Cannabidiol acts as molecular switch in innate immune cells to promote the biosynthesis of inflammation-resolving lipid mediators

Summary

Cannabinoids are phytochemicals from cannabis with anti-inflammatory actions in immune cells. Lipid mediators (LM), produced from polyunsaturated fatty acids (PUFA), are potent regulators of the immune response and impact all stages of inflammation. How cannabinoids influence LM biosynthetic networks is unknown. Here, we reveal cannabidiol (CBD) as a potent LM class-switching agent that stimulates the production of specialized pro-resolving mediators (SPMs) but suppresses pro-inflammatory eicosanoid biosynthesis. Detailed metabololipidomics analysis in human monocyte-derived macrophages showed that CBD (i) upregulates exotoxin-stimulated generation of SPMs, (ii) suppresses 5-lipoxygenase (LOX)-mediated leukotriene production, and (iii) strongly induces SPM and 12/15-LOX product formation in resting cells by stimulation of phospholipase A₂-dependent PUFA release and through Ca²⁺-independent, allosteric 15-LOX-1 activation. Finally, in zymosan-induced murine peritonitis, CBD increased SPM and 12/15-LOX products and suppressed pro-inflammatory eicosanoid levels in vivo. Switching eicosanoid to SPM production is a plausible mode of action of CBD and a promising inflammation-resolving strategy.

eTOC Blurb:

Peltner et al. reveal cannabidiol from cannabis as lipid mediator class-switching agent that stimulates innate immune cells for producing specialized pro-resolving mediators and suppresses pro-inflammatory leukotriene formation. Cannabidiol activates 15-lipoxygenase-1 via an allosteric site which causes SPM production in human macrophages and in zymosan-induced murine peritonitis in vivo.

Graphical Abstract

Introduction

Inflammation as crucial host response counteracts invading pathogens and tissue damage, and restores homeostasis.¹ Host-derived lipid mediators (LM) potently regulate all stages of inflammation, some with detrimental effects if chronically produced, while others are constitutively formed with beneficial impacts to normal tissue function.^2–4 LMs are formed from liberated polyunsaturated fatty acids (PUFA) mainly by cyclooxygenase (COX) and lipoxygenase (LOX) pathways (Figure 2A), where distal enzymes accomplish the formation of various functionally distinct bioactive mediators.⁵ Thus, the pro-inflammatory leukotrienes (LTs) and prostaglandins (PGs) are produced from arachidonic acid (AA) via 5-LOX and COX-1/2 as key enzymes, respectively, to mount and maintain inflammation. In contrast, the specialized pro-resolving mediators (SPMs) act as immunoresolvents to terminate the inflammatory process and promote tissue regeneration and the return to homeostasis.^3,6 SPMs comprise resolvins, protectins, and maresins derived from docosahexaenoic acid (DHA) and eicosapentaenoic acid (EPA), and the AA-derived lipoxins, with 12-/15-LOXs as major biosynthetic enzymes.^7,8 Aberrations in SPM formation and, thus, in the resolution process can lead to persistent and excessive inflammation, which worsens disease progression and fosters the development of chronic disorders.^7,8

Figure 2. Cannabinoids stimulate and induce LM formation in MDM.

(A) Quantitative LM pathway analysis and effects of CBD in exotoxin-stimulated M2-MDM. Cells were preincubated with CBD (10 μM) or vehicle (DMSO, 0.1%) for 15 min before challenge with SACM (1%) for 90 min. Formed LM were quantified in the supernatants. Node size represents the mean values in pg/2 × 10⁶ cells, and intensity of color denotes the fold change of CBD- versus vehicle-treated cells for each LM; n = 5. (B) Human M2-MDM were incubated with cannabinoids (10 μM) or vehicle (DMSO, 0.1%) for 90 min. Formed LM were quantified in the supernatants using UPLC-MS-MS. Results are given as a heatmap presenting the -fold change to vehicle control; n = 5. (C) PCA of analyzed LM from M1- and M2-MDM incubated with cannabinoids (10 μM) for 90 min. Results are shown in colors corresponding to M1 or M2 polarization and labeled with CBD or vehicle treatment. n = 3 (M1-MDM) or n = 5 (M2-MDM). (D) Selected LM produced in M2-MDM upon stimulation with CBD (10 μM) for 90 min given as means ± S.E.M of n = 35 (RvD5) or n = 25 (LTB4) independent experiments. ****p < 0.0001; versus vehicle. Data were log-transformed for statistical analysis; paired t-test. (E) PMNL (5 × 10⁶ cells), platelets (2.5 × 10⁸ cells), or M1- and M2-MDM (2 × 10⁶ cells) were incubated with vehicle (0.1% DMSO) or CBD (10 μM) for 90 min. Formed LM were quantified in the supernatants. 15-LOX products (sum of 17-HDHA, 15-HEPE and 15-HETE), RvD5 and LTB₄ are shown as -fold increase ± S.E.M of CBD- versus vehicle-treated samples (n.d. = not detectable). Data were log-transformed for statistical analysis; ratio-paired t-test, *p < 0.05, **p < 0.01; versus vehicle.

The major therapeutics to treat inflammation-related diseases are glucocorticoids and non-steroidal anti-inflammatory drugs (NSAIDs);^9,10 however, these agents dampen inflammation and immune responses, especially in early phases and fail to promote inflammation resolution. Additionally, these drugs frequently exert severe on-target side effects, e.g., the gastrointestinal toxicity of COX-inhibiting NSAIDs.^9–11 In this respect, enzyme manipulation in LM-producing cells for shifting the formation of pro-inflammatory LTs and PGs towards SPMs represents an alternative strategy with potential for pharmacotherapy of unresolved inflammation.^12–14 Neutrophils, monocytes and especially macrophages, are immune cells with marked capacities to release several inflammation-related LMs, and therefore, are attractive targets for LM-manipulating drugs to intervene with inflammatory disorders.^14–16 Macrophages occur as different phenotypes, such as pro-inflammatory M1-like cells at the onset of inflammation, but also as pro-resolving M2-like subtypes for termination and resolving the inflammatory process.¹⁷ In particular, M2 macrophages possess high capacities to produce SPMs due to robust expression of 15-LOX-1.^18–21 Several structurally diverse natural products, including the triterpenes acetyl-keto-β-boswellic acid (AKBA) and celastrol,^12,22 as well as benzylated chalcones and the biflavonoid 8-methylsocotrin-4’-ol act as LM class-switching molecules in macrophages promoting SPM formation.^23,24

Cannabis contains special bioactive terpenoids, collectively referred to as cannabinoids, including cannabidiol (CBD), Δ⁹-tetrahydrocannabinol (Δ⁹-THC), and cannabigerol (CBG), which displayed anti-inflammatory features in various experimental models.^25–27 In particular CBD, the major cannabinoid of fiber hemp, is devoid of psychoactive bioactivity but inhibits reactive oxygen species (ROS) formation and nuclear factor (NF)-κB signaling in microglial cells,²⁸ suppresses oxidative stress in neutrophils,²⁹ reduces tumor necrosis factor (TNF)-α production,³⁰ and modulates cytokine release in diverse cell types and in various in vivo models.^31,32 CBD may act through the endocannabinoid receptors (presumably via CB₂), by increasing anandamide signaling through inhibiting fatty acid amide hydrolase, or through PPARγ signaling, thereby conferring its impact on anxiety, epilepsy, neuropathic pain, and inflammation.^26,33–35 But how CBD and related cannabinoids affect inflammation-related LM formation is obscure. We recently reported that the cannabis-derived isoprenylated bibenzyl canniprene and the prenylflavonoids cannflavins A and B inhibit 5-LOX and microsomal PGE₂ synthase (mPGES)-1, thereby suppressing LT and PGE₂ formation.³⁶ Here, we show that CBD acts as molecular switch in innate immune cells to shift the formation of pro-inflammatory LTs and PGs to inflammation-resolving SPM, which provides a molecular basis for the anti-inflammatory properties of CBD.

Results

Cannabinoids inhibit 5-LOX and mPGES-1

To test if in analogy to canniprene and cannflavins,³⁶ cannabinoids inhibit 5-LOX and mPGES-1, eight prominent bioactive members³⁷ were screened for interference with human mPGES-1 and 5-LOX, employing well-documented cell-free assays,^38,39 as well as studying cellular 5-LOX inhibition in A23187-stimulated PMNL.⁴⁰ All tested compounds inhibited 5-LOX activity up to a concentration of 10 μM (IC₅₀ = 0.7 to 4.9 μM), while only CBG, CBGV and CBC suppressed mPGES-1 activity with IC₅₀ values <10 μM (Table 1). In A23187-stimulated PMNL, all compounds inhibited 5-LOX product formation (IC₅₀ = 1.5 to 7.6 μM), except CBD (IC₅₀ >10 μM). The 5-LOX inhibitor zileuton and the mPGES-1 inhibitor MK886 were active in these assays (not shown) as expected.^22,39 Of note, the formation of 12-LOX-derived 12-HETE and 15-LOX-derived 15-HETE in A23187-stimulated PMNL incubations were not at all suppressed by the cannabinoids. In contrast, especially CBD, CBDV and CBV caused > 2-fold increases of 15-HETE production. Together, several prominent cannabinoids act as inhibitors of 5-LOX and some of mPGES-1, supporting their anti-inflammatory features by interfering with pro-inflammatory eicosanoid formation.

Table 1. Effects of cannabinoids on the activities of isolated human recombinant 5-LOX and human mPGES-1 in microsomes as well as on 5-LOX product formation in A23187-stimulated PMNL.

IC₅₀ values (in μM) or residual activities at 10 μM (in %) are given as means ± S.E.M. of n = 3 independent experiments. CBD, cannabidiol; CBDV, cannabidivarin; CBG, cannabigerol; CBGV, cannabigerovarin; CBC, cannabichromene; CBN, cannabinol; CBV, cannabivarin; Δ⁹ -THC, tetrahydrocannabinol.

