A protocol is presented for the convenient and high-throughput isolation and enrichment of glandular capitate stalked and sessile trichomes from Cannabis sativa. The protocol is based on a dry, non-buffer extraction of trichomes using only liquid nitrogen, dry ice, and nylon sieves and is suitable for RNA extraction and transcriptomic analysis.
This paper presents a protocol for the convenient and high-throughput isolation and enrichment of glandular capitate stalked and sessile trichomes from Cannabis sativa. The biosynthetic pathways for cannabinoid and volatile terpene metabolism are localized primarily in the Cannabis trichomes, and isolated trichomes are beneficial for transcriptome analysis. The existing protocols for isolating glandular trichomes for transcriptomic characterization are inconvenient and deliver compromised trichome heads and a relatively low amount of isolated trichomes. Furthermore, they rely on expensive apparatus and isolation media containing protein inhibitors to avoid RNA degradation. The present protocol suggests combining three individual modifications to obtain a large amount of isolated glandular capitate stalked and sessile trichomes from C. sativa mature female inflorescences and fan leaves, respectively. The first modification involves substituting liquid nitrogen for the conventional isolation medium to facilitate the passage of trichomes through the micro-sieves. The second modification involves using dry ice to detach the trichomes from the plant source. The third modification involves passing the plant material consecutively through five micro-sieves of diminishing pore sizes. Microscopic imaging demonstrated the effectiveness of the isolation technique for both trichome types. In addition, the quality of RNA extracted from the isolated trichomes was appropriate for downstream transcriptomic analysis.
Glandular trichomes are hair-like structures present in plants that contain many secondary metabolites1 and represent a valuable bank of novel biosynthetic genes and enzymes2. In Cannabis, the biosynthesis of the important secondary metabolites, cannabinoids3 and terpenes4, is localized in the trichomes. Considering the role of trichomes in determining the quality of Cannabis both for medicinal and recreational uses, the study of trichome gene expression is of interest. To characterize the expression of trichome-specific genes, the trichomes of interest must first be isolated. Trichome isolation protocols were first described as early as 19925, and their latest developments have been recently reviewed2. In general, protocols for extracting glandular trichomes for transcriptomic characterization can be divided into two distinct sequential steps. The first step involves a thorough physical separation of the trichomes from the plant tissue. This step can be performed by using dry ice5, glass beads with a commercial apparatus6,7, grinding the plant material against a mesh sieve8, or vortexing the plant tissue in an isolation buffer9. The second step involves a more refined separation of the trichomes of interest from the microscopic plant residue and/or other trichome types. This step can be executed using density gradient centrifugation8,10 or sieves of various sizes7,9. Due to the extreme sensitivity of RNA in processed tissues to degrading agents, these two sequential steps are usually conducted in the ice-cold isolation medium, often in the presence of protein inhibitors4.
Conventional trichome isolation protocols require, in addition to the ice-cold temperatures, large amounts of isolation medium to ensure an efficient extraction procedure. The combination of these components results in an arduous, time-consuming isolation process that hinders high throughput. Presenting a straightforward, user-friendly alternative trichome isolation protocol is, therefore, likely to be beneficial for various aspects associated with trichome characterization. The present paper aims to offer an alternative protocol for isolating stalked and sessile glandular capitate trichomes from Cannabis sativa by combining and integrating several elements from the conventional protocols. These elements include dry ice5, the passing of the trichomes through several micro-sieves with decreasing pore sizes7,9, and substituting liquid nitrogen (LN) for the isolation medium8.
The novelty of the present trichome isolation protocol, as compared to conventional protocols, presents in a number of ways. This protocol is convenient, as it does not require hazardous components. The procedure can be conducted in the lab with minimal precautions and facilitates high throughput. Substituting LN for the standard liquid isolation medium ensures the integrity of the trichomes throughout the isolation process, enabling subsequent transcriptomic analysis. Upon sublimation of the LN and dry ice, the isolated trichomes are left free of harmful residuals. Further, the propensity of LN to sublimate at room temperature allows its generous use throughout the protocol. In contrast, using large volumes of conventional isolation medium generates practical difficulties in its handling. Finally, the protocol decreases the separation of the disc cell from the remaining fragile head structure of the glandular trichome, enabling the retention of the headspace content.
This protocol is presented in a detailed step-by-step fashion designed to assist the technical practice of isolating C. sativa glandular capitate trichomes. The protocol provides a manageable workflow that results in isolated trichomes with a high concentration and purity that are appropriate for downstream molecular analysis.
