Delonix regia Seed Germ as an Underutilized Biomass Resource: Nutritional Value, Safety, and Potential for Sustainable Protein Supply for Food Systems

Abstract

Global challenges in food security and sustainable biomass management highlight the need to diversify resource streams that can supply accessible, safe, and environmentally responsible protein. Delonix regia (flamboyant) seed germ (FG) represents an abundant but largely underutilized biomass resource in tropical and subtropical regions, where its seeds are routinely discarded as green waste. This study assesses the resource potential of FG by evaluating its nutritional composition, safety profile, and suitability for integration into sustainable protein provision strategies. The FG fraction was recovered from locally available seed residues and analyzed to determine their proximate composition, essential amino acid profile, and antinutrient content, providing insights into the qualitative attributes of this emerging resource. Safety was examined through in vivo acute toxicity assays and detailed histopathological evaluation of hepatic and renal tissues of CD1/ICR strain female mice, which revealed no morphological indicators of toxicity, inflammation, or cellular damage. The results indicate that FG contains a high protein concentration (78.35%) with a favorable essential amino acid pattern, supporting its potential as a renewable and locally accessible plant-based protein source. Beyond its nutritional value, the valorization of FG contributes to resource efficiency, waste-to-value pathways, and circular economy approaches by transforming an abundant municipal biomass residue into a functional component for sustainable food systems. Overall, the study underscores the feasibility of integrating FG into resource diversification strategies, enhancing protein availability while reducing environmental burdens associated with biomass disposal.

1. Introduction

Historically, meeting the population’s food needs has been achieved by expanding the area under cultivation as well as designing and adapting technological processes to increase production yields without taking the impact on the environment into account [1]. At present, human nutrition has evolved, and technologies developed to meet the demands of the population must adopt a sustainable approach, ensuring that they meet the needs of current and future generations while also preserving ecosystems and supplying food demands [2,3]. However, there are multiple factors that affect food systems and their sustainability, such as population growth, dietary differences and variability, overexploitation of natural resources, and increased biomass, among others [1]. Understanding the interaction of these variables provides relevant knowledge for designing strategies to achieve food security and sustainability, as they are related to several of the 169 targets of the 17 Sustainable Development Goals (SDGs) established for 2030 [4].

Within the impact that food has on the environment, it is widely recognized that meat production and consumption is one of the main causes of climate change due to the large amount of resources it consumes and the carbon emissions it produces. Despite this, it is estimated that livestock production will need to increase by 50–70% by 2050 in order to meet the needs of the world’s population [2]. Meanwhile, it has been reported that frequent consumption of cereals and legumes is not only associated with a reduction in chronic degenerative non-communicable diseases but also reduces the impact on the environment, leaving a significantly lower environmental footprint in terms of emissions, water use, and land use [5]. Current dietary recommendations are based on the double pyramid that links eating habits, health, and climate impact, where plant-based foods have a lower negative effect on the environment and greater health benefits [6].

Therefore, it is important to promote a change in habits and increase the consumption of plant-based proteins, exploring unconventional species to achieve greater diversification. Delonix regia (flamboyant tree) is a protein alternative since, due to its high protein content (12–25%) [7,8], it has been used in the feeding of tilapia, rabbits, and broiler chickens, among others. [7]. Delonix regia is widely distributed across tropical and subtropical regions of Latin America, the Caribbean, West Africa, Southeast Asia, and parts of India, where it is extensively planted as an ornamental tree for urban landscapes, highways, parks, and institutional campuses. In many cities, thousands of mature trees produce large quantities of lignocellulosic pods annually, each containing multiple protein-rich seeds. Although precise global production data are not systematically recorded due to its ornamental status rather than agricultural cultivation, urban forestry assessments indicate that seed biomass generation from ornamental leguminous trees can reach several kilograms per tree per season, representing a substantial yet unquantified biomass stream at a municipal scale [9,10].

As a legume, flamboyant seeds may develop secondary metabolites to protect themselves from predators, which, acting as antinutrients, could be harmful if consumed. However, it has been reported that with proper processing, these metabolites can be eliminated, or their concentration reduced so that they pose no risk to consumers [11]. Yusuf et al. [12] indicated the use of flamboyant seeds in the production of cosmetic products, edible oil, and traditional medicine, among others, without reporting any cases of adverse health effects or poisoning.

The flamboyant tree is mainly used as an ornamental plant due to the beauty of its flowers, so maintaining it represents an expense. It is estimated that caring for a single tree in the city of Colorado, US, for 40 years costs between 100 and 570 USD, and in general, each US city spends between 13 and 65 USD per year per tree [13]. In Ghana, the annual maintenance of a university’s forest area amounted to 8000 USD per year [14], and in the United Kingdom, the operating costs of maintenance, invasive species control, and labor expenses were 3600–3700 USD per hectare during the restoration stage [15].

From a biomass valorization perspective, flamboyant seeds represent a dispersed but recurrent seasonal resource embedded within urban green waste streams. In tropical cities with a high density of ornamental trees, pod fall occurs synchronously during dry seasons, generating concentrated periods of biomass accumulation. However, as with many urban bioresidues, the principal limiting factor for industrial utilization is not biochemical potential but logistical feasibility. The scattered spatial distribution of trees increases collection and transportation expenses, which often represent the largest economic barrier in waste-to-value systems [16]. Nevertheless, when integrated into organized municipal pruning and green waste management programs, these residues can become cost-effective feedstocks within circular bioeconomy frameworks.

