Elderberry and Linden Flowers Ethanol–Water Extracts: Extraction Type Effect, Analysis and Biological Activity Determination

Abstract

Elderberry (Sambucus nigra L.) and linden (Tilia cordata Mill.) flower extracts are known for their pro-healthy properties. Various extraction methods, both conventional (Soxhlet) and advanced (ultrasound assisted extraction, UAE, accelerated solvent extraction, ASE, and matrix solid phase dispersion, MSPD), were applied to obtain ethanol–water extracts from the plants. The biological profiles, the total polyphenol content (TPC) and total flavonoid content (TFC) in the extracts were determined spectrophotometrically. Almost 40 compounds were identified in extracts by means of HPLC-MS/MS. The contents of the chosen phenolic acids (chlorogenic, p-coumaric, protocatechuic, and gallic) and flavonoids (rutin, catechin, quercetin, kaempferol, apigenin, and naringenin) were determined by HPLC-DAD. It was observed that the elderflower extracts contained higher levels of the compounds investigated than did the linden extracts. Chlorogenic acid was the main phenolic acid in the majority of extracts form the elderberry flower, whereas in the linden extracts, it was protocatechuic acid. Nevertheless, rutin was the main component of both plant extracts. The cytotoxicity of the elderberry and linden flower extracts against VERO, FaDu, H1HeLa, and RKO cell lines were subsequently examined. The elderflower extracts exerted no cytotoxicity, whereas linden extracts showed selective cytotoxicity against FaDu and RKO cells with CC₅₀ of 54.35 and 46.27 µg/mL, respectively. The antiviral potential of the extracts against HHV-1, CVB3, and HRV14 were also examined. The results demonstrate antiviral activity against HHV-1, particularly for linden flower extract in concentration of 62.5 µg/mL.

1. Introduction

Secondary metabolites are low-molecular-weight organic compounds that do not contribute directly to primary plant functions. However, they are essential to plant communication with the surrounding world [1]. They may play different roles, e.g., pollination (volatile compounds), antifeedants (tannins), defense against microbes (isoflavones), and adaptations to terrestrial life (flavonoids) [1]. This group of plant metabolites is divided into many subclasses, and glycosides are one of them. They consist of a sugar moiety (formed from glucose or other sugars) and an aglycone. The latter can be flavonoids or phenolic acids. The basic skeleton of flavonoids consists of 15 carbon atoms forming two aromatic rings with a heterocyclic γ-pyrone ring connecting them. The latter, with a benzene ring, forms a structure called a “chromone”. Considering the structure and substituents, several classes of flavonoids are distinguished, including flavans, flavones, flavanols, and flavanones. The most common flavonoids in plants are derivatives of quercetin, kaempferol, and luteolin [1].

In plants in addition to glycosides, free aglycones, e.g., phenolic acids, are also found. This second group of glycosides, also common in the plant kingdom, is derived from phenolic acids of benzoic acid (gallic acid, protocatechuic acid) and cinnamic acid (caffeic acid, ferulic acid). These compounds can also form dimeric structures (e.g., chlorogenic acid). Both groups, flavonoids and phenolic acids, often occur in plants alongside one another.

The pro-healthy activity of plant extracts is commonly known. Various herbal infusions, extracts, and teas have a long history in the treatment of numerous diseases. As the extrahent, water or different alcohols were usually used. The first extractions were performed using simple equipment and were subsequently improved upon with time. Given temperature considerations, maceration and percolation, performed at room temperature, are the oldest modes of extraction [2,3,4]. Higher temperatures are applied to infusions and decoctions [4]. More complex extraction methods include reflux or Soxhlet apparatus extraction [5,6]. However, these methods consume both time and solvents, and application of high temperature could result in loss of some thermolabile compounds. Recently, extraction methods have employed ultrasound (ultrasound-assisted extraction, UAE) [7,8] or microwave irradiation (microwave-assisted extraction, MAE) [9,10]. Additionally, changes to the extrahent type and composition are employed. It can be a liquid at atmospheric pressure (e.g., water or organic compounds) [2,4,11,12] or at supercritical conditions (e.g., carbon dioxide in supercritical fluid extraction, SFC) [13,14]. Ionic liquids and surfactants can also be applied as extrahent components. [15,16] High-pressure and higher-temperature conditions are applied in accelerated solvent extraction (ASE) or pressurized liquid extraction (PLE) [3,17,18,19,20]. These techniques improve extraction yield and shorten extraction time. Another modern approach, the matrix solid-phase dispersion (MSPD), limits both the extrahent volume and the time. However, the sample size is a constraint [21,22,23,24]. The main intent of applying the aforementioned improvements was to obtain extracts containing higher levels of the required active compounds while avoiding their decomposition. Additional benefits of modern technique applications appear from the economic point of view (lowering the solvent use or time, application of green chemistry solvents). The complexity and price of the equipment and requirement of skilled staff (ASE, SFC) are their drawbacks. Additionally, there is the problem of transferring the analytical mode into preparative or even industrial type (MSPD).

Extractions are the foundational stage in phytochemistry, pharmacy, and the herbal industry. The extraction mode depends on the isolated compound types, the matrix (plant type), and the future extract assignment. Broad reviews on various types of extraction have been published recently [2,3,4,25,26,27]. Zhan et al. [26] presented general extraction techniques, various analytical methods of metabolite separations and their structure identification [26]. Bitwell’s manuscript is an excellent source of data on recent applications of modern techniques (e.g., ASE, UAE) for isolating various plant metabolites, including process conditions [4]. Outstanding and recent review was prepared by Bhadange and his team [3]. They presented in detail various isolation techniques for several groups of compounds (phenolics, flavonoids, alkaloids, and essential oils) and showed their potential applications across various brands. Sun et al. are the authors of another recently published paper [27]. They critically analyzed various extraction techniques, demonstrating how extraction parameters affect the biological activities of extracts. They indicated that a combination of techniques may increase extraction efficiency and protect bioactivity. Oubannin et al. [25] focused on various methods for extracting components of antioxidant activities from the plant material and then applying them as a natural alternative to the synthetic protective compounds in vegetable oils.

Elderberry (Sambucus nigra L., Flos Sambuci) and linden (Tilia cordata Mill. or Tilia platyphylos Scop., Inflorescentia Tiliae) are widespread in temperate climate zones, i.e., between 35° and 60° north/south latitude. Due to its ornamental qualities, the linden tree is often planted in parks and along roadsides while wild plants are found in deciduous forests. Elderberry bushes also naturally occur in Europe. Elderberry can be found in Europe’s north (the southern part of Scandinavia) and down into the Mediterranean regions [28,29]. The healing effects of both plants are known from folk medicine and have been reported in numerous studies [30,31,32,33]. Both plants exhibit antioxidant, anti-inflammatory, antiviral and many other biological activities. Such qualities are attributed to their contained phenolic compounds. According to the literature, the flavonoid content in elderflower raw materials is approximately 1.8%, and in linden about 1% [34]. Rutin (rutoside; quercetin 3-rhamnoside) is the most abundant constituent among the latter secondary metabolites.

The aim of the paper was to compare several extraction methods of secondary metabolites from the two plants (Sambucus nigra L. and Tilia cordata Mill.) regarding the extraction yield, the content of the secondary metabolites by HPLC-MS technique, as well as their qualitative content (HPLC-DAD technique). The total flavonoid (TFC) and total polyphenol (TPC) contents, as well as the anti-free radical activity of the extracts were also investigated. Moreover, some extracts were also examined for their antiviral activity (against Human Herpesvirus type 1 (HHV-1), Human Coxsackievirus B3 (CVB-3), and Human Rhinovirus-14 (HRV14$)) and for cell cytotoxicity. Despite the fact that the anti-viral activities of extracts from both plants are well researched, until this study there has been no available comparison data.

2. Results

2.1. Analysis of the Chemical Composition and Activity of Extracts

Plant extracts exhibit biological activity which can be determined using radical-scavenging capacity (RSC) or equivalent antioxidant activity (EAA). Thus, the effect of the extraction mode on these properties of the obtained extracts was evaluated. In the first stage, the RSC for all investigated extracts was determined. The mode of examination was based on the literature [35].

2.1.1. Evaluation of Radical Scavenging Capacity RSC

RSC values for both plant extracts obtained by various modes of extraction are presented in

Table 1. Evaluation of radical scavenging capacity, RSC.

Extraction Type	Linden Flower Extracts RSC [%]		Elderberry Flower Extracts RSC [%]
	RSC ± SD	RSD	RSC ± SD	RSD
UAE	83.42 ± 0.22	0.27	60.79 ± 0.30	0.01
Soxhlet	95.58 ± 0.00	0.00	92.42 ± 0.08	0.09
MSPD	95.75 ± 0.00	0.00	58.38 ± 0.30	0.51
ASE	92.83 ± 0.08	0.09	48.58 ± 0.16	0.32

UAE—ultrasound assisted extraction; Soxhlet—extraction with Soxhlet apparatus, MSPD—matrix solid phase dispersion, ASE—accelerated solvent extraction, application (experiments) number 3.

