Posted on

plant hormone changes in cannabis at seed set

Combined transcriptomic and metabolomic analysis reveals the potential mechanism of seed germination and young seedling growth in Tamarix hispida

Seed germination is a series of ordered physiological and morphogenetic processes and a critical stage in plant life cycle. Tamarix hispida is one of the most salt-tolerant plant species; however, its seed germination has not been analysed using combined transcriptomics and metabolomics.

Results

Transcriptomic sequencing and widely targeted metabolomics were used to detect the transcriptional metabolic profiles of T. hispida at different stages of seed germination and young seedling growth. Transcriptomics showed that 46,538 genes were significantly altered throughout the studied development period. Enrichment study revealed that plant hormones, such as auxin, ABA, JA and SA played differential roles at varying stages of seed germination and post-germination. Metabolomics detected 1022 metabolites, with flavonoids accounting for the highest proportion of differential metabolites. Combined analysis indicated that flavonoid biosynthesis in young seedling growth, such as rhoifolin and quercetin, may improve the plant’s adaptative ability to extreme desert environments.

Conclusions

The differential regulation of plant hormones and the accumulation of flavonoids may be important for the seed germination survival of T. hispida in response to salt or arid deserts. This study enhanced the understanding of the overall mechanism in seed germination and post-germination. The results provide guidance for the ecological value and young seedling growth of T. hispida.

Background

Seed germination and post-germination (early seedling growth) are controlled by various environmental factors and are important stages for the survival of higher plants under ambiently suitable environment [1]. Many physiological and morphological studies have been conducted on related processes, including plant pigment regulation [2,3,4], abiotic stress [5,6,7,8] and plant hormone regulation [9,10,11]. These stages require a large amount of energy and nutrients which can only be obtained from seed reserves because the germinated seeds cannot absorb minerals and produce energy through photosynthesis [12]. Seed germination begins with the water absorption by the stationary dry seed and is completed by radicle protrusion through the surrounding germ tissues. This process involves a series of orderly physiological and morphogenetic processes, such as energy conversion, nutrient consumption and metabolite changes [13]. After dry seeds have absorbed water, glycolysis, pentose phosphate pathways, and tricarboxylic acid (TCA) cycles are activated. Glycolysis and the TCA cycle provide most of the energy for seed germination [14]. Seed germination and early seedling growth are regulated through a complex network of signalling and gene expression regulation. For example, seed germination can be regulated by multiple plant hormones [15, 16], such as the antagonistic action of abscisic acid (ABA) and gibberellin (GA) [17] and the interaction of jasmonic acid (JA), indole-3-acetic acid (IAA) and other phytohormones [18].

Different plants may have similar molecular mechanisms, including plant hormonal behaviour, transcription and translation activation and radicle protrusion. However, different plant species also have some unique mechanisms, especially for reserve mobilisation and metabolic activation. These differences may be attributed to their different seed stocks [19]. With the rapid development of system biology and high-throughput sequencing, multi-omics technology has become an indispensable research tool in life science [20, 21]. This method can extensively analyse the cell life cycle in all aspects by identifying gene transcripts, metabolites and protein changes throughout growth and development from cell to tissue and the individual itself to understand the complex mechanism of plants and animals.

One of the extremely saline/alkali-tolerant plants, Tamarix hispida Willd. is a typical woody halophyte that forms a natural forest in 1% saline–alkali soil of desert environments. This species is also tolerant to drought stress and thus is an ideal material for cloning drought and saline tolerance-related genes and studying the saline tolerance mechanism of woody halophytes [22]. ThSAP30BP may play an important physiological role in the salt tolerance of T. hispida [23]. The 2-Cys peroxidase gene of this plant improves its tolerance to salt stress [24]. ThNAC7 induces the transcriptional levels of genes associated with stress tolerance to enhance salt and osmotic tolerance by increasing osmotic potential and enhanced ROS scavenging capability [25]. ThMYB13 may also play a role in salt stress tolerance in this transgenic plant [22]. The bZIP protein Thbzip1 of T. hispida is an ACGT elemental binding factor that enhances abiotic stress signalling in transgenic Arabidopsis thaliana [26]. However, the molecular mechanisms underlying the seed germination and post-germination of T. hispida have not been reported. Transcriptomic sequencing is an effective tool to understand complex molecular regulatory mechanisms and provide a new perspective on T. hispida seed germination. At present, this method is routinely used as an experimental platform and has made important contributions to the discovery and identification of genes involved in metabolic pathways [27]. Metabolomics can reveal the terminal products of the signalling pathway and consequently reflect the physiological state of an organism at a specific time. Metabolomes are quite similar to phenotypes and thus can provide detailed information about the intracellular activities regulated by metabolites. Therefore, metabolomics has been widely used in model plants and crops, such as A. thaliana [28,29,30], rice [31,32,33], soybean [34, 35] and many other medical plants [36,37,38].

In this study, the mechanisms underlying the germination and post-germination of T. hispida were investigated using an integrative transcriptomic and metabolomic approach. Gene Ontology (GO) enrichment analysis of differentially expressed genes (DEGs) indicated the differential involvement of biological processes in the six stages of seed germination and post-germination. The dynamic changes of the key genes in flavonoid biosynthesis and phytohormone-related were observed also observed. The Combined transcriptomic and metabolomic analysis showed the accumulation of flavonoids in post-germination. The results provide a valuable reference for the study of seed germination and its functions and effects on saline/alkali-tolerant plants.

Results

Stage definition and transcript assembly of seed germination and post-germination in Tamarix hispida

Seed germination can be divided into the following three stages according to Bewley’s definition [12]: rapid water imbibition, limited water absorption and increase in water uptake accompanied by embryonic axis elongation. For T. hispida, the dry seeds were labelled as stage 1 at 0 h, followed by the rapid increase in water uptake at 0.5 h as stage 2, a slow increase in water uptake at 5 h as stage 3 and hypocotyl extension period at 24 h as stage 4 (Fig. 1a and Additional file 1: Table S1). Post-germination comprised stage 5 with cotyledon unfolding at 144 h and stage 6 with four-true-leaf unfolding at 288 h (Fig. 1b). All these defined stages are shown in Fig. 1b.

Schematic of the seed germination and fresh weight curve of water absorption in Tamarix hispida. a Curve of fresh weight of seeds with time after water absorption. b Representation of morphological changes in seed germination and young seedling growth

Eighteen samples from six stages (1–6, three replicates for each stage) were sampled for RNA-seq. A total of 866,487,598 high-quality reads with Q30 higher than 90% were obtained after quality control, and 75,249 unigenes were de novo assembled using Trinity software. The length was over 1000 bp for 36,531 unigenes (48.6%) and over 1500 bp for 23,089 (30.1%) unigenes. The average gene length and N50 length were 1502 and 2031 bp, respectively. Functional unigene annotation was performed and mapped to NCBI non-redundant (Nr) (48,606, 64.6%), eggNOG database (39,772, 52.9%), Swiss-Prot (33,428, 44.4%), Kyoto Encyclopedia of Genes and Genomes (KEGG) (12,510, 16.6%) and GO (26,371, 35%) (Additional file 2: Fig. S1). A summary of RNA-seq data is shown in Table 1.

Differential expression and functions of the genes involved in seed germination and post-germination

Pairwise differential expression profiling analysis was conducted with the threshold of FDR ≤ 0.05 and absolute value fold change ≥ 2.0 using DEseq2 software to investigate the molecular basis of seed germination and post-germination [39]. Various numbers of DEGs among neighbouring stages were identified as follows: 1220 between stages 1 and 2 (736 up, 484 down), 1284 between stages 2 and 3 (830 up, 454 down), 19,439 between stages 3 and 4 (9206 up, 10,233 down), 12,728 between stages 4 and 5 (6458 up, 6270 down) and 955 between stages 5 and 6 (547 up, 408 down) (Fig. 2a, Additional file 1: Table S2). This result showed that the number of DEGs in the hypocotyl extension period (stage 3 versus stage 4) of seed germination and the cotyledon unfolding period (stage 4 vs. stage 5) of seed post-germination was higher than that in the other comparison groups.

Differential expressed genes (DEGs) during the six stages of Tamarix hispida seed germination and post-germination. a Numbers of up-regulated and down-regulated DEGs at adjacent stages. Enriched DEGs from Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. b Enriched KEGG pathways of DEGs between stages 1 and 4. c Enriched KEGG pathways of DEGs between stages 4 and 5. d Enriched KEGG pathways of DEGs between stages 5 and 6

GO and KEGG enrichment analysis for adjacent compared groups reflected the physiological response of germination and post-germination. For example, the most enriched DEGs were identified in the GO term of ‘cellular response to osmotic stress’ in the comparison between stages 1 and 2. Meanwhile, the most enriched DEGs were observed in the KEGG pathway of ‘metabolism of xenobiotics by cytochrome P450’ in the comparison between stages 1 and 2, indicating that cytochrome P450 may be important for the rapid increase in water uptake (Additional file 2: Fig. S2 and S3). Given that seed germination occurs from stage 1 to stage 4, further KEGG enrichment analysis was performed on 18,975 DEGs (stage 1 vs. stage 4). Among these enriched pathways, ‘phenylpropanoid biosynthesis’ (ko00940), ‘plant hormone signal transduction’ (ko04075) and ‘flavonoid biosynthesis’ (ko00941) were the top three significantly enriched categories (Fig. 2b, Additional file 1: Table S3) and thus have important roles in the seed germination of T. hispida. At early young seedling during post-germination, ‘phenylpropanoid biosynthesis’ and ‘flavonoid biosynthesis’ were the most significantly enriched in the periods of cotyledon unfolding (part I, stage 4 vs. stage 5) and true-leaf unfolding (part II, stage 5 vs. stage 6). Among the separate top five enriched pathways, ‘plant–pathogen interaction’ and ‘plant hormone signal transduction’ were significantly enriched in part I, and ‘flavone and flavonol biosynthesis’ and ‘anthocyanin biosynthesis’ were significantly enriched in part II (Fig. 2c and d, Additional file 1: Table S3). The enrichment of ‘anthocyanin biosynthesis’ term in true-leaf unfolding period indicated anthocyanin accumulation, which was consistent with the morphology of extremely dark red stem and root compared with those during cotyledon unfolding (Figs. 1b and 2d). These results showed that hormone signal transduction and flavonoid biosynthesis may be important for the seed germination and post-germination of T. hispida.

