Alternative Rooting Methods for Medicinal Cannabis Cultivation in Denmark—Preliminary Results

Abstract

1. Introduction

Cannabis sativa L. belongs to the family Cannabaceae, and it is among the first plants to be cultivated. Its uses include fiber, fuel, food, cosmetics, and medicine [1]. Drug-type cannabis contains a range of bioactive compounds, including cannabinoids, terpenoids, and alkaloids [2]. In January 2018, the program on medicinal cannabis was initiated in Denmark, making it possible for doctors to prescribe cannabis-based medicine to patients with unresolved medical conditions [3], thus fostering research locally [4]. Additionally, a license to cultivate medicinal cannabis has to be applied for, as selling dried cannabis inflorescences/flowers (cannabis bulk) is only allowed upon obtaining official approval for consistently producing high-quality inflorescences, which are measured as high, uniform levels of cannabinoids. Eleven companies have so far been given the final approval to sell bulk cannabis [5], and the trial was recently extended, making it permanently legal for companies to grow medicinal cannabis in Denmark [6]. Today’s medicinal cannabis production in Denmark is highly specialized and technical but simultaneously embedded in a framework that allows for limited use of some modern agricultural practices, for example, within vegetative propagation. In general, cannabis can be propagated through seed [7], vegetative cuttings [8], or in vitro propagation [9,10]. Propagation through vegetative cuttings is likely the most realistic practice in the Danish context, as it ensures consistent cannabinoid levels and similar growth patterns between plants. Caplan [11] evaluated the effects of several treatments (leaf tip removal, leaf number, position of cuttings, and rooting hormone) on the rooting success of cannabis cuttings. These authors recommend having at least three fully expanded leaves on cuttings, and the position of cuttings on mother plants had no effect on rooting success [11]. Knowledge of such propagation practices is important from a practical perspective since it is manual labor and, therefore, associated with considerable costs. In order to initiate rooting, cuttings are commonly exposed to auxins, a group of naturally occurring hormones that are important regulators in plants regarding the division, expansion, and differentiation of cells as reviewed by Enders et al. [12]. While such products are not explicitly mentioned in the official documents regarding Danish medicinal cannabis production [13,14], their prohibition has been confirmed [15]. Growers may then choose to import already rooted cuttings from elsewhere. Therefore, novel methods to induce root formation in cuttings of drug-type cannabis are needed in Denmark. Alternative rooting products for cannabis based on biological compounds are available on the market, but few have been tested scientifically. A study by Caplan [11] tested a willow (Salix alba) bark extract gel against an indole butyric acid (IBA) gel on drug-type cannabis cuttings with reported rooting successes in 40% and 84%, respectively. Willow bark extracts contain the phytohormone salicylic acid, which has been associated with improved root growth in plant cuttings, for example, in soybean [16]. Dialogue with Danish cannabis growers indicates that biostimulants, e.g., protein hydrolysates and algae extracts, are applied as alternative root agents, although with variable success. The current study focused on the root-inducing (Ri) ability that the rhizogenic bacterium Rhizobium rhizogenes possesses. This is a naturally occurring soil bacterium that exerts a pathogenic lifestyle conferred by the root-inducing plasmid (pRi) that strains within R. rhizogenes [17] and R. radiobacter as reviewed by Desmet et al. [18]). The transfer DNA (T-DNA) resembles the part of the plasmid that can horizontally transfer to the host plant genome and cause “hairy roots” to form at the infection site, which is the main symptom of successful T-DNA integration in the host [19,20]. Agropine strains of R. rhizogenes contain two separate T-DNA regions, namely, the left (T_L) and the right (T_R). T_L harbors 18 open reading frames (ORFs), including the root oncogenic loci (rol) genes rolA, rolB, rolC, and rolD, which are essential to initiate hairy root formation, whereas T_R contains a rolB homolog and two genes (aux1 and aux2) that are involved in auxin synthesis [21].