.

Radical scavenging capacity ranges from 48.58 ± 0.16% (elderberry flower and ASE extraction) to 95.75% (linden flowers and MSPD extraction). Such values indicate that both plant extracts exhibit strong (elderberry flower) or very strong (linden) free radical scavenging activity [34]. RSC values were higher for linden flowers compared to S. nigra. Since the extracts of both plants are prepared for the examination in a similar manner, this indicates that the first plant exhibits greater free radical scavenging capacity. Such observation is consistent with the literature [36,37].

Regarding the mode of the extraction, the RSC values for linden flower extracts range from 83.42 ± 0.22 to 95.75%, with the highest values observed for MSPD extracts and the lowest for UEA extracts. For elderberry flower extracts, the RSC range is from 48.58 ± 0.16% to 92.42 ± 0.08%. The lowest RSC was determined for ASE extracts, whereas the highest was for Soxhlet extracts. The results obtained are divergent; therefore, it is difficult to discuss the effect of the extraction mode on this parameter. However, such discrepancies can be found in the literature [38,39,40].

2.1.2. Evaluation of Equivalent Antioxidant Capacity of Extracts

Antioxidant activities of linden and elderberry flower extracts obtained by various modes of extraction expressed in equivalent antioxidant capacity of vitamin C (ascorbic acid) (EAC) are presented in

Table 2. Equivalent antioxidant capacity of extracts obtained by various extraction mode presenting as ascorbic acid equivalents (EACs).

Extraction Mode	Linden Flower Extracts EAC		Elderberry Flower Extracts EAC
	EAC ± SD [EAC mg/g]	RSD	EAC ± SD [EAC mg/g]	RSD
UAE	41.87 ± 0.14	0.32	28.18 ± 0.18	0.64
Soxhlet	49.83 ± 0.00	0.00	18.93 ± 0.02	0.10
MSPD	49.34 ± 0.00	0.00	26.70 ± 0.04	0.17
ASE	47.57 ± 0.05	0.10	20.9 ± 0.07	0.33

UAE—ultrasound assisted extraction; Soxhlet—extraction with Soxhlet apparatus, MSPD—matrix solid phase dispersion, ASE—accelerated solvent extraction, application (experiments) number 3.

.

Results presented in Table 2 show that both extracts have high antioxidant activity, as evidenced by DPPH* scavenging [EAC in the range from 18.93 ± 0.02 (elderberry flower Soxhlet extract) to 49.83 ± 0.0 mg/g of dry plant (linden Soxhlet extract)]. The equivalent of vitamin C (EAC) in mg/g of dry plant sample for this plant ranged from 41.87 ± 0.14 (for UAE) to 49.83 ± 0.00 (Soxhlet), while EAC of elderberry flower extracts was in the range 18.93 ± 0.02 (Soxhlet)–28.18 ± 0.18 (UAE). The order of the scavenging activity decreasing for linden extracts was Soxhlet > MSPD > ASE > UAE. For the elderberry flower, the order is as follows: UAE > MSPD > ASE > Soxhlet. Thus, for both extracts, the extraction mode affects their antioxidant activity. The effect of the extraction mode on the extract’s antioxidant activity is reported in the literature [38,39,40].

2.1.3. Evaluation of Total Polyphenol Content (TPC) of Extracts

The total polyphenols content (TPCs) determined with respect to gallic acid as standard for both plants and all extracts modes is presented in

Table 3. Total polyphenols content (TPCs) in linden and elderberry flower extracts obtained by various extraction modes expressed in gallic acid equivalent (GAE) in mg/g of dry plant sample.

Extraction Mode	TPC of Linden Flower Extracts		TPC of Elderberry Flower Extracts
	GAE [mg/g of dry plant sample] ± SD	RSD [%]	GAE [mg/g of dry plant sample] ± SD	RSD [%]
UAE	56.11 ± 0.09	0.16	126.74 ± 0.06	0.04
Soxhlet	39.41 ± 0.2	0.51	96.29 ± 0.05	0.06
MSPD	53.61 ±0.17	0.32	124.90 ± 0.06	0.05
ASE	55.63 ± 0.23	0.42	107.39 ± 0.14	0.13

UAE—ultrasound assisted extraction; Soxhlet—extraction with Soxhlet apparatus, MSPD—matrix solid phase dispersion, ASE—accelerated solvent extraction, application (experiments) number 3.

. The highest TPC was measured for elderberry flower UAE extract (126.74 ± 0.06 mg/g of dry plant sample) while the lowest TPC was for linden flower Soxhlet extract (39.41 ± 0.2 mg/g of dry plant sample). The elderberry flower extracts generally showed higher TPC, ranging from 96.29 ± 0.05 (Soxhlet) to 126.74 ± 0.06 (UAE) GAE mg/g. Almost all extraction modes yielded approximately twice the TPC for elderberry flower compared with linden flower extracts. As for the latter extracts, TPC ranged from 39.46 ± 0.2 (Soxhlet) to 56.11 ± 0.09 GAE mg/g (UAE). The order of enhancement of TPC values for linden extract with respect to the extraction mode was Soxhlet < MSPD < ASE < UAE. For elderberry flower extracts, the order was Soxhlet < ASE < MSPD < UAE. Thus, the highest TPC values were observed for the UAE mode. Moreover, regarding the TPC, the advanced modes of extraction (MSPD, ASE, and UAE) prevail over the conventional (Soxhlet).

2.1.4. Evaluation of Total Flavonoid (TFC) of Extracts

Total flavonoid content in the examined extracts was determined as the rutin equivalent (RE) using rutin as the standard. The same standard has been applied by Cimpoiu in [41]. As presented in

Table 4. Total flavonoid content (TFC) in linden and elderberry flower extracts obtained by various extraction modes expressed in rutin equivalent (RE, in mg/g of dry plant sample).

Extraction Mode	TFC of Linden Flower Extracts		TFC of Elderberry Flower Extracts
	RE [mg/g of dry plant sample] ± SD	RSD [%]	RE [mg/g of dry plant sample] ± SD	RSD [%]
UAE	10.52 ± 0.01	0.08	23.93 ± 0.17	0.72
Soxhlet	4.56± 0.04	0.78	17.28 ± 0.02	0.14
MSPD	8.34 ± 0.08	0.92	33.84 ± 0.29	0.84
ASE	13.38 ± 0.04	0.32	32.25 ± 0.02	0.08

UAE—ultrasound assisted extraction; Soxhlet—extraction with Soxhlet apparatus, MSPD—matrix solid phase dispersion, ASE—accelerated solvent extraction; application (experiments) number 3.

, significantly higher TFC values were obtained for elderberry flower extracts (TFC in the range 17.28 ± 0.02–33.84 ± 0.29 RE mg/g). The lowest value is for the Soxhlet, while the highest was for MSPD mode of extraction. TFCs for linden flower extracts were in the range 4.56 ± 0.04 (Soxhlet)–13.38± 0.04 (ASE). Similarly, TFC values for elderberry flower extracts were almost twice that of linden flower extracts. The extraction method was found to have a greater impact on the TFC values. Interestingly, the order of the increasing TFC RE for linden flower extracts was Soxhlet < MSPD < UAE< ASE. For the second plant, the flower extracts was Soxhlet < UAE < ASE < MSPD extraction. Similarly to TPC, advanced extraction mode application resulted in increasing the TFC values.

2.1.5. Comparison of Plant Extracts

As shown in Figure 1, TPC and TFC values determined for all elderberry flower extracts are significantly higher than those of linden extracts. Interestingly, the latter extracts exhibit higher EAC values. The higher content of some catechins and three phenolic acids [(p-coumaric, protocatechuic, and gallic acids (see the part quantitative determination of chosen metabolites)] was noted to be responsible for the higher EAC values of linden extracts [42,43].

It can be observed that the extraction mode affects the TPC, TFC, and EAC values. Thus, with respect to polyphenols (TPCs), the UAE appears to be the most effective method for both plants, while results for TFC and EAC are inconclusive and depend on the kind of plants. However, application of the advanced modes of extraction significantly increased these parameters.

2.2. HPLC-DAD-MS Analysis

Based on HPLC-MS data, 38 phenolic secondary metabolites were identified in elderberry flower extracts and 37 in linden extracts (see Supplementary Materials). For both plant extracts, our results were comparable to those reported for elderflower (33 compounds identified by Mikulic-Petkovsek [44]) and linden (33 compounds identified by Jabeur [45]).

