Genotype-Driven Proteomic Diversity in Macadamia Nuts: Implications for Allergenicity, Nutritional Quality, and Breeding Strategies

Abstract

Macadamia nuts are valued for their nutritional content, yet little is known about the factors that affect nut quality or the genetic contributions of parental lines. This study optimized a protein extraction protocol for lipid-rich macadamia nuts and applied SDS-PAGE and mass spectrometry-based proteomics to characterize nuts with known parental origin. A total of 431 high-confidence proteins were identified, with seed storage proteins (SSPs), primarily vicilin- and legumin-like globulins, comprising nearly 50% of total protein abundance. Distinct SDS-PAGE banding patterns among genotypes, including within maternal lines, indicated significant paternal influence on protein profiles. Correlation analysis confirmed that specific gel bands reveal the abundance of allergenic SSPs. Moreover, sequence coverage and post-translational modifications of SSPs varied across genotypes, indicating molecular divergence. These findings provide insights into the genetic basis of nut protein composition, inform allergenicity risk assessment, and support breeding strategies aimed at improving macadamia nut quality and developing hypoallergenic cultivars.

Introduction

Macadamia, a genus of Proteaceae, is native to Australia and is composed of four main species; M. integrifolia, M. jansenii, M. ternifolia, and M. tetraphylla. Since its introduction to Hawaii from Australia in the 1920s, macadamia has been successfully cultivated and commercialized as a nut crop. Commercial production primarily relies on two species, M. integrifolia and M. tetraphylla. Driven by the high demand for macadamia nuts, the industry has seen a rapid increase in production over the past 10 years, expanding cultivation into 18 countries. Beyond their economic values, macadamia nuts are renowned for their health benefits, including their ability to reduce low-density lipoprotein cholesterol and mitigate coronary artery disease risk factors. ,

Rich in fat (59.2–78.4%) and fiber (6.4–8.6%), macadamia nuts also contain moderate levels of protein (7.9–8.4%) and relatively little sugar (1.4–4.6%). Recent studies in nut proteomics demonstrate that characterizing seed protein profiles offers insights far beyond basic compositional analysis. In species such as peanut, walnut, and almond, detailed proteome profiling has enabled the identification of major seed storage proteinsvicilins and leguminsas key determinants of nutritional quality and allergenicity, informing both dietary evaluation and food safety risk assessment. , Although macadamia nuts contain less total protein than other tree nuts such as almonds and walnuts, they are particularly rich in 7S vicilin and 11S legumin storage proteins, which provide essential amino acids and contribute significantly to their nutritional quality. , Recent clinical findings have further expanded the allergen repertoire of macadamia by identifying additional IgE-reactive proteins, including oleosin, pectin acetylesterase, and aspartyl protease.

Proteomic approaches have also been used to evaluate how processing, such as roasting, influences protein structure, digestibility, and allergen potential, as well as to identify genotype-specific protein expression markers associated with stress tolerance and postharvest quality. , While proteomic applications in macadamia are still emerging, the identification of key seed storage proteins and functional enzymes in this species highlights its potential for advancing nutritional assessment, allergen management, and breeding for improved nut quality.

Previous observations in macadamia suggest that paternal genetic background may influence nut development, though the molecular basis underlying such effects remains poorly understood. While both cross- and self-pollination events occur in macadamia, molecular paternity testing using microsatellite markers has revealed an unexpected high selfing rate in some cultivars of this crop. This can be advantageous for conserving the beneficial genetic structure of the crop, which can result in more stable and reliable production, whereas cross pollination has been shown to increase yield in many fruit crops.

Characterizing the macadamia nut proteome and linking variation in protein abundance to nut paternity can help to better understand parental influences and uncover novel molecular markers linked to desirable traits, such as improved nutritional profiles, reduced allergenicity, and stress tolerance, providing a foundation for targeted breeding strategies. However, despite its importance, the macadamia nut proteome remains poorly studied compared to other major tree nuts. Recent advances in proteomics technology, combined with the availability of a chromosome-scale genome assembly, present an opportunity to fill this knowledge gap.

This study adopts a novel perspective by exploring how parental genetic contributions shape the macadamia nut proteome. By linking proteomic variation to nut paternity, it provides new insights into the molecular basis of genotype-dependent differences in protein composition, including storage protein abundance, post-translational modifications, and allergenic potential. These findings enhance our understanding of nut development and physiology and have important implications for improving nutritional quality, minimizing allergenicity, and identifying molecular markers to support precision breeding and cultivar selection in macadamia.