Despite the absence of formal production chains, the continuous generation of flamboyant seed residues across tropical urban regions suggests that sufficient availability exists to support localized valorization initiatives. Urban biomass mapping studies have demonstrated that ornamental tree residues can meaningfully contribute to decentralized bio-based production systems when supported by coordinated collection schemes and public–private partnerships [17]. Therefore, beyond its nutritional attributes, FG presents a realistic opportunity for regional-scale protein recovery strategies aligned with circular economy principles, particularly in tropical municipalities where biomass disposal already incurs management costs [1]. The objective of this study was to utilize flamboyant seeds as an underutilized biomass source to obtain protein-rich germ with nutritional value and food safety, evaluated through in vitro and in vivo tests.

2. Materials and Methods

2.1. Sample Collection

Mature flamboyant seed pods were collected in municipal parks in the north of Merida, Mexico (21°01′41.1″ N and 89°37′34.3″ W). An amount of 5 kg of seeds was extracted from their pods and collected, cleaned, and stored in polyethylene bags at room temperature. From these, 300 g batches of flamboyant seeds were taken and ground in a grinding mill (QGC Systems, Model 4-E, Phoenixville, PA, USA) until a seeds flour (FsF) with a particle size of 246 µm was obtained to determine the antinutrient content. Genetic and genomic information for Delonix regia is available in public databases. The complete plastid genome sequence of D. regia has been reported and is recorded in GenBank under accession number MN893243. The plastome exhibits a typical circular structure with a total length of 162,756 bp, comprising two inverted repeat regions (IRs; 25,544 bp each), a large single-copy region (LSC; 92,490 bp), and a small single-copy region (SSC; 19,178 bp). Phylogenetic analyses based on complete plastome sequences available in GenBank indicate that D. regia is nested within the subfamily Caesalpinioideae, supporting its taxonomic placement and genetic identity. All reagents were analytical grade and purchased from J.T. Baker (Phillipsburg, NJ, USA) and Merck (Darmstadt, Germany).

2.2. Flamboyant Germ (FG) Extraction

The flamboyant seeds were soaked in distilled water (1:5 w/v) at 70 °C for 4 h to obtain the germ (FG) by removing the hull and endosperm [18]. The FG obtained was dried at 40 °C for 24 h in an air-circulating oven (Imperial V Lab-Line Model 3476M, Boston, MA, USA) and milled (Thomas-Wiley Laboratory Mill Model 4, Swedesboro, NJ, USA) until obtaining a particle size capable of passing through a 100-mesh sieve (150 µm). Germ extraction yield (% dry basis) from Delonix regia seeds was determined after dehulling, and separation of the cotyledon fraction was calculated as the dry weight of recovered germ (g) in relation with the initial dry weight of whole seeds (g), and is expressed as a percentage.

2.3. Proximate Composition of Raw Material

The proximate composition of FsF and FG was determined as the content of moisture (925.09 method), crude protein was determined as nitrogen in the sample (954.01 method), crude fiber (962.09 method), fat (920.39 method), and ash (942.05 method). All determinations were carried out according to the procedures proposed by AOAC International [19]. Total carbohydrates were estimated by difference at 100% as a nitrogen-free extract (NFE).

2.4. Antinutritional Factors (AF) in FSF and FG

2.4.1. Tannins

The tannin content was measured following the procedures proposed by the International Organization for Standardization ISO 9648 [20]. An initial extraction was carried out as follows: 1 g of sample was suspended in 20 mL of dimethylformamide solution 75% (v:v) and agitated for 60 min, then centrifuged (HERMLE Labor Technik, Z300K, Wehingen, Germany) for 10 min at 1000× g. An amount of 1 mL of the supernatant was introduced into a test tube (A); 6 mL of distilled water and 1 mL of the ammonia solution (8 g/L) were added, and agitated in vortex. In another test tube, 1 mL of the supernatant was added (B), and 5 mL of distilled water and 1 mL of ferric ammonium citrate solution (3.5 g/L) were added and agitated in a vortex, then 1 mL of ammonia solution was added. The solutions A and B were transferred into measuring cells; the absorbance was measured 10 min after the end of the process at 525 nm in a VELAB spectrophotometer (VE-5100UV, Toluca, Mexico), and the result was the differences between absorbances. The calibration curve was prepared with tannic acid (773 Merck, New York, NY, USA), and the results are expressed as mg of tannic acid/g of sample.

2.4.2. Saponins

The saponins were determined following the technique proposed by Hiai & Nakajima [21]. An initial extraction was performed by weighing 0.5 g of each sample and adding 10 mL of 80% methanol:water (v:v) and leaving the mixture at a low stirring speed for 20 h. Subsequently, it was centrifuged at 1900× g for 10 min in a HERMLE (Labor Technik, Z300K, Wehingen, Germany) centrifuge, and the supernatant was collected. The precipitate was washed with 5 mL of 80% methanol, shaken in a vortex for 5 min, and centrifuged under the same conditions; this procedure was performed two more times, and the supernatants were pooled. In a test tube, 250 µL of the extract and 250 µL of vanillin reagent (800 mg/10 mL analytical grade ethanol) were added, placed in an ice bath, and 2.5 mL of concentrated sulfuric acid was added, and then the mixture was vortexed. The mixture was heated in a water bath at 60 °C for 10 min and cooled in an ice bath. The absorbance of the samples was measured at 520 nm in a VELAB spectrophotometer (VE-5100UV, Mexico). A calibration curve with diosgenin (0.5 mg/mL) as standard was used. The results are expressed as mg of diosgenin/g of sample.

