Caper bush (Capparis spinosa L.) bioactive compounds and antioxidant capacity as affected by adaptation to harsh soils

Abstract

Caper bush (Capparis spinosa) is a naturally grown species in different soils. To gain insight into the impact of various soil conditions on nutritional and phytochemical properties, aerial parts of caper bush (C. spinosa) were collected from gypsum and non-gypsum soils in southern Iran. Colorimetric analyses of antioxidant compounds (total phenolics and flavonoids) and antioxidant capacity tests (DPPH, FRAP, ABTS) were carried out, and intact aliphatic and indolic glucosinolates (predominant aliphatic glucocapparin) were analyzed by HPLC-DAD method. Based on the findings, plant parts and sites significantly impacted most parameters. The highest TPC values were observed in the petals in gypsum soil and the lowest in the non-gypsum soil seeds by 2317.78 and 635.06 mg/kg FW, respectively. Likewise, the highest TFC was recorded in the non-gypsum soil leaves and the lowest in the non-gypsum soil seeds by 401.06 and 55.61 Qu mg/kg FW, respectively. The highest and lowest FRAP values were observed in the leaves in gypsum (0.94) and the pistils in gypsum soil (0.80), respectively. Regarding ABTS values, the flags in the non-gypsum and gypsum sites showed the highest and lowest values of 89.51 and 78.40%, respectively. High DPPH values were recorded for most parts. The highest amount of glucocapparin was found in the pistils in gypsum, and the lowest was in the petals in gypsum soil by 35.81 and 21.65 µmol/g DW, respectively. The gypsum sites showed higher values for most of the studied parameters. The PCA results showed that pistils were associated with glucocapparin, petals with DPPH, and the leaves and sepals with TPC, FRAP, ABTS, and TFC. The majority of studied factors correlated well with TPC. Our results supported the potential of caper bush (C. spinosa) as a underexploited food rich in bioactivephytochemicals adapted to harsh soil conditions, with the potential for implementation in agroecosystems with adverse environmental conditions with the potential of better adaptation for securing the access to plant-derived foods.

Introduction

Environmental conditions strongly influence plant secondary metabolites¹. Drylands have harsh environments with large aridity and highly available elements such as sodium, calcium, and sulfur on the local scale, which plants have adapted by developing a high diversity of chemical traits, highlighted in climate change senarios². Gypsum soils are characterized by high levels of gypsum (CaSO₄·2H₂O), calcium (Ca), sulfur (S), and magnesium (Mg) which strongly affect plants’ chemical characteristics³. Plants adapted to gypsum soils fall into two categories: gypsovags, which can grow in both gypsum and non-gypsum soils, and gypsophiles, which primarily thrive in gypsum soils. Generally, gypsovags have lower concentrations of elements commonly found in gypsum soils, such as calcium and sulfur, compared to gypsophiles^4,5. Gypsum soil properties significantly affect antioxidant activity and phenolic compounds⁴. The response of phenolic compounds and antioxidant activity is species-specific⁶. The species that grow in gypsum soil may have higher phenolic and flavonoid content compared to salt marsh and dunes or lower than in serpentine soil^7,8. Additionally, gypsophiles have higher phenolic compounds and antioxidant activity than gypsovags⁹. However, in contrast, comparing gypsovags in two gypsum and non-gypsum soils showed higher phenolic compounds and antioxidant activity in non-gypsum soil⁹. Gypsum amendment can also positively affect antioxidant activity and secondary metabolites^10,11.