NOTE: The plant material used in this study consisted of four C. sativa ARO-Volcani strains (CS-11, CS-12, CS-13, and CS-14) that were grown in the Volcani Center, Israel, as described elsewhere11. Glandular capitate stalked trichomes were isolated from mature flowering inflorescences, and glandular capitate sessile trichomes were isolated from large fan leaves from mature non-flowering mother plants. All plant material was freshly picked and immediately stored at −80 °C.
CAUTION: Dry ice and LN are used throughout the protocol. These substances are extremely hazardous. The isolated trichomes can contain dry ice particles that can create dangerous gas pressure when inserted into sealed tubes; therefore, all caps should be needle-punctured. The use of protective goggles, appropriate lab wear, and gloves for handling at extremely low temperatures is highly advised.
1. Setup for the initial separation of trichomes from the plant material
2. Separation of trichomes from the plant material
3. Separation of glandular capitate trichomes (stalked and sessile) from other trichome types, such as bulbous and cystolithic trichomes, and debris
4. Passing trichomes through micro-sieves of decreasing pore size
5. Collection of the desired trichomes
6. Observation and analysis of the purified trichomes
The main modification included in this protocol compared to conventional trichome isolation protocols is substituting the standard isolation medium with LN. Using LN as an isolation medium allows a relaxed workflow, because as long as the samples are submerged in LN, metabolic degradation is not likely to occur. Furthermore, as the protocol avoids the hazardous components (i.e., aurintricarboxylic acid and β-mercaptoethanol) used in traditional trichome isolation medium, the work is not restricted to a chemical hood. Nevertheless, it is important to take the necessary precautions while handling dry ice and LN.
To assess the advantages of the present protocol, the trichome isolation results from the present protocol were compared with those obtained using a conventional trichome isolation procedure (which used an aqueous buffer, glass beads, and a fine sieve for trichome isolation)12, initially using a light microscope. A comparison of the components from different isolation phases via light microscope inspection illustrates the increasing trichome isolation degree as the isolation process progresses. In the initial stage of the isolation process (105-150 µm micro-sieve), pieces of leaf tissue were predominant in the isolated portion (Figure 9A), in contrast to the final isolation product, which was composed almost entirely of isolated trichomes (Figure 9C). Care should be taken to validate that the isolated samples are contamination-free (Figure 9B).
Although the trichome density was not quantified in this study, the quality of the isolated glandular capitate stalked and sessile trichomes in the final isolation step can be observed in Figure 10. The purity can be compared to that obtained with trichome isolation using a conventional protocol12 (Figure 10C). In the present protocol, the delicate stalked trichome head structure is retained, and the whole heads of the isolated stalked trichome are visible (Figure 10D), whereas using the conventional protocol, the complete trichome head structure is missing, and only the disc cells are present (Figure 10E). The yield can also be greatly increased since the use of dry ice and LN obviates the limit of plant material that can be extracted. An additional advantage of our non-aqueous protocol is that it uses only dry ice and LN, and, therefore, no residual components of the isolation medium are left once they evaporate. Using LN prevents the release of sticky secondary metabolites that can lead to the aggregation of the trichomes13. In the present protocol, the isolated trichomes in the final isolation step were recovered as a dry fine powder.
The extremely low temperatures maintained throughout the present protocol are likely to suffice in meeting the objectives of conventional buffer media-preventing RNA degradation (as long as the samples are kept submerged) and facilitating downstream molecular applications. In order to validate this assumption, we extracted RNA from the isolated trichomes and estimated the RNA integrity using commercially available kits. The obtained high RIN values (9.4 to 10) and distinct gel chromatography (Figure 11) clearly indicate a high RNA integrity. The RNA samples were further analyzed by RNAseq, and >20 M high-quality reads were obtained from all libraries.
In order to validate that, indeed, the mRNA from the isolated trichomes fraction is enriched in trichome-expressed genes, we compared our gene expression results for the inflorescence and corresponding isolated stalked trichomes to previous results presented by Livingston et al.14, which characterized the gene expression of purified trichomes of Cannabis (adapted from their Supplemental Table 1). Their protocol consisted of blending in aqueous buffer, filtering through sieves and, finally, trichome purification using a Percoll gradient. Trichome-expressed genes were characterized as those most highly correlated (p > 0.95) with the gene expression of cannabidiolic acid synthase (CBDAS), a trichome-specific gene marker14. The comparison of gene expression between the results obtained with the present protocol and those presented in Livingston et al.14 show that the 12 most highly enriched genes in our trichome fraction were also enriched in the Livingston et al. study14, including the trichome marker gene CBDAS (Table 1). Notably, a significantly higher enrichment ratio was obtained from the present protocol. These results confirm the validity of the present protocol for the study of trichome-enriched gene expression.
