Content of cannabinoids in clonally propagated industrial hemp

ABSTRACT

Cannabis sativa L. is an extremely variable species. Even within the same cultivar plants can significantly differ in the content and profile of cannabinoids. Therefore, the best method for production of uniform plants and standardized raw material is vegetative propagation using clones. The aim of this study was to determine the content of cannabidiolic acid (CBDA), cannabidiol (CBD), Δ⁹-tetrahydrocannabinolic acid (Δ⁹-THCA), Δ⁹-tetrahydrocannabinol (Δ⁹-THC), cannabichromene (CBC), cannabigerol (CBG), and cannabinol (CBN) in clonally propagated plants of industrial hemp. One hundred and thirty-nine plants representing 17 different hemp genotypes were regenerated in vitro, hardened, and grown in a vegetation hall until harvest. Single plants of each accession were analyzed using high-performance liquid chromatography with UV/diode-array detection (HPLC-DAD/UV). The results revealed significant variability in the total cannabinoid content (0.55–5.18% in dry weight) among tested genotypes and within the Epsilon 68 cultivar. The highest content of total CBD (4.410%) was recorded for EPS/40 genotype, while the level of total Δ⁹-THC was below the allowed threshold (0.3%). Therefore, we can conclude that some clonally propagated plants provided reproducible hemp material as a potential source of cannabidiol. The results of this study will be useful for breeding and early selection of hemp genotypes.

摘要

大麻是一个极易变异的物种. 即使在同一栽培品种中，大麻素的含量和分布也可能存在显著差异. 因此，生产统一植物和标准化原料的最佳方法是使用无性繁殖. 本研究旨在测定工业大麻克隆繁殖植物中大麻二酚酸（CBDA）、大麻二酚（CBD）、Δ9-四氢大麻酚酸（Δ9-THCA）、Δ9-THC、大麻色素烯（CBC）、大麻酚（CBG）和大麻酚（CBN）的含量. 代表17种不同大麻基因型的139株植物在体外再生、硬化并在植被大厅中生长，直到收获. 使用具有UV/二极管阵列检测的高效液相色谱法（HPLC-DAD/UV）分析每个登录的单株. 结果显示，在受试基因型之间和受试基因组内，大麻素总含量（干重为0.55–5.18%）存在显著差异. 结果显示，在测试的基因型和Epsilon 68品种中，大麻素总含量存在显著差异（干重为0.55–5.18%）. EPS/40基因型的总CBD含量最高（4.410%），而总Δ9-THC水平低于允许的阈值（0.3%）. 因此，我们可以得出结论，一些克隆繁殖的植物提供了一种可再生的大麻材料，作为大麻二酚的潜在来源。研究结果对大麻基因型的选育和早期选择具有一定的指导意义.

KEYWORDS:

• Cannabis sativa L. in vitro cultures
• genotype
• CBD
• Δ⁹-THC
• Δ9-THC
• variability

关键词:

• 体外培养
• 基因型
• Δ9-THC
• 可变性

Introduction

Hemp (Cannabis sativa L.) is a well-known multifunctional crop, with great medical and economic importance. The global legal cannabis market was estimated to be worth USD 10.81 billion in 2022 and is expected to reach USD 70.56 billion by 2031 (Global Medical Cannabis Market Report and Forecast 2023–2023–2031). In Europe, medical cannabis constitutes the majority of the legal cannabis market, and it is expected to grow to USD 140.5 million by 2025 (Consumer Goods & FMCG 2023). The main reason for the rapid growth of this market is the demand for legal cannabis products containing highly bioactive compounds – cannabinoids.