Furthermore, glucosinolates as sulfur containing compounds positively are influenced by sulfur supply which lead to increase in the levels of total glucosinolates⁴. Adding sulfur to the soil may enhance its availability, increasing total glucosinolates in Brassicaceae species like Brassica rapa, as well as individual glucosinolates such as glucoraphanin and glucoraphasatin in Brassica juncea L., and glucoraphanin of Brassica oleracea var. Italica. However, the effects of sulfur in gypsum-rich soils in drylands remain underexplored^12–14. However, there may be no significant differences in total glucosinolates, aliphatic, or indolic glucosinolates in different cultivars with very high sulfur fertilization¹⁵. Other parameters that can influence the role of sulfur in glucosinolates include nitrogen fertilization¹⁶, selenium fertilization¹⁷, and water stress^18,19. Glucosinolates are phytochemicals and secondary sulfur compounds synthesized from glucose and amino acids^20–22. They play essential roles in plant defense against diseases and pests and have shown anticancer activity^23,24. Glucosinolates (GLSs) are present in all parts of almost all species of plants in the order Brassicales. However, their content is higher in reproductive tissues (i.e., flowers and seeds) than in vegetative tissues^25–27. Different studies have reported high glucosinolate content in the fruits²⁸, seeds, leaves²⁹, and flowers³⁰ of Brassicaceae. Glucosinolates are natural substances in many plants, particularly in the Brassicaceae, Caricaceae, and Capparaceae families^31,32. Capparaceae is closely related to the Brassicaceae family and is rich in glucosinolates and flavonoids²³. However, there are limited studies on this family compared to Brassicaceae. Capparis spinosa is a well-known species of the Capparaceae family in the Mediterranean region. This perennial shrub has adapted to different habitats, including gypsum ecosystems^33,34. C. spinosa thrives in arid environments with scarce water and soils with high water-soluble salts, making it suitable for dryland restorations^33,35. Additionally, this species holds ecological and economic significance. It has culinary applications, such as pickles, and can be used as a vegetable^36,37.

The C. spinosa is a plant with high nutritional and bioactive values, offering potential health benefits³⁸. According to Bakr and El Bishbishy³⁹, C. spinosa is a rich source of sulfur compounds and also contains phenolic and flavonoid glycosides, rutin, and quercetin⁴⁰. It also includes many other bioactive compounds, such as alkaloids, steroids, terpenoids, and tocopherols⁴¹. Several studies have reported high polyphenol content and antioxidant activity in different parts, including buds^42,43, fruits^44,45, and leaves⁴⁶. Caper buds, especially the smaller ones, exhibit a higher polyphenol content than unripe fruits^42,47. Most studies have focused on specific plant parts, mainly buds and fruits, while some have reported that leaves and buds have higher polyphenols and flavonoids than fruits and seeds and not flower parts, which were reported as flower or flower buds^48,49. Several studies also reported the presence of glucosinolate compounds in various parts of C. spinosa, including methyl, isopropyl, and sec-butyl isothiocyanates in ripe fruits⁵⁰, indole-3-acetonitrile glycosides in the fruits⁵¹, glucocapparin in shoots, glucocapparin in seeds^52,53, glucobrassicin, neo-glucobrassicin, and hydroxy-glucobrassicin in leaves^54,55, as well as high concentrations of glucocapparin for leaves, flowers, and buds, glucobrassicin in leaves, 4-hydroxy glucobrassicin in buds and flowers, and neoglucobrassicin in low concentrations in buds and leaves⁵⁶. Furthermore, few studies on wild capers showed higher glucosinolate content than cultivated ones^56,57.

The largest population of C. spinosa for economic purposes is in the southwest of Zagros in the Kazeron region of Iran, with a high content of gypsum in the soil in most localities and high support for the economy of local people. To our knowledge, there are no studies on the impact of environmental factors on bioactive compounds and glucosinolate of wild capers, including C. spinosa, especially flower parts, in natural ecosystems like Iran in the different soil conditions present in this particular area of gypsum soils. Additionally, the effect of sulfur was studied as a fertilizer and not in the gypsum soil as very sulfur-rich habitats in the drylands. Therefore, the current study aimed to investigate the effect of gypsum and non-gypsum soil on glucosinolates content and composition, and antioxidant capacity of different parts of C. spinosa including leaves, flowers, and fruits to get insight on the chemical status of Caper bush in the different soils from an agri-food perspective.

Results

The chemical properties were significantly affected by both the sites and plant parts, except for FRAP, where only the sites and their interactions with plant parts were significant. Additionally, the effects of sites on Glucocapparin were also not significant. Plant parts had the highest effects on the properties, except for ABTS, where sites had the highest effects (Table 1).

Table 1.

Results of GLMMs for the effects of the gypsum sites, plant parts, and their interactions, on studied characters.

	Sites(S)		Parts (P)		S × P
Treatments	F	P	F	P	F	P
TPC (GAE mg/kg FW)	34.01	< 0.001	20101.62	< 0.001	328.41	< 0.001
TFC (QU mg/kg FW)	95.16	< 0.001	740.07	< 0.001	310.14	< 0.001
ABTS (%)	253.40	< 0.001	117.07	< 0.001	205.25	< 0.001
DPPH (%)	21.03	< 0.001	91.27	< 0.001	9.58	< 0.001
FRAP	0.49	0.491	14.16	< 0.001	1.42	0.242
Glucocapparin (µmol/gDW)	0.003	0.946	37.12	< 0.001	4.85	0.002

Bold values indicate significant effects of treatments.

