0% found this document useful (0 votes)

7 views20 pages

Cannabis Glandular Trichomes Alter Morphology and Metabolite Content During Flower

Research on the changes experimented by the Cannabis plant and its glandular trichomes during flowering

Uploaded by

glennorch

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

0% found this document useful (0 votes)

7 views20 pages

Cannabis Glandular Trichomes Alter Morphology and Metabolite Content During Flower

Research on the changes experimented by the Cannabis plant and its glandular trichomes during flowering

Uploaded by

glennorch

We take content rights seriously. If you suspect this is your content, claim it here.

Available Formats

Download as PDF, TXT or read online on Scribd

You are on page 1/ 20

1365313x, 2020, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/tpj.14516 by Readcube (Labtiva Inc.), Wiley Online Library on [18/10/2022].

See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
The Plant Journal (2020) 101, 37–56 doi: 10.1111/tpj.14516

Cannabis glandular trichomes alter morphology and

metabolite content during flower maturation
Samuel J. Livingston1,† , Teagen D. Quilichini1,2,†, Judith K. Booth3, Darren C. J. Wong4, Kim H. Rensing5,
Jessica Laflamme-Yonkman1, Simone D. Castellarin4, Joerg Bohlmann1,3,4, Jonathan E. Page1,2 and A. Lacey Samuels1,*
1
Department of Botany, University of British Columbia, Vancouver, British Columbia, Canada,
2
Anandia Laboratories Inc., Vancouver, British Columbia, Canada,
3
Michael Smith Laboratories, University of British Columbia, Vancouver, British Columbia, Canada,
4
Wine Research Centre, University of British Columbia, Vancouver, British Columbia, Canada, and
5
Fibics Inc., Ottawa, Ontario, Canada

Received 2 April 2019; revised 28 July 2019; accepted 19 August 2019; published online 30 August 2019.
*For correspondence (e-mail [email protected]).
†
Co-first authors, these authors contributed equally to this work.

SUMMARY
The cannabis leaf is iconic, but it is the flowers of cannabis that are consumed for the psychoactive and
medicinal effects of their specialized metabolites. Cannabinoid metabolites, together with terpenes, are pro-
duced in glandular trichomes. Superficially, stalked and sessile trichomes in cannabis only differ in size and
whether they have a stalk. The objectives of this study were: to define each trichome type using patterns of
autofluorescence and secretory cell numbers, to test the hypothesis that stalked trichomes develop from
sessile-like precursors, and to test whether metabolic specialization occurs in cannabis glandular trichomes.
A two-photon microscopy technique using glandular trichome intrinsic autofluorescence was developed
which demonstrated that stalked glandular trichomes possessed blue autofluorescence correlated with high
cannabinoid levels. These stalked trichomes had 12–16 secretory disc cells and strongly monoterpene-domi-
nant terpene profiles. In contrast, sessile trichomes on mature flowers and vegetative leaves possessed red-
shifted autofluorescence, eight secretory disc cells and less monoterpene-dominant terpene profiles. More-
over, intrinsic autofluorescence patterns and disc cell numbers supported a developmental model where
stalked trichomes develop from apparently sessile trichomes. Transcriptomes of isolated floral trichomes
revealed strong expression of cannabinoid and terpene biosynthetic genes, as well as uncharacterized genes
highly co-expressed with CBDA synthase. Identification and characterization of two previously unknown
and highly expressed monoterpene synthases highlighted the metabolic specialization of stalked trichomes
for monoterpene production. These unique properties and highly expressed genes of cannabis trichomes
determine the medicinal, psychoactive and sensory properties of cannabis products.

Keywords: Cannabis sativa, glandular trichomes, fluorescence microscopy, electron microscopy, metabolite
profiling, transcriptomics, cannabinoids, terpenes, DAPI.

INTRODUCTION

Cannabis sativa L. (cannabis) flowers are consumed for cannabidiolic acid (CBDA) (reviewed by Potter, 2009 and
medicinal and recreational purposes based on the proper- Pertwee, 2014). When heated, these molecules decarboxy-
ties of their specialized metabolites (i.e. cannabinoids and late to their bioactive metabolites, tetrahydrocannabinol
terpenes). The metabolites are abundantly produced in the (THC) or cannabidiol (CBD), respectively. THC interacts
glandular trichomes on female flowers, which represent with receptors in the human endocannabinoid system to
the valued ‘marijuana bud’, while male flowers are typi- produce psychoactive and therapeutic effects, while the
cally not consumed for recreational or medical purposes non-intoxicating CBD has distinct pharmacological proper-
due to the scarcity of glandular trichomes. Cannabinoids ties (Mechoulam, 1970; Pertwee, 2008). Cannabis terpenes
synthesized in the glandular trichomes on female include monoterpenes and sesquiterpenes, which are vola-
flowers include tetrahydrocannabinolic acid (THCA) and tile and contribute to the fragrance of cannabis flowers and

© 2019 The Authors 37

The Plant Journal © 2019 John Wiley & Sons Ltd
1365313x, 2020, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/tpj.14516 by Readcube (Labtiva Inc.), Wiley Online Library on [18/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
38 Samuel J. Livingston et al.

cannabis products (Hillig, 2004; Fischedick et al., 2010; Cannabis stalked trichomes resemble sessile/peltate tri-
Booth et al., 2017; Booth and Bohlmann, 2019; Mudge chomes superficially, but they are raised upon a large stalk
et al., 2019; Zager et al., 2019). Terpenes have been pro- (Potter, 2009). In both types, a disc of secretory cells
posed to act synergistically with cannabinoids to elicit the develop a large extracellular storage cavity formed by
pharmacological effects of cannabis consumption (Russo, delamination of the apical cell wall (Kim and Mahlberg,
2011). Despite the economic and medicinal importance of 1991, 2003). In fact, the morphological similarity between
cannabis glandular trichomes, the properties and relative stalked and sessile gland heads has resulted in some
contributions of the different types of cannabis glandular reports classifying these two trichome types as a single
trichomes remain poorly understood. group (Dayanandan and Kaufman, 1976). Additionally, a
On female cannabis flowers, three types of glandular tri- developmental relationship has been proposed, where ses-
chome have been described based upon their surface mor- sile trichomes on flowers may represent a premature stage
phology: bulbous, sessile, and stalked (Hammond and of stalked trichome development (Hammond and Mahl-
Mahlberg, 1973). Bulbous trichomes are the smallest in berg, 1977; Potter, 2009). It remains unclear if stalked tri-
size and produce limited specialized metabolites (Potter, chomes are distinct from sessile trichomes simply by the
2009). Sessile trichomes of cannabis sit on the epidermis presence or absence of a stalk, or whether stalked and ses-
with a short stalk and have a globose head comprised of a sile trichomes have additional unique characteristics.
multicellular disc of secretory cells and a subcuticular The goal of this study was to proceed beyond the known
metabolite storage cavity (Hammond and Mahlberg, 1977). surface features of cannabis glandular trichomes by char-
By comparison, stalked trichomes of cannabis have a simi- acterizing their internal anatomy, biochemistry and tran-
larly shaped, slightly larger, globose head elevated several scriptome. The first objective was to define the metabolite
hundreds of microns above the epidermal surface by a storage cavity characteristics of cannabis stalked and ses-
multicellular stalk (Mahlberg and Kim, 2004; Potter, 2009). sile trichomes using in vivo intrinsic fluorescence. The sec-
The relative contributions of the two major glandular tri- ond objective was to test the hypothesis that stalked
chome types, sessile and stalked, to the cannabinoid and trichomes on flowers develop from sessile-like precursors.
terpene profiles of cannabis flowers are unknown. Thirdly, we performed chemical and transcriptomic analy-
Anatomic and metabolic specialization of glandular tri- sis of stalked trichomes, and sessile-like premature stalked
chomes has been demonstrated in other well characterized trichomes, to determine whether metabolic specialization
glandular trichome systems, including Lamiaceae and occurs in cannabis trichomes, and to explore the gene co-
Solanaceae (Simmons and Gurr, 2005; Huang et al., 2008; expression networks of CBDA synthase during cannabinoid
Schilmiller et al., 2009). For these taxa, glandular tri- biosynthesis and storage.
chomes are considered to fall into two categories: peltate
trichomes, which are analogous to sessile as they sit on RESULTS
the organ surface, and capitate trichomes that have a long
Stalked glandular trichomes have unique blue intrinsic
stalk above the epidermis (Glas et al., 2012; Huchelmann
fluorescence
et al., 2017). The peltate type have eight secretory disc
cells that produce an abundance of metabolites, often In this work, we used ‘Finola’, a hemp variety known for its
enriched in monoterpenes, which accumulate in a large short stature and accelerated flowering time, making it
subcuticular cavity, as detailed for peppermint (Mentha 9 economically important and useful for research purposes
piperita) (Turner et al., 2000) and lavender (Lavandula pin- (van Bakel et al., 2011; Sawler et al., 2015). The female
nata) (Huang et al., 2008). The second category, capitate plants produced clusters of individual flowers (florets) sur-
glandular trichomes, have a large stalk holding a small rounded by bracts (Figure 1a). Each floret consisted of two
head of one or few apical secretory cells and produce protruding styles attached to an ovule, which was enclosed
metabolites that are not stored in a subcuticular cavity by a calyx covered with metabolite-rich trichomes (Fig-
(Glas et al., 2012). Within the genus Solanum, different ure 1b,c). Calyces and bracts were covered with non-glan-
subtypes of capitate trichomes produce specific metabo- dular hairs, as well as glandular trichomes. Trichome
lites such as acyl sugars or terpenes (Schilmiller et al., density was highest on calyces, therefore these were used
2009; McDowell et al., 2011). In cannabis, the stalked tri- for the characterization of glandular trichomes. Although
chomes produce more total cannabinoids than sessile tri- different trichomes on the calyx are apparent with conven-
chomes, which may be due to the relative differences in tional scanning electron microscopy (SEM) (Figure 1c;
trichome head diameter (Turner et al., 1978). The terpene Video S1), trichome morphology was sometimes damaged
profiles produced by individual trichome types have not by sample preparation. Therefore, we used cryo-SEM to
been investigated. Whether specialization of metabolite examine the trichomes in their native state, without fixa-
production exists between cannabis stalked and sessile tri- tion, to view the external morphology while retaining the
chomes remains unknown. metabolites of the storage cavity. The heads of stalked

© 2019 The Authors

The Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2020), 101, 37–56
1365313x, 2020, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/tpj.14516 by Readcube (Labtiva Inc.), Wiley Online Library on [18/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Cannabis glandular trichomes 39

Figure 1. The three types of glandular trichomes on

female cannabis flowers.
(a) (b)
(a) Flowering female ‘Finola’ plant; inset: isolated
floral cluster containing numerous calyces bearing
densely populated glandular trichomes. (b) Dissect- style calyx
ing microscope image of the calyx and styles of an
individual female flower. (c) Image of the calyx of
an individual female flower using conventional
SEM; scale bar 500 lm. (d–f) Cryo-SEM images of
the three types of cannabis glandular trichomes, (c)
classified as stalked (d), sessile (e), and bulbous (f);
scale bars 20 lm.
calyx
style

(d) (e) (f)

glandular trichomes were elevated above the epidermis on arrangements of metabolites were observed in sessile tri-
a multicellular stalk (Figure 1d) and sessile trichomes sat chomes from leaves (Figure 2d) and anthers (Figure 2e).
directly on the epidermis (Figure 1e). Bulbous trichomes Bulbous trichomes observed on leaves and florets had
had diminutive size and a small storage cavity (Figure 1f). blue-shifted cells and a very small red-shifted cavity (Fig-
To investigate how the trichomes accumulate lipidic ure 2f). The fluorescence emission and droplet arrange-
metabolites in the storage cavity, the internal structures of ments observed in ‘Finola’ trichomes are consistent with
each type of trichome were probed using multi-photon trichomes of high-THCA medicinal varieties ‘Purple Kush’
imaging. We developed a two-photon laser scanning fluo- and ‘Hindu Kush’ (Figure S1a–c). Pure cannabinoids fluo-
rescence microscopy approach, where a red-shifted pulsed resce with a peak emission at 430 nm (Hazekamp et al.,
laser is used to penetrate deep into live tissues and excite 2005), which falls within the broad 460 nm peak observed
the intrinsic fluorescence of metabolites. Spectral scans of from stalked glandular trichomes. The interpretation that
trichome cavities revealed substantial emission intensity the strong blue-shifted fluorescence of stalked trichomes
between 420–530 nm (Figure 2a). Two spectral emission reflects high cannabinoid content is supported by previous
windows were used for imaging: 420–460 nm, which was reports showing stalked trichomes have the highest
false-coloured teal; and 495–540 nm, which was false- cannabinoid storage of the three trichome types (Turner
coloured red (Figure 2a). Merging these spectral windows et al., 1978; Mahlberg and Kim, 2004; Potter, 2009). To fur-
produced a high resolution three-dimensional data set of ther examine the metabolite arrangements in storage cavi-
live cannabis trichomes, which revealed striking differ- ties, transmission electron microscopy was performed,
ences in fluorescence emission and the storage cavity revealing many small droplets of metabolites within the
internal morphology between stalked, sessile and bulbous cavity of sessile trichomes and a large central droplet
trichomes of calyces. Within the extracellular storage cav- within the cavity of stalked trichomes (Figure 2g,h). Com-
ity, stalked glandular trichomes contained one large dro- bined, our microscopy approaches show that trichomes on
plet with blue-shifted fluorescence and some smaller mature cannabis organs can be distinguished not only by
droplets in the surrounding dark non-fluorescent matrix whether they sit on the epidermis or a stalk, but also by
(Figure 2b). Sessile trichomes from calyces (Figure 2c) dis- their intrinsic fluorescence and the arrangements of
played numerous small droplets with red-shifted fluores- metabolites in their storage cavities. The strong blue aut-
cence within the cavity. Similar fluorescence and ofluorescence of the stalked glandular trichomes,

© 2019 The Authors

(a) Intrinsic Fluorescence of Trichomes (b)

