an absence of data. The ‘?’ for giant chromosomes in the class Litostomatea reflect an 1890 report of this feature in the genus Loxophyllum that has not been supported by subsequent studies (Ammermann, '87). The topology of the tree is redrawn from recent references and reflects the apparent non-monophyly of the class Nassophorea (Riley and Katz, '01, Gong et al., '09). The age of divergence between Phyllopharyngea and Spirotrichea is estimated to be at least 580 million years based on fossil evidence (Li et al., '07) while a molecular clock estimate the age of the ciliate clade at >1 billion years (Wright and Lynn, '97). Representative ciliate images redrawn from http://microscope.mbl.edu/.