Dynamic regulation of key gene families in flavonoid biosynthesis during seed germination and post-germination

The phenylpropanoid pathway is the upstream branch of flavonoid biosynthesis; hence, DEGs involved in phenylpropanoid biosynthesis (ko00940) and flavonoid biosynthesis (ko00941) were identified during the seed germination and post-germination of T. hispida. A total of 158 DEGs in phenylpropanoid pathway, such as genes encoding phenylalaninammo-nialyase (PAL), 4-coumarate-CoA ligase (4CL) and trans-cinnamate 4-monooxygenase (CYP73A), were overall up-regulated after stage 3 (Additional file 2: Fig. S4). Flavonoid biosynthetic genes were divided into two categories, namely, early biosynthetic genes (EBGs) responsible for the production of common precursors and late biosynthetic genes (LBGs) for the eventual products. The former mainly encodes naringenin-chalcone synthase (CHS), chalcone isomerase (CHI), naringenin 3-dioxygenase (F3H), flavonoid 3′-monooxygenase (F3’H) and flavonoid 3′,5′-hydroxylase (F3’5’H), and the latter mainly encodes flavonol synthase (FLS), dihydrokaempferol 4-reductase (DFR), anthocyanidin synthase (ANS), anthocyanidin reductase (ANR) and leucoanthocyanidin reductase (LAR) [40, 41]. Finally, 41 DEGs were identified in flavonoid biosynthesis pathway, including 35 EBGs and six LBGs (Fig. 3a, Additional file 1: Table S4). A clustering heatmap of gene expression divided these DEGs into six clusters. Except ThOMT-2 and ThF3’H-2 of cluster3, all of these genes were up-regulated upon hypocotyl extension. The EBGs of flavonoid biosynthesis were distributed in all clusters, and the LBGs were significantly enriched in cluster6 (p-value = 0.035) (Fig. 3b). The cluster4 genes were upregulated in stage 3, and the cluster6 genes were upregulated in stages 4 and 6. Further validation of ThCHS2 (TRINITY_DN7978_c0_g1) of cluster4 and ThCHS3 (TRINITY_DN8_c1_g1) of cluster6 by real-time quantitative reverse transcriptional PCR (qRT-PCR) experiment indicated the differential regulation of CHS genes (Fig. 3c). Compared with those in seed germination, more genes were involved in the phenylpropanoid and flavonoid biosynthesis pathways of post-germination and further contributed to flavonoid accumulation. This high level of flavonoid would help young seedlings to survive from extreme drought desert environments.

Analysis of DEGs in flavonoid biosynthesis pathway. a Mapping of enriched DEGs in flavonoid biosynthesis pathway (ko00941) [42]. The orange marks represent DEG enrichment. b Expression pattern of DEGs involved in flavonoid biosynthesis at different stages. c FPKM from RNA-seq and expression level from qRT-PCR of two chalcone synthase (CHS) genes. The expression at stage 1 was set as 1, and the relative expression level was calculated for several genes

Dynamic regulation of phytohormone-related DEGs during seed germination and post-germination

A total of 144 DEGs were significantly enriched in ‘plant hormone signal transduction’ (ko04075) in seed germination and post-germination, suggesting their important roles in the seed germination of T. hispida. Further classification of these hormone-related DEGs showed that 61 were involved in auxin (AUX) pathway, five in GA pathway, 45 in ABA pathway, eight in ethylene (ETH) pathway, eight in JA pathway and 17 in salicylic acid (SA) pathway (Additional file 1: Table S5). The 144 phytohormone-related DEGs were further divided into five expression clusters. Among which, cluster1 genes were upregulated in stages 1–3, cluster2 genes in stages 1–4, cluster3 genes only in stage 4, cluster4 genes in stages 4–6, and cluster5 genes in stages 5 and 6 (Fig. 4a).

Analysis of DEGs in plant hormone signal transduction a Heatmap of DEG expression of auxin, gibberellin (GA), abscisic acid (ABA), ethylene (ETH), jasmonic acid (JA) and salicylic acid (SA) signalling pathways at different stages. b DEG expression enriched in auxin, GA, ABA, ETH, JA and SA signalling pathways [42]; expression levels are indicated by the heatmap for different groups. The orange marks represent DEG enrichment. c FPKM from RNA-seq and expression level from qRT-PCR of phytohormone-related DEGs. The expression at stage 1 was set as 1, and the relative expression level was calculated for several genes

The DEGs (14/61) of AUX pathway were significantly enriched in cluster3 (17 genes) (p-value = 0.0005) and upregulated in stage 4, indicating the important role of auxin signalling in hypocotyl elongation. The DEGs (24/45) of ABA pathway were significantly enriched in cluster1 (45 genes) (p-value = 0.00016) and consistently upregulated in stage 1–3, implying the regulatory function of ABA in seed germination such as water uptake. The JA (6/8) and SA (12/17) genes were enriched in cluster5 (54 genes) (JA: p-value = 0.032, SA: p-value = 0.0035) and upregulated in stages 5 and 6 of post-germination. For example, jasmonate ZIM domain-containing protein (JAZ), NONEXPRESSER OF PR GENES 1 (NPR1) and transcription factor TGA (TGA) were significantly up-regulated during cotyledon expansion and true-leaf unfolding. The expression patterns of the identified phytohormone-related DEGs were mapped in plant hormone signal transduction (Fig. 4b). Five genes [IAA (TRINITY_DN37691_c0_g1), auxin responsive GH3 gene family (TRINITY_DN38209_c0_g1), PP2C (TRINITY_DN4920_c0_g1), serine/threonine-protein kinase SnRK2 (TRINITY_DN1976_c0_g1) and ABF (TRINITY_DN7836_c2_g100)] were randomly selected for qRT-PCR confirmation to further validate their differential transcription (Fig. 4c). The results suggested that different plant hormones play varying roles in the seed germination and post-germination of T. hispida.

Co-expression network analysis with WGCNA

A global co-expression network with all DEGs (46,538 genes) was constructed by weighted gene co-expression network analysis (WGCNA) to identify additional potential genes involved in seed germination and post-gemination. Thirteen co-expressed modules were obtained on the basis of the network concept (Fig. 5a), and stage-associated modules were identified by correlation analysis between module eigengenes and each stage (Fig. 5b). Among the positive stage-correlated modules with Pearson correlation coefficient (PCC) ≥ 0.60, three modules of ‘midnightblue’, ‘darkorange’ and ‘skyblue’ were correlated with stage 1; three modules of ‘white’, ‘cyan’ and ‘purple’ with stage 2; four modules of ‘darkturquoise’, ‘salmon’, ‘black’ and ‘darkred’ with stage 4; three modules of ‘green’, ‘lightcyan’ and ‘steelblue’ with stage 5; and one module of ‘green’ with stage 6. Except for an invalid ‘grey’, no module was correlated with stage 3 (Fig. 5b). GO and KEGG enrichment analyses of the gene sets merged from specific stage-associated modules were performed to further investigate the biological processes or pathways of the stage-associated modules. Different processes or pathways were enriched in germination and post-germination stages. In seed germination, ‘transposition, RNA-mediated’ GO term, and ‘apoptosis’ KEGG pathway were enriched in stage 2. Multiple GO processes, such as ‘translation’, ‘tricarboxylic acid cycle’ and ‘chloroplast organisation’ were enriched in stage 4. GO terms of ‘carbamoyl-phosphate synthase (glutamine-hydrolysing) activity’ and ‘protein autophosphorylation’ were the most enriched in stages 5 and 6, respectively. (Additional file 2: Fig. S5, Additional file 1: Table S6). Therefore, the co-expressed genes involved in multiple biological processes during stages 2 and 4 were essential for the seed germination and early seeding growth of T. hispida.

Weighted gene co-expression network analysis (WGCNA) of genes during the seed germination and young seedling growth of Tamarix hispida. a Dendrogram showing the co-expression modules identified by WGCNA across seed germination and post-germination. The major tree branches constitute 14 modules labelled with different colours. b Module-stage association (each row corresponds to a module, and each column represents a specific stage. The colour of each cell at the row-column intersection indicates the correlation coefficient between a module and the stage). Red indicates the positive correlation between the module and the stage, and blue indicates a negative correlation

Genes from three modules (‘purple’, ‘black’ and ‘green’) were further identified for transcriptomic analysis and visualised to understand the possible regulatory mechanism in the key developmental stages of seed germination and post-germination. GO enrichment analysis of the genes of ‘purple’ module revealed multiple terms related to phytohormone response and abiotic stress response, such as ‘response to heat’, ‘response to auxin’ and ‘response to abscisic acid’ (Additional file 2: Fig. S6, Additional file 1: Table S6). The node gene with a high connectivity within a module is defined as hub gene with an important role in different modules. In the ‘purple’ module, various transcription factors (TFs) with high connectivity were involved in hormone response and abiotic stress response. For example, DEHYDRATION-RESPONSIVE ELEMENT BINDING PROTEIN 2A (DREB2A, ThAP2–7) was involved in seed germination and induced by drought and osmotic stress [43], and DIVARICATA2 (DIV2, ThMYB-138) was required for ABA signalling and response to salt stress in Arabidopsis [44] (Fig. 6a). The ‘black’ module had the highest correlation with hypocotyl elongation (stage 4). Function analysis of high connectivity TFs showed that the hub genes of ‘black’ module were connected with plant organ development and flavonoid biosynthesis. For example, TRANSPARENT TESTA 2 (TT2, ThMYB-52) positively regulated anthocyanin accumulation in hypocotyls [45] and YABBY1 (YAB1, ThC2C2–44) and YABBY5 (YAB5, ThC2C2–4) controlled leaf blade development [46, 47] (Fig. 6b, Additional file 1: Table S6). Meanwhile, ‘phenylpropanoid biosynthesis’, ‘plant hormone signal transduction’ and ‘flavonoid biosynthesis’ were significantly enriched in ‘green’ module (Additional file 2: Fig. S6, Additional file 1: Table S6). Several genes in ‘green’ module encoded by WRKY DNA-BINDING PROTEIN (WRKY) TFs with high connectivity, such as WRKY DNA-BINDING PROTEIN 42 (WRKY42, ThWRKY-93), WRKY DNA-BINDING PROTEIN 33 (WRKY33, ThWRKY-2 and ThWRKY-70) and WRKY DNA-binding protein 4 (WRKY4, ThWRKY-56), were related to response to abiotic stress; their homologs are also involved in response to abiotic stress in Arabidopsis [48,49,50] (Fig. 6c).