Materials and Methods

Plant Material

Macadamia nuts previously used by Langdon et al. for DNA paternity testing were selected for the current study to investigate nut protein characteristics. The nuts were sourced from a Regional Varietal Trial established in 1984 within a large commercial orchard at Clunes, New South Wales, Australia. This trial comprises a diverse collection of macadamia cultivars, including Hawaiian cultivars developed at Hawaii Agricultural Research Station (HAES), such as HAES 791, HAES 344, and HAES 246; Australian cultivars developed at Hidden Valley Plantation (e.g., ‘A4’ and ‘A16’); and others such as Daddow ‘DAD’, Release ‘REL’, Own Venture ‘OV’, ‘NRG’, and the X series. For clarity, the naming convention used throughout this study follows the format ‘CultivarNumberRowID’ (e.g., 791R3) to indicate a specific tree of a given cultivar located in a particular row of the varietal trial, and ‘CultivarNumberNutID’ (e.g., 791N2) to indicate an individual nut collected from the corresponding maternal tree. Nuts were collected directly from multiple branches across the canopy of each tree to minimize sampling bias. After collection, the husks were removed and the seed coats were cracked and discarded. Nuts were immediately dried using self-indicating silica gel at a ratio of 10 times the wet weight. The dried nuts were then sealed in individual Ziplock bags until used for protein extraction.

Protein Extraction

To optimize protein extraction from macadamia for SDS-PAGE, a protocol was adapted from Rahman et al. with modifications to address the unique challenges presented by the lipid-rich nature of macadamia nuts. Initial trials tested water, methanol, and 70% ethanol as homogenization solvents to determine their effectiveness in removing excessive lipids. These trials also included sampling different regions of the nuts to evaluate protein distribution within a single nut.

The final optimized protocol involved slicing about 50 mg of nut tissue and placing it into a 2.0 mL microcentrifuge tube (Safe-Lock Tubes, Eppendorf, Germany) with a 5 mm stainless steel bead (Qiagen GmbH, Hilden, Germany) and 1.5 mL of methanol. Samples were homogenized using a Qiagen TissueLyser (Qiagen, Retsch GmbH, Germany) at 30 oscillations per second for 1 min. The homogenates were centrifuged at 25,155 × g at 4 °C for 30 min. The top clear solution and lipid layer were removed. The remaining protein pellet was freeze-dried and stored for experiments.

Trichloroacetic acid (TCA) precipitation followed the procedure described by Guo et al. with minor adjustments. To the dried protein pellet, 200 μL of 10 × TE (Tris EDTA pH 7.5), 200 μL of 0.3% (w/v) sodium deoxycholate, and 200 μL of 72% (w/v) TCA were sequentially added. Samples were vortexed between every step, incubated on ice for 1 h, and then centrifuged at 25,155 × g at 4 °C for 20 min. The supernatant was removed and discarded. The pellet was washed twice with 1 mL of 90% methanol and incubated at −20 °C overnight. Two additional washes were performed using acetonitrile. The final purified protein samples were air-dried in a fume hood.

Gel Electrophoresis

The dried nut pellets (∼2 mg), after homogenization with methanol and TCA precipitation, were suspended in 100 μL of loading buffer. The loading buffer was made by mixing 500 μL of Milli-Q water, 475 μL of 2× Laemmli sample buffer (Bio-Rad Laboratories, Hercules, California, USA), and 2.5 μL 2-mercaptoethanol. The sample solution was vortexed thoroughly to ensure complete dissolution of the protein pellet and then centrifuged at 25,155 × g at 4 °C for 20 min to remove insoluble debris. Proteins were separated by SDS-PAGE using a Mini-PROTEAN Tetra Cell system (Bio-Rad Laboratories, Hercules, California, USA). A 12% resolving gel and a 4% stacking gel were prepared using standard recipes. Each gel lane was loaded with 10 μL of the protein mixture. A molecular weight marker (Precision Plus Protein Dual Color Standards, Bio-Rad Laboratories) was included to estimate protein sizes.

The gels were run at 120 V for 90 min in 1× running buffer (25 mM Tris, 192 mM glycine, and 0.1% SDS, pH 8.3). After electrophoresis, gels were stained with Coomassie Brilliant Blue R-250 solution for 1 h and then destained in a solution containing 10% acetic acid and 40% methanol until protein bands were visible. Gel images were captured using a Gel Doc EZ Imager (Bio-Rad Laboratories). The intensity of protein bands was quantified using Bio-Rad Image Lab software (version 6.1, Bio-Rad Laboratories, Hercules, California, USA). The software’s lane and band analysis tools were employed, with background subtraction set to the rolling disk method and sensitivity adjusted to detect faint bands. Quantification results were exported for downstream analysis to assess protein abundance across samples. Protein bands that showed significant differences between cultivars were excised from the gels, destained, and digested with trypsin. The resulting peptides were subjected to nano-HPLC-MS/MS for downstream proteomic analysis.