Cannabinoid	Structure	Isolated enzyme (IC50)[μM]		PMNL
		mPGES-1	5-LOX	5-LOX (IC50) [μM]	12-LOX	15-LOX
					Residual activity (in %)
CBD		> 10	4.3 ± 1.5	> 10	131 ± 18	211 ± 20
CBDV		> 10	4.9 ± 1.4	7.6 ± 1.4	110 ± 15	219 ± 7
CBG		2.8 ± 0.8	0.7 ± 0.1	2.0 ± 0.2	152 ± 20	122 ± 20
CBGV		2.5 ± 0.6	1.8 ±0.5	2.6 ± 0.4	142 ± 17	167 ± 58
CBC		6.9 ± 0.2	1.6 ± 0.3	1.5 ± 0.5	122 ± 2	144 ± 26
CBN		> 10	4.3 ± 0.2	3.4 ± 0.1	131 ± 28	146 ± 22
CBV		> 10	2.3 ± 0.4	4.3 ± 4.0	104 ± 14	279 ± 53
Δ9-THC		> 10	1.6 ± 0.2	3.4 ± 0.4	100 ± 18	115 ± 55

Cannabinoids favorably modulate LM formation in exotoxin-stimulated macrophages

We investigated the 12/15-LOX-stimulatory effects of cannabinoids in more detail. Human MDM synthesize a large array of inflammation-related LM, where M1-MDM produce mainly pro-inflammatory LTs and PGs, while M2-MDM generate substantial amounts of SPMs and their monohydroxylated precursors (i.e., 17-HDHA, 15-HETE, 15-HEPE and 14-HDHA).^18,20,21 We investigated how cannabinoids (10 μM) would modulate LM profiles of M1- and M2-MDM exposed to bacterial exotoxins (e.g., α-hemolysin) contained in Staphylococcus aureus-conditioned medium (SACM), a suitable stimulus for robust activation of broad LM biosynthetic pathways.¹⁹ Pre-treatment (15 min) of M1-MDM with CBD or CBDV prior challenge with 1% SACM for 90 min (according to our previous studies)¹⁹ increased formation of 15-LOX products but also some 12-LOX and COX products were slightly elevated, while 5-LOX products (including RvD5) were decreased by all other cannabinoids (Figure 1A). Pre-treatment of M2-MDM with the cannabinoids (10 μM), especially with CBD, increased RvD5 and all other 12- and 15-lipoxygenated LM (i.e., 17-HDHA, 15-HETE, 15-HEPE, 14-HDHA, 12-HETE, and 12-HEPE), along with suppressed 5-LOX product formation (Figure 1B). Principal component analysis (PCA) of LM formed by SACM-stimulated MDM pre-treated with or without cannabinoids revealed clear separation of the cluster of M1- from the one of M2-MDM, with most distinct stimulatory effects given by CBD in M2-MDM (Figure 1C). Analysis of the effects of the cannabinoids on defined LM in M2-MDM as representatives of 15-LOX (i.e., 17-HDHA, 15-HETE), 12-LOX (i.e., 14-HDHA, 12-HETE) and dual 5-LOX/15-LOX (i.e., RvD5) confirmed that CBD strongly upregulates these LM (Figure 1D). Notably, Δ⁹-THC, the ring-closed version of CBD, was less efficient (14-HDHA) or inactive (17-HDHA, 15-HETE, 12-HETE), and all cannabinoids, except CBDV, suppressed SACM-induced LTB₄ formation in M2-MDM (Figure 1D). Together, CBD is the most efficient cannabinoid that upregulates 12/15-LOX-mediated LM formation in exotoxin-stimulated M2-MDM, accompanied by suppression of 5-LOX activity and reflects LM class-switching features. Quantitative pathway analysis for each AA-, EPA- and DHA-derived LM formed in SACM-stimulated M2-MDM visualizes the modulatory effects of CBD and the amounts of each LM produced (Figure 2A).

Figure 1. Cannabinoids favorably modulate LM formation in exotoxin-stimulated MDM.

(A-D) Human M1- and M2-MDM were preincubated with cannabinoids (10 μM) or vehicle (DMSO, 0.1%) for 15 min before challenge with SACM (1%) for 90 min. Formed LM were quantified in the supernatants. Data with M1-MDM (A) and M2-MDM (B) are given as heatmaps presenting the -fold change to vehicle control (100%). (C) PCA of analyzed LM. Results are shown in colors corresponding to M1 or M2 polarization and labeled with CBD or vehicle treatment. (D) Representative LM formed by M2-MDMs, expressed as pg/2 × 10⁶ cells. *p < 0.05; **p < 0.01; versus vehicle. Data are means ± S.E.M., n = 3 – 5. Data were log-transformed for statistical analysis; matched one-way ANOVA with Dunnett’s multiple comparisons test. n = 3 (M1-MDM) or n = 5 (M2-MDM).

Cannabinoids induce LM formation in human macrophages

Next, we investigated if the cannabinoids could induce LM formation in MDM on their own. Since agents that act as cellular stimuli may also hamper membrane integrity, we first analyzed the impact of cannabinoids (10 μM) on the release of cytosolic LDH within 1 or 3 h and noted some interference at 3 h, but still < 40% loss of membrane integrity compared to vehicle control (Figure S1). Incubation of resting M2-MDM with cannabinoids (10 μM) for 90 min caused formation of large quantities of 12/15-LOX products and RvD5, with an increase up to > 400-fold versus vehicle-treated cells, except for CBDV and CBGV (Figure 2B). In parallel, the release of AA, EPA and DHA as LM substrates was elevated. Note that the formation of 5-LOX and COX products was considerably less affected (Figure 2B). In resting M1-MDMs, the cannabinoids essentially diminished 17-HDHA and RvD5 production, while PUFA release and 5-HETE were upregulated, implying a differential impact of cannabinoids in these macrophage subtypes (Figure S2). PCA of all analyzed LM released from cannabinoid-treated MDM indicates a separation by M1 and M2 phenotypes when cells were stimulated with cannabinoids (Figure 2C). Again, as found for SACM-stimulated MDM, CBD caused most distinct and pronounced effects on the LM metabolome in resting M1- and M2-MDM (Figure 2C). Induction of 15-LOX product formation in M2-MDM at 3 μM CBD was not yet significant (data not shown) while 10 μM was an effective concentration and thus used for subsequent experiments. Comparison of the impact on LTB₄ as pro-inflammatory and on RvD5 as proresolving representative LM, each, clearly reflects the high capacity of CBD to stimulate RvD5 formation without any induction of LTB₄ biosynthesis (Figure 2D). Furthermore, CBD did not elevate the levels of LM-relevant enzymes in M1- and M2-MDMs during LM product formation (Figure S3).

To test how CBD impacts LM formation in other immunologically relevant cells that possess high capacities for LM formation, we incubated PMNL and platelets with 10 μM CBD for 90 min and determined their LM profiles. PMNL and platelets react comparably to M2-MDM with elevated formation of 12/15-LOX products when stimulated by CBD (Figure 2E and Table 2). TXB₂ formation was strongly increased by CBD in platelets and PMNL by 23-fold, each (Table 2). PMNL preparations are frequently contaminated by platelets, expressing 12-LOX and COX-1 (here: 3% platelets and 97% PMNL, as assessed by flow cytometry), which thus could contribute to 12-HETE formation. Indeed, platelets produced large amounts of 12-HETE which was strongly enhanced (28-fold) by CBD (Table 2). Of interest, in PMNL and also in M1-MDM, CBD (10 μM) caused production of 5-LOX-derived LTB₄, although these increases are much less pronounced than formation of 15-LOX products and SPMs in M2-MDM (Figure 2E and Table 2). Together, among several cannabinoids that act as stimuli for M2-MDM to selectively induce 12/15-LOX production, CBD is the most efficient derivative, confirming its feature as a LM class-switching compound.

Table 2. CBD elicits LM formation in human cells.

M1- and M2-MDM (2 × 10⁶ cells, each), platelets (2.5 × 10⁸ cells) or PMNL (5 × 10⁶ cells) were incubated with 10 μM CBD for 90 min and LM formed were quantified in the supernatant. Data are given in pg (LM) or in ng (PUFA) as means ± S.E.M. and as fold-change (f) versus vehicle; n = 3 (M1-MDM, platelets, PMNL) or n = 5 (M2-MDM). The LOD of 3 pg/sample was taken to express the fold increase for samples where the LM was not detectable (n.d.).

