AI-driven pan-proteome analyses reveal insights into the biohydrometallurgical properties of Acidithiobacillia

Abstract

Microorganism-mediated biohydrometallurgy, a sustainable approach for metal recovery from ores, relies on the metabolic activity of acidophilic bacteria. Acidithiobacillia with sulfur/iron-oxidizing capacities are extensively studied and applied in biohydrometallurgy-related processes. However, only 14 distinct proteins from Acidithiobacillia have experimentally determined structures currently available. This significantly hampers in-depth investigations of Acidithiobacillia's structure-based biological mechanisms pertaining to its relevant biohydrometallurgical processes. To address this issue, we employed a state-of-the-art artificial intelligence (AI)-driven approach, with a median model confidence of 0.80, to perform high-quality full-chain structure predictions on the pan-proteome (10,458 proteins) of the type strain Acidithiobacillia. Additionally, we conducted various case studies on de novo protein structural prediction, including sulfate transporter and iron oxidase, to demonstrate how accurate structure predictions and gene co-occurrence networks can contribute to the development of mechanistic insights and hypotheses regarding sulfur and iron utilization proteins. Furthermore, for the unannotated proteins that constitute 35.8% of the Acidithiobacillia proteome, we employed the deep-learning algorithm DeepFRI to make structure-based functional predictions. As a result, we successfully obtained gene ontology (GO) terms for 93.6% of these previously unknown proteins. This study has a significant impact on improving protein structure and function predictions, as well as developing state-of-the-art techniques for high-throughput analysis of large proteomic data.

Keywords: Acidithiobacillia; biohydrometallurgy; gene co-occurrence network analysis; protein structure prediction; proteome.

FIGURE 1

The statistical analysis of protein traits and structure confidences of the Acidithiobacillia proteome. (A) Mathematical modeling of the pangenome and core genome of Acidithiobacillia. (B) Bar chart showing functional proportions (based on COG categories) of different parts of the Acidithiobacillia pangenome (i.e., core, accessory, and unique). (C) Histogram showing the distribution of structure confidences of Acidithiobacillia pan-proteome. (D) Box plot showing the distribution of structure confidences of Acidithiobacillia pan-proteome among different COG categories with the average confidence values indicated.

FIGURE 2

A metabolic model of sulfur/iron utilization/biometallurgy related proteins in Acidithiobacillia based on genomic inference. The 3D protein structures illustrated in the figure were predicted by AlphaFold2. The case-study proteins discussed in this study are marked with orange rectangles.

FIGURE 3

The structures of Acidithiobacillia sulfate transporter (AfSULTR-ACK80903) and sulfide:quinone oxidoreductase (AfSQR-ACK80497). (A) The overall structure of AfSULTR monomer is comprised of transmembrane (TM) helices and anti-Sigma factor antagonist (STAS) domain (shown with red and purple colors, respectively). (B) Top panel: the AfSULTR dimer model formed by two identical monomers (shown with yellow and green colors, respectively). Bottom panel: electrostatic potential surfaces of the overall AfSULTR dimer calculated with adaptive Poisson-Boltzmann solver (APBS). (C) Superposition of AfSQR-ACK80497 onto human SQR (HmSQR, PDB: 6OIB).

FIGURE 4

Structure visualizations of Acidithiobacillia ferrous iron transporter (AfFeoB), iron oxidase and ferrochelatase. (A) The overall structure of AfFeoB monomer that consists of G domain, GDI domain, and transmembrane domain. (B) Electrostatic potential surfaces of the overall AfFeoB monomer calculated with adaptive Poisson-Boltzmann solver (APBS), which is rotated 90° rightward as indicated to reveal the negatively charged enriched region of AfFeoB that putatively binds the ferric cation. (C) The archetypical GTPase motifs G1–G5 (shown with yellow color) in the GTPase domain of AfFeoB that flank the nucleotide-binding pocket.

FIGURE 5

(A) The overall structure of Acidithiobacillia iron oxidase (Iro) with key conserved cysteine residues and aromatic residues that ligate/stabilize the [Fe(4)S(4)] cluster are shown in ball and stick format, and the free ferrous iron ion is represented by a red sphere. (B) Putative electron transfer pathway(s) in Iro identified by Emap using the structure from the last frame of MD simulation. (C) The overall topology of Acidithiobacillia ferrochelatase (ACK80603) monomer that is comprised of two similar domains (shown with yellow and green colors, respectively). The substrate protoporphyrin molecule is shown in ball and stick format. A free ferrous iron ion and a free magnesium ion are represented by red and pink spheres, respectively.

FIGURE 6

Cartoon representations and suggested key residues of (A) ACK80741, (B) Mycobacterium tuberculosis NrdH (PDB 4F2I), (C) thioredoxin from Escherichia coli TrxA (PDB 2TRX), (D) glutaredoxin 1 of Plasmodium falciparum (PDB 4HJM), (E) poxviral glutaredoxin (PDB 2HZF), and (F) human glutaredoxin (PDB 2FLS). The coloring is based on secondary structure (helix, red; sheet, blue; and loop, gray).

FIGURE 7

A graphical workflow of the computational processes.