Nano-HPLC-MS/MS Analysis

Protein samples were analyzed using an Eksigent, Ekspert NanoLC 400 uHPLC coupled to a TripleTOF 6600+ System (SCIEX, Canada) equipped with a PicoView nanoflow ion source (New Objective, USA). Protein extract (up to 5 μL) was injected onto a ChromXP C18-CL column (3 μm, 75 μm x 150 mm) (SCIEX, Canada). Mobile phase solvents consisted of solvent A; 0.1% formic acid in water, and solvent B; 0.1% formic acid in acetonitrile. Linear gradients of 5–30% solvent B were run over 120 min at 400 nL/min flow rate, followed by 30–90% solvent B for 3 min, and then 90% solvent B for 17 min, for peptide elution. The gradient was then returned to 5% solvent B for equilibration prior to the next sample injection. Column temperature was maintained at 45 °C throughout. The ion spray voltage was set to 2600 V, declustering potential at 80 V, curtain gas flow 25 psi, nebulizer gas 30 psi, and interface heater at 150 °C. The mass spectrometer was set to acquire 100 ms of full-scan TOF-MS data over the mass range 350–1500 m/z, followed by up to 50 50 ms full scan product ion data in information-dependent acquisition (IDA) mode over the mass range 100–1500 m/z. Ions observed in the TOF-MS scan exceeding a threshold of 100 counts and a charge state of +2 to +5 were set to trigger the acquisition of product ion MS/MS spectra of the resultant 50 most intense ions. Raw MS/MS spectral data were obtained from Analyst TF 1.7 software (SCIEX, Canada).

Protein Identification and Quantification

Spectral processing and database searching were performed in ProteinPilot 5.0.2 (SCIEX, Canada) using Paragon and Pro Group algorithms. The searches utilized a protein database derived from the first published Macadamia integrifolia genome assembly (Nock et al.; 34,274 protein-coding genes; available at Southern Cross University’s repository: 10.25918/5e320fd1e5f06). The encoded mzIdentML files were used for the downstream analysis. Scaffold 4.8.6 (Proteome Software, Portland, Oregon, USA) was used to validate MS/MS-based peptide and protein identifications and for quantification, using the same searching database previously used for ProteinPilot. According to the Peptide and ProteinProphet algorithms implemented into Scaffold, identifications were accepted if they could be established at a probability greater than 90 and 99.9%, for peptides and proteins, respectively, and if proteins contained at least two unique peptides. Proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony. Normalized spectral abundance factor (NSAF) was used for protein quantification.

Bioinformatics Analysis of Macadamia Nut Proteome

Functional annotation of the identified macadamia nut proteome was undertaken using the best BLASTP hits to sequences found in the National Center for Biotechnology Information (NCBI), using the UniprotKB/SwissProt Arabidopsis thaliana. The matched A. thaliana genome accessions (AGIs) were then submitted for GO enrichment analysis and data visualization using ClusterProfiler and Cnetplot packages in R version 4.2.1.

Results

Gel Electrophoresis

The effectiveness of different solvents in solubilizing macadamia nut protein and removing lipids was evaluated. Gel electrophoresis results indicated that methanol was the most effective solvent, as protein bands from methanol-soluble fractions resuspended after lipid removal were the sharpest and most distinct (Figure S1). These trials also included sampling different nut regions to evaluate protein distribution within a single nut, which showed no significant differences between regions (Figure S1). In subsequent experiments, methanol was chosen as the solvent of choice for homogenizing macadamia nuts to prepare samples for gel electrophoresis.

Nut proteins extracted from 13 individual nuts, all from the maternal tree cultivar ID 791R3, were analyzed. Interestingly, the protein banding patterns among these individual nuts were not identical, as shown in Figure S2. This variation suggested a potential influence of the paternal parent on the nut proteome, as the genetic contribution from the paternal side may lead to differences in protein abundance among progeny. To further validate the influence of paternal contribution on macadamia nut proteomes, we analyzed 54 nuts derived from five maternal trees with candidate paternal parents. The paternity of these nuts was determined directly from the results reported by Langdon et al., utilizing molecular analysis with microsatellite markers and a maximum likelihood approach to identify the most likely paternal parent based on allele frequencies and genotype matching.