2.4.3. Phytates

The obtention of phytates was carried out in Falcon conical centrifuge tubes. In total, 1 g of each sample was placed in each tube, and 50 mL of trichloroacetic acid (TCA) at 3% was added and shaken for 30 min; at the end of the time, the samples were centrifuged at 1000× g for 6 min in a HERMLE centrifuge (Labor Technik, Z300K, Wehingen, Germany). An amont of 10 mL of the supernatant of each sample was taken and placed in new conical centrifuge tubes with 4 mL of a solution of iron chloride (2.9 mg/mL), and the samples were heated in a water bath at 100 °C for 45 min. After this time, 4 to 5 drops of 3% sodium sulfate (3 g of sodium sulfate dissolved in 3% TCA) were added, and the heating continued for another 15 min. The samples were centrifuged at 1500× g for 6 min, and the supernatant was removed. 20 mL of 3% TCA was added to the precipitate and vortexed, heated to boiling for 10 min, and centrifuged at the same conditions; this process was performed two more times. The precipitate was analyzed to determine the content of phytates according to the technique proposed by Thompson & Erdman [22] without modifications. The samples were measured at 480 nm in a VELAB spectrophotometer (VE-5100UV, Mexico). A blank of the reagents only was measured, and the calibration curve was prepared from a standard 5 mg/mL ferric nitrate solution considering the 4:6 iron–phosphate ratio to express the results as mg of phytic acid/g sample.

2.4.4. Glucosinolates

The extraction of glucosinolates was carried out as follows: 500 mg of sample was added with methanol at 70% in a 1:15 w:v ratio, left in a water bath with agitation at 70 °C for 20 min, then centrifuged at 9700× g for 5 min in a HERMLE centrifuge (Labor Technik, Z300K, Wehingen, Germany), saving the supernatant. The process was repeated with the precipitate only, and the supernatants were then mixed. The samples were concentrated in a water bath at 80 °C with constant stirring. To determine the glucosinolate content, the technique proposed by Jezek et al. [23] was performed without modifications. A 5.60 mM sinigrin standard solution was used to construct the calibration curve (0–1 mM range) and measured at 420 nm in a VELAB spectrophotometer (VE-5100UV, Mexico). The results are expressed as µmoles of sinigrin/g of sample.

2.4.5. Trypsin Inhibitors

The trypsin inhibitors were extracted by weighing 1 g a sample in a 50 mL conical tube, adding 30 mL of 0.05 M HCl, and allowing the sample to rest for 24 h at 10 °C. This was centrifuged at 1000× g for 10 min in a HERMLE centrifuge (Labor Technik, Z300K, Wehingen, Germany). Then, the content of trypsin inhibitors in the sample was obtained following the technique proposed by Sobral & Wagner [24] without modifications. Trypsin inhibitors are expressed as TUI/mg of sample.

2.4.6. Cyanogenic Glucosides

For this determination, the 915.30 method proposed by the AOAC [19] was followed. Twenty grams of sample were weighed, and 20 mL of distilled water was added, leaving it to rest for 2 h. Subsequently, an additional 400 mL of distilled water and three drops of mineral oil were added. The dispersion was placed in a round-bottom flask, which was then placed in a steam distillation apparatus, where the condenser outlet collected 150 mL of the distillate in a flask containing 0.5 g of NaOH in 20 mL of water to neutralize the hydrocyanic acid. Subsequently, it was filled up to 250 mL with distilled water, and a 100 mL aliquot was taken, adding 8 mL of NH₄OH and 2 mL of KI (both at 5% w/v). The solution was titrated with 0.02 N AgNO₃. The results are expressed as mg of HCN/100 g of sample.

2.5. Evaluation of Nutritional Parameters

2.5.1. In Vitro Digestibility

For in vitro digestibility, an adaptation of the method proposed by Tavano et al. [25] was used. A 10 mL aqueous suspension was prepared with FG in a 1:15 (w:w) pepsin-substrate ratio, and the pH was adjusted to 4.5, followed by incubation for 30 min at 37 °C. The pH was then adjusted to 7, 2.5 mL of a pancreatin solution (0.4 g/100 mL in 1 M NaHCO₃) was added, and the mixture was incubated for 1 h at 37 °C. Throughout the hydrolysis period, the sample was kept in agitation at 350 rpm with a stir-pak (Cole-Palmer 50002-30, Vernon Hills, IL, USA). After this time, the samples were placed in a water bath at 100 °C to inactivate the enzymes and centrifuged at 12,000× g for 45 min in a HERMLE centrifuge (Labor Technik, Z300K, Wehingen, Germany). The soluble fraction was added to 24% trichloroacetic acid in a 1:1 (v:v) ratio, centrifuged at 6000 rpm, and the amount of nitrogen in the precipitate was determined using AOAC method 945.01 [19]. Digestibility % was calculated according to Equation (1).

$$\% \mathrm{Digestibility}={\displaystyle \frac{(Total protein)+(Nondigested protein)}{Total protein weight}}\times 100$$

(1)