Regulatory network of key phytohormones and flavonoid metabolites. Regulatory network of purple, black and green modules shown in a, b and c, respectively. The orange circles represent structural genes involved in plant hormone metabolism, the red circles represent structural genes involved in flavonoid metabolism, and the diamonds with different colours represent different families of transcription factors identified in the same module whose transcripts are correlated with the expression of structural genes

Combination of metabolic profiling and transcriptomic analysis

Many flavonoid-biosynthesis related genes were activated and enhanced in post-germination (Fig. 3). A widely targeted metabolomics analysis was performed on three stages, namely, slow water absorption (stage 3), hypocotyl elongation (stage 4) and cotyledon expansion (stage 5) to further understand the metabolic profiles in post-germination. The samples from three stages showed good triplications according to their principal component analysis (PCA) score plots, indicating that metabolite accumulation is stage-specific (Fig. 7a). A total of 1022 metabolites were identified in all samples. Within the threshold of variable importance in projection (VIP) > 1 and absolute value fold change > 2, the numbers of up-regulated and down-regulated metabolites in stage 3 versus stage 4 and stage 4 versus stage 5 were 205 and 149, 287 and 178, respectively. A total of 651 differentially expressed metabolites (DEMs) were identified and mainly comprised of flavonoids (133), phenolic acids (123), lipids (89), amino acids and derivatives (66), alkaloids (53), organic acids (45) and others (Fig. 7b, Additional file 1: Table S7). As the most abundant differentiated metabolite, the accumulation of flavonoid was consistent with the results of gene expression and functional enrichment in flavonoid biosynthesis during seed germination and post-germination.

Metabolome analysis of Tamarix hispida. a Principal component analysis (PCA) of metabolome data in stages 3, 4 and 5. b Statistics of different metabolites in stage 3 vs. stage 4 and stage 4 vs. stage 5 (VIP > 1, |log2(FC)| > 1). c Venn diagram of flavonoids in differential metabolites. d Schematic of a portion of flavonoid biosynthesis pathway in Tamarix hispida. Metabolic intermediates are marked in blue, and end products are marked in red. Heat maps show the transcriptional level of enzymes in the six stages and metabolite level in the three stages (3, 4 and 5). Gene sale bar corresponds to the range of relative transcriptional level of enzymes, and metabolite sale bar corresponds to the range of relative metabolite level

A combined transcriptomic and metabolomic analysis was performed in stages 3, 4 and 5 to investigate the potential regulation of metabolites. The results showed significantly different accumulation for 43 flavonoids in stage 3 versus stage 4 and stage 4 versus stage 5, 44 flavonoids in stage 3 versus stage 4 and 46 flavonoids in stage 4 versus stage 5 (Fig. 7c). With these identified differential flavonoids, the metabolic intermediates and end products of flavonoid biosynthesis pathway were mapped to a known KEGG pathway. The overall trends of flavonoid biosynthesis suggested a decrease in metabolic intermediates (such as naringenin, dihydrokaempferol, apigenin and hesperetin) but an increase in end products (such as quercetin 3-sulfonate, rhoifolin and baimaside) in seed germination and post-germination (Fig. 7d). The accumulation of end flavonoids in post-germination revealed the consistent upregulation of CHS gene regulation starting at stage 4 (Fig. 3), further indicating that chalcone synthase is a key and limited enzyme for flavonoid biosynthesis in the seed germination of T. hispida. During seed germination (stages 3 and 4), the metabolites located downstream of the flavonoid synthesis pathway, specifically hesperetin7-O-glucoside, quercetin 3-O-[beta-D-xylosyl-(1- > 2)-beta-D-glucoside] and baimaside, were significantly increased. Meanwhile, the metabolites located upstream of the flavonoid synthesis pathway, such as naringenin, dihydrokaempferol and isoquercitrin, were significantly decreased (Fig. 7d, Additional file 2: Fig. S7). The metabolite flow from naringenin to quercetin 3-sulfonate also confirms the increase and decrease in the metabolites located upstream and downstream of the flavonoid biosynthesis, respectively, in young seedling growth (stages 4 and 5). These results suggest that the metabolites located downstream of the flavonoid synthesis pathway are beneficial to seed germination and young seedling growth.

Discussion

T. hispida is a perennial shrub or small tree and a woody halophyte that serves as an excellent model for studies on resistance to abiotic stress. Seeds are at the crucial stage of biodiversity and agriculture for plant survival. If environmental conditions are only suitable for plant growth but not for seed germination and young seedling growth, then plants cannot survive in the community [51]. Therefore, the molecular mechanism of seed germination and young seedling growth in T. hispida must be explored to understand its adaptative evolution in extremely arid deserts.

Shortening the breeding cycle can help to meet the demand for the seedling supply of T. hispida. Transcriptomic and metabolomic analyses were performed to detect the changes in RNA level and metabolome levels during seed germination and young seedling growth. Differential transcriptome expression analysis revealed that the DEGs during seed germination and young seedling growth were significantly enriched in phenylpropanoid biosynthesis, plant hormone signal transduction and flavonoid biosynthesis pathway. Meanwhile, four-true-leaf turning red and anthocyanin biosynthesis pathway were significantly enriched for T. hispida seedlings. Flavones are synthesised by the flavonoid pathway, the downstream of phenylpropanoid biosynthesis [52]. These compounds are important in plant defence against biotic and abiotic stresses, such as oxidative damage [53] and UV stress [54]. Therefore, the DEGs and metabolites of flavonoid biosynthesis play an important role in the seed germination and young seedling growth of T. hispida.

In this study, the flavonoid biosynthesis genes were divided into six expression patterns. CHS is the first committed enzyme in the conserved flavonoid synthesis pathway [55], and three ThCHSs were identified in T. hispida. Clustering heatmap showed that two CHSs (ThCHS1 and ThCHS2) were highly expressed in stages 5 and 6, respectively, and ThCHS3 showed significantly increased expression, albeit in various levels, in stages 4 and 6. The expression model of ThCHSs is consistent with flavonoid biosynthesis in T. hispida, indicating that flavonoids are produced in large quantities during seed germination and young seedling growth. Meanwhile, the early and late genes of flavonoid biosynthesis pathway showed different expression trends. ThANR-1, ThLAR-1, ThDFR-1 and ThDFR-2 are the downstream genes of flavonoid biosynthesis pathway and were highly expressed in stages 4 and 6. However, some of the early genes of flavonoid biosynthesis pathway such as ThCHS-1, ThCHS-2, ThCHI-2 and ThF3H-2 were not expressed in stage 4. Therefore, the expression of flavonoid biosynthesis DEGs undergoes complex regulation during seed germination and young seedling growth.

Seed germination and young seedling growth are important processes affecting crop production and are influenced by a range of factors, including plant hormones [15].. The most important plant hormones for seed germination are ABA and GA, which have inhibitory and stimulatory effects on seed germination, respectively. In this study, ABA and GA expression significantly varied in seed germination and young seedling growth. For example, ThABFs and ThPP2Cs were down-regulated in seed germination but up- regulated in young seedling growth. Meanwhile, ThPIF and ThGID2 were up- regulated in seed germination. These findings indicate that ABA and GA play important roles in seed germination and young seedling growth. Auxin by itself is not a necessary hormone for seed germination but can interact with other hormones to affect seed germination and young seedling growth. For example, the release of ARF from repression by miRNA also affects ABA sensitivity during young seedling growth [56]. In addition, the DEGs of auxin pathway are vital for young seedling growth. As stress-response hormones, the upregulation of JA and SA pathways can be helpful to flavonoid accumulation during post-germination [57].

The ‘purple’ module was the largest module in WGCNA (19,095 genes, 41.03%) and was the most relevant to seed water uptake (stage 1: dry seed, stage 2: rapid water absorption, stage 3: slow water absorption). In this module, phytohormone and abiotic stress response related terms were significantly enriched, thereby supporting the importance of phytohormone in seed germination. The ‘green’ module (12,200 genes, 26.22%) is the second largest module, and its genes were positively relevant to post-germination (stage 5: cotyledon expansion, stage 6: four true leaf). Enrichment analysis showed that ‘phenylpropanoid biosynthesis’, ‘flavonoid biosynthesis’ and ‘plant hormone signal transduction’ were significantly enriched in this module, indicating that flavonoid is important to post-germination. As a transition node from seed germination to seedling development, hypocotyl elongation (stage 4) was significantly associated with four modules (‘darkred’, ‘black’, ‘salmon’ and ‘darkturquoise’). Among these four modules, ‘black’ (5221 genes, 11.22%) was the largest and the most relevant. ‘Phenylpropanoid biosynthesis’ and ‘flavonoid biosynthesis’ were also significantly enriched in this module, and its high connectivity node functional TFs were associated with plant organ development and flavonoid biosynthesis. These results further support the importance of flavonoids and phytohormones.

Conclusions

In this study, the molecular regulation of seed germination and post-germination in T. hispida was investigated using an integrated transcriptional and metabolomic method. GO and KEGG analysis showed that the pathways of plant hormone signal transduction, phenylpropanoid biosynthesis, and flavonoid biosynthesis were significantly enriched and thus have important roles in the two developmental periods. The gene families involved in the plant hormone pathway that was enriched in different expression clusters showed differential regulation in seed germination and post-germination, such as ABA for early germination stages, auxin for hypocotyl elongation, JA for cotyledon expansion and SA for four-true-leaf unfolding. Metabolomics showed that the final products of flavonoids such as 3 − O−[beta−D − xylosyl−(1− > 2) − beta−D − glucoside], quercetin, baimaside, rhoifolin, hesperetin− 7 − O − glucoside and quercetin− 3 − O − sulfonate accumulated in post-germination. RNA-seq and metabolomic analysis indicated the importance of flavonoid biosynthesis pathways and identified CHS as the key enzyme for the accumulation of final flavonoid products in post-germination. In addition, organic acids were involved in seed germination. All these results provide important insights into the cellular and metabolic changes underlying T. hispida seed germination and young seedling growth.

Materials and methods

Plant materials, experimental conditions and fresh weight measurements

The seeds of T. hispida were obtained from Alar City, Xinjiang Uygur Autonomous Region of China. The hairs of selected seeds were removed, and germination test was carried out in three replicates (100 seeds per replicate). The seeds were incubated in 3 mL of ultrapure water on two sheets of absorbent paper in a covered glass petri dish at 25 °C. At ambient temperature of 25 ± 1 °C, 100 germinated seeds (± 0.0001 g) were weighed every 15 min at the stage of rapid water absorption between 0 and 1 h. After sampling at a specific time, the excess water was immediately sucked up with absorbent paper, and the samples were quickly frozen in liquid nitrogen and stored at − 80 °C in a refrigerator for transcriptomic and metabolomic analyses. We declare that the research programme complies with relevant institutional, national and international guidelines and legislation, and we have permission to collect T. hispida seeds.

RNA isolation and cDNA library construction for RNA sequencing

Total RNA of samples was isolated and purified using Trizol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s procedure. The RNA was purified and quantified using NanoDrop 2000 (NanoDrop, Wilmington, DE, USA), and its integrity was assessed by Agilent 2100 with RIN number > 7.0. Total RNA was used to construction the RNA-seq libraries: mRNA was enriched from the total RNA using oligo (dT) magnetic beads, and the final average insert size for the final cDNA library was 350 bp (± 50 bp). Paired-end sequencing was performed on an Illumina Hiseq X-Ten (LC Bio, China) following the vendor’s recommended protocol.

