The C. elegans rab family: identification, classification and toolkit construction

Abstract

Rab monomeric GTPases regulate specific aspects of vesicle transport in eukaryotes including coat recruitment, uncoating, fission, motility, target selection and fusion. Moreover, individual Rab proteins function at specific sites within the cell, for example the ER, golgi and early endosome. Importantly, the localization and function of individual Rab subfamily members are often conserved underscoring the significant contributions that model organisms such as Caenorhabditis elegans can make towards a better understanding of human disease caused by Rab and vesicle trafficking malfunction. With this in mind, a bioinformatics approach was first taken to identify and classify the complete C. elegans Rab family placing individual Rabs into specific subfamilies based on molecular phylogenetics. For genes that were difficult to classify by sequence similarity alone, we did a comparative analysis of intron position among specific subfamilies from yeast to humans. This two-pronged approach allowed the classification of 30 out of 31 C. elegans Rab proteins identified here including Rab31/Rab50, a likely member of the last eukaryotic common ancestor (LECA). Second, a molecular toolset was created to facilitate research on biological processes that involve Rab proteins. Specifically, we used Gateway-compatible C. elegans ORFeome clones as starting material to create 44 full-length, sequence-verified, dominant-negative (DN) and constitutive active (CA) rab open reading frames (ORFs). Development of this toolset provided independent research projects for students enrolled in a research-based molecular techniques course at California State University, East Bay (CSUEB).

Figure 1. The complete list of PSI-BLAST hits using a sequence alignment of known C. elegan Rabs.

Genes are listed in descending order of RabF % identity (ID). The consensus sequence used to calculate RabF % ID is listed in the top row. For each hit, the portion of the alignment corresponding to each Rab Family motif (1–5) is provided. By contrast, the C-terminal sequence is only provided for those hits with one or two near terminal cysteines. For each C-terminus shown, the putative CBR interacting motif (CIM) is boxed if present. C-terminal cysteines are shaded orange. RabF consensus matches are shaded green and specific amino acids that suggest inclusion in a nonRab family are shaded yellow. The bold horizontal line separates the cluster of mostly known rabs from other monomeric GTPases. Numbers in parentheses indicate the number of amino acids omitted from the C-termini. RBH =  Reciprocal Best Hit.

Figure 2. A chladogram of Rab family members from C. elegans and H. sapiens.

The evolutionary history was inferred using the Neighbor-Joining phylogenetic reconstruction method. The tree is rooted with the natural outlying clade, Rab28. The optimal tree is shown with the percentage of replicate trees (>40) in which the associated genes cluster together in the bootstrap test (500 replicates) provided next to each branch. The tree is drawn to emphasize topology. The evolutionary distances were computed using the JTT amino acid substitution method and are in the units of the number of amino acid differences per site. Evolutionary analyses were conducted using MEGA5. Clades marked with red, orange or yellow circles indicate their degree of stability under a variety of phylogenetic reconstruction parameters (see text and methods for details). Red = 14/14, orange = 13/14 and yellow = 12/14 trees. Genes highlighted with black circles represent putative orphan C. elegan Rabs (lacking a human ortholog). For simplicity, closely related splice variants and well-supported human-specific clades were deleted (see methods for details).

Figure 3. Bootstrap scores for specific terminal clades from 13 additional phylogenetic reconstructions.

Thirteen additional phylogenetic analyses were performed using a combination of statistical methods. Phylogenetic reconstruction methods include Neighbor Joining (NJ), maximum likelihood (ML) or minimum evolution (ME). Amino acid substitution methods include Poisson (Po), JTT, Dayhoff (D), Equal Input (EI) and WAG. Gap deletion treatments include partial (Par) or pairwise (Pai) and rates and patterns of evolution include gamma distributed (+G), invariant sites (+I) or uniform (all others). All phylogenetic reconstructions were performed in MEGA5. Specific orthologous and/or paralogous clades include both worm and human Rab proteins. New orthologous clusters include bootstrap scores supporting the Rab44/4R79.2, Rab45/C33D12.6 and Rab23/ZK669.5 pairs. The new paralogous cluster includes the bootstrap scores that support the Rab44,4R79.2,Rab45 and C33D12.6 terminal clade.

