A global comparison of the human and T. brucei degradomes gives insights about possible parasite drug targets

Abstract

We performed a genome-level computational study of sequence and structure similarity, the latter using crystal structures and models, of the proteases of Homo sapiens and the human parasite Trypanosoma brucei. Using sequence and structure similarity networks to summarize the results, we constructed global views that show visually the relative abundance and variety of proteases in the degradome landscapes of these two species, and provide insights into evolutionary relationships between proteases. The results also indicate how broadly these sequence sets are covered by three-dimensional structures. These views facilitate cross-species comparisons and offer clues for drug design from knowledge about the sequences and structures of potential drug targets and their homologs. Two protease groups ("M32" and "C51") that are very different in sequence from human proteases are examined in structural detail, illustrating the application of this global approach in mining new pathogen genomes for potential drug targets. Based on our analyses, a human ACE2 inhibitor was selected for experimental testing on one of these parasite proteases, TbM32, and was shown to inhibit it. These sequence and structure data, along with interactive versions of the protein similarity networks generated in this study, are available at http://babbittlab.ucsf.edu/resources.html.

Figure 1. Global view of predicted active proteases of human and T. brucei showing sequence similarity relationships.

Protease sequences are represented as nodes, and similarity relationships between sequences better than the threshold (BLAST E-value ≤1e-5) are depicted as “edges” or lines between nodes. In the network are represented 594 human and 127 T. brucei sequences (total of 721 nodes and 10,188 edges). (A) Distribution by family of proteases. Nodes for human sequences are represented as circles and for T. brucei sequences as triangles, and are colored by MEROPS-associated family (see Methods). Families of some of the larger clusters are labeled, and the parasite-specific C51 and M32 clusters are circled in red. (B) Structure coverage of sequence space is broad in human and T. brucei. The same sequence similarity network as in panel A is shown except that it is color-coded by species and nodes are enlarged and designated by different shapes to denote if a crystal structure or model exists for that sequence. Node shapes: square = crystal structure; triangle = ModBase model; diamond = ModWeb model; small circle = no structure.

Figure 2. Distribution by catalytic type of peptidases predicted to be active in humans and T. brucei.

In humans, proteases of catalytic type S (where the catalytic moiety is serine) is dominant, but metallo (type M) and cysteine (type C) peptidases are also abundant. In contrast, in T. brucei, serine peptidases are less abundant, and cysteine and metallo proteases are equally prominent. Other main catalytic types in each organism include the threonine (type T) and aspartatic (type A) proteases. Catalytic types were assigned by catalytic type designated in the family of the closest BLAST hits to MEROPS sequences.

Figure 3. Structure similarity network of human and T. brucei proteases using crystal structures and models.

Nodes represent experimentally characterized (crystal structure) or modeled structures and edges represent pairwise structural similarity above the structural similarity threshold (FAST SN score ≥4.5). Nodes for 342 human and 71 T. brucei are shown in the network (total of 413 nodes and 7,234 edges). The two T. brucei-specific families (TbM32 and C51) highlighted in the sequence similarity network shown in Figure 1 are circled in red. (A) Nodes are colored by MEROPS-associated family, revealing cross-family structural relationships. Human structures are represented as circles and T. brucei as triangles. (B) The same structure similarity network as in panel A is painted by species and structure representation. Nodes are color-coded by species and node shape corresponds to type of structure representation for that sequence: square = crystal structure; triangle = ModBase model; diamond = ModWeb model. In contrast to T. brucei, there are a large number of experimentally characterized (crystal) structures for humans, but many T. brucei structures can be modeled.

Figure 4. Structural similarity network of human and T. brucei proteases labeled by clan.

The same network as in Figure 3 is colored here by assigned MEROPS clan (see Methods). One cluster is composed of multiple clans (MC, MF, MH, and CF).

Figure 5. The T. brucei M32 protease model shows active site similarity to a human protease ACE2.

The model of the T. brucei M32 protease (TbM32m, purple) is shown structurally aligned with crystal structure ACE2 (PDB code 1R4L, yellow). Depicted in ball-and-stick representation near the zinc ion are the metal binding residues and catalytic glutamate. ACE2 inhibitor MLN4760 is shown in green and ACE inhibitor lisinopril is in orange stick format (the position of which is from a structural alignment of ACE (1O86) with ACE2). The predicted steric clash of R273 in the ACE2 S1 pocket with lisinopril is marked with an arrow. The R273 CZ of ACE2 is predicted to be 1.5 Å from the lisinopril C9, so that a terminal nitrogen of R273 is in position to overlap with an oxygen of lisinopril. The arginine (R348) from TbM32m that is predicted to be close to the ACE2 R273 is also in ball-and-stick representation. The inset shows the overall structural similarity of the two proteins.

Figure 6. TbM32 is inhibited by 28FII (ACE2 inhibitor) and not by lisinopril (ACE inhibitor).

The chart shows results from a representative experiment with 1,10P (1,10 Phenanthroline, 100 µM), lisinopril (10 µM), and 28FII (10 µM). ** indicates significant difference from the control (DMSO vehicle) at p<0.005. The positive control 1,10P is a metal chelator that inhibits metallopeptidases.

Figure 7. Structure alignment of T. brucei C51 model (TbC51m) with a distant structure homolog, human Cathepsin F (CatF).

The superposition shows these two proteins have some general, overall structural similarities, but also large differences near the active site. The TbC51 model is colored in light orange, and the human CatF is in light green. While the catalytic Cys-His dyads are closely superimposed (depicted in ball-and-stick), a striking difference is marked by an arrow indicating the predicted steric clash between the CatF vinyl sulfone inhibitor (red) and the helix of TbC51 that partially obstructs the active site.