The fraction of identified phenolic acids and their derivatives was higher (16 compounds) in elderberry flower extracts, while the linden extracts contained 10 compounds from this group. Interestingly, apigenin and kaempferol-O-pentoside-O-deoxyhexoside were not detected in UAE extracts obtained from linden flowers.

High content of the phenolic acids in the elderberry flower extracts was confirmed by Przybylska-Balcerek [46]. These authors determined 15 phenolic acids and six flavonoids in the elderberry extract. In contrast, Mitic analyzed only three phenolic acids in linden extract [47].

HPLC-DAD analyses were conducted to assess the effect of extraction mode on the content of selected flavonoids and phenolic acids. Based on HPLC-MS experiments and our standards, the most common compounds in both groups were selected for further investigation. These were chlorogenic acid, p-coumaric acid, protocatechuic acid and gallic acid from the phenolic acids, while rutin, catechin, quercetin, kaempferol, apigenin and naringenin were for flavonoids. The concentrations of certain compounds in the extracts were too low to be quantified by our method (e.g., kaempferol and apigenin). The results are presented in

Table 5. Phenolic acid and flavonoids content in investigated S. nigra extracts.

Compound	Soxhlet Extraction		MSPD Extraction		ASE		UAE		Average Concentration
	Content ± SD [mg/g Dry Plant Material]	RSD%	Content ± SD [mg/g Dry Plant Material]	RSD%	Content ± SD [mg/g Dry Plant Material]	RSD%	Content ± SD [mg/g Dry Plant Material]	RSD %	[mg/g Dry Plant Material]
Chlorogenic acid	2.7701 ± 0.0230	0.0817	2.6593 ± 0.0006	0.0230	3.9446 ± 0.0009	0.0227	3.2556 ± 0.0001	0.0029	3.1574
p-coumaric acid	0.0052 ± 0.0000	0.1238	0.0094 ± 0.0000	0.2752	0.0098 ± 0.0000	0.0659	0.0096 ± 0.0000	0.0338	0.0085
Protocatechuic acid	0.0181 ± 0.0000	0.1110	0.0307 ± 0.0000	0.1311	0.0274 ± 0.0000	0.0734	0.0297 ± 0.0000	0.0677	0.0265
Gallic acid	0.0091 ± 0.0000	0.1927	0.0010 ± 0.0000	3.6355	0.0023 ± 0.0000	0.7530	0.0029 ± 0.0000	0.3037	0.0038
Rutin	3.585 ± 0.0008	0.0213	4.5030 ± 0.0020	0.0436	5.1029 ± 0.0007	0.0128	4.5235 ± 0.0003	0.0072	4.4286
Quercetin	0.0141 ± 0.0000	0.0969	0.0208 ± 0.0000	0.1693	0.0217 ± 0.0000	0.0813	0.0222 ± 0.0000	0.0794	0.0197
Naringenin	0.0504 ± 0.0000	0.0284	0.0415 ± 0.0000	0.0344	0.0353 ± 0.0000	0.1217	0.0425 ± 0.0000	0.0336	0.0424
Total content	6.4520		7.2656		9.1440		7.8860		7.6870

and

Table 6. Phenolic acid and flavonoids content in investigated T. cordata extracts.

Compound	Soxhlet Extraction		MSPD Extraction		ASE		UAE		Average Content
	Content ± SD [mg/g Dry Plant Material]	RSD%	Content ± SD [mg/g Dry Plant Material]	RSD%	Content ± SD [mg/g Dry Plant Material]	RSD%	Content ± SD [mg/g Dry Plant Material]	RSD %	[mg/g Dry Plant Material]
Chlorogenic acid	0.1156 ± 0.0011	0.9789	0.1456 ± 0.0001	0.0647	0.1433 ± 0.0002	0.1645	0.1199 ± 0.0000	0.0393	0.1311
p-coumaric acid	0.0142 ± 0.0002	1.1412	0.0154 ± 0.0000	0.0420	0.0144 ± 0.0000	0.0450	0.0125 ± 0.0000	0.0519	0.0141
Protocatechuic acid	0.1502 ± 0.0033	2.1934	0.2886 ± 0.0004	0.1464	0.1537 ± 0.0003	0.2224	0.1187 ± 0.0001	0.0847	0.1778
Gallic acid	0.0732 ± 0.0002	0.2166	0.1054 ± 0.0001	0.0669	0.0755 ± 0.0001	0.1401	0.0721 ± 0.0001	0.1222	0.0816
Rutin	0.8172 ± 0.0104	1.2685	1.4432 ± 0.0021	0.1437	0.8617 ± 0.0005	0.0633	0.5816 ± 0.0004	0.0750	0.9259
Quercetin	0.0106 ± 0.0000	0.3319	0.0111 ± 0.0001	0.5528	0.0083 ± 0.0000	0.1063	0.0070 ± 0.0000	0.2524	0.0092
Catechin	0.0627 ± 0.0013	2.1142	0.1464 ± 0.0001	0.0564	0.0981 ± 0.0000	0.0000	0.0647 ± 0.0001	0.1279	0.09298
Total content	1.2437		2.1561		1.3550		0.9766		1.4329

.

It turned out that the investigated plants differ not only in the number of secondary metabolites but also in their types. Thus, all elderberry flower extracts have significantly higher content of phenolic acids and flavonoids (sum from 6.4520 to 9.1410 mg/g) compared to linden extracts (sum from 0.9766 to 2.1561 mg/g). Rutin was the major component of the flavonoid group of all extracts, both elderberries and linden; however, plant extracts differed regarding the group of phenolic acids. Herein chlorogenic acid was the major component of all S. nigra flower extracts, while protocatechuic acid was the major component of linden extracts with the exception for UAE extract for which the major component was chlorogenic acid.

Kaempferol, apigenin, and naringenin levels in linden extracts were below the limit of detection of the analytical method used. Regarding elderflower extracts, the same problem was observed with kaempferol, apigenin, and catechin. Results for elderberry flower extracts are consistent with those reported in the literature [48,49]. The current literature on linden flower extracts also confirms our results [50,51,52].

Concerning the elderberry flower extracts and the concentrations of the investigated compounds, the order of decreasing concentration was rutin, chlorogenic acid, naringenin, protocatechuic acid, quercetin, gallic acid, and p-coumaric acid (see Table 5). For linden extracts, the descending order of the investigated compounds was rutin, protocatechuic acid, chlorogenic acid, gallic acid, catechin, p-coumaric acid, and quercetin. Rutin, a major secondary metabolite of the polyphenol group in both plant extracts, was also reported in [34].

Regarding the extraction modes and elderberry flower extracts, the highest contents of chlorogenic acid, p-coumaric acid, and rutin were obtained with ASE, whereas for the remaining compounds it depended on the extraction mode. Accordingly, Soxhlet extraction application resulted in increasing concentration of gallic acid and naringenin, MSPD in protocatechuic acid, and UAE in quercetin. The Soxhlet extraction method yielded the lowest content of half of the investigated compounds (p-coumaric, protocatechuic acid, rutin, and quercetin). The lowest contents of chlorogenic and gallic acids were observed in the MSPD extracts. Regarding naringenin, it was in ASE mode wherein decreasing flavonoid content was determined. The highest total content of the investigated compounds was observed in ASE extracts, whereas the lowest was observed in Soxhlet extracts. The order of decreasing total content was ASE > UAE > MSPD > Soxhlet. The effect of extraction on the identified compounds in elderberry flower extract was reported by Floares [53].

The results for the content of individual compounds in various linden extracts are more conclusive (see Table 6). Thus, extracts obtained by MSPD showed the highest concentrations of the investigated metabolites, in mg/g of dry plant material. The lowest contents were observed for the majority of the tested compounds in the UAE extracts. The order of the effect of extract type on decreasing the total content of investigated compounds was MSPD > ASE > Soxhlet > UAE.

2.3. Cytotoxicity Towards Non-Cancerous and Cancer-Derived Cells

For the assessment of cytotoxicity and antiviral activity, only one extraction mode for both plants, namely UAE, was chosen. The highest values of TPC were the reason for the choice of such proceedings.

The cytotoxicity results are presented in

Table 7. Cytotoxicity of S. nigra and T. cordata extracts.

Extract	VERO	FaDu	H1HeLa	RKO
S. nigra	>1000 **	421.27 * ± 19.40	462.47 ± 12.59	220.07 ± 14.38
T. cordata	312.82 ± 18.76	54.35 ± 2.14	140.97 ± 1.35	46.27 ± 4.66

* CC₅₀—50% cytotoxic concentration, μg/mL, mean ± SD). ** CC₅₀ was above the highest tested concentration.