1600
Average Intensity

1400 Sessile
1200 Stalked
1000
800
600
400
200
0
400 420 440 460 480 500 520 540 560 580
Wavelength (nm)
(c) (d)

(g) (h)
(e) (f)
♂

Figure 2. Glandular trichomes exhibit distinct intrinsic fluorescence and metabolite organizations.
(a) Plot of mean SD emission intensity against wavelength (nm) shows trend towards lower wavelength emissions from stalked trichome cavities. (b-f) Multi-
photon microscopy images of the glandular trichomes on cannabis plants, including female calyx stalked (b), calyx sessile (c), leaf sessile (d), male antherial ses-
sile (e), and calyx bulbous (f), revealing distinct organization and intrinsic fluorescence of the metabolites stored in the storage cavity. (g, h) Transmission elec-
tron micrographs of sessile glandular trichome (g) with many droplets in storage cavity, and a stalked glandular trichome (h) with large central droplet of
metabolites in storage cavity. Scale bars 25 lm.

contrasted with red-shifted autofluorescence in sessile and samples were fixed, cleared, and stained with DAPI to
bulbous trichomes, provide diagnostic characteristics for show nuclei (Figure 3). Sessile trichomes consistently had
each trichome type. exactly eight cells on mature leaves (100%, n = 21; Fig-
ure 3a) and on mature calyces (78.6%, n = 14; Figure 3b).
Stalked glandular trichomes have many secretory disc
In contrast, the large stalked glandular trichomes on
cells
mature calyces had the largest and most variable number
If stalked trichomes are simply sessile trichomes elevated of cells, typically 12–16 cells (Figure 3c). Because the
on a multicellular stalk, then the arrangement and number inconspicuous storage cavities of bulbous trichomes could
of secretory disc cells forming the base of the globose not be distinguished from undifferentiated pre-secretory
head should be similar between stalked and sessile. Pre- trichomes after fixation and clearing, only stalked and ses-
liminary observations suggested that cannabis floral glan- sile trichomes with mature cavities were included in this
dular trichomes had ‘a flattened disc of few to many cells’ analysis. Stalked trichomes on mature calyces had statisti-
(Hammond and Mahlberg, 1973). Subsequent reports indi- cally significantly more cells in their secretory discs than
cated that sessile (Potter, 2009) and stalked trichomes the sessile trichomes of either mature calyces or vegetative
(Dayanandan and Kaufman, 1976) have eight cells, or that leaves (Figure 3d). In addition, the trichome disc cell num-
stalked trichomes have eight to thirteen cells (Hammond bers of a high-THCA cannabis variety, ‘Purple Kush’, were
and Mahlberg, 1977). However, the number of cells in dif- quantified to test if the larger number of cells within
ferent trichome types was difficult to resolve (Potter, 2009). stalked trichomes was also observed in ‘drug-type’ vari-
To test the hypothesis that the different trichome types eties. As with the hemp variety, sessile trichomes had
had different numbers of cells in their secretory disc, exactly eight cells while stalked trichomes had the largest

© 2019 The Authors

Figure 3. Stalked glandular trichomes on mature

cannabis flowers have a proliferation of cells in (a) (b)
their secretory disc.
(a–c) Multi-photon microscopy images of disc cell
nuclear labelling for leaf sessile (a), calyx sessile (b)
and calyx stalked (c) glandular trichomes; scale bars
25 lm. (d) Box plot showing disc cell number of
leaf sessile (n = 21), calyx sessile (n = 14), and calyx
stalked (n = 22) trichomes. Center line indicates
median; box limits indicate 1st (lower) and 3rd (up-
per) quartiles; whiskers indicate minimum and max-
imum data points; black circles indicate outliers;
*P < 0.001, non-parametric pair-wise comparisons
using Dunn’s procedure with a Bonferroni correc-
tion.

(c) (d) Secretory Disc Cells/Trichome

20 *
*
18

Disc Cell Number

16
o

14
12
o

10
o

8
Leaf Calyx Calyx
Sessile Sessile Stalked

and most variable number of cells in the secretory disc Floral stalked trichomes develop from trichomes that
(Figure S1d). These data indicated that cannabis sessile appear sessile
and stalked glandular trichomes are distinct, with stalked
The hypothesis that stalked trichomes develop from sessile
glandular trichomes having greater numbers of secretory
trichomes was historically difficult to test using surface
disc cells.
morphology alone. Time-lapse movies of intact calyces on
Stalked glandular trichomes dominate the mature female dense floral clusters of intact plants, or isolated calyces in
flowers tissue culture, proved technically challenging to achieve
within the constraints of regulatory permits. Therefore, we
To test the hypothesis that stalked trichomes may develop
re-examined the trichomes of immature calyces using our
from sessile trichomes on cannabis florets, the distribu-
diagnostic criteria for sessile and stalked trichomes, rea-
tions of stalked and sessile trichomes on the cannabis
soning that if sessile trichomes were maturing into stalked
calyx were mapped during development. SEM with an
trichomes, then the intrinsic fluorescence properties and
image tiling and collection system (Atlas) was used to
droplet organization in the storage cavity should change
obtain maps of whole calyces that could be magnified to
from the red-shifted small droplets of sessile trichomes on
single trichome resolution (Video S1). Stalked trichomes
calyces, vegetative leaves, and anthers (Figure 2c–e) to the
were scarcely observed on short calyces (Figure 4a),
blue-shifted large droplets seen in stalked trichomes (Fig-
whereas more mature, longer calyces possessed an abun-
ure 2b). The intrinsic fluorescence properties of trichome
dance of stalked trichomes (Figure 4b). Quantification of
cavities were imaged with two-photon microscopy (Fig-
trichomes using the Atlas system revealed a positive corre-
ure 4d). The largest droplet in the cavity and the ratio of
lation between calyx length and the percentage of stalked
total cavity fluorescence intensity in the ‘teal channel’ of
trichomes present (R2 = 0.74) (Figure 4c). Throughout the
420–460 nm over the ‘red channel’ of 495–540 nm were
development of the calyx, the proportion of stalked tri-
measured for stalked and sessile trichomes (Figure 4e). A
chomes increased from less than 30% to 80–90% for the
KruskalWallis test was conducted to determine if there
most mature calyces. These data support the hypothesis
were differences in fluorescence ratio and droplet size
that stalked trichomes develop from sessile glandular tri-
between stalked trichomes from calyces, sessile trichomes
chomes. In addition, the high proportion of stalked tri-
from calyces and sessile trichomes from vegetative leaves.
chomes on mature florets indicates that the characteristics
Fluorescence ratios and droplet size differed significantly
of the stalked trichomes will dominate the pharmacological
between the three glandular trichome groups,
and sensory properties of cannabis flowers.
© 2019 The Authors
The Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2020), 101, 37–56
1365313x, 2020, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/tpj.14516 by Readcube (Labtiva Inc.), Wiley Online Library on [18/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
42 Samuel J. Livingston et al.

(a) (b) (c) Stalked Glandular Trichomes Increase with Calyx Maturation
100
90

Percentage Stalked
80
70
60
50 R² = 0.7427
40
30
20
10
0
0 1 2 3 4 5 6 7 8
Length of Calyx (mm)

(d) (f)

(e) Intrinsic Fluorescent Properties of Cannabis Glandular Trichomes (g)

6
Cavity Fluorescence Ratio

Leaf Sessile
5
Calyx Sessile
4 Calyx Stalked

0
0 20 40 60 80 100 120
Diameter of largest droplet in cavity (µm)

Figure 4. Sessile-like trichomes on calyces develop into stalked glandular trichomes.

(a, b) Atlas SEM images of an immature (a) and a mature (b) calyx; scale bars 0.5 mm. (c) Positive correlation of percentage of stalked trichomes and calyx
length. (d) Multi-photon microscopy image of three sessile trichomes on a calyx; scale bar 50 lm. (e) Quantitative measurements of cavity fluorescence ratio
(teal channel signal/red channel signal) of leaf sessile (n = 28), calyx sessile (n = 45), and calyx stalked glandular trichomes (n = 53). Dots represent individual tri-
chome measurements. (f) Nuclear labelling with DAPI of sessile trichomes on an immature calyx; scale bar 50 lm. (g) Box plot revealing disc cell numbers in
leaf sessile (n = 21), mature calyx sessile (n = 14), mature calyx stalked (n = 22), and immature calyx sessile (n = 32). Center line indicates median; box limits
indicate 1st (lower) and 3rd (upper) quartiles; whiskers indicate minimum and maximum data points; black circles indicate outliers; letters above boxes indicate
statistically significant groups (P < 0.001) determined by post-hoc non-parametric pair-wise comparisons using Dunn’s procedure with a Bonferroni correction
for multiple comparisons.

v2(2) = 80.588, P < 0.0005 and v2(2) = 90.310, P < 0.0005 of stalked trichomes. If the sessile trichomes on an imma-
respectively (Figure S2). Post-hoc pair-wise comparisons ture calyx are actually premature stalked trichomes, they
using Dunn’s procedure with a Bonferroni correction for are predicted to have greater than eight cells. If they are
multiple comparisons revealed statistically significant dif- similar to sessile trichomes on vegetative leaves (which do
ferences in fluorescence ratio and droplet size among each not produce stalked trichomes), they are predicted to con-
set of the three trichome types (Figure S2). Plotting teal:red tain only eight cells. The majority of sessile trichomes on
fluorescence ratio as a function of droplet size (Figure 4e) immature calyces had greater than eight cells (Figure 4f,g).
demonstrated that the sessile trichomes on the calyx dis- Even pre-secretory trichomes on calyces that lacked a stor-
played intermediate cavity droplet arrangements and fluo- age cavity were more likely to contain greater than eight
rescence ratios between the sessile trichomes of a cells (75%, n = 32) than to have only eight cells (25%,
vegetative leaf and the stalked trichomes of a calyx. There- n = 32). These data indicate that the majority of premature
fore, the fluorescence properties of the metabolites are trichomes sitting on the epidermal surface of developing
consistent with many of the apparently sessile trichomes calyces have cell numbers similar to stalked glandular tri-
on calyces representing premature stalked trichomes. chomes on mature calyces.
Examining the cell numbers in the secretory disc pro- Although most of the sessile trichomes on immature
vided a second line of evidence in support of the hypothe- flowers appear destined to become stalked as the calyx
sis that a subpopulation of sessile trichomes on matures, a scant subpopulation of sessile trichomes per-
developing flowers are a premature developmental stage sisted. On mature calyces, the secretory disc cell numbers
© 2019 The Authors
The Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2020), 101, 37–56
1365313x, 2020, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/tpj.14516 by Readcube (Labtiva Inc.), Wiley Online Library on [18/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Cannabis glandular trichomes 43

of sessile trichomes with no stalk were more variable (79% To specifically test if floral stalked trichomes had distinct
had eight cells, and 21% had greater than eight cells; chemical profiles from sessile trichomes, the metabolite-
n = 14) than mature vegetative leaves (Figure 4g). A bino- rich contents of individual trichomes’ storage cavities were
mial logistic regression was performed to ascertain the sampled using a microcapillary probe. The characteristic
effect of disc cell number on the likelihood of a trichome to blue intrinsic fluorescence of the stalked trichomes, excited
have a stalk on mature leaves and calyces. The logistic by a hand-held ultraviolet light (Figure 5e,f), provided a
regression model was statistically significant, and the useful diagnostic indicator to guide the microcapillary
model correctly classified 88% of cases (v2(1) = 33.687, probe to one trichome and extract the storage cavity con-
P < 0.0005). Based on our model, we could predict that tents (Figure 5g). The cavity contents were pooled and
86% of secretory trichomes with more than eight cells are analyzed with gas chromatography and liquid chromatog-
premature stalked trichomes, while 89% of secretory tri- raphy to profile the terpenes and cannabinoids, respec-
chomes with eight cells are sessile trichomes, like those of tively. As with the whole calyces, terpene profiles in
vegetative leaves. Taken together, these data and the mod- samples collected from stalked trichomes were dominated
eling indicate that stalked glandular trichomes develop by monoterpenes (92%), with a monoterpene:sesquiter-
from premature stalked trichomes with greater than eight pene ratio over 12 (Figures 5h and S4). As previously
cells. noted (Turner et al., 1978), sessile trichomes on the epider-
mis of mature calyces could not be sampled without cross-
Calyx glandular trichomes are rich in cannabinoids, with
contaminating with stalked trichomes, so sessile tri-
monoterpene-dominant terpene profiles
chomes’ storage cavity contents were sampled from
If the majority of floral sessile trichomes are in the process anthers and vegetative leaves. Consistent with the whole
of becoming stalked, then the sessile trichomes on imma- vegetative leaf results, the individual sessile trichomes
ture calyces are predicted to be biochemically more similar from vegetative leaves sampled by microcapillary con-
to stalked trichomes on mature calyces, and share fewer tained very low monoterpenes compared to sesquiterpe-
characteristics with sessile trichomes on other mature nes (Figures 5h and S4). Sessile trichomes from anthers
organs such as vegetative leaves and anthers. We tested also had lower monoterpene:sesquiterpene ratios than the
this prediction by sampling the metabolites of stalked and stalked trichomes (Figure 5h). Microcapillary sampling of
sessile trichomes using two methods: whole-organ immer- stalked and sessile trichomes revealed similar cannabinoid
sion in pentane (Figure 5a–d) and individual trichome sam- profiles between the trichome types (Figure 5i), in agree-
pling using a pulled glass microcapillary tube guided by ment with the whole-organ cannabinoid profiles. These
ultraviolet (UV) light (Figure 5e–i). Whole-organ immersion data indicated that mature calyces, and the stalked glandu-
of immature calyces (defined as no conspicuous stalked tri- lar trichomes that cover their surfaces, accumulate abun-
chomes, <4 mm calyx length), and mature calyces (defined dant cannabinoids and monoterpene-rich terpenes that are
as >70% stalked trichomes, >4 mm calyx length) revealed a distinct from sessile trichomes of anthers or vegetative
substantial increase in monoterpenes during calyx matura- leaves that have lower monoterpene:sesquiterpene ratios.
tion (Figure 5a). The terpene profile of individual ‘Finola’ Therefore, the bona fide sessile trichomes found on
hemp plants was highly variable (Figures S3 and S4), as anthers and vegetative leaves, which do not develop into
previously reported (Booth et al., 2017). However, the pro- stalked trichomes, have chemically distinct terpene profiles
portion of monoterpenes to sesquiterpenes was consis- from the premature stalked and mature stalked glandular
tently high in surface extracts from both immature and trichomes of cannabis flowers.
mature calyces, in contrast to the sesquiterpene-rich meta-
Isolated floral trichomes’ transcriptome is enriched in
bolic profile from vegetative leaves (Figures 5b and S3).
cannabinoid and monoterpene biosynthetic genes
Unlike the variation in monoterpene:sesquiterpene ratios,
individual cannabinoid components were similar between Transcriptomic studies from isolated trichomes have pro-
vegetative leaves and flowers, although floral tissue con- vided important insights into the gene expression underly-
tained more total cannabinoids than leaf (Figure 5c,d). ing the specialized metabolism specifically within
These chemical analyses indicate that on the flower, the glandular trichomes (Lange, 2015a; Huchelmann et al.,
transitory population of sessile trichomes are biochemi- 2017; Zager et al., 2019). A previous study of the cannabis
cally similar to stalked glandular trichomes, supporting a transcriptome examined various organs of cannabis
developmental model that the majority of sessile tri- including whole flowers at early, mid and late stages of
chomes are actually premature stalked trichomes. In addi- flowering (van Bakel et al., 2011); however, these floral
tion, the different terpene profiles of flowers compared transcriptome datasets included a mixture of glandular tri-
with vegetative leaves suggests specialization of terpene, chome types in addition to the underlying floret tissues. To
but not cannabinoid, metabolism in different cannabis explore potential differences in gene expression among
glandular trichomes. the isolated floral trichome types, trichomes were