Cutadapt [58] was used to remove the reads containing adaptor contamination, low quality bases and undetermined bases. Sequence quality was then verified using FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). After the adaptor and low-quality sequences were removed, the clean reads were assembled into expressed sequence tag clusters (contigs) and de novo assembled into transcripts by using Trinity [59] in the paired-end method. On the basis of the similarity and length of the sequence, the longest transcript was selected as a single gene for subsequent analysis.

Functional annotation of DEGs

The function of the unigenes was annotated by alignment with the NCBI non-redundant (NR) and SwissProt databases using Blastx (v2.10.1) with a threshold e-value of 10 − 5 . The proteins with the highest hits to the unigenes were used to assign functional annotations. With SwissProt annotation, GO classification was performed by mapping the relation between SwissProt and GO terms. The unigenes were mapped to the KEGG database to annotate their potential metabolic pathways [42, 60, 61] using eggNOG-mapper (v2.0.1–14) [62].

All clean reads were aligned onto pomegranate genome. The trimmed mean of M values (TMM) method was used to calculate the unigene expression abundance. DEGs were detected by DESeq2 (v1.22.2) [39] with the absolute value of log2(fold change) value > 1 and a false discovery rate (FDR) < 0.05 as selection criteria. GO enrichment and KEGG pathway enrichment analyses of DEGs were performed using R (v4.0.3) package clusterProfiler (v3.18.1) [63].

Weighted correlation network analysis and gene network visualisation

Co-expression networks were constructed using the WGCNA (v1.69) package in R (v3.6.1) [64] with the soft threshold = 12 and the minModuleSize = 400. Eigengene values were calculated for each module and used to test associations with each germination stage. Networks were visualised using Cytoscape v.3.8.2 [65].

Metabolomic analysis

The sample preparation, extract analysis, metabolite identification, and quantification were performed at Wuhan Metware Biotechnology Co., Ltd. (Wuhan, China) (http://www.metware.cn/) following their standard procedures [66, 67].

Biological samples were freeze-dried by vacuum freeze-dryer (Scientz-100F) and then crushed using a mixer mill (MM 400, Retsch) with a zirconia bead for 1.5 min at 30 Hz. Briefly, 100 mg of lyophilised powder was dissolved in 1.2 mL of 70% methanol solution and vortexed for 30 s for every 30 min for six times in total. The sample was placed in a refrigerator at 4 °C overnight. Following centrifugation at 12,000 rpm for 10 min, the extracts were filtrated (SCAA-104, 0.22 μm pore size; ANPEL, Shanghai, China, http://www.anpel.com.cn/) for UPLC-MS/MS analysis.

The sample extracts were analysed using an UPLC-ESI-MS/MS system (UPLC, SHIMADZU Nexera X2, https://www.shimadzu.com.cn/; MS, Applied Biosystems 4500 Q TRAP, https://www.thermofisher.cn/cn/zh/home/brands/applied-biosystems.html) under the following analytical conditions: UPLC column, Agilent SB-C18 (1.8 μm, 2.1 mm * 100 mm); and mobile phase, solvent A of pure water with 0.1% formic acid and solvent B of acetonitrile with 0.1% formic acid. Sample measurements were performed with a gradient program as follows: initial gradient of 95% A, 5% B; a linear gradient to 5% A within 9 min; 5% A, 95% B for 1 min; and adjusted 95% A, 5.0% B within 1.1 min and kept for 2.9 min. The flow velocity was set as 0.35 mL per minute, the column oven was set to 40 °C, and the injection volume was 4 μL. The effluent was alternatively connected to an ESI-triple quadrupole-linear ion trap (QTRAP)-MS.

LIT and triple quadrupole (QQQ) scans were acquired on a triple quadrupole-linear ion trap mass spectrometer (Q TRAP), AB4500 Q TRAP UPLC/MS/MS System, equipped with an ESI Turbo Ion-Spray interface, operating in positive and negative ion modes and controlled by Analyst 1.6.3 software (AB Sciex). The ESI source operation parameters were as follows: ion source, turbo spray; source temperature 550 °C; ion spray voltage (IS) 5500 V (positive ion mode)/− 4500 V (negative ion mode); ion source gas I (GSI), gas II (GSII) and curtain gas (CUR) were set at 50, 60 and 25.0 psi, respectively; and high collision-activated dissociation (CAD). Instrument tuning and mass calibration were performed with 10 and 100 μmol/L polypropylene glycol solutions in QQQ and LIT modes, respectively. QQQ scans were acquired as MRM experiments with collision gas (nitrogen) set to medium. DP and CE for individual MRM transitions were conducted with further DP and CE optimisation [67]. A specific set of MRM transitions were monitored for each period according to the metabolites eluted within this period.

Significant differences in relative metabolite content were tested by orthogonal partial least squares discriminant analysis (OPLS-DA) with a threshold of variable important in projection (VIP) value > 1 and an absolute value of absolute value of fold change value > 2 using R software version 4.0.3 (http://www.r-project.org/).

qRT-PCR analysis

Seven DEGs were selected for qRT-PCR analysis with actin (β-actin) as internal reference gene. Primers were designed through NCBI official website (https://www.ncbi.nlm.nih.gov/) and are listed in Additional file 1: Table S8.

The RNA extracted from T. hispida was used to synthesise first-strand cDNA with SweScript RT I First Strand cDNA Synthesis Kit following the manufacturer’s instructions. qRT-PCR was performed with a 2*Universal Blue SYBR Green qPCR Master Mix kit in accordance with the manufacturer’s instructions. The experimental conditions were set as follows: 40 cycles at 95 °C for 30 s (predegeneration), 95 °C for 15 s, 60 °C for 10 s and 72 °C for 30 s. The mRNA expression level of the genes was calculated with the 2 −ΔΔCt method. Each plant sample was analysed three times (each replicate contained three technical replicates).

What are Plant Growth Hormones?

Here’s a crash course on the five different classes of plant hormones, and why knowing a bit about them and their uses can improve your yields.

I am sure you’ve heard the old saying about how one bad apple spoils the bunch? Well, it’s true, and the cause is a hormonal imbalance.

In nature, the first ripe apple of the season drops to the ground and begins to decompose. During the decomposition process, the apple releases a gas called ethylene. Ethylene is a plant growth hormone, or phytohormone, that triggers nearby apples to ripen and fall to the ground.

The sweet smell of all those decomposing apples attracts foraging animals, who eat the apples and spread the seeds far and wide. Ethylene and other plant growth hormones are vital to all aspects of plant growth and development, and a little knowledge and understanding about them and their uses can improve your yields.

When a plant sets flowers, the direction it stretches, the size of its fruits, when it drops them to the ground and virtually every other aspect of plant development is controlled by hormones.

Environmental signals trigger production of these powerful compounds. The hormones are produced either in the leaves, flowers, shoots, roots or fruits, depending on which hormone is in production.

They are made on the smooth endoplasmic reticulum within plant cells, then transported through the cell walls and circulated throughout the plant. Whether part of the normal life cycle or in response to stress, the interacting plant hormones are responsible for all growth changes.

Plant hormones are grouped into five classes depending on their chemical makeup and what they cause to occur, or prevent from occurring: abscisic acid, auxins, cytokinins, gibberellins and ethylene.

The plural is used when listing three of these agents of change because there is not one single molecule, but a group of them that have similar functions and molecular compositions. Let’s look closer at each class of hormone.

Abscisic Acid

Abscisic acid (ABA) is Mother Nature’s natural timer. ABA builds up in developing seed coats, so when a seed falls to the ground, the slow dissipation of abscisic acid causes the seed to break dormancy. This is vital in frozen climates because if a seed were to germinate too early, set roots and begin to grow, it would perish.

Certain plants, such as pines, have high amounts of ABA stored in their seed coats and need to be stratified (forced germination by mimicking winter conditions) for several months prior to germination. Some short-lived annuals, on the other hand, have low levels of ABA and can pop overnight.

ABA also helps regulate the respiration process during times of drought. In a back-and-forth communication between roots and leaves, ABA is produced and used to modify potassium and sodium levels in the guard cells, causing the stomata to close and the plant to save water. (See: 10 Facts on Abscisic Acid)

Auxins

Next on the roster are the powerful root, shoot and fruit regulators known as auxins. High auxin levels result in cell wall plasticity, allowing growing cells to stretch out. Bigger cells mean bigger tissues, which result in bigger organs, and bigger organs result in larger fruits and fantastic flowers.

Auxins are also responsible for phototropism, or the way plants grows towards the light. By regulating which cells elongate and which don’t, the plant is able to grow directionally. (See: 10 Facts on Auxins)

Cytokinins

Cytokinins are as important as auxins, especially considering levels of both are kept relatively even. A simplified explanation: if one level is 50%, the other level is 50%. If one level rises to 60%, the other drops to 40%. If cytokinin levels are low, the plant produces vegetative growth.

As cytokinin levels increase—and auxin levels decrease—a plant transitions into the flowering stage. Higher cytokinin levels cause plants to grow bushier, with shorter internodal spacings. There is a good chance the secret ingredient in your favorite fertilizer is a mix of auxins and cytokinins, balanced precisely to induce your garden to grow one way or another. (See: 10 Facts on Cytokinins)

Ethylene

The next plant hormone is ethylene, a gas produced as pectin breaks down in the cell walls of ripening or rotting fruits. The release of ethylene gas by one rotten apple triggers nearby apples to ripen prematurely, spoiling the entire bunch.

Ethylene also plays a role in phototropism, stem growth—lower levels correspond to thicker stem growth—and takes part in the initiation of leaf development. (See: 10 Facts on Ethylene)

Gibberellins

Last but not least are gibberellins, or gibberellic acid (GA). This class of hormones has a lot of responsibilities. Gibberellins cause seeds to start growing after germination and help seedlings manage food storage while still developing photosynthetic leaves. During vegetative growth, GA also causes stretching, or large internodal spacing.

Flowering plants affected by the daylight period length are induced to flower by adjusting GA levels. Gibberellins are often used in growth-promoting products. (See: 10 Facts on Gibberellins)

Making Plant Hormones Work for You

These five classes of hormones work synergistically to trigger all the necessary physiological processes plants undergo to complete their life cycles and ensure another generation.

Cytokinins and auxins balance out the switch from vegetative to flowering growth, gibberellins and abscisic acid work together to promote heavier fruiting, and ethylene and auxins coordinate to cause the dropping of leaves.

Low doses of gibberellic acid promotes growth, while high levels inhibit it. A cocktail of chemicals is constantly flowing through a growing plant, the recipe of which is ever-fluctuating.

Speaking of recipes, how about the ones that are in some of your favorite grow products? Often the true active ingredient is not listed on the label, which is why that bottle of potassium sulfate seems to make magic happen.

