In Search of High-Yielding and Single-Compound-Yielding Plants: New Sources of Pharmaceutically Important Saponins from the Primulaceae Family.

So far, only a few primrose species have been analyzed regarding their saponin composition and content. Moreover, the roots of only two of them are defined by the European Union (EU) Pharmacopoeia monograph and commercially utilized by the pharmaceutical industry. Thus, this study intended to find some new sources of main triterpene saponins from Primulae radix, namely primulasaponins I and II together with the closely related sakurasosaponin. Using isolated standards, UHPLC-ESI-HRMS served to assess over 155 Primulaceae members qualitatively and quantitatively. Nine examples of plants accumulating over 5% of primulasaponin I in their roots were found. Among them, in one case, it was found as the almost sole secondary metabolite with the concentration of 15-20% (Primula grandis L.). A reasonable content of primulasaponin II was found to be typical for Primula vulgaris Huds. and P. megaseifolia Boiss. & Bal. The sakurasosaponin level was found in seven species to exceed 5%. The finding of new, single and rich sources of the abovementioned biomolecules among species that were never analyzed phytochemically is important for future research and economic benefit. The chemotaxonomic significance of the occurrence of these three saponins in Primulaceae is discussed.

1. Introduction

Primulaceae Batsch. is a large family covering many perennial and herbaceous plants. Members of this family are widely distributed in the Northern Hemisphere, particularly in the meadows and rocky valleys of the Alps, Apennines, Pyrenees, Himalayas and North American Cordillera but also in dry regions of Kazakhstan, Iran and Turkey. Today, some of the primroses are known exclusively from herbaria specimens while several exist as endangered endemics. Since the time of Charles Darwin and A. K. Bulley, primulaceous plants have been a point of interest for botanists and alpine enthusiasts as well as for the gardeners who help to maintain threatened species and discover new ones [1,2,3,4,5]. Thanks to them, over 200 Primulaceae-related species remain in the plant trade apart from decorative varieties. Due to these and the fact that most Primula species possess partly recognized chemical composition [6,7,8], they provide an interesting field for phytochemical research.

According to the European Pharmacopoeia, the official primrose root (Primulae radix) may be obtained from Primula veris L. or P. elatior (L.) Hill. and is well known as an efficient secretolytic and expectorant drug [9,10,11]. It was widely introduced to European academic medicine as a surrogate of the American Senega root during WWI. It is still popular as a component of simple and complex pharmaceutical formulations [9]. It has been proven that the main saponins and active components of the official drug are primulasaponins: I (=primulasaponin A, PSI) and II (PSII), sometimes mentioned as primula acids due to the acidic character of their glycone part [9,12].

Primula sieboldii E. Morren is common in the Far East region (Japan, Amurland, Manchuria and Korea). The phytochemical value of this plant is connected with the triterpenoid biosynthesis yielding almost a single saponin, sakurasosaponin (SSI) [13], a close structural relative of PSI and PSII (Figure 1). However, ethnomedicinal applications of P. sieboldii to treat cough and bronchitis are not widely known [14].

Figure 1

Structures of primulasaponins I, II and sakurasosaponin (PSI, PSII and SSI, respectively).

Besides the well-known activities of saponins such as expectorant or vasoprotectant, an increasing number of researchers are interested in the evaluation of their promising positive interactions with, for example, chemotherapeutics [15,16,17]. Their usage in pharmacy, cosmetology and the food industry as natural and efficient emulsifiers or foaming agents is also desirable [18,19,20].

The glycosides and glycoside-esters of oleanane type are among the most widely distributed groups of saponins. Despite their wide occurrence, primulasaponins I, II and sakurasosaponin were previously not the object of any clinical study. However, they belong to 13,28-epoxyoleanane saponins that are of particular medical interest. Epoxidized oleanane derivatives were found to be active as enzyme inhibitors [21], anti-mycobacterial and anti-protozoan compounds [22,23] and selectively cytotoxic molecules [24,25]. It was demonstrated that the partially deglycosylated metabolites of long-chain 13,28-epoxyoleanane saponins also display antitumor activities at a level similar to their prodrug digested by intestinal flora [26].

Primulasaponins I and II, as well as sakurasosaponin, may serve as a readily available model compound for research (including structural modifications) in the mentioned areas of medicine. Particularly interesting are aldehydes such as ardisiacrispin B that can display, for example, cytotoxic effects in multi-factorial drug-resistant cancer cells [25,27]. The researchers know that the problem of the unavailability of large amounts of a single specialized metabolite is usually a hurdle in semi-synthesis. The number of unwanted by-products increases dramatically with a decrease in the purity of the substrate. Herewith, the use of single-metabolite yielding plants could be a strategy to work around this problem.

High-yielding sources of specialized natural compounds are not very frequent. The primary phytochemical education teaches that such compounds are usually found in concentrations that do not exceed 3–5% of the dry mass of plant material. Positive exceptions include quinine (>10% [28]) and some new sources of theobromine (>6%, Camelia ptilophylla [29]). Among phenylpropanoids, eugenol reaches 10–12% in cloves (at the level of >80% in essential oil). The highest yielding non-alkaloid substances usually occur in plants as complex mixtures (e.g., ~10% of tannins mixture in oak galls, ~6% of saponins mixture in horse-chestnut seeds). The LC-MS proven concentrations of single saponins in the Primulaceae family are usually about 2–6% [7,30].

The pharmaceutical industry prefers single, pure and well-defined biomolecules to serve as the standards. The most readily available compounds or the most active ones are used for the quality determination of raw herbal drugs as well as the reference for clinical studies. Single compounds can be conveniently evaluated, packed and dosed. On the other hand, the purification of natural compounds is usually bothersome and expensive, especially in the complex group of saponins. The lack of broad introduction of modern and rapid strategies for saponins’ determination results in perpetuating semi-quantitative, non-selective methods based on measurements of simple physicochemical properties (e.g., the new monograph included in European Pharmacopoeia (2.8.24) describes the “foam test” to evaluate the quality of saponin drugs [31]).

The objective of this study was to evaluate the distribution and average concentrations of the abovementioned saponins in commercially available Primulaceae species by the use of a cost-effective UHPLC-HRMS method [32]. The second target of this study was to find any species producing the main active compounds in significant quantities preferably as sole substances. The final goal was to identify primulas that could be further used as substitutes for cowslip and oxlip in medical use or utilized as a raw material to efficiently produce significant quantities of selected saponin compounds for commercial purposes or pharmacological deep-research needs.

2. Materials and Methods

2.1. Chemicals

LC-MS grade solvents were purchased from Merck (Darmstadt, Germany) and Sigma-Aldrich (St. Louis, MO, USA) while those of analytical grade were from Chempur (Piekary Śląskie, Poland) and POCh (Lublin, Poland).