Protein extracts from these nuts were separated by SDS-PAGE, and the intensity of protein bands was quantified using Bio-Rad Image Lab software. Heatmaps were generated to visualize the intensity differences of distinct bands across nuts (Figure ). The heatmap revealed that even among nuts with the same maternal parent, the intensity of protein bands varied significantly.

1.

Heatmap showing protein band intensity variations among macadamia nuts derived from five known maternal lines. Nut protein extracts from 54 individual nuts, each with candidate paternal parentage determined by microsatellite marker analysis, were analyzed by gel electrophoresis. Band intensities were quantified using Bio-Rad Image Lab software. Rows represent distinct protein bands, columns represent individual nuts, and the color gradient indicates relative intensity. Bands B1–B5 correspond to the selected bands indicated in the original gel images provided in Figure S2.

Nuts from the same maternal parent and sharing the same candidate paternal parent, such as maternal tree 246R1 with candidate paternal parent ID 344, exhibited highly similar protein band intensity patterns. Conversely, nuts from different maternal–paternal combinations, such as 791R3 with candidate paternal ID 344 and the same maternal parent with paternal ID 791, showed substantial differences in band intensity, indicating distinct proteomic profiles.

Characterization of Macadamia Nut Proteome by Mass Spectrometry Analysis

To capture a broader representation of proteomic diversity and investigate potential influences of parental contributions, samples were selected based on distinct protein banding patterns observed in gel electrophoresis. Although the selected samples (791N2, 246N2, and 344N6) derived from different maternal sources, the variation in their banding patterns was hypothesized to represent proteomic diversity that could be associated with distinct genetic and physiological characteristics and provided an opportunity to explore paternal genetic contributions. Protein samples of the three selected samples were subjected to LC-MS/MS analysis.

In total, 431 proteins were identified at a protein threshold of 99.9% probability with minimum two unique peptides identified from 85,615 spectra across the three macadamia nut samples (Table S1). These proteins ranged in theoretical molecular weight from 7 to 258 kDa (Table S1). Sequences of the identified macadamia nut proteins were submitted to NCBI, and the functional annotation of these proteins was undertaken using the best BLASTP hits with e-value <10^–5 against the UniprotKB/SwissProt Arabidopsis thaliana database, which resulted in the export of 402 annotated proteins (Table S2). Protein abundance was estimated using the NSAF method.

In terms of the relative abundance of the identified proteins, nutrient reservoir proteins, composed of seed storage proteins, including legumins and vicilins, accounted for almost 50% of all proteins detected (Figure nutrient reservoir proteins). The proteins with the second highest abundance were those in the category of lipid metabolic process (∼11%), followed by carbohydrate metabolic process (∼9%) in the third place. Proteins involved in the oxidation–reduction process, stress response, and cellular homeostasis had similar abundance in macadamia nuts, accounting for 4–5% of the total protein content. In addition, proteins responsible for protein metabolic processes, such as amino acid metabolism, translation, protein binding, and degradation comprised 1–3% of the detected proteins, whereas the remaining categories such as cell wall organization, signal transduction, and proteasome, nucleotide, and glutathione secondary metabolic processes individually made up less than 1% (Figure ).

2.

Functional categorization and relative abundance of identified macadamia nut proteins. Proteins identified from the macadamia nut samples were functionally annotated using the best BLASTP hits (e-value <10^–5) against the UniProtKB/SwissProt Arabidopsis thaliana database. The chart shows the relative abundance of proteins categorized into functional groups based on Gene Ontology (GO) terms. Protein abundance was calculated using the NSAF method, and abundance values were averaged across the three macadamia nut samples.

Of all the proteins identified, 320 were present in at least two out of three macadamia nuts as indicated by the inner area of the blue dashed line in the Venn diagram in Figure A. Gene ontology (GO) enrichment analysis was then performed on the matched Arabidopsis accessions of these proteins. A total of 267, 287, and 270 proteins were assigned to at least one biological process (BP), cellular component (CC), and molecular function (MF) GO annotation term, respectively (Table S3). Moreover, the top 20 enriched GO terms of BP, CC, and MF are represented in Figure B, Figure C, and Figure D, respectively.

3.

Comparative and functional analyses of the macadamia nut proteome. (A) Venn diagram showing the number of proteins shared between three nuts. Enrichment network depicting the top 20 most significantly enriched GO terms of (B) BP, (C) CC, and (D) MF (hypergeometric test with Bonferroni correction, p-value <0.01) in the nut proteome. Each GO term is represented as a circle; the size of the circle reflects the number of the proteins, and the color indicates the statistical significance.