		M1-MDM			M2-MDM			platelets			PMNL
		veh.	CBD	f	veh.	CBD	f	veh.	CBD	f	veh.	CBD	f
SPM	RvD5	6.3 ± 0.8	n.d.	0.5	6.5 ± 1.1	427 ± 210	66	n.d.	n.d.	1.0	5.8 ± 0.1	64 ± 46	11
	RvE4	n.d.	n.d.	1.0	n.d.	21 ± 10	7.0	n.d.	29 ± 20	10	19 ± 2.7	206 ± 146	11
	LXA4	n.d.	n.d.	1.0	n.d.	n.d.	1.0	n.d.	n.d.	1.0	8.5 ± 2.1	61 ± 30	7.2
	PDX	n.d.	n.d.	1.0	n.d.	9.3 ± 4.1	3.1	5.7 ± 1.0	35 ± 7.7	6.1	n.d.	n.d.	1.0
	MaR1	n.d.	n.d.	1.0	n.d.	n.d.	1.0	n.d.	n.d.	1.0	16 ± 1.2	37 ± 15	2.3
12-/15-LOX products	17-HDHA	116 ± 38	91 ± 25	0.8	25 ± 4.7	3710 ± 2012	148	161 ± 82	718 ± 329	4.5	22 ± 6.1	101 ± 42	4.6
	15-HEPE	15 ± 3.5	18 ± 3.2	1.2	n.d.	1778 ± 955	593	41 ± 20	291 ± 190	7.1	n.d.	69 ± 32	23
	15-HETE	106 ± 35	244 ± 59	2.3	31 ± 7.6	21124 ± 11851	681	384 ± 145	4079 ± 2370	11	38 ± 6.0	738 ± 260	19
	14-HDHA	7.0 ± 1.1	12 ± 4.2	1.7	12 ± 3.1	728 ± 412	61	6240 ± 3816	18832 ± 4944	3	34 ± 6.8	149 ± 60	4.4
	12-HEPE	n.d.	15 ± 1.1	5.0	4.9 ± 0.2	237 ± 113	48	3914 ± 3009	43389 ± 15248	11	25 ± 5.7	599 ± 303	24
	12-HETE	9.7 ± 0.3	111 ± 25	11	12 ± 1.1	2644 ± 1573	220	55931 ± 22171	1583931 ± 41360	28	404 ± 47	10476 ± 3953	26
5-LOX products	7-HDHA	19 ± 0.5	27 ± 1.1	1.4	n.d.	187 ± 76	62	19 ± 0.7	19 ± 0.1	1.0	19 ± 0.3	66 ± 44	3.5
	5-HEPE	10.0 ± 0.1	146 ± 12	15	11 ± 1.3	82 ± 21	7.5	n.d.	n.d.	1.0	47 ± 27	363 ± 304	7.7
	5-HETE	11 ± 0.6	733 ± 33	67	24 ± 5.3	605 ± 230	25	22 ± 3.2	18 ± 14	0.8	456 ± 222	2825 ± 1944	6.2
	t-LTB4	n.d.	124 ± 15	41	n.d.	45 ± 12	15	53 ± 13	580 ± 36	11	45 ± 13	1053 ± 708	23
	LTB4	n.d.	276 ± 69	92	n.d.	20 ± 1.3	6.7	n.d.	n.d.	1.0	132 ± 51	3125 ± 2018	24
COX products	PGD2	14 ± 3.0	122 ± 12	8.7	20 ± 5.7	107 ± 41	5.4	50 ± 29	986 ± 481	20	3.9 ± 1.0	88 ± 46	23
	PGE2	613 ± 156	6794 ± 523	11	49 ± 17	338 ± 94	6.9	80 ± 41	6050 ± 2812	76	17 ± 4.9	590 ± 336	35
	PGF2α	160 ± 25	547 ± 152	3.4	19 ± 5.9	146 ± 29	7.7	27 ± 16	243 ± 152	9.0	6.3 ± 0.5	38 ± 19	6.0
	TXB2	1365 ± 82	3207 ± 644	2.3	537 ± 222	2520 ± 735	4.7	4239 ± 2478	97755 ± 52378	23	175 ± 85	4073 ± 2162	23
other	13-HDHA	12 ± 1.5	32 ± 3.4	2.7	6.3 ± 0.6	38 ± 11	6.0	47 ± 31	468 ± 252	10	5.5 ± 0.4	5.4 ± 0.7	1.0
	10-HDHA	8.4 ± 0.6	17 ± 0.8	2.0	5.8 ± 0.9	29 ± 9.5	5.0	159 ± 97	396 ± 105	2.5	5.5 ± 0.8	11 ± 3.7	2.0
	4-HDHA	n.d.	24 ± 6.0	8.0	11 ± 1.2	42 ± 12	3.8	n.d.	n.d.	1.0	n.d.	n.d.	1.0
	18-HEPE	17 ± 1.9	27 ± 3.3	1.6	13 ± 1.1	36 ± 5.8	2.8	61 ± 30	389 ± 211	6.4	11 ± 1.3	21 ± 2.6	1.9
	11-HEPE	n.d.	25 ± 2.4	8.3	n.d.	29 ± 11	10	25 ± 18	565 ± 357	23	n.d.	9.4 ± 3.9	3.1
	11-HETE	16 ± 3.1	465 ± 20	29	9.2 ± 1.4	473 ± 153	51	395 ± 223	10175 ± 3340	26	7.3 ± 1.5	60 ± 21	8.2
	5,15-diHETE	n.d.	n.d.	1.0	31 ± 4.5	1193 ± 729	38	228 ± 81	6352 ± 740	28	96 ± 25	2451 ± 1584	26
		M1-MDM			M2-MDM			platelets			PMNL
		veh.	CBD	f	veh.	CBD	f	veh.	CBD	f	veh.	CBD	f
PUFA	AA	18 ± 1	817 ± 410	44	26 ± 6	1410 ± 407	53	37 ± 14	142 ± 100	3.8	29 ± 2	89 ± 5	3.0
	EPA	2 ± 0	82 ± 29	33	3 ± 1	226 ± 59	71	3 ± 1	13 ± 11	4.9	2 ± 0	5 ± 1	2.3
	DHA	19 ± 3	78 ± 18	4.1	12 ± 3	163 ± 44	13	4 ± 2	8 ± 6	2.0	8 ± 2	8 ± 2	1.0

Effects of CBD on MDM polarization, LM-enzyme expression, and cytokine release

In addition to the direct modulatory effects on LM formation and the related enzymatic pathways, CBD may also impact MDM polarization and, thus, the expression of LM-biosynthetic enzymes and of other inflammatory mediators (i.e., cytokines) under long-term exposure. Since incubation of unpolarized M0_M-CSF and M0_GM-CSF with 10 μM CBD for 48 h caused cytotoxic effects (Figure S4), we used 3 μM as a non-cytotoxic CBD concentration. CBD had no significant effect on the expression of CD80 and CD54 (M1-markers) or CD163 and CD206 (M2-markers) on the surface of MDM during polarization to the M2-phenotype (Figure 3A). Also, the capacity of MDMs to release cytokines such as IL-6, IL-10, and TNF-α during polarization was only slightly reduced by CBD (Figure 3B and S4). Moreover, no change in the protein levels of cPLA₂, COX-1, COX-2, 5-LOX, 15-LOX-1, and 15-LOX-2 was apparent between CBD- and vehicle-treated M1- and M2-MDM (Figure 3C). Nevertheless, exposure of MDM to 1% SACM upon CBD treatment slightly decreased LM release, e.g., of COX products in M2-MDM but not in M1-MDM (Figure 3D). LTB₄ in M1- or M2-MDM, however, was not altered by CBD (Figure 3D).

Figure 3. Effects of CBD on MDM polarization, LM-enzyme expression and cytokine release.

(A-D) M0_GM-CSF or M0_M-CSF were pretreated with CBD (3 μM) or vehicle (DMSO, 0.1%) for 15 min before polarization to M1-(24 h) or M2-(48 h) MDMs, respectively. (A) Expression of the M1 markers CD54 and CD80 as well as on the M2-markers CD163 and CD206 in M2-MDMs, analyzed by flow cytometry. Cell populations are represented as pseudocolor dot plots and mean fluorescence of living CD14⁺ cells. Data are means + S.E.M.; n.s., not significant (p > 0.05), ratio-paired t-test. (B) Cytokine release during M1-MDM polarization was measured by ELISA; n.s., not significant (p > 0.05), ratio-paired t-test. Data are means ± S.E.M.; n.d. not detectable. (C) Relative LM-enzyme expression on the protein level in M1- and M2-MDM was assessed by Western blot. Data are means ± S.E.M. (D) LM production after MDM polarization and subsequent stimulation with 1% SACM for 90 min was analyzed. 15-LOX products (sum of 17-HDHA, 15-HEPE and 15-HETE), 5-LOX products (sum of 7-HDHA, 5-HEPE, 5-HETE, t-LTB₄, et-LTB₄ and LTB₄) and COX products (sum of PGE₂, PGD₂, PGF_2α, TXB₂) in ng/2 × 10⁶ cells are shown as means ± S.E.M. Data were log transformed for statistical analysis. *p < 0.05; paired t-test. n = 3 (M1-MDM), n = 6 (M2-MDM).

CBD causes 15-LOX-1 activation and alters 5-LOX regiospecificity in HEK293 cells

To explore which LOX isoenzymes are affected in MDM upon stimulation with CBD, HEK293 cells stably expressing human recombinant 15-LOX-1, 15-LOX-2 or 5-LOX and FLAP were incubated with CBD (10 μM) for 3 h in the presence of exogenous PUFA to ensure substrate availability for LM production (Figure 4A). In HEK293 cells expressing 15-LOX-1, CBD increased 17-HDHA formation by 31-fold, and also the 12/15-LOX products 15-HEPE and 14-HDHA was elevated about 12-fold, while other LM were hardly affected. In comparison, in HEK293 cells expressing 15-LOX-2, the formation of 12/15-LOX products, except 15-HETE, was less elevated by CBD.

Figure 4. Molecular signaling involved in CBD-induced 15-LOX-1 activation.

(A) CBD activates 15-LOX-1 and alters 5-LOX regiospecificity. HEK293 cells (10⁶), stably transfected with LOXs, were stimulated with vehicle (0.1% DMSO) or CBD (10 μM) for 3 h in the presence of AA, EPA, and DHA (1 μM, each). LM were assessed in the supernatants. Results are presented in a heatmap as -fold change versus vehicle control (100%). (B, C) M2-MDM were incubated with (B) pyrrophenone (“cPLA2-i.”, 2 μM), or (C) AA, EPA and DHA (1 μM, “FA”) for 15 min before stimulation with 10 μM CBD for 90 min. Formed LM in the supernatants were quantified. Data of representative LM (expressed as pg/2 × 10⁶ cells) are given as means ± S.E.M.; n = 3–5; **p < 0.05, **p < 0.01, ***p < 0.001. Data were log-transformed for statistical analysis; paired one-way ANOVA with Tukey’s multiple comparisons test, n = 5 (B) or ratio paired t-test, n = 3 (C). (D, E) M2-MDM were incubated with 10 μM CBD or vehicle (DMSO, 0.1%) and the incubation was stopped after the indicated times. (D) Formed LM in the supernatants were quantified. Data are expressed as fold increase versus vehicle control, given as means ± S.E.M.; n = 3. (E) Cells were fixed, permeabilized, incubated with antibodies against 5-LOX (red) and 15-LOX-1 (cyan blue) and analyzed by immunofluorescence microscopy; scale bar = 10 μm. Results shown for one single cell are representative for approximately 100 individual cells analyzed in three independent experiments. (F) M2-MDM were stained with Ca²⁺-sensitive Fura-2/AM and stimulated with vehicle (0.1% DMSO), CBD (10 μM) or positive control (ionomycin, 2 μM). Fluorescence ratio was used to infer intracellular Ca²⁺ concentrations over time. Graphs are representatives for three individual experiments. (G) M2-MDM were incubated with 1 mM EDTA and 20 μM BAPTA/AM for 15 min before stimulation with vehicle (0.1% DMSO) or CBD (10 μM) for 90 min. Formed LM were quantified in the supernatants. Data are means ± S.E.M., n = 5. Data were log-transformed for statistical analysis; paired one-way ANOVA with Tukey’s multiple comparisons test (only the comparison between CBD with and without BAPTA/EDTA is shown). (H) M2-MDM were incubated with vehicle (0.1% DMSO) or with EDTA (1 mM) and BAPTA (20 μM) for 15 min before stimulation with vehicle (0.1% DMSO), CBD (10 μM) or SACM (1%) for 90 min. Cells were stained and analyzed by immunofluorescence microscopy; scale bar = 10 μm. Results shown for one single cell are representative for approximately 100 individual cells analyzed in n = 3 independent experiments. (I) CBD docks to 15-LOX-1. The AF2 model of human 15-LOX-1 is shown in cartoon rendering with the catalytic domain in blue, PLAT domain in grey, and orange sphere representing the catalytic iron. There were three main allosteric locations where CBD (stick rendering) docked with two positions in the interdomain groove (highlighted red or black oval with pink carbon sticks or yellow carbon sticks) and a third location above the catalytic iron (green oval with orange carbon sticks). Potential amino acids that interact with the docked CBD are highlighted in boxes. The aliphatic tail of CBD (green oval with orange sticks) is in an uncommonly restrained and bent shape.

