Journal Pre-proofs

Tetrahydrocannabinolic acid A (THCA-A) reduces adiposity and prevents met-

abolic disease caused by diet-induced obesity

Belén Palomares, Francisco Ruiz-Pino, Martin Garrido-Rodriguez, M. Eugenia

Prados, Miguel A. Sánchez-Garrido, Inmaculada Velasco, María J. Vazquez,
Xavier Nadal, Carlos Ferreiro-Vera, Rosario Morrugares, Giovanni Appendino,
Marco A Calzado, Manuel Tena-Sempere, Eduardo Muñoz

PII: S0006-2952(19)30392-2
DOI: https://doi.org/10.1016/j.bcp.2019.113693
Reference: BCP 113693

To appear in: Biochemical Pharmacology

Received Date: 29 September 2019

Accepted Date: 5 November 2019

Please cite this article as: B. Palomares, F. Ruiz-Pino, M. Garrido-Rodriguez, M. Eugenia Prados, M.A. Sánchez-
Garrido, I. Velasco, M.J. Vazquez, X. Nadal, C. Ferreiro-Vera, R. Morrugares, G. Appendino, M.A. Calzado, M.
Tena-Sempere, E. Muñoz, Tetrahydrocannabinolic acid A (THCA-A) reduces adiposity and prevents metabolic
disease caused by diet-induced obesity, Biochemical Pharmacology (2019), doi: https://doi.org/10.1016/j.bcp.
2019.113693

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will
undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing
this version to give early visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Elsevier Inc. All rights reserved.

Tetrahydrocannabinolic acid A (THCA-A) reduces adiposity and prevents

metabolic disease caused by diet-induced obesity

Belén Palomares1,2,3, Francisco Ruiz-Pino1,2,3, Martin Garrido-Rodriguez1,4, M. Eugenia

Prados5, Miguel A. Sánchez-Garrido1,2,3, Inmaculada Velasco1,2,3, María J. Vazquez1,2,3,

Xavier Nadal6, Carlos Ferreiro-Vera6, Rosario Morrugares1,2,3, Giovanni Appendino7,

Marco A Calzado1,2,3, Manuel Tena-Sempere1,2,3, *, and Eduardo Muñoz1,2,3,*

*Equal contribution
1Maimonides Biomedical Research Institute of Córdoba (IMIBIC), Córdoba, 14004,

Spain; 2Department of Cellular Biology, Physiology and Immunology, University of

Córdoba, Córdoba; 3University Hospital Reina Sofía, Córdoba, Spain; 4Innovative

Health Group, Madrid, Spain; 5Emerald Health Biotechnology España, Córdoba, Spain;
6Phytoplant Research, Córdoba, Spain; 7Dipartimento di Scienze del Farmaco,

Università del Piemonte Orientale, Novara, Italy.

Corresponding author: Eduardo Muñoz, MD, PhD. Instituto Maimónides de

Investigación Biomédica de Córdoba, Universidad de Córdoba, Avda Menéndez Pidal

s/n. 14004. Córdoba, Spain. Phone: +34 957213766. Email: fi1muble@uco.es

Running title: THCA-A prevents obesity induced by high fat diet

Keywords: Cannabis, Δ9-THCA-A, PPARγ, obesity, inflammation

Category: Full Length Article

Abbreviations: BAT, Brown Adipose Tissue; FRET, Fluorescence Resonance Energy

Transfer; HFD, High Fat Diet; hMSC, Human mesenchymal stem cells; iWAT, inguinal

White Adipose Tissue; LBD, Ligand Binding Domain; MetS, Metabolic syndrome;

1
PPAR, Peroxisome Proliferator-Activated Receptor; RGZ, rosiglitazone; Δ9-THC, Δ9-

tetrahydrocannabinol; UCP-1, uncoupling protein-1.

2
Abstract

Medicinal cannabis has remarkable therapeutic potential, but its clinical use is limited

by the psychotropic activity of Δ9-tetrahydrocannabinol (Δ9-THC). However, the

biological profile of the carboxylated, non-narcotic native precursor of Δ9-THC, the Δ9-

THC acid A (Δ9-THCA-A), remains largely unexplored. Here we present evidence that

Δ9-THCA-A is a partial and selective PPARγ modulator, endowed with lower

adipogenic activity than the full PPARγ agonist rosiglitazone (RGZ) and enhanced

osteoblastogenic effects in hMSC. Docking and in vitro functional assays indicated that

Δ9-THCA-A binds to and activates PPARγ by acting at both the canonical and the

alternative sites of the ligand-binding domain. Transcriptomic signatures in iWAT from

mice treated with Δ9-THCA-A confirmed its mode of action through PPARγ.

Administration of Δ9-THCA-A in a mouse model of HFD-induced obesity significantly

reduced fat mass and body weight gain, markedly ameliorating glucose intolerance and

insulin resistance, and largely preventing liver steatosis, adipogenesis and macrophage

infiltration in fat tissues. Additionally, immunohistochemistry, transcriptomic, and

plasma biomarker analyses showed that treatment with Δ9-THCA-A caused browning

of iWAT and displayed potent anti-inflammatory actions in HFD mice. Our data

validate the potential of Δ9-THCA-A as a low adipogenic PPARγ agonist, capable of

substantially improving the symptoms of obesity-associated metabolic syndrome and

inflammation.

3
1. Introduction

Preparations of Cannabis sativa have been used since the earliest written records of

medical history; the potential efficacy of cannabis extracts and cannabinoids being

reported in multiple medical conditions. Almost 150 cannabinoids have been isolated

from Cannabis [1] with Δ9-tetrahydrocannabinol (Δ9-THC) and cannabidiol being the

best investigated class members. These cannabinoids are produced and stored in the

plant as acidic precursors (cannabinoid acids) [2]. Decarboxylation requires heating [3],

and remarkably Δ9-THCA-A, the acidic precursor of Δ9-THC, is not psychotropic, and

its binding to cannabinoid receptors is highly controversial [4]. We have recently shown

that Δ9-THCA-A, but not its decarboxylation product, Δ9-THC, is a potent activator of

the peroxisome proliferator-activated receptor-γ (PPARγ) endowed with

neuroprotective activity [5]. Given the key metabolic roles of PPARγ, these findings

provided a rationale to investigate the possibility that Δ9-THCA-A based therapies

could be beneficial for the management of metabolic disorders.

Obesity is a major risk factor for the development of multiple diseases [6, 7]

including prominently cardiovascular and metabolic disorders, such as type-2 diabetes

and metabolic syndrome (MetS) [8]. Given the conspicuous lack of effective therapies

for the integral management of obesity, considerable attention is currently given to the

identification of novel pharmacological agents globally targeting its deregulated

substrates and other related pathways, as thermogenesis, to regain energy homeostasis

[9-11].

In fact, growing interest has been paid recently to elucidate the physiological

roles and eventual therapeutic use of adaptive thermogenesis in the control of body

weight and energy homeostasis [12, 13]. Studies in this area have been directed not only

to the brown adipose tissue (BAT), where canonical adaptive thermogenesis occurs in

4
mammals, including humans [14], but also to the capacity of the white adipose tissue

(WAT) to undergo browning, a process whereby adipocyte precursor cells acquire

brown-like features, becoming beige or brite adipocytes [15]. Brown and beige

adipocytes are defined by a large set of mitochondria and high expression of uncoupling

protein-1 (UCP-1), as major molecular switch to enhance heat dissipation at the expense

of lower ATP synthesis at the respiratory chain, which is the hallmark of thermogenic

activity [16].

PPARγ is a nuclear receptor that plays key role in regulating a number of

biological functions including lipid and glucose metabolism [17]. PPARγ is also

involved in the natural and adaptive immune response, as well as in additional

biological activities like the browning of white fat [18]. PPARγ ligands include a wide

array of natural and synthetic molecules, with glitazones, a class of compounds

extensively used in type-2 diabetes, being the archetypal activators. Glitazones, like

rosiglitazone (RGZ), are considered full PPARγ ligand agonists, and this profile is

inextricably associated to both antidiabetic activity and undesirable side effects, like

weight gain, edema, and osteoporosis [19]. However, while the therapeutic potential of

PPARγ agonists remains high, interest has substantially shifted towards partial ligands

that have opened novel research avenues for the identification and/or development of

second-generation PPARγ agonists.

2. Materials and Methods

2.1 Δ9-THCA-A isolation

Δ9-THCA-A was purified at >95% from the proprietary Cannabis variety Moniek

(CPVO/20160114) using a Counter Current Chromatography (CCC) by Phytoplant

Research S.L. (Córdoba, ES). An Agilent liquid chromatography set-up (Model 1260,

5
Pittsburgh, PA, United States) consisting of a binary pump, a vacuum degasser, a

column oven, an autosampler and a diode array detector (DAD) equipped with a 150

mm length, 2.1 mm internal diameter, 2.7 mm pore size Poroshell 120 EC-C18 column

was used for the quality control of the purified cannabinoids. The analysis was

performed using water and acetonitrile both containing ammonium formate 50 mM as

mobile phases. Flow-rate was 0.2 mL/min and the injection volume was 3 µL.