It is sometimes possible to regenerate plants from transformed tissue, resulting in complete, transgenic plants recognized as the hairy root or Ri phenotype [18]. Such plants are considered naturally transformed, and this phenomenon has been reported to occur in nature, e.g., tobacco (Nicotiana glauca) [22], cultivated sweet potato (Ipomoea batatas) [23], and Linaria [24]. Using naturally occurring wild-type bacterial strains is, therefore, considered a non-GMO (genetically modified organism) method in the European Union so long as they remain unmodified [25]. Several general traits that often follow a natural transformation include dwarfed stature, wrinkled leaves, increased rooting ability, altered flowering, and reduced fertility (reviewed by Cassanova et al. [26] and Desmet et al. [18]). Additionally, transformed tissues can produce elevated levels of medically important secondary metabolites, e.g., anthraquinones, saponins, flavonoids, alkaloids, anthocyannins, and terpenes [27,28]. A rooting deficiency of almond (Prunus dulcis) cuttings of the cultivar “Supernova” was overcome after transformation with wild-type R. rhizogenes [29]. In order to fully realize such potential for medicinal cannabis, transformation and regeneration would be necessary, but only the former has been demonstrated in C. sativa by Feeney and Punja [30] and Wahby et al. [31]. Thus, natural transformation can potentially serve as an important tool in medicinal cannabis, effectively bypassing the legal issues that genetic engineering implies. Agronomically useful traits regarding rooting can also be obtained without altering shoot properties. Lambert and Tepfer [32] investigated chimeric apple cuttings (transgenic roots, WT aerial part) to overcome a rooting deficiency, which was achieved on cuttings of thee-year-old trees (rootstock M9b) after inoculation with R. rhizogenes (A4 strain). In the current study, we investigated rooting performance in chimeric cannabis cuttings in a similar manner to Lambert and Tepfer [32]. We specifically aimed to investigate whether R. rhizogenes inoculation could improve rooting in cannabis cuttings versus IBA or treated solely with water.

2. Results

In the current study, the rooting performance of cannabis cuttings was compared between three treatment groups in two repetitions displaced in time. The rooting success, meaning whether there was visible root(s) after 25 days of growth or not, was recorded for each cutting and used to evaluate whether all data across repetitions and cultivars could be combined (Table 1). The highest rooting success within treatments for both repetitions was that of IBA (98 ± 3%). This was followed by a 95 ± 5% and 86 ± 14% success rate for the Rhizobium rhizogenes (A4) and H₂O treatment, respectively (Table 1). In repetition 1, 83 out of 96 cuttings rooted (86%), whereas all cuttings of repetition 2 rooted successfully. Rooting failure was most pronounced in the H₂O treatment, where 25% of the “Bruce Banner” and 0% of the “Motherlode Kush” cuttings rooted (Table 1). As rooting averages differed between repetitions, it was hypothesized that these differences were more related to repetitions rather than cultivars. Therefore, the data was further examined before being combined or separated. A chi-squared test between cultivars in both repetitions was significant (p ≤ 0.0102), indicating a difference in rooting success among them (data not shown). The poor rooting success of “Bruce Banner” (25%) and “Motherlode Kush” (0%) in repetition 1 was removed from the data before running the test again. This rendered the expected rooting success in alignment with the observed frequencies, ultimately providing the basis for combining data across cultivars (p ≤ 0.2786).

Finally, the rooting success between repetitions was analyzed, excluding “Bruce Banner” and “Motherlode Kush” cuttings from repetition 1, which was significantly different (p ≤ 0.0093), suggesting that data across repetitions should not be combined. Therefore, this analysis provided the basis for combining data across cultivars but not across repetitions.

3. Discussion

The aim of this work was to investigate R. rhizogenes as an alternative rooting agent to IBA in drug-type cannabis cuttings. Quantitative biometrics were presented to analyze differences in rooting abilities between treatments. A key finding was that R. rhizogenes (A4) and H₂-treated cuttings proved to have longer and well-established shoots, yet with shorter roots than IBA-treated cuttings. Moreover, results from repetition 1 showed improved rooting abilities of IBA-exposed cuttings compared to A4 and H₂O cuttings, as displayed by the significantly higher root/shoot dry weight ratio of IBA cuttings (0.22 ± 0.01) compared to A4 (0.14 ± 0.001) and H₂O (0.13 ± 0.02) cuttings (Figure 1c). This was also the case in repetition 2 with root/shoot dry weight ratios, measuring 0.32 ± 0.02 (IBA), 0.11 ± 0.00 (A4), and 0.12 ± 0.00 (H₂O) (Figure 2c). Altered proportions of biomass in roots and shoots after IBA exposure was also found by Hunt et al. [33] in a study on vegetative pine cuttings (Pinus elliottii var. elliottii × P. caribaea var. hondurensis). These were either dipped in 16.000 mg L⁻¹ IBA for 5 s or water as a control and after 13 weeks, root/shoot dry weight ratios were significantly different for clones with IBA and clones without. In another study, beech (Fagus sylvatica) seedlings exposed to >1000 mg L⁻¹ IBA had significantly greater root/shoot dry weight ratios (1.28) than untreated controls 105 days after exposure [34].