Cannabis sativa L. plants synthesize about 600 compounds including cannabinoids, flavonoids, terpenoids, terpenes, alkaloids, and polyphenols. So far, more than 150 cannabinoids have been identified in hemp (Hanuš et al. 2016). The term phytocannabinoids is often used to discriminate the plant-produced cannabinoids from the endocannabinoids which are a natural part of the endocannabinoid system in mammals. Phytocannabinoids interact with the endocannabinoid receptors (CB1, CB2, and GPR₅₅) and some other receptors, e.g.: transient potential vanilloid receptor (TRPV) or serotonin receptor (5-TH2A, 5-TH3A) (Stasiłowicz et al. 2021). Cannabinoids have found medical application in epilepsy, sclerosis multiplex (SM), and in alleviating side effects of chemotherapy such as vomiting and nausea, pain, and appetite loss (Stasiłowicz et al. 2021). They have also shown documented pharmacological potential in inflammatory bowel diseases (IBDs), Parkinson’s disease, Tourette’s syndrome, schizophrenia, glaucoma, and coronavirus disease (Raj et al. 2021). Beneficial properties of C. sativa are mainly attributed to two active compounds: the hallucinogenic ∆⁹-tetrahydrocannabinol (Δ⁹-THC) and the non-psychoactive cannabidiol (CBD). Apart from these two predominant phytocannabinoids, cannabigerol (CBG), cannabichromene (CBC), cannabinol (CBN), and cannabivarin (CBCV) are also often reported. In most European countries hemp cultivation is restricted by law due to the anti-drug policy, and the limit of 0.2% of Δ⁹-THC content was implemented in the European Union (EU Regulation 1307/2013). However, a new limit of 0.3% Δ⁹-THC content came into force on 1 January 2023 according to the new Common Agricultural Policy (CAP). For practical and legal reasons prohibiting the cultivation of high ∆⁹-THC hemp, the cultivars were classified as three main chemotypes: drug-type with higher (>0.3%) levels of ∆⁹-THC and CBD content lower than 0.5%, an intermediate type (chemotype II) with similar levels of each and fiber-type (chemotype III) with higher CBD content (>0.5%), and THC content lower than 0.3% (Grassi and McPartland 2017). Other chemotypes IV and V are also classified: chemotype IV has a low ∆⁹-THC level but a relatively high content of CBG (cannabigerol) and chemotype V – synthesize trace amounts of cannabinoids (Mandolino and Carboni 2004; Meijer and Hammond 2005). Whereas the ratio of ∆⁹-THC/CBD is constant during the plant’s life (Meijer et al. 2003; Pacifico et al. 2008), the content of phytocannabinoids may be very variable. Presence of ∆⁹-THC and CBD is genetically determined by genes coding the enzymes tetrahydrocannabinolic acid synthase (THCAS) and cannabidiolic acid synthase (CBDAS). THCAS and CBDAS are responsible for the production of tetrahydrocannabinolic acid (Δ⁹-THCA) and cannabidiolic acid (CBDA), the native forms of Δ⁹-THC and CBD, respectively. The presence of THCAS and CBDAS (defined as B_T and B_D allele, respectively) differs among cultivars and among individuals within the same cultivar (Borroto Fernandez et al. 2020; Pacifico et al. 2006). Depending on hemp cultivar, from 0.97% up to 35% of individuals carrying the B_T allele (homozygotic or heterozygotic) were determined in fiber hemp (Borroto Fernandez et al. 2020). However, the presence of the B_T allele is not obligatory for Δ⁹-THC production, because it is assumed that both synthases THCAS and CBDAS are potent to produce residual CBDA and Δ⁹-THCA (Melzer, McCabe, and Schilling 2022). Additionally, pseudogenes and gene copy number of THCAS and CBDAS also contribute to phytochemical diversity in hemp (Vergara et al. 2019; Weiblen et al. 2015).

Beyond genetics, the content of cannabinoids is affected by sex (Nagy et al. 2019), ontogeny (Ingallina et al. 2020) and variety of environmental factors (Campbell et al. 2019). Among them light intensity and spectrum (Danziger and Bernstein 2021a; Magagnini, Grassi, and Kotiranta 2018), temperature (Galic et al. 2022), drought (Caplan, Dixon, and Zheng 2019; Sheldon et al. 2021), nutrients (Bernstein et al. 2019; Saloner and Bernstein 2021, 2022a, 2022b; Shiponi and Bernstein 2021), heavy metal (Husain et al. 2019) and plant growth-promoting rhizobacteria (Pagnani et al. 2018) have been reported. Exposure to heavy metals significantly increased total CBD and THC content in hemp plants (Husain et al. 2019). Drought (Caplan, Dixon, and Zheng 2019), low phosphorus supplementation (Shiponi and Bernstein 2021), UV radiation (Giupponi et al. 2020; Jenkins 2021), and blue light (~450–520 nm) combined with a short photoperiod (Magagnini, Grassi, and Kotiranta 2018) may increase cannabinoid concentration. On the other hand, high level of phosphorus (30 mg L⁻¹) and nitrogen (100 ppm NH₄) (Saloner and Bernstein 2021, 2022b) had an adverse impact on THCA and CBDA accumulation. The importance of plant density and architecture for chemical quality and standardization in drug-type medical cannabis was revealed by Danziger and Bernstein (2021a, 2021b, 2022). They demonstrated that increased density of planting reduced concentration of cannabinoids, however cannabinoid yield per cultivation area was not affected. The response to plant density and architecture modulation was cannabinoid-specific. Under high-density growth, cultivation practice such as defoliation or removal of branches increased light penetration to the bottom of the plants and improved chemical standardization. It should be noted that different cultivars may respond differentially to the same environmental factors (Campbell et al. 2019; Danziger and Bernstein 2021a; Herppich et al. 2020; Poniatowska et al. 2022).