Chromatographic peaks were recorded at 210 nm. All determinations were carried out at

35 ºC. All samples were analyzed in duplicate. The result of Δ9-THCA-A purity 95.42%

and THC impurity 1.32% was calculated as weight (%) versus a commercial standard

from THC Pharm GmbH (Frankfurt, Germany) and Cerilliant (Round Rock, Texas,

USA).

2.2 Cell lines and luciferase assays

HEK-293T and 3T3-L1 cells were obtained from the ATCC (Manassas, VA, USA).

HEK-293T (1 x 105) cells were seeded in 24-well plates and transiently co-transfected

with the indicated constructs containing human LBDs (GAL4-PPARγ-LBD, GAL4-

PPARδ-LBD, GAL4-PPARα-LBD and GAL4-luc) using Roti©-Fect (Carl Roth,

Karlsruhe, Germany). After treatments, the luciferase activities were measured using

Dual-Luciferase Assay (Promega, Madison, WI, USA).

2.3 In vitro adipocyte and osteoblast differentiation and real-time PCR

Human mesenchymal stem cells (hMSCs) derived from bone marrow were seeded in α-

MEM containing 10% FBS, 2 mM Glutamine, 1 ng/ml bFGF, and antibiotics, and

adipocyte (AD) and osteoblast (OM) differentiation was performed as described [20].

Treatment with RGZ (1 μM) and Δ9-THCA-A (1, 5 and 10 μM) in the presence and the

absence of T0070907 (5 μM) started at the same time as the differentiation process.

After 14 days of differentiation, the mRNA was analyzed by qPCR and, after 21 days,

6
adipogenesis and osteoblastogenesis were analyzed by Oil red O and Alizarin red

staining respectively as described previously [20]. The primers used are listed in Table

1.

2.4 Animal studies

Six-week old male C57BL6 mice (from Charles Rivers Laboratories; l’Arbresle,

France) were housed under constant conditions of light and temperature with free access

to food and water. All procedures were reviewed and approved by the Ethics Committee

of the University of Cordoba and carried out in accordance with European Union

Directive 2010/63/EU for the use and care of experimental animals. At 8 weeks of age,

mice were randomly assigned in two groups (N = 20) and fed either high-fat diet

(HFD), D12451 (Research Diets, New Brunswick, NJ, USA; 45%, 20%, and 35%

calories from fat, protein and carbohydrate, respectively) or standard diet (CD), A04

SAFE Diets (Augy, France; 8,4%, 19,3% and 72,4% calories from fat, protein, and

carbohydrate, respectively) for 15 weeks. Of note, A04 SAFE was used as control diet

in our experiments since this is used routinely in our Animal Department and has a

percentage of calorie content from fat sources grossly similar to that of other control

diets (e.g., D12450B diet from Research Diets, with 10% calories from fat), with an

stable nutritional composition in terms of percentage of main nutrients and calorie

sources. Body weight (BW) gain, terminal BW, and daily energy intake were monitored

once weekly during the first 12 weeks, and twice a week during treatment period. The

latter was calculated from mean food ingestion per week using the energy density index

provided by the manufacturer (3.34 kcal/g for CD or 4.73 kcal/g for HFD). In order to

assess the metabolic effects of Δ9-THCA-A, mice were treated daily by intra-peritoneal

injection of this compound (20 mg/Kg dissolved in ethanol/cremophor/saline 1:1:18) for

three weeks, from week 12, in CD and HFD groups (n = 10/group). The amount of fat

7
mass and adiposity in all set of experiments were measured by Quantitative magnetic

resonance (QMR) scans, using the EchoMRI™ 700 analyzer (Houston, TX, USA,

software v.2.0), before initiation of the treatments with Δ9-THCA-A and at the end of

the experimental procedures. Intraperitoneal glucose and insulin tolerance tests, as well

as triglyceride determinations, were conducted as described previously [20].

In another experiment, eight-week old male C57BL6 mice fed with CD were

treated with Δ9-THCA-A (20 mg/Kg i.p.), with or without the selective PPARγ

inhibitor, T0070907 (5 mg/Kg i.p.), for 3 weeks (n = 10/group) [5]. Pair-aged mice

(n=10) treated with vehicle served as controls.

2.5 Immunohistochemistry and protein analysis by Western Blots.

Liver tissue sections (5 µm) were stained with haematoxylin and eosin (H&E) and

evaluated for steatosis according to the Kleiner system [21]. A semi-quantitative score

was assigned to describe the extent of steatosis (0, <5%; 1, 5–33%; 2, 33–66%; and 3,

>66%). IHC analysis of iWAT tissue sections (7 µm) was carried out as described

previously [20]. The antibodies used in this study were F4/80 antibody (1:50; MCA497,

Bio-Rad Laboratories, Hercules, CA, USA) and UCP-1 antibody (1:500; ab10983,

Abcam, Cambridge, UK). UCP-1 expression by western blot was analyzed as described

previously [20]. Differentiated 3T3-L1 cells in adipogenic medium were preincubated

with either Δ9-THCA-A or RGZ and treated with 50ng/ml TNFα for 30min and the

expression of PPARγ Ser272 and total PPARγ analyzed with the antibodies anti-PPARγ

Ser272 (1:200, bs-4888R, Bioss, Woburn, MA, USA) and anti-PPARγ (D1:1.000, 2435,

Cell Signaling Technology, Danvers, MA, USA). Protein levels were normalized to β-

actin (1:5000 dilution, A5060, Sigma Aldrich, St. Louis, MO, USA).

2.6 Determination of hormonal, metabolic and inflammatory markers.

8
Leptin, insulin, glucagon, glucagon-like peptide 1 (GLP-1) and adiponectin were

measured in serum using quantitative Bio-Plex Pro™ Mouse Diabetes 8-Plex

immunoassay (#171F7001M; Bio-Rad Laboratories) and Bio-Plex Pro Mouse Diabetes

Adiponectin assay #171F7002M (Bio-Rad Laboratories) according to the

manufacturer’s instructions. For other biomarkers, serum samples from mice were

pooled (n = 6 mice per group) and assayed for cytokine and adipokine expression using

the Proteome Profiler Mouse XL Cytokine Array and the Proteome Profiler Mouse

Adipokine Array (R&D Systems, Minneapolis, MN, USA) according to the

manufacturer’s protocols. Spot density was determined using Quick Spots image

analysis software (R&D Systems).

2.7 RNA-Seq and bioinformatic analysis

For each group, total RNA was extracted from iWAT, prepared, pooled, and run on an

Agilent Bioanalyzer system to confirm quality (RNA integrity number >8).

Transcriptome libraries were then constructed using poly-A selection with the TruSeq

Stranded mRNA Library Prep Kit (Cat. No. RS-122-2101, Illumina, San Diego, CA,

USA). In brief, 300 ng of total RNA from each sample was used to construct a cDNA

library, followed by sequencing on the Illumina HiSeq 2500 with single end 50 bp reads

and ~30 millions of reads per sample. The FASTQ files were pre-processed with

Trimmomatic (v0.36) to remove adapter sequences and aligned to the mm10 assembly

of the mouse genome using HISAT2 (v2.1.0). The counts per gene matrix were

obtained from the alignments with featureCounts (v1.6.1) using the in-built RefSeq

annotation for the mm10 genome assembly. After filtering genes with less than 15 reads

across samples, the raw counts were analyzed with DESeq2 (v1.20.0) to obtain the

regularized log transformed expression matrix and the differential expression analysis

results. We used a threshold of an absolute fold change ≥ 2 and an adjusted P value ≤

9
0.01 to consider a gene as differentially expressed in any comparison. Heatmaps were

generated using the scaled mean of the regularized log expression for each group with

the R package ComplexHeatmap (v1.20.0). The gene set enrichment analysis (GSEA)

[22] and the over-representation analysis were performed using the R package

ClusterProfiler (v3.10.1) [23]. For GSEA, genes were pre-ranked using the log2

transformed fold change. The KEGG pathway database and Gene Ontology (Biological

Process) annotation were used to group genes by biological function. All the P values

were adjusted using the Benjamini and Hochberg correction to control the false

discovery rate (FDR). RNA-seq data have been deposited in the Gene Expression

Omnibus databank (accession no. GSE129573).

2.8 Docking analysis

Ligand docking, and binding properties were calculated by using the AutoDock4 [24]

and the Vina software [25] with the virtual screening tool PyMOL [26]. The receptor

models used were the PDB references 5Y2O57 [27], 4EMA56 [28] and 5LGS58 [29].

Search space for the docking was set around the binding sites described previously [30,

31].

2.9 PPARγ coactivator assay

Time-resolved fluorescence resonance energy transfer (TR-FRET) coactivator assay for

PPARγ was performed by employing LanthaScreen™ kit (ThermoFisher Scientific®,

#A15126) according to the manufacturer’s protocol

2.10 Statistical analysis.

In vitro data are expressed as mean ± SD with a minimal of 3 to 4 independent

experiments. In vivo results are represented as mean ± SEM and the determinations

were conducted with a minimal total number of 6-10 animals per group. Statistical

analyses were performed on data distributed in a normal pattern, using ANOVA

10
followed by the post-hoc Tukey’s test, when relevant. P < 0.05 was taken as the

minimum level of significance. Statistical analysis was performed using GraphPad

Prism® version 6.01.