Significantly enriched (Bonferroni adjusted p-value <0.01) GO terms associated with BP showed a big cluster of metabolic and cellular processes. These include ADP metabolic process, ATP metabolic process, ATP generation from ADP, glycolytic process, gluconeogenesis, pyruvate metabolic process, nucleoside diphosphate metabolic process, nucleotide metabolic process, purine-containing compound metabolic process, and monocarboxylic acid metabolic process (Figure B and Table S3). Identified proteins were also enriched in GO pathways of response to chemicals and environmental stimulus, such as response to metal ion, response to heat, response to cold/freezing, response to water stress, response to salt stress (Figure B and Table S3). The macadamia nut proteome was also significantly enriched in terms for seed development, lipid storage, protein metabolic process, response to oxidative stress and cell redox homeostasis (Figure B and Table S3).

Significantly enriched CC GO terms shown in Figure C indicated enrichment of terms such as secretory vesicle, apoplast, peroxisome, microbody, stromule, lipid droplet/storage body, and protein storage vacuole, also in the nonmembrane-bound ribosome, intracellular protein-containing complex, extracellular region apoplast, and cell wall. Furthermore, the enriched molecular function GO terms (Figure D and Table S3) of macadamia nut proteins are protein folding, structural molecule activities, and binding such as protein binding, ATP binding, copper ion binding, and catalytic activities including oxidoreductase, lyase, isomerase, and antioxidant activities.

Proteomic Analysis of Excised Protein Bands Reveals Variability in Macadamia Nut Storage Proteins

To explore the underlying causes of differences observed in gel electrophoresis banding patterns among macadamia nuts from different parents, protein bands that showed significant differences among the three samples were excised from the gels and subjected to independent nano-HPLC-MS/MS analysis. These bands were selected based on their differential appearance across the three samples (791N2, 246N2, and 344N6), which were previously shown to have distinct overall proteomic profiles. Visual inspection of SDS-PAGE revealed that bands B2 and B5 were absent in 246N2, while bands B4 and B5 were significantly fainter in 344N6 (Figure A).

4.

SDS-PAGE banding variation and MS-based correlation analysis link gel patterns to underlying storage protein composition in macadamia nuts. (A) Gel electrophoresis showing distinct banding patterns among macadamia nuts 246N2, 344N6, and 791N2. Protein bands with significant differences were excised for proteomic analysis. Lanes with the same sample label represent independent technical replicates from the same nut extract. Marker sizes (kDa) are indicated on the right. (B) Correlation matrix showing Pearson correlations between SDS-PAGE band intensities and MS-quantified protein abundance for selected seed storage protein families. Each dot represents the correlation between a specific protein and a gel band; dot size corresponds to the magnitude of the correlation coefficient (|r|), and asterisks indicate significance (*p < 0.05; **p < 0.01). Protein names are color-coded by family, and the green box highlights key correlations of interest.

LC-MS/MS analysis identified the major proteins in these bands as 11S legumin-like and 7S vicilin-like globulins, the principal SSPs in macadamia. While both types of proteins were consistently identified, their sequence coverage and post-translational modifications (PTMs) varied across genotypes (Figure S3), suggesting structural or functional divergence. For example, certain bands in 246N2 showed higher sequence coverage of 11S and 7S proteins, while the same bands in other samples had reduced coverage or distinct PTMs. 11S globulin peptides were mainly detected in bands near ∼67 kDa (e.g., B2), rather than as dissociated acidic/basic chains, whereas 7S vicilins and vicilin-like antimicrobial peptides appeared in multiple bands (Table S4). These results indicate that both abundance and modification differences contribute to the unique banding profiles of each genotype.

To assess whether protein band intensities observed in SDS-PAGE gels corresponded to the abundance of specific seed storage proteins identified by MS, a correlation analysis was performed between gel-derived band intensities and MS-based protein abundances of key storage protein families (legumin-type globulins, vicilin-type globulins, oleosins, lipocalins, oil body proteins, and late embryogenesis abundant proteins). The correlations are visualized in Figure B, with a green box highlighting the region of interest. The analysis revealed that band 1 and band 2 intensities were positively correlated with the abundance of vicilin-like seed storage proteins (particularly MintSC108g00230 and MintSC108g00240) and lipocalins, while showing negative correlations with legumin-type globulins and oleosins. In contrast, bands 3 and 4 exhibited an inverse pattern, with higher correlations to legumin-type globulins and oleosins, and negative associations with vicilins and lipocalins. These results show that different bands reflect the presence of distinct protein types. Furthermore, bands 5–7 showed moderate to strong correlations with lower-abundance vicilin-like proteins (Mint03g06120, Mint14g04820, and Mint02g04520) and a suite of late embryogenesis abundant proteins. Notably, band 5 was positively correlated with these proteins, while bands 6 and 7 displayed negative correlations, indicating a possible transition in protein composition categories across these gel regions.