© 2019 The Authors

Figure 5. Mature calyces and microcapillary-sam-

(a) (c)
pled stalked glandular trichomes have monoter-
pene-dominant terpene profiles.
(a) Absolute levels of monoterpenes and sesquiter-
penes from developing whole calyces dipped in sol-
vent showing increases in monoterpenes as calyces
mature. (b) The ratio of monoterpenes:sesquiterpe-
nes is high in both immature calyces, with predomi-
nantly premature stalked trichomes, and mature
calyces, with stalked trichomes. Leaves, which pro-
duce exclusively sessile trichomes, have more
Immature calyx Immature calyx
sesquiterpenes. (c) Absolute levels of cannabidiolic
(b) (d) acid, mean SE, n = 5 plants with three technical
replicates per plant. (d) Relative amounts of total
cannabinoids in immature calyx and mature calyx
compared to low levels in leaf; n = 5 plants with
Mean relative peak area

three technical replicates per plant. (e–g) A floral

cluster (e), illuminated with ultraviolet light (f) to
produce blue fluorescence from stalked glandular
trichomes to assist microcapillary sampling of
resin, as shown in (g). (h) The ratio of monoterpe-
nes:sesquiterpenes from microcapillary sampling of
Immature calyx Immature calyx storage cavities from anther sessile trichomes, flo-
ral stalked trichomes or leaf sessile trichomes. (i)
(e) (f) (g) The cannabinoid profiles of microcapillary-sampled
storage cavities of floral stalked trichomes (100 cav-
ities each) and leaf sessile trichomes (200 cavities
each), n = 3 plants with three technical replicates
per plant.

(h) (i)

physically isolated from ‘Finola’ flowers, and enriched by successfully mapped to both ‘Finola’ (FN) and ‘Purple
type using a Percoll step centrifugation. Trichomes were Kush’ (PK) reference transcriptomes (van Bakel et al., 2011;
separated from mid-stage flowers into three categories: Data S1; Table S1) were 7.8 M.
the very small bulbous; pre-stalked trichomes with inter- As a first step to understand the important metabolic
mediate size and red-shifted fluorescence; and mature pathways within each trichome type, highly expressed
stalked with their larger size and blue-shifted intrinsic fluo- genes were shortlisted and their functional annotation (i.e.
rescence. The pre-stalked category would include the MapMan BIN categories) assessed. Among the most highly
minor subfraction of trichomes that would persist as ses- expressed genes (ranked by descending fragment per kilo-
sile at maturity. RNA quality was assessed by gel elec- base of transcript per million (FPKM)) were those involved
trophoresis and NanoDrop spectroscopy. Nine cDNA in mitochondrial electron transport, lipid metabolism, sec-
libraries encompassing three biological replicates for bul- ondary metabolism (especially isoprenoids and ter-
bous, pre-stalked trichomes and mature stalked were con- penoids), glycolysis, citric acid cycle, and redox
structed and sequenced on an Illumina HiSeq 2500. In homeostasis in all trichome types (Data S1 and Figure S5).
total, 98 million (M) paired-end reads (101 bp) were Interestingly, no statistically significant differences
obtained after adaptor removal, trimming, and quality fil- between stalked and pre-stalked trichomes were found
tering with an average of 10 M paired-end reads/sample. (false discovery rate (FDR) > 0.05; Data S1). To determine
The average number of surviving reads that were which genes were differentially expressed in the floral

© 2019 The Authors

glandular trichomes during metabolite production, stalked trichomes. These findings are consistent with our chemical
(ST) and pre-stalked (PS) fractions were compared to iso- data, which do not support cannabinoid profile specializa-
lated bulbous trichomes (BU), which were considered a tion amongst stalked and sessile trichomes (Figure 5).
less metabolically-productive epidermal outgrowth control. As gene expression of plant specialized metabolic path-
Analysis using MAPMAN annotated BIN categories ways often show signatures of co-regulation (Schilmiller
showed higher enrichment for categories such as sec- et al., 2012; Lange et al., 2015b; Celedon et al., 2016), we
ondary metabolism, lipid metabolism, glycolysis and trans- performed gene co-expression analysis to prioritize novel
port in ST-BU compared to PS-BU (Figures 6a and S6 and candidate genes potentially relevant to cannabinoid and ter-
Data S2). Differential gene expression analysis revealed pene metabolism as well as metabolite export and storage
statistically significant differences (FDR < 0.05) between in the trichome. Our transcriptome compendium for co-ex-
the glandular trichomes that produce high amounts of pression consisted of the isolated trichome dataset, as well
metabolites (ST or PS), and the less productive bulbous tri- as the root, shoot, stem, and flowers (pre-, early- and mid/
chomes (Table S2 and Data S3). mature-flower) datasets of van Bakel et al. (2011). Using
Most genes involved in cannabinoid and monoterpene CBDA synthase as a query ‘guide’ gene, the top 300 co-ex-
biosynthesis (Figure 6b) were detected in all trichome pressed gene partners were retrieved (ranked by descend-
types (Figure 6c,d). The isolated trichomes expressed ing PCC values) (Data S4). These included candidate genes
higher FPKM levels of the cannabinoid and terpene biosyn- for enzymes of the hexanoate pathway that are currently
thetic genes in glandular trichomes compared to the uncharacterized, such as fatty acid omega-6 desaturase
whole-organ transcriptomic data on mid-flower, root, and (FAD). Some of the most highly co-expressed genes are
shoot from van Bakel et al. (2011) (Figure 6c,d). The iso- ATP-binding cassette (ABC) transporters of the ABCG sub-
lated glandular trichomes had higher expression of the family, which have been demonstrated to export lipids in a
genes encoding enzymes of the two biochemical pathways variety of plant systems (Hwang et al., 2016). In addition, a
leading to cannabinoid biosynthesis: geranyl diphosphate suite of genes encoding cell wall-modifying enzymes was
(GPP) production by the methylerythritol phosphate (MEP) co-expressed with CBDAS, which may play a role in the
pathway, as well as olivetolic acid production by extension expansion of the storage cavity cell wall to accommodate
and cyclization of hexanoate (Figure 6c,d). Our data sup- metabolite accumulation and storage. The co-expression
port the proposed pathway for hexanoyl-CoA biosynthesis relationships of CBDAS and selected highly co-expressed
by the action of lipoxygenase (LOX), hydroperoxide lyase genes were plotted for root, shoot, stem, flower and iso-
(HPL), and acyl-activating enzyme (AAE) (Stout et al., 2012) lated trichome data sets (Figure 7a). The co-expressed
(Figure 6c,d; Table S3). The genes encoding enzymes that genes were highly specific to glandular trichomes and often
act on hexanoyl-CoA, tetraketide synthase (TKS, Taura showed little to no expression in vegetative tissues (Fig-
et al., 2009; Gagne et al., 2012) and olivetolic acid cyclase ure 7a). These analyses identify co-expressed genes with
(OAC, Gagne et al., 2012) were expressed at significantly the cannabinoid pathway, expanding our knowledge of co-
higher levels in cannabis trichomes relative to all other expressed genes related to metabolite export and cell wall
genes (background) (Figure 6c,d and Data S3). Geranyl expansion of cannabis floral trichomes.
diphosphate synthase (GPPS), which occupies a central Three of the genes that were most highly co-expressed
position in the biosynthesis of both cannabinoids and ter- with CBDAS (PCC ≥ 0.99) encode terpene synthases (TPS).
penes, was expressed at significantly higher levels in These were cross-referenced with known TPS genes from
stalked and sessile glandular trichomes compared with ‘Finola’ (Booth et al., 2017) and other cannabis strains
bulbous (Figure 6c,d and Table S3). Genes encoding the (Zager et al., 2019). The most highly co-expressed tran-
final step of the cannabinoid pathway, which produce can- scripts mapped to TPS3 (Table S4), a myrcene synthase
nabigerolic acid (CBGA) by the combination of olivetolic (Booth et al., 2017). The product of this enzyme, myrcene,
acid and GPP by a prenyltransferase (PT), were highly was a highly abundant monoterpene detected with gas
expressed, as was cannabidiolic acid synthase (CBDAS) chromatography (GC) in this study (Figures S3 and S4).
(Taura et al., 1996; Figure 6c,d and Table S2). Expression The next two most highly co-expressed candidate terpene
of transcripts for CBDAS and predicted upstream precursor synthases (Figure 7a), CsTPS37FN and CsTPS38FN (named
pathways appeared to be coordinated, as many of the lat- following Zager et al., 2019), were not previously charac-
ter genes were strongly correlated (average Pearson’s pro- terized. Phylogenetic analysis placed these enzymes under
duct-moment correlation coefficient [PCC] >0.8) with the TPS-b clade of terpene synthases, suggesting they are
CBDAS expression (Figure 6e). This analysis revealed that monoterpene synthase enzymes (Figure S7). To exploit the
the transcripts belonging to the proposed cannabinoid transcriptome data and test the in vitro function of these
biosynthetic pathway are highly expressed in all floral putative TPS, CsTPS37FN and CsTPS38FN were produced
glandular trichomes, with select transcripts highly by heterologous expression in E. coli, and their activity
expressed in stalked trichomes relative to bulbous was tested with GPP as substrate. For CsTPS37FN, the

© 2019 The Authors

(a)
ST - BU PS - BU (b) Hexanoate MEP (c) (d)
Pathway Pathway
PK FN PK FN PK FN PK FN
PT PT
Photosynthesis Desaturase DXS CBDAS CBDAS
Major CHO metabolism GPPS_lsu GPPS_lsu
DXR GPPS_ssu
Minor CHO metabolism GPPS_ssu
Glycolysis LOX MCT IPP isomerase IPP isomerase
Fermentation HPL HPL
Gluconeogenesis/glyoxylate cycle CMK AAE1 AAE1
OPP HPL LOX1 LOX1
MDS AAE3 AAE3
TCA/org transformation
Desaturase Desaturase
Mito e- transport/ATP synthesis AAE HDS MCT MCT
Cell wall DXR
HDR DXR
Lipid Metabolism Hexanoyl-CoA HDR HDR
N-Metabolism DXS DXS
TKS IPP isomerase
Amino acid metabolism CMK CMK
S-Assimilation GPP synthase HDS HDS
OAC MDS MDS
Metal Handling (lsu, ssu)
Secondary metabolism Olivetolic acid PT GPP OAC OAC
Hormone metabolism TKS TKS

Pre-stalked
Stalked
Mid-flower
Bulbous
-stalked

Shoot
Root
Stalked
Mid-flower
Bulbous
Shoot
Root
Co-factor and vitamin metabolism
TPSb Log2 FPKM Raw Z-score
Tetrapyrrole synthesis CBGA 0 5 10 15 –1 0 1
Stress THCAS CBDAS
Redox
Polyamine metabolism THCA CBDA Monoterpenes
Nucleotide metabolism (e) Trichome Trichome
PCC with CsCBDAS_FN

Biodegradation of xenobiotics bg MEP GPP HEX OA CB

C1-metabolism 1.0
Misc
RNA 0.5
DNA
Protein 0.0
Signalling
Cell −0.5
Development
Transport −1.0
-log 10 (FDR)

GPP synthase (lsu)

Background

GPP synthase (ssu)

CHI−like

OLS
CMK

DXR

DXS

HDS

MCT

MDS

AAE1

AAE3

HPL

LOX1

PT
IPP isomerase
+
_
1 3 5

Gene

Figure 6. Transcriptomic analyses shows strong commitment to specialized metabolism by isolated cannabis trichomes.
(a) Enriched MapMan BIN functional categories of stalked (ST) versus bulbous (BU) and pre-stalked (PS) versus BU trichomes. Red/orange and blue colour
depicts scores (expressed as log10 FDR) of enriched categories for significantly upregulated and downregulated genes in each comparison. Both ‘Purple Kush’
(PK) and ‘Finola’ (FN) transcriptome annotations were used to complement each other’s enrichment results. (b) Genes of the cannabinoid biosynthetic pathway
from production of hexanoyl-CoA (green), olivetolic acid (orange), MEP pathway (purple), geranyldiphosphate (GPP- blue) and cannabinoids (red). Expression of
selected genes in (b), presented as Log2FPKM (c) and raw Z-score (d) to highlight strong expression and enrichment in isolated trichomes, respectively. The
whole-organ data of van Bakel et al. (2011) are shown for context, as specialized metabolite-related transcripts are enriched in isolated trichomes. (e) Pearson’s
product-moment correlation coefficient (PCC) of known genes involved in cannabinoid biosynthesis with CBDAS. Comparison of the observed PCC against all
other genes (bg, background) shows strong transcriptional correlation of cannabinoid pathway biosynthetic genes with CBDAS.

