Genome of Qualea grandiflora Mart. (Vochysiaceae) Reveals Multi-Level Aluminium Handling Mechanisms in a Cerrado Hyperaccumulator Species

1. Introduction

Aluminum (Al) is a major abiotic stressor that inhibits plant growth and development by disrupting critical physiological processes, including respiration, photosynthesis, and gene expression [1,2]. In response to Al toxicity, plants activate oxidative stress metabolism to mitigate cellular damage [2,3]. However, certain species exhibit Al tolerance or even dependence, employing mechanisms such as Al complexation with organic acids (e.g., citrate, malate, and oxalate), leading to accumulation exceeding 1 g Al·kg⁻¹ dry mass [2,3]. Notably, in the Brazilian Savanna (Cerrado), several species not only accumulate Al but also require it for optimal growth [1,2]. Among these, the Vochysiaceae family is particularly significant, as all its species are Al accumulators [2].

Qualea grandiflora Mart. (Vochysiaceae, Myrtales; APG IV, 2016), commonly known as "pau-terra," is a widely distributed Al-accumulating species in the Cerrado. This diploid (2n = 22) [4,5], hermaphroditic tree disperses seeds during the dry season (July–September) [6] and exhibits remarkable adaptability to varying light conditions, nutrient availability, and soil pH [7], making it a candidate for reforestation of degraded areas. Beyond its ecological role, Q. grandiflora has garnered pharmaceutical interest due to its antibacterial [8], antiulcer [9], anticonvulsant, analgesic, and antioxidant properties [10].

Given its high Al accumulation and tolerance, Q. grandiflora has been extensively studied. Andrade et al. [11] demonstrated that Al accumulates in chloroplasts without impairing their function. Furthermore, anatomical and physiological studies revealed that exogenous Al does not inhibit seed germination, with ~60% of Al stored in the seed translocated to seedling leaves [12,13]. Histochemical analyses localized Al primarily in cotyledons, suggesting a potential interaction with lipid and protein reserves, which may contribute to this species' ecological success [13]. Transcriptomic analyses of Al-exposed Q. grandiflora leaves revealed upregulation of genes linked to cell wall biosynthesis, primary/secondary metabolism, phytohormones (brassinosteroids, salicylic acid), chloroplast biogenesis, defence responses and Al transport, reinforcing its Al-dependent growth [14].

A well-documented Al tolerance mechanism involves Al-activated exudation of organic acids (O.As.; e.g., malate, citrate) into the rhizosphere, mediated by the ALMT (aluminium-activated malate transporter) and MATE (multidrug and toxic compound extrusion) gene families [15,16]. While transcriptomic studies on Al-accumulating species like Psychotria rubra (Rubiaceae) have identified Al-responsive genes [17], the absence of comparative analyses under varying soil acidity limits mechanistic insights.

A critical barrier to understanding Al-responsive mechanisms in non-model plants is the lack of reference genomes. Although next-generation sequencing (NGS) has expanded genomic resources for non-model species [18], no genome assembly exists for Q. grandiflora. To address this gap, we present the first de novo genome assembly of Q. grandiflora, enabling the identification of Al-tolerance and accumulation-related genes. This work establishes a foundational genomic resource for studying Al adaptation in Cerrado flora and advances our understanding of metallophyte evolution.

2. Results and Discussion

2.1. Genome Sequencing and Assembly

Whole-genome sequencing of Qualea grandiflora was conducted using the Illumina HiSeq 2500 platform with 100 bp paired-end libraries, generating approximately 57.7 Gb of raw data. Quality filtering (Q-score >30) yielded 536,103,004 high-quality reads (99.6% retention rate) from an initial 537,656,590 reads. The assembly achieved approximately 81% coverage of the estimated 500 Mb genome (Table 1), representing superior coverage compared to other Al-accumulating Cerrado species like Caryocar brasiliense: 45.7%; and Eugenia dysenterica: 56.7% [19].