A hierarchical clustering heatmap was generated based on the NSAF of all identified proteins. Six clusters were identified, among which C1, C3, and C5 represent the proteins with relatively high abundance in nuts 246N2, 344N6, and 791N2, respectively (Figure A). GO enrichment analyses of proteins involved in clusters C1, C3, and C5 indicated that these proteins were clustered to varied biological processes, potentially contributing to the physiological and quality differences observed among the nuts (Figure B). For instance, proteins with higher abundance in 344N6 (Figure A, cluster C1) were significantly enriched in metabolic and cellular processes such as ATP/ADP metabolic process, purine nucleotide/ribonucleotide metabolic processes, and glycolytic process. In contrast, proteins abundant in 791N2 (Figure A, cluster C3) were strongly enriched in oxidative stress responses, such as cellular response to reactive oxygen species, reactive oxygen species metabolic process, cellular oxidant detoxification, and cell redox homeostasis. Furthermore, proteins with higher abundance in 246N2 (Figure A, cluster C5) were enriched in pathways related to environmental stress responses, including water stress and endoplasmic reticulum stress, indicating potential links to environmental conditions during development.

5.

Comparative studies showing quantitative differences between macadamia nut proteomes. (A) Heatmap showing the hierarchical relationship (Ward’s method) between the relative abundance (NSAF) of the identified proteins from three macadamia nuts. (B) Enrichment network representing the top 20 most significantly enriched biological process GO terms (hypergeometric test with Bonferroni correction, p < 0.01). Each GO term is represented as a circle, and the size of the circle reflects the number of the proteins. Different clusters and their contributions to the overall enrichment are represented by different colors and proportions.

Discussion

Compared to the extensive body of research on the lipid composition of macadamia nuts, , studies characterizing their proteomes remain limited. Early studies, such as those by Sathe et al., investigated various methods to solubilize proteins from edible nuts, including macadamia. They optimized conditions for protein solubility using different extraction solvents and electrophoretic techniques, with sodium borate buffer performing best for most nuts. However, the primary focus of this work was to understand protein properties in the context of food allergenicity and functional food development. Likewise, Ghorbani and Morgan developed an ELISA method for detecting trace amounts of macadamia proteins in food, a crucial advancement for allergen labeling and ensuring consumer safety. Marcus et al. discovered a new family of antimicrobial peptides (MiAMP2) derived from a 7S globulin precursor protein, which exhibited strong antifungal activity against plant pathogens, highlighting the defensive roles of certain macadamia proteins. Rost et al. conducted the first label-free shotgun proteomics analysis of macadamia nuts, identifying 947 proteins. However, their use of a generic NCBI protein database, Viridiplantae (for all green plants), and the acceptance of only single-peptide based identifications, while increasing the number of identified proteins, also introduced a high risk of false positives as it is quite common that peptides can be shared by multiple proteins.

Nock et al. made a significant breakthrough in macadamia genomics by publishing the first chromosome-scale genome assembly for macadamia, predicting 34,274 protein-coding genes. This genomic resource provides a robust foundation for constructing an accurate and comprehensive protein database to support macadamia proteomics studies. In this study, we used this database for peptide identification, ensuring high confidence in protein-level assignment. For functional annotation, however, we performed BLASTP against the Arabidopsis UniProtKB/SwissProt database, which allowed downstream GO enrichment analysis based on well-characterized model plant annotations, not yet available for macadamia.

Based on ontological pathway assignment, proteins involved in carbohydrate metabolism, amino acid biosynthesis, protein processes, nucleic acid metabolism, stress response, and energy metabolism were well-represented in the proteome in this study (Figure ). Notably, proteins associated with carbohydrate metabolism constituted a significant portion (9.4% relative to total abundance). These proteins were mapped to key metabolic pathways including glycolysis/gluconeogenesis, pentose phosphate, TCA cycle, pyruvate metabolism, fructose and mannose metabolism, and starch and sucrose metabolism.

Additionally, the glyoxylate cycle, a modified formed of TCA cycle, was also identified. Unique to this pathway, the enzymes isocitrate lyase and malate synthase were detected (Table S3). The presence of these enzymes suggested that the macadamia nuts actively utilize lipid reserves during germination and seedling establishment, as the glyoxylate cycle is critical for converting stored lipids to usable energy and carbon sources. Further support for this is the presence of oleosin and caleosin proteins, which function in docking oil bodies to glyoxysomes and mobilizing triacylglycerol, underscoring the importance of lipid metabolism in the nut.