Plant growth hormones are sometimes listed on the label; other times, they are not. If a hormone is made synthetically, it is called a plant growth regulator, or PGR.

Two common PGRs seen on product labels are the auxins indole-3-butyric acid (IBA) and napthaleneacetic acid (NAA), which you have likely used in rooting hormone products. (Read More: The Science Behind Root Growth Promoters)

These two PGRs mimic the natural hormone IAA (indole-3-acetic acid), and initiate the formation of a callus and then root development. Using a product with IBA or NAA will ensure that time spent taking and rooting cuttings is not wasted.

PGRs can be used like a tools in a tool box, adjusting a garden’s growth however a grower sees fit. A common problem indoor and greenhouse growers face is running out of room, which can be remedied by using PGRs that inhibit stem elongation.

You may have heard of products that use paclobutrazol, flurprimidol or trinexapac-ethyl, which stop stems from stretching by inhibiting gibberellin biosynthesis.

Depending on the timing of their use, you can keep vegetative plants shorter for longer, or make a flowering plant produce short, tight internodes with increased lateral branching. PGRs can be useful tools if used properly, but due to health and safety reasons, some PGRs are meant to be used on ornamental crops only.

If you decide to use them, make sure they are safe to use on your intended crop, use the recommended safety equipment, pay attention to re-entry periods, and do a post-harvest rinse of all produce.

If you want the benefits of a PGR, but don’t want to use synthetic chemicals, Mother Nature has it all figured out. Potent phytohormones are produced in plants, fungi and algae. They can be directly applied to the garden, or extracted and concentrated into easy-to-use liquids and powders. Several are in products you may already be using.

Willow bark powder is a great natural rooting hormone due to the high amounts of salicylic acid present in the bark, which promotes root initiation. This natural miracle worker also plays a part in inducing systematic acquired resistance, causing a plant to bulk up its entire defense system, and reducing chances of future disease or pest problems. A foliar spray of willow bark water will toughen up your plants and keep them stronger for longer.

Another natural plant growth hormone source is yeast, which produces the auxin indole-3-acetic acid. Yeast extracts are probably on the list of unlisted ingredients in products that make your garden blast off. Sprouted seed teas (SSTs) are starting to gain popularity amongst probiotic farmers.

A sprouting seed is packed with abscisic and gibberellic acids, as well as a bunch of other bioactive enzymes and beneficial proteins. These teas are made by soaking seeds, often barley or rye, in water until they sprout their radicle (the first bit of root that emerges from a germinating seed), blending the sprouting seeds into a slurry, or just collecting the water they are soaked in.

This biologically active liquid can be used in a root drench or foliar spray. Plants will respond to the hormones with root and shoot development, cell elongation and heavy flowering.

Another common source of plant hormones is kelp. Kelp products contain auxins, gibberellins and cytokinins, causing plants to grow more leaves, as well as stimulating flowering, increasing lateral branching, developing more roots and dividing more cells.

Different products will have different concentrations and ratios depending on the extraction and concentration process and the type of kelp used. High cytokinin levels cause giant kelp to grow up to 2 feet per day, though that kind of growth might get out of hand even in the largest of warehouse gardens.

Conclusion

Plant growth hormones are like tools in a tool chest. Using the right product at the right time allows growers to tailor their gardens how they see fit.

One can induce vertical growth and leaf development to meet vegetative growth goals, and then stop vertical growth and promote lateral branching and flower initiation to finish big, or hold clones a couple weeks longer by halting growth altogether.

Wise use of phytohormones can bring your garden to the next level—just remember to watch out for the bad apples.

Impact of Different Phytohormones on Morphology, Yield and Cannabinoid Content of Cannabis sativa L.

The impact of exogenously applied plant growth regulators (PGR), 1-naphthalenaecetic acid (NAA), 6-benzylaminopurine (BAP), and a mixture of both (NAA/BAP-mix), was investigated in regard to plant height, length of axillary branches, number of internodes, biomass yield and cannabinoid content of three different phytocannabinoid-rich (PCR) Cannabis genotypes. The results showed that total plant height was significantly reduced under the application of NAA (28%), BAP (18%), and NAA/BAP-mix treated plants (15%). Axillary branch length was also significantly reduced by 58% (NAA) and 30% (NAA/BAP-mix). BAP did not significantly reduce the length of axillary branches. The number of internodes was reduced by NAA (19%), BAP (10%), and the NAA/BAP-mix (14%) compared to the untreated control. NAA application influenced the plant architecture of the tested cv. KANADA beneficially, resulting in a more compact growth habitus, while inflorescence yield (23.51 g plant −1 ) remained similar compared to the control (24.31 g plant −1 ). Inflorescence yield of v. 0.2x and cv. FED was reduced due to PGR application while cannabinoid content remained stable. Overall, the application of PGR could be used on a genotype-specific level to beneficially influence plant architecture and optimize inflorescence yield per unit area and thus cannabinoid yield, especially in the presence of space limitations under indoor cultivation.

1. Introduction

Cannabis sativa L. has a long history of cultivation for medicinal and food purposes as well as a source of textile fibers [1,2]. Five chemotypes of Cannabis were recognized and classified based on their cannabinoid profile and concentration: Chemotype I has a high Δ 9 -tetrahydrocannabinol/cannabidiol (THC/CBD) ratio (>1); plants with an intermediate ratio (≈1) are defined as chemotype II; fiber-type plants with a low THC/CBD ratio (<1) are defined as chemotype III; plants containing cannabigerol acid (CBGA) as their main cannabinoid are defined as chemotype IV [3], and chemotype V contains almost no cannabinoids [4,5,6,7].

Over the next few years, the cultivation of C. sativa is expected to rise based on adaptations to regulatory frameworks throughout the world, which may promote the constitution of new companies and exploit products obtainable from C. sativa. Medical cannabis was legalized in Germany in March 2017 [8]. As such, the regulatory framework comprises a policy that provides broad access to medical Cannabis. Today, Germany is the leading medical Cannabis prescriber in Europe, followed by Italy and the Netherlands [8]. The use of CBD in nutraceuticals, cosmetics, and pharmaceuticals has renewed interest in the effects of nonpsychotropic cannabinoids in particular [9,10]. Cultivation of C. sativa L. was banned in many states due to its psychoactive drug component Δ 9 -THC [11]. Industrial hemp genotypes, which comply with the 0.2% THC threshold set by the European Union legislation, can be cultivated without restrictions by farmers within the EU [12]. Breeding efforts for medicinal purposes focused on CBD-enhanced chemotypes, called phytocannabinoid-rich (PCR) Cannabis. PCR chemotypes target contents of more than 10% CBD and less than 0.2% THC, with a minimal range of variation. However, to optimize yield and the overall content of nonpsychoactive cannabinoids, a better understanding of the relationship between morphology and flower formation is necessary.

The target for high-quality Cannabis production includes a continuous and uniform inflorescence yield and the production of a specific cannabinoid compound [13]. Since the pharmaceutical industry requires the highest quality, it is necessary to ensure consistency in the cannabinoid profile of Cannabis plants and the quality of female flowers intended for such use. To optimize inflorescence yield per unit area and thus cannabinoid yield, indoor cultivation systems with minimum space requirements are needed. Outdoor cultivation seems to be economically feasible for the isolation of pure compounds or extract preparations. Indoor cultivation has advantages in terms of quality assurance and hygienic standards, allowing contamination to be eliminated and homogeneous Cannabis batches to be produced under controlled conditions [14]. Even though indoor systems are expensive operations, interest regarding the efficient use of available growing space is rising.

Exogenously applied growth regulators, which are chemical analogues to phytohormones, can influence the height and side-branching of plants, resulting in greater biomass and seed production [15,16]. Naturally, they occur in plants, acting as signaling compounds at low concentrations [17]. The first discovered and best understood phytohormones are auxins (IAA) and cytokinins (CK) [18], which are classified as growth promotors according to their main function [19]. IAA, synthesized in the shoot apex in young leaves and transported basipetally to the roots [20], have keys role in the maintenance of apical dominance, as well as responsibility for cell elongation. Their main effects include rooting, stimulation, and inhibition of axillary bud outgrowth [21]. CKs are involved in meristem activity regulations [22], plant shape determination, plant adjustment to side conditions, and responses to the environment [23]. IAA and CK act either antagonistically or synergistically to control developmental processes, such as the formation and maintenance of meristem [24]. C. sativa plants grow vertically, focusing on the growth of one dominant main shoot, with smaller side branches surrounding it, producing small buds and resulting in a low inflorescence yield. IAA from intact shoot apexes inhibit axillary branching, whereas CK, induced by removing the shoot apex, stimulates axillar branching [25]. Hence, apical dominance is regulated by IAA and CK [25]. Additionally, shoot branching depends on the genotype, growth stage, and environmental factors, including day-length, light intensity, temperature, and nutrition [26,27]. To generate a large variety of plant forms, shoot branching is a major determinant of the plant architecture, regulated by endogenous and environmental cues [23]. Synthetic compounds with similar activities to endogenous plant hormones are called “plant growth regulators” (PGR) [28]. For example, 1-naphthalenaecetic acid (NAA) and 6-benzylaminopurine (BAP) are synthetic analogues of the endogenous phytohormones IAA and CK, respectively. The effect of phytohormones on the architecture of C. sativa L. has not yet been studied in detail [23].

It is hypothesized that both the removal of the shoot apex and the additional exogenous application of PGR can stimulate axillary side-branching and influence the number of short side branches. The targeted plant architecture would be a compact and bushy canopy to promote optimal air circulation and light utilization. Further, shorter side branches give plants more stability during bud development. Nevertheless, the total yield cannot be rated only by the number and weight of inflorescence, therefore the content of cannabinoids is also of great interest [13].

The aim of this study was to evaluate the impact of exogenously applied NAA, BAP, and a mixture of both PGR on plant architecture of three different PCR Cannabis genotypes. Total plant height, axillary branch length after removal of the shoot apex, and number of internodes were investigated in detail. Furthermore, the biomass yield of inflorescence and leaves, as well as cannabinoid content, were determined.

2. Materials and Methods

2.1. Experimental Setup

A greenhouse experiment was set up to test the impact of exogenously applied plant growth regulators (PGR) on morphological characteristics of different phytocannabinoid-rich (PCR) Cannabis genotypes, namely cv. KANADA, cv. FED, and cv. 0.2x. The experiment was conducted at the University of Hohenheim, Germany, beginning on 28 November 2018 and finishing on 4 May 2019. Genotypes were kindly provided by the company Ai Fame, Switzerland. Genotypes were treated every fortnight with either 1-naphthaleneacetic acid (NAA) concentrated to 10 mg L −1 , or with 6-benzylaminopurine (BAP) concentrated to 50 mg L −1 , or a 1:5 mixture of 50% of NAA and 50% of BAP solution (NAA/BAP-mix). The experiment included a control group (control) that was sprayed with deionized water on the respective dates. NAA and BAP were both purchased from Sigma-Aldrich, St. Louis, USA. The first application took place 6 days after planting (DAP) and the last application was at 90 DAP. The applied solution amounted for each application 7.6 mL until 32 DAP and was increased to 11.0 mL for the following applications between day 32 and day 81. On the application days 88 and 90 DAP, the applied solution was increased to 27.0 mL for each application to spray the whole aboveground biomass of the plant.