To confirm that 15-LOX-1 mediates CBD-induced LM formation in M2-MDM, we selectively inhibited 15-LOX-1 by BLX-3887.²⁰ In fact, BLX-3887 (1 μM) strongly diminished CBD-induced RvD5 and 12/15-LOX product formation (Figure S5). The selectivity of BLX-3887 for the 15-LOX-1 isoform was confirmed in A23187-activated HEK293 cells stably expressing 15-LOX-1. As expected, BLX-3887 suppressed 12/15-LOX product formation, whereas in HEK293 expressing the 15-LOX-2 paralogue, LM formation was not decreased by BLX-3887 (Figure S5).

Molecular signaling involved in CBD-induced 15-LOX-1 activation.

We studied the underlying mechanisms of CBD-induced LM formation through 15-LOX-1 in more detail. In stimulated macrophages, the release of AA and EPA as substrates for COX/LOXs is commonly mediated by cPLA₂α.⁴¹ To investigate if cPLA₂α is involved in CBD-induced LM formation, we pre-incubated M2-MDM with the selective cPLA₂ inhibitor pyrrophenone,⁴² before stimulation with CBD. The amounts of AA, EPA, and DHA were all significantly impaired by cPLA₂ inhibition, where the release of C20 PUFAs AA and EPA was more efficiently suppressed as compared to the C22 PUFA DHA (Figure 4B). In parallel, the AA-derived 15-HETE was significantly inhibited by pyrrophenone, whereas DHA products such as RvD5 and 17-HDHA were not significantly reduced (Figure 4B). Therefore, greater substrate availability could simply account for elevated 15-LOX product formation. However, co-addition of CBD together with exogenous AA, EPA and DHA (1 μM, each) still markedly increased 15-LOX product formation (Figure 4C), suggesting that CBD also stimulates the 15-LOX-mediated conversion of exogenous PUFA to LM. Addition of PUFAs to M2-MDM in the absence of CBD hardly enhanced formation of 15-LOX products within 90 min (Figure 4C).

Besides release PUFA release, LM production depends on LOX activation and its subcellular redistribution that grants access to PUFA in activated cells.^43,44 Upon adequate stimulation of human MDM, 5-LOX and 15-LOX-1 move from soluble cell compartments to either the nuclear membrane (5-LOX) or to yet unidentified membranous structures.²¹ We studied 5-LOX and 15-LOX subcellular localization in CBD-activated M2-MDM using immunofluorescence microscopy. CBD caused translocation of cytosolic 15-LOX-1 to membranous structures starting after 15 min, with the enzyme remaining there up to 90 min (Figure 4D). The localization of nucleosolic 5-LOX did instead not change at any time point. In contrast, exposure of M2-MDM to 1% SACM caused translocation of both LOXs (Figure 4H), as expected.¹⁹ Analysis of the temporal formation of 15-LOX products in M2-MDM stimulated by CBD showed a delayed pattern, where 17-HDHA, 15-HETE and RvD5 were markedly produced only after 60 but not after 30 min (Figure 4E).

Since elevated intracellular Ca²⁺ levels ([Ca²⁺]_i) trigger LOX activation in MDM,^19,21,23 we studied the effect of CBD on [Ca²⁺]_i. Upon exposure of M2-MDM to CBD, the [Ca²⁺]_i increased, which however was much less pronounced as compared to the Ca²⁺-ionophore ionomycin (positive control) (Figure 4F). We then studied whether Ca²⁺ is per se required for the CBD-induced production of 15-LOX-derived LM by employing BAPTA-AM and/or EDTA to chelate intra- and/or extracellular Ca²⁺. After Ca²⁺ depletion of M2-MDM, both CBD-induced 15-LOX-1 translocation and 15-LOX product formation were still evident, although a tendency for lower 15-LOX product levels was obvious (Figure 4G and 4H). This contrasts the Ca²⁺-dependent induction of 15-LOX-1 translocation and product formation upon stimulation with SACM (Figure 4H).¹⁹ In addition to the rise in [Ca²⁺]_i, distinct MAPK cascades trigger LM formation, for example, by phosphorylation of serines in cPLA₂ and 5-LOX.^44,45 The p38 MAPK inhibitor skepinone-L and the ERK-1/2 activation inhibitor U-0126 failed to suppress CBD-induced LM formation in M2-MDM (Table S1), excluding the contribution of the MAPK in these processes. Moreover, the specific CB₂ receptor antagonist SR 144528 did not block CBD-induced LM formation in M2-MDM (Table S2).

Identification of putative binding sites for CBD in human 15-LOX-1 and 5-LOX by molecular docking

We employed a similar docking strategy for CBD to human LOXs as we recently used for binding celastrol to 5-LOX,²² and AKBA to an AlphaFold2 (AF2) model of human 15-LOX-1.¹² Overall, AutoDock Vina positioned CBD in the familiar allosteric location in a deep groove at the interface of the C2-like (or Polycystin-1, Lipoxygenase, Alpha-Toxin (PLAT)) and catalytic domains for both 15-LOX-1 (Figure 4I) and 5-LOX (Figure S6).¹³ Additionally, a distinct and non-overlapping docking location in this interdomain groove (Figure 4I and S6) was selected for potential binding of CBD to the LOXs. Hydrophobic burial of CBD into the interdomain groove is the predominant molecular interaction influencing the docking score with the hydroxyls from the benzenediol moiety sharing H-bonds with main chain interactions with E108 or K133 from 5-LOX (Figure S6) or with main chain interactions with H127, V105, or S107 from 15-LOX-1 (Figure 4I). A third potential docking site for CBD to 15-LOX-1 was found at an allosteric location on the catalytic domain; however, the aliphatic tail of the cannabinoid is restrained to a bent position for hydrophobic burial (Figure 4I). Overall, CBD docks to 15-LOX-1 and 5-LOX in similar locations in the interdomain cleft, which supports the validity of these potential binding sites.

CBD promotes the LM class switch in zymosan-induced peritonitis in vivo

To test if CBD-induced 15-LOX-1 activation and the shift from pro-inflammatory to proresolving LM was evident also in vivo, we explored the effects of CBD in murine zymosan-induced peritonitis as suitable animal model.¹⁹ First, we studied if 15-LOX-expressing murine peritoneal macrophages (PMs) as resident cells in the peritoneum are susceptible to CBD and generate 12/15-LOX products. In PMs, isolated from peritoneal lavage of naïve mice, CBD (10 μM) strongly elevated SPM production, particularly protectins and maresins, along with increased 15-HETE, 14-HDHA, and 12-HETE levels within 90 min (Figure 5A), whereas COX products were hardly elevated without marked LTB₄ production (Table S3).

Figure 5. CBD modulates LM formation in murine macrophages and in zymosan-induced mouse peritonitis in vivo.

(A) Macrophages were isolated form the peritoneal cavity of mice and then treated with CBD (10 μM) or vehicle (0.1% DMSO) for 90 min. Formed LM in the supernatants were analyzed. Data, expressed in pg/ 5 × 10⁵ cells, are given as means ± S.E.M., individual LM or LM groups, namely protectins (sum of PD1 and PDX) and maresins (sum of MaR1 and MaR2) are shown. Paired Student`s t-test, *p < 0.05, ***p < 0.001. (B) Experimental timeline of mouse peritonitis. Mice received CBD (10 mg/kg), indomethacin (10 mg/kg) or vehicle (0.5 mL, DMSO 2%) i.p., each, 30 min prior to induction of peritonitis by zymosan (1 mg/mouse, i.p.). After 4 h, animals were sacrificed and the exudates were obtained by peritoneal lavage. (C) Cell numbers in the exudates were counted and are shown as means ± S.E.M. of n = 6 individual animals. **p < 0.01; vs. veh.; paired one-way ANOVA with Dunnett’s multiple comparisons test. (D, E) LM were isolated from the exudates and analyzed. (D) Groups of LM in the exudates after (means of n = 6 individual animals per group) shown in a heatmap presenting the −fold change of CBD or indomethacin vs. vehicle treatment, each. 5-LOX products include 5-HETE, 5-HEPE, t-LTB₄, LTB₄; COX products include PGE₂, PGD₂, TXB₂, PGF_2α; 15-LOX products include 17-HDHA, 15-HETE, 15-HEPE; 12-LOX products include 14-HDHA, 12-HETE, 12-HEPE; SPM include PD1, PDX, RvD5, and MaR1; PUFA include AA, EPA, and DHA. (E) Selected LM expressed in pg/mL exudates, n = 6 mice per group. Data points are from individual animals. *p < 0.05; **p < 0.01, ***p < 0.001, **** p < 0.0001; versus vehicle; one-way ANOVA with Dunnett’s multiple comparisons test where possible or unpaired t-test (MaR1); n = 6 individual animals.