A study by Lin et al. used weighted gene coexpression network analysis to identify nut development-associated modules significantly linked to lipid metabolism, particularly to fatty acid biosynthesis, elongation, and desaturation. Consistent with these findings, our proteomic analysis of mature nuts identified key enzymes involved in these pathways (Figure and Table S3), such as acetyl-coA carboxylase1 (ACC1) and enoyl-[acyl-carrier-protein] reductase [NADH] 1 (MOD1), which catalyze the initial and final steps of the de novo synthesis of fatty acids.

The diversity of fatty acid profiles among macadamia nuts has been demonstrated in a study of 197 accessions, which revealed significant variation in the abundance and activity of fatty acid desaturases (FADs), as these enzymes play a key role in determining fatty acid saturation. Previous studies have shown that manipulating FAD expression through overexpression or gene knockout can dramatically alter fatty acid saturation levels and therefore affect the quality of the seeds. − In this study, stearoyl-[acyl-carrier-protein] 9-desaturase 7 (FAB2) was the only desaturase identified. Interestingly, its abundance varied significantly between nuts of different parents (NSAF: 5.08 × 10^–05, 1.68 × 10^–03, and 3.10 × 10^–04) (Table S1 and Figure ). This variability suggested potential differences in fatty acid composition among parents, which could influence both the nutritional value and functional properties of macadamia nuts.

Enzymes involved in cyanogenesis, a defense mechanism in members of the Proteaceae family, were identified in this study with variable abundance between cultivars. Specifically, two β-glucosidases, integral to cyanogenic glycoside biosynthesis, were detected (Table S1). Cyanogenesis is a key defense strategy in macadamia, releasing toxic hydrogen cyanide upon tissue disruption to deter herbivores. Transcriptomic studies have shown a high expression of a β-glucosidase gene in macadamia leaf, shoot, and flower bud. , The observed variability in β-glucosidase abundance among nuts may reflect genetic diversity, including potential paternal contributions, and could have significant implications for food safety. Notably, the highest β-glucosidase (e.g., MintSC3165g00010) abundance was observed in 791N2 (Table S1). This finding is particularly interesting given that HAES 791 is the only major commercial cultivar with confirmed ancestry to M. ternifolia, a species known for its bitter kernels and elevated cyanogenic glycoside content. While cyanogenesis is essential for plant defense, the residual levels of toxic cyanogenic compounds in edible nuts need to be carefully managed to ensure consumer safety, as improper processing or consumption of nuts with high cyanogenic compound levels could pose health risks.

As reproductive tissue, macadamia nuts serve as a nutrient reservoir, supporting plant development. Although macadamia nuts are renowned for their high unsaturated lipid content, they are also a good source of plant protein. This study identified 19 seed storage proteins (SSP), comprising two 12S legumin-type and five vicilin-like globulins (Table S1), which together make up 44–49% of the total protein in the nut (Figure ). Abundance of these SSPs in macadamia nuts from this study aligns with findings from other tree nut species, where such proteins constitute 50–90% of the total protein in plant seed. − A previous study reported a much lower proportion of SSPs (8%) in macadamia nuts. However, our SDS-PAGE gel (Figure A and Figure S2), in which major bands were excised and confirmed by LC-MS/MS analysis, clearly shows a much higher abundance of storage proteins. This discrepancy may arise from differences in extraction protocols, cultivar variation, or the detection sensitivity of analytical methods used in the earlier study. In many plant species, 11S globulins are resolved into acidic (∼37–42 kDa) and basic (∼20 kDa) chains under SDS–PAGE conditions. In macadamia, however, LC-MS/MS mapping of SDS-PAGE bands (Table S4) indicated that 11S peptides were predominantly recovered from ∼67 kDa regions (e.g., B2) rather than from these lower molecular weight positions. This implies that the macadamia 11S globulins remain intact or partially assembled during SDS electrophoresis-PAGE, potentially reflecting unique biochemical stability or quaternary structure. Such structural resilience may have implications for processing behavior and digestibility and warrants further investigation.