The assembly pipeline compared ABySS 2.0 and DISCOVAR de novo, with ABySS 2.0 producing superior results. The final assembly comprised 406,048,254 bp distributed across 277,998 scaffolds, with scaffold lengths ranging up to 600,552 bp (Table 1). Assembly metrics included an N50 of 2,499 bp, L50 of 35,899, and GC content of 36.3%. Repetitive elements constituted 50.17% (203,730,046 bp) of the assembled genome, consistent with patterns observed in other plant genomes (Table 1).

BUSCO assessment against the Embryophyta database (1,440 orthologs) revealed 63.1% complete genes (including 1.3% duplicated), 13.1% fragmented, and 23.8% missing (Table 2). This completeness level meets established standards for non-model plant genomes, where >50% completeness is considered acceptable and more likely to appear in draft assemblies given the challenges of genome size, complexity, and phylogenetic divergence [20,21]. The 63.1% completeness particularly supports the utility of this assembly for gene-centric analyses, as noted in similar non-model plant genome projects [22].

2.2. Annotation and Prediction of Qualea grandiflora Genome

The structural annotation of the Q. grandiflora genome identified 38,034 genes, including 37,691 protein-coding mRNA genes and 343 tRNA genes, with an average gene length of 1,353.38 bp and a total of 114,847 exons (Table 3 and Table S1). To assess gene completeness, protein-coding sequences were evaluated using BUSCO, revealing that 58.4% of genes were complete, 2.6% were duplicated, 17.0% were fragmented, and 24.6% were missing (Table 2).

For functional annotation, predicted proteins were analysed using InterProScan to identify conserved domains and motifs, followed by comparison against the InterPro database. Additionally, a BLAST search was performed against the UniRef90 Viridiplantae protein database to assign putative functions.

Gene Ontology (GO) analysis classified 7,754 genes into Biological Process (BP), 3,686 into Molecular Function (MF), and 3,447 into Cellular Component (CC) categories (Figure 2). Within BP, the most frequent terms include organic substance metabolic process (4.2%), primary metabolic process (4.0%), nitrogen compound metabolic process (3.7%), cellular metabolic process (4.0%), biosynthetic process (1.8%), gene expression (1.6%), and transport (1.3%). The MF category was dominated by protein binding (2.0%), catalytic activity acting on proteins (2.3%), carbohydrate derivative binding (2.4%), oxidoreductase activity (2.6%), hydrolase activity (2.8%), small molecule binding (3.3%), transferase activity (3.5%), ion binding (5.0%), and organic/heterocyclic compound binding (5.6%). For CC, most genes were linked to intracellular anatomical structures (9.0%), organelles (7.1%), membranes (6.8%), cytoplasm (5.9%), intrinsic membrane components (4.3%), and cell periphery (1.8%). A summary of these GO term distributions is provided in Figure 2.

2.3. Candidate Genes from the Qualea grandiflora in Response to Al

To identify genes that could be associated with aluminium (Al) metabolism in Q. grandiflora, we compared known Al-responsive gene sequences from the literature against the Q. grandiflora genome using Blast2GO and BLAST/NCBI [23]. Sequences with the highest similarity and lowest e-values were selected for further analysis (Table S2). Several key genes linked to Al tolerance were identified, including those involved in Al transport, organic acid metabolism, plant growth regulation, the methylation cycle, and cell wall synthesis.

Figure 2. Gene ontology (GO) analysis of gene sequences of Qualea grandiflora Mart. (Vochysiaceae). A) Qualea grandiflora sequences associated with some Biological Process categories. B) Main molecular function categories that of gene sequences of Q. grandiflora according to GO analysis. C) Cellular component categories of some Q. grandiflora genes as determined by GO analysis.

Figure 2. Gene ontology (GO) analysis of gene sequences of Qualea grandiflora Mart. (Vochysiaceae). A) Qualea grandiflora sequences associated with some Biological Process categories. B) Main molecular function categories that of gene sequences of Q. grandiflora according to GO analysis. C) Cellular component categories of some Q. grandiflora genes as determined by GO analysis.