Genotypes and treatments were randomly allocated to 36 plants according to a row–column design, which was established with four rows and three columns per replicate. Thus, three complete replicates existed.

2.2. Synthetic Growth Regulator Preparation

For the PGR solutions, 30 mg NAA was dissolved in 3 L of distilled water and 150 mg of BAP was dissolved in 3 L distilled water, with 15 mL Tween added to each solution as a surfactant. The NAA/BAP-mixture (NAA/BAP-mix) was prepared using 50% NAA solution and 50% BAP solution. NAA, BAP, and the NAA/BAP-mix were applied by evenly spraying all leaves of the plants. The control group was treated with the same amount of deionized water to simulate the spraying effect on the leaf surface. No Tween was added to the control water.

2.3. Plant Material

The experimental plants were generated by vegetative propagation by cutting only the apical tips of standardized mother plants. The cuttings were dipped into a rooting hormone (0.25% 4-(3-Indolyl)-butyric acid) and gently cultivated in 25 mm × 25 mm slabs, filled up with a growing media mixture of 50% seedling substrate (Klasmann-Deilmann GmbH, Geeste, Germany) and 50% sand. The clones were sprayed with water four times a day to reach a relative humidity above 90%. After 14 days and adequate root growth, the cuttings were transplanted in seedling substrate (Klasmann-Deilmann GmbH, Geeste, Germany) into a pot that was 90 mm in diameter. The shoot apexes were removed to 9 internodes for genotype KANADA and 0.2x-genetic and 11 internodes for the auto-flowering genotype FED. The experimental plants were transferred at 15 DAP into 13 × 13 cm square pots, at 50 DAP into 18 × 18 cm square pots, and finally at 74 DAP into 10 L containers, in a growing media mixture consisting of 15% black peat, 20% fraction 1, 25% milled peat, 20% GF medium, 10% pine bark, 10% leca, 1 kg m −3 horn chips, and 1 kg m −3 NPK 12-14-24 (Klasmann-Deilmann GmbH, Geeste, Germany). Under an indoor vegetative life cycle of 18 h, sunlight was supplemented with artificial lightning using Gavita high-pressure sodium (HPS) lamps, i.e., E-Series DE FLEX EU Lamp (750 W, 400 V, 1500 μmol s −1 ), Aalsmeer, Netherlands. The experiment was irrigated by a drip irrigation system and fertilized four days a week with 0.2% of Plantaactiv 18-12-18 Type A during the vegetative growth cycle and with 0.2% of Plantaactive 10-20-30 Type B during the generative growth cycle, which was purchased from Hauert (Grossaffoltern, Switzerland). The temperature during the vegetative growth stage varied from 23.7 °C to 27.6 °C. Relative humidity varied between 22.4% and 47.5%. At 129 DAP, the experimental plants were moved into a climate chamber to a 12-h photoperiod to initiate floral development. The temperature during the generative growth cycle varied from 17.9 °C to 24.1 °C. Relative humidity varied between 30.8% and 89.2%.

2.4. Measurements

Measurements took place for every plant each seventh day for a total period of 132 days. Plants were measured for their total height and length of axillary branches. Depending on the genotype, 9 (for KANADA and 0.2x-genetic) or 11 (for FED) branches were measured. Nodes of each axillary branch were counted for the KANADA and 0.2x-genetic genotypes. The autoflowering characteristics and dense foliage of genotype FED made it impossible to count axillar branch nodes.

2.5. Plant Samples

Genotypes were harvested when 70% of the pistils had darkened. Gland heads of trichomes are clear or slightly amber at the beginning of the growth cycle. Prior to harvest, when cannabinoid levels reach their maximum, they turn cloudy. The state of trichomes was monitored with binoculars. Genotype FED was harvested at 137 DAP. Harvest of genotype KANADA and 0.2x-genetic took place at 142 and 156 DAP, respectively. Inflorescence and leaves were dried at a temperature of 20 °C for 14 days. After the drying process, dry matter was weighed and recorded in gram per single plant to determine the dry weight (DW). Subsequently, the dried plant material was ground with an ultracentrifugal mill (Retsch, Type ZM 200, Haan, Germany) to acquire a homogeneous powder, with a particle size of 1 mm. The residual moisture of each samples was measured with a moisture analyzer (DBS 60-3 of Kern and Sohn GmbH, Balingen, Germany).

2.6. Extraction and Quantification of Cannabinoids by HPLC Analysis

Quantitative analysis of cannabinoids, particularly cannabidiol (CBD), cannabidiolic acid (CBDA), Δ 9 -tetrahydrocannabinol (THC), and Δ 9 -tetrahydrocannabinolic acid (THCA), was performed, according to Lehmann and Brenneisen [29] with slight modifications after Burgel et al. [30]. For the cannabinoid extraction, 100 ± 10 mg of the grinded sample was dissolved in 100 mL methanol 90%/chloroform 10% (v/v) (9 + 1) composite.

An external calibration of cannabinoid quantification was performed according to Burgel et al. [30], using one standard (CAN1) containing the target compounds (CAN1: THC 2%, CBD 2%, THCA 10%, CBDA 10%). The reference cannabinoids, CBD, THC, and THCA, were purchased from Lipomed (Arlesheim, Switzerland), and was purchased from CBDA from Echo Pharmaceuticals BV (Weesp, The Netherlands).

Data were processed using ChemStation Software for LC Rev. B.04.03-SP2 (Agilent, Santa Clara, CA, USA). The retention time of the respective chromatographic target peak, was compared with the chromatographic peak of the reference to carry out a quantitative analysis. The UV spectra was used to preliminarily allocate the chromatographic peak to the reference spectra visually. The identity of the target cannabinoid was proven if the deviation of the retention time of the chromatographic peak was ≤0.5 min and the optical spectra comparison did not show any difference.

To calculate the respective cannabinoid content C TS in mass percent [ % m / m ], Equation (1) was used, where A TS is defined as the peak area of the standard analyst, B TS is defined as the peak area of the sample analyst in μV × s, V is defined as the volume, E W T S h i j k l as the weight portion of the product in mg, and F ijkl as the residual moisture of the product in %m/m. Indices are defined for the i-th genotype in the j-th row, the h-th column of the k-th replicate, and the l-th treatment; thus, the calculation was performed for each plant.

2.7. Statistical Analysis

A mixed model approach was used to analyze all traits, which were determined by the measurement of single plants. Thus, the dry weight of the leaves and inflorescence, estimation, and statistical interference of the cannabinoids present in the dried plant material were analyzed by

where y hijkl is the observation of the i-th genotype in the j-th row, the h-th column of the k-th replicate, and the l-th treatment, μ is the intercept, δ i is the fixed effect of the i-th genotype, τ l is the fixed effect of the l-th treatment, δ τ i l is the fixed interaction effect of the corresponding main effects, b k is the fixed effect of the k-th replicate, e hijkl is the plant or error effect associated with observation y hijkl , and r jk and c hk are the random row and column effects within the k-th replicate, respectively. Normal distribution and homogeneous variance of residuals were checked graphically via residual plots. If needed, the data were logarithmically transformed to fulfill the requirement concerning homogeneous variance and normal distribution. In this case, estimates were back transformed for presentation purposes only. Standard errors were back transformed using the delta method.

Total plant height was measured weekly for 20 weeks. The number of internodes and the length of axillar branches were measured weekly for 14 weeks. Thus, repeated measures were taken and Model (2) was extended by the factor measurement with 20 or 14 levels, as follows:

where t m is the effect of the m-th measurement and all other effects are defined analogous to Model (2) for each measurement m. As repeated measures from each row, column, and plant were taken, a first-order autoregressive variance–covariance structure with heterogeneous variance was assumed for these random effects, allowing for a serial correlation between observations taken from the same row, column, or plant. The variance–covariance structure was simplified to a homogeneous variance first-order autoregressive structure if this decreased the AIC [31], thus resulting in a better model fit. After finding significant differences via the global F-test, Fishers LSD test was performed for multiple comparisons. A letter display was used to present the results of the multiple comparisons [32]. All statistical analyses were conducted using the statistical software SAS version 9.4 (SAS Institute, Cary, NC, United States). Figures were also generated using the statistical software SAS version 9.4 and Excel 2013 (Microsoft Corporation, Washington, DC, USA).

3. Results

3.1. Plant Height

In the following section, plant height is described for different plant growth regulator (PGR) treatments (NAA, BAP, and NAA/BAP-mix) separately in comparison to control plants. The height of the treated plants showed significant interactions between treatments and measurements over time. No significant interactions were observed between treatment and genotype. Plant heights of NAA-treated and BAP-treated plants were significantly reduced compared to the control plants, starting 13 and 20 days after planting (DAP), respectively. At 132 DAP, NAA- and BAP-treated plants indicated heights of 69.20 ± 4.43 cm and 79.53 ± 4.43 cm, respectively, compared to the control with 96.47 ± 4.43 cm ( Figure 1 A,B). Plants treated with the NAA/BAP-mix showed the same trend (81.74 ± 4.43 cm). A significant reduction in plant height was observed between 20 DAP and 62 DAP compared to the control. No significant growth reduction was determined between 69 and 104 DAP ( Figure 1 C). Finally, the treatments resulted in 28% (NAA), 18% (BAP), and 15% (NAA/BAP) shorter plants in comparison to the control ( Figure 1 A–C).

Mean plant height over all three tested genotypes treated with (A) 1-naphthalenacetic acid (NAA), (B) 6-benzylaminopurine (BAP), and (C) NAA/BAP-mix, compared to the nontreated control over a time period of 132 days after planting (DAP). (D) Mean plant height of genotype KANADA, 0.2x-genetic, and FED over 132 days. Means covered with at least one identical lowercase letter did not differ significantly at α = 0.05. The arrows show the period (6 to 90 DAP) during which application took place.

Across all treatments, the KANADA genotype significantly indicated the highest plants at 132 DAP (113.01 ± 3.83 cm), followed by genotype FED (72.45 ± 3.83 cm), whereas genotype 0.2x-genetic indicated the shortest plants, with a final plant height of 59.88 ± 3.83 cm ( Figure 1 D).