Generally, industrial hemp accumulates lower amounts of cannabinoids compared to drug-type cultivars, with predominance of CBD over other cannabinoids. Low Δ⁹-THC cultivars are also in demand on the market, because not all patients and consumers accept the psychotic effect associated with medical marijuana treatment. Long-term use of high Δ⁹-THC hemp can be addictive and lead to cognitive impairment, especially in young people. Therefore, industrial hemp is also considered as a source of valuable cannabinoids, particularly CBD (Glivar et al. 2020; Pexová Kalinová et al. 2021). However, phenotypic and genetic variations within industrial hemp are high, and even within the same cultivar individuals might display different chemotypes (Hillig and Mahlberg 2004; Pacifico et al. 2006; Welling et al. 2016). On the other hand, the raw material for pharmaceutical production has to meet strict requirements: batch-to-batch consistency and standardized content of cannabinoids (Chandra et al. 2017). Therefore, the best way to achieve production of uniform plants and standardized raw material is vegetative propagation using clones (Fischedick et al. 2010). Vegetative propagation can be achieved using cuttings, as is practiced in drug-type cultivars, or by in vitro cultures (Chandra et al. 2017). In vitro cultures can offer rapid multiplication, large-scale production, controlled conditions of cultivation, and a lack of environmental restrictions (e.g., exposure to pathogens and pests), but above all they provide true-to-type plants. Hence, they can be useful in reproduction of valuable genotypes for production of standardized raw material or for breeding purposes, providing the parent material for crossbreeding.

In our study, clonally propagated plants of industrial hemp derived from individual seedlings were regenerated in vitro, hardened, and cultivated in the vegetation hall under semi-controlled conditions. Cloned plants from each of the 17 accessions were used as material for the phytochemical study on phenotypic variation. This preliminary study is a part of a program aimed at evaluation and selection of hemp genotypes with higher CBD content for further breeding programs. The main purpose of this study was to obtain clonally propagated plants of industrial hemp genotypes and to determine the content of CBDA, CBD, Δ⁹-THCA, Δ⁹-THC, CBC, CBG, and CBN using the HPLC-DAD/UV method.

Material and methods

Plant material

Hemp seed cultivars (Epsilon 68, Henola), accession K290, and hybrids – Carmagnola × K 290 (marked 1565) and Beniko × Carmagnola (marked 1491) – were obtained from the Fibrous Plants Gene Bank of the Institute of Natural Fibres and Medicinal Plants (INF&MP-NRI, Poznań, Poland). Seeds of the Henola variety were provided by Prof Henryk Burczyk from the seed plantation (Experimental Station, INF&MP, Pętkowo). Basic characteristics of hemp cultivars and the hybrid components are presented in Table 1.

Table 1. Breeding information on tested hemp accessions and the parent components of hybrids (Carmagnola × K 290 (1565) and Beniko × Carmagnola (1491)).

Cultivar/accession	Origin	Purpose	Genotypic expression	Vegetative cycle (days)	Plant height (cm)	CBD content (%)	THC content (%)
Henola	Poland	Grain	Monoecious	115–120	150–200	0.07–1.00	<0.02
Epsilon 68	France	Fiber/Grain	Monoecious	140–150	250–350	0.51–1.04	0.13–0.03
K290	Ukraine	Fiber/Grain	Dioecious	130–140	300–350	1.25–2.03	0.04
Carmagnola	Italy	Fiber/CBD	Dioecious	140–160	200–250	0.13–3.80	0.01–0.24
Beniko	Poland	Fiber/Grain	Monoecious	140–160	250–300	0.39–1.50	0.01–0.08

Data were obtained from Fibrous Plants Gene Bank INF&MP. K290 is a local landrace from Dnipro province.

(former Dnepropetrovsk province).

In vitro propagation

Seeds were sterilized in 70% (v/v) ethanol for 1 min followed by rinsing in commercial bleach-ACE solution (3:1) for 20 min. After sterilization seeds were washed three times in sterile, deionized water, placed in a half-strength MS medium (Murashige and Skoog 1962), and cultivated at 25°C in darkness for 3 days, then for 7 days in light. In vitro cultures were established from sterile and individually marked seedlings. From each single seedling the genotype was derived and cloned using shoot tip explants. Shoot tips were removed from the regenerated plants at the age of 14–21 days and cultured for the next 2 weeks to stimulate the growth of lateral shoots, as described by Wróbel et al. (2022).

Acclimatization conditions

The regenerated plants with well-developed roots were removed from the vessels and washed in autoclaved water, then placed in pots with sterilized soil (standard garden soil without additives) under glass covers and grown at 25 ± 1°C (18/6 photoperiod, 80–100 μmol m² s⁻¹). After 1 week, the glass covers were replaced with plastic covers and plants were progressively exposed to the environmental humidity. Plantlets were irrigated twice a week. Plants were hardened for 3–4 weeks and then transferred to the pots in the vegetation hall.

Pot experiment in the vegetation hall

The pot experiment was carried out in 2021, from May 25 until September 13, inside the vegetation hall in the Experimental Station in Pętkowo (Wielkopolskie, Poland N: 52°12ʹ09ʹʹ; E: 17°15ʹ21ʹʹ). The vegetation hall was not heated or lighted, except solar source. The temperature was maintained in the range of 15–25°C, and humidity 40–60% during the entire period of the experiment. The pots were randomly arranged in the vegetation hall.