Notably, transmembrane transporter genes such as ALMT (aluminium-activated malate transporter), MATE (multidrug and toxic compound extrusion family), ABC (ATP-binding cassette transporter), and NRAT (natural resistance–associated macrophage protein 1) were detected [3,15,24,25]. Transcriptomic analysis revealed that four ABC transporter genes were differentially expressed under Al exposure, with one upregulated in Q. grandiflora leaves [14]. In contrast, NRAT family genes showed no differential expression in response to Al [12,26]. Surprisingly, most Al-tolerance-related genes in Q. grandiflora were not upregulated under Al stress; instead, many were downregulated in both leaves and roots [12,26]. This suggests that their regulation may depend on additional endogenous or environmental factors beyond Al exposure alone [27].

Genes involved in essential cellular processes were also identified, including SAMS2_1 and SAMS2_2 (S-adenosyl methionine synthetase 2), which catalyse the synthesis of S-adenosyl methionine (SAM) – a critical methyl group donor for DNA, RNA, protein, and cell wall methylation, as well as ethylene and polyamine biosynthesis [28,29,30]. Intriguingly, SAMS2 was upregulated in Q. grandiflora roots under Al stress [12,26]. Additionally, we identified SAHH2 (S-adenosylhomocysteine hydrolase 2) and adenosine kinase (ADK; IPR001810), both crucial for maintaining SAM-dependent methylation reactions [31,32,33].

Cell wall-related genes were also prominent, including PME29, a pectin methylesterase involved in modulating pectin methylation – a key determinant of cell wall rigidity [32,34,35]. Given that the cell wall is a primary site of Al accumulation and toxicity [3,34,36], the upregulation of PME genes in Al-treated roots [12,26] suggests enhanced cell wall remodelling as a compensatory mechanism.

This study identified Al-responsive genes in Q. grandiflora, highlighting potential mechanisms of tolerance and physiological effects of this metal in this species. However, functional characterization of these genes and further investigation into their regulatory dynamics in the presence of Al are needed to fully elucidate their roles.

3. Material and Methods

3.1. Plant Material and DNA Extraction

Qualea grandiflora seeds were collected from a Cerrado stricto sensu area (14° 42' 19.31" S; 47° 40' 11.81" W) in the municipality of Água Fria de Goiás, Goiás, Brazil. Samples from the mother plants were herborized and incorporated into the Herbarium of the University of Brasília (UB); Voucher: 217284.

After germination, leaves from Q. grandiflora seedlings were frozen in liquid nitrogen and transported to the Plant Biotechnology Laboratory at the University of Brasília (UnB), for total DNA extraction. Total DNA was extracted using the Qiagen DNeasy Plant Kit (Hilden, Germany). The concentration and quality of the DNA were assessed using a NanoDrop 1000 (Thermo Fisher Scientific, Wilmington, USA), a Bioanalyzer 2100 (Agilent Technologies Inc., Santa Clara, USA), a Qubit fluorometer (Life Technologies, Thermo Fisher Scientific, Wilmington, USA), and a 1% agarose gel (Figure 1).

Figure 1. Quality and integrity of the total DNA extracted from Qualea grandiflora Mart. A) Agarose gel showing the integrity of Q. grandiflora DNA. B) Bioanalyzer data on Q. grandiflora DNA quality and integrity (by LACTAD).

Figure 1. Quality and integrity of the total DNA extracted from Qualea grandiflora Mart. A) Agarose gel showing the integrity of Q. grandiflora DNA. B) Bioanalyzer data on Q. grandiflora DNA quality and integrity (by LACTAD).

3.2. Genome Sequencing and Assembly

Sequencing was performed on an Illumina HiSeq 2500 platform (San Diego, CA, USA), and two paired-end libraries were constructed using the TruSeq DNA Nano Low Throughput Library Prep Kit (Illumina, San Diego, CA, USA), with read sizes ranging from 100 to 300 bp.