3. Results

3.1. Δ9-THCA-A is a selective and non-adipogenic PPARγ ligand agonist that

induces iWAT browning through a PPARγ-dependent pathway

We have previously shown that Δ9-THCA-A is a PPARγ agonist at nanomolar

concentrations [5]. Herein, we have studied the selectivity of Δ9-THCA-A on different

PPARs, showing that, when compared to the full ligand agonist rosiglitazone (RGZ),

Δ9-THCA-A is a partial ligand activator for PPARγ, devoid of PPARα and PPARδ

transcriptional activities (Figure 1A).

To further investigate the effect of Δ9-THCA-A on PPARγ in vivo, mice were

treated with Δ9-THCA-A in the presence or not of T0070907, and transcriptomic

analyses in inguinal white adipose tissue (iWAT) were performed. Differential

expression analysis of both conditions versus control mice identified a total of 3719

genes with an adjusted P ≤ 0.01 and an absolute fold change ≥ 2 (Figures 1B and 2A).

From the 850 genes changing in response to THCA, 72 belonged to the PPAR signaling

or thermogenesis pathways. On the other hand, in presence of the antagonist T0070907,

from the 3245 with significant changes, only 26 were found in those pathways (Figure

1C and 2A). UCP-1 gene, a key marker of the iWAT browning process, was found

among the top 25 genes upregulated by Δ9-THCA-A in a T0070907-sensitive manner

(Figures 1C). Moreover, western blotting and immunohistochemistry showed that Δ9-

THCA-A induced the expression of UCP-1 protein in iWAT (Figure 1D, 1E). As we

shown in Figures 2A and 2B, PPAR and thermogenesis pathways share a group of

11
commons genes indicating that both present similar regulatory mechanisms. In line with

such putative thermogenic activation, Δ9-THCA-A treatment caused a trend for

suppression of body weight gain that was independent of food intake changes but

blocked by co-administration of T0070907 (Figure 3). In addition, we identified 70

upregulated genes linked to NF-κB and cytokine-cytokine-receptor signaling, whose

expression was largely prevented by treatment with Δ9-THCA-A (Fig 4A, C)

We also found 317 genes that were modified by Δ9-THCA-A in a way

insensitive to T0070907 (Figure 4A). To functionally evaluate this gene cluster, we

performed an over-representation analysis using the KEGG pathways and Gene

Ontology annotations (Figure 4B-C). Among them, upregulation of genes related with

the fatty acid metabolism and downregulation of genes belonging to cGMP-PKG,

calcium signaling pathways and muscle tissue development, was found. Docking

analysis revealed that Δ9-THCA-A may exert its PPARγ activity by acting at both

canonical and alternative binding sites of the PPARγ LBD. Δ9-THCA-A binds to the

alternative site by interacting with Ser342 at the β-sheet (Ω-loop) through its

carboxylate group (Figure 5). However, in fluorescence resonance energy transfer

(FRET) assays we found that Δ9-THCA-A does not affect PPARγ co-regulators

interaction (Figure 6). Taken together, our results showed that the bioactivity of Δ9-

THCA-A was mediated by the PPARγ canonical pathway as well as by other pathways.

The ability of Δ9-THCA-A to influence MSCs differentiation into adipocytes or

osteoblasts was studied in vitro, after culture of MSCs in adipogenic medium. Figure

7A-B show that MSC treated with Δ9-THCA-A contained fewer and smaller lipid

droplets compared to RGZ treatment. Moreover, Δ9-THCA-A induced lower expression

of the adipogenic differentiation markers, PPARγ, aP2a, ADIPOQ, LPL and CEBPA, as

compared to cells treated with RGZ (Figure 7C). In contrast, we found that Δ9-THCA-A

12
enhanced osteoblast mineralization as well as the expression of the osteogenic

differentiation markers, Runx2, SP7, IBS and ALP, (Figure 7D-F). These data indicate

that Δ9-THCA-A qualifies as a partial PPARγ ligand significantly less adipogenic than

RZG and with an enhanced osteoblasts differentiation capacity.

3.2. Δ9-THCA-A ameliorates HFD-induced metabolic perturbations and iWAT

inflammation

Next, we analyzed the effects of Δ9-THCA-A (20 mg/kg BW/day) in a mouse model of

HFD-induced obesity. Feeding a HFD for 12-wks resulted in a significant increase in

body weight over CD controls (BW; Figure 8A), together with enhanced fat mass

(35.89 ± 2.63 g vs. 16.19 ± 1.45 g in CD; P>0.001) and adiposity index, calculated as

ratio between fat mass and fat + lean mass (36.43 ± 2.62 vs. 16.56 ± 1.49 % in CD;

P=0.001). Δ9-THCA-A treatment caused a marked suppression of BW gain in HFD

mice (Figure 8B), observed also, albeit with lesser amplitude, in CD mice. The

increased adiposity caused by HFD was reversed by Δ9-THCA-A treatment, which

caused also a significant suppression of the adiposity index in control lean mice, as

determined by Student t-test between the CD and CD+Δ9-THCA-A groups (Figure 8C).

In contrast, treatment with Δ9-THCA-A failed to significantly modify daily food intake

neither in the HFD (0.89 ± 0.08 in Δ9-THCA-A-treated vs. 1.01 ± 0.06 kcal/g BW/d in

vehicle-treated mice; P=0.2626) or the CD (1.08 ± 0.06 in Δ9-THCA-A-treated vs. 1.04

± 0.05 kcal/g BW/d in vehicle-treated mice; P=0.0617) groups.

Δ9-THCA-A improved also glucose homeostasis in HFD-induced obese mice. While

HFD exposure for 15-wks evoked an elevation of basal glucose levels, worsened

glucose tolerance after glucose bolus injection (Figure 8D) and reduced insulin

sensitivity (Figure 3E), treatment of HFD mice with Δ9-THCA-A for 3-wks resulted in

lowering of basal glycemia, and markedly improved glucose profiles, both in glucose

13
and insulin tolerance tests, which displayed better profiles than those of CD mice

without pharmacological intervention. Moreover, positive effects of Δ9-THCA-A in

terms of glucose tolerance and insulin sensitivity were also detected in lean control mice

(Figure 8D-E). This was associated to a significant lowering of basal insulin levels after

Δ9-THCA-A administration to CD and HFD mice (Figure 8F), possibly reflecting a

state of enhanced insulin sensitivity. Δ9-THCA-A treatment of obese mice largely

prevented liver fat infiltration caused by HFD and markedly reduced the steatosis score

(Figure 8G-H). In addition, Δ9-THCA-A significantly decreased serum triglyceride

levels both in obese and lean mice (Figure 8I).

The iWAT transcriptomic profile in control and HFD mice, untreated or treated

with Δ9-THCA-A, was next investigated. Differential expression analysis revealed a

total of 1387 genes overcoming the cutoff of an adjusted P ≤ 0.01 and an absolute fold

change ≥ 2 in any of the two comparisons (Figure 9A). Among them, KEGG pathway

analyses revealed that genes involved in NF-κB signaling and cytokine-cytokine

receptor were upregulated in HFD mice, which matches their inflammatory phenotype

in iWAT [33]. As described above in figures 2A and 2B, the group of common genes

between cytokine-cytokine receptor and NF-B signaling pathways indicate the shared

regulatory mechanisms between both processes. Additionally, genes belonging to the

insulin receptor signaling pathway showed lower expression in HFD mice, in line with a

state of insulin resistance. Interestingly, Δ9-THCA-A treatment of HFD mice reduced

the expression of genes belonging to the inflammatory pathways, partially recovering

the expression of those linked to the insulin signaling process (Figure 9A-B). In total,

DEGs analysis revealed that 1014 genes including those related to NF-κB signaling and

cytokine-cytokine receptor (70 genes) were modified in HFD mice and normalized in

Δ9-THCA-A-treated HFD mice (Figure 9C). Finally, to confirm the anti-inflammatory

14
profile of Δ9-THCA-A, we analyzed by qPCR the top 25 upregulated inflammatory

genes in the iWAT of HFD mice (Figure 9D), confirming the increased expression of

key components of this gene-set, including TNFα, ICAM-1, CD4, CXCL-16, CCK22,

CXCR5 and CXCR2 (Figure 9E).

In keeping with their obese phenotype, HFD mice also displayed features of

adipose tissue enlargement and inflammation. Thus, a significant increase in adipocyte

volume was observed, accompanied by signs of macrophage infiltration of WAT,

assessed by F4/80 staining of clusters of macrophages surrounding dead adipocytes, in

the so-called crown-like structures (CLS) (Figure 10A-B). Δ9-THCA-A administration

fully prevented adipocyte enlargement and the appearance of CLS in the iWAT of

HFD-induced obese mice, while substantially increasing UCP-1 content in iWAT

(Figure 10A-C).

