Identification of putative cis-regulatory elements in Cryptosporidium parvum by de novo pattern finding

Abstract

Background: Cryptosporidium parvum is a unicellular eukaryote in the phylum Apicomplexa. It is an obligate intracellular parasite that causes diarrhea and is a significant AIDS-related pathogen. Cryptosporidium parvum is not amenable to long-term laboratory cultivation or classical molecular genetic analysis. The parasite exhibits a complex life cycle, a broad host range, and fundamental mechanisms of gene regulation remain unknown. We have used data from the recently sequenced genome of this organism to uncover clues about gene regulation in C. parvum. We have applied two pattern finding algorithms MEME and AlignACE to identify conserved, over-represented motifs in the 5' upstream regions of genes in C. parvum. To support our findings, we have established comparative real-time -PCR expression profiles for the groups of genes examined computationally.

Results: We find that groups of genes that share a function or belong to a common pathway share upstream motifs. Different motifs are conserved upstream of different groups of genes. Comparative real-time PCR studies show co-expression of genes within each group (in sub-sets) during the life cycle of the parasite, suggesting co-regulation of these genes may be driven by the use of conserved upstream motifs.

Conclusion: This is one of the first attempts to characterize cis-regulatory elements in the absence of any previously characterized elements and with very limited expression data (seven genes only). Using de novo pattern finding algorithms, we have identified specific DNA motifs that are conserved upstream of genes belonging to the same metabolic pathway or gene family. We have demonstrated the co-expression of these genes (often in subsets) using comparative real-time-PCR experiments thus establishing evidence for these conserved motifs as putative cis-regulatory elements. Given the lack of prior information concerning expression patterns and organization of promoters in C. parvum we present one of the first investigations of gene regulation in this important human pathogen.

Figure 1

Flow-chart illustrating methodology. Pattern-finding was carried out in two iterations, first de novo and a second time using information obtained from comparative real-time PCR expression profiles.

Figure 2

Motifs identified upstream of oocyst wall and large secretory proteins. (a) Upstream regions of genes encoding cryptosporidial oocyst wall proteins, and the occurrences of the most significant upstream motif shared by all of these upstream regions. The positions of the motifs are drawn to scale. All positions are shown with respect to the translational start. Solid black symbols denote a motif located on the reverse strand. Sequence-logo displaying the information content for the over-represented motif. (b) Upstream regions of genes encoding cryptosporidial large secretory proteins, and the occurrences of the most significant upstream motif shared by all of these upstream regions. Sequence-logo displaying the information content for the over-represented motif. Expression profiles for both families of genes were published elsewhere (Abrahamsen et al. 2004; Templeton et al. 2004).

Figure 3

Results of motif and expression analyses. (a) Motifs and expression profiles associated with genes involved in nucleotide metabolism. Schematic representation of the upstream regions are shown for each gene. The location of 4 different candidate motifs are indicated by the use of four different shapes. The single motif found in each gene of the group is indicated by a circle. The locations of three additional candidate motifs, each associated with a sub-set of sequences are indicated by the remaining shapes drawn on the upstream regions and as indicated to the left of each sequence logo. Solid black shapes indicate motifs found on the reverse strand. Comparative real-time PCR profiles for sub-sets of each group of genes organized by expression profile over a 72 h period are shown as sub-sets 1–3 Each sub-set is associated with a single candidate motif. (b) Motifs and expression profiles associated with genes involved in DNA replication. (c) Motifs and expression profiles associated with genes involved in glycolyis.