In repetition 2, the root and shoot growth measurements generally showed similar results for A4 and H₂O treatment groups and several simultaneous significant differences compared to IBA (Figure 2a,c–e,g). Cuttings exposed to IBA exhibited increased rooting (Figure 2c–e) as well as decreased shoot growth (Figure 2a,g). Average total root lengths for A4 and H₂O cuttings were not statistically significant in both repetitions of this experiment (Figure 1d and Figure 2d), suggesting no improved rooting abilities of cannabis cuttings inoculated with R. rhizogenes. Comparable results were reported by Hatta [35] on softwood cuttings of two jujube (Ziziphus jujuba) cultivars (“Contorta” and “Li”). Ten weeks after inoculation with strains A4 and TR105 of R. rhizogenes, neither strain had an effect on the length of the longest root compared to water controls, although the longest roots of “Li” cuttings were significantly longer than those of “Concorta” cuttings (p ≤ 0.05), suggesting different innate rooting abilities between cultivars [35]. A similar cultivar difference was found when measuring the number of roots per cutting, which also revealed significantly more roots (p ≤ 0.05) in cuttings inoculated with the TR105 strains, whereas the A4 strain yielded no more roots on average than water controls [35]. Increased total root lengths (p ≤ 0.01) and root dry weights (p ≤ 0.05) have been shown in mint (Mentha piperita) cuttings upon inoculation with R. rhizogenes (strain A16), although root dry weights were increased more by inoculation with Bacillus megatorium (strain M3) than R. rhizogenes. In the present work, root dry weights of A4 cuttings were the same as those of H₂O cuttings in both repetitions (Figure 1d and Figure 2d).

The effect of IBA exposure on shoot growth in cannabis cuttings has apparently not yet been illustrated in the scientific literature. In this work, A4 and H₂O-treated cuttings grew significantly longer during 25 days of aeroponic propagation (114 ± 7 mm and 103 ± 7 mm) than IBA cuttings (43 ± 5 mm). Longer shoots are preferred by the industry as mature plants can be obtained earlier. The non-significant relationship between A4 and H₂O cutting shoot growth indicated a growth retarding effect of IBA rather than improved shoot growth in A4 cuttings. This could have advantageous economic implications should cuttings not treated with hormone reach maturity faster than those exposed to the hormone, thereby contributing to completing a full growing cycle more effectively. Complete immersion of woody ornamental cuttings in IBA solutions is known to cause shoot growth retardation. Initial growth alterations upon different hormone exposures may, however, be compensated at a later growth stage, eventually yielding similar marketable plants (reviewed by Blythe et al. [36]). In beech, cuttings exposed to 250–1000 mg L⁻¹ IBA for 10 min showed increased shoot lengths after 54 days, while cuttings exposed to 2000 mg L⁻¹ IBA were no different from cuttings without hormone, possibly indicating a threshold value of IBA concentration, duration of the exposure, or a combination of both [34]. Consequently, these authors concluded that the optimum duration of IBA exposure in beech cuttings depended on IBA concentration [34]. Cannabis cuttings in this experiment were exposed to a 100 mg L⁻¹ IBA solution following the “long basal soak” protocol for herbaceous cuttings by Kroin [37], which provided guidelines for hormone solution concentrations and several ways of exposing plant material. This method was also employed by Gehlot [38] on vegetative propagation of neem (Azadirachta indica), although the actual exposure time was not listed. Nevertheless, across a range of hormone solution concentrations (250; 500; 750; 1000; 1500 mg L⁻¹), 250 mg L⁻¹ IBA provided the best results, e.g., the highest rooting success (80%) and highest number of roots (71), compared to other hormone concentrations [38]. In the current study, Kroin’s [37] “long basal soak” method was chosen over more time-efficient methods because it was suspected that bacterial infection of cuttings would be less likely to occur in seconds or minutes of inoculation.

In order to assess whether R. rhizogenes improved the rooting performance of cannabis cuttings, it was essential to determine if inoculation had led to the infection of the cuttings. A total of 12 out of 32 (38%) cuttings tested showed bands for EF1α and rolB but not for virD, thus meeting the criteria for inferring that an infection, leading to hairy root formation, had occurred (Figure 4). It could, perhaps, be beneficial to evaluate the infection success over gradients of bacterial concentrations or durations of the inoculation step. In Arabidopsis thaliana, transformation with Agrobacterium tumefaciens (strain GV3101(pmP90)) has been attempted at a range of bacterial concentrations (OD₆₀₀ ≤ 0.15–1.75), with little reported effects on transformation rates (0.21 ± 0.05–0.57 ± 0.15%) [39]. Bacterial cultures used for that experiment were grown to the stationary phase (ca. 18–24 h growth) and subsequently diluted to different concentrations, and an additional late stationary phase culture (84 h growth; adjusted to OD₆₀₀ ≤ 0.8) transformed as efficiently (0.50 ± 0.05%) as the younger cultures [39].