. When tested on non-cancerous VERO cells, the CC₅₀ (50% cytotoxic concentration) of S. nigra extract could not be established because the viability of treated cells remained above 60% even at the highest tested concentration of 1000 µg/mL. According to the classification of plant extract cytotoxicity based on the guidelines of the National Cancer Institute (NCI) [54] and our previously published paper [55], if the CC₅₀ value is above 500 µg/mL, the extracts show no cytotoxicity. Hence, it can be concluded that S. nigra extract exerts no cytotoxic effect on VERO cells. Moreover, based on the aforementioned classification, this extract exhibited weak cytotoxic effects (CC₅₀ within the range of 201–500 µg/mL) against cancer-originating cells, namely FaDu, H1HeLa, and RKO (Table 7). In contrast, T. cordata extract exhibited weak cytotoxicity towards VERO cells. In addition, across all cancer-derived cells, a moderate cytotoxicity (CC₅₀ ranging from 20 to 200 µg/mL) was observed. The highest cytotoxicity was observed for T. cordata extract against FaDu and RKO cells, with CC₅₀ values of 54.35 and 46.27 µg/mL, respectively. Notably, T. cordata extract showed selective anticancer cytotoxicity against FaDu and RKO, exhibiting SI values of 5.76 and 6.76, respectively, when compared to VERO cells. Additionally, the analysis of dose–response curves presented in Figure 2 highlights the selective cytotoxicity of this extract against cancer-derived cells, particularly FaDu and RKO. Singularly, the viability of FaDu and RKO cells was decreased by T. cordata extract at 125 µg/mL to 13% and 22% (as compared with the cell control), while the VERO cells showed a mean viability of 70%.

2.4. Evaluation of Antiviral Potential

The first step in evaluating the antiviral activity of the S. nigra and T. cordata extracts towards HHV-1, CVB3, and HRV14 was assessing their influence on the virus-induced cytopathic effect. All three viruses included in this research are capable of inducing cytosidal infection in permissive cells. An example of non-infected VERO cells is shown in Figure 3A, while an example of HHV-1-induced CPE in VERO cells (HHV-1 virus control) is presented in Figure 3B. Acyclovir, a standard antiherpesviral drug, managed to inhibit the formation of virus-induced CPE (Figure 3C). Moreover, S. nigra extracts managed to decrease the CPE formation dose-dependently (Figure 3D–F), and this effect was most profound when the dose of 250 µg/mL was used. However, even at this concentration, there were some noticeable signs of viral infection. These included cell rounding and ballooning (Figure 3D). T. cordata extract at 62.5 µg/mL exerted particularly more potent CPE inhibition, with only minor signs of cell rounding (Figure 3G). However, at a lower dose of 31.25 µg/mL, there was only a minor effect on the HHV-1-induced CPE (Figure 3H). The assessment of anti-CVB3 activity showed that neither extract affected virus-induced CPE formation in VERO cells (Figure 4C,D), and the monolayer resembled the CVB3 virus control (Figure 4B). Ribavirin, a broad-spectrum antiviral, at 500 µg/mL, exerted a noticeable decrease in CPE, but the monolayer density was lower than that observed in the cell control (Figure 4A). Similar observations were made during anti-HRV14 experiments, where virus-infected H1HeLa cells treated with either of the extracts (Figure 5C,D) also resembled the HRV14 virus control (Figure 5B). Ribavirin was also tested on HRV14-infected cells, and at 500 µg/mL, it exerted an inhibitory effect on CPE development, but with noticeable signs of cell detachment and rounding (Figure 5E). Antiviral activity of ribavirin against CVB3 and HRV14 was dose-dependent, and at a lower concentration (250 µg/mL), there were only minor effects on CPE development (Figure 4F and Figure 5F). It should be emphasized that it was not possible to select a more effective antiviral substance against CVB3 and HRV14, as there are no approved antiviral drugs against infections caused by these viruses.

The replication of viruses in the permissive cells allows the production of infectious viral progeny. To assess the influence of S. nigra and T. cordata extracts on the production of HHV-1, CVB3, or HRV14 infectious virions, the end-point virus titration method was applied, and the results are presented in Figure 6. Accordingly, S. nigra managed to reduce the infectious titer of HHV-1 dose-dependently (Figure 6A). At a concentration of 250 µg/mL, the reduction (Δlog) reached 1.69 log. As the S. nigra concentration decreased to 125, 62.5, and 31.625 µg/mL, the reduction also gradually dwindled to 1.13, 0.56, and 0.24 log, respectively. Even more potent inhibition of HHV-1 infectious titer was induced by T. cordata extract at 62.5 µg/mL, reaching 2.73 log (Figure 6B). However, T. cordata extract at 31.25 induced only a 0.58 log reduction. Acyclovir at 60 µg/mL abolished the HHV-1 replication, and no infectious HHV-1 progeny was detected. The HHV-1 infectious titer in the virus control samples ranged between 4.51 and 4.64 logCCID₅₀/mL. Neither of the tested extracts induced a significant reduction in the infectious titer of CVB3 replicating in VERO cells (Figure 6C). Interestingly, ribavirin reduced the CVB3 infectious titer by only 0.42 log, comparable to the inhibition induced by S. nigra at 250 µg/mL (0.54 log). Both extracts at 62.5 µg/mL exerted a minor decline in the infectious titer of HRV14 replicating in H1HeLa cells, ranging from 0.32 to 0.6 log for S. nigra and T. cordata, respectively (Figure 6D).

Since both tested extracts demonstrated antiviral potential against HHV-1 in viral CPE inhibition assays and reduced the viral infectious titer, it was decided to evaluate the impact of these extracts on the viral load of HHV-1, based on measurements of HHV-1 DNA in samples collected from CPE assays using Real-Time PCR. Amplification of the HHV-1 UL54 gene in the virus-infected VERO cells treated with S. nigra and T. cordata extracts and acyclovir is presented in Figure 7A. Results of relative quantitation (ΔCq) of HHV-1 viral load, presented in Figure 7B, indicate that S. nigra at 250 µg/mL reduced the viral load by 1.35 log. A dose-dependent decrease in the reduction rate was observed, with S. nigra at 31.25 µg/mL reaching only a 0.2 log decrease in viral load. These results correspond with the abovementioned reduction in HHV-1 viral infectious titer induced by S. nigra. Importantly, qPCR amplification also confirmed the antiherpesviral potential of T. cordata, which reduced viral load by 2.59 log, and also corresponded with the reduction in viral infectious titer (2.73 log). Acyclovir reduced the HHV-1 viral load by 4.96 log compared to the virus control. The melting curve analysis performed after qPCR amplification showed a peak at 84.5–85 °C in all tested samples, confirming that the same amplicon was present in all samples.

The results obtained demonstrate antiviral potential against HHV-1, particularly with respect to the extracts of T. cordata. Corresponding results of infectious titer and viral load reduction may indicate that this activity results from the inhibition of intracellular replication of HHV-1, probably at early stages of its replication, and due to the inhibition of viral nucleic acid transcription or replication [56].

3. Discussion

The biological activity of plant extracts can be determined using various means, among others phytochemical assays. Such research usually starts with in vitro antioxidant activity measurements, either spectrophotometrically in RSC or using compounds of established antioxidant activity (e.g., Trolox, vitamin C, TEAC, EAC). The same technique is applied to determine total flavonoid (TFC) or total polyphenol (TPC) concentrations. TFC is determined based on the reaction of complexation with aluminum or ferric salt by flavonoids and is expressed as quercetin, rutin, catechin, or apigenin equivalents. TPC is based on the phenolic compound’s ability to reduce oxidizing agents and form colorful complexes. The Folin–Ciocalteau reagent is employed in this assay. The results are presented as equivalents of the standards. Gallic acid, caffeic acid, or catechin can be used as the standards [43,57].

The research is supplemented by HPLC analysis of the extract, which determines specific compounds qualitatively and quantitatively. As linden and elderberry flower extracts are known for their biological activities, their examination for such action seems obvious.

In this study the antioxidant activities of elderberry and linden flower ethanolic–water extracts were investigated based on the reaction with DPPH*. Vitamin C solution was applied as the standard.

Thus, the effect of the extraction method on the antioxidant activity of investigated extracts was examined in our research. In our research, free radical scavenging activities of elderberry flower extracts ranging from 48.58 to 60.78% were determined for ASE, MSPD, and UAE modes. The literature indicates that higher percentage ranges result in higher anti-free radical activity [35]. Measured % values for most extraction methods indicate that elderberry flower extracts exhibit strong antioxidant activity. This parameter for Soxhlet extract, which showed the highest RSC (92.42%) and the strongest antioxidative activity, can be explained by the higher mass of the raw plant material (2.5g) used in the extraction process. Determination of RSC was based on the percentage of color inhibition and was not linked to the amount of the standard equivalent per g of the sample (as in EAC). Thus, samples of lower raw material mass [e.g., 1g (UAE and ASE) or 0.5 g (MSPD)] exhibited lower RSC.