Next, we studied if CBD would increase SPM production also in vivo. Male CD1 mice were treated with CBD or indomethacin (10 mg/kg; i.p.) for 30 min prior to induction of peritonitis by injection of zymosan (1 mg/mouse, i.p.). After 4 h of zymosan injection, mice were sacrificed, LM were assessed in the peritoneal lavage and infiltrated leukocytes were counted (Figure 5B). Cell infiltration was significantly reduced in the peritoneal cavity of mice that were pre-treated with indomethacin but not when pre-treated with CBD (Figure 5C). Cursory analysis of produced LMs in CBD-treated mice revealed a notable increase of 12/15-LOX products and SPM, but a slight decrease in COX/5-LOX product formation and PUFA release (Figure 5D). More detailed analysis of individual LMs showed significant elevation of PD1, PDX, MaR1 and RvD5 when mice were pre-treated with CBD (Figure 5E). Changes in COX or 5-LOX product formation due to CBD treatment was not obvious, but indomethacin strongly reduced PG formation along with increased LTB₄ levels (Figure 5E), as expected. Only negligible changes in 12/15-LOX product and SPM production were observed when mice were pre-treated with indomethacin.

Discussion

By employing comprehensive LM metabololipidomics with human innate immune cells,^20,21 we found that CBD promotes the formation of SPM and related 12/15-LOX products. CBD was the most potent compound within a set of cannabinoids which included also Δ⁹-THC. SPMs are potent endogenous anti-inflammatory immunoresolvents that promote inflammation resolution, and function in host defense, pain, organ protection, and tissue remodeling.^3,6 Of interest, CBD not only upregulates the formation of SPM in activated leukocytes, but also induces their biosynthesis in resting cells. Our data propose CBD as an allosteric switch that directly activates 15-LOX-1 without stimulating other LOX isoenzymes. Conversely, 5-LOX-mediated LT formation was not stimulated but rather inhibited by CBD. Thus, CBD favorably impacts LM formation by switching from pro-inflammatory eicosanoids to inflammation-resolving SPM, which might be a plausible mode of action underlying the well-recognized anti-inflammatory properties of CBD.²⁶ In fact, CBD reduced PGE₂ levels in carrageenan- or complete Freund’s adjuvant-induced rat paw edema.^46,47 Nevertheless, detailed or mechanistic studies of the impact of CBD on SPM-biosynthetic pathways are so far missing.

Cannabis and cannabinoids, especially CBD, are well recognized anti-inflammatory agents. CBD is devoid of psychoactive effects and is the major cannabinoid of fiber hemp.⁴⁸ Several preclinical findings have suggested the potential of CBD for the management of inflammatory diseases like asthma,⁴⁹ and neuro neurodegenerative disorders with inflammatory events, such as Parkinson’s and Alzheimer’s disease.^50,51 Thus, 5-LOX-derived LTs, whose production is suppressed by CBD, are major mediators of asthma,⁴⁴ and SPM, which are upregulated by CBD, are currently evaluated for the treatment of Parkinsońs and Alzheimer’s disease.⁵² Nabiximols (an extract of cannabis with defined Δ⁹-THC and CBD (1:1) contents) is approved for reducing spasticity in multiple sclerosis patients,⁵³ but the efficacy of cannabis for other indications is still under debate, with potential uses for pain relieve and treatment of inflammatory bowel diseases.^54,55 CBD reduced inflammation in several experimental in vivo models, including experimental colitis, pulmonary inflammation, collagen-induced arthritis, hepatic ischemia-reperfusion injury, β-amyloid-induced neuroinflammation, and autoimmune encephalomyelitis.²⁶ Notably, in these types of models, SPMs displayed beneficial features leading to inflammation resolution.⁷ These findings strengthen the hypothesis that CBD may mediate its pro-resolving effectiveness by elevating SPM through activation of 15-LOX-1, demonstrated in SACM-stimulated macrophages in vitro and in zymosan-induced murine peritonitis in vivo. For our cellular studies, CBD was employed at 10 μM, which is a physiologically relevant concentration, as CBD levels in murine plasma after oral application of 115 mg/kg CBD ranged from 2 – 3.5 μg/mL (corresponding to 6 – 11 μM CBD).⁵⁶ Interestingly, CBD and CBD-containing extracts accelerated wound healing with a yet unclarified mechanistic basis;⁵⁷ elevation of SPM that are known to stimulate wound healing,⁵⁸ might be a reasonable explanation.

CBD′s anti-inflammatory effects are explained so far by suppression of ROS formation and oxidative stress, impaired NF-κB signaling with reduced COX-2 and iNOS expression, and beneficial modulation of TNFα, IL1β, IL-6, IFN-γ, and IL-2.^28–31 Several targets have been suggested for CBD, especially CB receptors, although their role in conferring CBD functions is unclear. For example, allosteric antagonistic activities of CBD on CB receptors,⁵⁹ and indirect mechanism via the release of the endogenous CB agonists arachidonylethanolamide and 2-arachidonylglycerol were proposed.^60,61 Furthermore, CBD interacts with multiple non-canonical pathways and protein targets, including adenosine A2A receptor, 5-HT_1A, TRPV1, GPR55, and PPARγ, thereby conferring its impact on anxiety, epilepsy, neuropathic pain, and inflammation.^26,33 5-LOX was proposed before as target of CBD,⁶² indicating that LOXs are involved in CBD actions, which is supported by our recent findings, adding 15-LOX-1 to the list of CBD targets, independent of the above-mentioned proteins. Notably, despite its role in biosynthesizing pro-inflammatory LTs, 5-LOX is involved in the production of RvD5 in M2-MDMs implying an anti-inflammatory role in this respect.¹⁴ The divergent expression of the LOX pathways in M1-MDM (5-LOX dominating) and M2-MDM (15-LOX-1 dominating) are seemingly causative for the differential effect of CBD on the two phenotypes.

CBD induces two different but consecutive events in SPM biosynthesis, namely (i) the release of PUFA as substrates for COXs and LOXs, and (ii) the activation of 15-LOX-1 for PUFA oxygenation. Previous reports showed that CBD and Δ⁹-THC stimulated AA release in platelets, neuroblastoma cells, and lung fibroblasts.²⁶ Also, CBD increased n-3-PUFA levels in skeletal muscle in a rat obesity model, proposed to attenuate inflammation.⁶³ These findings support elevated PUFA release in MDM by CBD, likely due to stimulation of cPLA₂ activity. However, CBD was not explored in previous studies to channel PUFAs into bioactive LM production. Interestingly, potent anti-inflammatory and pro-resolving effects of anabasum, a synthetic analog of Δ⁹-trans-THC, were reported, accompanied by reduced LTB₄ formation and elevated biosynthesis of SPM (i.e., LXA₄, LXB₄, RvD1, RvD3) in a human model of self-resolving acute inflammation.⁶⁴ Therefore, cannabinoids like Δ⁹-THC or CBD may not just simply inhibit pro-inflammatory responses but might also stimulate inflammation resolution along with tissue regeneration, observed in skeletal muscle.⁶⁵

Traditionally, inflammatory disorders are treated with anti-inflammatory drugs, where NSAIDs are most frequently used.¹⁰ NSAIDs act by lowering PG formation via inhibition of COX enzymes, with marked on-target side effects.^2,10 By blocking PG formation, NSAIDs attenuate beneficial effects of PGs on disease outcome, including pro-resolving actions of PGD₂,^66,67 and modulation of macrophage polarization towards the resolving M2 phenotype by PGE₂.⁶⁸ In addition, NSAIDs may shunt PUFA towards formation of other pro-inflammatory LMs, mainly LTs,²⁰ causing aspirin- or NSAID-exacerbated respiratory disease.⁶⁹ Compared to NSAIDs, CBD reduced LTB₄ formation in activated cells and to some extend in zymosan-induced peritonitis, while PG formation was not suppressed in either model. Interestingly, CBD markedly elevated COX-derived TXB₂ in human platelets but also in human and murine macrophages, indicating activation also of downstream enzymes, such as TXA synthase.

New discoveries have recently inspired the development of strategies to promote resolution of inflammation and tissue repair through activation of pro-resolving mechanisms using SPMs as immunoresolvents in order to circumvent inflammation-prolonging and adverse effects of NSAIDs.^3,6,11 Several experimental studies with rodents showed that inflammation resolution can be accomplished by direct SPM administration.⁷ SPMs limit excessive PMNL tissue infiltration, counter-regulate pro-inflammatory eicosanoid and cytokine release, and stimulate local macrophage-mediated efferocytosis,3 and are currently explored as pharmacological treatment in clinical studies.^6,70 However, administration of SPMs to patients is demanding in terms of adjusting the correct concentration, ensuring metabolic stability, and targeting the concrete site of inflammation, which hampers clinical applicability. Therefore, pharmacological approaches to achieve elevated SPM levels in tissues by stimulating endogenous SPM formation are attractive alternative strategies. This might be accomplished by LM “class-switching” compounds that shift the biosynthesis from PG and LTs towards SPM.¹⁴ CBD certainly fulfills the requirement of a LM “class-switching” compound.

We studied how CBD leads to 15-LOX product formation in intact cells. Agents that cause cellular LOX activation typically elicit receptor-dependent (α-hemolysin) or -independent (A23187) elevation of [Ca²⁺]_i. Ca²⁺ then binds cPLA₂ and LOXs, mediates their membrane binding, and stimulates enzymatic catalysis.^43,44 Ca²⁺ depletion failed to abolish CBD-induced LM formation in M2-MDMs implying that not only Ca²⁺-dependent group IV cPLA₂-α is operative. Although our data with pyrrophenone suggest cPLA₂-α as likely candidate, this inhibitor may affect also other PLA₂s. Note that pyrrophenone mainly suppressed AA- and EPA-derived LM but much less DHA-derived ones, supporting the involvement of another PLA₂ liberating DHA. It is known that during inflammation resolution, different PLA₂ might act with some stereospecificity on different membrane-tethered PUFAs.⁴¹ Furthermore, inhibition of 15-LOX-1 activity in M2-MDMs by BLX-3887 gave not only reduced 15-/17-hydroxylated PUFA species but also diminished 12- and 14-lipoxygenated products, implying that 15-LOX-1 generates the latter LM. Nevertheless, CBD may also activate the platelet-type 12-LOX that is abundant in platelets and produces large amounts of 12-HETE in these cells, which was strongly enhanced by CBD.