Finally, changes in the circulating levels of key metabolic hormones were

assessed in HFD-induced obese mice, treated or not with Δ9-THCA-A. In line with their

obese phenotype, HFD mice showed increased leptin and glucagon levels, and

decreased GLP-1 concentrations vs. lean CD animals. Treatment with Δ9-THCA-A

fully normalized leptin, GLP-1 and glucagon levels, and evoked a modest, albeit

significant increase in adiponectin concentrations in HFD mice (Figure 10D-G). Alike,

proteome profiler arrays targeting a comprehensive set of circulating adipokines and

cytokines confirmed the increase of the levels of leptin, together with other adipose

born-factors, such as Oncostatin M and Serpin E1, in HFD mice, which were decreased

by Δ9-THCA-A, while adiponectin concentrations were increased. Likewise, a set of

cytokines including Endostatin, IGFBp-5, PCSK9, Adipsin and IGFBP-3 were

suppressed by HFD but increased by Δ9-THCA-A in HFD mice, while an opposite

pattern was observed for CCL11 and CCL6 (Figure 10H).

15
4. Discussion

Obesity has become a major societal problem, now recognized as the major preventable

risk factor for all-cause mortality [34, 35]. Glitazones, such as RGZ, are synthetic full

PPARγ agonists that have been used for decades for the treatment of type-2 diabetes.

However, the use of glitazones in diabetic patients has dropped significantly over the

past years due to a series of adverse side effects that include bone loss and osteoporosis,

as well as fluid retention. Bone loss could be bound to the effect of glitazones on the

lineage commitment of MSC towards osteoblasts and adipocytes, as RGZ has been

shown to suppress osteoblastogenesis and induce adipocyte differentiation [36]. The

development of more balanced, partial PPARγ agonists, devoid of the side effects

showed by the currently marketed PPARγ full agonists, is considered a major challenge

for drug discovery [37]. We provide evidence that Δ9-THCA-A is a partial PPARγ

agonist with a lower adipogenic capacity than glitazones and endowed also with

osteoblastogenesis-enhancing properties. In in vivo animal experiments, Δ9-THCA-A

could successfully ameliorated adiposity and reversed the metabolic and inflammatory

complications associated to diet-induced obesity.

Docking analysis revealed that Δ9-THCA-A may exert its PPARγ activity by acting at

both canonical and alternative binding sites of the PPARγ LBD. Δ9-THCA-A binds to

the alternative site by interacting with Ser342 at the β-sheet (Ω-loop) through its

carboxylate group (supplementary information). This may explain the observation that

Δ9-THCA-A is at least 20-fold more potent than Δ9-THC to activate PPARγ [5]. Our

data also suggest that Δ9-THCA-A may stabilize the Ω -loop region and inhibit

phosphorylation of PPARγ Ser273, thus inducing anti-diabetic activity. However, in

FRET assays we found that Δ9-THCA-A does not affect PPARγ co-regulators

16
interaction. Nevertheless, it has been shown that others partial agonists such as INT131

and BVT-13 do not induce significant binding or displacement of PPARγ co-regulators

indicating that recruitment co-activators is not essential for the signalling pathways

activated partial PPARγ agonists such as A9-THCA-A as opposite to full PPARγ

agonists [38-40]. Interestingly, SR2595, an inverse PPARγ agonist that blocks

phosphorylation of Ser273, could promote osteoblastogenesis in cultured hMSCs [41].

Thus, our in vitro and in vivo data strongly suggest that the binding of Δ9-THCA-A to

the PPARγ orthosteric site accounts for most of its effects via this nuclear receptor.

However, signaling through the alternative site or other pathways could also be of

biological relevance; in lean mice, 317 genes were modified by Δ9-THCA-A but

unaffected by T0070907. Whether this pathway is linked or not to the PPARγ

alternative site or to other pathways, such as CB1 signaling, remains to be investigated.

There is growing evidence of a link between obesity and inflammation, and Δ9-THCA-

A showed a very potent anti-inflammatory profile in HFD mice. Activation of innate

and adaptive immune response has been observed in the fat tissue of obese individuals

[42-44], and we identified 70 upregulated genes linked to NF-κB and cytokine-

cytokine-receptor signaling, whose expression was largely prevented by treatment with

Δ9-THCA-A. Within them, T- and B cell markers, as well as macrophage-derived

cytokines and chemokines, were identified, showing that both the innate and the

adaptive immune responses account for the inflammatory status in the fat tissue of HFD

mice. There is strong evidence that PPARγ inhibits NF-κB activation through several

mechanisms, repressing NF-κB-mediated transcription of pro-inflammatory cytokines

in immune and non-immune cells [45]. Thus, Δ9-THCA-A could dampen inflammation

mainly through the PPARγ pathway, opening the possibility of the use of this

phytocannabinoid in other chronic or acute inflammatory diseases.

17
The efficacy of Δ9-THCA-A to alleviate a wide spectrum of metabolic and hormonal

derangements linked to diet-induced obesity was evaluated in a validated preclinical

model of metabolic syndrome, namely, 12-wk exposure to HFD [20]. This model fully

recapitulates the cardinal manifestations of obesity and its major complications,

including an increase of body weight, fat mass, adipocyte area and adiposity index,

accompanied by perturbed glucose homeostasis and insulin resistance, as well as

enhanced liver lipid deposition and steatosis. Notably, a regimen of 3-wk treatment with

an effective daily dose of Δ9-THCA-A was sufficient to reverse such adverse metabolic

profile, causing a marked suppression of body weight gain and adiposity, significantly

lowering basal glucose and insulin levels, and fully preventing HFD-induced glucose

intolerance and insulin resistance. Moreover, circulating triglyceride levels were

reduced and steatosis scores were substantially improved by 3-wk Δ9-THCA-A

treatment in HFD mice, substantially reversing the metabolic phenotype caused by the

high fat content diet. Likewise, the major hormonal and cytokine alterations caused by

HFD were reversed by Δ9-THCA-A, with a significant decrease in circulating leptin and

glucagon, and a significant increase in serum GLP-1 and adiponectin levels. Such a

switch towards an anti-inflammatory state was also detected by our adipo-/cytokine

arrays, which confirmed the reversal of most of the pro-inflammatory humoral

alterations caused by HFD. Interestingly, the beneficial effects of Δ9-THCA-A were not

only observed in HFD conditions, but also, albeit to a lesser extent, in lean animals fed

a control diet, in which 3-wk treatment with Δ9-THCA-A was capable to reduce body

weight and adiposity, as well as basal glucose, insulin and triglyceride levels, together

with a significant enhancement of glucose tolerance and insulin sensitivity. All these

features define an optimal pharmacological profile of Δ9-THCA-A for globally handling

the metabolic syndrome linked to obesity.

18
The beneficial effects of Δ9-THCA-A on body weight and metabolic status are

somewhat at odds with the reported orexigenic and lipogenic activity of cannabinoids

[46, 47]. In our studies, Δ9-THCA-A failed to cause any detectable change in food

intake, supporting a distinct pharmacological profile compared to Δ9-THC, whose

feeding-promoting effects underlie its use in the management of cachexia [48]. The lack

of orexigenic effect together with the substantial body weight loss induced by Δ9-

THCA-A suggests a potential use of this compound to induce adaptive thermogenesis.

Indeed, chronic treatment with Δ9-THCA-A caused the upregulation of a large set of

genes belonging of the thermogenesis pathway, with a prominent increase in the UCP-1

content in iWAT form lean and HFD mice. Altogether, these findings suggest that,

acting at least partially via PPARγ, Δ9-THCA-A causes browning of the white adipose

tissue, and this activity could mechanistically underlie at least some of its beneficial

effects on energy and metabolic homeostasis. Notwithstanding, it cannot be excluded

that the 11-carboxylic metabolites of THCA-A, known to be produced in vivo [4], may

mediate at least part of the metabolic effect of THCA-A administration in our obese,

HFD mice. Altogether, these features add to our current efforts to identify

pharmacological agents capable to activate thermogenesis without undesired side effects

and suggest that Δ9-THCA-A, as well as non-decarboxylated Cannabis sativa extracts,

are worth of consideration for the management of obesity and metabolic diseases.

19
Acknowledgments

This research was funded by grants SAF2017-87701-R (EM) and BFU2017-83934-P

(MTS) (Ministerio de Economía y Competitividad, Spain; co-funded with EU funds

from FEDER Program); Project P12-FQM-01943 (M.T.-S.; Junta de Andalucía, Spain).

CIBER Fisiopatología de la Obesidad y Nutrición is an initiative of Instituto de Salud

Carlos III. None of the funding bodies played any role in the study design, data

collection and analysis, the decision to publish, or the preparation of the manuscript. BP

is a predoctoral fellow supported by the i-PFIS program, Instituto de Salud Carlos III

(IFI15/00022; European Social Fund “investing in your future”).

A version of the manuscript has been deposited in the preprint server BioRxiv

(https://doi.org/10.1101/622035).

Conflicts of Interest: The authors declare no conflict of interest in relation to the

contents of this work.

20
[1] L.O. Hanus, S.M. Meyer, E. Munoz, O. Taglialatela-Scafati, G. Appendino,

Phytocannabinoids: a unified critical inventory, Natural product reports 33(12) (2016)

1357-1392.

[2] Z. Krejci, M. Horak, F. Santavy, [Hemp (Cannabis sativa)-an antibiotic drug. 3.

Isolation and constitution of two acids from Cannabis sativa], Die Pharmazie 14 (1959)

349-55.

[3] M. Wang, Y.H. Wang, B. Avula, M.M. Radwan, A.S. Wanas, J. van Antwerp, J.F.