In addition, we compared the expression data of five actin genes of Cannabis, since actin is frequently used as a reference gene for transcriptome studies (Table 2). Our results for both the isolated stalked and sessile trichomes were comparable to those reported by the Livingston et al. study14.
Finally, we calculated the expression and trichome enrichment data for the chlorophyll a-b binding protein family genes, which are presumably not trichome enriched (Table 3). Indeed, the 12 members of this gene family showed a trichome enrichment factor of <1, serving as a negative control for trichome enrichment. These combined results indicate the unique expression patterns of the inflorescence and isolated trichome fractions and support the trichome integrity and quality of the isolated trichome fraction.
Figure 1: Setup for loading the 1 L glass beaker. The 1 mm screen door mesh is secured to the sides of the glass beaker via a few rubber bands (not shown). The 1 L glass beaker should be loaded in an upright position. Please click here to view a larger version of this figure.
Figure 2: Setup of the flour sifter. The flour sifter is fitted with a 350 µm micro-sieve via rubber bands at the bottom (not shown). Please click here to view a larger version of this figure.
Figure 3: Setup of the first step of trichome isolation. Please click here to view a larger version of this figure.
Figure 4: Setup of the glass beaker. The glass beaker fitted with three layers of 1 mm screen-door (mosquito) mesh secured via rubber bands at the opening. The opening of the beaker is pointing downward. Please click here to view a larger version of this figure.
Figure 5: Setup of the micro-sieve cone structure used for trichome isolation. Please click here to view a larger version of this figure.
Figure 6: Setup for the dipping and shaking motion of the micro-sieve containing isolated trichomes. The horizontal/vertical movement should resemble a tea bag infusion. The setup is identical for each mesh size used. Please click here to view a larger version of this figure.
Figure 7: Setup for transferring the final isolated trichomes. The final isolated trichomes are transferred to a labeled 1.5 mL tube. All the utensils should be pre-chilled with LN. Please click here to view a larger version of this figure.
Figure 8: Flowchart outlining the protocol. The isolation of glandular capitate trichomes from C. sativa involves three steps: (1) the initial detachment of the trichomes from the plant source and passage via two sieves (1 mm and 350 µm), (2) passing the plant material via five micro-sieves of decreasing pore sizes (150 µm, 105 µm, 80 µm, 65 µm, and 50 µm), and (3) the collection of the isolated trichomes into a pre-chilled 1.5 mL tube. Retained tissue from any of the micro-sieve stages may similarly be collected. Please click here to view a larger version of this figure.
Figure 9: Microscopy images of the isolated trichomes using the current protocol. (A,B,C) Glandular capitate sessile trichomes from C. sativa fan leaves. (A) Mid-isolation, from micro-sieves with pore sizes ranging from 105 µm to 150 µm. Note the retention of large amounts of green, non-trichome material. (B) End of isolation, from micro-sieves with pore sizes ranging from 50 µm to 65 µm, contaminated with plant source (red arrow), and (C) free of contamination. Scale bars = (A) 100 µm, (B,C) 50 µm. Please click here to view a larger version of this figure.
Figure 10: Microscopy images of the isolated trichomes. (A,B) Glandular capitate trichomes from C. sativa: (A) stalked type, isolated using the current protocol, and (B) sessile type, isolated using the current protocol. (C) Stalked trichomes isolated using a conventional protocol. The low yield and lack of uncompromised heads of the trichomes are apparent. (D,E) Stalked glandular capitate trichomes isolated using (D) the present protocol and (E) the conventional protocol. (F) Glandular stalked trichomes isolated from 65 µm to 105 µm micro-sieves (close-up). Note the high isolation of stalked trichomes, with detached heads and stalk parts. The red arrows indicate cystolithic trichomes, the blue arrows indicate a few stalk parts from the glandular capitate stalked trichomes, and the yellow arrows indicate separated disc cells. Scale bars = (A,C,D,E,F) 100 µm, (B) 50 µm. Please click here to view a larger version of this figure.