Recently, AKBA was identified as a direct 15-LOX-1 activator, operative at an allosteric site on 15-LOX-1, and markedly stimulated 15-LOX product and SPM formation in M2-MDM.¹² Our current data imply that CBD utilizes a similar mechanism to induce SPM production by allosteric 15-LOX-1 activation. Multiple potential allosteric binding locations for CBD binding to 15-LOX-1 were identified by docking studies, with the interdomain cleft between the PLAT and catalytic domains being the most probable binding site. Further docking studies of CBD to 5-LOX supported these finding with CBD overlapping with the binding site for AKBA that was solved in a co-crystal structure with 5-LOX.¹³ Thus, CBD binds to the same interdomain cleft of these LOXs like ABKA and both compounds inhibit 5-LOX¹³ but activate 15-LOX-1¹², possibly by differential allosteric modulations from formation of the substrate access portals to controlling the positioning of PUFA substrates in the active sites, which for 5-LOX might be detrimental but beneficial for 15-LOX-1. For 5-LOX, AKBA in addition shifted the regiospecificity from a 5- towards a 12-lipoxygenating enzyme¹³, which was not the case for CBD. Of note, CBD was shown by others to inhibit recombinant 15-LOX in a cell-free assay (IC₅₀ = 2.56),⁷¹ which opposes 15-LOX-1 activation by CBD in M2-MDMs. Such opposing modulation of LOXs (i.e., 5-LOX) was also found for AKBA^12,13, suggesting that the cellular environment (e.g., redox tone, membrane interactions, 15-LOX-1 post-translational modification) may shift 15-LOX-1 inhibition towards activation. Fully delineating the molecular interaction of CBD with these human LOXs will require further biophysical and structural studies, but the CBD docking sites overlap with the determined allosteric regulation site of AKBA in 5-LOX.

In summary, we show that CBD favorably modulates LM biosynthesis in inflammation-related cells and in an experimental animal model, suppressing pro-inflammatory LTs but promoting the formation of SPM, which represents a plausible mechanism underlying the anti-inflammatory properties. Our findings suggest that CBD activates 15-LOX-1, thereby inducing the LM class switch in order to accomplish the resolution of inflammation.

Limitations of the study

Although we performed comprehensive LM metabololipidomics with macrophages, glutathione-conjugated LM including cys-LT and peptide-conjugates in tissue regeneration (CTRs) of SPM, i.e., MCTRs, PCTRs and RCTRs, could not be measured by our UPLC-MS-MS method. In addition to i.p. application, CBD should be tested when given p.o. to mice, to better estimate the pharmacological relevance, and ex vivo analysis of LM formed by PMs from CBD-treated mice remains to be determined. Moreover, besides molecular docking, analysis of physical binding of CBD to 5-LOX and 15-LOX-1 proteins as co-crystal structures is desirable. Finally, the identity of the PLA₂ isozyme that liberates DHA in response to CBD is unclear and remains to be identified.

Significance

Cannabis and their bioactive ingredients, the cannabinoids, are well-recognized anti-inflammatory remedies with continuously increasing popularity for treating inflammation-related disorders including Alzheimer’s disease, inflammatory bowel diseases, and inflammatory skin disorders. Particularly, CBD has received enormous attention as the major cannabinoid of fiber hemp, devoid of psychoactive bioactivity but with considerable anti-inflammatory actions in numerous experimental in vivo models of inflammation. Although various mechanisms and molecular targets potentially underlying the anti-inflammatory features of CBD were proposed, the concrete state of knowledge in this respect is still vague. Moreover, the impact of CBD on LM that are crucial for regulating inflammation, is scarcely elucidated. Here, we established CBD as lipid mediator class-switching agent that stimulates innate immune cells for production of SPM but suppresses pro-inflammatory leukotriene biosynthesis. From a bioactivity screening of eight major cannabis-derived cannabinoids including the psychoactive Δ⁹-tetrahydrocannabinol, CBD clearly emerged as the most efficient compound, and 5-lipoxygenase and 15-lipoxygenase could be revealed as molecular CBD targets. Comprehensive metabololipidomics analysis in different human macrophage phenotypes showed that in exotoxin-stimulated cells, CBD suppresses 5-lipoxygenase-mediated leukotriene formation but upregulates formation of SPM by stimulating 15-lipoxygenase-1. Intriguingly, CBD elicits SPM production in resting cells due to allosteric 15-lipoxygenase-1 activation and enhanced substrate supply. Studies using zymosan-induced murine peritonitis confirmed increased SPM production and suppressed eicosanoid formation by CBD in vivo. Conclusively, our results reveal CBD as an allosteric switch that directly activates 15-lipoxygenase-1 for SPM production without stimulating other lipoxygenase isoenzymes; in contrast, 5-lipoxygenase-mediated leukotriene formation is inhibited by CBD. This favorable biosynthetic shift from pro-inflammatory LM to inflammation-resolving SPM might be a plausible mode of action underlying the well-recognized anti-inflammatory properties of CBD and represents a molecular strategy to accomplish a local environment that is beneficial to promote inflammation resolution.

STAR METHODS

KEY RESOURCES TABLE

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Dr. Oliver Werz (oliver.werz@uni-jena.de).

Materials availability

This study did not generate new unique reagents. Antibodies, reagents, cell lines and animals used for experiments were obtained from commercial or internal sources as reported in the Key Resources Table.

Key Resource Table

Reagent & Resource	Source	Identifier, Catalog number
Antibodies
Alexa Fluor 488 goat anti-rabbit IgG (H+L)	Thermo Fisher Scientific	Cat#A11034; RRID:AB_2576217
Alexa Fluor 555 goat anti-mouse IgG (H+L)	Thermo Fisher Scientific	Cat#A21424; RRID:AB_141780
APC anti-human CD206 (clone 19.2)	BD Biosciences	Cat#550889; RRID:AB_398476
APC-H7 anti-human CD80 (clone L307.4)	BD Biosciences	Cat#561134; RRID:AB_10565974
FITC anti-human CD14 (clone M5E2)	BD Biosciences	Cat#555397; RRID:AB_395798
IRDye 680LT goat anti-mouse	Li-Cor Biotechnology	Cat#926–68020; RRID:AB_10706161
IRDye 800CW goat anti-mouse	Li-Cor Biotechnology	Cat#926–32210; RRID:AB_621842
IRDye 800CW goat anti-rabbit	Li-Cor Biotechnology	Cat#926–32211; RRID:AB_621843
mouse monoclonal anti-15-LOX-1	Abcam	Cat#ab119774; RRID:AB_10901109
mouse monoclonal anti-5-LOX	BD Bioscience	Cat#610694; RRID:AB_2226941
mouse monoclonal anti-β-actin	Cell Signaling	Cat#3700; RRID:AB_2242334
PE anti-human CD163 (clone GHI/61)	BD Biosciences	Cat#556018; RRID:AB_396296
PE-Cy7 anti-human CD54 (clone HA58)	Biolegend	Cat#353115; RRID:AB_2715943
rabbit monoclonal anti-COX-2	Cell Signaling	Cat#12282; RRID:AB_2571729
rabbit polyclonal anti-15-LOX-2	Abcam	Cat#ab23691; RRID:AB_447612
rabbit polyclonal anti-5-LOX	Dr. Olof Radmark, Karolinska Institutet, Stockholm, Sweden	N/A
rabbit polyclonal anti-COX-1	Cell Signaling	Cat#4841; RRID:AB_2084807
rabbit polyclonal anti-cPLA2α	Cell Signaling	Cat#2832; RRID:AB_2164442
rabbit polyclonal anti-β-actin	Cell Signaling	Cat#4967; RRID:AB_330288
Bacterial and virus strains
Staphylococcus aureus 6850wt strain	kindly provided by Dr. Lorena Tuchscherr, University Hospital Jena, Germany	N/A
Chemicals, Peptides and Recombinant Proteins
11β-PGE2	Cayman Chemical	Cat#14510; CAS: 38310–90-6
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT)	Sigma	Cat#M5655; CAS: 298–93-1
A23187	Cayman	Cat#Cay11016–10; CAS: 52665–69-7
arachidonic acid	Cayman Chemical	Cat#Cay90010; CAS: 506–32-1
BAPTA AM	Cayman Chemical	Cat#15551; CAS: 126150–97-8
BLX-3887	Cayman Chemical	Cat#27391; CAS: 934758–70-0
Brain-heart-infusion broth	Sigma	Cat#53286
cannabichromene (CBC)	Appendino et al. [37]	N/A
cannabidiol (CBD)	Appendino et al. [37]	N/A
cannabidivarin (CBDV)	Appendino et al. [37]	N/A
cannabigerol (CBG)	Appendino et al. [37]	N/A
cannabigerovarin (CBGV)	Appendino et al. [37]	N/A
cannabinol (CBN)	Appendino et al. [37]	N/A
cannabivarin (CBV)	Appendino et al. [37]	N/A
d4-LTB4	Cayman Chemical	Cat#Cay320110; CAS: 124629–74-9
d4-PGE2	Cayman Chemical	Cat#Cay314010; CAS: 34210–10-1
d5-LXA4	Cayman Chemical	Cat#Cay10007737; CAS: 1622429–53-1
d5-RvD2	Cayman Chemical	Cat#Cay11184; CAS: 1881277–33-3
d8–5S-HETE	Cayman Chemical	Cat#Cay334230; CAS: 330796–62-8
d8-AA	Cayman Chemical	Cat#Cay390010; CAS: 69254–37-1
DHA	Cayman Chemical	Cat#90310; CAS: 6217–54-5
Dulbecco’s modified Eagle’s high glucose medium with glutamine (DMEM)	GE Healthcare Life Sciences	Cat#FG0435
EPA	Cayman Chemical	Cat#90110; CAS: 10417–94-4
Fetal calf serum (FCS)	Sigma	Cat#F7524
Fura-2/AM	Thermo Fisher	Cat#F1221
geneticin	Roth	Cat#2039.3; CAS: 108321–42-2
GM-CSF	Peprotech	Cat#300–23; GenPept: P04141
hygromycin B	Roth	Cat#1287.1; CAS: 31282–04-9
IFNγ	Peprotech	Cat#300–02; GenPept: P01579.1
IL-4	Peprotech	Cat#200–04; GenPept: P05112
indomethacin	Cayman Chemical	Cat#70270; CAS: 53–86-1
ionomycin	Cayman Chemical	Cat#10004974; CAS: 56092–81-0
Lipofectamine™ LTX Reagent with PLUS™ Reagent	Invitrogen	Cat#15338030
lipopolysaccharide (E. coli)	Sigma	Cat#L3129
Lymphocyte separation medium	Promocell	Cat#C-44010
M-CSF	Peprotech	Cat#300–25; GenPept: P09603
methyl formate	Sigma	Cat#291056; CAS: 107–31-3
mouse serum	Invitrogen	Cat#10410
Non-immune goat serum	Invitrogen	Cat#50–062Z
penicillin/streptomycin	Sigma	Cat#P0781
PGB1	Cayman Chemical	Cat#Cay11110; CAS: 13345–51-2
PGH2	Cayman Chemical	Cat#17020; CAS: 42935–17-1
ProLong Diamond Antifade Mountant with DAPI	Thermo Fisher Scientific	Cat#P36971
pyrrophenone	Cayman Chemical	Cat#13294; CAS: 341973–06-6
recombinant human IL-1β	Preprotech	Cat#200–01B
RPMI 1640	Sigma	Cat#R8758
skepinone-L	Cayman Chemical	Cat#16974; CAS: 1221485–83-1
SR144528	MedChemexpress	Cat#HY-13439
tetrahydrocannabinol (Δ9-THC)	Appendino et al. [37]	N/A
triton X-100	Sigma	Cat#T8787; CAS: 9036–19-5
U-0126	Cayman Chemical	Cat#70970; CAS: 109511–58-2
zymosan A (Saccharomyces cerevisiae)	Sigma	Cat#Z4250; CAS: 58856–93-2
Experimental Models: Cell lines
Human: A549 cells	ATCC	CCL-185
Human: HEK293 cells	ATCC	CRL-153
Human: HEK293_5-LOX/FLAP	Gerstmeier et al. [73]	N/A
Human: HEK293_15-LOX-1	Gilbert et al. [13]	N/A
Human: HEK293_15-LOX-2	Gilbert et al. [13]	N/A
Experimental Models: Organisms/Strains
Escherichia coli BL21 (DE3)	New England Biolabs	C2525l
Male CD-1 mice	Charles River Laboratories	N/A
Critical commercial assays
Human TNFα ELISA kit	R&D Systems	DY210
Human IL-10 ELISA kit	R&D Systems	DY217
Human IL-6 ELISA kit	R&D Systems	DY206
Zombie Aqua™ Fixable Viability Kit	Biolegend	423102
CytoTox 96®Non-Radioactive Cytotoxicity assay	Promega	G1780
Software and Algorithms
AxioVision Se64 Rel.. 4.9	Carl Zeiss	https://www.zeiss.de/mikroskopie/downloads/axiovision-downloads.html
FlowJo X Software	BD Biosciences	https://www.flowjo.com/solutions/flowjo/downloads
Analyst software 1.6.3	AB Sciex	https://sciex.com/products/software/analyst-software
GraphPad Prism 8	GraphPad Software Inc	https://www.graphpad.com/scientific-software/prism/
Odyssey 3.0 software	LI-COR	https://www.licor.com/bio/products/software/image_studio/index.html
OriginPro 2021	OriginLab Corporation	https://www.additive-net.de/de/software/produkte/originlab/originpro