Prior to assembly, Illumina reads were filtered and trimmed using Trimmomatic v.0.20 [37] to remove the TruSeq adapters. For the trimmed reads, FastQC (v0.11.3) [38] was used to plot quality scores and sequence length distribution. K-mer analysis was performed in the libraries. Next, genome size estimation was conducted using GenomeScope [39] with Jellyfish v.2.0 [40] to count and build the histogram of K-mers.

The genome was assembled using ABySS v2.0 [41] and DISCOVAR de novo [42]. Both assemblers follow the classic De Bruijn graph approach.

To assess the quality of the de novo genome assembly, we used BUSCO v3.0.2 [43] against the Embryophyta ortholog database. BUSCO classified the sequences as complete and single-copy (S), complete and duplicated (D), fragmented (F), or missing (M). In addition to assessing the relative gene completeness of the assemblies, the measurement of duplicated hits is useful for evaluating the extent to which the assembler may falsely expand genomic regions, particularly those that are highly polymorphic [43].

3.3. Genome Annotation

For gene prediction and annotation, the Funannotate pipeline v1.8.1 [44], AUGUSTUS v2.5.5 [45] SNAP [46], and GlimmerHMM [47] were used. First, Funannotate cleaned up and sorted the sequences. Then, the RepeatModeler v2.0/RepeatMasker v4.0.1 software identified repetitive elements [48]. Additionally, Funannotate was trained using RNA-Seq data from Q. grandiflora (SRA: SRR5248188, SRA: SRR5248189, SRA: SRR5248190, SRA: SRR5248191, SRA: SRR5248192, SRA: SRR5248193) and the Uniprot Myrtales protein sequence database. AUGUSTUS was trained using alignments from the BUSCO dataset [43].

Functional annotations of genes were predicted based on the SwissProt database [49] and the Eukaryotic Orthologous Groups (KOG) protein database [50], using an E-value threshold of <1e-5. Gene classification was performed using InterProScan (v5.20-59.0) for Gene Ontology (GO) terms [51] were evaluated using Blast2GO software [52].

All proteins were assigned based on their similarity to the Pfam database [53] and compared for carbohydrate-active enzyme domains (CAZymes) using GlimmerHMM v3.0.472 [47]; and were also compared to the MEROPS [54] and eggNOG v4.5 [55] databases using BLASTP (version 2.7.1) with an E-value of 1e-5 [56].

The analysis of metabolic pathways was performed by searching the KEGG database (Kyoto Encyclopedia of Genes and Genomes) with an E-value cut-off of 1e-5 using the BlastKOALA mapping platform (KEGG Orthology and Links Annotation) [57]. Functional annotation was based on the following parameters: the Plants taxonomic group and the family eukaryotes database as reference.

3.4. Al-Responsive Genes

Gene sequences associated with Al responses were selected from the literature, combined with BLAST results from NCBI (National Center for Biotechnology Information) (http://blast.ncbi.nlm.nih.gov/Blast.cgi). Batch BLAST similarity searches were performed automatically on the database. We used RNA-Seq data from Q. grandiflora (SRA: SRR5248188, SRA: SRR5248189, SRA: SRR5248190, SRA: SRR5248191, SRA: SRR5248192, SRA: SRR5248193).

4. Conclusions and Perspectives

This study is a high-significant genome sequencing of Q. grandiflora, achieving a total genome coverage of 81% with excellent metrics, particularly considering the lack of a reference genome and the non-model status of the species (BioProject PRJNA786741). Additionally, 11 gene families potentially involved in Al response mechanisms were identified in Q. grandiflora.

Thus, this study also provides:

The first comprehensive genome resource for an Al-hyperaccumulating Cerrado species

Evidence for both conserved (e.g., transporter families) and unique (e.g., constitutive expression) Al-tolerance strategies

Candidate genes for functional validation, particularly:

PME29-mediated cell wall modification

SAM-dependent methylation pathways

Key unresolved questions to be addressed include:

The role of protein-level regulation in Al metabolic responses in Q. grandiflora and other native Al-accumulating species

Ecological implications of constitutive vs. inducible tolerance

Evolutionary origins of Al hyperaccumulation in Vochysiaceae