dominant product was terpinolene, while recombinant similarities between the mature stalked and floral prema-
CsTPS38FN produced (E)-b-ocimene (Figure 7b). Therefore, ture stalked trichomes’ transcriptomes provide an addi-
the co-expression analysis with CBDAS revealed previously tional line of evidence supporting the model that the
uncharacterized monoterpene synthases, and their activity majority of calyx sessile trichomes develop into stalked tri-
in producing terpinolene or (E)-b-ocimene is consistent chomes as the flowers mature.
with the monoterpenes detected by GC-MS (Figures S3
DISCUSSION
and S4). There was good agreement between the monoter-
pene-rich terpene profiles of floral stalked trichomes identi- Glandular trichomes represent ‘natural cell factories’
fied in our chemical analyses and the highly expressed (Huchelmann et al., 2017), producing large quantities of
monoterpene synthase genes (Table S4). These data pro- specialized metabolites. In cannabis, the anatomic and
vide another line of evidence that the stalked glandular tri- metabolic distinctions between glandular trichome types
chomes have specialized to produce not only large on different plant organs were unclear. The findings pre-
amounts of cannabinoids but monoterpene-rich terpene sented here reveal that the sessile and stalked trichomes of
profiles as well. cannabis differ not only in whether they sit on a large stalk
Overall, the transcriptomic analysis demonstrates that or directly on the epidermal surface, they also have distinct
the glandular trichomes on cannabis flowers are strongly fluorescent properties, number of cells in their secretory
dedicated to cannabinoid and terpene production. The disc, and terpene metabolite profiles (Figure 8). The
© 2019 The Authors
The Plant Journal © 2019 John Wiley & Sons Ltd, The Plant Journal, (2020), 101, 37–56
1365313x, 2020, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/tpj.14516 by Readcube (Labtiva Inc.), Wiley Online Library on [18/10/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Cannabis glandular trichomes 47

Tissue
(a) Root Early flower Bulbous trichome
Shoot Mid flower Pre-stalked trichome
Stem Preflower Stalked trichome

OAC CMK DXR FAD

12 FN11431.1 12 FN06062.1 12 FN10385.1 12 FN23649.1

9 9 9 9
Log2FPKM Co-expressed genes (Target)

6 6 6 6

3 3 3 3

PCC: 0.976 PCC: 0.964 PCC: 0.972 PCC: 0.983

0 0 0 0 0

12
0

ABC TPS3 TPS37 TPS38

12 FN06246.1 15 FN10459.1 15 FN15171.1 12 FN20433.1

12 12
9 9
9 9
6 6
6 6
3 3
3 3
PCC: 0.964 PCC: 0.996 PCC: 0.990 PCC: 0.991
0 0 0 0
0

12
0

Log2FPKM CsCBDAS_FN (Query)

(b) ×10 4
8
CsTPS37FN
d
7

4
i.s.

3
Total Ion Intensity

1
a b c
0
×10 4
CsTPS38FN
3.5
e
3

2.5

1.5

0.5

0
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Acquisition Time (min)

Figure 7. Co-expression analysis using CBDAS as a query reveals numerous genes putatively involved in metabolite biosynthesis, transport and storage.
(a) Co-expression relationship of CBDAS with selected highly co-expressed genes in various tissues and organs. The Pearson’s product-moment correlation
coefficient (PCC) depicts the co-expression strength of each interaction. (b) Products of CsTPS37FN (top panel) and CsTPS38FN (bottom panel) when incubated
with GPP. a: a-pinene, b: bpinene, c: limonene, d: terpinolene, e: (E)-b-ocimene. i.s., internal standard, isobutylbenzene.

stalked glandular trichomes of mature flowers have a glob- present in all cannabis trichome types (Mahlberg and Kim,
ular head consisting of an enlarged disc of greater than 2004). In light of the data that supports stalked trichome
eight secretory cells, which produce blue-shifted autofluo- development from sessile precursors on flowers, we
rescent secretions rich in cannabinoids and monoterpenes. hypothesize that co-incidental accumulation of monoterpe-
Conversely, sessile trichomes found on vegetative leaves nes and cannabinoids during stalk development remodels
have exactly eight secretory cells, produce red-shifted aut- the apoplastic storage cavity from a highly compartmental-
ofluorescent secretions with less cannabinoids and higher ized state to fewer, larger droplets at maturity. The autoflu-
proportions of sesquiterpenes. We also show that stalked orescence and two-photon microscopy revealed the in vivo
glandular trichomes are a terminal stage of development metabolite storage patterns of cannabis glandular tri-
arising from sessile-like glandular trichomes on cannabis chomes. Given the successful fluorescence activated cell
floral tissues. Trichome-specific transcriptomes generated sorting of tomato type VI glandular trichomes (Bergau
here support this developmental model. et al., 2016), the cannabis trichomes’ autofluorescence
Investigations into glandular trichome structure and observed here also may serve as a potential diagnostic fea-
development using intrinsic fluorescence microscopy are ture for future studies.
still in their infancy. Recent studies have found that the In cannabis, both sessile and stalked glandular tri-
glandular trichomes of some species, including tomato chomes secrete and store varying proportions of monoter-
(Bergau et al., 2015, 2016), Japanese catnip (Liu et al., penes and sesquiterpenes, in addition to the
2018), and lavender (Huang et al., 2008) exhibit autofluo- terpenophenolic cannabinoids. The cannabis ‘sessile’ glan-
rescent secretory cells and/or metabolites stored within dular trichome is morphologically like the canonical peltate
their trichome cavities. A report using isolated snap-frozen glandular trichomes (reviewed previously by Werker, 2000;
stalked glandular trichomes of cannabis revealed autofluo- Schilmiller et al., 2008; Glas et al., 2012; Tissier, 2012;
rescence in the secretory disc cells and their stalks, but not Huchelmann et al., 2017). These trichomes are well-charac-
within the storage cavity (Ebersbach et al., 2018). In con- terized in the Lamiaceae family, the members of which
trast to these authors’ findings, which used a longer excita- produce economically important monoterpene-rich essen-
tion wavelength, our two-photon in vivo imaging of intact tial oils including peppermint, lavender and basil. Interest-
cannabis glandular trichomes revealed striking patterns of ingly, our data demonstrate that the sessile trichomes of
autofluorescence within the storage cavities of both stalked cannabis leaves, while they contain both monoterpenes
and sessile types on flowers and leaves. Of particular inter- and sesquiterpenes, do not follow the monoterpene-domi-
est is the observation that a highly compartmentalized cav- nant pattern of other peltate-type trichomes.
ity was found in sessile trichomes of cannabis leaves and The monoterpene-rich cannabis stalked glandular tri-
flowers, but not the stalked trichomes that dominate the chomes have gland heads similar in structure to sessile- or
mature flowers. Our data contrasts with previous reports peltate-type trichomes, however, they sit on top of a multi-
using TEM that suggested a ‘vesiculated’ compartment is cellular stalk. Cannabis stalked glandular trichomes can be

Monoterpenes >> Sesquiterpenes

Proliferation of disc cells (> 8)

Monoterpenes < Sesquiterpenes

Eight disc cells

Epidermal surface

Cannabis sessile trichome Cannabis prestalked trichome Cannabis stalked trichome

Figure 8. Unique features of stalked, pre-stalked and sessile glandular trichomes of cannabis. Bona fide sessile glandular trichomes (left) contain numerous
small droplets of red-shifted intrinsic fluorescence, eight disc cells and sit directly on the epidermal surface. The sessile trichomes produce greater amounts of
sesquiterpenes relative to monoterpenes. Stalked glandular trichomes (far right) develop from pre-stalked trichomes (center). Both pre-stalked and stalked tri-
chomes have a greater number of disc cells, and they produce greater amounts of monoterpenes relative to sesquiterpenes. As the multicellular stalk lifts the
glandular trichome head above the epidermis, the extracellular storage cavity is filled with a large droplet of blue-shifted intrinsic fluorescence, which is corre-
lated with high cannabinoid content. Yellow indicates cuticle and cell wall surrounding the disc cells and storage cavity.

contrasted with capitate-type glandular trichomes of both quality was not different across cannabis trichome types
the Lamiaceae and the Solanaceae, which have an elon- and development, but the quantity clearly increased in flo-
gated stalk that subtends only one or few secretory cells, ral stalked glandular trichomes.
and their specialized metabolites are often secreted onto In contrast with the cannabinoids, specialization of ter-
the epidermal surface by pores formed in the cuticle (Tis- penes was found in our microcapillary sampling, with
sier, 2012; Huchelmann et al., 2017). Capitate-type VI monoterpenes detected in higher proportions in floral
tomato glandular trichomes secrete acyl sugars and terpe- stalked trichomes compared to vegetative and male sessile
nes into a small intercellular cavity between four secretory trichomes. Microcapillary sampling of individual sessile tri-
cells (Bergau et al., 2015). Therefore, with respect to the chomes was previously performed (Malingre et al., 1975),
proliferated secretory cells on top of the stalk and their however the authors did not report terpene identity or
ability to store large amounts of volatile compounds in a compare with stalked trichomes. Most cannabis plants
subcuticular cavity, the cannabis stalked glandular tri- sampled have different amounts of the set of terpene com-
chomes are atypical of capitate glandular trichomes in pounds detected in this study (i.e. myrcene, a-pinene,
other species. (E)-b-ocimene, terpinolene, limonene, b-caryophyllene, and
Our quantitative analysis and modeling show that a-humulene) (Mediavilla and Setinemann, 1997; Meier and
stalked glandular trichomes arise from sessile-like prema- Mediavilla, 1998; Hillig, 2004; Potter, 2009; Aizpurua-Olai-
ture stalked glandular trichomes on floral tissues, and this zola et al., 2016; Booth et al., 2017). Our sampling of fresh
relationship may be predicted by the number of secretory material resulted in a higher monoterpene to sesquiter-
disc cells upon cavity formation. These findings reveal that pene ratio than was previously reported in samples that
cannabis stalked glandular trichomes represent a terminal were dried for hours, weeks, or months (Ross and ElSohly,
state of differentiation for floral sessile glandular tri- 1996; Hillig, 2004; Aizpurua-Olaizola et al., 2016). Using
chomes, as opposed to the drastically different morphol- fresh material, in which the more volatile monoterpenes
ogy and developmental trajectories of capitate and peltate are largely retained, gives a more accurate picture of the
glandular trichomes in other species. The similar results compounds produced in the glandular trichomes of canna-
obtained from two-photon imaging and analysis of disc bis plants. The monoterpene-rich stalked glandular tri-
cell numbers among the hemp-type and medicinal-type chomes dominate the mature cannabis flower, so
varieties of cannabis confirm the relevance of our observa- understanding the molecular basis of variations in
tions in diverse cannabis varieties. Recently, it was shown monoterpene profiles of this trichome type is important for
that cannabis varieties grown for medical consumption understanding their impact on the sensory properties of
have larger gland heads on flowers when compared to cannabis products.
cannabis strains grown for industrial fiber production The transcriptomic and chemical analyses presented
(Small and Naraine, 2016), indicating selective pressure for here show that the five most highly expressed terpene syn-
larger glandular trichomes capable of producing greater thases from stalked glandular trichomes have monoter-
amounts of cannabinoids and other secretory products. pene synthase activities that are consistent with the
The larger heads in medicinal-type varieties may be the monoterpenes detected by microcapillary sampling of indi-
outcome of greater disc cell numbers in the gland head. vidual stalked trichomes and the whole calyx extracts. In a
Previous studies of cannabis specialized metabolites previous study of monoterpene synthases from ‘Finola’
have characterized the cannabinoid and terpene profiles of (Booth et al., 2017), the terpene synthase capable of pro-
many cannabis strains, though the trichome-specific contri- ducing terpinolene was not detected. In this study, the
butions to the overall plant’s chemotype were not most highly expressed terpene synthase in the transcrip-
resolved. Stalked glandular trichomes are abundant on tomes of the isolated glandular trichomes was character-
mature flowers (Turner et al., 1977; Potter, 2009; Small and ized as the missing terpinolene synthase, completing the
Naraine, 2016), and they store cannabinoids in higher set of terpene synthases accounting for the major
abundance than sessile trichomes (Turner et al., 1978). A monoterpenes in the model cannabis strain ‘Finola’.
study using laser capture microdissection sampling of dif- The most highly expressed genes in isolated cannabis
ferent floral trichomes has been performed (Happyana trichomes demonstrated not only a strong commitment to
et al., 2013), however, non-glandular hairs were incorrectly cannabinoid and terpene biosynthesis, but also high
classified as ‘capitate-sessile trichomes’, and therefore did expression of genes related to primary metabolism, such
not compare the sessile glandular trichomes as defined by as glycolysis, TCA cycle, and ATP synthesis (Figure 6). As
others (Hammond and Mahlberg, 1973; Potter, 2009) to demonstrated in a multi-omics study of tomato glandular
stalked trichomes. In our own microcapillary sampling, trichomes, differentially expressed genes for these path-
sessile and stalked trichomes were both dominated by ways are consistent with the high demands of metabolic
CBDA, as expected for a hemp cannabis variety, with mar- productivity in glandular trichomes (Balcke et al., 2017).
ginal variations in other cannabinoids. The cannabinoid While transcript abundance alone is insufficient to