3.2. Number of Internodes of Axillar Branches

PGR treatments significantly influenced the number of internodes of axillary branches of the genotypes (KANADA and 0.2x-genetic). Internodes of the autoflowering genotype FED could not be measured. Axillary branches of NAA-treated plants showed an inhibited number of internodes after 14 days of application (20 DAP) and resulted in an average number of 17 internodes per axillary branch compared to the control, which showed 21 internodes per branch ( Figure 2 A). The same trend was observed for the BAP-treated and NAA/BAP-mix-treated plants. BAP and NAA/BAP-mix did not affect the number of internodes per axillary branch during the first 20 days of application (26 DAP). Between 26 and 90 DAP, a reduction of internodes was observed and resulted in 19 and 18 internodes per axillary branch compared to the control ( Figure 2 B,C). For the final measurement, all PGR were shown to inhibit the average number of internodes during the vegetative period compared to the control. No significant differences were observed between the treatments ( Figure 2 A–C).

Mean number of internodes per axillary branch over two tested genotypes (KANADA and 0.2x-genetic) treated with (A) NAA, (B) BAP, and (C) NAA/BAP-mix, compared to the nontreated control over a time period of 90 days. (D) Mean number of internodes per axillary branch of genotype KANADA and 0.2x-genetic across time. Means covered with at least one identical lowercase letter did not differ significantly at α = 0.05. The arrow shows the time point (6 DAP) of the first application.

Across all treatments, genotype KANADA showed an inhibited number of nine internodes until 34 DAP compared to genotype 0.2x-genetic, which showed inhibition of 10 internodes. Finally, at 90 DAP, genotype KANADA demonstrated a significantly higher number of internodes per axillary branch (20 internodes) compared to genotype 0.2x-genetic, which showed 18 internodes ( Figure 2 D).

3.3. Length of Axillary Branches

PGR treatments significantly influenced the length of axillary branching. NAA-treated plants started after seven days of application (13 DAP) to inhibit the growth of axillary branches compared to the control. After 84 days of application (90 DAP), the axillar branching growth was reduced and resulted in an average length of 19.23 ± 2.25 cm ( Figure 3 A). After seven days of application (13 DAP), BAP-treated plants showed a significant reduction in axillar branching over the vegetative period and resulted in 36.46 ± 4.21 cm at 90 DAP compared to the control (45.26 ± 5.41 cm) ( Figure 3 B). A similar trend was observed for plants treated with the NAA/BAP-mix, with a final length of axillary branches of 31.63 ± 3.77 cm compared to the control ( Figure 3 C). At 26 DAP and 34 DAP, NAA-treated plants were shown to be influenced to a greater extent compared to the BAP-treated and NAA/BAP-mix-treated plants, respectively. NAA resulted in the highest axillary branch length reduction with 26.03 cm, while BAP and the NAA/BAP-mix showed 8.80 cm and 13.63 cm length reductions compared to the control, which was a significant finding ( Figure 3 A–C).

Median length of axillar branches in cm over all three tested genotypes treated with (A) NAA, (B) BAP, and (C) NAA/BAP-mix compared to the nontreated control each over a time period of 90 days. (D) Mean length of axillar branches of genotype KANADA and 0.2x-genetic across time. Means covered with at least one identical lowercase letter did not differ significantly at α = 0.05. The arrow shows the time point (6 DAP) of the first application.

A genotype-specific difference in axillary branch length across all treatments was observed. Genotype KANADA indicated a final length of axillary branches of 53.64 ± 5.29 cm, which were the longest axillary branches compared to genotype 0.2x-genetic (26.22 ± 2.59 cm) and FED (22.46 ± 2.06 cm) at 90 DAP ( Figure 3 D).

3.4. Yield Parameters

Dry weight (DW) yield of inflorescence per plant showed significant interactions between genotype and treatment. Accordingly, DW means of the different treatments are described separately for each genotype. Control plants of genotype KANADA reached the highest inflorescence yield with 24.31 g plant −1 , followed by plants treated with NAA (23.51 g plant −1 ) and BAP (22.97 g plant −1 ). Plants treated with a mixture of NAA and BAP (NAA/BAP-mix) indicated the lowest DW yield of inflorescence, with 14.79 g plant −1 compared to the control ( Table 1 ). Control plants of 0.2x-genetic indicated the highest inflorescence yield with 20.83 g plant −1 , followed by plants treated with BAP (15.88 g plant −1 ), which showed no significant difference compared to NAA-treated and NAA/BAP-mix-treated plants, with results of 7.61 g plant −1 and 8.57 g plant −1 , respectively ( Table 1 ). Control plants of the autoflowering genotype FED showed the highest inflorescence yield. with 32.10 g plant −1 , which was a significant finding. NAA-treated plants indicated the lowest yield (5.39 g plant −1 ) compared to the control ( Table 1 ).

Table 1

Mean inflorescence dry weight (DW) in g plant −1 of genotypes KANADA, 0.2x-genetic, and FED, treated with NAA, BAP, and NAA/BAP-mix and a nontreated control. Results are presented as mean values ± standard error of mean (mean ± SEM). Means of treatments in one column followed by at least one identical lowercase letter are not significantly different, as indicated by the LSD test (α = 0.05). Means of genotypes in one row followed by at least one identical uppercase letter are not significantly different, as indicated by the LSD test (α = 0.05). The p-values correspond to global F tests for differences between the levels of the mentioned genotypes, treatments, or their interactions.

Trait Treatment Genotype
KANADA 0.2x-genetic FED
Inflorescence DW
[g plant -1 ]
Control 24.31 ± 3.06 aA 20.83 ± 3.79 aA 32.10 ± 3.76 aA
NAA 23.51 ± 3.10 abA 7.61 ± 3.10 bB 5.39 ± 3.10 cB
BAP 22.97 ± 3.11 abA 15.88 ± 3.11 abA 17.61 ± 3.13 bA
NAA/BAP-mix 14.79 ± 3.13 bAB 8.57 ± 3.11 bB 19.12 ± 3.80 bA
p-values
Genotype [G] 0.0041
Treatment [T] 0.0003
G × T Interaction 0.0430

DW yield of leaves per plant showed significant differences between treatments and genotypes. On average, genotype KANADA showed the highest DW yield of leaves (25.14 g plant −1 ) over the treatments compared to 0.2x-genetic (18.54 g plant −1 ) and FED (17.76 g plant −1 ; Table 2 ), which was a significant finding.

Table 2

Mean leaves dry weight (DW) in g plant −1 of genotypes KANADA, 0.2x-genetic, and FED and mean leaves dry weight (DW) in g plant −1 of genotypes KANADA, 0.2x-genetic, and FED, treated with NAA, BAP, and NAA/BAP-mix and a nontreated control. Results are presented as mean values ± standard error of mean (mean ± SEM). Means of genotypes or treatments in one row followed by at least one identical lowercase letter are not significantly different, as indicated by the LSD test (α = 0.05). The p-values correspond to global F tests for differences between the levels of the mentioned genotypes, treatments, or their interactions.

Trait Genotype
Leaves DW
[g plant -1 ]
KANADA 0.2x-genetic FED
25.14 ± 1.94 a 18.54 ± 2.07 b 17.76 ± 2.21 b
Treatment
Control NAA BAP NAA/BAP-mix
25.6 ± 2.60 a 12.43 ± 2.24 b 22.96 ± 2.24 a 20.79 ± 2.43 a
p-values
Genotype [G] 0.0337
Treatment [T] 0.0045
G × T Interaction 0.1914

Plants treated with NAA indicated the lowest DW yield (12.43 g plant −1 ) of leaves compared to the control and the other treatments over the three genotypes, ranging from 25.60 g plant −1 (control) to 20.79 g plant −1 (NAA/BAP-mix; Table 2 ), which was a significant finding.

3.5. Cannabinoid Content

Cannabinoid acids occur in planta and undergo decarboxylation to their neutral forms upon heating or pyrolysis. The following results for cannabidiol (CBD) refer to the sum of cannabidiolic acid (CBDA) and CBD analyzed. Inflorescence and plant leaves treated with NAA, BAP, or NAA/BAP-mix did not show any statistical differences between the plants exposed to growth regulators and the control plants. Significant differences in CBD content were only between genotypes. KANADA indicated the highest content of CBD in inflorescence (10.33%) and leaves (7.03%), followed by the inflorescence (7.91%) and leaves (6.77%) of 0.2x-genetic. The lowest content was measured in the inflorescence and leaves of the autoflowering genotype FED, with 6.34% and 5.59%, respectively ( Table 3 ).

Table 3

Mean content of cannabidiol (CBD) in mass percent [%m/m] of genotypes KANADA, 0.2x-genetic, and FED. CBD was analyzed in inflorescence and leaves. Results are presented as mean values ± standard error of mean (mean ± SEM). Means in one row followed by at least one identical letter are not significantly different as indicated by LSD test (α = 0.05). The p-values correspond to global F tests for differences between the levels of the mentioned genotypes, treatments, or their interactions.

Trait Genotype
KANADA 0.2x-genetic FED
CBD [%m/m]
Inflorescence 10.33 ± 0.30 a 7.91 ± 0.30 b 6.34 ± 0.31 c
Leaves 7.03 ± 0.18 a 6.77 ± 0.20 a 5.59 ± 0.28 b
p-values Inflorescence
Genotype [G] 0.0226
Treatment [T] 0.4411
G × T Interactions 0.1072
p-values Leaves
Genotype [G] 0.0026
Treatment [T] 0.8755
G × T Interactions 0.7891

4. Discussion

Exogenous application of plant growth regulators (PGR; 1-naphthaleneacetic acid (NAA), 6-benzylaminopurine (BAP), and a mixture of both (NAA/BAP-mix)) was shown to have an impact on modifying the plant architecture of C. sativa.