The pots of capacity of 160 dm³ (50 cm high and 30 cm in diameter) were filled with 21 kg of the soil. Soil analysis demonstrated a mechanical composition of sandy clay, slightly acidic (pH = 5.9) with a high (8.6 mg·100 g⁻¹ of soil) magnesium content, moderate phosphorus, and potassium content (13.5 and 18.2 mg·100 g⁻¹ of soil, respectively); boron (1.15 mg·100 g⁻¹ of soil), manganese (54.7 mg·100 g⁻¹ of soil), copper (2.9 mg·100 g⁻¹ of soil), zinc (6.7 mg·100 g⁻¹ of soil); and low iron content (423 mg·100 g⁻¹ of soil), and a 1.64% content of humus. Plants were not fertilized. Hardened plants derived from in vitro cultures were transplanted into pots (3–4 per pot). Each genotype was represented by 5–10 regenerated (cloned) plants. Plants were irrigated (200 ml per plant) twice a day during the intensive growth phase, and once per day during the flowering phase. The percentage of acclimatized plants was calculated 1 month after transplant and before harvest. In vitro regenerated plants, acclimatized plantlets and growing plants in the vegetation are presented in Figure 1 (a-c).

Figure 1. a) in vitro regenerated plant (EPS/40), b) acclimatization of plantlets, c) plants growing in the vegetation hall.

Plant material collection and HPLC-DAD/UV analysis

Sampling of the female/monoecious plants was carried out at the beginning of the seed maturation stage. Inflorescences were cut (±30 cm) and packed in paper bags. Henola genotypes were harvested on July 30^th (66 days after planting), the other genotypes on September 13^th, 2021 (111 days after planting). Each sample contained material collected from a single plant. Harvested plant material was dried at 40°C ± 0.1°C in a laboratory dryer. Content of cannabinoids (CBDA, CBD, Δ⁹-THCA, Δ⁹-THC, CBC, CBG, and CBN) was analyzed by a certified laboratory (Cannalabs by EkotechLAB, Poland) using the HPLC-DAD/UV method. Dried plant material was sieved using a 0.5 mm mesh sieve and grounded. Samples were extracted by adding 10 mL of ethanol (96%) to 100 mg of the accurately weighed ground material. Then, the samples were incubated for 30 min in an ultrasonic bath at 50°C. The supernatant was removed and filtered through a syringe filter (0.2 μm) into the vials. Analyses were performed using a Thermo Dionex Ultimate 3000 chromatograph with diode array detector (DAD RS 300). Separation was carried out on a Dr Maisch ReproSil XR 120 (3.5 µm × 50 mm × 3 µm) column. The mobile phase A was water + 0.1% formic acid (v/v), while the mobile phase B was methanol + 0.05% formic acid (v/v). The elution procedure was as follows: 1–6.8 min, 60% B→73% B; 6.8–9.3 min, 73% B→95% B; 9.3–9.4 min, 95% B→60% B; 9.4–10 min, 60% B. The volume of the samples was 1 μL, flowrate 1 mL/min, with detection at 228 nm, and the column temperature was 40°C. The amount of each cannabinoid (%w/w) in the original sample was calculated using calibration curves of standard compounds. All cannabinoid standards were purchased in LGC (Dr Ehrenstorfer®, Augsburg, Germany). The limit of quantification for all cannabinoids was 0.01%, and the limit of detection was 0.003%. The total concentration of Δ⁹-THC and CBD in %w/w [%mg of cannabinoid/100 mg herbal drug] was calculated as follows:

$\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}{\mathrm{\Delta }}^9-\mathrm{T}\mathrm{H}\mathrm{C}=\mathrm{\%}{\mathrm{\Delta }}^9-\mathrm{T}\mathrm{H}\mathrm{C}+\left(\mathrm{\%}{\mathrm{\Delta }}^9-\mathrm{T}\mathrm{H}\mathrm{C}\mathrm{A}\mathrm{x}0.877\right)$

$\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}\mathrm{C}\mathrm{B}\mathrm{D}=\mathrm{\%}\mathrm{C}\mathrm{B}\mathrm{D}+\left(\mathrm{\%}\mathrm{C}\mathrm{B}\mathrm{D}\mathrm{A}\mathrm{x}0.877\right)$

The conversion factor 0.877 represents a loss of molecular mass during the decarboxylation reaction.

Statistical analysis

Statistical analysis was performed using Statistica 12.0 (StatSoft Inc.). Phytochemical measurements were repeated at least three times for each tested sample and compound. These data were analyzed using ANOVA, and the statistical significance was determined by applying Tukey’s test and Fisher’s test with a p value of 0.05.

Results and discussions

Clonal propagation and obtaining the raw material for analysis

Clonal propagation ensures a constant cannabinoid profile and a high uniformity of the raw material, which allows adherence to strict pharmaceutical standards (Chandra et al. 2017; Fischedick et al. 2010). Although different in vitro protocols have been reported (Monthony et al. 2021), the low regeneration rate, poor replicability, and genotype-dependent response of the explants are the main limitations for in vitro propagation. In this study, we used a slightly modified regeneration protocol, developed previously in our laboratory (Wróbel et al. 2022). This protocol involves direct regeneration from shoot tips taken from lateral shoots without the use of cytokinins. Applying this regeneration protocol, 150 plants representing 17 different genotypes were regenerated in vitro (Figure 1).