Figure 11: Analysis of the integrity of RNA extracted from the trichomes. Chromatographs and gels of extracted RNA samples from glandular capitate stalked and sessile trichomes isolated using the present protocol. Results for the RNA from (A–D) the stalked trichomes of the four cultivar lines used in this study, CS-11, CS-12, CS-13, and CS-14, respectively, and (E–H) the sessile trichomes, CS-11, CS-12, CS-13, and CS-14. Please click here to view a larger version of this figure.
Table 1: Comparison of the trichome-enriched gene expression obtained with RNAseq using the present protocol to the results obtained by Livingston et al.14. The 12 genes displaying the highest correlation to cannabidiolic acid synthase (CBDAS) expression by Livingston et al.14 (from their Supplemental Table Data S4, adapted with permission from Wiley publishers), considered a trichome-specific gene marker, were chosen for comparison. Please click here to download this Table.
Table 2: Comparison of the gene expression (FPKM) of five Cannabis actin genes of isolated stalked and sessile trichomes from the present protocol (our gene expression data from Cannabis variety CS-1111 (Var CS-11)) with previously published results adapted from Livingston et al.14. The results of Livingston et al.14 were presented based on the Finola FN reference genome, and their data were translated to the CS10.2 reference genome. Please click here to download this Table.
Table 3: Gene expression (FPKM) of Cannabis chlorophyll a-b binding genes of whole inflorescences and isolated stalked trichomes from the present protocol (our gene expression data from Var CS-11). Please click here to download this Table.
Compared to the currently available trichome isolation protocols, two main modifications are described in the present protocol. These include the detachment of the trichomes from the plant material using dry ice in the initial step and substituting LN for the commonly used liquid buffer medium. The first modification for C. sativa trichome purification is based on an earlier protocol that introduced the use of crushed dry ice to detach the trichomes from geranium pedicels5. While traditional trichome isolation protocols generally use a small (50 mL) test tube, a 1 L glass beaker was used in this study with a generous amount of crushed dry ice, allowing a larger volume of initial plant material (up to 10 g) to be processed. Furthermore, in the present protocol, two additional passages of the trichomes through micro-sieves were introduced in the first isolation step. The first step employs a vertical up-and-down vigorous saltshaker-like motion, and the second employs a horizontal sifting motion through the micro sieve in the flour sifter. The combined horizontal and vertical sieving motions are suggested to promote an enhanced separation based on the trichome size and shape.
The second, most significant modification addresses the isolation medium in which the passage through the micro-sieves is conducted. In the present protocol, the conventional aqueous isolation medium is completely replaced by LN. A previous protocol8 also used LN in the first step for trichome isolation, separating trichomes from the plant material (by grating the flower material against a mesh sieve), but continued with a liquid buffer extraction and Percoll density separation. Substituting LN in place of the conventional isolation medium eliminates the need for special procedures associated with the toxic components of the conventional isolation medium, allowing a relaxed workflow and high throughput.
A key feature of the present protocol relies on the propensity of dry ice and LN to sublimate quickly at room temperature. This feature enables their generous application throughout the isolation process. In contrast, using a large amount of the conventional isolation buffer would lead to technical complications concerning handling its large volumes.
While the isolation process in the present protocol promotes the retention of the fragile head structure of the glandular capitate stalked trichome, the conventional protocols promote the separation of the disc cells from the remaining gland material, which is easily washed away, leaving only the disc cells. This phenomenon can clearly be seen in our results comparing a conventional isolation protocol to our new protocol (Figure 10D,E). Besides the microscopic observation, the quality analysis of the RNA extracted from the trichomes (isolated using this protocol) clearly indicates that the RNA integrity is retained during the isolation process and that it is suitable for transcriptomic analysis.
Furthermore, the gene expression results from our isolated trichome fraction indicate that the fraction is highly enriched in established trichome-expressed genes (Table 1) and, reciprocally, that non-trichome-expressed genes are unrepresented (Table 3).
The main limitation of the present protocol is its exclusive suitability for isolating trichomes according to their propensity to pass through micro-sieves. While the protocol cannot be directly applied to trichome isolation using a density gradient, it is suitable for isolating the trichomes prior to the density gradient step, if necessary. In this study, the trichomes isolated using this protocol did not undergo a proteomic characterization, and their suitability for such an application needs to be verified. However, as the RNA samples extracted from both stalked and sessile trichomes exhibited no degradation, as indicated by the chomatographs and high RIN values, it can be assumed that the trichomes isolated with the present protocol are likely to meet the requirements needed for proteomic analysis.