Total phenolic content and total flavonoid content

The total phenolic content (TPC) and total flavonoid content (TFC) varied significantly among different plant parts and locations (Table 1). The highest TPC values were observed in the flags (non-gypsum: 2307.90 ± 13.7 mg/kg FW; gypsum: 1721.48 ± 5.6 mg/kg FW), petals (non-gypsum: 2257.28 ± 20.7 mg/kg FW; gypsum: 2317.78 ± 21.7 mg/kg FW), and fruits (non-gypsum: 1126.42 ± 4.45 mg/kg FW; gypsum: 1921.48 ± 5.65 mg/kg FW). Conversely, the lowest values were found in the seeds (non-gypsum: 635.06 ± 1.4 mg/kg FW; gypsum: 684.44 ± 10.0 mg/kg FW) and pistils (non-gypsum: 1083.21 ± 10.1 mg/kg FW; gypsum: 1328.89 ± 10.5 mg/kg FW), regardless of whether the soil was gypsum or non-gypsum. TPC levels for petals, pistils, and fruits significantly increased in gypsum soil, with increases of 2.7%, 22.7%, and 70.58%, respectively. In contrast, the levels for sepals and flags decreased by 11.3% and 25.4%, respectively (Fig. 1).

Fig. 1.

Effects of the sites and plant parts on biochemical parameters (mean ± SE, n = 3). Uppercase letters are the results of the T- tests for the sites and lower-case letters are the results of the Tukey test for the plant parts.

The highest TFC was recorded in the leaves at both sites (non-gypsum: 401.06 ± 3.0 Qu mg/kg FW; gypsum: 372.27 ± 7.8 Qu mg/kg FW). At the gypsum site, the fruits showed the next highest TFC value (367.73 ± 12.03 Qu mg/kg FW), while the non-gypsum site exhibited much lower values (73.79 ± 4.0 Qu mg/kg FW). The sepals at both sites had similar TFC levels (non-gypsum: 270.76 ± 4.0 Qu mg/kg FW; gypsum: 287.4 ± 6.60 Qu mg/kg FW), and the petals at the non-gypsum site also showed significant TFC (272.27 ± 4.5 Qu mg/kg FW). The next highest TFC values were found in the pistils and flags, while seeds at both sites had the lowest TFC values (non-gypsum: 55.61 ± 5.4 Qu mg/kg FW; gypsum: 98.03 ± 5.4 Qu mg/kg FW). At the gypsum site, TFC decreased in the leaves, petals, and flags by 7.1%, 28.3%, and 41.6%, respectively, while it increased in the fruits and sepals by 398.3% and 6.2%, respectively (Fig. 1).

Antioxidant capacity

The antioxidant activities significantly varied among different treatments (see Table 1). The highest and lowest FRAP values were observed in the leaves (non-gypsum: 0.91 ± 0.02, gypsum: 0.94 ± 0.02) and pistils (non-gypsum: 0.83 ± 0.01, gypsum: 0.80 ± 0.01), respectively, indicating strong differences. Other plant parts generally showed no significant differences among them. Additionally, there were no differences in FRAP values between the two sites (Fig. 1).

High DPPH values were recorded for the seeds (non-gypsum: 75.93 ± 1%, gypsum: 94.73 ± 0.1%), leaves (non-gypsum: 86.26 ± 1%, gypsum: 97.47 ± 0.2%), fruits (non-gypsum: 91.03 ± 0.6%, gypsum: 86.45 ± 0.5%), petals (non-gypsum: 88.69 ± 0.3%, gypsum: 88.79 ± 0.8%), and flags (non-gypsum: 92.39 ± 1%, gypsum: 90.35 ± 1.4%), with no significant differences observed. However, the sepals (non-gypsum: 78.46 ± 5.8%, gypsum: 76.71 ± 1.2%) and pistils (non-gypsum: 46.88 ± 3.4%, gypsum: 59.06 ± 2.2%) exhibited significantly lower values compared to the other parts. In the gypsum site, the pistils, leaves, and seeds showed significant increases compared to the non-gypsum site (26.0%, 13.0%, and 24.8%, respectively, Fig. 1).