2.2. Plant Material

The roots of the Primulaceae representatives were obtained from various botanical collections and nurseries. A total amount of 157 taxa were gathered. Of 111 Primula species and varieties, 60 belonged to Aleuritia, 11 to Auganthus, 30 to Auriculastrum and 10 to Primula subgenera. Additionally, the following primrose relatives were collected: 24 Androsace members together with 5 Cortusa, 2 Dionysia, 2 Hottonia, 2 Lysimachia, 2 Omphalogramma, 6 Soldanella and 3 Vitaliana taxa. Some were repeated in consecutive years. Plants were authenticated [1,2,3,4,33] and documented by photography by the author (M.W.). The plants were harvested in the late summer stage of growth except for lysimachias and hottonias (spring/summer). As a comparison, four trade samples of primrose root deposited in the collection of the Department of Pharmacognosy and Herbal Drugs were used.

The roots were carefully washed, separated from rhizomes and leaves and dried at room temperature in the shade. Vouchers of plant substances were deposited in the Department of Pharmacognosy and Herbal Drugs of Wroclaw Medical University. The precise list of all plant samples and their origin (donators) is combined in Appendix A Table A1.

Table A1

List of Primulaceae species used for screening in this study.

Taxon (a)	Acronym	Taxon (b)	Acronym
genus Primula L.		sg. Aleuritia, sct. Minutissimae
sg. Aleuritia, sct. Aleuritia, ssct. Aleuritia		Primula primulina	PPRI_P_2014
Primula halleri	PHAL_P_2013, PHAL_K_2016	sg. Aleuritia, sct. Muscarioides
Primula scandinavica	PSCA_B_2013	Primula concholoba	PCON_K_2014
Primula scotica	PSCO_K_2016	Primula muscarioides	PMUS_P_2013
sg. Aleuritia, sct. Aleuritia, ssct. Algida		Primula vialii	PVIA_B_2013
Primula algida	PALG_K_2014	sg. Aleuritia, sct. Oreophlomis
Primula darialica	PDAR_K_2014	Primula auriculata	PAUA_K_2015
sg. Aleuritia, sct. Armerina		Primula luteola	PLUT_K_2015
Primula fasciculata	PFAS_K_2014	Primula rosea ‘Gigas’	PROG_B_2013
Primula involucrata	PINV_P_2015	Primula warschenewskiana	PWAR_K_2014
Primula munroi ssp. yargongensis (->P. involucrata in [1])	PMUY_P_2015	sg. Aleuritia, sct. Petiolares, ssct. Edgeworthii
Primula yargongensis (->P. involucrata in [1])	PYAR_P_2015	Primula moupinensis	PMOU_GP_2014
Primula zambalensis	PZAM_K_2014	Primula moupinensis ssp. barkamensis	PMOB_K_2015
sg. Aleuritia, sct. Capitatae		sg. Aleuritia, sct. Petiolares, ssct. Griffithii
Primula capitata ssp. mooreana	PCAM_B_2013	Primula calderiana ssp. calderiana	PCAL_K_2014
Primula glomerata	PGLO_K_2014	Primula tanneri ssp. nepalensis	PTNN_K_2014
sg. Aleuritia, sct. Crystalophlomis		sg. Aleuritia, sct. Petiolares, ssct. Petiolares
Primula purdomii	PPUR_P_2014	Primula boothi var. repens	PBOR_K_2014
sg. Aleuritia, sct. Crystalophlomis, ssct. Crystalophlomis		sg. Aleuritia, sct. Petiolares, ssct. Sonchifolia
Primula chionantha	PCHI_K_2014	Primula sonchifolia ssp. sonchifolia	PSON_K_2014
Primula chionantha ssp. chionantha	PCHC_K_2015	sg. Aleuritia, sct. Proliferae
Primula chionantha ssp. sinopurpurea	PCHS_K_2015	Primula beesiana	PBES_B_2013
Primula graminifolia (->P. chionantha in [1])	PGRM_P_2015	Primula bulleyana	PBUL_B_2013
Primula longipetiolata	PLON_K_2014	Primula japonica	PJAP_B_2013
Primula macrophylla	PMAC_K_2015	Primula prolifera	PPRO_K_2015
Primula orbicularis	PORB_K_2015, PORB_P_2015	Primula wilsonii var. anisodora	PWIA_K_2015
sg. Aleuritia, sct. Crystalophlomis, ssct. Maximowiczii		sg. Aleuritia, sct. Pulchella
Primula maximowiczii var. maximowiczii	PMAX_K_2014	Primula pulchella	PPUL_K_2015, PPUL_P_2015
Primula tangutica	PTAN_P_2013	Primula sharmae	PSHA_K_2016
Primula woodwardii	PWOD_P_2013, PWOD_P_2014	Primula stenocalyx	PSTE_P_2013, PSTE_P_2014
sg. Aleuritia, sct. Davidii		sg. Aleuritia, sct. Sikkimensis
Primula bergenioides	PBER_K_2015	Primula firmipes	PFIR_K_2014
Primula ovaliifolia	POVA_K_2015	Primula florindae	PFLO_B_2013
sg. Aleuritia, sct. Denticulata		Primula florindae, red flowered	PFLR_K_2015
Primula denticulata ‘Alba’	PDEA_B_2013	Primula ioessa	PIOE_P_2013
Primula denticulata	PDEN_K_2016	Primula sikkimensis	PSIK_K_2014
Primula monticola	PMON_K_2014, PMON_K_2015	Primula waltonii	PWAL_K_2014
Taxon (c)	Acronym	Taxon (d)	Acronym
sg. Aleuritia, sct. Soldanelloides		sg. Auriculastrum, sct. Auricula, ssct. Erythrodosum
Primula reidii	PREI_K_2016	Primula daoensis	PDAO_K_2014
Primula reidii var. williamsii ‘Alba’	PREW_F_2015	Primula hirsuta	PHIR_P_2013
sg. Aleuritia, sct. Yunnanensis		sg. Auriculastrum, sct. Auricula, ssct. Rhopsidium
Primula florida (P. blinii)	PBLI_P_2013, PBLI_P_2014	Primula allionii	PALL_F_2015
Primula rupicola	PRUP_K_2016	Primula integrifolia	PINT_P_2013
sg. Auganthus, sct. Bullatae		Primula tyrolensis	PTYR_TK_2015
Primula bullata var. rufa	PBUR_K_2014	sg. Auriculastrum, sct. Auricula, hybrids
Primula forrestii	PFOR_K_2014	Primula venusta	PVEN_B_2013
sg. Auganthus, sct. Cortusoides, ssct. Cortusoides		Primula allioni × Primula villosa	PXAV_P_2013
Primula cortusoides	PCOR_B_2013	sg. Auriculastrum, sct. Cuneifolia
Primula polyneura	PPOL_K_2015	Primula cuneifolia	PCUN_K_2016
Primula sieboldii	PSIE_K_2014	Primula cuneifolia ssp. heterodonta	PCUH_K_2016
sg. Auganthus, sct. Cortusoides, ssct. Geraniifolia		sg. Auriculastrum, sct. Dodecatheon
