Data-driven large-scale genomic analysis reveals an intricate phylogenetic and functional landscape in J-domain proteins

Abstract

The 70-kD heat shock protein (Hsp70) chaperone system is a central hub of the proteostasis network that helps maintain protein homeostasis in all organisms. The recruitment of Hsp70 to perform different and specific cellular functions is regulated by the J-domain protein (JDP) co-chaperone family carrying the small namesake J-domain, required to interact and drive the ATPase cycle of Hsp70s. Besides the J-domain, prokaryotic and eukaryotic JDPs display a staggering diversity in domain architecture, function, and cellular localization. Very little is known about the overall JDP family, despite their essential role in cellular proteostasis, development, and its link to a broad range of human diseases. In this work, we leverage the exponentially increasing number of JDP gene sequences identified across all kingdoms owing to the advancements in sequencing technology and provide a broad overview of the JDP repertoire. Using an automated classification scheme based on artificial neural networks (ANNs), we demonstrate that the sequences of J-domains carry sufficient discriminatory information to reliably recover the phylogeny, localization, and domain composition of the corresponding full-length JDP. By harnessing the interpretability of the ANNs, we find that many of the discriminatory sequence positions match residues that form the interaction interface between the J-domain and Hsp70. This reveals that key residues within the J-domains have coevolved with their obligatory Hsp70 partners to build chaperone circuits for specific functions in cells.

Keywords: Hsp40 co-chaperones; J-domain proteins; artificial neural networks; large-scale data analysis; protein homeostasis.

Fig. 1.

Structural views of the J-domain and domain architectures of JDPs. (A) Schematic view of the canonical Hsp70 cycle. (B) Structural view of the Hsp70-J-domain complex [PDB ID: 5NRO (20)]. NBD denotes the Hsp70 nucleotide-binding domain, and SBD the Hsp70 substrate-binding domain. The four helices (α1–α4) forming the J-domain and the characteristic HPD motif are highlighted. (C) Cytosolic class A JDP of Saccharomyces cerevisiae (Ydj1) in its constitutive homodimeric form [the structure is a combination of the separate J-domain, PDB ID 5VSO (21), CTDs PDB ID 1NLT (22), and dimerization domains, PDB ID 1XAO (23)]; in lack of a known experimentally determined structure, the G/F-rich linker between the J-domain and the first C-terminal domain has been hand-drawn on one of the two protomers. The different domains have been colored only on one protomer. The general architecture of class A JDPs is highlighted below the structure (G/F: glycine/phenylalanine rich linker; CTD: C-terminal, substrate-binding domain; ZFLR: cysteine-rich, zinc-finger-like region; DD: dimerization domain). (D) Class B JDP of Thermus thermophilus in its constitutive homodimeric form [PDB ID 1C3G (24)]. The different domains have been colored only on one protomer. The general architecture of class B JDPs is highlighted below the structure. (E) Escherichia coli HscB [class C; PDB ID 1FPO (25)] with its architecture (HSCB_C: C-terminal domain of HscB). (F) Mus musculus Erdj5 [class C; PDB ID 5AYK (26)] and its architecture.

Fig. 2.

Characterization of the JDP dataset. (A) Growth of the number of JDP sequences found in the UniProt Database over the last three decades. The red dashed line shows an exponential fit. (B) Class proportions in the dataset. (C) Amino acid frequencies of the G/F linker region of class A JDPs. (D) Amino acid frequencies of the G/F linker region of class B JDPs. (E) Number of JDPs in the different kingdoms, highlighted by their class (A: green; B: blue; C: red). (F) The 12 most abundant J-protein domain architectures. Numbers on the left denote the number of sequences with this architecture in our dataset, and, on the right, we report the corresponding human, yeast, plant, or bacterial proteins. (G) Highlight of the evolutionary conservation of the cysteine-rich proto-ZFLR in ER-resident JDPs in six metazoans and two plants.

Fig. 3.

Two-dimensional representations of J-domains. JDP sequences are projected in two dimensions by a nonlinear transformation (UMAP) and colored according to different schemes. (A) J-domains are identified as being of class A (green), B (blue), or C (red) JDPs. (B) J-domains are identified as being of bacterial (red), eukaryotic (blue), or archaea (green) origin. (C) The five most abundant C subclasses are highlighted (see Fig. 2F for representative members). The circles in each panel focus on the group identified as HscB-like JDPs. (D) Zoom on the encircled region in the three other panels, highlighting the further differentiation of the eukaryotic ones into fungi, metazoans, and plants (bacterial HscB sequences are shown in grey background). For visual clarity, a random subset of 50% of all points are displayed.

Fig. 4.

Identification of the sequence and structure positions most relevant for the classification. (A) The relevance score for each position and for different classification tasks is represented in a heatmap. For each position, the relevance score averaged over the different tasks is represented, relative to the average over all positions, above the heatmap (HPD are not considered as due to their characteristic conservation they have negligible contributions to classification). Interface positions in the bacterial J-domain/DnaK complex (20) dimer are highlighted in blue. (B) Comparison of the relevance scores between interface and noninterface positions in the J-domain/DnaK hetero (20). The P-value was computed by the two-sided Mann–Whitney test with continuity correction. (C) Interresidue interaction network between bacterial J-domain/DnaK highlighting the 3 most relevant residues for each classification task. Bottom row: Only the interface residues for DnaK are shown. Blue nodes: Interface J-domain residues. Top 3 rows: Noninterface J-domain residues. The colored rings around the nodes indicate which J-domain residues are most relevant for each classification task. Links in the network are structural contacts between pairs of residues, measured on the crystal structure (20) (Methods). (D) The five top scores identified in the top panel are highlighted on the structure of the complex (20); blue: five most relevant residues on the J-domain; red: the interaction surface on DnaK. The five overall most relevant positions on the J-domain participate in the interaction surface between the two proteins.