Kołodziej and Drożdżał investigated elderberry flowers and fruits ethanolic and methanolic extracts from plants growing at various part of Poland using RSC and DPPH* assay [58]. Their results showed higher anti-free radical activity in flowers and ethanolic extracts than in fruits and methanolic extracts. Considering the linden extracts obtained by various extraction modes, all extracts showed very strong free radical scavenging activity (above 80%). The differences between the three extraction methods (Soxhlet, MSPD, and ASE) are insignificant (lower than 3%), while the difference for the UAE is significantly higher (around 12%). High values of free radical scavenging for the linden extracts were also measured for supercritical CO₂ extraction by Pieczykolan et al. [59].

The RSC values are significantly higher for the linden than for elderflower extracts. Such a comparison is presented for the first time.

Interestingly, we found that the extraction method significantly influences the RSC activities of extracts only for elderberry flowers. Indeed, this parameter did not significantly affect the linden flower extract (a 12% difference in RSC), whereas it substantially changed the RSC for elderberry flower extracts (43.83%). Such differences in activities were also observed in the literature by Dzięcioł et al. [38,39], Lim et al. [40], and Cimpoiu et al. [41].

Antioxidant activities measured using equivalents of standards with established activities is another assay that describes the biological activity of the extracts. In our tests based on the literature, we chose ascorbic acid (vitamin C) as the standard [(EAC; equivalent antioxidant capacity of vitamin C (ascorbic acid))] [42]. This standard was also successfully applied by Tadesse et al. [60].

Consistent with previous antioxidant assays (RSC), linden extracts exhibited higher and more consistent EAC values (41.87–49.83 EAC). For elderberry flower extracts, EAC values were more variable and extraction mode-dependent. Such an effect was also described by Dziecioł et al. [39]. Regarding the Sambucus nigra (elderberry flower) teas, such protocol (the assay with EAC) was presented by Viapiana and Wesołowski [61]. Considering linden extract, the EAC mode was used by Bardakci et al. in research on the effect of the brewing material on the biological activities of linden infusions [62].

The literature presented only one data point, which compared both plant flower extracts [63]. The authors’ research confirmed our findings of higher antioxidant activity in linden flower extracts. However, the research employed the ABTS assay [62].

The antioxidant activities of elderberry flower extracts were measured using Trolox equivalents by Tobaszewska and Sikora [36]. In their study, they investigated the effect of elderberry flower stabilization on the extract’s antioxidant activity. Therein they compared fresh, frozen, air-flow-dried, and freeze-dried flowers. It turned out that the highest antioxidant activity was observed for the last.

Antioxidant activity can be expressed as IC₅₀, the concentration at which 50% of free radicals are scavenged. Comparison of the antioxidant activity of linden flowers, carob, and clove ethanolic extracts with DPPH and ABTS assays expressed as IC₅₀ were presented by Güller et al. [64]. Among the plants investigated, clove extract showed the highest antioxidant activity. Higher, similar IC₅₀ values were determined for linden and carob extracts.

The spectrophotometric determination of the total flavonoid content (TFC) and total polyphenol content (TPC) was the next step of the investigation. Based on the methodology presented in [65] we compared TFC (GAE) and TPC (RE) of the plant’s extracts examined. In doing so, significantly higher values of both abovementioned parameters were determined for elderflower extracts. Such a comparison for elderflower and linden flower extracts was presented for the first time.

Güller et al. also compared TPC and TFC for linden flowers, carob, and clove extracts [64]. The first parameter was determined in gallic acid equivalents, GAE, while the latter was undertaken in quercetin equivalents, QE. In this research, the highest TPC was observed for clove, while the lowest was observed for linden extract. A quite contrary result was obtained for TFC, and linden extracts contained the highest QE among the plants investigated.

The stabilization mode also affected the total flavonoid (TFC) and total polyphenol (TPC) contents. Thus, the latter, measured in catechin equivalents, also showed the highest value for free-drying stabilized material. Conversely, the highest TFC, expressed as catechin equivalents, was determined for air-flow drying, while the lowest was for frozen [36].

The effect of the various extraction techniques (maceration, maceration with shaking, ultrasound-assisted extraction, and reflux extraction) on the biological activities of extracts was examined by Dzięcioł et al. [38] regarding maca root extracts and RSA (radical scavenging activity). They noticed that reflux extraction was the most efficient. Moreover, except for the process mode, the effect of the extrahent and the extraction time was crucial.

Effects of various parameters including the mode of the extraction effect on the antioxidant profile of the Sambucus nigra extracts were explored by Floeres et al. [53]. The authors compared conventional and accelerated (ultrasound- and microwave-assisted) extraction methods. The results showed that the accelerated modes resulted in higher TPC. Our results confirmed these determinations. Therefore, the TPC of both plants’ extracts was highest when ultrasound-assisted extraction was used.

Regarding the evidenced TFC and our results application of the modern modes (MSPD or UAE) also brings about higher values of this parameter for both plant extracts.

The qualitative and quantitative analyses of the investigated extracts were performed using the HPLC method with MS-MS detection (for qualitative analysis) or DAD detection (for quantitative analysis). Our research regarding the previous analysis (see Supplementary Materials) confirmed the literature data regarding the secondary metabolites both in linden and elderberry flower extracts [44,45].

Elderberry flower extracts were richer in phenolic acids and their derivatives than linden extracts.

Extraction method used affects the amount of isolated secondary metabolites. Several papers have investigated such relationships, particularly with respect to phenolic compounds [40,43,44,45,50,51]. For example, Oniszczuk et al. compared the extraction process modes on the amount of the secondary metabolites isolated [50,66]. They investigated the efficiency of Soxhlet, USE, ASE, and MSPD extractions of selected polyphenolic compounds (phenolic acids and flavonoids) [66] using Equisetum as the plant material, or UAE, ASE. MAE. Soxhlet and HRE were used for linden [50]. The highest extraction efficiencies for many of the compounds investigated were achieved with the USE. However, for some compounds, e.g., rutin and quercetin, the highest extraction efficiency was obtained using ASE, whereas for quercetin and tilioside, this came about when MASE was applied [50]. Thus, the extraction efficiency of some compounds depended on their structure (functional group).

Contrary to the results reported by Oniszczuk [50] for linden extracts, our research showed the highest extraction efficiency, determined as the amount of compounds isolated, was for the MSPD mode while the lowest was indicated for the UAE mode. However, both tests differed in the composition of the extrahent (ethanol and water, 70:30% v/v in our research, and 80% of methanol in water in [50]).

The effect of the extraction mode on the extraction efficiency of elderberry flower extracts was investigated by Vujanović et al. [67]. The authors compared MAE, UAE, and maceration as extraction methods and found that MAE was the most efficient. In turn, three methods (conventional, USE, and microwave-assisted extraction) for extracting secondary metabolites from elderberry flowers were compared by Floares et al. [53]. The highest efficiency was observed with microwave-assisted extraction (advanced extraction mode), whereas the lowest was seen with conventional extraction. Our results are consistent with the abovementioned paper since the highest extraction efficiency was achieved with the advanced (ASE) mode.

Linden flowers are commonly used as a traditional medicine to relieve cold, cough, and sore throat symptoms [68,69,70]. Previously, Tilia platyphyllos flower methanolic extracts and their subfractions were evaluated for potential anti-influenza activity but were found to be ineffective [71]. Moreover, T. cordata also did not exert anti-influenza activity [72]. In contrast, Tilia amurensis honey inhibited influenza A virus replication in murine macrophages, probably by regulating innate immune response in these cells [73]. There are reports on the antibacterial activity of linden tea [74] and essential oils [75]. However, surprisingly, reports on the antiviral activity of linden flower extracts are very scarce. In this respect, our study aims to fill this gap and provide information on the antiviral study of linden flower extract towards HHV-1, CVB3, and HRV14. Our results established that T. cordata extract at 62.5 µg/mL potently decreased the HHV-1-induced CPE, reduced the infectious titer by 2.73 log, and the viral load by 2.59 log. However, no significant activity was observed against two RNA viruses included in this research.

T. cordata flower extract was previously found to exert a selective cytotoxic activity against the human pancreatic cancer cell line (MIA PaCa-2; CRM-CRL-1420), with the cellular viability reduced by more than 50% at 500 µg/mL. However, an extract from Tilia rubra flowers exhibited higher cytotoxicity against MIA PaCa-2. T. rubra extracts also significantly reduced the growth rate of MIA PaCa-2 spheroids and induced a G2/M cell cycle arrest [76]. Our study also evaluated the cytotoxicity of linden flower extract towards cell lines derived from three cancer types. Accordingly, across all these cell lines, observed selective cytotoxicity was compared to VERO cells. Interestingly, the linden flower extract showed significantly higher cytotoxicity towards FaDu, H1HeLa, and RKO than previously reported for MIA PaCa-2, with CC₅₀ values ranging from 46.27 to 140.97 µg/mL.