quantitatively determine flux of central carbon metabolites watered with Peter’s Excelâ (15-5-15). To induce flower develop-
through metabolic pathways (Xie et al., 2008; Schwender ment, the light cycle was switched to 12/12, and plants were
watered with Maxibloom (5-15-14).
et al., 2014), the highly expressed central carbon and spe-
cialized metabolism BIN categories observed in cannabis Two-photon microscopy
glandular trichomes, in addition to the abundance of spe-
Fresh plant material was harvested from flowering ‘Finola’, ‘Purple
cialized metabolites stored by these structures, indicate a
Kush’ and ‘Hindu Kush’, and with the use of an Olympus SZX10
need for an abundance of carbon precursors and reducing stereomicroscope, was dissected into vegetative leaves (1 mm2
power. The absolute flux, identity, and source of carbon squares) and flowers (whole immature calyces- <4 mm length, no
precursors required for metabolite production in cannabis observable stalked trichomes; whole mature calyces- >4 mm
will require the types of detailed metabolomic analyses length, >70% stalked trichomes). Dissected samples were affixed
to double-sided scotch tape on a 25 9 75 mm microscope slide
and flux modeling that have been done in tomato (Balcke
(Fisher) and mounted in distilled water prior to imaging on an
et al., 2017) and peppermint (Johnson et al., 2017) glandu- Olympus Multi-photon Laser Scanning Microscope FV1000 MPE
lar trichomes. using an Olympus XLPLN 25X WMP water-dipping objective.
This study moves our knowledge of the properties of can- For spectral analysis (lambda scans) of live trichomes, brief exci-
nabis trichomes into the molecular age from the limited tation at 720 nm with 12% laser power and 800 HV on the photo-
understanding based on the pioneering descriptive work of multiplier tube detector was used. To minimize damage to the
storage cavity, 4 lsec pixel1 dwell time and 256 9 256 pixel den-
the trichomes’ exteriors (Hammond and Mahlberg, 1973, sity was applied and cavities that burst were excluded. Emission
1977; Potter, 2009). Our findings support the hypothesis that was collected from 400 to 600 nm with a 10 nm bandwidth. Stor-
the stalked glandular trichomes that densely cover the eco- age cavity fluorescence was captured by tracing a region of interest
nomically important cannabis flowers develop from transi- (ROI) on the intensity projection (n = 12 cavities per trichome type).
tory sessile-like precursors during floral development, and ROI emission was plotted using Olympus FV10- ASW v3.01.
All subsequent 2P imaging was acquired with 720 nm excita-
are anatomically and biochemically distinct from sessile tri-
tion, and emission was captured using a BFP/GFP/RFP/DsRed filter
chome types on anthers and vegetative leaves. The spectro- cube in two channels, RXD1 (420–460 nm, false-coloured teal) and
scopic and morphological features of ‘Finola’ glandular RXD2 (495–540 nm, false-coloured red). Images were processed
trichomes were also observed in medicinal high-THCA can- and analyzed using ImageJ/FIJI (Schneider et al., 2012). Images
nabis varieties, suggesting the conclusions drawn from the are presented as maximum intensity Z-projections.
hemp-type ‘Finola’ variety may also be applied to other vari- Largest droplet diameter measurements were performed by
scanning through the Z-stack of images containing the extracellu-
eties of cannabis. Uncovering the unique properties of
lar cavity of each trichome, identifying the largest droplet and
these economically and biotechnologically important struc- measuring the greatest edge-to-edge length within the droplet.
tures, and the genes that are highly expressed within them, For cavity fluorescence ratio measurements, maximum intensity
provides new opportunities for molecular breeding, tar- Z-projections of images containing only cavity fluorescence were
geted engineering, informed harvest timing and optimized produced and measured using FIJI. Fluorescence emission from
the trichome cavity was quantified by comparing emission inten-
processing of this important plant.
sity in the cavity to a ROI containing no plant tissue within the
same image. Corrected total cavity fluorescence (CTCF) of each
EXPERIMENTAL PROCEDURES emission channel was determined by the following equation:
CTCF = Area Integrated Density of cavity – (Area of cavity 9
Plant growth mean pixel fluorescence of background ROI). Cavity fluorescence
ratio was determined by dividing the CTCF of the teal channel by
Female pistillate Cannabis sativa L. plants of the auto-flowering
the CTCF of the red channel.
hemp-type variety ‘Finola’ were soil-grown from seed at Anandia
Laboratories in Vancouver, British Columbia, in a Health Canada- Statistical analysis of the fluorescence and droplet diameter
permitted research laboratory. Seeds were planted into a soil mix- measurements were performed using the Statistical Package for
ture containing 1.5 tablespoons of FlorikoteTM (15-5-15) per 1 L the Social Sciences (SPSS v.25.0). A KruskalWallis test revealed
scoop of soil (Sungro, Sunshine Mix #4). Plants were grown under that distributions of both fluorescence ratio measurements and
T5 linear fluorescent lamps (Plusrite, FL54T5/865/HO), using an 18 droplet diameters were not similar for all trichome groups, as
h/6 h cycle (hours of light/hours of dark). Plants were watered with assessed by visual inspection of the boxplots (Figure S1).
tap water for the first 2 weeks after planting and subsequently Disc cell counting
watered with Peter’s Excelâ (15-5-15) at a concentration of
0.05 tbsp L1 (tablespoons per litre) for the duration of vegetative From three plants of the ‘Finola’ and ‘Purple Kush’ varieties of
growth. Sterile bamboo stakes were used to support plant weight cannabis, at least three biological replicates of vegetative leaves
to best maintain an upright position when necessary. Biological (1 mm2 dissections) and whole calyces (3–6 mm) were dissected
controls were applied to the plants on a weekly basis. Female Can- and chemically fixed in 4% paraformaldehyde in PME buffer
nabis sativa L. plants of the marijuana-type varieties ‘Purple Kush’ (25 mM PIPES, 5 mM MgSO4, 5 mM EGTA) overnight at 4°C. After
and ‘Hindu Kush’ were grown via clonal propagation. Clones were three washes in Tris-buffered saline/Tween (TBST), the tissues
rooted in rockwool, then transferred directly into soil in a growth were transferred to a mixture of chloral hydrate/glycerol/water
chamber (BC Northern Lights) under LED lights (BC Northern (8:1:2) and incubated for at least 3 days at room temperature on a
lights, 3000K 80 CRI spectrum). The plants were subjected to vege- rotary shaker. Length of clearing required was dependent on size
tative growth for 2–3 weeks using an 18 h/6 h light cycle and of tissue, and each sample was observed daily until the epidermis

was translucent. The tissues were transferred to 0.2–1 lg ml1 Toronto, Canada) equipped with a UV-light attachment (NIGHT-
DAPI (Sigma, Oakville, Canada) in distilled water, wrapped in tin- SEA Stereo Microscope Fluorescence Adapter, Electron Micro-
foil, and incubated at room temperature for at least 10 min. Sam- scopy Sciences). Extracts were collected by breaking resin-filled
ples were rinsed for 24 h in distilled water at room temperature, microcapillaries directly into GC vials (Supelco headspace vials
followed by sample preparation and image acquisition as with polytetrafluoroethylene (PTFE) silicone caps, Sigma, Oakville,
described in the live-cell 2-P section above. Images were analyzed Canada kept on dry ice during dissections, then stored at 80°C.
using FIJI, where cells were counted using identifiable nuclei. Sta-
tistical analysis was performed using SPSS v.25.0. Solid-phase microextraction (SPME) GC-MS

Cryo-SEM analysis Following UV-light assisted microextraction of trichome metabo-

lites, terpene content was analyzed by the solid-phase microextrac-
After bract removal, dissected female cannabis flowers were tion technique on an Agilent 7890A/5975C Inert XL MSD Triple Axis
attached to a universal freeze fracture specimen holder (Leica gas chromatograph mass spectrometer equipped with an Agilent
Microsystems, Concord, ON, Canada) with Tissue-Tek (VWR GC Sampler 80 autosampler. Supelco SPME 50/30 lm DVB/CAR/
25608-930, Mississauga, ON, Canada) and frozen under high vac- PDMS fiber and Agilent DB-WAX Column (30 m, 0.25 mm internal
uum in a cryo preparation unit (Emitech K1250, Quorum Tech- diameter, 0.25 lm film thickness) were used. The SPME GC-MS
nologies, Puslinch, ON, Canada) Using the Leica EM VCT100 cryo cycle included a 300 sec pre-incubation, 40°C incubation on SPME
transfer system, samples were shuttled from the preparation unit GC agitator unit, 600 sec extraction (fiber exposure), and 300 sec
to a cryo-scanning electron microscope (Hitachi S-4700 Field Emis- desorption, with an oven program of 35°C for 4 min, then increase
sion, Krefeld, Germany), pre-cooled to 140°C. Sublimation at 4° per min to 110°C, then increase 3° per min to 150°C, then
105°C for 20 min was applied to remove surface ice. Images increase 25° per min to 230°C and hold for 3 min. The helium flow
were captured with 1.0 kV and 15–20 mm working distance. In rate was 0.9 ml min1 and injection was split with a ratio of 4:1.
total, nine flowers were examined, prepared in three independent The MS acquisition was performed in electron ionization mode
experiments. with a mass range from 33.0 to 400.0. Selected ion monitoring was
performed simultaneously for masses 93, 121, 136, 161, 189 and
Sample preparation for stitched SEM 204 with a dwell of 35. Data processing was performed with the
Enhanced Chem Station and mass spectra were matched against
Individual cannabis flowers were fixed in 0.025 M PIPES buffer
NIST/WILEY library spectra (W9N08).
with 2% v/v glutaraldehyde at pH 7.0 overnight at 4°C. Samples
were rinsed in buffer twice then dehydrated with an ascending Microcapillary-sampled trichome cannabinoid analysis via
grade ethanol series of 30, 50, 70, 95, and 100% for 20 min each, ultraperformance liquid chromatography tandem mass
with two additional 100% ethanol exchanges. This was followed
by critical-point drying with solvent-substituted liquid CO2 spectrometry (UPLC)-MS/MS
(Autosamdri 815B, Tousimis, Rockville, MD, USA), and sample Following SPME GC-MS, 50 ll of solvent (80 ACN:20 H2O) was
mounting on aluminum specimen stubs. Samples were rotary added to each GC vial and vortexed vigorously to dissolve
coated with 5–12 nm of platinum (Cressington 208C high resolu-
metabolites associated with microcapillary fragments. The solu-
tion sputter coater) and imaged in variable pressure mode with a
tion was separated from glass fragments by careful pipetting and
Zeiss EVO or Sigma SEM at 10 kV and 2.5 nA at a working dis- placed in a high recovery silanized glass autosampler vial (Cana-
tance between 20 and 30 mm. Mosaic secondary electron (SE) dian Life Sciences). The UPLC-MS/MS apparatus (Waters Acquity
images of entire calyces were collected at 500 nm per pixel and
H-Class with Xevo TQD) was equipped with an APCI probe in posi-
stitched using Zeiss Atlas 5 software. tive detection mode and a BEH C8 column (length 2.1 9 100 mm,
Transmission electron microscopy i.d. 1.7 lm). Sample separation was performed with a mobile
phase A: 0.1% formic acid (FA) in liquid chromatography mass
Samples were fixed and dehydrated as described for conventional spectrometry (LC-MS) grade H2O and mobile phase B: 0.1% FA in
SEM. After ethanol dehydration, samples were embedded with an LCMS grade acetonitrile, a flow rate of 0.6 ml min1, and injection
ascending grade Spurr’s resin series of 10, 20, 40, 60, 80, 95 and volume of 5 ll, autosampled at 15°C. The program was as follows:
100% for at least 3 h each. Two additional 100% resin exchanges 0 min: 48% B; 0–5 min: 48–83% B; 5–5.01 min: 83–95.2% B; 5.01–
were performed over 24 h, then samples were embedded in fresh 6 min 95.2% B; 6–6.01 min 95.2–47.7% B. A cannabinoid standards
100% Spurr’s resin in Beem capsules (Ted Pella, Redding, CA, mixture (see Table S5, Cayman Chemical Company, Ann Arbor,
USA) and polymerized at 60°C for 48 h. Here, 70 nm gold sections MI, USA) at concentrations ranging from 100 to 100 000 ng ml1
were cut using a Leica Ultramicrotome UCT, mounted on copper was used to generate a calibration curve, and the mixture at con-
grids coated with 0.3% formvar, and post-stained with 2% uranyl centrations of 1000, 5000 and 25 000 ng ml1 was run between
acetate and lead citrate for 20 and 10 min, respectively. Imaging sample sets for quality control. Data processing was performed
of sections was performed using a Hitachi 7600 transmission elec- with MassLynx v4.1 and comparison with standards’ retention
tron microscope operating at 80 kV. times and mass spectra.

UV-light assisted microextraction Trichome type-specific enrichment by filtration and

density gradient centrifugation
Using microcapillaries (Microcaps; Drummond, Broomall, PA,
USA) pulled to a fine point with a micropipette puller (Inject-Matic, Using previously developed trichome isolation procedures (Ger-
Geneva, Switzerland) and a micromanipulator (MK1; Singer Instru- shenzon et al., 1992; Sallets et al., 2014) as a guide, 18–20 g of
ments, Somerset, UK), metabolites were extracted specifically cannabis floral clusters (bud) was freshly removed from a female
from 100 male antherial sessile, 200 female foliar sessile or 100 plant. The majority of flowers on each selected plant had calyces
female floral stalked glandular trichomes. Glandular trichome between the immature (<4 mm length with no observable stalked
identification was aided by a stereomicroscope (SZX10; Olympus, trichomes) and mature (>4 mm length with >70% stalked