Primula heucheriifolia	PHEU_K_2015	Primula austrofrigida	PAUS_K_2014
Primula jesoana	PJES_K_2014	Primula clevelandii	PCLE_K_2016
Primula kisoana	PKIS_K_2015	Primula conjugens	PCJG_K_2014
Primula palmata	PPAL_K_2014	Primula jeffreyi	PJEF_K_2015, PJEF_K_2016
sg. Auganthus, sct. Obconicolisteri		Primula latiloba	PLLB_K_2014
Primula obconica	POBC_K_2014	Primula meadia	PMEA_K_2014
sg. Auganthus, sct. Reinii		Primula pauciflora	PPAU_K_2014
Primula takedana	PTAK_E_2014	Primula tetrandra	PTET_K_2014
sg. Auriculastrum, sct. Amethystina		sg. Auriculastrum, sct. Parryi
Primula amethystina var. brevifolia	PAMB_K_2015	Primula angustifolia	PANG_K_2016
sg. Auriculastrum, sct. Auricula, ssct. Arthritica		Primula parryi	PPAR_K_2014
Primula glaucescens ssp. longobarda	PGLL_P_2013, PGLL_P_2014	Primula rusbyi	PRUS_K_2014
Primula spectabilis	PSPE_P_2013	sg. Primula, sct. Primula
sg. Auriculastrum, sct. Auricula, ssct. Auricula		Primula elatior	PELA_K_2014
Primula auricula	PAUR_B_2013	Primula elatior var. amoena	PAMO_P_2014
sg. Auriculastrum, sct. Auricula, ssct. Brevibracteatum		Primula juliae	PJUL_B_2013
Primula carniolica	PCAR_K_2014, PCAR_K_2015	Primula megaseifolia	PMEG_K_2015
Primula latifolia ‘Alba’	PLAA_P_2013	Primula veris (syn. P. officinalis)	PVRS_K_2014
Primula marginata	PMGN_P_2013	Primula vulgaris	PVUL_K_2014
sg. Auriculastrum, sct. Auricula, ssct. Chamaecallis		sg. Primula, sct. Sredinskya
Primula minima	PMIN_I_2014	Primula grandis	PGRA_K_2014, PGRA_K_2016
Primula minima f. niveum	PNIV_P_2014	sg. Primula, hybrids
sg. Auriculastrum, sct. Auricula, ssct. Cyanaster		Primula margotae ‘Garryarde Guinevere’	PMAG_B_2013
Primula glutinosa	PGLU_P_2013	sg. Sphondyllia
		Primula × kewensis	PXKE_K_2016
		Primula verticillata	PVRT_K_2014
Taxon (e)	Acronym	Taxon (f)	Acronym
genus Androsace L.		sct. Aizodium
sct. Andraspis		Androsace bulleyana	ABUL_F_2015
Androsace albana	AALB_K_2019	genus Cortusa L.
sct. Aretia, ssct. Aretia		Cortusa matthioli	CMAT_P_2013
Androsace cylindrica	ACYL_K_2015	Cortusa matthioli ssp. matthioli	CMAT_K_2015
Androsace lehmannii	ALEH_P_2014	Cortusa matthioli ssp. caucasica	CCAU_K_2015, CCAU_K_2016
Androsace mathildae	AMTH_TK_2015	Cortusa matthioli ssp. sachalinensis	CSAC_P_2013, CSAC_K_2015
sct. Aretia, ssct. Dicranothrix		Cortusa matthioli ssp. turkestanica	CTUR_P_2013
Androsace lacteal	ALAC_TK_2015	genus Dionysia Fenzl.
Androsace laggeri (= A. carnea var. laggeri)	ACAL_P_2014	Dionysia khatamii	DKHA_F_2015
Androsace obtusifolia	AOBT_P_2014	Dionysia zschummelii	DZSH_F_2015
sct. Chamaejasme, ssct. Hookerianae		genus Hottonia L.
Androsace limprichtii	ALIM_P_2014	Hottonia inflata hb	HOIN_MO_2014
sct. Chamaejasme, ssct. Mucronifoliae		Hottonia palustris hb	HOPA_PG_2014
Androsace mariae var. tibetica	AMAT_P_2014	genus Lysimachia L.
Androsace mucronifolia	AMUC_TK_2015	Lysimachia nummularia ‘Gold’	LYNG_PG_2014
Androsace sempervivoides	ASPV_B_2015	Lysimachia thyrsiflora	LYTH_MO_2014
sct. Chamaejasme, ssct. Strigillosae		genus Omphalogramma (Franch.) Franch.
Androsace spinulifera	ASPI_P_2015	Omphalogramma delavayi	ODEL_PO_2019
Androsace strigillosa	ASTR_P_2014	Omphalogramma tibeticum	OTIB_K_2019
sct. Chamaejasme, ssct. Sublanatae		genus Soldanella L.
Androsace adenocephala	AADE_F_2015	sct. Crateriflorae
Androsace nortonii	ANOR_P_2014	Soldanella alpina	SALP_K_2016
sct. Chamaejasme, ssct. Villosae, series Chamaejasmoidae		Soldanella carpatica	SCAR_K_2014
Androsace brachystegia	ABRA_P_2014	Soldanella cyanaster	SCYA_K_2014
Androsace chamaejasme ssp. carinata	ACHC_P_2014	Soldanella dimoniei	SDIM_K_2014
Androsace zambalensis	AZAM_P_2014	Soldanella villosa	SVIL_K_2014
sct. Chamaejasme, ssct. Villosae, series Euvillosae		sct. Tubiflorae
Androsace dasyphylla	ADAS_P_2014	Soldanella minima	SMIN_F_2015, SMIN_TK_2015
Androsace robusta ssp. purpurea	AROP_P_2014	genus Vitaliana Sesl.
Androsace sarmentosa	ASAR_K_2015	Vitaliana primuliflora	VPRI_B_2015
sct. Pseudoprimula		Vitaliana primuliflora ssp. assoana	VPAS_TK_2015
Androsace elatior	AELA_F_2015	Vitaliana primuliflora ssp. praetutiana	VPPR_B_2015, VPPR_B_2017
sct. Douglasia
Androsace montana (= Douglasia montana)	AMON_F_2015

—herb was used instead of roots; sg.—subgenus, sct.—section, ssct.—subsection, ssp.—subspecies, var.—variety. The meaning of acronyms used to avoid long plant names in storage and during research is as follows: First letter—genus, three following letters—taxon (species; optionally together with variety), next separated letter or two—donator abbreviation and year of collection in the end. Donators abbreviations: B—Bergenia, Nursery, Paweł Weinar, Kokotów, Poland; E—Edrom, Nursery, Terry and Cath Hunt, Coldingham, UK; F—Floralpin, Nursery, Frank Schmidt, Waldenbuch, Germany; G—Private Collection, Gunhild and Thorkild Poulsen, Aldershvile, Denmark; I—Alpine Garden Mt. Patscherkofel, Institute of Botany, Innsbruck University, Peter Daniel Schlorhaufer, Innsbruck, Austria; K—Kevock Garden, Nursery, Stella and David Rankin, Lasswade, UK; MO—Mayla Ogrody, Nursery, Dawid Stefaniuk, Siedlakowice, Poland; P—Josef and Bohumila Plocar, Nursery, Švihov, Czech Republic; PG—Planta Garden, Krzysztof Sternal, Dobra, Poland; PO—Pottertons, Nursery, Jackie and Robert Potterton, Nettleton, UK; TK—Private Collection, Tomasz Kubala, Poland.

