Evolutionary origin of the protozoan parasites histone-like proteins (HU)

---

Abstract

The histone-like proteins (HU) belong to a family of DNA architectural proteins that stabilize nucleoprotein complexes. We found a putative HU protein (TgGlmHMM_3045) in Toxoplasma gondii genome that was homologous to the bacterial HU protein. This putative sequence was located in the scaffold TGG_995361 of the chromosome 10. The sequence included the prokaryotic bacterial histone-like domain, KFGSLGlRRRGERVARNPRT (ID number PS00045). HU protein sequences were also found in Plasmodium falciparum, Neospora caninum, Theileria parva and Theileria annulata. We found that the homology of the putative HU protein in Apicomplexa was greater with bacterial histone-like proteins than with eukaryotic histone proteins. The phylogenetic tree indicated that the putative HU protein genes were acquired in Apicomplexa by means of a secondary endosymbiotic event from red algae and later they were transfered from the apicoplast organelle to the nuclear genome.

Keywords: histone-like protein, Apicomplexa, Toxoplasma gondii, endosimbiotic, apicoplast, lateral transfer gene

---

Introduction

The histone-like proteins (HU) play a role in DNA replication, recombination, repair and regulation of gene transcription. HU protein has been shown to be important for optimal survival of the cells in stationary phase and under various stress conditions [Balandina et al., 2001]. Bacteria contain HU proteins that share some properties with the eukaryotic histones and are a major component of the bacterial nucleoid [Nash and Robertson, 1981]. However, unlike eukaryotic histones, the domain of the bacterial histone-like proteins (BHL) has not been shown to interact as a unit with DNA to form complexes analogously [Hübscher et al., 1980].

The Apicomplexa, the ciliates and the order of dinoflagellates, form the infrakingdom Alveolata [Cavalier-Smith, 2003]; they possess some chromosomal proteins, similar to histones. Thus, one putative HU protein sequence of prokaryotic origin has been described in Toxoplasma gondii, in Plasmodium yoelii, and also in Plasmodium berghei [Kobayashi et al., 2002]. This situation generates an intriguing question about the origin of the histones in the apicomplexas. There is evidence that the evolutionary development of histones and nucleosomes leads to an efficient packaging of DNA into structured chromosomes, which were probably a key step in the evolution of eukaryotes from prokaryotes [Minsky et al., 1997]. We were interested in examining the presence of putative histone-like protein in Toxoplasma gondii and other Apicomplexa genomes and to determine their phylogenetic relations in order to obtain more information about their evolutionary origin. The high phylogenetic relation of histone-like protein genes between prokaryotic nucleoids and apicoplast suggest that this gene could be a therapeutic target for pathogenic Apicomplexa in the future.

---

Materials and methods

The Toxoplasma gondii and Plasmodium sequences were downloaded from ToxoDB (www.ToxoDB.org) and PlasmoDB databases (www.plasmodb.org), respectively [Kissinger et al., 2002; Kissinger et al., 2003]. The BHL domain from Escherichia coli was used as query sequence. Other protein sequences were downloaded from GenBank with their corresponding GI numbers indicated along with each taxon. Redundant sequences were removed. PSI-BLAST searches were performed with WU-Blast2 server at the European Bioinformatics Institute server to look for homologous proteins in other members of the Apicomplexa [Lopez et al., 2003]. Search for domains and functional motifs were performed on genomic Smart and Prosite database [Schultz et al., 1998; Hofmann et al., 1999]. Sequence alignments were made using ClustalW program (www.ebi.ac.uk/clustalw) [Thompson et al., 1994].

For the phylogenetic trees, we performed a Minimum Evolution (ME), Maximum Parsimony (MP) and UPGMA analysis. The trees were constructed with the MEGA 3.0 software (www.megasoftware.net) with the following settings: 100 bootstrap replicates and 42535 seed; a close neighbor interchange was used as search option (level = 1) with an initial neighbor-joining tree, MaxTrees = 1. Gaps were considered as complete deletions. JTT substitution model was used [Jones et al., 1992]. Evolution rates among sites were considered uniform. We used the method performed by Chan et al. 2006, which identified a group with an extra N-terminal domain (approximately 20 amino acids) rich in basic residues when compared to the canonical HU [Wong et al., 2003]. They identified this group as the "long HU' (e. g. the dinoflagellate HCc3) vs. the 'short HU". Therefore, the tree was rooted with the HU sequences from: cyanobacteria; long and short HU forms from γ/β-proteobacteria; dinoflagellates; algae and protozoan.