Epsilon 68 genotypes included line EPS/40 obtained previously as well as the new genotypes of this cultivar. The multiplication rate varied depending on the tested genotype and ranged from 1.8 to 4.5 shoot tips per donor plant calculated in a two-week cycle. These results are consistent with previously obtained values, with the multiplication rate 3.0 (Wróbel et al. 2022). Plantlets after transplanting in the vegetation hall grew vigorously and 93% of them successfully survived to harvest (Figure 1c). Samples (139) collected from individual plants were then analyzed using the HPLC-DAD/UV method. Chromatograms of Δ⁹-THCA and CBDA are presented in Figure 2.

Figure 2. a) HPLC profile of EPS/40 genotype at λ = 228 nm, b) HPLC profile of K290/136 genotype at λ = 228 nm. 1—CBD, 2—CBDA (pink colored), 3—Δ⁹-THC, 4—Δ⁹-THCA.

Differences in the content of CBDA, CBD, ∆⁹-THCA, ∆⁹-THC, CBG, and CBC are summarized in Table 2. The sum of cannabinoid content (THCA, THC, CBD, CBDA, CBC and CBG) is presented in Figure 3.

Figure 3. Sum (%) of total cannabinoids (CBDA, CBD, Δ⁹-THCA, Δ⁹-THC, CBC and CBG) in tested genotypes. Mean values in column with the same letter are not significantly different at p = .05 (tukey’s multiple range test). Genotypes: HN—Henola; 1491—Beniko xCarmagnola, 1565—Carmagnola x K 290; EPS—Epsilon 68.

Table 2. Content of CBDA, CBD, ∆⁹-THCA, ∆⁹-THC, CBC, CBG, and CBN in tested genotypes.

No	Genotype	n	Mean content % (±SD) DW
			CBDA	CBD	∆^9-THCA	∆^9-THC	CBC	CBG
1	HN/2	7	0.749 ± 0.082^ab	0.216 ± 0.060^bcd	0.014 ± 0.005^a	0.014 ± 0.005^a	0.034 ± 0.008^cde	<0.01
2	HN/35	8	0.568 ± 0.060^a	0.110 ± 0.028^ab	0.014 ± 0.005^a	<0.01	0.010 ± 0.00^a	<0.01
3	HN/39	10	0.478 ± 0.097^a	0.046 ± 0.021^a	0.018 ± 0.006^a	<0.01	0.010 ± 0.00^a	<0.01
4	HN/64	7	0.619 ± 0.114^a	0.173 ± 0.073^bcd	0.020 ± 0.000^ab	0.010 ± 0.000^a	0.016 ± 0.005^ab	<0.01
5	1491/78	10	2.000 ± 0.283^de	0.208 ± 0.047^bcd	0.052 ± 0.006^abc	0.017 ± 0.005^a	0.038 ± 0.006^e	0.016 ± 0.005^a
6	1491/5	9	3.311 ± 0.554^f	0.249 ± 0.046^cde	0.082 ± 0.014^bc	0.022 ± 0.004^a	0.031 ± 0.003^bcde	0.020 ± 0.006^ab
7	K290/K1	9	2.027 ± 0.571^de	0.266 ± 0.133^de	0.058 ± 0.033^abc	0.018 ± 0.008^a	0.036 ± 0.017^de	0.017 ± 0.005^a
8	K290/136	5	2.176 ± 0.287^de	0.200 ± 0.081^bcd	0.824 ± 0.200^d	0.298 ± 0.080^b	0.036 ± 0.011^cde	0.018 ± 0.005^ab
9	1565/8	7	1.649 ± 0.211^cde	0.106 ± 0.024^ab	0.043 ± 0.005^abc	0.010 ± 0.000^a	0.017 ± 0.005^abc	0.050 ± 0.008^c
10	1565/26	10	1.632 ± 0.360^cd	0.353 ± 0.134^e	0.039 ± 0.017^ab	0.026 ± 0.007^a	0.031 ± 0.01^bcde	0.027 ± 0.005^b
11	EPS/1	9	1.876 ± 0.108^de	0.244 ± 0.054^cd	0.039 ± 0.003^abc	0.017 ± 0.005^a	0.021 ± 0.003^abcd	<0.01
12	EPS/6	8	2.158 ± 0.440^de	0.221 ± 0.064^bcd	0.053 ± 0.014^abc	0.018 ± 0.005^a	0.043 ± 0.017^e	0.020 ± 0.00^ab
13	EPS/10	9	1.860 ± 0.392^de	0.167 ± 0.037^bcd	0.051 ± 0.011^abc	0.013 ± 0.005^a	0.034 ± 0.012^de	<0.01
14	EPS/19	8	1.214 ± 0.121^bc	0.170 ± 0.046^bcd	0.034 ± 0.007^ab	0.012 ± 0.004^a	0.016 ± 0.005^ab	0.010 ± 0.00^a
15	EPS/21	10	2.159 ± 0.175^e	0.151 ± 0.021^bc	0.055 ± 0.010^abc	0.013 ± 0.005^a	0.031 ± 0.011^bcde	<0.01
16	EPS/24	5	3.058 ± 0.324^f	0.252 ± 0.038^cde	0.066 ± 0.009^abc	0.016 ± 0.005^a	0.026 ± 0.005^abcde	<0.01
17	EPS/40	8	4.723 ± 0.542^g	0.269 ± 0.004^de	0.103 ± 0.010^c	0.019 ± 0.010^a	0.069 ± 0.011^f	<0.01