2.3. Preparation of Samples

For HPLC evaluation, each sample was precisely weighed (100 mg), transferred into a sealed vial and extracted with 2 mL of 70% MeOH for 15 minutes in an ultrasonic bath (Bandelin, Berlin, Germany) at 25 °C and 50% of the power. 1 mL of each extract was filtered using 0.22 μm PTFE single-use syringe filters (Merck-Millipore, Darmstadt, Germany), diluted 100 times with acetonitrile/water mixture (1/1, v/v; LC-MS class) and stored at 4 °C before analysis [32].

2.4. Preparation of Standards

The standards of primulasaponin I and II were isolated from the commercial primrose root (complies with Pharmacopoeia requirements; Galke GmbH, Gittelde, Germany). The procedure was as follows: 50 g of the root was homogenized (A11; IKA, Königswinter, Germany) and extracted twice with 0.5 L of 70% methanol. The resulting extract was concentrated using a rotary evaporator (Büchi, Flawil, Switzerland) at 40 °C, diluted with water, applied on a Diaion HP-20 SPE column (Sigma-Aldrich, St. Louis, MO, USA) and eluted with water/methanol of an increasing gradient. Fractions eluted with 50% MeOH, containing a mixture of saponins (702 mg; 1.4% of starting material), were combined, concentrated and subjected to LC on silica (Merck, Darmstadt, Germany) in ethyl acetate/acetic acid/water (5/1/1, v/v/v) as the mobile phase [16]. Later, the eluates were collected according to their TLC profiles. For TLC analysis, the same solvent system was used as for LC on silica. Finally, the combined eluates were purified by solid-phase extraction (SPE) on the Chromabond C18 column (Macherey-Nagel, Düren, Germany) giving as a result pure saponins PSI (33 mg) and PSII (73 mg).

The standard of sakurasosaponin was isolated from commercial P. sieboldii roots (Kevock Garden, Lasswade, UK). 1.60 g of the powdered root was extracted by 1-day maceration with 70% methanol. The extract was diluted with water and subjected to SPE on the Chromabond C18 column. The fraction eluted with 70% methanol was collected, evaporated to dryness (60 mg) and finally purified by preparative TLC on an RP-18 W plate (Macherey-Nagel, Düren, Germany), giving as a result 25 mg of SSI.

The saponins were stored as powders at 4 °C before the NMR and HRMS analysis. An aliquot of isolated saponins was then subjected to HRMS, 1H- and 13C NMR spectroscopy, including 2D experiments and compared positively against assignments from the literature [12,13].

The stock solutions containing 2.0 mg/mL, 2.4 mg/mL and 2.2 mg/mL of PSI, PSII and SSI respectively were prepared in acetonitrile/water (1/1, v/v) and stored at 4 °C before the analytical use.

2.5. General NMR and HRMS Experimental Procedures

1H, 13C and 2D NMR spectra were obtained on a Bruker Avance 300 NMR spectrometer (Bruker BioSpin, Rheinstetten, Germany), operating at 300 MHz and 75 MHz respectively at 300 K, using standard pulse programs and methanol-d4 (Armar AR, Döttingen, Switzerland) as well as DMSO-d6 (Armar AR) as the solvent. HRMS measurements of isolated compounds were conducted on the ESI-qTOF Maxis instrument (Bruker Daltonics, Bremen, Germany) while for UHPLC-HRMS detection ESI-qTOF Compact (Bruker Daltonics) was used. The instruments were operated in negative mode and calibrated with the Tunemix mixture (Bruker Daltonics) with m/z standard deviation below 0.5 ppm. In the case of the UHPLC-MS measurements a calibration segment was introduced at the beginning of every single run. The mass accuracy of saponin standards was within 3 ppm. The analysis of the obtained mass spectra was carried out using Data Analysis and Quant Analysis (Bruker Daltonics) software. The main instrumental parameters were as follows: scan range 50–2200 m/z, nebulizer pressure 1.5 bar, dry gas (N2) 7.0 L/min, temperature 200 °C, capillary voltage 2.2 kV, ion energy 5 eV, collision energy 10 eV, low mass set at 200 m/z. The samples were dissolved in acetonitrile/water (1:1, v/v) containing 0.1% HCOOH.

2.6. Purity of Standards

The purity of isolated standards of saponins was ascertained using a 1H-qNMR method with maleic acid (Fluka, Buchs, Switzerland) as the internal standard with declared 99.94% content. Precisely weighed samples of each saponin were mixed with the precisely weighed standard and dissolved in 1 mL of methanol-d4 (Armar AR) in two repetitions. 600 μL of each was used for analysis in a 5 mm tube on 700 MHz Bruker apparatus equipped with TXI CP (Bruker BioSpin). The analysis was conducted according to the procedure of the Polish Center for Technology Development, based on [34].

2.7. LC-MS Analytical Conditions

The Thermo Scientific UHPLC Ultimate 3000 apparatus (Thermo Fisher Scientific, Waltham, MA, USA) consisted of an LPG-3400RS quaternary pump with a vacuum degasser, a WPS-3000RS autosampler and a TCC-3000SD column oven. The ESI-qTOF Compact (Bruker Daltonics, Bremen, Germany) was connected as the MS detector. The separation was achieved on a Kinetex C-18 column (150 × 2.1 mm) of 2.6 μm particle size, core-shell type (Phenomenex, Torrance, CA, USA). The gradient elution system consisted of 0.1% HCOOH in water (mobile phase A) and 0.1% HCOOH in acetonitrile (mobile phase B). At the flow rate of 0.3 mL/min, the following elution program was used: 0→1 min (2%→30% B), 1→31 min (30%→60% B), 31→31.5 min (60%→100% B), 31.5→35.5 min (100% B). The column was equilibrated for 7 min before the next analysis. Blanks were added after each run to avoid any sample carryover. All analyses were carried out isothermally at 30 °C. The injection volume for samples and standard solutions was 5 μl. Each analysis was calibrated in the first segment of analysis and performed in duplicate. The method was as in Reference [32].