2.5.2. Amino Acid Composition

The profile content of Asp + Asn, Glu + Gln, Ser, His, Gly, Thr, Arg, Ala, Pro, Tyr, Val, Met, Cys, Ile, Leu, Phe, and Lys was quantified as follows: 5–10 mg of protein content in the sample was weighed and placed in a digestion tube. Acid digestion (6 N HCl) and drying were performed on the samples, and a derivatization process with diethylethoxymethylenemalonate for 45 min at 50 °C was performed. The resolution was carried out using a binary gradient system of 25 mM sodium acetate solution with 0.02% sodium azide (pH 6) and acetonitrile with a constant flow of 0.9 mL/min at 18 °C, according to the methodology proposed by Alaiz et al. [26] without modifications. The equipment used was an Agilent HP 1100 series with automatic injection and a VWD detector, and a C18 reverse-phase chromatographic column (Nova Pack brand, particle size 4 μm, 300 × 3.9 mm) was used. For Trp determination, according to the proposal by Yust et al. [27], a basic hydrolysis with 4 N NaOH was required, after which an isocratic elution system consisting of 25 mM sodium acetate, 0.02% sodium azide pH 6, and acetonitrile (91:9), with a delivering flow of 0.9 mL/min, was performed. For this determination, the same equipment and column were used.

2.5.3. Essential Amino Acid Proportion (EAP)

To calculate the proportion of EA in relation to total amino acids, the amino acid profile obtained in FG was used, according to Equation (2) proposed by Chavan et al. [28].

$$\mathrm{EAP}=\left({\displaystyle \frac{\mathrm{I}\mathrm{l}\mathrm{e}+\mathrm{L}\mathrm{e}\mathrm{u}+\mathrm{L}\mathrm{y}\mathrm{s}+\mathrm{C}\mathrm{y}\mathrm{s}+\mathrm{P}\mathrm{h}\mathrm{e}+\mathrm{T}\mathrm{y}\mathrm{r}+\mathrm{T}\mathrm{h}\mathrm{r}+\mathrm{T}\mathrm{r}\mathrm{p}+\mathrm{V}\mathrm{a}\mathrm{l}+\mathrm{H}\mathrm{i}\mathrm{s}}{Ala+Asn+Arg+Gly+Ile+Leu+Lys+Met+Cys+Phe+Tyr+Thr+Trp+Val+His+Pro+Ser }}\right)$$

(2)

2.5.4. Amino Acid Score (AS)

Equation (3) was used to determine the chemical score, following the pattern for a standard protein proposed by the FAO [29] and reported by Chavan et al. [28], where Ile, Leu, Lys, Met + Cys, Phe + Tyr, Thr, Trp, and Val had values of 4.00, 7.04, 5.44, 3.52, 6.08, 4.00, 0.96, and 4.96, respectively.

$$\mathrm{AS}={\displaystyle \frac{mg of amino acid per g test protein }{mg of amino acid per g of FAO standard pattern}}$$

(3)

2.5.5. Calculated Protein Efficiency Ratio (C-PER)

The C-PER was calculated using Equations (4) and (5), proposed by Lee et al. [30], with a reference standard of 2.5 corresponding to casein.

C-PER = 0.08084_[X7] − 0.1094

(4)

C-PER = 0.06320_[X10] − 0.1539

(5)

where:

X7 = Σ of the values of the amino acids Tre, Val, Met, Ile, Leu, Phe, and Lys in g/100g of protein.

X10 = Σ of the values of the amino acids His, Arg, Trp, Tre, Val, Met, Ile, Leu, Phe, and Lys in g/100g of protein.

2.5.6. Biological Value

The estimated biological value was calculated using Equation (6), which was proposed by Morup & Olesen and reported by Chavan et al. [28] using the values for an ideal protein according to the FAO [29].

$$\mathrm{BV}={10}^{2.15}\times q_{Lis}^{0.41}\times q_{Fen+Tyr}^{0.60}\times q_{Met+Cys}^{0.77}\times q_{Tre}^{2.4}\times q_{Trp}^{0.21}$$

(6)

where:

q = ${\displaystyle \frac{a_i of the sample }{a_i of the reference pattern}}$ when the a_i of the sample is ≤ than the a_i of the reference pattern.

q = ${\displaystyle \frac{a_i of the reference pattern}{a_i of the sample }}$ when the a_i of the sample ≥ than the a_i of the reference pattern.

a_i = mg of amino acid per 1 g of protein.

2.6. In Vivo Evaluation

2.6.1. Acute Toxicity

Approval for the acute toxicity evaluation was obtained on 4 October 2021 from the Committee for the Care and Use of Laboratory Animals of the Faculty of Medicine of the Universidad Autónoma del Estado de Morelos (CCUAL-FM-UAEM), with identification code 007/2021, in accordance with Mexican Official Standard NOM-062-ZOO-1999 [31]. Female mice of the CD1/ICR strain with a body weight of 25–30 g at the time of reception were used for this evaluation. During the first week, the animals were in the adaptation process and were then randomly divided into groups of three animals (n = 3) per residence (group).

The in vivo model was performed according to Lorke’s proposal [32], which consists of two stages. In the first phase, FG doses of 10, 100, and 1000 mg/kg of animal weight were evaluated, along with a control group that was only given distilled water via oropharyngeal gavage. For administration, the FG fraction was suspended in distilled water immediately before dosing to ensure homogeneity and prevent sedimentation. The suspension was prepared to allow for a standardized administration volume of 10 mL/kg body weight. The suspension was continuously vortexed prior to each administration to maintain uniform dispersion of the material.