Mean values in column with the same letter are not significantly different at p = .05 (Tukey’s multiple range test). HN – Henola; 1491—Beniko x Carmagnola, 1565—Carmagnola x K 290; EPS – Epsilon 68, n – number of plants.

Total content of CBD, ∆⁹-THC and ratio of total ∆⁹-THC/total CBD, as well as total CBD/total ∆⁹-THC, are presented in Table 3. Coefficient of variation calculated for CBDA and THCA is presented in Table 4.

Table 3. Content of total CBD and total THC, ratio of CBD/THC and ratio of THC/CBD.

Genotype	n	Mean content % (±SD) DW		Ratio of total CBD/THC	Ratio of total THC/CBD
		Total CBD	Total THC
HN/39	10	0.465 ± 0.103^a	0.026 ± 0.006^a	17.88	0.06
HN/35	8	0.608 ± 0.073^a	0.022 ± 0.005^a	27.64	0.04
HN/64	7	0.715 ± 0.153^ab	0.028 ± 0.001^ab	25.54	0.04
HN/2	7	0.872 ± 0.080^ab	0.025 ± 0.005^ab	34.88	0.03
EPS/19	8	1.234 ± 0.140^bc	0.041 ± 0.007^ab	30.10	0.03
1565/8	7	1.552 ± 0.205^cd	0.048 ± 0.004^ab	32.33	0.03
1565/26	10	1.784 ± 0.244^de	0.060 ± 0.012^abc	29.73	0.03
EPS/10	9	1.798 ± 0.368^de	0.058 ± 0.011^abc	31.00	0.03
EPS/1	9	1.889 ± 0.147^de	0.051 ± 0.007^abc	37.04	0.03
1491/78	10	1.962 ± 0.288^de	0.063 ± 0.008^bc	31.14	0.03
K290/K1	9	2.043 ± 0.515^de	0.068 ± 0.025^bcd	30.04	0.03
EPS/21	10	2.044 ± 0.159^de	0.059 ± 0.010^abc	34.64	0.03
K290/136	5	2.108 ± 0.181^de	1.021 ± 0.132^e	2.06	0.48
EPS/6	8	2.113 ± 0.337^e	0.064 ± 0.009^bcd	33.02	0.03
EPS/24	5	3.034 ± 0.321^f	0.074 ± 0.012^bcd	27.49	0.04
1491/5	9	3.153 ± 0.524^f	0.094 ± 0.015^cd	33.54	0.03
EPS/40	8	4.410 ± 0.491^g	0.109 ± 0.010^d	40.46	0.02

HN-Henola. 1491 - Beniko x Carmagnola. 1565 - Carmagnola x K 290; EPS-Epsilon 68; n- number of plants. Values of total CBD and total THC are mean ± SD of three repetitions for chemical analysis. Mean values in column with the same letter are not significantly different at p = .05 (Tukey’s multiple range test).

Table 4. Coefficient of variation calculated for THCA and CBDA.

Genotype	n	CV (%)
		CBDA	THCA
HN/39	10	20.3	33.3
HN/35	8	10.6	35.7
HN/64	7	18.4	0.0
HN/2	7	10.9	35.7
EPS/19	8	10.0	20.6
1565/8	7	12.8	11.6
1565/26	10	22.1	22,1
EPS/10	9	21.1	21.6
EPS/1	9	5.8	7.7
1491/78	10	14.2	11.5
K290/K1	9	28.2	38.7
EPS/21	10	8.1	15.6
K290/136	5	13.2	24.3
EPS/6	8	20.4	26.4
EPS/24	5	10.4	13.6
1491/5	9	16.7	17.1
EPS/40	8	11.5	9.7

HN-Henola. 1491 - Beniko x Carmagnola. 1565 - Carmagnola x K 290; EPS-Epsilon 68; n- number of plants.