The isolation degree (from the initial plant source and other types of trichomes) of the glandular capitate sessile trichomes from the fan leaves is very high, as the fan leaves lack the glandular capitate stalked trichome type15,16, and the cystolithic and bulbous trichomes are easily isolated from the sessile trichomes due to their structural and size differences. However, with regard to the isolation of glandular capitate stalked type trichomes, it is best to address their isolation degree as enrichment, as the plant source, the mature female inflorescence, contains all four trichome types, in addition to the pre-stalk type14. It stands to reason that most of the isolated glandular capitate trichomes are of the stalked type, as their elevated position (in relation to the epidermal-bound sessile trichomes) enhances the likelihood of them colliding with a dry ice particle and becoming detached from the plant. Furthermore, the juncture connecting the head and stalk parts of the stalked trichome is considered a point of weakness associated with the abscission of the gland head17. Thus, it would be accurate to refer to the stalked trichome portion as a highly enriched fraction with potentially low contamination of the sessile type. However, further analysis was not conducted to validate this assumption.
The suitability of the present protocol to extract glandular trichomes from other plant species is yet to be determined. The high trichome density in C. sativa presumably contributes to the success of the present protocol. However, it seems unlikely that this is the sole reason, as a relatively large plant material source (up to 10 g per isolation cycle) can be effectively processed. Other studies have presented detailed protocols for isolating different trichome types from other plants, for example, rosette leaf trichomes from Arabidopsis thaliana18,19. While the applicability of the present protocol for isolating other trichome types was not studied in this work, elements presented in this work, especially the substitution of LN for the isolation buffer, are likely to improve the trichome isolation process.
The authors have nothing to disclose.
The authors acknowledge financial support from CannabiVar Ltd. All plant material was generously provided by Professor Hinanit Koltai from the Volcani Center, Israel.
Bioanalyzer RNA Pico 6000 chip | Agilent, Germany | Reorder number 5067-1513 | Lab-on-a-chip system |
Transsonic-310 | Elma, Germany | D-78224 | Ultrasonic cleaning unit |
TruSeq RNA Sample Prep Kit v2 | Illumina, USA | RS-122-2001 | Sample preperation for RNA sequencing library |
Spectrum Plant Total RNA Kit | SIGMA-ALDRICH, USA | STRN50-1KT | Plant Total RNA Kit |
Nylon micro-sieve with a mesh size of 350 µm (40 x 40 cm or larger than the circumference of the flour sifter) | Sinun Tech, Israel | r0350n350210 | Nylon screen aperture |
Nylon micro-sieve with mesh size of 150 µm (size of 30 x 30 cm) | Sinun Tech, Israel | r0150n360465 | Nylon screen aperture |
Nylon micro-sieve with mesh size o 105 µm (size of 30 x 30 cm) | Sinun Tech, Israel | r0105n320718 | Nylon screen aperture |
Nylon micro-sieve with mesh size o 80 µm (size of 30 x 30 cm) | Sinun Tech, Israel | r0080n370465 | Nylon screen aperture |
Nylon micro-sieve with mesh size o 65 µm (size of 30 x 30 cm) | Sinun Tech, Israel | r0065n340715 | Nylon screen aperture |
Nylon micro-sieve with mesh size o 50 µm (size of 30 x 30 cm) | Sinun Tech, Israel | r0080n370465 | Nylon screen aperture |
Up to 10 g of frozen plant material (stored in -80 oC or liquid nitrogen) | |||
Suitable gloves for handling low temperatures | |||
Safety goggles | |||
1 mm screen door (mosquito) mesh (strip of 30 x 100 cm) | |||
Large strainer (colander) with holes approximately 5 mm | |||
1 L glass beaker | |||
1 block of dry ice (0.5-1 kg) | |||
Hammer and hard flat object | |||
Two 5 L plastic containers | |||
Rubber bands | |||
Large flour sifter or sieve strainer- preferably one with a detachable plastic ring on the circumference | |||
Several large and small round bottom stainless steel containers. One of them should be larger than the flour sifter's circumference (approximately 40 cm in diameter), to minimize the loss of the sifted mass outside the round bottome stainless steel container | |||
Pre-chilled (via liquid nitrogen) stainless steel spoon, spatula, and scoopula | |||
Clean plate | |||
Several clothespins | |||
Pre-chilled (via liquid nitrogen) labeled 1.5 mL tubes with holes poked on the lid with a sterile needle | |||
Two containers of liquid nitrogen | |||
1 cm wide painting brush |