Based on the number of animals that died at each dose level during this initial screening phase, a decision was made whether to escalate to higher dose levels. In accordance with the method, if no mortality or only limited mortality occurred, higher doses were selected for the second phase; conversely, if significant mortality was observed, intermediate doses would be considered. As no mortality was observed in the first phase, the study proceeded to evaluate the higher doses (1600, 2900, and 5000 mg/kg body weight). The animals were monitored for a period of 15 days and then euthanized. The 3R-Refinement principles, which advocate for the Replacement, Reduction, and Refinement of animal tests, were considered. Replacement and Reduction focus on exploring alternative methods and minimizing the number of animals used, and Refinement pertains to techniques that alleviate pain, distress, and suffering experienced by animals during experimentation. A total of 18 mice was used, three for each concentration evaluated, and randomly distributed into six groups of three animals per treatment.

2.6.2. Histological Analyses

Liver and kidney samples from each dead mouse used in the study (3 per group) were placed in 10% neutral phosphate-buffered formalin. The fixation was done in 4% paraformaldehyde and then sectioned; 5 µm thick sections were stained with hematoxylin and eosin [33]. The histological characteristics of tissues were evaluated using a Leica Qwin image-analyzer system on a Leica DMLS microscope (Leica DM750 Wetzlar, Germany).

2.7. Statistical Analysis

The results were processed using descriptive statistics and expressed as the mean ± SD. The data were analyzed using statistical software Statgraphics Centurion XV (15.2.06) and subjected to an independent sample Student’s t-test to determine if there was a difference between groups according to the methods reported by Montgomery [34]. Significant differences were considered to be those with a value of p < 0.05.

3. Results

3.1. Germ Extraction Yield

The germ extraction yield of Delonix regia seeds across independent replicates ranged from 20.48% to 21.32% db, with a mean of 21.10 ± 0.54. The yields for mucilage and hull were of 40.82% ± 1.13 and 34.78% ± 1.27, respectively. The calculated 95% confidence interval for flamboyant germen extraction was 20.45–21.35%, indicating low experimental dispersion and good reproducibility under laboratory-scale conditions.

3.2. Proximate Composition

The proximal composition indicated statistically significant differences (p < 0.05) between FsF and FG in all chemical parameters evaluated (

Table 1. Proximal composition in flours of flamboyant seed (FsF) and germ (FG) (% d.b.).

Component	FsF	FG
Moisture	9.50 a ± 0.21	3.61 b ± 0.17
Crude protein	20.09 a ± 0.89	78.35 b ± 0.45
Crude fiber	16.83 a ± 0.82	10.58 b ± 0.86
Fat	3.32 a ± 0.11	0.66 b ± 0.01
Ash	4.06 a ± 0.03	6.24 b ± 0.10
Carbohydrates as NFE	55.68 a ± 0.03	4.17 b ± 0.40

All analysis was performed by triplicate. The data represents the average value ± standard deviation. ^a,b Letters in the same row indicate significant statistical differences (p < 0.05).

). The process for obtaining FG led to an increase in crude protein content from 20.09% to 78.35%. In contrast, carbohydrates estimated as NFE showed a decrease from 55.65% to 4.17%. This relationship between increased protein and decreased carbohydrates represents an advantage for the use and exploitation of flamboyant seeds as a non-conventional source of protein, allowing for the exploration of the protein present in FG, as it also contained lower amounts of crude fiber, fat, and ash.

3.3. Antinutritional Factors

The antinutrient content present in FsF and FG is shown in

Table 2. Antinutrient content in flamboyant seed (FsF) and germ (FG).

Antinutrient	FsF	FG
Tannins (mg/g of sample)	3.10 b ± 3.00	7.05 a ± 4.36
Saponins (mg/g of sample)	17.543 b ± 0.21	28.130 a ± 0.17
Phytates (mg/g of sample)	9.4 a ± 1.10	3.26 b ± 17.81
Glucosinolates ¥	0.17 a ± 0.82	0.14 b ± 0.86
Trypsin inhibitors *	1.20 b ± 0.11	3.91 a ± 0.01
Cyanogenic glucosides (mg/g of sample)	0.00 a ± 0.00	0.00 a ± 0.00

The data represents the average value ± standard deviation. ^¥ Glucosinolates were expressed as µmol of sinigrin/g of sample. * Trypsin inhibitors were expressed as TUI/mg of sample. ^a,b Letters in the same row indicate significant statistical differences (p < 0.05).

. Statistically significant differences (p < 0.05) were observed after FG extraction in all components, except for cyanogenic glycosides, which were not found in either FsF or FG. It was observed that when the germ was removed from the seed, the glucosinolate content decreased by around 18%, and in the case of phytates, the reduction was approximately 65%. In the case of tannins, saponins, and trypsin inhibitors, their concentration increased by approximately 200%, 160%, and 300%, respectively. These results suggest that these types of antinutritional factors are mainly found in the germ of the seed, explaining why their values increased when only this isolated fraction was analyzed.

3.4. Evaluation of Nutritional Parameters

Table 3. Amino acid composition and nutritional parameters of flamboyant germ (FG).