2.8. Validation of the Analytical Method

The UHPLC–MS assay was validated with respect to the specificity, linearity, precision, accuracy and stability.

2.8.1. Specificity

Concentrations of the saponins were calculated using areas of peaks from EIC chromatograms extracted for 1103.564 ± 0.01 m/z (PSI, 15.67 min ± 0.50 min), 1235.606 ± 0.01 m/z (PSII, 14.07 min ± 0.50 min) and for 1249.622 ± 0.01 m/z (SSI, 15.10 min ± 0.50 min). The concentrations were measured in triplicate. Each sample was evaluated manually for the absence of closely eluting interferences (within ± 0.50 min of the nominal retention time) using the abovementioned extraction and UHPLC-MS conditions. Because every taxon possessed different patterns of interfering compounds, validation parameters were analyzed using standards.

2.8.2. Linearity, Range and Limits of Analysis

The linearity was achieved by assaying a series of the mixed standard solution (PS I, PS II, SSI; consisting of 20 analytic points for each analyte; in a range of 0.050-250 µg/mL) in duplicate over three consecutive days. The efficient calibration equation to assure the calibration curve fit in the whole range of detector response as well as to include different saponin content in samples was proposed to be:

Using Statistica 12.5 software (Tulsa, OK, USA), the variable parameters of equation (n, A, B) together with coefficients of correlation (r) and coefficients of determination (r2) were calculated for each curve. The limits of detection (LOD) and quantification (LOQ) were evaluated by the signal-to-noise approach with the use of the lowest concentration. All results are given in Table 1.

Table 1

Parameters of calibration Equation (1), for primulasaponins I (PSI), II (PSII) and sakurasosaponin (SSI), together with values of r, r2, LOD and LOQ.

Standard	n ± SD	A ± SD	B ± SD	r	r2	LOD [ng/mL]	LOQ [ng/mL]
PSI	0.890 ± 0.072	4,574,451 ± 83,044	24.73 ± 1.26	0.9998	0.9996	6.7	20.3
PSII	0.776 ± 0.110	3,274,550 ± 100,794	23.41 ± 2.24	0.9994	0.9988	6.4	19.4
SSI	0.943 ± 0.110	3,951,182 ± 86,066	20.29 ± 1.62	0.9995	0.9990	7.4	22.5

2.8.3. Precision, Accuracy and Stability

Precision, accuracy and stability were determined using standards at low, medium and high concentration levels (1, 10 and 100 µg/mL for all saponins, corresponding to 0.2%, 2.0% and 20% of root dry mass). All concentration levels were measured in triplicate. The precision and accuracy were tested once a day and repeated for three consecutive days. Intra- and interday precision were defined as the relative standard deviation (RSD), while the accuracy was determined by the relative error (RE %). The stability of analyzed saponins was assessed using standards stored at room temperature for 14 days in three repetitions. All results are gathered in Table 2.

Table 2

Validation parameters for UHPLC-MS assay for primulasaponins I (PSI), II (PSII) and sakurasosaponin (SSI).

Standard	Repeatability (RSD)/(Intra-Day Precision)			Intermediate Precision (RSD)/(Inter-Day Precision)
Level	1 µg/mL	10 µg/mL	100 µg/mL	1 µg/mL	10 µg/mL	100 µg/mL
PSI	2.3	2.0	0.4	5.6	1.6	1.2
PSII	4.0	1.7	0.7	6.1	1.5	1.1
SSI	2.8	1.2	1.0	3.0	1.2	0.9
Standard	Accuracy (%RE)			Stability (RSD)/(14 days)
Level	1 µg/mL	10 µg/mL	100 µg/mL	1 µg/mL	10 µg/mL	100 µg/mL
PSI	+21.2	+3.5	+0.1	3.2	3.3	2.6
PSII	+30.6	+5.5	−0.7	10.1	1.8	4.2
SSI	+15.4	+4.8	+0.4	6.8	2.2	1.4

2.9. Clustering of Saponin Concentrations

Using Statistica 12.5 software (Tulsa, OK, USA), clustering analysis (CA) was performed to group the samples according to the content of each saponin. The results are presented at Figure 2 and Table 3.

Figure 2

Visual representation of clustering analysis (CA) of analyzed Primulaceae taxa; based on the content of primulasaponins I, II and sakurasosaponin (PSI, PSII and SSI).

Table 3

List of outstanding groups resulting from clustering analysis; corresponds with Figure 2.

Sample Acronym	Group	Sample Acronym	Group	Sample Acronym	Group	Sample Acronym	Group
PFOR_K_2014	1	HOIN_MO_2014	2	PCHS_K_2015	3	PCHI_K_2014	6
POBC_K_2014	1	PZAM_K_2014	2	PAUA_K_2015	3	PCHC_K_2015	6
PTAK_E_2014	1	PCOR_B_2013	2	PROG_B_2013	3	PLON_K_2014	6
PCLE_K_2016	1	PPOL_K_2015	2	PCJG_K_2014	3	PMAC_K_2015	6
PTET_K_2014	1	PSIE_K_2014	2	PJEF_K_2015	3	PWAR_K_2014	6
PANG_K_2016	1	PHEU_K_2015	2	PELA_K_2014	3	PR2	6
		PPAL_K_2014	2	PMEG_K_2015	3
		PCUN_K_2016	2	PVRS_K_2014	3	PVUL_K_2014	7
		PCUH_K_2016	2	PR3	3	PR1	7
		PAUS_K_2014	2	PR4	3
		PMEA_K_2014	2	PMAG_B_2013	3	PGRA_K_2014	8
		PPAU_K_2014	2			PGRA_K_2016	8

PR1-4—samples of commercial trade Primulae radix, declared to fulfill pharmacopoeial requirements. Background color corresponds with botanical classification: blue—sg. Auganthus, yellow—sg. Auriculastrum, sct. Dodecatheon, green—sg. Aleuritia, sct., Crystallophlomis, pink—sg. Aleuritia, sct. Oreophlomis, grey—sg. Primula, sct. Primula, brown—sg. Primula, sct. Sredinskya, transparent—others.

3. Results

3.1. Method Development

In our study, only species offered year by year in common trade in Europe were collected, trying to obtain at least 2–3 species from each available botanical section or subsection. Finally, 157 taxa were collected to perform the screening (collected in Table A1). The roots of the authenticated plants were sequentially dried, milled, extracted with 70% MeOH, filtered and diluted to obtain samples applicable for the assay. The universal extraction strategy was based on observations during the isolation process: SPE fractions eluted with 70% MeOH from C18 bed were the richest in isolated saponins. Ultrasound-assisted extraction (UAE) is a common strategy for the rapid extraction of both fresh and dry plant material. It was first-choice when small amounts of plant material were available.