Parcher, M.A. ElSohly, I.A. Khan, Decarboxylation Study of Acidic Cannabinoids: A

Novel Approach Using Ultra-High-Performance Supercritical Fluid

Chromatography/Photodiode Array-Mass Spectrometry, Cannabis Cannabinoid Res

1(1) (2016) 262-271.

[4] G. Moreno-Sanz, Can You Pass the Acid Test? Critical Review and Novel

Therapeutic Perspectives of Delta(9)-Tetrahydrocannabinolic Acid A, Cannabis

Cannabinoid Res 1(1) (2016) 124-130.

[5] X. Nadal, C. Del Rio, S. Casano, B. Palomares, C. Ferreiro-Vera, C. Navarrete, C.

Sanchez-Carnerero, I. Cantarero, M.L. Bellido, S. Meyer, G. Morello, G. Appendino, E.

Munoz, Tetrahydrocannabinolic acid is a potent PPARgamma agonist with

neuroprotective activity, Br J Pharmacol 174(23) (2017) 4263-4276.

[6] Guidelines (2013) for managing overweight and obesity in adults. Preface to the

Expert Panel Report (comprehensive version which includes systematic evidence

review, evidence statements, and recommendations), Obesity 22 Suppl 2 (2014) S40.

[7] J. Karczewski, E. Sledzinska, A. Baturo, I. Jonczyk, A. Maleszko, P. Samborski, B.

Begier-Krasinska, A. Dobrowolska, Obesity and inflammation, European cytokine

network 29(3) (2018) 83-94.

21
[8] G.A. Bray, W.E. Heisel, A. Afshin, M.D. Jensen, W.H. Dietz, M. Long, R.F.

Kushner, S.R. Daniels, T.A. Wadden, A.G. Tsai, F.B. Hu, J.M. Jakicic, D.H. Ryan,

B.M. Wolfe, T.H. Inge, The Science of Obesity Management: An Endocrine Society

Scientific Statement, Endocrine reviews 39(2) (2018) 79-132.

[9] M.H. Tschop, B. Finan, C. Clemmensen, V. Gelfanov, D. Perez-Tilve, T.D. Muller,

R.D. DiMarchi, Unimolecular Polypharmacy for Treatment of Diabetes and Obesity,

Cell metabolism 24(1) (2016) 51-62.

[10] T.D. Muller, C. Clemmensen, B. Finan, R.D. DiMarchi, M.H. Tschop, Anti-

Obesity Therapy: from Rainbow Pills to Polyagonists, Pharmacological reviews 70(4)

(2018) 712-746.

[11] C. Clemmensen, B. Finan, T.D. Muller, R.D. DiMarchi, M.H. Tschop, S.M.

Hofmann, Emerging hormonal-based combination pharmacotherapies for the treatment

of metabolic diseases, Nature reviews. Endocrinology 15(2) (2019) 90-104.

[12] R.K.C. Loh, B.A. Kingwell, A.L. Carey, Human brown adipose tissue as a target

for obesity management; beyond cold-induced thermogenesis, Obesity reviews : an

official journal of the International Association for the Study of Obesity 18(11) (2017)

1227-1242.

[13] J. Liu, Y. Wang, L. Lin, Small molecules for fat combustion: targeting obesity,

Acta pharmaceutica Sinica. B 9(2) (2019) 220-236.

[14] M. Lopez, R. Nogueiras, M. Tena-Sempere, C. Dieguez, Hypothalamic AMPK: a

canonical regulator of whole-body energy balance, Nature reviews. Endocrinology

12(7) (2016) 421-32.

[15] M. Giralt, F. Villarroya, White, brown, beige/brite: different adipose cells for

different functions?, Endocrinology 154(9) (2013) 2992-3000.

22
[16] K. Ikeda, P. Maretich, S. Kajimura, The Common and Distinct Features of Brown

and Beige Adipocytes, Trends in endocrinology and metabolism: TEM 29(3) (2018)

191-200.

[17] P. Tontonoz, B.M. Spiegelman, Fat and beyond: the diverse biology of

PPARgamma, Annual review of biochemistry 77 (2008) 289-312.

[18] H. Wang, L. Liu, J.Z. Lin, T.R. Aprahamian, S.R. Farmer, Browning of White

Adipose Tissue with Roscovitine Induces a Distinct Population of UCP1(+) Adipocytes,

Cell metabolism 24(6) (2016) 835-847.

[19] M. Ahmadian, J.M. Suh, N. Hah, C. Liddle, A.R. Atkins, M. Downes, R.M. Evans,

PPARgamma signaling and metabolism: the good, the bad and the future, Nature

medicine 19(5) (2013) 557-66.

[20] B. Palomares, F. Ruiz-Pino, C. Navarrete, I. Velasco, M.A. Sanchez-Garrido, C.

Jimenez-Jimenez, C. Pavicic, M.J. Vazquez, G. Appendino, M.L. Bellido, M.A.

Calzado, M. Tena-Sempere, E. Munoz, VCE-004.8, A Multitarget Cannabinoquinone,

Attenuates Adipogenesis and Prevents Diet-Induced Obesity, Scientific reports 8(1)

(2018) 16092.

[21] D.E. Kleiner, E.M. Brunt, M. Van Natta, C. Behling, M.J. Contos, O.W.

Cummings, L.D. Ferrell, Y.C. Liu, M.S. Torbenson, A. Unalp-Arida, M. Yeh, A.J.

McCullough, A.J. Sanyal, N. Nonalcoholic Steatohepatitis Clinical Research, Design

and validation of a histological scoring system for nonalcoholic fatty liver disease,

Hepatology 41(6) (2005) 1313-21.

[22] A. Subramanian, P. Tamayo, V.K. Mootha, S. Mukherjee, B.L. Ebert, M.A.

Gillette, A. Paulovich, S.L. Pomeroy, T.R. Golub, E.S. Lander, J.P. Mesirov, Gene set

enrichment analysis: a knowledge-based approach for interpreting genome-wide

23
expression profiles, Proceedings of the National Academy of Sciences of the United

States of America 102(43) (2005) 15545-50.

[23] G. Yu, L.G. Wang, Y. Han, Q.Y. He, clusterProfiler: an R package for comparing

biological themes among gene clusters, Omics : a journal of integrative biology 16(5)

(2012) 284-7.

[24] G.M. Morris, R. Huey, W. Lindstrom, M.F. Sanner, R.K. Belew, D.S. Goodsell,

A.J. Olson, AutoDock4 and AutoDockTools4: Automated docking with selective

receptor flexibility, Journal of computational chemistry 30(16) (2009) 2785-91.

[25] O. Trott, A.J. Olson, AutoDock Vina: improving the speed and accuracy of

docking with a new scoring function, efficient optimization, and multithreading, Journal

of computational chemistry 31(2) (2010) 455-61.

[26] E.H. Baugh, S. Lyskov, B.D. Weitzner, J.J. Gray, Real-time PyMOL visualization

for Rosetta and PyRosetta, PloS one 6(8) (2011) e21931.

[27] M.A. Lee, L. Tan, H. Yang, Y.G. Im, Y.J. Im, Structures of PPARgamma

complexed with lobeglitazone and pioglitazone reveal key determinants for the

recognition of antidiabetic drugs, Scientific reports 7(1) (2017) 16837.

[28] M.V. Liberato, A.S. Nascimento, S.D. Ayers, J.Z. Lin, A. Cvoro, R.L. Silveira, L.

Martinez, P.C. Souza, D. Saidemberg, T. Deng, A.A. Amato, M. Togashi, W.A. Hsueh,

K. Phillips, M.S. Palma, F.A. Neves, M.S. Skaf, P. Webb, I. Polikarpov, Medium chain

fatty acids are selective peroxisome proliferator activated receptor (PPAR) gamma

activators and pan-PPAR partial agonists, PloS one 7(5) (2012) e36297.

[29] G. Brusotti, R. Montanari, D. Capelli, G. Cattaneo, A. Laghezza, P. Tortorella, F.

Loiodice, F. Peiretti, B. Bonardo, A. Paiardini, E. Calleri, G. Pochetti, Betulinic acid is

a PPARgamma antagonist that improves glucose uptake, promotes osteogenesis and

inhibits adipogenesis, Scientific reports 7(1) (2017) 5777.

24
[30] T. Waku, T. Shiraki, T. Oyama, K. Maebara, R. Nakamori, K. Morikawa, The

nuclear receptor PPARgamma individually responds to serotonin- and fatty acid-

metabolites, The EMBO journal 29(19) (2010) 3395-407.

[31] T.S. Hughes, P.K. Giri, I.M. de Vera, D.P. Marciano, D.S. Kuruvilla, Y. Shin, A.L.

Blayo, T.M. Kamenecka, T.P. Burris, P.R. Griffin, D.J. Kojetin, An alternate binding

site for PPARgamma ligands, Nature communications 5 (2014) 3571.

[32] J.H. Choi, A.S. Banks, T.M. Kamenecka, S.A. Busby, M.J. Chalmers, N. Kumar,

D.S. Kuruvilla, Y. Shin, Y. He, J.B. Bruning, D.P. Marciano, M.D. Cameron, D.

Laznik, M.J. Jurczak, S.C. Schurer, D. Vidovic, G.I. Shulman, B.M. Spiegelman, P.R.

Griffin, Antidiabetic actions of a non-agonist PPARgamma ligand blocking Cdk5-

mediated phosphorylation, Nature 477(7365) (2011) 477-81.