	Content (g/100 g of Protein)	FAO Recommendation ¥ [35]	Amino Acid Score
Essential amino acids
His	2.23 ± 0.26	1.6	>1
Ile	3.25 ± 0.13	3.0	>1
Leu	7.74 ± 0.01	6.1	>1
Lys	4.97 ± 0.05	4.8	>1
Thr	2.78 ± 0.13	2.5	<1
Trp	1.06 ± 0.01	0.66	>1
Val	6.74 ± 0.14	4.0	>1
Met	0.51 ± 0.09	*	*
Cys	0.84 ± 0.08	*	*
Tyr	3.22 ± 0.24	**	**
Phe	4.75 ± 0.21	**	**
* SAA	1.35	2.3	<1
* AAA	7.97	4.1	>1
Non-essential amino acids
Asp + Asn	6.84 ± 0.30
Glu + Gln	19.67 ± 0.32
Ser	4.35 ± 0.10
Gly	4.56 ± 0.42
Arg	8.62 ± 0.07
Ala	6.29 ± 0.09
Pro	11.55 ± 0.76
Nutritional Parameters
In vitro digestibility		43.63%
EAP		38.10%
C-PER[X7]		2.36
C-PER[X10]		2.54
BV		24.63 ± 0.86

¥ Recommended daily intake for the older children, adolescents, and adults. * SAA = Met + Cys, ** AAA = Tyr + Phe.

shows the amino acid profile of FG, as well as the in vitro digestibility value, EAP, AS, C-PER, and BV. The value of protein digested in vitro with pepsin and pancreatin indicated that flamboyant germ was digested by approximately 50%. When comparing the essential amino acid content with the FAO’s recommended daily requirements for children, adolescents, and adults, only the sulfur amino acids (Met + Cys) did not meet these requirements. When evaluating the amino acid score using the FAO standard for a standard protein, threonine and methionine + cysteine had a value lower than 1 and were thus considered to be the limiting amino acids in flamboyant germ. C-PER is a tool that allows us to assess the quality of a food protein when it is incorporated into the tissue of the consumer, and it reflects in the weight gain through muscle generation. FG has C-PER values that are considered desirable, as they exceed a value of 2. This is also linked to BV, as it measures the efficiency of protein absorption into body tissues and depends particularly on essential amino acids being present in adequate amounts, where essential amino acid levels were possibly responsible for the value obtained in this study (24.63). However, it should be noted that this data was obtained using mathematical model estimates, so it is recommended that studies be conducted on in vivo models to confirm the value obtained.

3.5. In Vivo Evaluation of Acute Toxicity

In stage one of the acute toxicity assessment, no animal deaths were observed for the administered doses of FG of 0, 10, 100, and 1000 mg/kg of body weight. In stage two of the evaluation, concentrations of 1600, 2900, and 5000 mg/kg of body weight were analyzed, again with no deaths. It should be noted that throughout the observation period, there were no changes in weight, sustained piloerection, opacity in the eyes, changes in gait, or physiological changes such as the appearance of protuberances.

3.6. Histological Analysis

In the case of histopathologies, normal glomeruli were identified in the renal parenchyma sections, with no evidence of mesangial proliferation or sclerosis (

Table 4. Histopathologies of the kidney and liver from the acute in vivo toxicity study of flamboyant germ (FG).

First Phase Dose (mg/kg of Body Weight)
	Control	10	100	1000
Kidney
Liver
Second phase dose (mg/kg of body weight)
	1600	2900	5000
Kidney
Liver

). The tubules were observed to be lined by cubic epithelium with nuclei containing open chromatin and discrete nucleoli, with no evidence of necrosis. There was no acute or chronic inflammatory infiltration. In the case of hepatic parenchymal sections, portal triads consisting of artery, vein, and bile duct without acute or chronic inflammatory infiltrate were identified. The hepatocytes that make up the hepatic lobules were arranged in trabeculae around a central vein; no evidence of liver damage such as ballooning or apoptosis was identified, nor was there any acute or chronic inflammatory infiltrate.

4. Discussion

4.1. Germ Extraction Yield

Compared with other plant-derived seeds processed by manual decortication and fractionation, the germ recovery of Delonix regia (~21% on a dry basis) was higher. For example, in cereal grains like maize, which are more intensively studied, germ recovery from wet milling ranges from ~3% to ~6% of whole-kernel mass, highlighting the lower fractionation yields often obtained in such systems for cereals [36]. However, compared with other underutilized arboreal seeds processed, the yield was lower on impact milling, and in the screening process for Saponaria vaccaria, the germ fraction was reported at ~35% of whole-seed weight, reflecting typical outcomes in laboratory sample fractionation processes for non-major crops [37]. Although direct comparisons across species and processing methods are constrained by anatomical and technological differences, the relatively high germ fraction obtained from flamboyant seeds suggests that its seed structure is favorable for decortication, cotyledon isolation, and germen extraction, positioning it in the upper intermediate range of recoverable embryo/cotyledon yields when compared to other understudied food seed and arboreal seed systems.

4.2. Proximate Composition

The protein content in FsF quantified in this study (20.09%) was higher than that reported for commonly consumed legumes such as Phaseolus vulgaris varieties like pinto bean (15.41%), navy bean (14.98%), black bean (15.24%), kidney bean (15.35%), Vigna unguiculata (13.22%), Cicer arietinum (14.35%), and Lens culinaris (17.86%), but lower than the values for Lupinus albus (25.85%) and Glycine max (28.62%) [30]. The carbohydrates in FsF (55.68%) consist mainly of mucilage, which is recognized for its beneficial techno-functional properties [12,18], increasing the value and comprehensive use of this seed otherwise considered to be organic waste.