Because the reference substances of all mentioned saponins were not simply available, the standards were isolated from pharmacopoeial primrose roots (PSI and PSII) and the roots of Primula sieboldii (SSI). Their structures (Figure 1) were elucidated based on the following HRMS and NMR experiments. High-resolution mass spectrum measured for the investigated creamy amorphous solids in negative ion mode revealed the following main peaks: for PSI [M–H]− at 1103.5643 m/z (calculated for C54H87O23; 1103.5644), for PSII [M–H]– at 1235.6054 m/z (calculated for C59H95O27; 1235.6066) and for SSI [M–H]− at 1249.6241 m/z (calculated for C60H97O27 1249.6223). The formulas generated for detected ions corresponded well to neutral molecules C54H88O23 (PSI), C59H96O27 (PSII) and C60H98O27 (SSI) and the MS/MS fragmentation of the glycone part was the additional proof. All HRMS and MS/MS spectra are attached to the Supplementary Figures S1 and S2. The detailed 1H and 13C NMR spectra, combined with COSY, HSQC and HMBC experiments, finally confirmed the identities of isolated compounds (Supplementary Figures S3–S5) with regard to the literature [12,13]. Table S1, summarizing NMR measurements, is attached to Supplementary Material. The purity level that is needed for the circumspect usage of isolated compounds in the quantitative analysis was calculated by qHNMR as 88.91% (PSI), 78.14% (PSII) and 91.89% (SSI).

Limited by the fact that the PSI, PSII and SSI saponins are not applicable for efficient and reliable HPLC-UV analysis due to the lack of conjugated double bonds except carbonyl groups, we used a previously developed simple UHPLC-ESI-MS analytical method for their precise quantitative evaluation [32]. On its basis, a qualitative method was proposed to rapidly estimate the concentrations. A non-standard curve (1) (Table 1) was fit to cover the response of the detector in a broad range of concentrations (0.050–250 µg/mL, relative to 0.01–50% of root dry mass). The correlation coefficients were all of 0.999. The LOQ values were between 19 and 23 ng/mL.

The full validation of the quantitative method was intentionally not performed at this level of research; however, some data are collected in Table 2 to supply the assay. Values of inter-day and intra-day precision are acceptable, while low accuracy at the lowest level suggested the measurement uncertainty and need to use higher concentrations. The standards seemed to be stable in two weeks at room temperature.

The MS detector was calibrated before each UHPLC analysis to guarantee accuracy. Samples were separated by blank analyses to exclude the possibility of overlapping. The low flow of chromatographic solvents together with the relatively fast gradient program resulted in eco-friendliness. The primarily considered faster way of HRMS quantitative analysis in the “direct injection mode” (without LC) was rejected because of the detection of ions isobaric to PSI and SSI in some of the first chromatograms. These ions were sufficiently separated from analytes and did not affect the area readings.

3.2. Distribution of Saponins

Figure 2 and Table 3 show the result of K-means clustering (CA). Table 4 shows in the form of a heatmap the average concentrations of analyzed saponins in roots of 157 Primulaceae members (> 165 samples). Only the exemplars with the saponin level exceeding 0.20% were presented with percentages (calculated from root dry mass) in the Appendix A, Table A2. The concentrations of PSI, PSII and SSI were under the LOD or the compounds were present in trace amounts in representatives of genera Cortusa L. (also considered as a part of the Primula L. [35]), Dionysia Fenzl., Lysimachia L., Omphalogramma Franch., Soldanella L., Vitaliana Sesl. and in a part of Androsace L.

Table 4

A heatmap of average primulasaponins I and II and sakurasosaponin distribution in Primulaceae family. Highest concentration—red, lowest concentration—green.

Taxon	n	PSI	PSII	SSI	Taxon (Continuation)	n	PSI	PSII	SSI	Taxon (Continuation)	n	PSI	PSII	SSI
genus Primula L., sg. Aleuritia					genus Primula L., sg. Auganthus					genus Androsace L.
sct. Aleuritia, ssct. Aleuritia	3				sct. Bullatae	2				sct. Aizodium	1
sct. Aleuritia, ssct. Algida	2				sct. Cortusoides, ssct. Cortusoides	3				sct. Andraspis	1
sct. Armerina	5				sct. Cortusoides, ssct. Geraniifolia	4				sct. Aretia, ssct. Aretia	3
sct. Capitatae	2				sct. Obconicolisteri	1				sct. Aretia, ssct. Dicranothrix	3
sct. Crystalophlomis	1				sct. Reinii	1				sct. Chamaejasme, ssct. Hookerianae	1
sct. Crystalophlomis, ssct. Crystalophlomis	7				genus Primula L., sg. Auriculastrum					sct. Chamaejasme, ssct. Mucronifoliae	3
sct. Crystalophlomis, ssct. Maximowiczii	3				sct. Amethystina	1				sct. Chamaejasme, ssct. Strigillosae	2
sct. Davidii	2				sct. Auricula, ssct. Arthritica	2				sct. Chamaejasme, ssct. Sublanatae	2
sct. Denticulata	3				sct. Auricula, ssct. Auricula	1				sct. Chamaejasme, ssct. Villosae
sct. Minutissimae	1				sct. Auricula, ssct. Brevibracteatum	3				series Chamaejasmoidae	3
sct. Muscarioides	3				sct. Auricula, ssct. Chamaecallis	2				series Euvillosae	3
sct. Oreophlomis	4				sct. Auricula, ssct. Cyanaster	1				sct. Douglasia	1
sct. Petiolares, ssct. Edgeworthii	2				sct. Auricula, ssct. Erythrodosum	2				sct. Pseudoprimula	1
sct. Petiolares, ssct. Griffithii	2				sct. Auricula, ssct. Rhopsidium	3
sct. Petiolares, ssct. Petiolares	1				sct. Auricula, hybrids	2				genus Cortusa L.	6
sct. Petiolares, ssct. Sonchifolia	1				sct. Cuneifolia	2				genus Dionysia Fenzl.	2
sct. Proliferae	5				sct. Dodecatheon	8				genus Hottonia L.	2
sct. Pulchella	3				sct. Parryi	3				genus Lysimachia L.	2
sct. Sikkimensis	6				genus Primula L., sg. Primula					genus Omphalogramma (Franch.) Franch.	2
sct. Soldanelloides	2				sct. Primula	6				genus Soldanella L.	6
sct. Yunnanensis	2				sct. Sredinskya	1				genus Vitaliana Sesl.	3
genus Primula, sg. Sphondyllia	2				sct. Primula, hybrids	1

—herb was used instead of roots; sg.—subgenus, sct.—section, ssct.—subsection, ssp.—subspecies, var.—variety, n—number of analyzed taxa in genus, section or subsection.