Data and Code Availability

• All data reported in this paper will be shared by the lead contact upon request.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

Experimental Model and Study Participant Details

Human cells

Leukocyte concentrates derived from freshly withdrawn blood (16 I.E. heparin/mL blood) of healthy adult male and female volunteers (18 – 65 years, without details about ancestry, race or ethnicity) were provided by the Department of Transfusion Medicine at the University Hospital of Jena, Germany. The experimental procedures were approved by the local ethical committee (approval no. 5050–01/17) and were performed in accordance with the guidelines and regulations. Written informed consent was obtained from patients. According to previously published procedures,⁷² polymorphonuclear leukocytes (PMNL) and peripheral blood mononuclear cells (PBMC) were isolated by density gradient centrifugation using lymphocyte separation medium (C-44010, Promocell, Heidelberg, Germany) after sedimentation of erythrocytes by dextran. Platelet-enriched plasma was taken from the supernatant after density gradient centrifugation, diluted with PBS pH 5.9 (4:1 v/v) and centrifuged (2100 × g, 15 min, room temperature). The pelleted platelets were resuspended in a 1:1 v/v mixture of PBS pH 5.9 and NaCl solution (0.9% m/v) and washed two more times. Finally, platelets were resuspended in PBS pH 7.4 containing CaCl₂ (1 mM). To isolate monocytes, PBMC were seeded in cell culture flasks (Greiner Bio-one, Frickenhausen, Germany) in PBS pH 7.4 with CaCl₂ and MgCl₂ (Sigma-Aldrich, Steinheim, Germany). After 1 h at 37 °C and 5% CO₂ for adherence of the monocytes, the medium was discarded and replaced with RPMI 1640 (Thermo Fisher Scientific, Schwerte, Germany) containing heat-inactivated fetal calf serum (FCS, 10% v/v), penicillin (100 U/mL), streptomycin (100 μg/mL) and L-glutamine (2 mmol/L).

Transfection of HEK293 cells was performed by using pcDNA3.1 plasmids and lipofectamine according to the manufacturer’s protocol (Invitrogen, Darmstadt, Germany) and as reported before.⁷³ HEK293 cell lines stably expressing human recombinant 15-LOX-1 or 15-LOX-2 were selected using geneticin (400 μg/mL) as reported,¹³ while HEK293 cell lines stably expressing human recombinant 5-LOX and FLAP were selected using geneticin (400 μg/mL) and hygromycin B (200 μg/mL), as reported elsewhere.⁷³

Animals

Adult (6–8 weeks) male CD1 mice (Charles River, Calco, Italy) were housed at the animal care facility of the Department of Pharmacy of the University of Naples “Federico II” and kept under controlled environment (i.e., temperature 21 ± 2 °C and humidity 60 ± 10%) and provided with normal chow and water ad libitum. Prior to experiments, mice were allowed to acclimate for 4 days and were subjected to 12 h light/dark schedule. Treatments were conducted during the light phase. The experimental procedures were approved by the Italian Ministry and carried out in accordance with the EU Directive 2010/63/EU and the Italian DL 26/2014 for animal experiments and in compliance with the ARRIVE guidelines and Basel declaration including the 3 R concept.

Method Details

Cell viability assays

For analysis of the effects of test compounds on cell viability, cells were incubated with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, 5 mg/mL, 20 μL; Sigma-Aldrich, Munich, Germany) for 2 – 3 h at 37 °C (5% CO₂) in the darkness. The formazan product was solubilized with sodium dodecyl sulfate (SDS, 10% in 20 mM HCl) and the absorbance was monitored at 570 nm (Multiskan Spectrum microplate reader, Thermo Fisher Scientific, Schwerte, Germany).

For analysis of the effects of the compounds on cell integrity, the release of lactate dehydrogenase (LDH) was assessed using CytoTox 96^®Non-Radioactive Cytotoxicity assay according to the manufactureŕs (Promega, Mannheim, Germany) instructions. After treatment, cells were centrifuged at 400×g (5 min, 4 °C) and the supernatants were diluted to appropriate LDH concentrations. Then, the absorbance was measured at 490 nm using a NOVOstar microplate reader (BMG Labtechnologies GmbH, Offenburg, Germany). Cell integrity was calculated according to the manufacturés guidelines.

Purification and activity assay of mPGES-1

Interleukin (IL)-1β-stimulated A549 cell microsomes were used as source for mPGES-1 to assess its activity in a cell-free assay. After stimulation with IL-β (1 ng/mL) for 48 h, A549 cells were harvested and sonicated. As reported previously,³⁹ the resulting homogenate was subjected to differential centrifugation at 10,000 × g for 10 min at 4°C. Afterwards the supernatant was centrifuged at 174,000 × g for 1 h at 4 °C. After resuspending the microsomal fraction (pellet) in 1 mL of homogenization buffer (0.1 M potassium phosphate buffer, pH 7.4, 60 μg/mL soybean trypsin inhibitor, 1 mM phenylmethanesulfonyl fluoride, 250 mM sucrose, 2.5 mM glutathione, 1 μg/mL leupeptin), the total protein content was determined. The microsomes were subsequently resuspended in potassium phosphate buffer (0.1 M, pH 7.4) containing 2.5 mM glutathione. Cannabinoids at different concentrations or vehicle (DMSO 0.1%; vehicle control) were added and the reaction (4 °C, 100 μL total volume) was started by adding 20 μM of PGH₂. After 1 min, 100 μL of stop solution (40 mM FeCl₂, 80 mM citric acid, and 10 μM 11β-PGE₂ as internal standard) was added at 4 °C. Following purification of PGE₂ and 11β-PGE₂ by solid-phase extraction, the amounts of PGE₂ were quantified via RP-HPLC as previously described.³⁹

Determination of 5-LOX product formation in cell-free assays

Human recombinant 5-LOX was expressed in E. coli BL21 (DE3) transformed with pT35-LO plasmid and purified using affinity chromatography on an ATP-agarose column as previously published.⁴⁰ 5-LOX (0.5 μg) was preincubated with cannabinoids at different concentrations or vehicle (0.1% DMSO) in 1 mL PBS pH 7.4 containing EDTA (1 mM). After 15 min, samples were pre-warmed for 30 s at 37 °C before CaCl₂ (2 mM) and AA (20 μM) were added to initiate 5-LOX product formation. After 10 min at 37 °C, 1 mL of ice-cold methanol containing PGB₁ (200 ng) as standard and 530 μL PBS containing 0.06 M HCl were added and 5-LOX products were extracted via solid-phase extraction.³⁸ The resulting methanol-eluate was analyzed for all-trans isomers of LTB₄ and 5-H(p)ETE by RP-HPLC utilizing a C-18 Radial-PAK column (Waters, Eschborn, Germany) as previously reported.³⁸