[33] M.F. Gregor, G.S. Hotamisligil, Inflammatory mechanisms in obesity, Annu Rev

Immunol 29 (2011) 415-45.

[34] M. Bluher, Obesity: global epidemiology and pathogenesis, Nature reviews.

Endocrinology 15(5) (2019) 288-298.

[35] A. Hruby, F.B. Hu, The Epidemiology of Obesity: A Big Picture,

PharmacoEconomics 33(7) (2015) 673-89.

[36] A.A. Amato, F. de Assis Rocha Neves, Idealized PPARgamma-Based Therapies:

Lessons from Bench and Bedside, PPAR research 2012 (2012) 978687.

[37] C.K. Tan, Y. Zhuang, W. Wahli, Synthetic and natural Peroxisome Proliferator-

Activated Receptor (PPAR) agonists as candidates for the therapy of the metabolic

syndrome, Expert Opin Ther Targets 21(3) (2017) 333-348.

[38] I.M. Chrisman, M.D. Nemetchek, I.M.S. de Vera, J. Shang, Z. Heidari, Y. Long,

H. Reyes-Caballero, R. Galindo-Murillo, T.E. Cheatham, 3rd, A.L. Blayo, Y. Shin, J.

Fuhrmann, P.R. Griffin, T.M. Kamenecka, D.J. Kojetin, T.S. Hughes, Defining a

25
conformational ensemble that directs activation of PPARgamma, Nature

communications 9(1) (2018) 1794.

[39] T. Ostberg, S. Svensson, G. Selen, J. Uppenberg, M. Thor, M. Sundbom, M.

Sydow-Backman, A.L. Gustavsson, L. Jendeberg, A new class of peroxisome

proliferator-activated receptor agonists with a novel binding epitope shows antidiabetic

effects, The Journal of biological chemistry 279(39) (2004) 41124-30.

[40] U. Kintscher, M. Goebel, INT-131, a PPARgamma agonist for the treatment of

type 2 diabetes, Current opinion in investigational drugs 10(4) (2009) 381-7.

[41] D.P. Marciano, D.S. Kuruvilla, S.V. Boregowda, A. Asteian, T.S. Hughes, R.

Garcia-Ordonez, C.A. Corzo, T.M. Khan, S.J. Novick, H. Park, D.J. Kojetin, D.G.

Phinney, J.B. Bruning, T.M. Kamenecka, P.R. Griffin, Pharmacological repression of

PPARgamma promotes osteogenesis, Nature communications 6 (2015) 7443.

[42] K.J. Chung, M. Nati, T. Chavakis, A. Chatzigeorgiou, Innate immune cells in the

adipose tissue, Reviews in endocrine & metabolic disorders 19(4) (2018) 283-292.

[43] Y. Zhao, L. Lin, J. Li, Z. Xiao, B. Chen, L. Wan, M. Li, X. Wu, C. Hin Cho, J.

Shen, CD4(+) T cells in obesity and obesity-associated diseases, Cellular immunology

332 (2018) 1-6.

[44] S.M. Reilly, A.R. Saltiel, Adapting to obesity with adipose tissue inflammation,

Nature reviews. Endocrinology 13(11) (2017) 633-643.

[45] X. Zhang, H.A. Young, PPAR and immune system--what do we know?, Int

Immunopharmacol 2(8) (2002) 1029-44.

[46] V. Di Marzo, S.K. Goparaju, L. Wang, J. Liu, S. Batkai, Z. Jarai, F. Fezza, G.I.

Miura, R.D. Palmiter, T. Sugiura, G. Kunos, Leptin-regulated endocannabinoids are

involved in maintaining food intake, Nature 410(6830) (2001) 822-5.

26
[47] D. Cota, G. Marsicano, M. Tschop, Y. Grubler, C. Flachskamm, M. Schubert, D.

Auer, A. Yassouridis, C. Thone-Reineke, S. Ortmann, F. Tomassoni, C. Cervino, E.

Nisoli, A.C. Linthorst, R. Pasquali, B. Lutz, G.K. Stalla, U. Pagotto, The endogenous

cannabinoid system affects energy balance via central orexigenic drive and peripheral

lipogenesis, The Journal of clinical investigation 112(3) (2003) 423-31.

[48] A.N. Verty, I.S. McGregor, P.E. Mallet, Paraventricular hypothalamic CB(1)

cannabinoid receptors are involved in the feeding stimulatory effects of Delta(9)-

tetrahydrocannabinol, Neuropharmacology 49(8) (2005) 1101-9.

Figure 1. Characterization of Δ9-THCA-A as a selective PPARγ agonist. (A)

Receptor-specific transactivation by Δ9-THCA-A. HEK-293T cells were co-transfected

with the plasmids encoding nuclear receptors (GAL4-PPARγ, GAL4-PPARα and

GAL4-PPARδ) and their cognate luciferase reporter (GAL4-luc). After transfection,

cells were treated with Δ9-THCA-A (10 μM) and receptor-specific agonists for 6 hours.

Control (white bars), Δ9-THCA-A (blue bars) and specific ligands for each receptor

(grey bars): RGZ (1 μM) for PPARγ, WY14643 (5 μM) for PPARα, and GW0742 (5

μM) for PPARδ. Results are expressed as the fold induction ± SD (n = 3-5) relative to

untreated control. *P < 0.05, **P < 0.01 and ***P < 0.001 agonist ligands or Δ9-THCA-

A treatment vs. control (ANOVA followed by Tukey’s test). (B) Pie charts indicating

the number and proportion of differentially expressed genes included in the KEGG

pathways of interest for each comparison. (C) Heatmap of the top 25 genes induced by

Δ9-THCA-A inside the PPAR signaling or thermogenesis pathways that are not

27
differentially expressed in the Δ9-THCA-A+T0070907 vs control comparison. (D, E)

Representative Western blot images of UCP-1 protein expression and immuno-

histochemistry with anti-UCP-1 antibodies in iWAT tissue (original magnification × 10,

scale bar: 200 μm) (n = 3).

Figure 2. Transcriptomic analysis of the PPAR dependent activity of Δ9-THCA-A

in iWAT. (A) Heatmap of all the differentially expressed genes (absolute fold change ≥

2 and an adjusted P value ≤ 0.01) in 9-THCA-A versus control or 9-THCA-

A+T0070907 versus control comparisons. The color represents the scaled mean of log

transformed expression. The column annotations indicate the sample group and the

points at the right side highlight the position of genes belonging to the KEGG pathways

of interest. (B) Gene set enrichment analysis results for the KEGG pathways of interest.

The left side enrich plots indicate the position of the genes belonging to each pathway in

the pre-ranked list per comparison. The right-side bar plots represent the normalized

enrichment score (NES) and significance of the GSEA result *P ≤ 0.05; **P ≤ 0.01;

***P ≤ 0.001.

Figure 3. Effect of 3 weeks treatment of Δ9-THCA-A combined or not with of

T0070907 on body weight gain and food intake. (A) BW gain, and (B) Daily food

intake (Kcal/g BW/day) during the treatment, for the three experimental groups. Data

are presented as mean ± SEM (n= 6 mice per group).

Figure 4. Transcriptomic analysis of the PPAR independent activity of Δ9-THCA-

A in iWAT. (A) Heatmap of the 317 genes that are modified in the same direction by

both 9-THCA-A and 9-THCA-A+T0070907 treatments. The color represents the

28
scaled mean of log transformed expression. The column annotations indicate the sample

group. (B) Over-represented KEGG pathways and Gene Ontology (Biological Process)

terms in the clusters of up or down regulated genes by both treatments. The presence of

a point indicates the over representation (Fisher Exact Test Adjusted P ≤ 0.05) of a

pathway or term (Y axis) in a group of genes (X axis).

Figure 5. Ligand docking, and binding properties of Δ9-THCA-A to PPARγ. (A)

9
PPARγ LBD structure 5Y2O ser342 bound to Δ -THCA-A (blue), B.E. VINA

Kcal/mol= -8.8; B.E. AutoDock Kcal/mol= -10.58; Predicted Ki 17.45 nM. (B) PPARγ
9
LBD structure 4EMA ser342 bound to Δ -THCA-A (blue), B.E. VINA Kcal/mol= -7.8;

B.E. AutoDock Kcal/mol= -8.2; Predicted Ki 982.28 nM. (C) PPARγ LBD structure
9
5LSG ser289 in Helix 3 bound to Δ -THCA-A in the orthosteric site (blue), B.E. VINA

Kcal/mol= -7.5; B.E. AutoDock Kcal/mol= -8.8; Predicted Ki 302.1 nM. (D) PPARγ
9
LBD structure 5LSG bound to RGZ in the orthosteric site (red), and Δ -THCA-A in the

alternative site (blue).