After extracting the germ from the endosperm and separating the husk from the hydrated seed, the protein content was found to be over 70%, which classifies FG as a protein concentrate [38]. This value was obtained without any additional treatment, which raises the possibility of increasing the protein value to an isolate with 90% protein if technological processes such as isoelectric point fractionation or micellization are applied. Currently, the use of protein-rich concentrates and isolates from animal-based foods is showing a clear downward trend due to dietary preferences, allergies, vegetarianism, and other factors. Therefore, exploring alternative plant-based protein sources such as legumes allows for a wider range of choices for consumers [39].

4.3. Antinutritional Factors

Cyanogenic glycosides were not present in FsF or FG, which is consistent with the findings of Kumar et al. [40], where cyanides are originally absent in legumes. Furthermore, cooking is a process that destroys cyanogenic glycosides; although these compounds were not found in the raw seed, the heat treatment used for germ extraction (70 °C) would guarantee the absence of such compounds. The glucosinolate content was lower than that reported for broccoli (12.51 µmol/g) [41], while the concentration of phytates present in FG (3.26 mg/g of sample) was within the range found in legumes such as peas (3.1–7.1 mg/g), chickpeas (2.8–13.6 mg/g), and lentils (2.55–12.2 mg/g) [42]. For glucosinolates and phytates, the hydration process at 70 °C for 4 h resulted in a decrease in these compounds, as reported by Das et al. [11], where this type of processing allows for the reduction or elimination of antinutritional factors in edible legumes.

The tannin content in FG was within the range indicated for different lentil varieties (3.73–10.20 mg/g) [43]. For saponins, an intake range based on the consumption of different foods of 2500 mg/kg of body weight has been proposed [44]; the concentration of these compounds present in FG would allow consumption of between 200 and 400 g per day without exceeding the permitted limit. The levels of trypsin inhibitors in FG were within the range reported for different soybean varieties (2–20 ITU/mg) [45]. All antinutrients quantified in FsF and FG were within the amounts reported for commonly consumed legumes, so their consumption would not have any adverse health effects.

From a safety standpoint, reference or tolerable intake levels for specific antinutritional compounds have been proposed. For phytates, although no official maximum limit has been established, dietary intakes between 0.3 and 2.6 g/day are considered typical in plant-based diets without adverse effects in healthy populations [46]. This means the concentration detected in FG (3.26 mg/g) was within safe intake ranges. Regarding saponins, acceptable daily intake estimates suggest levels up to 10 mg/kg body weight/day without toxicological effects [44], which supports the safety margin of FG consumption obtained in this study. Trypsin inhibitor levels in FG were comparable to those reported for thermally processed soybean products, where heat treatment effectively reduces protease inhibitory activity below physiologically relevant thresholds [45]. For tannins, moderate dietary exposure, generally <1 g/day, is not associated with adverse health outcomes and may even exert antioxidant benefits [47]. Therefore, the antinutrient profile of FG did not exceed recognized safety reference ranges and was comparable to widely consumed legumes.

Flamboyant germ (FG) is emerging as an unconventional plant-based protein source with high potential for strengthening food security, particularly in contexts where access to animal-based proteins is limited or environmentally costly. The use of underutilized seeds made it possible to transform agroforestry waste into an ingredient with added value. The safety associated with its antinutritional factor profile and the possibility of its integration into food formulations support its viability as a sustainable alternative.

4.4. Evaluation of Nutritional Parameters

All protein amino acids were found in FG (Table 3), with a profile similar to that reported for Oyedeji et al. [48], where the difference was the high content of Met and Cys (1.08 and 1.82 g/100 g, respectively) compared to the values in this study (0.51 and 0.84 g/100 g of protein). These two amino acids were the only ones that did not meet the daily requirements for children, adolescents, and adults stipulated by the FAO [35]. The presence of taste-active amino acids such as Glu + Gln, Gly, and Ala suggests that flamboyant protein would be a good alternative in the production of sensorially acceptable foods [48].

The digestibility of FG was lower compared to that reported by Ohanenye et al. [49] for soybean protein, with around 56%. These same authors mention that there are factors that could influence the value obtained, such as amino acid composition, the three-dimensional structure of the protein, whether any components of the seed, like the husk, are removed, and the technique used for the determination, among others. The high protein concentration (>70%) obtained in FG positions it within the range of commercial plant protein concentrates currently used in food systems. Plant protein concentrates and isolates derived from soybean, pea, and fava bean are widely incorporated into meat analogs, protein-enriched beverages, bakery products, and extrusion-based snacks [50]. Given its amino acid profile and functional characteristics, FG protein could potentially be explored for similar applications, particularly in regions where flamboyant biomass is available. Additionally, the processing required to obtain the protein-rich fraction suggests potential scalability using conventional wet fractionation technologies.

Agyekum et al. [51] indicated that the application of treatments such as soaking, heating, and nixtamalization could improve digestibility. The EAP value for FG (38.10%) was similar to those reported for edible legumes such as peas, chickpeas, lentils, broad beans, and common beans, with values of 38.59, 40.55, 39.59, 40.42, and 33.64%, respectively [52]. The C-PER₇ and C-PER₁₀ values were higher than 2, indicating that FG protein is of high quality. In the case of C-PER₁₀, it had a similar value to casein (2.5), which is the protein used as a reference standard [53]. According to Osunsanmi et al. [54], the BV indicates the amount of protein that will be utilized by the body, with the FG utilizing approximately one-quarter of the protein consumed. This low level should be taken cautiously, as it was calculated mathematically and may not take into account all the biochemical variables involved in protein utilization. Osunsanmi et al. [54] improved BV values by applying processes such as fermentation, defatting, and the obtaining of a protein isolate at the isoelectric point of the raw material (Parkia biglobosa), achieving values of ~70. These treatments could be an option for improving the yield of FG. However, techno-functional characterization, such as solubility, emulsifying capacity, foaming properties, and gelation behavior under food-processing conditions, would be necessary to validate its industrial suitability.