3.2.1. Distribution of Primulasaponins

The presence of both PSI and PSII was established to be typical only for the genus Primula L. and subgenus Primula (0.1–6.2% and 0–3.3% respectively). PSII was not found in considerable amounts elsewhere. Moreover, the second compound was found in significant amounts, exceeding 2% in roots of only two species, P. vulgaris Huds. and P. megaseifolia Boiss. & Bal. Contrary to PSII, the saponin PSI was also present in Primula L., subgenus Aleuritia (Duby) Wendelbo, in sections Crystallophlomis (Rupr.) Federov (0.1–9.5%), Davidii Balf. f. (<1%), Oreophlomis (Rupr.) Federov (1.7–5.2%) and Petiolares Pax (about 1%). Among Proliferae-belonging taxa, only P. japonica reached the level of 1% of PSI. In members of subgenera Auganthus (Link) Wendelbo and Auriculastrum Schott only trace amounts of PSI were found except for Primula parryi A. Grey roots (~1%). Primula grandis L., a single representative of Primula L., section Sredinskya Stein was newly found to concentrate the highest amounts of PSI among all samples (15–20%). After this observation, a similar level of concentration was also found in its remarkably fleshy rhizomes.

Some years ago, the genus Dodecatheon L. was proposed to be included in Primula L. based on genetic analyses [36]. Our findings on PSI concentrations in 7 species from the Dodecatheon L. were generally similar to its relative subgenus Auriculastrum. However, two former Dodecatheon members—P. conjugens (Greene) Mast & Reveal and P. jeffreyi (van Houtte) Mast & Reveal—were quite abundant in PSI saponin (3.9% and 2.9% respectively).

3.2.2. Distribution of Sakurasosaponin

Sakurasosaponin, primarily detected and described only in P. sieboldii E. Morren, was found to be typical for almost all of its analyzed relatives from the subgenus Auganthus (Link) Wendelbo (range of 1.1–9.1%). Its presence in the subgenus Auriculastrum Schott, section Cuneifolia Balf. f. (2.3–3.3%) and former genus Dodecatheon L. (0.7–8.6%) is noteworthy. Particularly high amounts of SSI were newly discovered in P. forrestii Balf. f. (~8%), P. obconica Hance and P. takedana Tatew. (both >8%). P. jesoana Miq was found to be the only species from the subgenus Auganthus in which sakurasosaponin was absent. Relatively small amounts of SSI were detected in both Hottonia specimens (0.6–2.3%) and several representatives of Androsace (0.1–1.3%).

3.2.3. Results of Clustering Analysis

K-means clustering led to the separation of a dataset into eight groups of different numbers (Figure 2). The most numerous group 5 consists of samples containing no one of analyzed compounds at a concentration higher than 0.2%; the second numerous group 4 contains samples with poor concentrations of all three saponins. Group 2 collects the samples with medium amounts of the sakurasosaponin (mainly from subgenus Auganthus—section Cortusoides and subgenus Auriculastrum—sections Cuneifolia and Dodecatheon) while samples containing large amounts of this compound are gathered in group 1 (exemplars from subgenera Auganthus—P. forrestii, P. obconica, P. takedana—and Auriculastrum—P. angustifolia, P. clevelandii and P. tetrandra). Group 3 consists of samples with an medium content of primulasaponin I, groups 6 and 7—high content of this compound (in case of group 7, accompanied by significant amounts of primulasaponin II). Samples very rich in single primulasaponin I from subgenus Primula, section Sredinskya are in group 8. Ascriptions of samples to groups 1–3 and 6–8 are presented in Table 3.

4. Discussion

The lack of high-yielding or single-compound-yielding plant sources of active constituents is a common problem for institutions that develop standardization procedures of herbal drugs and those involved in deep research or clinical studies. In order to improve the quality and the value of medicinally important crops, several attempts are made in different fields. Some try to modify environmental conditions. Others try modern genetic manipulations. The recent trends in this area are also focused on the importance of soil microorganisms and mycorrhizal fungi [37,38]. The old strategy to increase the production of crucial secondary metabolites is to conduct organized selection of better-yielding cultivars, varieties or hybrids. The latter approach may still work successfully since 21st-century knowledge on medicinal plants is still rudimentary.

To find new and abundant sources of primulasaponins I and II and closely related sakurasosaponin, over 155 taxa of Primulaceae members were analyzed by the newly designed universal method. Most of them were not previously mentioned in the literature in the phytochemical context. Limited by the fact that the PSI, PSII and SSI saponins are not applicable for efficient HPLC-UV analysis, a simple UHPLC-ESI-MS analytical method was developed for this purpose. Previously UHPLC-APCI-MS, in comparison with UHPLC-ELSD, was used for PSI and PSII quantification in pharmacopoeial herbal drugs [7]. For assay purposes, the standards of both primulasaponins (PSI and PSII) were isolated from the pharmacopoeial primula root while sakurasosaponin (SSI) was isolated from the roots of P. sieboldii by LC and FC. The identity of compounds was confirmed by comparison of spectrometric and spectroscopic data (HRMS, MS/MS and NMR) with the literature [12,13], while their purity was determined by qHNMR.

According to the literature, the saponin content in Primula species may vary from 2% to 12% but the exact range of expectation is not defined in the monographs [9,10,11]. Our study shows that the roots of non-pharmacopoeial primroses may be an alternative source for PSI isolation, sometimes exceeding the level of 10%. Some of the selected species exhibit vigorous growth and are not very difficult to propagate (P. conjugens, P. chionantha, P. macrophylla, P. vulgaris). The other primarily promiscuous well-growing ones (e.g., from the section Proliferae) do not serve this purpose due to the lack of saponins mentioned above. Primula grandis L., a single representative of the section Sredinskya, should be considered as a particularly valuable source of this saponin. The reasons are as follows: almost complete exclusiveness of PSI in the saponin profile of this source (Figure S1), its unusual concentration exceeding 15–20% of both root and rhizome dry mass, together with significant growth and hardiness of this species.

Sakurasosaponin, which was previously reported only from P. sieboldii, appears to be typical for many Auganthus and Dodecatheon members. For example, P. takedana and P. obconica may yield reasonable amounts of this compound with almost no coexisting saponins. The structure of this compound (SSI) is quite similar to those of PSI and PSII. The above, together with the ethnomedicinal applications of P. sieboldii, should lead soon to the consideration of SSI as a substitute for PSI and PSII or their mixtures (in the form of extracts).

The information presented in the Figure 2 and Table 3 and Table 4 could be valuable as supplementary data for taxonomists analyzing phylogenetic relationships in this family [39,40]. Some previously underestimated species should be recognized as remarkably useful for the pharmaceutical industry.