Figure 6. Effects on PPARγ-ligand interactions in presence of Δ9-THCA-A in

LanthaScreen assays. (A) TR-FRET assay was employed to study coactivator peptide

recruitment to the human PPARγ LBD in response to RGZ or Δ9-THCA-A. Data are

expressed as a percentage of the maximum RGZ response. Three experiments were

performed, and representative graphs are shown. (B) Corepressor peptide displacement

to human PPARγ LBD in response to RGZ or Δ9-THCA-A was examined by TR-FRET

assay. Data are expressed as a percentage of the maximum recruitment in the absence of

ligand. Representative data from three experiments are shown. (C) Alternative site

binding for Δ9-THCA-A in PPARγ was tested by using increasing concentrations of

29
RGZ or Δ9-THCA-A both in the absence or the presence of 5 µM T0070907 PPARγ

antagonist. TRAP220 data are expressed as a percentage of the maximum RGZ

response, while SMRT data are expressed as a percentage of the maximum recruitment

in the absence of ligand. Representative graphs from three experiments are show

Figure 7. Effect of Δ9-THCA-A on MSCs adipogenic and osteoblastogenic

differentiation. MSCs were cultured under adipogenic medium (AM) in the presence of

RGZ or Δ9-THCA-A. (A) Representative images of the cells stained with Oil Red O

assayed by light microscopy (×10) after 21 days of differentiation. (B) Quantification of

the stained lipid droplets in Oil Red O (ORO+ cells) was performed measuring

absorbance at 540nm. (C) mRNA levels of adipogenic markers were analyzed by qPCR

after 14 days of differentiation. (D) MSCs were cultured under osteoblastogenic

medium (OM) in the presence of Δ9-THCA-A and mineralization detected by Alizarin

red staining was assessed by gross appearance after 21 days of differentiation. (E)

Quantification of the eluted Alizarin Red stain measuring absorbance at 405nm. (F)

Gene expression of osteoblastogenic markers analyzed by qPCR. Results represent the

mean ± S.D (n = 3). *P < 0.05, **p < 0.01 and ***P < 0.001 RGZ or Δ9-THCA-A vs.

AM; Δ9-THCA-A vs. OM. (ANOVA followed by Tukey’s test).

Figure 8. Effect of administration of Δ9-THCA-A on metabolic and hormonal

parameters in a mouse model of HFD-induced obesity. (A) Body weight (BW)

evolution of adult male mice fed for 12-weeks with high fat diet (HFD) or the

corresponding control diet (CD). (B) BW change in HFD and CD mice treated for three

weeks with Δ9-THCA-A or vehicle; values are referenced to BW at the beginning of

treatment (taken as 0). (C) Percentage of adiposity, at the end of treatments in the four

30
experimental groups. (D-E) Glucose and Insulin tolerance tests in CD and HFD mice

treated with Δ9-THCA-A or vehicle for three weeks. (F) Basal insulin levels at the end

of the three-week treatment period are shown for the four experimental groups. (G)

Liver sections with hematoxylin and eosin (H&E) staining (original magnification x10,

scale bar: 200 μm). (H) Steatosis scores (n = 6 mice per group) and (I) plasma levels of

triglycerides. Values correspond to means ± SEM (n= 5-8 mice per group). *P<0.05,

**P<0.01, ***P<0.001 Δ9-THCA-A-treated mice or HFD mice vs. control (CD) mice;

#P<0.05, ##P<0.01, ###P<0.001 Δ9-THCA-A-treated HFD mice vs. HFD mice treated

with vehicle (ANOVA followed by Tukey’s test).

Figure 9. Transcriptomic analysis of Δ9-THCA-A effects in the iWAT of HFD

mice. (A) Heatmap of all the differentially expressed genes (absolute fold change ≥ 2

and an adjusted P value ≤ 0.01) in HFD (brown box) versus control (black box) or

HFD+Δ9-THCA-A (green box) versus control comparisons. The color represents the

scaled mean of log transformed expression. The column annotations indicate the sample

group and the points at the right side highlight the position of genes belonging to the

KEGG pathways of interest. (B) Gene set enrichment analysis results for the KEGG

pathways of interest. The left side enrich plots indicate the position of the genes

belonging to each pathway in the pre-ranked list per comparison. The right-side bar

plots represent the normalized enrichment score (NES) and significance of the GSEA

result *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. (C) Pie charts indicating the number and

proportion of differentially expressed genes included in the KEGG pathways of interest

for each comparison. (D) Heatmap of the top 25 genes induced by HFD (brown box)

inside the NF-κB or cytokine-cytokine receptor pathways that are not differentially

expressed in the HFD+Δ9-THCA-A (green box) vs control (black box) comparison. (E)

31
Gene expression of pro-inflammatory genes were measured by qPCR. Results are

presented as mean ± SEM (n= 4-8 mice per group). *P<0.05, **P<0.01 HFD mice vs.

control mice; #P<0.05, ##P<0.01 Δ9-THCA-A-treated HFD mice vs. HFD mice treated

with vehicle (ANOVA followed by Tukey’s test)

Figure 10. Effects of Δ9-THCA-A on iWAT browning, adiposity and circulating

factors, with key roles in metabolic homeostasis in CD and HFD animals. (A)

Crown Like Structures (CLS) and browning in iWAT. Representative

immunohistochemical detection of anti-F4/80 and anti-UCP-1 antibodies (original

magnification x20, scale bar: 100 μm), (B) Quantification of adipocyte area (n = 6

animals per group), (C) UCP-1 protein levels determined by western blotting in iWAT

tissue (n=3). Hormonal markers linked to energy and metabolic homeostasis assayed:

(D) Leptin, (E) GLP-1, (F) Glucagon and (G) Adiponectin. (H) Heatmap showing the

plasma profile of cytokines and adipokines. Values correspond to means ± SEM (n= 5-8

mice per group). *P<0.05, ***P<0.001 HFD mice vs. control (CD) mice; #P<0.05,

###P<0.001 Δ9-THCA-A-treated HFD mice vs. HFD mice treated with vehicle

(ANOVA followed by Tukey’s test).

Highlights

32
- Δ9-THCA-A is a partial PPARγ ligand agonist with low adipogenic activity
- Δ9-THCA-A enhances osteoblastogenesis in bone marrow derived mesenchymal stem
cells.
- Δ9-THCA-A reduces body weight gain, fat mass, and liver steatosis in HFD-fed mice
- Δ9-THCA-A improves glucose tolerance, insulin sensitivity, and insulin profiles in
vivo
- Δ9-THCA-A induces browning of iWAT and has a potent anti-inflammatory activity

Table 1. Primers used in this study.

Gene Forward (5´  3´) Reverse (5´  3´)

h-PPARγ2 GCGATTCCTTCACTGATACACTG GAGTGGGAGTGGTCTTCCATTAC
h-LPL GGCGCTACCTTGAGATAGAGTTCTG TGTTTTCTACAGGGTGCTTTAGATGAC
h-aP2a CCAGGAATTTGACGAAGT TCTCTTTATGGTGGTTGATT
h-CEBPA CCTTGTGCCTTGGAAATGCAAAC CTGCTCCCCTCCTTCTCTCA
h-ADIPOQ CATGACCAGGAAACCACGACTC CCGATGTCTCCCTTAGGACCA
h-RUNX2 TGGTTAATCTCCGCAGGTCAC ACTGTGCTGAAGAGGCTGTTG
h-ALP CCAACGTGGCTAAGAATGTCATC TGGGCATTGGTGTTGTACGTC
h-SP7 AGCCAGAAGCTGTGAAACCTC AGCTGCAAGCTCTCCATAACC
h-IBSP AGGGCAGTAGTGACTCATCCG CGTCCTCTCCATAGCCCAGTGTTG
h-HPRT ATGGGAGGCCATCACATTGT ATGTAATCCAGCAGGTCAGCAA
m-TNFα CTACTCCCAGGTTCTCTTCAA GCAGAGAGGAGGTTGACTTTC
m-ICAM1 GTGGCGGGAAAGTTCCTG CGTCTTGCAGGTCATCTTAGGAG
m-CD4 TCCTTCCCACTCAACTTTGC AAGCGAGACCTGGGGTATCT
m-CXCL16 CGTTGTCCATTCTTTATCAGGTTCC TTGCGCTCAAAGCAGTCCA
m-CCL22 AAGACAGTATCTGCTGCCAGG GATCGGCACAGATATCTCGG
m-CXCR5 ACTCCTTACCACAGTGCACCTT GGAAACGGGAGGTGAACCA
m-CXCR2 CACCGATGTCTACCTGCTGA CACAGGGTTGAGCCAAAA
m-GAPDH TGGCAAAGTGGAGATTGTTGCC AAGATGGTGATGGGCTTCCCG

33
 Lean mice

9-THCA-A PPAR Thermogenesis

HFD obese mice Thermogenesis

Weight
Inflammation

Liver steatosis

A B
80 ***
800
8
60 *** Control
Legend
*** 600
6 THCA vs Control THCA+T007 vs Control
40 Agonist
Legend
400
4 ***
Fold Induction

20 200 THCA
Legend
15 2
* 1155
* 778 3219
26
10 100 72
1

5 **
5
5 **

Inside PPAR signaling or

0 0
0
Rest of DEGs
Thermogenesis
PPARγ PPARα PPARδ

A B
Modified by THCA and THCA+T007 Over-represented KEGG pathways
(317 genes)
D
C
A B Control THCA THCA + T007
Modified by THCA and THCA+T007 Over-represented KEGG pathways
36
(317 genes)
UCP-1