Regarding ABTS values, the flags from the non-gypsum site showed the highest value (non-gypsum: 89.51 ± 0.6%), while the flags from the gypsum site exhibited the lowest value (gypsum: 78.40 ± 1.5%). The leaves (89.60 ± 0.2%) and seeds (87.91 ± 0.2%) from the gypsum site, as well as the sepals (85.66 ± 0.8%) and petals (85.48 ± 0.5%) from the non-gypsum site, also demonstrated high ABTS values. The lowest ABTS value was recorded in the seeds from the non-gypsum site (49.18 ± 0.9%). The ABTS values for the pistils, seeds, leaves, and fruits from the gypsum site significantly increased (4.5%, 78.7%, 10.7%, and 12.5%, respectively), while the flags showed a decrease at the gypsum site (12.4%, Fig. 1).

Glucosinolates

The various treatments had a significant impact on the levels of individual glucosinolates. The content of glucocapparin varied considerably based on the treatments (Table 1). The highest glucocapparin levels were found in the pistils (non-gypsum: 35.6 ± 1.3, gypsum: 35.81 ± 0.9 umol/g DW), followed by the sepals (non-gypsum: 31.58 ± 0.8, gypsum: 29.39 ± 2.4 umol/g DW), leaves (non-gypsum: 30.10 ± 0.7, gypsum: 31.62 ± 1.9 umol/g DW), and fruits (non-gypsum: 27.59 ± 1.9, gypsum: 33.35 ± 0.3 umol/g DW). In contrast, the petals (non-gypsum: 22.80 ± 1.1, gypsum: 21.65 ± 0.7 umol/g DW) and flags (non-gypsum: 22.10 ± 0.7, gypsum: 23.60 ± 2.1 umol/g DW) exhibited lower values at both sites. Notably, there were contrasting increases and decreases in the levels observed in the fruits and seeds at the gypsum site compared to the non-gypsum site, with changes of 20.9% and 18.3%, respectively (Fig. 1).

Most glucosinolate compounds were found in low concentrations in the plant parts located at the gypsum site, with the exception of the leaves and pistils, which were also observed at the non-gypsum site. Specifically, n-propyl-glucosinolate was detected in the pistils (21 µmol/g DW), sepals (10.05 µmol/g DW), flags (2.74 µmol/g DW), and petals (3.42 µmol/g DW) at the gypsum site, as well as in the leaves (5.33 ± 3.2 µmol/g DW) at the non-gypsum site. N-butyl-glucosinolate was identified in the pistils (2.97 µmol/g DW), sepals (2.37 µmol/g DW), and petals (1.38 µmol/g DW) at the gypsum site. Meanwhile, glucobrassicin was present in the pistils (0.09 µmol/g DW) and petals (0.62 µmol/g DW) at the gypsum site, as well as in the leaves at both sites (non-gypsum: 0.57 ± 0.12 µmol/g DW; gypsum: 1.29 ± 0.42 µmol/g DW). Neo-glucobrassicin was exclusively found in the leaves at both sites (non-gypsum: 0.84 ± 0.42 µmol/g DW; gypsum: 3.60 ± 0.80 µmol/g DW). Additionally, hydroxy-glucobrassicin was detected in the pistils of both sites (non-gypsum: 0.19 ± 0.02 µmol/g DW; gypsum: 1.79 ± 0.56 µmol/g DW), as well as in the fruits (0.19 µmol/g DW) and leaves (0.75 ± 0.04 µmol/g DW) at the gypsum site (see Table 2).

Table 2.

Glucosinolate (GLS) contents in different plant parts of C. spinosa in two gypsum sites (mean ± SD).