The current work suggests that cultivars “Herijuana”, “Bruce Banner”, “Motherlode Kush”, and “Wild Thailand” can be infected by R. rhizogenes, whereas “Hindu Kush”, “California Orange”, “The Pure”, and “Big Bud” are less susceptible under the reported conditions (Figure 4). Agrobacterium-mediated transformation of C. sativa was first shown using A. tumefaciens in tissue culture by Feeney and Punja [30] on four fiber type cultivars “Uniko-B”, “Kompolti”, “Anka”, and “Felina-34”. Additional fiber types cultivars “Futura77”, “Delta-405”, and “Delta-Llosa” and drug-type cultivars “CAN0111” and “CAN0221” were successfully transformed with R. rhizogenes (strains A4 and AR10) with varying, though insignificant, successes, ranging from 43–98% [31]. Additionally, “CAN0221” was inoculated with eight wild-type R. rhizogenes strains, which were all able to induce hairy roots on seedlings without statistically detectable differences in susceptibility toward strains [31]. It should be mentioned that despite being susceptible to R. rhizogenes infection, differences in response to infection occur between cultivars in vitro, for example, by frequency of root induction or the number of roots [40].

The rooting success of C. sativa cuttings was used to assess whether all biometric data between repetitions 1 and 2 could be combined. A series of Chi-squared tests provided the basis for combining data across cultivars but not repetitions, suggesting that the conditions between repetitions differed too much to analyze all data collectively (data not shown). In the aeroponic growth units used in this study, cuttings were positioned in close proximity in a fixed grid pattern (≤10 cm between stems of cuttings). As a result, cuttings measuring 130–150 mm in length would most likely benefit from further trimming of fan leaves to avoid excessive shading of one another. Removing leaf tips could also reduce shading, but this was advised against in cannabis cuttings by Caplan et al. [11], who found an 18% decrease in rooting success. In mint (Mentha piperita) cuttings, Kaymak et al. [41] found no significant effect of cutting lengths (10, 15, 20 cm) on rooting success after inoculation with several PGPR, including R. rubi (strain A16). R. rubi inoculation provided the highest rooting success (89%) compared to water controls (78%) [41]. In repetition 2 of the current experiment, 100% rooting success was achieved in all treatments and cultivars with an average initial cutting length of 81 ± 1 mm, suggesting a potential threshold length for cannabis cuttings using aeroponic growth units.

Successful rooting of water-treated cuttings in repetition 2 further suggested that root initiation in vegetative cannabis cuttings will occur under optimal growing conditions. Water-treated cuttings of this experiment required no laboratory facilities or inputs that were incompatible with the official growing framework; therefore, the findings of this work demonstrated a protocol for vegetatively propagating drug-type cannabis in a manner that is suitable for growers in Denmark. This assumes “optimal growing conditions”, which are largely yet to be determined for cannabis grown in indoor, controlled environments [42]. Drawing parallels from hemp production could be biased as it is mostly an outdoor crop grown for its fiber rather than psychoactive compounds [43].

4. Materials and Methods

5. Conclusions

Rooting performance was best in IBA-treated cuttings with increased root length, root dry weight, and root/shoot dry weight ratio in cuttings compared to those treated with R. rhizogenes or water. However, IBA inhibited shoot growth, which could be a problem later during commercial production establishment. Superior shoot performance was obtained in R. rhizogenes (A4) and H₂O-treated cuttings, thus showing advantages for the later establishment of commercial production of cannabis. These results are preliminary as a deeper analysis could be obtained on WinRhizo. A variation in cultivar susceptibility toward R. rhizogenes was noted in 38% of inoculated cuttings that were successfully infected. The aeroponic growth unit was successful in generating rooted cutting in all tested treatments; therefore, it is an alternative to cannabis producers.

Funding Statement

This research received no external funding.

Author Contributions

Conceptualization, B.T.F., J.K.S. and H.L.; methodology, J.K.S.; formal analysis, J.K.S.; writing—original draft preparation, B.T.F. and J.K.S.; writing—review and editing, B.T.F. and H.L.; visualization, J.K.S.; supervision, B.T.F. and H.L. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Raw data available upon request.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

References