55 Tubulin

E
CONTROL
Control THCA
THCA THCA+T007
THCA + T007
(UCP-1)
(UCP-1)
iWAT

C
iWAT

Over-represented Gene Ontology

(Biological Process) terms
200µm 200µm 200µm

C
Over-represented Gene Ontology
(Biological Process) terms

Scaled averaged Log expression

Scaled averaged Log expression

Figure 1
A B THCA vs Control

THCA+T007 vs Control

Figure 2
 1,5
1.5

Kcal/g BW/day
1 .1
0

0,5
0.5

0 .0
0

Figure 3
A B
Modified by THCA and THCA+T007 Over-represented KEGG pathways
(317 genes)

C
Over-represented Gene Ontology
(Biological Process) terms

Scaled averaged Log expression

Figure 4
A S342 S342 B
H3
H3

β3
β3

C H3 H3 S342 D
S289

β3

Figure 5
A B

Figure 6
A CONTROL AM RGZ 1μM B
50
50
***

40
40
***

% ORO+
30 ***

ORO+
30

20
20

%
THCA 1μM THCA 5μM THCA 10μM
10
10

00
--
C AAM
M R++ZG
AM +
TH+ 1C +
+M
TµA
HTC + ++1M0 µ M
+CM5Aµ
1HµA
RGZ (µM) --
RGZ ( M ) -
- 1
1
-- -- --
THCA (µM) - - - 1 5 10
THCA ( M ) - - - 1 5 10

12000
12000 ***
C
8000
8 000
7000
7 000 **
x p r e s s i o n ( %(%)
)

6000
6 000
R e l a t i v e eexpression

5000
5 000 *

4000
4 000
Relative

3000
3 000

2000
2 000
***
*** ***
*** *** ***
*** *** ***
1000
1 000 ** ***
*** * *

0
0
PAM
AgMCPPog- ngR+
-Pg +P
tAM
rTP+
ZG
go+H
+
lgTC
1H
+TLA+1A
uC
H
P
+MC
Lu+5LC
+LA M
P -oLPM
Lu-P1 0
L
L +R
n+LuP
tL
AM
rM
T+
ZG
LPHlLT+
+o +C1
F
H
TA +HM1A
uAB
+C
+uFC
C
F
+ A
AB
5
FM-oAB
AB
-u 1F
M0+
n+RAB
tF
u
AM+
rT
ZG
AB
M
+
+CC1H
oHlT+ TA+HM
uE
+CB +CC
1A
C
+uA
E5M -B
-Euo
E1C
B
M0+R
n
B
+ E
C +
tuAM
rT
ZG
B
+E
M
oHlT++C
B AD
1H+C
TA
u
+HQ
M +AD
1AD
A
C
+uC
AD
A
5M-ou1Q
-Q AD
n
Q
M+AD
+0R
tu
AM+oH
rQ
T
ZG
+MlT++C1H
Q u+C
TA
+ HM +uA5Mu1M
1+
A
C 0uM
RGZ (µM) - - 1 - - - - - 1 - - - - - 1 - - - - - 1 - - - - - 1 - - -
RGZ ( M ) - - 1 - - - - - 1 - - - - - 1 - - - - - 1 - - - - - 1 - - -
THCA (µM) - - - 1 5 10 - - - 1 5 10 - - - 1 5 10 - - - 1 5 10 - - - 1 5 10
THCA ( M ) - - - 1 5
PPARγ2 10 - - - 1
LPL 5 10 - - - 1 5
aP2a 10 - - - 1 5
CBPA 10 - - - 1 5 10
ADIPOQ
PPAR- 2 LPL aP 2a CEBPA A D IP O Q
D E 800
800
CONTROL OM THCA 1μM THCA 5μM THCA 10μM
e d ( %(%)

600
600
)

*
Red
A liz a r in R

400
400
Alizarin

200
200

400
400 00
F OM
O M -
-C +H
O
T
+
+TAH+
MC 1T AM+C5 A
CµH µM
+
+1 0 µ M + +
***
*** THCA
R G Z ((µM)
M ) -- -- 1 5 - 10-
expression (%)

1 -
R e la tiv e e x p r e s s io n ( % )

300
300 THCA ( M ) - - - 1 5 10
*
**

200
200
*
Relative

100
100

00
OM - +TCHA
O M C
- TOH+M
+
T
+CH
1A+
µ
CM
+ A+
5 µ - OH++M
+1M0C-µTM
+
TC H
A
T+ C
H +
1A
µ
CM
+ A+
5 µ --TM
+1M0Cµ
++M
OH +A
TC HT+C
H1+
A
µ
C
+MA+
5 -µ-TM
µ+1M0C ++M
OH TC+
H
A
T+C
H1+
A
µ
C A+µ+1M0 µ M
+M
5
THCA (µM) -
THCA ( M ) -
-- 1 5 10 - -- 11 5
1 5 10 -
10 - -- 11 55 10
5 10 -
- - - 11 55 10
10 - 10

RUNX2 SP7 IBSP ALP

Figure 7
RUNX2 SP7 IB S P ALP
 A B C
Treatment time (weeks)
13 13.5 14 14.5 15 15.5 50
50
45 CD
HFD 2 ** 40 ***

(Fat/Fat+Lean)
Δ Body Weight (g)
** 40
*

% Adiposity
Body Weight (g)

**
40 ** 0
** * 30
30 ###
**
35 ** -2 ###
**
* 20
** 20
**
** -4 ###
30 ** *
* -6 10
10
CD ###
25 -8 CD + THCA 00
HFD ### ### HFD - - + +
20 -10 HFD + THCA
0 2 4 6 8 10 12
THCA (20mg/kg) - + - +
Weeks

D E F
CD CD+ + THCA 240 CD
CD CD+ THC-A
CD + THCA(20mg/kg)
600
600 CD CD THC-A (20mg/kg) 240
HFD
HFD HFD
HFD + THCA
+ THC-A (20mg/kg) 10000
10000
HFD
HFD HFD+ THC-A
HFD + THCA(20mg/kg)
210
210
(mg/dl)
(mg/dl)

500

Insulin (pg/mL)
500 8000
8000
*
Glucose (mg/dl)
Glucose (mg/dl)

180
180
400
400 6000
6000 #
Glucose

*
Glucose

150
150
300
300 ** 4000
4000

120 ##
** 120
200
200 *
2000
2000
** #
### #
90
90
*
* 00
100
100 00
0 20 60 120 0
0 20
20 60
60 120
120 HFD
HC
FDO N -T- RTOH-L-C AHH+
F+FDD ++
+T H C A
0 20 60 120 Time (min)
Time (min) T H C A ( 2 0 m g /k g )
Time (min) Time (min) THCA (20mg/kg) -- ++ -- ++

G VEHICLE THCA
H I
*** 300
33 300
Triglycerides (mg/dL)
CONTROL

Steatosis score

#
S te a t o s is s c o r e

22 200
200

200µm 200µm *
###
11 100
100
HFD

00 00
HFD C-C
H F DT H -DA H( 2-F +g+
- 0DmH F+/k H)C++A
TDg ( 2 0 m g /k g ) HFD C-C
H F DT H -- 0DmH+g+
-DA H( 2F F+/k H)C+
TDg +A ( 2 0 m g /k g )

200µm 200µm T H C A ( 2 0 m g /k g ) - + - + T H C A ( 2 0 m g /k g ) - + - +
THCA (20mg/kg) - + - + THCA (20mg/kg) - + - +

Figure 8
A B HFD vs Control

HFD+THCA vs Control

C D
HFD vs Control HFD+THCA vs Control

944 525
70 23

Inside NF-kappa B signaling or

Rest of DEGs
Cytokine-Cytokine receptor

E
C o nControl
tro l HHFD
FD H F D + T H C A ( 2 0 m g /k g )
HFD+THCA
60
60
60 *
40
40 *
20 40
R e l a t i v e expression

20
e x p r e s s io n

2*0
R e la tiv e e x p r e s s io n

**
10
10 * * **
**
10 * ** #
** *
Relative

55 #
** #
*
# # ##
# ##
5
00 #
# # ##
f #
m 4 6 2 ## 5 2
tn a c
d l1 l2 R R
ic c c C C
c
x c
C
X
C
X Figure 9
0
16

2
f

4
am

2
tn

l2
cd

R
cl

cc

XC

XC
ic

cx

C
A CD HFD HFD+THCA B
2000
2000

Adipocyte area (µm 2)

(F4/80)
***
iWAT

1500
1500

1000
1000 ###
100µm 100µm 100µm

500
500
(UCP-1)
iWAT

00
HFD - + +
100µm 100µm 100µm THCA (20mg/kg) - - +

C CD HFD HFD + THCA

36
UCP-1
iWAT

50 Tubulin

D E
Leptin GLP-1
50000
50000 120
120
Leptin (pg/mL)

GLP-1 (pg/mL)

*** #
40000
40000

80
80
30000
30000 H
20000
20000 *
40
40
###
10000
10000

00 00
HFD - + + HFD - + +
THCA (20mg/kg) - - + THCA (20mg/kg) - - +

F G
Glucagon Adiponectin
400
400 12
12
Adiponectin (pg/mL)

* *
Glucagon (pg/mL)

300
300
88

200
200

44
100
100

00 00
HFD - + + HFD - + +

THCA (20mg/kg) - - + THCA (20mg/kg) - - +

Figure 10