		Aliphatic glucosinolates			Indolic glucosinolates
Parts	Sites	Glucocapparin	n-propyl- glucosinolate	n-butyl- glucosinolate	Hydroxy-glucobrassicin	glucobrassicin	Neo-glucobrassicin	Total GLSs
Pistils	Gypsum	35.81 ± 1.60	21.00	2.97	1.79 ± 0.79	0.09	ND	61.7 ± 5.61
	Non-gypsum	35.6 ± 2.22	ND	ND	0.19 ± 0.03	ND	ND	35.8 ± 2.25
Sepals	Gypsum	29.39 ± 4.12	10.05	2.37	ND	ND	ND	41.8 ± 13.92
	Non-gypsum	31.58 ± 1.36	ND	ND	ND	ND	ND	31.6 ± 1.36
Flags	Gypsum	23.6 ± 3.59	2.74	ND	ND	ND	ND	26.01 ± 10.43
	Non-gypsum	22.1 ± 1.15	ND	ND	ND	ND	ND	22.1 ± 1.15
Petals	Gypsum	21.65 ± 1.17	3.42	1.38	ND	0.62	ND	27.1 ± 3.44
	Non-gypsum	22.8 ± 1.89	ND	ND	ND	ND	ND	22.8 ± 1.89
Fruits	Gypsum	33.35 ± 0.52	ND	ND	0.75 ± 0.07	ND	ND	34.1 ± 0.45
	Non-gypsum	27.59 ± 3.33	ND	ND	ND	ND	ND	27.6 ± 3.33
Leaves	Gypsum	31.62 ± 3.26	ND	ND	0.19	1.29 ± 0.72	3.60 ± 1.39	36.7 ± 4.19
	Non-gypsum	30.1 ± 1.19	5.33 ± 5.51	ND	ND	0.57 ± 0.22	0.84 ± 0.73	39.2 ± 7.73
Seeds	Gypsum	24.17 ± 0.62	ND	ND	ND	ND	ND	24.2 ± 0.62
	Non-gypsum	29.56 ± 1.32	ND	ND	ND	ND	ND	29.6 ± 1.32

Multivariate analysis

Our results showed positive correlations between most parameters, except for glucocapparin, which negatively correlated with TPC and DPPH. TPC was correlated with all parameters except FRAP, and TFC correlated with ABTS and FRAP. Among the antioxidant activities, only DPPH and FRAP were correlated (Table 3).

Table 3.

The correlation coefficient between biochemical characters for C. spinosa.

Trait	TPC	TFC	ABTS	DPPH	FRAP	Glucocapparin
TPC (GAE mg/kg FW)	1
TFC (QU mg/kg FW)	0.65***	1
ABTS (%)	0.55***	0.53***	1
DPPH (%)	0.35*	0.30	0.23	1
FRAP	0.25	0.62***	0.15	0.37*	1
Glucocapparin (µmol/gDW)	-0.35*	0.13	-0.11	-0.63***	0.01	1

* P < 0.1, *P < 0.05, ***P < 0.0001.

The principal component analysis (PCA) results show that the first axis accounts for 45.5% of the variations. In comparison, the second axis accounts for 24.2% of the variations (Fig. 2). All parameters exhibited a strong positive correlation with Dim.1, except for glucocapparin, which showed a negative correlation. TFC and glucocapparin displayed strong positive correlations, while DPPH showed negative correlations with Dim.2 (Table 4). The presence of pistils was explained by Glucocapparin, which was found on the positive side of PCA axis 2 and the negative side of PCA axis (1) TFC, FRAP, ABTS, and TPC positively correlated with the gypsum site’s sepals, leaves, and fruits on the positive side of PCA axis 1 and (2) They also showed a negative correlation between the seeds, leaves of gypsum fruits, and leaves of non-gypsum site on the negative side of PCA axes 1 and 2. Additionally, the petals and flags of the non-gypsum site were explained by DPPH on the positive side of PCA axis 1 and the negative side of PCA axis 2 (Fig. 2). Sites, plant parts, and their interactions significantly affected the biochemical compositions, according to the PERMANOVA results (Table 5).

Fig. 2.

PCA results for 7 plant parts in 2 sites.

Table 4.

Correlations between chemical parameters for C.spinosa and principal component axes. *P < 0.05, ***P < 0.001.

Traits	Principal component
	Dim. 1	Dim. 2	Dim. 3	Dim. 4	Dim. 5
TPC (GAE mg/kg FW)	0.82***	− 0.01	− 0.35	− 0.43	0.01
TFC (QU mg/kg FW)	0.81***	0.50***	0.07	− 0.09	0.18
ABTS (%)	0.67***	0.18	− 0.55	0.44	− 0.14
DPPH (%)	0.66***	− 0.56***	0.32	0.23	0.31
FRAP	0.60***	0.30	0.68	0.03	− 0.29
Glucocapparin (µmol/gDW)	− 0.38*	0.88***	0.04	0.10	0.22

Table 5.

Examining the significant effects of the biochemical in the sites, plant parts, and their interactions using the PERMANOVA.

Treatments	Df	Sum of Sq.	F	R 2	P-value
Part	1	184.51	44.8	0.69	0.001
Site	1	22.04	10.8	0.28	0.001
Part×Site	1	23.27	13.3	0.21	0.001

Chemical parameters were categorized into distinct groups, with glucocapparin in the first and the remaining in the second clusters. Treatments were categorized into distinct clusters, with pistils from both sites in the first and second classes comprising two subclasses: one cluster containing leaves and sepals and the other containing petals and flags. The fruits and seeds of gypsum and non-gypsum were situated in distinct clusters (Fig. 3).