Elderberry fruit juice was also reported to have anti-influenza activity [77]. Methanolic extracts from leaves and flowers of S. nigra exhibited antiviral activity against dengue virus 2 (DENV-2), and this activity was observed predominantly in virucidal assays at 400 µg/mL [78]. There is also a report on anti-influenza and anti-HHV-1 activity of an infusion prepared from a herbal composition of Sambucus nigra L. flowers, aerial parts of Hypericum perforatum L., and roots of Saponaria officinalis L. [79] Moreover, there are reports that elderberry supplementation can effectively treat upper respiratory symptoms associated with the common cold and influenza [80] and COVID-19 (coronavirus disease 2019) [81]. However, some concerns have been raised about the potential overstimulation of the human immune system and an increased risk of a cytokine storm [80]. Sinupret^®, a herbal medicinal product made of five medicinal plants, including primrose flowers (Primula vulgaris L.), gentian root (Gentiana lutea L.), sorrel herb (Rumex acetosa), elderberry flowers (Sambucus nigra), and vervain herb (Verbena officinalis L.), was found to exert noticeable antiviral activity against influenza A, parainfluenza type 3, respiratory syncytial virus (RSV), human rhinovirus B subtype 14 (HRV 14), coxsackievirus subtype A9 (CVA9), and adenovirus C subtype 5 (Adeno 5) [82]. Since most studies focus on the antiviral activity of herbal mixtures containing elderberry flowers, it is crucial to establish whether this activity is also exerted by extracts obtained solely from this plant, and, importantly, against both DNA and RNA viruses. In our studies, the elderberry flower extract also exerted a noticeable antiherpesviral effect at 250 µg/mL, decreasing HHV-1-induced CPE, reducing the viral infectious titer by 1.69 log and viral load by 1.35 log. Only a minor effect on the infectious titer of CVB3 or HRV-14 was observed. However, it is worth noting that, due to different extraction methods, the reported content of bioactive compounds in S. nigra fruits and flowers shows substantial variability, which could affect biological activity, including antiviral activity [83].

Both linden and elderberry flowers are used to treat symptoms of upper respiratory tract infections. Thus, our studies explored their antiviral effect against HRV-14, one of the most common causes of the common cold. However, no significant activity against this virus or against another tested RNA virus, CVB3, was observed. The antiviral activity against HHV-1 was more potent in the case of linden than in elderberry flowers. Although T. cordata and S. nigra extracts studied herein show the presence of similar phenolic compounds, some noticeable differences may be responsible for disparities in antiviral activity. For example, kaempferol-O-rhamnoside was found exclusively in linden, and previous silico studies suggested that kaempferol-O-rhamnoside derivatives may be inhibitors of Monkeypox and Marburg viruses [84]. In fact, kaempferol and its derivatives were shown to inhibit DNA viruses [83], including kaempferol-3-O-rutinoside and quercetin 3-O-rutinoside, which inhibited HHV-1 [84,85], and were identified in both linden and elderflower. Also, quercetin and its derivatives exhibit versatile antiviral effects against various viruses, including hepatitis B and C viruses, coronaviruses, human respiratory syncytial virus (RSV), herpesviruses, influenza, picornaviruses, and retroviruses [86]. Hence, disparities in antiviral potential may result from differences in the amounts of these compounds or from synergistic interactions with other bioactive secondary metabolites. On the other hand, proanthocyanidins have been reported as antiherpesviral agents [87], and procyanidin dimers, trimers, and tetramers were found only in linden. Interestingly, A-type proanthocyanidins have been proposed as novel broad-spectrum antiviral agents capable of inhibiting the replication of many human viruses, including enveloped and non-enveloped DNA and RNA viruses [88].

4. Materials and Methods

4.1. Materials

Rutin, apigenin-7-glycoside, naringenin-7-glycoside, isocytoside, diosmin, genistein and isoquercetin were purchased from Roth (Karlsruhe, Germany). In contrast 2-hydroxyflavanone, naringenin, coumarin, acacetin, hesperetin, luteolin and 6,7 dihydroxycoumarin were procured from Sigma-Aldrich (Saint Luis, MO, USA). Flavanone and kaempferol originated from Alfa Aesar, Ward Hill, MA, USA, while myricetin was from Merck Millipore. Acetonitrile, acetic acid, and Folin—Ciocaulteu reagent were from Chempur (Piekary Śląskie, Poland), while acetone, ethanol, chloroform, 2-propanol, ethyl acetate, and sodium carbonate were purchased from Polskie Odczynniki Chemiczne S.A. (Gliwice, Poland). Dichloromethane and methane were from Sigma-Aldrich (Steinheim, Germany), and citric acid was obtained from Merck Millipore (Darmstadt, Germany). Regarding MSDP extraction equipment and material used and sourced were: BAKERBOND Octadecyl gel 40 µm (J.T. Baker, Radnor, PA, USA), TLC chromatographic plates (10 × 20 cm) TLC Silica gel 60 F254 and HPTLC chromatographic plate 10 cm × 20 cm (HPTLC Silica gel 60 F254 Premium Purity, Merck Millipore (Darmstadt, Germany). HPLC Zorbax Eclipse XDB-C18, columns (particle diameter 5 μm, 3 × 150 mm or 4.6 × 150 mm) were from Agilent Technologies (Agilent Technologies, Santa Clara, CA, USA).

4.2. Equipments

The following equipment accomplished the extraction processes: Soxhlet extractions (Büchi Labortechnik, Eseen, Germany), ultrasonic assisted extraction (ultrasonic bath (Emag, Mörfelden-Walldorf, Germany), ASE (ASE 100 extractor, Dionex, Sunnyvale, CA, USA), MSPD (SPE system containing of vacuum pump and SPE chamber).

The total flavonoid content (TFC), total polyphenols content (TPC), and antioxidant activity of the investigated extracts were determined using the spectrophotometer UV–Vis (ThermoFisher Scientific, Waltham, MA, USA).

The quantitative analyses of the extracts were performed by utilizing Agilent 1290 Infinity II LC coupled with an Agilent 6470 Triple Quadrupole MS system (Agilent, Santa Clara, CA, USA). The mobile phases comprise an aqueous acetic acid (5%) and acetonitrile solution.

4.3. Methods

4.3.1. Plant Material Origin and Preparation

The plant material for the research was commercially available a small-leaved linden flower (Latin: Tilia cordata) and an elderflower (Latin: Sambucus nigra). The linden flower was grown in Grodzisk by the “Dary natura” company, while the elderflower was purchased from the “Flos” brand, which runs a plantation in Mokrsko. To ensure appropriate quality and durability, the plant material was stored in the same dry conditions and at room temperature throughout the research time.

The plant material was crushed into powder using at mortar.

4.3.2. Extraction Procedures

The extrahent composition (ethanol and distilled water, 70:30% v/v) was chosen based on the literature [53,89].

Ultrasound Assisted Extraction (UAE)

A total of 1 g of the plant material powder, both elderberry flower and linden flowers, was poured into the amber glass bottles and 10 mL of the extrahent (70% of ethanol in distilled water) was added. The vessel was introduced into an ultrasound bath filled with warm (50 °C) distilled water. The extraction process was conducted for 30 min (at 75% of the optimal equipment power). The extract was then gently poured at the weighted baker. The solid plant material was extracted again twice, using 10 mL of the extrahent. The extracts were filtered through paper filters, combined and evaporated under atmospheric pressure using a hotplate at a temperature of 50 °C. The dried extracts were, afterwards, dissolved in an ethanol and distilled water mixture (50:50% v/v, 10 mL) and subsequently filtered through a 0.1 μm pore diameter filter, into an amber glass flask. The solutions were stored in a refrigerator at 5 °C.

Some extracts were prepared with citric acid (1% w/v) as the extrahent component. The main procedure was the same for these extracts.

Soxhlet Apparatus Extraction

A total of 2.5 g of the plant material powder was placed into tissue paper thimbles. The samples were then put into the Soxhlet apparatus. The extraction was done using 100 mL of an extrahent (70% ethanol-distilled water mixture). The continuous extraction process was performed for 4.5 h. The obtained extract was filtered and evaporated to dryness under a hood under atmospheric pressure using a hotplate at 50 °C. The dry extract was then dissolved into 10 mL of a mixture of ethanol and water (50:50% v/v) and filtered through a 0.1 μm pore diameter filter to an amber glass flask. The solutions were stored in the refrigerator at 5 °C.