---

Results and discussion

We found a complete sequence (TgGlmHMM_3045) in ToxoDB that codes for a putative HU protein. The sequence is located and encoded in the nuclear genome (chromosome 10 scaffold TGG_995361). Furthermore there is no copy of this gene neither in apicoplast nor in mitochondrial toxoplasma genomes. There is evidence that massive gene transfer from the plastid to the nuclear genome has occurred in the dinoflagellates [Bachvaroff et al., 2004; Hackett et al., 2004] and the apicomplexans [Roos et al., 2002; Huang et al., 2004]. HUs are known to be present in some plastid genomes [Sato et al., 2003] and may be some of the many genes included in this transfer from the plastid genome. The Toxoplasma HU sequence contains 235 aa, and exhibits a molecular weight of 25.170 kDa. Analysis on the complete apicomplexan HU sequences revealed the presence of plastid targeting signal peptide of 33 aa [Emanuelsson et al., 1999; Zuegge et al., 2001], and revealed the presence of a conserve BHL domain of 80 aa. The vast majority of apicoplast proteins including all proteins involved in metabolic activities are encoded in the nuclear genome and post-translationally imported into the apicoplast [Roos et al., 2002]. The BHL domain has significant homologies to the prokaryotic HU [Wong et al., 2003], furthermore the protein included the histone like motif ID number of the PS00045 Prosite Data base (Fig. 1).

We performed a PSI-BLAST search for the sequence TgGlmHMM_3045 using this last one as a query. Seven HU proteins were found in the following Apicomplexa: Plasmodium falciparum, P. vivax, P. berghei, Neospora caninum, Theileria parva and T. annulata (Fig. 1). Consensus ME trees were constructed from 100 sets of JTT distance matrices generated from bootstrapping (Fig. 2). Fast et al., 2001, suggest that apicomplexan and dinoflagellate plastids appear to be the result of a single endosymbiotic event, but our results showed that the two groups of HU from alveolate do not cluster together. The HU sequences from dinoflagellate cluster with a group of long HU from gamma/beta-proteobacteria, forming a clade that branched off from the proteobacteria short HU (Fig. 2); but the HU sequences from protozoan and algae cluster with the cyanobacteria. This result is similar to those obtained by Chan et al. 2006. The HU sequences from apicomplexan clade branches off, with a node of low bootstrap value (44), from the plastid-encoded HU homologue of Guillardia theta (a green algae). The low bootstrap value could be originated by the highly derived nature of the HU sequence from G. theta.

In a previous work, Chan et al., 2006, suggested that the apicomplexan HU clade branches off from the plastid-encoded HU of Cyanidioschyzon merolae (red algae) but they did not include red algae in their analysis. Our result differed when we included the HU sequence from G. theta (Fig. 2). Additionally, we constructed the phylogenetic tree without the HU sequence from C. merolae and our bootstrap value increased considerably. We performed a detailed analysis using the chloroplast genome database (chloroplast.cbio.psu.edu) and we found that the HU sequence from Toxoplasma has a Percent_id = 35.955 with the HU sequence from G. theta and a Percent_id = 34.090 when we compared it with the HU sequence from C. merolae.

The origin of the apicoplast (the plastid of apicomplexan) genome, whether from a green or a red plastid, has received much recent attention [Köhler et al., 1997; Waller et al., 2003; Funes et al., 2004]. Thus, Chan et al. 2006, assumed in his discussion that the ancestral alveolate acquired their plastids from red algae. However, the most accepted hypothesis is that T. gondii grouped this organellar genome with cyanobacteria and plastids, showing consistent clustering with green algal plastids. Taken together, these observations indicate that the Apicomplexa acquired a plastid by secondary endosymbiosis, probably from a green alga [Köhler et al., 1997]. When we used Maximum Parsimony (MP) and UPGMA phylogenetical test using the same JTT substitution model, both trees showed different evolutionary histories branch off Apicomplexa HU with red algae (C. merolae) (Fig. 3a and 3b). Although the result of the ME test showed a green algae HU origin in Apicomplexa, both, the MP and UPGMA test, showed a red algae origin with a higher boostrap value than those of the ME test on the basal node of the cluster of Apicomplexa, resulting in values of 50 and 59 respectively (Fig. 3a and 3b). This result confirmed those obtained by Chan et al., 2006.

In conclusion our phylogenetic analysis showed that the most probably origin of HU protein in Apicomplexa was from cyanobacteria through an intermediate step including an endosymbiotic event with red algae.