Content of CBDA, CBD, ∆⁹-THCA, ∆⁹-THC, CBG, and CBC

Differences in content of CBDA, CBD, ∆⁹-THCA, ∆⁹-THC, CBG, and CBC between tested genotypes were significant (Table 2). CBDA was the dominant cannabinoid, at the level 0.48–4.72%, whereas the content of other cannabinoids was significantly lower: 0.05–0.35% of CBD, 0.01–0.82% of Δ⁹-THCA, 0.01–0.3% of Δ⁹-THC, 0.01–0.5% of CBG, and 0.01–0.07% of CBC. Significant variation in CBDA content among tested genotypes was found. The majority of genotypes (10) contained a medium level of this compound at the level 1–3%. Three genotypes (EPS/40, EPS/24 and 1491/5) reached the highest level of CBDA (>3%); among them EPS/40 was characterized by a significantly high content (4.723%±0.542) of this compound. Similar or even higher level of CBDA (6.282%) for this French cultivar was reported by Vásquez-Ocmín et al. (2021). Δ⁹-THCA content also varied significantly among tested genotypes. Most of the tested genotypes contained low (0.02–0.10%) Δ⁹-THCA level. All Henola genotypes reached the lowest content of Δ⁹-THCA, i.e., ≤0.02%. In this respect, K290/136 genotype was characterized by the highest content (0.8%) of this cannabinoid. Unfortunately, there are no data available on the variability within this accession. Carboxylic acidic forms of cannabinoids (CBDA and THCA) are predominant and native for hemp, but they are nonenzymatically degraded into the neutral form which occurs naturally in plants or during heating or drying (Taura et al. 2007). In this study, the percentage of CBD in the sum of CBDA and CBD ranged from 5.4% to 22.4%, whereas Δ⁹-THC represented 15.5% to 50% of total Δ⁹-THC, respectively. The percentage of neutral forms is mainly dependent on the temperature of storage and drying. These values are comparable with the results obtained by Glivar et al. (2020) or Eržen et al. (2021). CBC was detected at a low level of 0.01%–0.04% in the majority of the tested samples, but EPS/40 genotype was characterized by the highest content of this cannabinoid (0.069%±0.011) and the result achieved statistical significance.

CBG was determined in a trace amount (≤0.01%) in most of the samples, but seven genotypes (1491/5, 1491/78, 1565/8, 1565/26, K290/K1, K290/136, and EPS/6) accumulated CBG at the level 0.02%–0.05%. Genotype 1565/8 was characterized by a relatively high content (0.05%±0.008) of this cannabinoid. CBN was not detectable at a level below 0.01%. This might imply that degradation of ∆⁹-THCA during drying or storage resulted in a trace amount of CBN – the side product of this process (Fischedick et al. 2010).

Total CBD and total ∆⁹-THC content and ratios of CBD/∆⁹-THC and ∆⁹-THC/CBD

Heterogeneity in total content of CBD and ∆⁹-THC was also observed in this study. A higher content of total CBD (>3%) was recorded for two genotypes – EPS/40 and 1491/5, but the highest level of this cannabinoid (4.410%±0.491) was determined for EPS/40 genotype (Table 3). Comparable or even higher content of CBD (4–9.6%) was found by Glivar et al. (2020) in bracts or different parts of inflorescences of Antal, Carmagnola, Helena, and Tiborszallasi cultivars. Our results are consistent with previously determined content of total CBD (3.9%–5.4%) in Epsilon 68 genotypes (Wróbel et al. 2022). In this study, we found significant variability within the Epsilon 68 cultivar and confirmed that the EPS/40 line accumulates relatively constant content of CBD under semi-controlled conditions (4.410%±0.491) as grown in the field (5.0134%±0.1182) (Wróbel et al. 2022). Notably, the Δ⁹-THC value did not exceed the allowed threshold (0.3%) in this study and previously. This is a good prognostic sign for this accession, and it should be considered as one of the criteria for selection in the breeding process.

The total Δ⁹-THC concentration is important information for growers and breeders due to the different national legal regulations. A new limit of 0.3% was set by Polish law in 2022 (Dz 2022 poz. 764). It should be stressed that Δ⁹-THC values observed for most single plants were below the allowed threshold and ranged from 0.02% to 0.08%. In this study, one accession K290/136 genotype (1.021%±0.132) exceeded the allowed THC limit. This may be due to the naturally occurring inter-individual variability within this local Ukrainian landrace. Unfortunately, there are not any data on variability in the K290 accession. An increased level of THC in single plants has been recorded previously in some industrial cultivars: Carmagonola (Pacifico et al. 2006), Felina 34 and Białobrzeskie (Mechtler, Bailer, and de Hueber 2004) and K-573 accession (Mańkowska et al. 2015). It is not excluded that during in vitro process an individual (seedling) representing a different chemotype was found. Over 200 seedlings of the K290 accession were obtained, and only a few of them were randomly selected for further micropropagation, among them seedling number 136 (K290/136). However, more studies are needed in the subsequent seasons for a definite conclusion.