Matrix Solid Phase Dispersion, MSPD, Extraction

A total of 0.5 g of the powdered plant material and 2 g of the C-18 modified silica gel grains were poured into an agate mortar and processed for 15 min. The matrix obtained in this mode was gently placed in the column equipped with a security filter and then compacted using the plunger. Another security paper filter was placed on top of the column. The column prepared this way was inserted into the elution socket in the SPE chamber’s cover connected to the pump. In the center of the chamber, a 10 mL volumetric flask was employed as the extract receiver.

The eluent was eluted with a mixture of ethanol and water with a volume of 10 mL (70%:30% v/v), being gradually added to the front of the column. The procedure was supported by a vacuum of −0.01 MPa and was performed twice for linden flower and elderflower, using 1 g of plant substance and 4 g of sorbent. The final extracts were evaporated to dryness on a heating plate at 50 °C, and the residues were dissolved in a mixture of ethanol (5 mL) and water (5 mL). The obtained extract was filtered with a 0.1 µm filter and stored in a dark glass bottle in a cool place.

Accelerated Solvent Extraction, ASE

The plant powder of the 1 g mass was placed into the extraction cell of the ASE 100 apparatus. A mixture of ethanol and distilled water (70% + 30% v/v) was applied as an extrahent. The extraction process conditions are presented in the table below (

Table 8. ASE process conditions.

Parameter	Value
Pressure [bar]	100
Temperature [°C]	70
Stationary time [min]	5
Flow volume [%]	80
Cleaning time [s]	60
Number of cycles	3

).

4.3.3. Extracts Composition and Bioactivity Determination

Total Polyphenol Content Determination

Total polyphenol content (TPC) was determined spectrophotometrically based on the prescription presented in [64] with some improvements. The basing extract solution was diluted with distilled water around 10–20 times depending on the extract and plant type, and precisely 100 μL of extract was mixed with 1000 μL of freshly prepared Folin–Ciocalteu reagent (diluted 1/10 with redistilled water). After 5 min, 1000 μL of a 7.5% Na₂CO₃ solution was added. The sample was then placed in the dark at room temperature for 60 min of incubation. After this time, the sample absorption was measured at 760 nm of light wavelength. Deionized water was applied as a blank. The calibration curve was prepared using the gallic acid solution standards in the range of 7.81 μg/mL to 500 μg/mL measurement (both extract samples as calibration curve points), which was triplicated.

A plot was subsequently constructed with the linear relationship (y = 0.0027x + 0.0427, R² = 0.9961) based on absorption vs. gallic acid concentrations. This equation was afterwards employed as a means of determining the polyphenols content in the extract sample (presented as milliequivalents of gallic acid). The TPC of the extraction type was then established using the formula given by Alara [90] and expressed as milliequivalents of gallic acid (GAE) per g of dry extract (mg GAE/g of dry extract).

Total Flavonoid Content

Total flavonoid content (TFC) was determined spectrophotometrically based on the prescription presented in [65] with some improvements. The basing extract solution was diluted with water around 7–10 times (depending on the extract and plant type), and precisely 500 μL of extract was mixed with 200 μL of freshly prepared aluminum chloride solution. Methanol was then added to the mixture in such a way that the final volume was 2000 μL. Next, the sample was placed in the dark at room temperature for 40 min of incubation. After this time, the sample absorption was measured at 415 nm of the light wavelength. A solution of aluminum chloride without standard was utilized as a blank. The calibration curve was prepared through the use the rutin solution standards in the range of 31.3 μg/mL to 500 μg/mL [66,67]. All measurements (both extract samples as calibration curve points) were triplicated.

A plot was constructed with the linear relationship (y = 0.0028x + 0.014, R² = 0.9983) based on absorption vs. rutin concentrations. This equation was used to determine the flavonoid content in the extract sample as presented as the milliequivalents of rutin. The TFC of the extraction type was established using the formula given by Alara [90] and expressed as milliequivalents of rutin (RE) per g of dry extract (mg RE/g of dry extract).

Evaluation of Radical Scavenging Capacity

Evaluation of radical scavenging capacity (RSC) of extracts was established based on the prescription in [91] with some changes. A methanolic solution of Table 2,2-diphenyl-1-picrylhydrazyl hydrate radical, DPPH* (Sigma–Aldrich, Steinheim, Germany) (2 mL, 6 × 10⁻⁵ M) was spiked with 20 μL of diluted extract (1:10). Decreasing absorbance of DPPH was measured spectrophotometrically at 515 after the 30 min of incubation at room temperature in dark. Methanolic solution of DPPH without extract was used as the blank sample. The RSC was calculated using the following formula:

RSC = ((absorption of blank sample − absorption of extract sample)/absorption of blank sample) × 100%

Equivalent Antioxidant Capacity Evaluation

Evaluation of antioxidant capacity of extracts occurred by applying the prescription in [91] with some changes. A methanolic solution of Table 2,2-diphenyl-1-picrylhydrazyl hydrate radical DPPH* (Sigma–Aldrich, Steinheim, Germany) (2 mL, 6 × 10⁻⁵ M) was spiked with 20 μL of diluted extract. Decreasing absorbance of DPPH was then measured spectrophotometrically at 515 nm after 30 min of incubation at room temperature in the dark. A calibration curve was established via the standard solutions of vitamin C in the range 0.01–0.5 mg/mL. Afterwards, a plot was constructed based on absorption vs. vitamin C concentrations by means of the linear relationship (y = −1.2062x + 0.6261, R² = 0.9979). This equation was subsequently applied to determine the equivalent antioxidant capacity of the extract sample presented as the milliequivalents of vitamin C per g of extract sample.

4.3.4. High-Performance Liquid Chromatography HPLC-DAD

Regarding the qualitative analysis of the extract phenolic compounds, see the Supplementary Materials.

Extracts were analyzed by means of an Agilent 1290 Infinity LC System (Santa Clara, CA, USA) with a diode array detector (DAD). Column Zorbax Eclipse Plus-C18 column (4.6 × 100 mm, 3.5 µm, Agilent, Santa Clara, CA, USA) was employed. The column oven was set at 40 °C, while the injection volume was 20 µL. The elution solvents were 1% acetic acid in water (A) and methanol (B) pumped at a flow rate of 1.5 mL min⁻¹. The analytes were eluted according to the following gradient (only B % values are presented, the rest of the mobile phase composition consisted of solvent A): 0–13 min from 90% to 15% B; 13–26 min linear gradient from 15% to 25% B; 26–31 min linear gradient from 25% to 30% B. The UV absorption spectra of the standards and samples were measured between 280 and 360 nm. The chosen detection wavelengths are different for different analytes and represent their absorbance maxima or are close to them.

Quantitative Analyses of Investigated Compounds

The amount of the investigated compounds was investigated through applying HPLC-DAD method and were based on calibration curve equations. The calibration range, equations, Pearson coefficients, LOQ and LOD values are presented in

Table 9. Figures of merit of the phenolic compounds’ determination by HPLC-DAD.

Compound	Linear Range [µg/mL]	Linear Equation	Pearson Coefficient	LOD [µg/mL]	LOQ [µg/mL]
Quercetin	1–100	y = 4.018x − 1.352	0.999	6.63	20.10
Chlorogenic acid	5–1000	y = 1.500x − 8.342	0.999	36.03	109.20
Catechin	5–150	y = 0.854x − 2.678	0.999	6.95	21.10
p-Coumaric acid	1–50	y = 10.92x − 6.977	1.000	0.75	2.30
Protocatechuic acid	2.5–100	y = 3.515x − 1.590	0.999	5.15	15.60
Gallic acid	1–50	y = 4.011x + 1.411	0.999	1.03	3.10
Rutin	50–1500	y = 0.648x − 10.91	0.999	38.49	116.60
Naringenin	5–200	y = 4.944x + 6.315	0.999	9.99	30.30
Caempferol	5–300	y = 2.2976x − 0.3059	0.999	27.12	82.19
Apigenin	5–200	y = 5.969x − 16.93	0.994	48.31	146.39

.