Auxins (IAA) promote stem elongation and maintain apical dominance, including inhibition of axillary bud outgrowth [21]. Hence, an increase in plant height, which leads to taller plants on average, through exogenous NAA treatment was expected. However, the results showed reduced plant height. Between 41 to 83 DAP, a reduction in plant growth in NAA-treated plants was larger compared to NAA/BAP-mix-treated plants. Mendel et al. [23] reported a nonsignificant inhibition of the average plant height of C. sativa over a time period of 56 days, in which plants were exposed to different NAA concentrations (5 mg L −1 , 10 mg L −1 , and 20 mg L −1 ). In contrast, Lalge et al. [33], documented an increase in total plant height of C. sativa plants treated with 10 mg L −1 NAA, whereas concentrations of 5 mg L −1 and 20 mg L −1 showed no significant effect on plant height [33]. Our study indicated that NAA-treated plants showed the highest axillary side-branch reduction, with an average length of 19.23 cm, compared to BAP treated plants (36.46 cm), NAA/BAP-mix treated plants (31.36 cm), and control plants (45.26 cm). However, this was in accordance with the measured number of internodes per axillary side branch, which indicated 19% (four internodes) reduction in the number of internodes per axillary side branch after NAA application compared to the control, whereas no significant differences between treatments were observed. Lalge et al. [33] and Mendel et al. [23] indicated increased lateral branching of plants subjected to NAA treatment at all concentrations that affected the plant. Lalge et al. [33] argued that buds of C. sativa demonstrate decreased sensitivity to the inhibitory effects of IAA in apical dominance. Since the apical dominance in our study was broken before application, this may explain the different response of the plant to NAA application, whereas Mendel et al. [23] explained the inhibitory effect in terms of plant height and the significant promoting effect in terms of axillary side-branch length by an over-optimal NAA concentration and an imbalance in the hormonal endogenous balance. Other studies showed that high IAA concentrations induced ethylene synthesis. The release of this gas could be responsible for the decrease in stem elongation [34]. Further, synthetic auxin (NAA) in C. sativa could interact with a different set of receptors than the most common and natural auxin indole-3-acetic acid (IAA) [23]. In the case of plant height and axillary side-branch growth, even low concentrations (5 mg L −1 ) were sufficient to achieve active stem elongation. Higher concentrations of NAA (10 mg L −1 and 20 mg L −1 ) eliminated and restored the hormonal effect. This biphasic behavior of the reaction could depend on adjustment of the ratio, with concentrations of exogenous auxins exceeding the minimum effective dose [23].

PGR, especially growth retardants including CKs, are generally applied in order to obtain short and compact plants [35]. CKs are involved in meristem activity regulation [22], stem elongation inhibition, and plant flowering, among other things [35]. In apical dominance, CK has an antagonistic effect to the IAA effects of. Studies of other species showed increased CK levels (25-fold within 24 h) after decapitation of the shoot apex [36]. Therefore, axillary bud outgrowth is correlated with CK concentration, which is locally biosynthesized in the nodal stem [37]. Further, it is well known that direct application of CKs to axillary buds promotes their outgrowth, even in intact plants [38]. The aim of exogenous CK application is to obtain well-branched plants without removing the apical meristem. The removal of the apical shoot apex is a common pruning technique to control the way plants grow, either to restrict plant height, maximize yield [39], or as a management tool to optimize space utilization for indoor cultivation.

Decapitation of the shoot apex to reach a defined number of side branches, in combination with exogenously applied synthetic analogues of CK (BAP), met expectations and resulted in an 18% reduction in total plant height after 90 days of application in comparison to control plants. The average length of axillary side branches showed a reduction trend of around 20% compared to the control. However, no significant reduction in axillary side branches was shown. The measured number of internodes of the axillary side branches was reduced by 9% on average compared to the control. In contrast, Mendel et al. [23] and Lalge et al. [33] reported that treatments with BAP did not affect the total plant height of C. sativa at concentrations of 10 mg L −1 , 25 mg L −1 , and 50 mg L −1 . A total plant height of 200 to 220 cm was documented by Lalge et al. [33] 56 days after the first application took place for all concentrations, including the control, in accordance with a study by Leite et al. [40], where exogenously applied CK was not effective in modifying evaluated plant growth parameters of other plant species (Glycine max L. Merr.). Further, a strongly significant stimulation of axillary side-branch growth was documented at the highest BAP concentration (50 mg L −1 ). Control plants measured 1 cm in side-branch length on average, whereas BAP-treated plants measured 10 cm in side-branch length on average 56 days after the first treatment took place [23]. Lalge et al. [33] stated a concentration-dependent increase in axillary branch length of BAP-treated plants. The highest dosage (50 mg L −1 ) showed the most significant results, with around 50% increase in average axillary side-branch length compared to the control, 56 days after first application. These findings were in contrast to the present results, which can be explained by a genotype-specific reaction of PGR. While Mendel et al. [23] and Lalge et al. [33] used an industrial hemp fiber variety called Bialobrzeskie, of chemotype III, a phytocannabinoid-rich (PCR) chemotype was used in the present study. Further, by removing the shoot apex, a change in IAA/CK balance was expected.

A similar trend applied to plants treated with a mixture of both (NAA/BAP-mix). In line with NAA and BAP treatments, a reduction (15%) in total plant height after 84 days of application (90 DAP) was measured compared to the control. The length of axillary side branches was reduced by 31% on average, with the number of internodes reduced by 14% compared to nontreated control plants. IAA and CK interact in a complex manner to control plant growth [41]. Experiments with the model plant Arabidopsis thaliana showed a homeostatic regulatory feedback loop model, in which CK functioned as a positive regulator of IAA biosynthesis and IAA repressed CK biosynthesis [41,42,43]. It remains unclear whether this model can be extrapolated to the functioning of the whole plant, in which IAA and CK concentrations are modified naturally or in response to exogenous application [44]. Nevertheless, the present study showed shorter plants with a reduced length of axillary branches compared to the control, but not as much as plants which were only treated with NAA.

Plant architecture was influenced using PGR. The aim was to modify the plant morphology in order to generate small, compact plants for various indoor growing systems. While NAA-treated plants showed a short habitus, including significant the shortest axillary side branches with a reduced number of internodes. BAP-treated plants and plants treated with the NAA/BAP-mix also showed shorter habitus and demonstrated shorter axillary side branches with reduced numbers of internodes. It is important to note that the use of PGR did not reduce biomass yield and the content of cannabinoids. Results showed that the impact of PGR on yield of inflorescence was dependent on the interaction between the genotype and the treatment. NAA- and BAP-treated plants of genotype KANADA showed an equally high yield of inflorescence DW than the control plants, whereas 0.2x genotype showed an equally high DW yield, but only for BAP-treated plants. However, genotype FED indicated a lower DW yield of inflorescence in PGR treatments. It is important to mention that Cannabis genotypes respond differently to production conditions, as reported by Backer et al. [45]. CBD contents in inflorescence and leaves showed no impact on PGR. Considering that a minimal range of variation is aimed for and a THC content of < 0.2% must not be exceeded, the fact that PGR applications have no influence on cannabinoid content may be advantageous.

Above all, the inflorescence yields are decisive, since a higher content of cannabinoids is expected. Stout et al. [46] reported the highest CBD levels in Cannabis flowers, with lower amounts in leaves. Cannabinoids can be extracted from the reproductive plant parts and foliage. Inflorescence has higher concentrations of cannabinoids than foliage material, however foliage parts comprise the lager biomass of the Cannabis plant [47]. The PCR genetics used in the present study showed lower leaf DW compared to inflorescence DW. Leaf DW yield depended on the genotype and treatment, where genotype KANADA showed the highest DW leaf yields and genotypes 0.2x and FED were not that profitable. The use of PGR did not reduce leaf yield for BAP-treated and NAA/BAP-mix-treated plants.

In summary, the results showed that the used PCR genetics reacted differently to PGR applications. Genotype KANADA was shown to be suitable as a treatment with synthetic auxin (NAA) to adapt plant architecture to corresponding indoor conditions. Short plant height was characterized by stability in the generative phase and avoided breakdown of side branches. A defined number of axillary side branches by apical bud removal and a reduced side-branch length guaranteed homogeneous flower development by uniform exposure. Furthermore, better harvesting conditions were given due to uniform plant height and homogeneous plant development. Despite a more compact habitus, the flower yield was not reduced and the KANADA genotype showed the highest CBD content compared to the other genotypes. Treatment with synthetic CK (BAP) was shown to be disadvantageous regarding the PCR genetics used, because axillary side-branch length could not be significantly shortened. Treatment with a mix of both showed no beneficial effects in any of the three genotypes, as the DW of inflorescence was reduced by the treatment. Only leaf DW was not reduced. If leaf yield is the purpose of use, BAP treatment may be appropriate. In this case, genotype KANADA is also recommended due to the significant finding of the highest leaf DW.

The use of specific PGR is permitted in fruit growing, horticulture, and field-crop cultivation [17,48,49]. In the cultivation of medicinal grade Cannabis production, there are currently no approved PGR to modify plant architecture. Since the use of PGR in production under good manufacturing practice (GMP) or good agricultural and collection practice (GACP) guidelines is not clearly defined, the use of PGR in the cultivation of nonmedical Cannabis could be of importance for cosmetics or nutraceuticals, for example. Nevertheless, minimum effective chemical inputs, including fertilizers, growth regulators, pesticides, and herbicides, should be achieved and well documented to secure the marketability of the product [50]. When using synthetic PGR, it is very important to know exactly the appropriate application methods and concentrations to avoid possible residues in the final product, as high concentrations of IAA are toxic. Because of these properties, compounds with auxin-like activities were developed and can be applied as herbicides [51]. Synthetic IAAs are more stable than endogenous IAA, because the compounds show reduced metabolic turnover [52]. Thus, natural phytohormones could be an alternative to synthetic ones.

5. Conclusions

The results of this study showed that exogenously applied plant growth regulators (PGR), namely 1-naphthaleneacetic acid (NAA), 6-benzylaminopurine (BAP), and a mixture of both (NAA/BAP-mix), had significant impacts on the plant architecture of C. sativa. Phytocannabinoid-rich (PCR) Cannabis genetics reacted in a genotype-specific manner to PGR applications. The use of NAA led to a more compact plant architecture with a consistently high inflorescence yield for the genotype KANADA; cannabidiol (CBD) content was not affected. A beneficial effect on the autoflowering genotype (FED) could not be confirmed. Genotypes 0.2x-genetic and FED showed reduced inflorescence yield due to PGR application. The use of PGR opens up a very interesting field and requires further study to test the use of PGR at different concentrations. Although exogenously applied PGR might be a cultural practice in the future, further studies to screen more PCR-genetics on their specific reactions to PGR applications are required to develop new genotype-specific indoor cultivation systems.

Acknowledgments

The authors would like to thank the technical assistance of Theresa Thiel, for the great support of sample processing. Thanks to Joan Espel Grekopoulos of the company AI FAME, for support in the laboratory, where a part of the chemical analysis where done.

Author Contributions

Conceptualization, L.B., D.S., and S.G.-H.; methodology, L.B. and S.G.-H.; software, L.B. and J.H.; validation, L.B. and J.H.; formal analysis, L.B. and J.H.; investigation, L.B.; resources, L.B. and D.S.; data curation, L.B.; writing—original draft preparation, L.B.; writing—review and editing, S.G.-H. and J.H.; visualization, L.B.; supervision, S.G.-H.; project administration, S.G.-H.; funding acquisition, S.G.-H. All authors read and agreed to the published version of the manuscript.

Funding

This research was funded by the German Federal Ministry for Economic Affairs and Energy within the Central Innovation Program for SMEs (16KN050543).

Conflicts of Interest

The authors declare no conflict of interest. The funders had no role in the design of the study, in the collection, analysis, or interpretation of data, in the writing of the manuscript, or in the decision to publish the results.