Uniformity of cannabinoid content is important information about the tested accessions and can be evaluated with standard deviation (SD); however more relevant is coefficient of variation (CV). In this regard, a very high uniformity in Δ⁹-THCA and CBDA content was found for EPS/1 and EPS/40 genotypes (Table 4). CV calculated for EPS/1 genotype were 7.7% and 5.8% and 9.7% and 11.5% for EPS/40, respectively. On the opposite side, significantly higher CV values (38.7% and 28.2% respectively) were noted for K290/K1 or 1565/26 lines. Potential causes of the variability within genotype could be uneven access of plants to light in the vegetation hall. Light quality and intensity play a key role. The significant increase in the amount of cannabinoids extracted from inflorescences from bottom to top of the stem was demonstrated by Namdar et al. (2018). Effect of micro-climates in plant shoots due to an increase in inter-shoot shading was found recently (Danziger and Bernstein 2021a, 2022). Shadow or weak light penetration in inflorescences was correlated with lower cannabinoid concentration. On the other hand, defoliation increased cannabinoid concentrations locally due to better light penetration. Spatial gradients of light intensity in plant canopies intensify with plant height. In this study, differences in plant height resulted from differences between cultivars, e.g., Henola and Epsilon 68 (Table 1) and individual plants. Another source of variability may have been temperature fluctuations (15–25°C) under the semi-controlled conditions. Considering that the cloned plants represented different individuals derived from different cultivars and accesions, a varied response to light and temperature was possible. Different responses of the cultivars to the same environmental factors were documented (Herppich et al. 2020; Poniatowska et al. 2022). Therefore, further research on the effects of different light and temperature is worth considering.

The ratio of total Δ⁹-THC/total CBD as well as the ratio of total CBD/total Δ⁹-THC is used for chemotype evaluation. Δ⁹-THC/CBD values lower than 1.0 were attributed to the fiber-type cultivars. In this study, all tested genotypes fulfilled this criterion (Table 3). However, according to the classification based on CBD/Δ⁹-THC ratio, two different chemotypes are represented: III – typical for fiber type and II – intermediate. Chemotype III is represented by almost all genotypes, with a characteristic low Δ⁹-THC content (<0.3%) and prevalent CBD content >0.5%. Intermediate chemotype II is represented by K290/136 genotype with CBD/Δ⁹-THC ratio 2.06 and total content of Δ⁹-THC above 1%. Single plants expressing different chemotypes (II and III) were identified within cultivars Carmagnola and Fibranova (Pacifico et al. 2006), Felina 34, Białobrzeskie, and Beniko (Mechtler, Bailer, and de Hueber 2004). A study on the distribution of chemotypes in industrial hemp confirmed the presence of at least some individuals with Δ⁹-THC-predominant or intermediate phenotype within 23 cultivars (Borroto Fernandez et al. 2020). Genetic analysis determined the frequency of B_T allele-carrying individuals from 0.97% to 35% depending on cultivar. It means that if the size of the tested sample is sufficiently large (100 or more individuals), one is very likely to find individuals (genotypes) with an active B_T allele (B_D/B_T or B_T/B_T genotype) expressing different chemotypes. Therefore, random selection of seeds (individuals) during the in vitro process might lead to clonal propagation of individuals expressing different chemotypes in this study.

The average CBD/Δ⁹-THC ratio is about 33:1, which is consistent with the results obtained by Glivar et al. (2020). However, this ratio for two genotypes (EPS/40 and HN/35) was more than 40:1. Toth et al. (2020) suggested that it might be difficult to develop a cultivar that accumulates more than 6% of CBD content while maintaining the allowed low Δ⁹-THC level based on the CBD/Δ⁹-THC ratio 20:1. Our results are consistent with this claim, but further studies and experiments are needed to verify this assumption.

Conclusions

In this study, in vitro cultures were used for clonal propagation of industrial hemp. One hundred and thirty-nine plants representing 17 different hemp genotypes were regenerated in vitro, hardened, and grown until harvest in the vegetation hall. The results of HPLC-DAD/UV analysis revealed large variability in the total content of cannabinoids (0.55–5.18%) among tested genotypes as well as within the Epsilon 68 cultivar. The highest content of total CBD (4.410%±0.491) was recorded for EPS/40 genotype, while the maintained level of THC was below the legally allowed threshold (0.3%). Some accessions (EPS/1 and EPS/40) expressed constant cannabinoid profile and high uniformity. Therefore, we can conclude that some clonally propagated plants provided reproducible hemp material as a potential source of cannabidiol. Additionally, we confirmed that EPS/40 genotype accumulates higher (>4%) content of CBD under semi-controlled conditions as grown in the field and therefore is a good candidate for selection and material for breeding practice.

Highlights

• Seventeen different hemp genotypes were regenerated in vitro, hardened, and grown in a vegetation hall.

• Single plants were analyzed using high-performance liquid chromatography with UV/diode-array detection (HPLC-DAD/UV).

• Significant variability in the total cannabinoid content (0.55–5.18% in dry weight) among tested genotypes was found

• The highest content of total CBD (4.410%) was recorded for EPS/40 genotype

• Clonally propagated plants can provide Cannabis material as a potential source of cannabidiol.

Ethical approval

We confirm that all the research meets ethical guidelines and adheres to the legal requirements of the study country. The research does not involve any human or animal welfare-related issues.

Acknowledgements

The authors would like to express their gratitude to Professor Henryk Burczyk for providing seeds of the Henola cultivar and to Dr Karolina Wielgus for her assistance and support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was supported by the Polish Ministry of Agriculture and Rural Development, resolution of the Council of Ministers no. DSR.nw.070.4.2021 and no. 171/2017.