4.5. Evaluation of Acute Toxicity

In vivo analysis conducted in a murine model showed that flamboyant germ (FG) does not pose a toxicity risk for human consumption, as it had a lethal dose (LD₅₀) greater than 5000 mg/kg of body weight. This was supported by the absence of histopathological alterations in key organs such as the liver and kidneys, where no signs of acute or chronic inflammation were found. These findings confirm the safety of FG as a food ingredient and reinforce its potential for use in formulations intended for human consumption. In addition, the traditional use of different parts of the flamboyant in folk medicine [55]; the cosmetics, pharmaceutical, and food industries [12]; and its use in traditional preparations such as iregi in Nigeria [56] support its historical acceptance and functionality in different cultural and productive contexts.

The use of FG as a protein source would offer significant advantages from a food security perspective, given the widespread distribution of flamboyant trees in tropical, subtropical, and temperate regions, as well as their underutilization as a plant resource. Currently, this species is mainly used for ornamental purposes, generating large volumes of biowaste associated with its hard pods, particularly during dry seasons [57]. The comprehensive use of the seed would increase its added value, considering not only the germ but also components such as mucilage, with potential applications as a carrier of bioactive compounds and as a stabilizer in the food industry [18]. The pods can be used to produce biodiesel [12] and polymers [57], promoting the comprehensive use of the biomaterial and preventing its accumulation as waste in areas where flamboyant trees grow.

In the current context, plant-based proteins represent a strategic alternative to animal proteins, whose production systems are associated with high environmental costs, intensive water consumption, loss of ecosystems, and their contribution to climate change [58]. In contrast, plant proteins offer more accessible, sustainable options that are compatible with health-oriented dietary trends, which has driven the development and consumption of protein supplements derived from legumes (soybean, chickpea, lupin, and faba bean), cereals (rice, wheat, millet, maize, and sorghum), pseudocereals (amaranth, quinoa, and buckwheat), and seeds (chia, sesame, pumpkin, and sunflower) [59].

Considering the global population’s projected growth by 2050, reaching nearly 10 billion people, diversifying protein sources is essential to ensure adequate, sustainable, and resilient diets [60]. In this scenario, the use of flamboyant seeds to obtain unconventional proteins takes on special relevance, transforming a material currently considered waste into a valuable food resource. This strategy will not only contribute to strengthening food security but also align with circular economy principles and sustainability by reducing waste, optimizing available plant resources, and contributing positively to mitigating global ecological deterioration and climate change. The transition from ornamental biomass residue to commercial protein ingredient would require structured collection systems, quality standardization, and regulatory assessment for food-grade approval, which remain to be developed. Nevertheless, similar valorization pathways have successfully been implemented for other underutilized legume residues, demonstrating technical feasibility when supported by coordinated supply chains [50].

Despite the promising results obtained, the study presents some limitations. The acute toxicity assessment was limited to short-term exposure in a murine model, and sub-chronic or chronic toxicity evaluations were not performed. Protein digestibility and biological value were estimated using in vitro and predictive models rather than in vivo nitrogen balance studies. Techno-functional properties relevant to industrial food applications should be comprehensively evaluated. Finally, although the availability of flamboyant biomass suggests valorization potential, logistical and economic feasibility analyses at a municipal scale were beyond the scope of the present work. Future research should therefore include long-term safety assessments, protein functionality studies under processing conditions, and life cycle or techno-economic analyses to fully validate FG as a scalable plant-based protein ingredient.

5. Conclusions

The flamboyant tree seed (Delonix regia) showed high potential for incorporation into human food systems, allowing for the valorization of seeds currently considered plant waste. This approach contributes directly to food security (SDG 2) by increasing the availability of plant-based proteins and reducing dependence on conventional protein sources with a greater environmental impact. Flamboyant germ (FG)’s favorable amino acid profile, acceptable antinutritional levels, and absence of acute toxicity (LD₅₀ > 5000 mg/kg body weight) confirm its safety and nutritional feasibility for food applications. Although digestibility and biological value were moderate, these limitations are typical of plant matrices and can be improved through technological processing.

This study provides a scientific foundation for integrating underutilized urban biomass into circular bioeconomy strategies. The valorization of flamboyant germ demonstrated how ornamental tree residues can be repositioned as unconventional protein sources, contributing to food security, waste reduction, and climate-resilient food systems. Future research should focus on techno-functional characterization of FG protein under industrial processing conditions, optimization of protein isolation and digestibility-enhancing treatments, evaluating techno-economic assessments of seed collection and processing logistics, conducting life-cycle assessments, and regulatory evaluation for food-grade commercialization. Addressing these aspects will be critical to transitioning from laboratory-scale validation to scalable implementation within sustainable resource management frameworks. Promoting plant-based proteins would help mitigate environmental impacts, aligning with actions to combat climate change (SDG 13). The findings position the flamboyant tree germ as an unconventional, harmless, and sustainable plant protein resource with the potential to be integrated into food innovation policies and the development of resilient food systems.