Fig. 3.

Heat map for the seven plant parts in two studied sites using biochemical traits. Mean values refer to colors from minimum displayed in dark blue to maximum represented in yellow.

Discussion

Along with morphological adaptations, species that grow in harsh drylands on gypsum soils also adapt at the biochemical level, and understanding these adaptative modifications would help to know how these species function and how can be resilient alternatives for food security purposes in a context of global warming and agri-food shortage. The research findings indicate that C. spinosa contains high levels of bioactive and chemical compounds in various parts, regardless of soil type. The chemical composition varied primarily in different plant parts and was less affected by soil type. Glucosinolates were high in all plant parts, with Glucocapparin being the predominant compound. Although gypsum did not increase Glucocapparin levels, it strongly impacted other glucosinolates in the respective sites.

Caper bush (C. spinosa) had higher antioxidant capacity and glucosinolate contents in gypsum soils

Consistent with our first hypothesis, this study showed differences in glucosinolates and antioxidant activity between two soil types. While only ABTS exhibited differences across sites irrespective of plant parts, substantial differences were observed for each part. Gypsum soil demonstrated higher antioxidant activity of ABTS and DPPH in the leaves, fruits, seeds, and pistils and higher phenolic and flavonoid content in the leaves, fruits, and some flower parts. These results contradict Çekiç et al.‘s, (2018) findings, which indicated that three studied gypsovags have lower antioxidant and phenolic content in gypsum soil than in non-gypsum soil. However, they do align with results that demonstrate higher antioxidant activity and phenolic content in gypsum soil compared to non-gypsum soil⁷, as well as with studies showing that increasing sulfur and gypsum in the soil leads to higher antioxidant activity and phenolic content^10,11.

Additionally, gypsum soil influences glucosinolates. Glucocapparin was the dominant compound, accounting for 83.3% of gypsum and 97.3% of non-gypsum without significant differences. These results show that C. spinosa has glucosinolates in high concentrations in different soil types. Other studies also showed that Glucocapparin is the predominant glucosinolate compound in C. spinosa, accounting for more than 95%⁵³. Glucosinolates with low concentrations were observed mainly in gypsum, except for a few observations in the leaves and pistils in non-gypsum. Although there are no studies for certain species in gypsum and non-gypsum habitats for comparison, several studies showed that sulfur and gypsum amendments increase glucosinolate compounds^13,14,58. The increases in antioxidant activity and glucosinolates could be linked to higher environmental stress, such as drought and the high sulfur content of gypsum soil⁴. More studies are needed on gypsum habitat species to draw better conclusions.

Flowering parts are highly variable in antioxidant capacity and glucosinolate contents

In line with our second hypothesis, we observed varying antioxidant activity and glucosinolate compounds across different plant parts. The leaves, fruits, and some flower parts exhibited the highest phenolic and flavonoid content, while the seeds had the lowest. All plant parts showed high antioxidant activity, with the leaves demonstrating the highest values and the seeds and pistils showing the lowest, regardless of the site. Other studies have also indicated that flower buds and leaves have higher polyphenol and flavonoid content^42,43,46. These studies are consistent with our findings, which show that leaves and buds have higher polyphenols and flavonoids than fruits and seeds^48,49.