Figure 4. Comparative analysis of intron position among diverse Rab subfamily members.

A) Cladogram indicating evolutionary relationships of 18 species examined here –. Ecdy. = Ecdysozoa, Chromal. = Chromalveolata. For species abbreviations see Methods. B) An ML tree of Opisthokonts created from the MSA used to map intron positions. Bootstrap support (100 replicates) is indicated for each subfamily cluster. C) For each subfamily, the number of times a Subfamily Specific Conserved Intron Position (SSCIP) involving the indicated number of species was observed (gray bars), compared to what is expected by chance (black diamonds). The difference between observed and expected is statistically significant where indicated. *P(Monte Carlo) <0.05. ***P(Monte Carlo)≤0.00001. The Rab31, 6, 5, 22, 34, 21 and 23 subfamilies include 17, 18, 17, 9, 9, 10, 14 and 12 species, respectively. D) and E) Heat map indicating number of introns within Rab31 (D) or Rab6 (E) that match SSCIPs from Rab31, 5, 22, 21, 6, 34 and 23. The circled number indicates the number of introns present in the MSA for each gene. % equals the percentage of introns that are shared with the true SSCIP. C56E6.2 (D) and Y71H2AM.12 (E) are highlighted red. Genbank Descriptions (if any) and RABDB! classifications are included. Classification abbreviations include: HypoRabX1 (H.RabX1), HypoRabX2 (H.RabX2), HypoRabX3 (H.RabX3) and MetazoaRabX3 (M.RabX3). F) A pairwise comparison of intron position conservation between specific genes (Rab31 at left, Rab6 at right) and their corresponding set of SSCIPs. Black diamonds plot the probability that a specific number of intron matches would be expected by chance for each set of conditions. Chart 1 plots a comparison of 5 introns with 7 SSCIPs (5×7). Chart 2∶4×7. Chart 3∶2×7. Chart 4∶4×6. Observed values for a subset of genes are indicated with P values estimated from the Monte Carlo simulation data (See text and methods). Species abbreviations are as in A. C) and F) 72 protosplice sites assumed.

Figure 5. A flow chart describing the lab module involving verification and modification of ORFeome Rab clones.

Steps 1 through 4 were done in parallel to steps A through D. Two peer-review steps at 3 and D were included to minimize mistakes in primer design and sequence analysis of WT ORFeome clones. Abbreviations: Gene of Interest (GOI), Constitutive Active (CA), Dominant Negative (DN), Restriction Fragment Length Polymorphism (RLFP), Polyacrylamide Gel Electrophoresis (PAGE), Human Ras (HRAS), Wild-Type (WT).

Figure 6. Full-length ORF sequence of tag-312 (Rab45) isolates reveals a different splice pattern than predicted.

A) A nucleotide alignment of tag-312 (FlOCS), genomic DNA and the predicted tag-312 ORF at two regions where exon/intron splice junction differences were found. In the top alignment, FlOCS reveals a new intron, splitting predicted exon 7 into exons7a and 7b. In the bottom alignment, FlOCS reveals an alternate 5′ splice donor and 3′ splice acceptor for intron 8. Compare FlOCS-supported 5′ and 3′ splice sites boxed in bold to predicted 5′ and 3′ splices sites (boxed, not bold). B) Two multiple sequence alignments of Rab45 subfamily members spanning the two regions described in 5A above demonstrate the impact of FLOCS-supported gene structure differences. The alignment includes proteins from the nematodes, Caenhorabditis elegans, tag-312(FlOCS) and NP508523.1, Caenhorabditis brenneri, Caenhorabditis remanei and Ascaris suum in addition to Xenopus laevis (frog), Homo sapiens (human), Monodelphis domestica (opossum) and Anolis carolinensis (lizard). The intron that splits exon 7 into two creates a 15 amino acid deletion that is conserved among all species examined (top). The alternate intron 8 creates an Indel in a region of Rab45 that is conserved among nematodes only.