trichomes) stages of development. This developmental range was further cleaned by ethanol precipitation by adding DNase/RNase-
selected to capture sessile, stalked and bulbous trichomes simul- free H2O to a total volume of 500 ll, 50 ll of sodium acetate
taneously and within the secretory phase of metabolite produc- (3 M, pH 5.2) and 1 ml ice cold 100% ethanol. Following 1 h
tion. Trichomes were exposed by excising buds with scissors. incubation at 20°C, RNA was pelleted (4°C, 18 000 g, 30 min)
Plant tissues were placed in a pre-chilled BeadBeater (BioSpec and rinsed with ice cold 70% ethanol and pelleted (4°C,
1107900, Bartlesville, OK, USA) with Amberlite XAD4 (1 g g1 18 000 g, 15 min). Each dried RNA pellet was resuspended in
plant material; Sigma) and 4°C isolation buffer (0.6% w/v methyl 30 ll RNase-free H2O. RNA quality was assessed on a 0.8%
cellulose, 1% w/v PVP3 MW 40000, 25 mM HEPES, 200 mM sor- agarose gel and NanoDrop spectrophotometer. For gel analysis,
bitol, 10 mM sucrose, 10 mM KCl, 5 mM MgCl2-6H20, 0.5 mM each sample was combined with formamide (see Masek et al.,
KH2PO4) to full volume. The sealed chamber was submerged in 2005) and denatured at 65°C, 5 min prior to loading. RNA with a
ice for 30 min. All subsequent steps were performed at 4°C. Plant NanoDrop OD260/280 of 1.8–2 and OD260/230 of 1.8 or greater were
materials were blended with three short pulses separated by deemed acceptable.
30 sec cooling intervals and filtered through a 250-lm nylon Sequencing libraries were prepared from 500 ng of total RNA
mesh. The filtration residue was rinsed with cold wash buffer using the TruSeq Stranded mRNA Sample Preparation kit (Illu-
(consisting of all isolation buffer components, except methyl cellu- mina, San Diego, CA, USA) by the Sequencing and Bioinformatics
lose and PVP3) to obtain 1 L of filtrate. Second and third filtration Consortium (Dept. of Pharmaceutical Sciences, UBC Vancouver).
steps were performed with pre-wetted 102 and 54 lm mesh nylon Libraries were multiplexed and sequenced on a HiSeq 2500 in
mesh, respectively, leaving sessile and stalked trichome heads on rapid run mode, generating paired-end 100 bp reads. Raw FASTQ
the filter surface. Trichomes were gently collected in round-bot- reads were trimmed and quality filtered with trimmomatic v0.35m
tom tubes (Fisher Scientific 14-959-10B, Ottawa, ON, Canada) with (Bolger et al., 2014) using default settings except the following;
a 10 ml transfer pipette and wash buffer. To obtain bulbous tri- LEADING:20, TRAILING:20, SLIDINGWINDOW:4:20, AVGQUAL:20,
chomes from the filtrate, three additional filtrations through and MINLEN:40. Alignment of filtered reads against the Cannabis
54 lm nylon mesh were performed, followed by a final filtration sativa reference transcriptomes, namely canSat3 and finola1 (van
through 25 lm mesh (Bubble Bag OGS1). The bulbous trichome Bakel et al., 2011), were performed using the bowtie2 v2.3.0 pro-
fraction was obtained from the final filter surface. gram using default settings (Langmead and Salzberg, 2012). Read
All culture tubes containing trichome fractions were balanced summarization was performed with htseq-count v0.7.0 software
by weight and pelleted in a swing bucket centrifuge (Eppendorf (Anders et al., 2015) with default parameters using the reference
5810R; 420 rpm, 4 min at 4°C). The pellets were gently resus- transcriptome annotation of canSat3 and finola1. Differential
pended in 6 ml (for sessile and stalked-enriched fractions) or expression analysis was performed using the DESeq2 program
2 ml (for bulbous-enriched fractions) of wash buffer. To separate (Love et al., 2014). An FDR <0.05 defines differentially expressed
sessile and stalked trichome heads (consisting of a secretory transcripts between the tissue contrasts. Transcript abundance
disc with intact storage cavity), trichomes were layered onto a were expressed as fragments per kilobase of transcript per million
100/35/15 density gradient consisting of 3 ml of 100% Percoll mapped reads (FPKM) using DESeq2 and the length of predicted
(Sigma-Aldrich P1644), 4 ml of 35%, and 4 ml of 15% Percoll in PK and FN transcripts. Enrichment of MapMan BIN categories
wash buffer. Multiple density gradients with trichomes were pre- (Thimm et al., 2004) of differentially expressed genes were evalu-
pared, balanced and spun at 300 rpm for 15 min, 4°C. After cen- ated for enrichment using Fisher’s exact test adjusted with FDR
trifugation, a sessile-enriched upper band and stalked-enriched for multiple hypothesis correction according to previously estab-
lower band was visible. Each band was collected in separate lished workflows (Wong et al., 2017) using MapMan BIN-anno-
round-bottom tubes, resuspended in 14 ml wash buffer, pelleted tated PK and FN transcriptomes. MapMan BIN categories were
(420 rpm, 4 min, 4°C) and gently resuspended in 5 ml of wash considered significantly enriched in respective comparisons at a
buffer to remove Percoll. FDR < 0.05.
To harvest enriched trichome fractions, 0.5 ml aliquots of each To enable robust comparisons between our trichome-enriched
trichome type were collected in 2 ml microcentrifuge tubes, leav- transcriptomes and the ones obtained from van Bakel et al. (2011)
ing 1 ml for trichome type scoring. Each tube was diluted with (i.e. root, shoot, stem, and flowers), the same RNA-seq processing
1 ml wash buffer, spun at full speed in a chilled microcentrifuge, workflow was performed and summarized as a whole. RNA-seq
the supernatant removed and the pellet was frozen in liquid nitro- gene co-expression networks were constructed using log2-trans-
gen. All pellets were stored at 80°C. formed FPKM values using the Pearson’s correlation coefficient as
To assess trichome type separation efficacy, samples from each the metric of similarity. A ‘guide’ gene co-expression analysis was
gradient fraction were placed on a slide beneath a coverslip performed on selected candidate(s) and their top 300 co-ex-
(22 9 40 mm) for visual scoring with a Leica DMR microscope. pressed genes (ranked by descending PCC value) were analyzed
The presence of each trichome type with an intact storage cavity further.
was recorded, along with the number of trichome discs (lacking
intact cavities). Scoring was performed for a single pass along the Atlas analysis
length of a coverslip using a 920 objective and UV filter to aid tri-
Image analysis was performed using the Atlas browser export
chome type identification. Under brightfield and UV light, sessile
software. The length of each calyx was measured using the point-
trichome heads appeared granular and fluoresced green, respec-
to-point measurement tool along the medial axis of the calyx from
tively. In contrast, stalked trichome heads appeared homogenous
base to tip. Glandular trichomes were counted within 500–
and fluoresced blue.
1000 lm2 areas of the mid-section of each calyx, with each area
RNA extraction, sequencing and transcriptome analysis containing a minimum of 25 trichomes. Trichomes were anno-
tated as a stalked glandular trichome if the presence of a stalk
Using trichomes separated by type, RNA extractions were per- could be seen at high magnification. Sessile glandular trichomes
formed with a RNeasy Plant Mini Kit (Qiagen, Germantown, MD, were annotated based on the absence of a stalk, and a minimum
USA), including the DNase digestion. Each RNA sample was gland diameter of 25 lm.

Whole calyces were immersed in 500 ll of pentane with

TPS cloning and characterization
31 g L1 isobutyl benzene as internal standard. Resin was
Total RNA was isolated from ‘Finola’ flowers using the Invitrogen extracted by vigorous vortexing with glass beads followed by
PureLink Plant RNA reagent (www.thermofisher.com). RNA quality shaking for 4 h at room temperature. Tissue was dried at 60°C for
and concentration was measured using the Bioanalyzer 2100 RNA 16 h, then weighed. The same procedure was followed for leaves,
Nanochip assay (www.agilent.ca). Full-length sequences of but using 4 ml of solvent.
FN15171.1 and FN20433.1 were obtained using 50 Rapid Amplifica- Compounds were identified using the same GC-MS equipment
tion of cDNA ends (50 -RACE) (www.clontech.com). 50 Plastid target as above, but with the following temperature program: 50°C for
peptides were predicted using the TargetP algorithm (Emanuels- 3 min, then increased at 10°C per min to 150°C, then increased at
son et al., 2000). Sequences were truncated to the RRX8W motif 15°C per min to 320°C, and held for 5 min. Identifications were
(Bohlmann et al., 1998) and cloned into the pASK IBA37 vector made by comparison with authentic standards and NIST/WILEY
(www.iba-lifesciences.org) with a 50 6X-HIS tag. library spectra. Monoterpenes were quantified using standard
Vectors were transformed into E. coli strain BL21DE3 for curves of a-pinene, myrcene, limonene and linalool. Sesquiterpe-
heterologous protein expression. Bacterialcultures were grown in nes were quantified using b-caryophyllene, (E)-b-farnesene, and
50 ml Luria Broth containing ampicillin (100 mg ml1). Cultures bisabolol standard curves. No standards were available for
were grown at 37°C at 200 rpm until they reached OD600 0.6, then cannabinoids, so THC and CBD were identified by retention index
at 18°C for another 2 h. Protein production was then induced and quantified by relative peak area.
using 200 lg L1 anhydrous tetracycline, and the cultures were
shaken for a further 18 h before harvesting by centrifugation. Whole-organ cannabinoid analysis via UPLC-MS/MS
Recombinant protein was purified using the GE Healthcare His Following GC-MS, 10 ll of pentane extract was evaporated to dry-
SpinTrap kit (www.gelifesciences.com) according to the manufac- ness under a thin stream of nitrogen and resuspended in 1 ml
turer’s instructions. Binding buffer for purification was 20 mM 2- acetonitrile (CAN). The UPLC-MS/MS apparatus (Waters Acquity
[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES) (pH H-Class with QDa) was run in electrospray ionization (ESI) detec-
7.5), 500 mM NaCl, 25 mM imidazole, and 5% glycerol. Cells were tion mode and a Raptor ARC-C18 column (length 2.1 9 150 mm,
lysed in binding buffer supplemented with Roche complete pro- i.d. 2.7 lm). Sample separation was performed with a mobile
tease inhibitor tablets (lifescience.roche.com) and 0.1 mg ml1 phase A: 0.1% FA in LCMS grade H2O and mobile phase B: 0.1%
lysozyme. Elution buffer was 20 mM HEPES (pH 7.5), 500 mM NaCl, FA in LCMS grade ACN, a flow rate of 0.6 ml min1, and injection
350 mM imidazole, and 5% glycerol. Purified protein was desalted volume 5 ll, autosampled at 15°C. The program was as follows:
through Sephadex into TPS assay buffer. 0 min: 72% B; 1–3 min: 72–84.5% B; 3–3.6 min: 84.6–100% B; 3.8–
In vitro assays were performed at 500 ll volume by incubating 5 min 100–72% B. A cannabinoid standards mixture (Cerilliant, see
purified protein with isoprenoid diphosphate substrates as previ- Table S6) at concentrations ranging from 40 to 20 000 ng ml1
ously described (O’Maille et al., 2004). TPS assay buffer was was used to generate a calibration curve, and the mixture at con-
25 mM HEPES (pH 7.5), 100 mM KCl, 10 mM MgCl2, 5% glycerol, centrations of 1000, 5000 and 25 000 ng ml1 was run between
and 5 mM DTT. GPP was dissolved in 50% methanol and used at sample sets for quality control. Standards (Syst. Suit inje. 1–5 at
32 lM. Assays were overlaid with 500 ll pentane containing 320 ng ml1) were analyzed. Data processing was performed with
31 g L1 isobutyl benzene as internal standard. MassLynx v4.1 and comparison with standards’ retention times
Product identification was performed using an Agilent 7890A and mass spectra.
gas chromatographer with a 7683B series autosampler and a
7000A TripleQuad mass spectrometer. Ionization was at 70 eV ACCESSION NUMBERS
electrospray with a flow rate of 1 ml min1 of He. The column
was an Agilent VP-5MS (30 m, 250 lm internal diameter, 0.25 lm All raw sequencing data were deposited in the NCBI
film). The temperature program was: 50°C for 1 min, then Sequence Read Archive (http://www.ncbi.nlm.nih.gov/sra)
increased at 10°C per min to 280°C, then held for 5 min. Injection under the BioProject and SRA accession of PRJNA483805
was 1 ll pulsed splitless. Products were identified by comparison and SRP155904, respectively. Sequences for TPS37 and
with authentic standards.
TPS38 were deposited into NCBI GenBank (https://www.
Whole-organ metabolite analysis ncbi.nlm.nih.gov/genbank/) with the accession numbers
MK614216 and MK614217 respectively.
Metabolite analysis was performed on whole organs from 5
‘Finola’ plants harvested at week 7 post-seed germination (imma-
ACKNOWLEDGEMENTS
ture calyces) and week 13 post-seed germination (mature calyces
and mature vegetative leaves). Three floral clusters were har- Funding for this project came from the Canadian Natural Sciences
vested for each time point, one from an axillary bud of a lateral and Engineering Research Council Discovery Grants to ALS, SDC,
branch from the first node (oldest flowers), one from an axillary and JB; and a Mitacs Elevate Post-doctoral Fellowship, in partner-
bud near the shoot apex, and one from the tip of the main cola ship with Anandia Laboratories, to TDQ. The authors thank Lina
(youngest flowers). Three replicate calyces were dissected from Madilao for GC method development, John Coleman for UPLC,
each cluster and used for metabolite analysis. At least three vege- Sunita Sinha at the Sequencing and Bioinformatics Consortium
tative leaves were harvested from the shoot base at the oldest (Department of Pharmaceutical Sciences, UBC Vancouver) and
branching nodes. Calyces were immature if there were no con- Leo Law and Shisen Wang for technical assistance. The authors
spicuous stalked trichomes and the calyx was less than 4 mm in would like to thank the many Anandia employees involved in
length, while mature calyces contained >70% stalked trichomes plant cultivation and maintenance throughout the duration of the
and greater than 4 mm in length, as assessed on an Olympus project. The authors thank Elizabeth Samuels, Dr. Raju Datla, and
SZX10 stereomicroscope. Vegetative leaves were devoid of Tonya Severson for constructive comments on the manuscript.
stalked glandular trichomes. The laboratory of Professor Nelly Pante kindly provided the