Moreover, there was substantial glucocapparin content in every part of the plant. There were high values for the fruits and leaves, and the flower parts had varying amounts that ranged from highest to lowest. Pistil and sepal amounts ranged from 78 to 85 and 83.9% to more than 98% for fruits and seeds, depending on the part of the plant. Moreover, large levels of glucocapparin were found in leaves, flowers, buds, seeds, shoots, and buds^52,53,56. Our findings support those of Maldini et al. (2016), who found that flowers exhibited higher levels of Glucocapparin (364.13 ± 1.21 mg/100 g FW) than buds (295.5 ± 5.3 mg/100 g FW) and leaves (240.18 ± 2.43 mg/100 g FW) and in contrast to Bor et al. (2009) findings on C. ovata, where glucocapparin levels were higher in the seeds (39.35 ± 0.09 µmol/g DW) and leaves (25.56 ± 0.11 µmol/g DW) than in flowers (5.17 ± 0.10 µmol/g DW). According to specific research, small buds have higher levels of glucocapparin (14.07 µmol/g FW) compared to larger buds (5.41 µmol/g FW)⁵⁹. Furthermore, taking into account glucosinolates in low concentration, observations were made of n-propyl-glucosinolate and glucobrassicin in flowers and leaves, hydroxy-glucobrassicin in flowers, leaves, and fruits, and N-butyl-glucosinolate only in flowers and Neo-glucobrassicin only in leaves. Maldini et al. (2016) also reported on glucobrassicin levels in leaves (22.22 ± 0.77 mg/100 g FW), flowers (8.37 ± 0.13 mg/100 g FW), and buds (19.07 ± 0.85 mg/100 g FW). Additionally, they reported neoglucobrassicin levels in leaves (0.21 ± 0.02 mg/100 g FW), flowers (0.4 ± 0.13 mg/100 g FW), and bud (0.29 ± 0.04 mg/100 g FW) as well as 4-hydroxy glucobrassicin levels in leaves (6.56 ± 0.41 mg/100 g FW), buds (41.82 ± 1.52 mg/100 g FW), and flowers (38.52 ± 0.76 mg/100 g FW). Additionally, Bor et al. (2009) demonstrated glucobrassicin in C. ovata at the leaves (5.78 ± 0.02 µmol/g DW), flowers (0.81 ± 0.01 µmol/g DW), buds (2.53 ± 0.03 µmol/g DW), and seeds (3.82 ± 0.01 µmol/g DW). They also demonstrated 3-butenyl glucosinolate in the leaves (0.34 ± 0.02 µmol/g DW), flowers (0.23 ± 0.01 µmol/g DW), buds (1.17 ± 0.01 µmol/g DW), and seeds (1.65 ± 0.03 µmol/g DW).

The total content of glucosinolates ranged between 22.1 µmol/g (flags of Non-gypsum) and 61.6 µmol/g (pistils of gypsum) with an average of 32.7 (µmol/g DW). In comparison with other glucosinolate-containing plants from Brassicaceae, the amounts found in our study were comparable with results found in Brussels sprouts (25.1 µmol/g of dry mass) or seeds of Lesquerella fendleri (27.5 µmol/g), B. carinata (111 µmol/g), C. sativa (38 µmol/ g), L. fendleri (70 µmol/g), and L. sativum (127 µmol/g), Crambe abyssinica(115 µmol/g), Brassica napus (rapeseed) (15 µmol/g), or Brassica nigra (mustard) (130 µmol/g)⁵².

In our study, we used mature flowers rather than buds, which may account for variations in the results compared to other studies. To our knowledge, our results are the first to provide data on flower parts, including petals, sepals, pistils, and flags, in comparison to previous studies that only reported on flowers or flower buds of C. spinosa^48,49. Our results revealed surprisingly different compounds in flower parts, with variable amounts observed at two sites. For example, the ABTS in flags in soil types showed the highest and lowest values and variable records of glucosinolate compounds in the two study sites. These results highlight the important role of flower parts in C.spinosa as sources of bioactive compounds and their sensitivity to environmental conditions.

Conclusions

C. spinosa (caper bush) can thrive in various soil conditions, particularly gypsum. This suggests that C. spinosa exhibits stronger bioactive and antioxidant activity and higher levels of GLSs in gypsum soil due to the more challenging conditions of this habitat, such as high sulfur content. The variable results in flower parts indicate the need for further study of flowers better to understand the plant’s response to environmental conditions and to identify plant parts with high bioactive compounds. The high quality of fruits and seeds in both soil conditions indicates that this species could be widely used for restoring harsh lands for agricultural food production and ensuring sustainable food security in the face of global warming, which is expected to impact cropping areas worldwide.

Materials and methods

Study area and plant material

The sampling sites were located in the Abkenar rangeland (N 29º 27ʹ 22.722ʺ, E 51º 45ʹ 37.536ʺ; Kazerun city, Fars province, Iran). Two sites with similar climatic and topographic conditions, including gypsum and non-gypsum sites, were chosen, with the gypsum site being a relict gypsum mine. The average altitude was 768 m above sea level, and the dominant plants were mostly Astragalus, Ziziphus, Alhagi, Peganum harmala, and annuals. The average annual precipitation and average annual temperature were 346 mm and 23 ºC, respectively, and the region’s climate is semi-arid using the Emberger method.