4.4. Cell Lines and Culturing

This research utilized a panel of cell lines that included non-cancerous VERO cells (ATCC, CCL-81, and monkey kidney fibroblasts) and cancer-derived cell lines: FaDu (human hypopharyngeal squamous cell carcinoma, ATCC, HTB-43), H1HeLa (human cervical adenocarcinoma, ATCC, CRL-1958), and RKO (human colon cancer, ATCC, CRL-2577). VERO cells were cultured in Dulbecco Modified Eagle Medium (DMEM, Corning, Tewksbury, MA, USA), whereas cancer cell lines were maintained in Modified Eagle Medium (MEM, Corning). All media were supplemented with antibiotics (Penicillin-Streptomycin Solution, Corning) and fetal bovine serum (FBS, Corning; 10% FBS for cell passaging and 2% FBS for cell maintenance and experiments). Trypsin and phosphate-buffered saline (PBS) were obtained from Corning, while MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) was sourced from Sigma (Sigma-Aldrich, St. Louis, MO, USA). Cells were incubated in a 5% CO₂ environment at 37 °C (CO₂ incubator, Panasonic Healthcare Co., Tokyo, Japan). Stock solutions of T. cordata and S. nigra extracts were prepared by dissolving them in cell culture-grade DMSO (PanReac Applichem, Darmstadt, Germany) to achieve a concentration of 50 mg/mL, which were then stored at −23 °C until required for the experiments.

4.5. Cytotoxicity Evaluation

The cytotoxicity was evaluated by way of a microculture tetrazolium-based assay, following the methodology detailed previously [92]. In summary, selected cell line monolayers in 96-well plates were treated with serial dilutions of extracts in cell media for 72 h. After incubation, the cell media were discarded, plates were rinsed with PBS, and MTT-supplemented media were added for an additional 3 h incubation. The resultant formazan precipitate was then dissolved, and absorbance was measured at 540 and 620 nm using a Synergy H1 Multi-Mode Microplate Reader (BioTek Instruments, Inc. Winooski, VT, USA) equipped with Gen5 software (ver. 3.09.07; BioTek Instruments, Inc.). The gathered data were exported to GraphPad Prism (version 10.2), where 50% cytotoxic concentration values (CC₅₀) were derived from dose–response curves (non-linear regression). Selectivity indexes (SIs) were also computed (SI = CC₅₀VERO/CC₅₀CancerCells). The differences in CC₅₀ values across cell lines were subjected to statistical analysis using GraphPad Prism (one-way ANOVA, Tukey’s multiple comparisons test, alpha = 0.05). Moreover, non-toxic concentrations towards VERO and H1HeLa were selected for further use in antiviral studies. Non-toxic concentrations were herein defined as concentrations decreasing the cellular viability by not more than 10%. Importantly, the selected concentrations could not influence the morphology of the cellular monolayer, as compared with the non-treated cell control.

4.6. Antiviral Assay

Antiviral activity was assessed against Human Herpesvirus type 1 (HHV-1, Orthoherpesviridae, Alphaherpesvirinae, Simplexvirus, ATCC, VR-260) and Human Coxsackievirus B3 (CVB3, Picornaviridae, Enterovirus, Human enterovirus B, ATCC VR-30) propagated in VERO cells, as well as against Human Rhinovirus-14 (HRV-14, Picornaviridae, Enterovirus, Rhinovirus B, ATCC VR-284) replicating in H1HeLa, as previously described [93,94]. The antiviral assays included the influence of extracts on the formation of virus-induced cytopathic effect (CPE), reduction in viral infectious titer, and, for anti-HHV-1 experiments, the semi-quantitative assessment of the viral load reduction using Real-Time PCR. Briefly, the VERO or H1HeLa cells growing in a monolayer in 48-well plates were infected with HHV-1, CVB3, or HRV-14 in 100-fold CCID₅₀/mL (CCID₅₀–50% cell culture infectious dose) and incubated for 1 h. Uninfected cells were left as a cell control. Acyclovir and ribavirin were used as standard antiviral drugs. Subsequently, the cells were washed with PBS to remove unadsorbed viral particles, and the extracts in non-cytotoxic concentrations were added. Incubation was continued until CPE (cytopathic effect) was observed in the virus control (VC; HHV-1-infected, untreated cells). Afterwards, the 48-well plates were observed and documented with an inverted microscope (CKX41, Olympus Corporation, Tokyo, Japan) equipped with a camera (Moticam 3+, Motic, Hong Kong, China) and software for image documentation (Motic Images Plus 2.0, Motic). The plates were then thrice frozen (−76 °C) and thawed, and samples for virus titration and viral DNA isolation were collected.

Samples obtained from antiviral assays underwent an end-point dilution assay to determine the HHV-1, CVB3, and HRV14 infectious titers [94]. In brief, VERO (for HHV-1 and CVB3) or H1HeLa (for HRV14) cells monolayer in 96-well plates were incubated for 72 h with ten-fold dilutions of samples (three replicates) in cell media. Daily observations were undertaken to assess the progression of CPE. Following incubation, all media were discarded, and the virus infectious titer for each sample was evaluated based on cellular viability by applying the previously outlined MTT method. The difference (Δlog) between the infectious titer of selected virus (logCCID₅₀/mL) in the samples treated with tested extracts or reference drugs and the virus control (VC) from the same experiment was calculated using the formula (Δlog = logCCID₅₀VC − logCCID₅₀Sample). Data analysis was performed through GraphPad Prism.

Real-time PCR (qPCR) assessed the effects of T. cordata and S. nigra extracts on HHV-1 replication in infected cells, focusing on the variations in HHV-1 viral load between extract-treated samples and the VC. The DNA was isolated (QIAamp DNA Mini Kit, QIAGEN GmbH)) from pooled samples collected in three replicates of antiviral assays. Subsequently, qPCR amplification employing SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories, Life Science Group, Hercules, CA, USA) and primers (UL54F–5′CGCCAAGAAAATTTCATCGAG 3′, UL54R–5′ ACATCTTGCAC CACGCCAG 3′) on the CFX96 (Bio-Rad Laboratories) thermal cycler was performed. The amplification process consisted of the following steps: initial activation of polymerase at 98 °C for 3 min; cycling repeated 40 times, which included DNA denaturation at 95 °C for 10 s, annealing and synthesis at 60 °C for 30 s, followed by fluorescence acquisition; and finally, melting curve analysis conducted between 65 °C and 95 °C. The viral loads of HHV-1 in the samples were assessed against the VC using the relative quantity (ΔCq) method, analyzed through CFX Manager™ Dx Software (version 3.1.3090.1022) from Bio-Rad Laboratories [90]. To determine the sensitivity of the qPCR, dilutions of the HHV-1 virus DNA isolate were prepared at 10-, 100-, and 1000-fold and subsequently analyzed.

5. Conclusions

The results of the conducted investigations showed that all plant extracts demonstrated biological activity. Regarding radical scavenging capacity (RSC), the measured data indicated strong (elderberry) or very strong (linden) values.

The mode of obtaining the extracts, which resulted in RSC and EAC, was crucial only for elderflower extracts, while results for all linden extracts were comparable. Similar discrepancies for the plants were observed for the EAC parameter. As for linden flower extracts, they showed close EAC values, whereas elderberry flower extracts differed.

Regarding the total polyphenols content (TPC) for both extracts and total flavonoid content (TFC) of linden extracts, the results confirmed that advanced modes of extraction (MSPD, ASE, and UAE) resulted in higher values of the above-mentioned parameters compared to conventional (Soxhlet) modes.

With respect to the analysis of the extracts, the results confirmed the literature data, and around 40 compounds, i.e., flavonoids, phenolic acids, and their derivatives, were identified in all extracts obtained by various methods for both investigated plants. Interestingly, elderberry extracts contained more phenolic acids and their derivatives than did the linden extracts.

Rutin was the main identified compound in both extracts and across all extraction modes. Considering phenolic acids, chlorogenic acid was the main component from this group in elderberry extracts, while protocatechuic acid was the main component in linden and Soxhlet, MSPD and ASE extracts. Meanwhile, higher amounts of chlorogenic acid were determined for UAE extract.

The elderberry extracts contained significantly higher amounts (concentration) of the investigated compounds from flavonoids and phenolic acids. Application of the advanced extraction methods (MSPD, ASE, and UAE) resulted in a higher level of the examined compounds.

Both plant extracts (elderberry and linden) are known for their anti-inflammatory activity in the upper respiratory tract. However, in our tests, the antiviral activity against HRV-14, one of the most common causes of the common cold, was insignificant. A similar effect was observed against other tested RNA viruses, CVB3. Still, we observed that antiviral activity against HHV-1 was more potent in the case of linden than elderflower. In contrast the elderflower extract exerted a noticeable antiherpesviral effect at 250 µg/mL, decreasing the HHV-1-induced CPE, reducing the viral infectious titer by 1.69 log and viral load by 1.35 log. Only a minor effect on the infectious titer of CVB3 or HRV-14 was observed.

The cytotoxic influence of elderberry flower and linden flower extracts on VERO, FaDu, H1HeLa, and RKO cell lines was examined. Cytotoxicity assays showed that elderflower extract was not toxic to VERO cells and exhibited only minor toxicity toward cancer-derived cells. Conversely, linden flower extract was weakly cytotoxic to VERO cells but exhibited moderate cytotoxicity across all cancer-derived cells.