equipment for microcapillary sampling. The technical assistance Data S3. Complete list of transcripts in stalked and pre-stalked tri-
of the UBC Bioimaging Facility is gratefully acknowledged. chomes that are differentially expressed compared to bulbous tri-
chomes.
CONFLICT OF INTEREST Data S4. Top 300 most highly co-expressed genes with CBDAS.
Video S1. Example of ATLAS multi-scale SEM imaging of canna-
Jonathan Page is the CSO of Aurora Cannabis Inc., a can-
bis calyx.
nabis license producer and biotechnology company. Kim
Rensing is an employee at Fibics, Inc., the software com- REFERENCES
pany producing the Atlas software used in this study. Aizpurua-Olaizola, O., Soydaner, U., Ozt € u € rk, E., Schibano, D., Simsir, Y.,
Navarro, P., Etxebarria, N. and Usobiaga, A. (2016) Evolution of the
AUTHOR CONTRIBUTIONS cannabinoid and terpene content during the growth of Cannabis sativa
plants from different chemotypes. J. Nat. Prod. 79, 324–331.
SJL, TDQ, JEP and ALS conceived and designed the exper- Anders, S., Pyl, P.T. and Huber, W. (2015) HTSeq—a Python framework to
iments. Cryo-SEM imaging was performed by TDQ. TEM work with high-throughput sequencing data. Bioinformatics, 31, 166–169.
van Bakel, H., Stout, J.M., Cote, A.G., Tallon, C.M., Sharpe, A.G., Hughes,
was performed by SJL. 2P live-cell imaging and analysis T.R. and Page, J.E. (2011) The draft genome and transcriptome of Canna-
was performed by SJL and TDQ. DAPI imaging and analy- bis sativa. Genome Biol. 12, R102.
sis was performed by SJL and JLY. Conventional SEM was Balcke, G.U., Bennewitz, S., Bergau, N., Athmer, B., Henning, A., Majovsky,
P., Jimenez-Go mez, J.M., Hoehenwarter, W. and Tissier, A. (2017) Mul-
imaged by SJL, TDQ, and KHR and analyzed by SJL. Micro- tiomics of tomato glandular trichomes reveals distinct features of central
capillary sampling was performed and analyzed by TDQ. carbon metabolism supporting high productivity of specialized metabo-
Whole-organ dipping was performed and analyzed by JKB. lites. Plant Cell, 29, 960–983.
Bergau, N., Bennewitz, S., Syrowatka, F., Hause, G. and Tissier, A. (2015)
In vitro TPS-b characterization was performed by JKB. Tri- The development of type VI glandular trichomes in the cultivated tomato
chome isolation and enrichment for RNA-seq was prepared Solanum lycopersicum and a related wild species S. habrochaites. BMC
by TDQ and transcriptome analysis by DCJW. SJL, TDQ Plant Biol. 15, 289–304.
Bergau, N., Santos, A.N., Henning, A., Balcke, G.U. and Tissier, A. (2016)
and ALS wrote the manuscript with major input from JKB, Autofluorescence as a signal to sort developing glandular trichomes by
KHR, DCJW, SDC, JB and JEP. flow cytometry. Front. Plant Sci. 7, 949.
Bohlmann, J., Meyer-Gauen, G. and Croteau, R. (1998) Plant terpenoid syn-
SUPPORTING INFORMATION thases: molecular biology and phylogenetic analysis. Proc. Natl Acad.
Sci. USA, 95, 4126–4133.
Additional Supporting Information may be found in the online ver- Bolger, A.M., Lohse, M. and Usadel, B. (2014) Trimmomatic: a flexible trim-
sion of this article. mer for Illumina sequence data. Bioinformatics, 30, 2114–2120.
Figure S1. Two-photon imaging and disc cell counting of high- Booth, J.K. and Bohlmann, J. (2019) Terpenes in Cannabis sativa- from
THCA medicinal cannabis varieties ‘Purple Kush’ and ‘Hindu plant genome to humans. Plant Sci. 284, 67–72
Booth, J.K., Page, J.E. and Bohlmann, J. (2017) Terpene synthases from
Kush’.
Cannabis sativa. PLoS One, 12, e0173911.
Figure S2. Statistical analyses (distributions and pair-wise compar- Celedon, J.M., Chiang, A., Yuen, M.M.S., Diaz-Chavez, M.L., Madilao, L.L.,
isons) of droplet size and fluorescence ratio measurements. Finnegan, P.M., Barbour, E.L. and Bohlmann, J. (2016) Heartwood speci-
Figure S3. Terpene profiles and chromatograms from whole-organ fic transcriptome and metabolite signatures of tropical sandalwood (San-
solvent extractions. talum album) reveal the final step of (Z)-santalol fragrance biosynthesis.
Plant J. 86, 289–299. https://doi.org/10.1111/tpj.13162.
Figure S4. Terpene profiles and chromatograms from microcapil-
Dayanandan, P. and Kaufman, P.B. (1976) Trichomes of Cannabis sativa
lary-extracted trichome sampling. (Cannabaceae). Am. J. Bot. 63, 578–591.
Figure S5. Summary of highly expressed genes according to their Ebersbach, P., Stehle, F., Kayser, O. and Freier, E. (2018) Chemical fingerprint-
functional categories in cannabis glandular trichomes. ing of single glandular trichomes of Cannabis sativa by coherent anti-
Figure S6. Summary of enriched MAPMAN functional categories Stokes raman scattering (CARS) microscopy. BMC Plant Biol. 18, 275.
of differentially expressed cannabis glandular trichome genes. Emanuelsson, O., Nielsen, H., Brunak, S. and von Heijne, G. (2000) Predict-
ing subcellular localization of proteins based on their N-terminal amino
Figure S7. Maximum likelihood phylogeny of terpene synthase acid sequence. J. Mol. Biol. 300, 1005–1016.
amino acid sequences. Fischedick, J.T., Hazekamp, A., Erkelens, T., Choi, Y.H. and Verpoorte, R.
Table S1. Summary of RNA-seq analysis metrics. (2010) Metabolic fingerprinting of Cannabis sativa L., cannabinoids and
terpenoids for chemotaxonomic and drug standardization purposes. Phy-
Table S2. Differentially expressed genes whose expression was
tochemistry, 71, 2058–2073.
higher in stalked or sessile trichomes compared with bulbous tri-
Gagne, S.J., Stout, J.M., Liu, E., Boubakir, Z., Clark, S.M. and Page, J.E.
chomes. (2012) Identification of olivetolic acid cyclase from Cannabis sativa
Table S3. Highly expressed cannabinoid biosynthetic genes in reveals a unique catalytic route to plant polyketides. Proc. Natl Acad. Sci.
cannabis trichomes. USA, 109, 12811–12816.
Table S4. Terpene synthase gene expression and characterization Gershenzon, J., McCaskill, D., Rajaonarivony, J.I.M., Mihaliak, C., Karp, F.
and Croteau, R. (1992) Isolation of secretory cells from plant glandular
in isolated glandular trichomes.
trichomes and their use in biosynthetic studies of monoterpenes and
Table S5. Xevo TQD conditions for the analysis of 10 cannabi- other gland products. Anal. Biochem. 200, 130–138.
noids. Glas, J.J., Bernardus, C.J.S., Alba, J.M., Escobar-Bravo, R., Schuurink, R.C.
Table S6. QDa conditions for the analysis of 10 cannabinoids. and Kant, M.R. (2012) Plant glandular trichomes as targets for breeding
or engineering of resistance to herbivores. Int. J. Mol. Sci. 13, 17077–
Data S1. Complete list of transcripts in stalked, pre-stalked and
17103.
bulbous trichomes.
Hammond, C.T. and Mahlberg, P.G. (1973) Morphology of glandular hairs of
Data S2. Detailed list of enriched MAPMAN functional categories Cannabis sativa from scanning electron microscopy. Am. J. Bot. 60, 524–
of stalked versus bulbous and pre-stalked versus bulbous. 528.

Hammond, C.T. and Mahlberg, P.G. (1977) Morphogenesis of capitate glan- Potter, D.J. (2009). The propagation, characterisation and optimisation of
dular hairs of Cannabis sativa (Cannabaceae). Am. J. Bot. 64, 1023. cannabis as a phytopharmaceutical. Thesis, Kings College London. Avail-
Happyana, N., Agnolet, S., Muntendam, R., Van Dam, A., Schneider, B. and able from: https://ia801604.us.archive.org/25/items/CANNABISSATIVAA
Kayser, O. (2013) Analysis of cannabinoids in laser-microdissected tri- SAPHYTOPHARMACEUTICAL/THE%20PROPAGATION%2C%20CHARAC
chomes of medicinal Cannabis sativa using LCMS and cryogenic NMR. TERISATION%20AND%20OPTIMISATION%20OF%20CANNABIS%
Phytochemistry, 87, 51–59. 20SATIVA%20AS%20A%20PHYTOPHARMACEUTICAL.pdf (Accessed
Hazekamp, A., Peltenburg, A., Verpoorte, R. and Giroud, C. (2005) Chro- October 25 2018).
matographic and spectroscopic data of cannabinoids from Cannabis Ross, S.A. and ElSohly, M.A. (1996) The volatile oil composition of fresh
sativa L. J. Liq. Chromatogr. Relat. Technol. 28, 2361–2382. and air-dried buds of Cannabis sativa. J. Nat. Prod. 59, 49–51.
Hillig, K.W. (2004) A chemotaxonomic analysis of terpenoid variation in Russo, E.B. (2011) Taming THC: potential cannabis synergy and phyto-
Cannabis. Biochem. Syst. Ecol. 32, 875–891. cannabinoid-terpenoid entourage effects. Br. J. Pharmacol. 163, 1344–
Huang, S.-S., Kirchoff, B.K. and Liao, J.-P. (2008) The capitate and peltate 1364.
glandular trichomes of Lavandula pinnata L. (Lamiaceae): histochemistry, Sallets, A., Beyaert, M., Boutry, M. and Champagne, A. (2014) Compara-
ultrastructure and secretion. J. Torrey Bot. Soc. 135, 155–167. tive proteomics of short and tall glandular trichomes of Nicotiana taba-
Huchelmann, A., Boutry, M. and Hachez, C. (2017) Plant glandular tri- cum reveals differential metabolic activities. J. Proteome Res. 13,
chomes: natural cell factories of high biotechnological interest. Plant 3386–3396.
Physiol. 175, 6–22. Sawler, J., Stout, J.M., Garner, K.M., Hudson, D., Vidmar, J., Butler, L.,
Hwang, J.U., Song, W.Y., Hong, D. et al. (2016) Plant ABC transporters Page, J.E. and Myles, S. (2015) The genetic structure of marijuana and
enable many unique aspects of a terrestrial plant’s lifestyle. Mol. Plant, 9, hemp. PLoS One, 10, e0133292.
338–355. Schilmiller, A.L., Last, R.L. and Pichersky, E. (2008) Harnessing plant tri-
Johnson, S.R., Lange, I., Narayanan, S. and Lange, B.M. (2017) Bioenerget- chome biochemistry for the production of useful compounds. Plant J.
ics of monoterpenoid essential oil biosynthesis in nonphotosynthetic 54, 702–711.
glandular trichomes. Plant Physiol. 175, 681–695. Schilmiller, A.L., Schauvinhold, I., Larson, M., Xu, R., Charbonneau, A.L.,
Kim, E.S. and Mahlberg, P.G. (1991) Secretory cavity development in glan- Schmidt, A., Wilkerson, C., Last, R.L. and Pichersky, E. (2009) Monoterpe-
dular trichomes of Cannabis sativa L. Am. J. Bot. 78, 220–229. nes in the glandular trichomes of tomato are synthesized from a neryl
Kim, E.S. and Mahlberg, P.G. (2003) Secretory vesicle formation in the diphosphate precursor rather than geranyl diphosphate. Proc. Natl Acad.
secretory cavity of glandular trichomes of Cannabis sativa L. (Cannaba- Sci. USA, 106, 10865–10870.
ceae). Mol. Cells, 15, 387–395. Schilmiller, A.L., Charbonneau, A.L. and Last, R.L. (2012) Identification of a
Lange, B.M. (2015a) The evolution of plant secretory structures and emer- BAHD acetyltransferase that produces protective acyl sugars in tomato
gence of terpenoid chemical diversity. Annu. Rev. Plant Biol. 66, 139–159. trichomes. Proc. Natl Acad. Sci. USA, 109, 16377–16382.
Lange, I., Poirier, B.C., Herron, B.K. and Lange, B.M. (2015b) Comprehensive Schneider, C.A., Rasband, W.S. and Eliceiri, K.W. (2012) NIH Image to Ima-
assessment of transcriptional regulation facilitated metabolic engineer- geJ: 25 years of Image Analysis. Nat. Methods, 9, 671.
ing of isoprenoid accumulation in Arabidopsis. Plant Physiol. 169, 1595– Schwender, J., Ko € nig, C., Klapperst€uck, M. et al. (2014) Transcript abun-
1606. dance on its own cannot be used to infer fluxes in central metabolism.
Langmead, B. and Salzberg, S.L. (2012) Fast gapped-read alignment with Front. Plant Sci. 5, 668.
Bowtie 2. Nat. Methods, 9, 357–359. Simmons, A.T. and Gurr, G.M. (2005) Trichomes of Lycopersicon species
Liu, C., Srividya, N., Parrish, A.N., Yue, W., Shan, M., Wu, Q. and Lange, and their hybrids: effects on pests and their natural enemies. Agric. For.
B.M. (2018) Morphology of glandular trichomes of Japanese catnip (Schi- Entomol. 7, 265–276.
zonepeta tenuifolia Briquet) and developmental dynamics of their secre- Small, E. and Naraine, S.G.U. (2016) Size matters: evolution of large drug-
tory activity. Phytochemistry, 150, 23–30. secreting resin glands in elite pharmaceutical strains of Cannabis sativa
Love, M.I., Huber, W. and Anders, S. (2014) Moderated estimation of fold (marijuana). Genet. Resour. Crop Evol. 63, 349–359.
change and dispersion for RNA-seq data with DESeq2. Genome Biol. 15, Stout, J., Boubakir, Z., Ambrose, S.J., Purves, R.W. and Page, J.E. (2012)
550. The hexanoyl-CoA precursor for cannabinoid biosynthesis is formed by
Mahlberg, P.G. and Kim, E.S. (2004) Accumulation of cannabinoid in glan- an acyl-activating enzyme in Cannabis sativa trichomes. Plant J. 71, 353–
dular trichomes of Cannabis (Cannabaceae). J. Ind. Hemp, 9, 15–36. 365.
Malingre, T., Hendriks, H., Batterman, S., Bos, R. and Visser, J. (1975) The Taura, F., Morimoto, S. and Shoyama, Y. (1996) Purification and characteri-
essential oil of Cannabis sativa. Planta Med. 28, 56–61. zation of cannabidiolic-acid synthase from Cannabis sativa L. Biochemi-
Masek, T., Vopalensky, V., Suchomelova, P. and Pospisek, M. (2005) Dena- cal analysis of a novel enzyme that catalyzes the oxidocyclization of
turing RNA electrophoresis in TAE agarose gels. Anal. Biochem. 336, 46– cannabigerolic acid to cannabidiolic acid. J. Biol. Chem. 271, 17411–
50. 17416.
McDowell, E.T., Kapteyn, J., Schmidt, A. et al. (2011) Comparative func- Taura, F., Tanaka, S., Taguchi, C., Fukamizu, T., Tanaka, H., Shoyama, Y.
tional genomic analysis of Solanum glandular trichome types. Plant and Morimoto, S. (2009) Characterization of olivetol synthase, a polyke-
Physiol. 155, 524–539. tide synthase putatively involved in cannabinoid biosynthetic pathway.
Mechoulam, R. (1970) Marihuana chemistry. Science, 160, 1467–1469. FEBS Lett. 583, 2061–2066.
Mediavilla, V. and Setinemann, S. (1997) Essential oil of Cannabis sativa L. Thimm, O., Bla€sing, O., Gibon, Y., Nagel, A., Meyer, S., Kru€ ger, P., Silbig, J.,
strains. J. Int. Hemp Assoc. 4, 80–82. Mu € ller, L.A., Rhee, S.Y. and Stitt, M. (2004) MAPMAN: a user-driven tool
Meier, C. and Mediavilla, V. (1998) Factors influencing the yield and the to display genomics data sets onto diagrams of metabolic pathways and
quality of hemp (Cannabis sativa L.) essential oil. J. Int. Hemp Assoc. 5, other biological processes. Plant J. 37, 914–939.
16–20. Tissier, A. (2012) Glandular trichomes: what comes after expressed
Mudge, E.M., Brown, P.N. and Murch, S.J. (2019) The Terroir of Cannabis: sequence tags? Plant J. 70, 51–68.
terpene metabolomics as a tool to understand Cannabis sativa selec- Turner, J.C., Hemphill, J.K. and Mahlberg, P.G. (1977) Gland distribution
tions. Planta Med. 85, 781–796. and cannabinoid content in clones of Cannabis sativa L. Am. J. Bot. 64,
O’Maille, P.E., Chappell, J. and Noel, J.P. (2004) A single-vial analytical and 687–693.
quantitative gas chromatography-mass spectrometry assay for terpene Turner, J.C., Hemphill, J.K. and Mahlberg, P.G. (1978) Quantitative determi-
synthases. Anal. Biochem. 335, 210–217. nation of cannabinoids in individual glandular trichomes of Cannabis
Pertwee, R.G. (2008) The diverse CB1 and CB2 receptor pharmacology of sativa L. (Cannabaceae). Am. J. Bot. 65, 1103–1106.
three plant cannabinoids: delta9-tetrahydrocannabinol, cannabidiol and Turner, G.W., Gershenzon, J. and Croteau, R. (2000) Development of peltate
delta9-tetrahydrocannabivarin. Br. J. Pharmacol. 153, 199–215. glandular trichomes of peppermint. Plant Physiol. 124, 665–679.
Pertwee, R.G. (ed). (2014) Handbook of Cannabis. Oxford, UK: Oxford Werker, E. (2000) Trichome diversity and development. Adv. Bot. Res. 31, 1–
University Press. 35.