Three replications of flowers, leaves, and fruit samples were taken at random from both sites under healthy C. spinosa shrubs in the summer season, transported to the laboratory, and stored at 4 °C. The leaves and fruits were dried using a freeze dryer at -80 and ground. Seven plant parts, including the sepals, petals, pistils, flags, leaves, fruits, and seeds, were considered for biochemical analysis.

The hydroalcoholic extracts were prepared to analyze the total polyphenol content and antioxidant capacity of Capparis spinosa L. formulations following the protocol described by Gironés-Vilaplana et al.⁶⁰. In short, raw materials and formulations were mixed with five volumes of methanol/formic acid/water (25:1:24).

Colorimetric assays for reducing and antioxidant capacity

We assessed TPC, TFC, DPPH, ABTS, and FRAP. The total phenolics content (TPC) was analyzed using the Folin-Ciocalteu colorimetric technique with gallic acid as the standard⁶¹. The extract’s absorbance was measured at a wavelength of 750 nm and reported as (GAE mg/kg FW). The total flavonoid content (TFC) was quantified using the aluminum chloride colorimetric method, with quercetin as the standard⁶². The absorbance of the samples was measured at a wavelength of 510 nm and quantified as (QU mg/kg FW). The DPPH free radical scavenging activity was assessed utilizing the Brand-Williams et al.⁶³ method with slight modifications at a wavelength of 515 nm. The approach of Re et al.⁶⁴ assessed ABTS radical scavenging activity with minor modifications at 734 nm. Pulido et al.⁶⁵ method was used for FRAP assay to quantify the ferric-reducing capacity of plants at a wavelength of 700 nm. A BioTek 800 TS Elisa microplate reader was used for the colorimetric method.

Glucosinolates

The extraction and analysis of glucosinolates and phenolic compounds was carried out following the method reported by Baenas et al.⁶⁶, with the modifications included in Abellán et al.⁶⁷. Briefly, 100 mg of lyophilized plant material was extracted in 1 mL 70% v/v MeOH, at 70 °C for 30 min, with occasional vortex during the extraction. After that, samples were placed in slurry ice bath for 5 min. to stop reaction, centrifuged for 15 min and 10,000 rpm at room temperature. The supernatant was recovered and filtered with a 0.22 µmØ PVDF membrane filter. The intact glucosinolates were analysed in an HPLC-DAD system, using a Luna C18 (250 mm × 4.6 mm, 5 μm) column with a Phenomenex “Securityguard” pre-column with a C18 cartridge (Phenomenex, Macclesfield, UK). Water: formic acid (99:1, v/v), and acetonitrile were used as mobile phases A and B, respectively, with a flow rate of 1mL/min. The linear gradient started with 1% of solvent B, reaching 17% solvent B at 15 min up to 17 min, 25% at 22, 35% at 30, and 50% at 35, which was maintained up to 45 min. The glucosinolates were identified following their UV spectra (227 nm), and the order of elution previously described for similar acquisition conditions. Glucosinolates were quantified using sinigrin and glucobrassicin, as external standards for aliphatic and indolic compounds, respectively (Phytoplan, Germany).

Statistical analysis

To assess the effects of sites, plant parts, and their interaction treatments on the biochemical parameters of C. spinosa, we used a generalized linear mixed-effects model (GLMM) with a Gaussian error distribution⁶⁸. In the analysis, sites and plant parts were fixed factors, with each plant sample as a random intercept. All analyses were done in R v4.4.1 (R Foundation for Statistical Computing, Vienna, AT) using the package lme4. We assessed the significance of effects by comparing the full model with a model in which a specific effect was withheld using likelihood ratio tests via the methods implemented in afex (Singmann et al., 2015). Values were reported as means ± standard error (SE) from triplicate experiments. Principle component analysis (PCA) was used to evaluate the relations between biochemical traits and correlations between the traits. PERMANOVA was performed to test for significant differences in biochemical composition between sites, plant parts, and their interactions. Multivariate analyses were conducted with the vegan⁶⁹ and factoextra packages^69,70.

Plant material

The collection of plant samples in this study was done according to legislation and formal permission of Iran Natural Resources and Watershed Management organization.

Author contributions

E.Y: Writing – original draft, Resources, Investigation. M.A: Supervision, Resources, Project Administration, Writing – review & editing, Validation, Conceptualization. T.A. A: Writing – review & editing, Validation, Methodology. D. A. M: Writing – review & editing, Supervision, Project administration, Conceptualization.

Data availability

The datasets generated during this study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.