Frequently, phytochemical studies are limited to a few species and use different methods. Here, by using a unified approach, a reasonable number of plant species was tested. The samples were supplied for seven consecutive years (2013-2019) to cover the maximal number of Primulaceae species available in commercial trade. Attempts were made to sufficiently cover the number of representatives in each section. It allowed us to select some notable cases. In the light of the presented results, ornamental hybrids and varieties of Primula species belonging to the subgenera Primula and Auganthus were selected to continue the systematic exploration. This sizeable comparative study is a good starting point for further research covering seasonal and environmental saponin variation as well as for establishing metabolomic relationships among primroses.

This work presents the first systematic study of three 13,28-epoxy-oleanane-type saponins (PSI, PSII, SSI) in the genus Primula by UHPLC-ESI–MS. Although the mentioned saponins were previously not a subject of any clinical study, the discovery of abundant and single-compound sources of primulasaponins and sakurasosaponin should be a milestone in studies on their activities. As pure substances, they can be used directly or for structural modifications. The area of semisynthesis of new enzyme inhibitors and antibacterial or anti-protozoan agents is fertile. One direction for the application of their derivatives is a study on their cytotoxicity or modulation of action of well-known cytotoxic or antimicrobial agents.

Taxon (a)	Average % of Dry Mass				Taxon (b)	Average % of Dry Mass				Taxon (c)	Average % of Dry Mass
	PSI	PSII	SSI			PSI	PSII	SSI			PSI	PSII	SSI
genus Primula					sg. Auganthus, sct. Bullatae					sg. Auriculastrum, sct. Cuneifolia
sg. Aleuritia, sct. Armerina					P. bullata var. rufa	nd	nd	1.07		P. cuneifolia	Nd	nd	2.31
P. zambalensis	t	nd	2.97		P. forrestii	t	nd	8.00		P. cuneifolia ssp. heterodonta	Nd	nd	3.28
sg. Aleuritia, sct. Crystallophlomis					sg. Auganthus, sct. Cortusoides					sg. Auriculastrum, sct. Dodecatheon
P. chionantha	5.57	nd	nd		P. cortusoides	nd	nd	4.37		P. austrofrigida	T	nd	2.62
P. chionantha ssp. chionantha	6.35	nd	nd		P. heucherifolia	t	nd	2.24		P. clevelandii	t	nd	8.59
P. chionantha ssp. sinopurpurea	4.31	nd	nd		P. kisoana	t	nd	1.68		P. conjugens	3.86	nd	1.26
P. graminifolia	1.22	nd	nd		P. palmata	t	nd	2.13		P. jeffreyi	2.85	nd	1.00
P. longipetiolata	9.48	nd	nd		P. polyneura	t	nd	2.22		P. latiloba	t	nd	0.77
P. macrophylla	7.06	nd	nd		P. sieboldii	t	nd	4.20		P. meadia	t	nd	3.56
P. orbicularis	0.60	nd	nd		sg. Auganthus, sct. Obconicolisteri					P. pauciflora	t	nd	5.08
P. purdomii	t	nd	nd		P. obconica	t	nd	8.59		P. tetrandra	t	nd	6.90
sg. Aleuritia, sct. Davidii					sg. Auganthus, sct. Reinii					sg. Auriculastrum, sct. Parryi
P. bergenioides	0.73	nd	nd		P. takedana	t	nd	9.09		P. angustifolia	0.35	nd	8.60
P. ovaliifolia	t	nd	nd		sg. Auriculastrum, sct. Auricula					P. parryi	1.13	nd	0.44
sg. Aleuritia, sct. Oreophlomis					P. allionii	t	nd	nd		P. rusbyi	t	nd	0.85
P. auriculata	3.24	nd	nd		P. allionii × P. villosa	t	nd	nd		sg. Primula, sct. Primula
P. luteola	1.75	nd	nd		P. auricula	t	nd	nd		P. elatior var. amoena	t	nd	nd
P. rosea ‘Gigas’	4.06	nd	nd		P. carniolica	t	nd	nd		P. elatior	3.30	nd	nd
P. warschenewskiana	5.21	nd	nd		P. glaucescens ssp. longobarda	t	nd	0.48		P. elatior rhiz	4.63	nd	nd
sg. Aleuritia, sct. Petiolares					P. glutinosa	t	nd	nd		P. juliae	0.79	t	nd
P. boothi var. repens	1.04	nd	nd		P. hirsuta	t	nd	nd		P. juliae rhiz	0.21	t	nd
P. moupinensis	t	nd	nd		P. integrifolia	t	nd	nd		P. megaseifolia	3.24	2.65	nd
P. moupinensis ssp. barkamensis	1.17	nd	nd		P. latifolia ‘Alba’	t	nd	nd		P. veris (syn. P. officinalis)	2.84	t	nd
P. sonchifolia ssp. sonchifolia	0.67	nd	nd		P. marginata	t	nd	nd		P. vulgaris	6.23	3.26	nd
sg. Aleuritia, sct. Proliferae					P. minima	t	nd	nd		P. vulgaris rhiz	3.30	2.27	nd
P. besiana	t	nd	nd		P. minima f. niveum	t	nd	nd		sg. Primula, sct. Sredinskya
P. japonica ‘Alba’	1.05	nd	nd		P. tyrolensis	t	nd	nd		P. grandis (PGRA_K_2014)	23.88	nd	nd
P. prolifera	t	nd	nd		P. venusta	t	nd	nd		P. grandis (PGRA_K_2016)	15.72	nd	nd
sg. Aleuritia, sct. Pulchella										P. grandis rhiz (PGRA_K_2016)	14.57	nd	nd
P. pulchella	t	nd	t

Taxon (d)	Average % of Dry Mass
	PSI	PSII	SSI
sg. Primula, hybrids
P. margotae ‘GarryardeGuinevere’	4.80	0.32	nd
P. margotae rhiz	2.21	0.23	nd
sg. Primula, trade samples
Primulae radix 1 (PR1)	8.60	4.90	nd
Primulae radix 2 (PR2)	6.16	t	nd
Primulae radix 3 (PR3)	2.66	t	nd
Primulae radix 4 (PR4)	2.18	t	nd
sg. Sphondylia, hybrids
P. × kewensis	0.33	nd	nd
genus Androsace
sct. Aretia
A. lehmanii	nd	nd	1.12
sct. Chamaejasme
A. adenocephala	nd	nd	t
A. brachystegia	nd	nd	t
A. chamaejasme ssp. carinata	nd	nd	t
A. limprichtii	nd	nd	0.36
A. robusta ssp. purpurea	nd	nd	0.76
A. sarmentosa	nd	nd	1.28
A. strigillosa	t	nd	t
genus Hottonia
H. inflata hb	0.36	nd	2.27
H. palustris hb	0.31	nd	0.62

—herb or —rhizome was used instead of roots; sg.—subgenus, sct.—section, ssp.—subspecies, var.—variety, nd—below LOD, t—between LOQ and 0.2%. Values are averages of triplicates. Standard deviations lower than 5% in each case.