Wong, D.C.J., Gutierrez, R.L., Gambetta, G.A. and Castellarin, S.D. (2017) to multiple levels of metabolic control in sweet basil glandular tri-
Genome-wide analysis of cis-regulatory element structure and discovery chomes. Plant J. 54, 349–361.
of motif-driven gene co-expression networks in grapevine. DNA Res. 24, Zager, J.J., Lange, I., Srividya, N., Smith, A. and Lange, B.M. (2019) Gene
311–326. networks underlying cannabinoid and terpenoid accumulation in Can-
Xie, Z., Kapteyn, J. and Gang, D.R. (2008) A systems biology investigation nabis. Plant Physiol. 180, 1877–1897. https://doi.org/10.1104/pp.18.
of the MEP/terpenoid and shikimate/phenylpropanoid pathways points 01506.

Cannabis Flowers Ph. Eur
100% (1)
Cannabis Flowers Ph. Eur
6 pages
Potter Thesis 2009
100% (1)
Potter Thesis 2009
255 pages
The Propagation, Characterisation and Optimisation of Cannabis Sativa L As A Phytopharmaceutical
No ratings yet
The Propagation, Characterisation and Optimisation of Cannabis Sativa L As A Phytopharmaceutical
255 pages
PPM Cannamimetics White Paper
No ratings yet
PPM Cannamimetics White Paper
33 pages
Jonathan Ott: The Delphic Bee: Bees and Toxic Honeys As Pointers To Psychoactive and Other Medicinal Plants.
No ratings yet
Jonathan Ott: The Delphic Bee: Bees and Toxic Honeys As Pointers To Psychoactive and Other Medicinal Plants.
4 pages
The Plant Journal - 2019 - Livingston - Cannabis Glandular Trichomes Alter Morphology and Metabolite Content During Flower
No ratings yet
The Plant Journal - 2019 - Livingston - Cannabis Glandular Trichomes Alter Morphology and Metabolite Content During Flower
20 pages
5 - Conceptos Basicos Sobre Cannabis Medicinal - Nicole Halçartegaray
No ratings yet
5 - Conceptos Basicos Sobre Cannabis Medicinal - Nicole Halçartegaray
88 pages
Journal Pone 0242633
No ratings yet
Journal Pone 0242633
26 pages
Trichomes of Cannabis Sativa L.
No ratings yet
Trichomes of Cannabis Sativa L.
15 pages
Braich Et Al. 2019, Transcriptome Atlas
No ratings yet
Braich Et Al. 2019, Transcriptome Atlas
12 pages
Trichomes: What Is A Trichome
No ratings yet
Trichomes: What Is A Trichome
7 pages
Cannabis Plant Botanical Aspects
No ratings yet
Cannabis Plant Botanical Aspects
10 pages
Trichomes of Cannabis Sativa (Cannabaceae)
No ratings yet
Trichomes of Cannabis Sativa (Cannabaceae)
14 pages
Plants-12-03646-V2 Plants
No ratings yet
Plants-12-03646-V2 Plants
21 pages
Exogenous Application of Stress Related
No ratings yet
Exogenous Application of Stress Related
13 pages
Cannabis Botany
No ratings yet
Cannabis Botany
10 pages
Gland Distribution and Cannabinoid Conte
No ratings yet
Gland Distribution and Cannabinoid Conte
8 pages
Chemical Ecology of Cannabis
No ratings yet
Chemical Ecology of Cannabis
17 pages
The Morphology Inflorescence Yield and Secondary Metabolite 2024 Industri
No ratings yet
The Morphology Inflorescence Yield and Secondary Metabolite 2024 Industri
11 pages
David Potter
No ratings yet
David Potter
53 pages
Gene Expression Profile Analysis of Tobacco Leaf Trichomes
No ratings yet
Gene Expression Profile Analysis of Tobacco Leaf Trichomes
9 pages
Effects of Ethephon On Terpenoids in Cannabis Sativa L. in Vegetative Stage
No ratings yet
Effects of Ethephon On Terpenoids in Cannabis Sativa L. in Vegetative Stage
10 pages
2024-0170RV MedChem
No ratings yet
2024-0170RV MedChem
25 pages
Indoor Vs Outdoor 2023
No ratings yet
Indoor Vs Outdoor 2023
15 pages
Molecules 27 05868
No ratings yet
Molecules 27 05868
18 pages
Parallel Evolution of Cannabinoid Biosynthesis: Nature Plants
No ratings yet
Parallel Evolution of Cannabinoid Biosynthesis: Nature Plants
31 pages
Cannabis Flower
No ratings yet
Cannabis Flower
4 pages
Cannabis Phenolics and Their Bioactivities
No ratings yet
Cannabis Phenolics and Their Bioactivities
44 pages
FLR 3.1 Sorima
No ratings yet
FLR 3.1 Sorima
3 pages
Aizpurua-Olaizola Et Al 2016 - Evolution of The Cannabinoid and Terpene Content During The
No ratings yet
Aizpurua-Olaizola Et Al 2016 - Evolution of The Cannabinoid and Terpene Content During The
8 pages
Molecules 27 01719 With Cover
No ratings yet
Molecules 27 01719 With Cover
15 pages
BCP 84 2463
No ratings yet
BCP 84 2463
5 pages
Molecules 24 02113 PDF
No ratings yet
Molecules 24 02113 PDF
12 pages
Microbioma Cannabis Revision Biblografica
No ratings yet
Microbioma Cannabis Revision Biblografica
16 pages
The Hexanoyl-Coa Precursor For Cannabinoid Biosynthesis Is Formed by An Acyl-Activating Enzyme in Cannabis Sativa Trichomes
No ratings yet
The Hexanoyl-Coa Precursor For Cannabinoid Biosynthesis Is Formed by An Acyl-Activating Enzyme in Cannabis Sativa Trichomes
13 pages
Plants 11 00140 v2
No ratings yet
Plants 11 00140 v2
16 pages
Ramanetal2017 Cannabismorpho-Anatomy Bookchapter
No ratings yet
Ramanetal2017 Cannabismorpho-Anatomy Bookchapter
18 pages
Plant Science: Judith K. Booth, Jörg Bohlmann T
No ratings yet
Plant Science: Judith K. Booth, Jörg Bohlmann T
6 pages
The Cream of The Crop - Biology, Breeding and Applications of Cannabis Sativa
No ratings yet
The Cream of The Crop - Biology, Breeding and Applications of Cannabis Sativa
46 pages
Cannabis Grow Course 101 - Module 2 - Prepare (Lesson 3)
No ratings yet
Cannabis Grow Course 101 - Module 2 - Prepare (Lesson 3)
5 pages
2017 Quimiotipos Ethan Russo
No ratings yet
2017 Quimiotipos Ethan Russo
9 pages
A New ESI-LC/MS Approach For Comprehensive Metabolic Profiling of Phytocannabinoids in Cannabis
No ratings yet
A New ESI-LC/MS Approach For Comprehensive Metabolic Profiling of Phytocannabinoids in Cannabis
15 pages
Cannabis 101
No ratings yet
Cannabis 101
27 pages
Heterologous Production of Cannabis Sativa-Derived Specialised Metabolites of Medicinal Significance
No ratings yet
Heterologous Production of Cannabis Sativa-Derived Specialised Metabolites of Medicinal Significance
30 pages
Report Pe Drugs Cannabis
No ratings yet
Report Pe Drugs Cannabis
33 pages
Development and Validation of A Fast Gas Chromatography/mass Spectrometry Method For The Determination of Cannabinoids in Cannabis Sativa L
No ratings yet
Development and Validation of A Fast Gas Chromatography/mass Spectrometry Method For The Determination of Cannabinoids in Cannabis Sativa L
31 pages
From Cultivar To Chemovar II
No ratings yet
From Cultivar To Chemovar II
15 pages
Poster Congreso Internacional IACM 2022
No ratings yet
Poster Congreso Internacional IACM 2022
1 page
Research Journal
No ratings yet
Research Journal
6 pages
CBD-THC Relacion Con Quimiotipo
No ratings yet
CBD-THC Relacion Con Quimiotipo
13 pages
Identification of Olivetolic Acid Cyclase
No ratings yet
Identification of Olivetolic Acid Cyclase
15 pages
INDCRO6524
No ratings yet
INDCRO6524
7 pages
American J of Botany - 2004 - Hillig - A Chemotaxonomic Analysis of Cannabinoid Variation in Cannabis Cannabaceae
No ratings yet
American J of Botany - 2004 - Hillig - A Chemotaxonomic Analysis of Cannabinoid Variation in Cannabis Cannabaceae
10 pages
A Chemo Taxonomic Analysis of Cannabinoid
No ratings yet
A Chemo Taxonomic Analysis of Cannabinoid
10 pages
Methods For Quantification of Cannabinoids
No ratings yet
Methods For Quantification of Cannabinoids
10 pages
Russo, 2017 THC, CBD, CBG, CBC, CBN, CBDV, Cbda
No ratings yet
Russo, 2017 THC, CBD, CBG, CBC, CBN, CBDV, Cbda
68 pages
Sample Coa Report11684211
No ratings yet
Sample Coa Report11684211
6 pages
Position Paper-Manila
No ratings yet
Position Paper-Manila
4 pages
Jcmah MS Id 555829
No ratings yet
Jcmah MS Id 555829
5 pages
USVI Marijuana Legalization Bill
No ratings yet
USVI Marijuana Legalization Bill
72 pages
Hallucinogenic Plants A Golden Guide
No ratings yet
Hallucinogenic Plants A Golden Guide
164 pages
Borougerdi Uta 2502D 12618
No ratings yet
Borougerdi Uta 2502D 12618
285 pages
CBD Pachaging Market
No ratings yet
CBD Pachaging Market
2 pages
High Times - December 2017
No ratings yet
High Times - December 2017
152 pages
Terpene Infused Batch2
No ratings yet
Terpene Infused Batch2
3 pages
CBDepot - Eu Product Specification
No ratings yet
CBDepot - Eu Product Specification
16 pages
Industrial Hemp
No ratings yet
Industrial Hemp
37 pages
Russo 2007 History Chem Biodiversity
No ratings yet
Russo 2007 History Chem Biodiversity
36 pages
What Is An Eighth, Quarter, Half, or Ounce of Weed (Visual Guide) Key To Cannabis
No ratings yet
What Is An Eighth, Quarter, Half, or Ounce of Weed (Visual Guide) Key To Cannabis
1 page
Conceptual Review On Vijaya Cannabis Sativa Linn A
No ratings yet
Conceptual Review On Vijaya Cannabis Sativa Linn A
11 pages
GSCBPS 2023 0393
No ratings yet
GSCBPS 2023 0393
6 pages
Blueberry Burst Gummies
No ratings yet
Blueberry Burst Gummies
1 page
DEA Clarification of The Marihuana Extract Rule
No ratings yet
DEA Clarification of The Marihuana Extract Rule
2 pages
Cannabis
No ratings yet
Cannabis
12 pages
CBD Topical Guide - DR
No ratings yet
CBD Topical Guide - DR
5 pages
Arizona
No ratings yet
Arizona
19 pages
Need For Standardization of Cannabis Sativa L. (Hemp, Medical Marijuana and Marijuana As A Drug) in Slovenia
No ratings yet
Need For Standardization of Cannabis Sativa L. (Hemp, Medical Marijuana and Marijuana As A Drug) in Slovenia
18 pages
The Dangers of Marijuana Wax
No ratings yet
The Dangers of Marijuana Wax
2 pages
Examen Stiinte Economice Limba Engleza
No ratings yet
Examen Stiinte Economice Limba Engleza
3 pages
State Law Power What The Maps Portend For 2025 Presentation
No ratings yet
State Law Power What The Maps Portend For 2025 Presentation
42 pages
Pharmacognosy Final Year Note Created by Milantirtha Mete (In Charge)
100% (1)
Pharmacognosy Final Year Note Created by Milantirtha Mete (In Charge)
15 pages
Peak Digital Magazine 2021
No ratings yet
Peak Digital Magazine 2021
23 pages
Edibles List Magazine - The Technology Issue - No. 44
No ratings yet
Edibles List Magazine - The Technology Issue - No. 44
33 pages
Cannabis Sativa-1st Paper
No ratings yet
Cannabis Sativa-1st Paper
10 pages
rws11 q3 Mod3 Explicit and Implicit Claims in Written Text Final
No ratings yet
rws11 q3 Mod3 Explicit and Implicit Claims in Written Text Final
31 pages