SSPs not only provide nutritional value but also contribute to the major allergenic components in most tree nuts. Vicilins and legumins, in particular, are responsible for IgE-mediated responses and cross-sensitization across nut and legume species. They are also known to be potent allergens in macadamia nuts, analogous to major allergenic proteins in other nuts and legumes. In macadamia, these correspond to the allergens Mac i 1 (vicilin) and Mac i 2 (legumin). Although 2S albumins are major allergens in several other nut and seed species, no convincing 2S albumin candidates were detected in our data set. This is consistent with earlier proteomic work on macadamia that also reported no direct matches to cross-species 2S albumins despite extensive coverage of other storage proteins. Biologically, macadamia seeds are dominated by vicilin- and legumin-type globulins, whereas 2S albumins, being small, hydrophobic, and of relatively low abundance in this species, may also be less amenable to recovery by gel-based fractionation and LC–MS/MS workflows. Similar vicilin and legumin proteins in other crops exhibit high allergenicity: For example, Ara h 1 (a 7S vicilin) and Ara h 3 (an 11S legumin) of peanut are major peanut allergens, with Ara h 1 sensitization found in ∼95% of peanut-allergic patients. , Likewise, walnut allergy is largely mediated by homologous SSPs (e.g., vicilin Jug r 2 and legumin Jug r 4). In soybean, the vicilin-class Gly m 5 and legumin-class Gly m 6 represent the principal allergens and are strongly associated with severe allergic reactions. The prominence of vicilin- and legumin-like proteins (Figure ) underscores their allergenic potential, mirroring the situation in peanuts and tree nuts where seed storage globulins dominate the allergen.

The observed variability in SSP profiles among macadamia genotypes carries significant implications for allergenicity management. Distinct SDS-PAGE banding patterns revealed natural variation in the abundance of major SSPs, including vicilin-like and legumin-like globulins (Figure A), mirroring genotype-dependent shifts in Ara h 1 (vicilin) and Ara h 3 (legumin) levels reported in peanut. Densitometric analysis further demonstrated that band intensities correlated strongly with MS-based quantification of these allergenic proteins (Figure B), supporting the use of gel profiles as a rapid, cost-effective screening tool to estimate allergen abundance, consistent with findings in peanut proteomic studies. These correlations suggest that gel profiles could serve as a preliminary screening tool to identify macadamia lines with lower levels of allergenic SSPs before undertaking more extensive and costly analyses.

The ability to manipulate seed storage protein abundance in macadamia has important implications for nut allergies. Harnessing this natural variation in SSP content could inform breeding strategies aimed at manipulating food allergenicity in macadamia. This variation was observed among samples selected to represent different parental genotypes, underscoring the contribution of parental genetics to SSP abundance and composition (Figure ). Leveraging such natural variation may guide breeding strategies aimed at lowering allergen content. In particular, profiling parental protein expression patterns could facilitate the predictive selection of hypoallergenic cultivars, an approach successfully explored in other nut crops. In this context, SDS-PAGE banding profiles, combined with untargeted LC-MS/MS, provide an effective first-tier screening approach for detecting major shifts in allergenic SSP abundance, while also revealing information on protein integrity and assembly not accessible through targeted MS alone. Targeted approaches may be used, where relevant, to complement this broad screening and enable more precise quantification of specific allergens.

Taken together, our findings highlight the critical role of proteomic variability in shaping macadamia nut physiology, nutritional quality, and defense mechanisms. By linking gel banding pattern differences to proteomic data, we provide evidence of how genetic variability, including potential parental contributions, influences key traits such as lipid metabolism. These findings lay the groundwork for further exploration of macadamia nut biology. Future research should focus on integrating proteomics with other omics approaches to better understand the genetic and environmental factors driving these differences. Insights gained could inform breeding programs to enhance macadamia nut quality, food safety, and adaptability to changing environmental conditions, contributing to both scientific knowledge and agricultural sustainability.

The raw mass spectrometry proteomics data generated in this study have been deposited to Southern Cross University research portal (https://researchportal.scu.edu.au/) and are available under DOI: 10.25918/data.486.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jafc.5c07184.

• (Figure S1) optimization of protein extraction from macadamia nuts and SDS-PAGE analysis comparing the effectiveness of different solvents for lipid removal during protein extraction from macadamia nuts; (Figure S2) SDS-PAGE gel showing protein banding patterns of macadamia nuts derived from maternal tree 791R3; and (Figure S3) mass spectrometry analysis of excised SDS-PAGE bands revealing genotype-dependent variation in seed storage protein structure (PDF)

• (Table S1) List of 431 identified proteins from three macadamia nut samples with protein probability ≥99.9% and ≥2 unique peptides; (Table S2) functional annotation of macadamia nut proteins identified by BLASTP against the Arabidopsis Uniprotkb/Swissprot database (e-value <10^–5); (Table S3) gene ontology biological process, cellular component, and molecular function annotations for macadamia nut proteins matched to Arabidopsis accessions (XLSX)

This research is supported by seed funding from SCU, Australia.

The authors declare no competing financial interest.