Determination of LM formation in human PMNL

Following isolation, PMNL were immediately resuspended in PBS pH 7.4 containing glucose (1 mg/mL) and CaCl₂ (1 mM) at a density of 5 × 10⁶ cells/mL. To determine the influence of cannabinoids on 5-LOX product formation in PMNL, cells were pre-treated with cannabinoids at different concentrations or with vehicle (DMSO; 0.1%) for 15 min at 4 °C and afterwards stimulated to produce LMs by addition of A23187 (2.5 μM) at 37 °C. The reaction was stopped by adding 1 mL of ice-cold methanol containing PGB₁ (200 ng) as internal standard. After centrifugation at 2000 × g for 10 min, solid phase extraction and RP-HPLC-analysis of 5-LOX products (LTB₄ and its all-trans isomers and 5-H(p)ETE) as well as 12-HETE and 15-HETE was carried out as described above. 20-OH-LTB₄ levels were below the limit of detection. PMNL preparations are often contaminated by platelets that express platelet-type 12-LOX and COX-1 and thus, contribute to 12-HETE and PG/TX formation. Analysis of the PMNL fraction by flow cytometry for (CD41-positive) platelets revealed that only 3% of the whole cell population were platelets and 97% PMNL.

To determine the stimulatory potential of CBD, PMNL (resuspended in PBS pH 7.4 containing glucose (1 mg/mL) and CaCl₂ (1 mM) at a density of 5 × 10⁶ cells/mL) were treated with CBD (10 μM) or Staphylococcus aureus-conditioned medium (SACM) (1%) at 37 °C for 90 min. SACM was produced as previously described.¹⁹ In brief, to prepare SACM, bacteria (methicillin-susceptible S. aureus strain 6850) were grown for 24 h in medium, diluted to an OD600nm of 0.05 and grown for another 24 h, pelleted for 5 min at 3350×g, and sterile-filtered through a Millex-GP filter unit (0.22 μm; Millipore) prior to use. The reaction was stopped by addition of 2 mL of ice-cold methanol containing deuterated LM standards (200 nM d8–5S-HETE, d4-LTB₄, d5-LXA₄, d5-RvD2, d4-PGE₂ and 10 μM d8-AA; Cayman Chemical/Biomol GmbH, Hamburg, Germany), and samples were processed for LM analysis using UPLC-MS-MS as described below.

Determination of LM formation in human platelets

Freshly isolated platelets (250 × 10⁶ cells/mL) were resuspended in PBS pH 7.4 (1 mL) containing CaCl₂ (1 mM). Vehicle (DMSO; 0.1%), CBD (10 μM) or SACM (1%) was added and cells were incubated for 90 min at 37 °C. The reaction was stopped by adding the mixture into 2 mL ice-cold methanol containing deuterated LM standards as described above.

Generation of human MDM and incubation for LM formation

Differentiation of monocytes to macrophages and their polarization to M1- and M2-like macrophage phenotypes was carried out as recently described.²¹ Briefly, PBMC were incubated with either 20 ng/mL GM-CSF or M-CSF (Cell Guidance Systems Ltd., Cambridge, UK) for 6 days in RPMI 1640 supplemented with FCS, L-glutamine, penicillin, and streptomycin as described above. This yielded M0_GM-CSF and M0_M-CSF monocyte-derived macrophages (MDM) respectively. Afterwards, LPS (100 ng/mL) and IFN-γ (20 ng/mL; Peprotech, Hamburg, Germany) were added for 24 h to obtain M1-MDM, whereas IL-4 (20 ng/mL; Peprotech) was added to generate M2-MDM within 48 h.

To study if cannabinoids can induce LM formation, M1- or M2-MDM (2 × 10⁶/mL) were incubated with vehicle (DMSO 0.1%), cannabinoids (10 μM), or SACM (1%) in PBS containing 1 mM CaCl₂ for 90 min at 37 °C and 5% CO₂. Afterwards, the reaction was stopped by adding 2 mL of ice-cold methanol containing deuterated LM standards (200 nM d8–5S-HETE, d4-LTB₄, d5-LXA₄, d5-RvD2, d4-PGE₂ and 10 μM d8-AA; Cayman Chemical/Biomol GmbH, Hamburg, Germany), samples were then processed for LM analysis using UPLC-MS-MS as described below. To study if cannabinoids modulate agonist-induced LM formation, M2-like MDM (2 × 10⁶/mL) were pre-treated with cannabinoids (10 μM) under the conditions specified above for 15 min prior to incubation with SACM (1%) for 90 min. The incubations were stopped as described above. To investigate the influence of different compounds on the effect of CBD, M2-MDM (10⁶/mL) were pretreated with pyrrophenone (2 μM; Cayman Chemical/Biomol GmbH), BLX-3887 (1 μM; Cayman Chemical/Biomol), skepinone-L (3 μM; Cayman Chemical/Biomol), U-0126 (10 μM; Cayman Chemical/Biomol), BAPTA (20 μM; Cayman Chemical/Biomol), EDTA (1 mM), or a mixture of AA, EPA, and DHA (1 μM; Cayman Chemical/Biomol GmbH, Hamburg, Germany) for 15 min prior to incubation with CBD for 90 min. The reaction was stopped as indicated above.

To assess long-term effects of CBD on LM formation during and after macrophage polarization, M0_GM-CSF and M0_M-CSF MDM were pre-treated with CBD (3 μM) for 15 min before addition of LPS and IFNγ for polarization to M1-MDM or addition of IL-4 for polarization to M2-MDM. After 48 h, the supernatants of the cultures were added to 2 mL ice-cold methanol containing deuterated standards as indicated above. RPMI was replaced with PBS containing CaCl₂ (1 mM) and SACM was added (1%). After 90 min, the reaction was stopped and samples were processed for LM analysis using UPLC-MS-MS as described below.

Generation of murine peritoneal macrophages (PMs) incubation for LM formation

Resident PMs were obtained after lavage of the peritoneal cavity of the mice with 7 mL of cold Dulbecco’s modified Eagle’s medium (DMEM) containing 5 U/mL heparin. PMs were centrifuged at 500 × g at 4 °C for 5 min and incubated for 24 h in RPMI 1640 supplemented with FCS, L-glutamine, penicillin, and streptomycin as described above at 37 °C prior to treatment. Cells (5 × 10⁵) were treated with CBD (10 μM) or vehicle (DMSO; 0.1%) in PBS pH 7.4 containing CaCl₂ (1 mM) for 90 min at 37 °C. Supernatants were collected and frozen at −80 °C. Formed LM were isolated by SPE and analyzed by UPLC-MS-MS as described below.

Zymosan-induced peritonitis in mice

Peritonitis in mice was induced as described before.⁷⁴ Mice (n = 6/group) received intraperitoneally (i.p.) CBD (10 mg/kg), indomethacin (10 mg/kg), or vehicle (2% DMSO in saline) in 0.5 mL saline/mouse 30 min prior to induction of peritonitis by injection of zymosan (1 mg/mouse in 0.5 mL saline, i.p.). After 4 h of peritonitis, mice were euthanized in a saturated CO₂ atmosphere and peritoneal lavage was performed by washing the peritoneal cavity with 3 mL ice-cold PBS and subsequent centrifugation (18,000×g, 5 min, 4 °C) immediately after killing mice with CO₂. Cells in the lavage were counted using a light microscope in a Burker’s chamber after vital trypan blue staining. Cell-free supernatants of the lavage was immediately frozen and stored for the analysis of LM levels via UPLC-MS-MS as described below.

LM metabololipidomics by UPLC-MS-MS

Samples obtained from MDM, PMs, platelets, and PMNL containing deuterated LM standards were kept at −20 °C for at least 60 min to allow protein precipitation. The extraction of LM was performed as recently published.⁷² In brief, after centrifugation (1200 × g; 4 °C; 10 min) acidified H₂O (9 mL; final pH = 3.5) was added and samples were extracted on solid phase cartridges (Sep-Pak^® Vac 6cc 500 mg/6 mL C18; Waters, Milford, MA, USA). Samples were loaded on the cartridges after equilibration with methanol followed by H₂O. After washing with H₂O and n-hexane, samples were eluted with methyl formate (6 mL). The solvent was fully evaporated using an evaporation system (TurboVap LV, Biotage, Uppsala, Sweden) and the residue was resuspended in 150 μL methanol/water (1:1, v/v) for UPLC-MS-MS analysis. LM were analyzed with an Acquity^™ UPLC system (Waters, Milford, MA, USA) and a QTRAP 5500 Mass Spectrometer (ABSciex, Darmstadt, Germany) equipped with a Turbo V^™ Source and electrospray ionization. LM were eluted using an ACQUITY UPLC^® BEH C18 column (1.7 μm, 2.1 mm × 100 mm; Waters, Eschborn, Germany) heated at 50 °C with a flow rate of 0.3 mL/min and a mobile phase consisting of methanol-water-acetic acid at a ratio of 42:58:0.01 (v/v/v) that was ramped to 86:14:0.01 (v/v/v) over 12.5 min and then to 98:2:0.01 (v/v/v) for 3 min.²⁰ The QTRAP 5500 was run in negative ionization mode using scheduled multiple reaction monitoring (MRM) coupled with information-dependent acquisition. The scheduled MRM window was 60 s, optimized LM parameters were adopted,²⁰ with a curtain gas pressure of 35 psi. The retention time and at least six diagnostic ions for each LM were confirmed by means of an external standard for each and every LM (Cayman Chemical/Biomol GmbH). Quantification was achieved by calibration curves for each LM. Linear calibration curves were obtained for each LM and gave r² values of 0.998 or higher. The limit of detection for each targeted LM was determined as described.²⁰ For UPLC-MS-MS analysis, the quantification limit was 3 pg LM in reconstituted methanol/water after SPE, calculated for the entire sample/incubation, and this value was taken to express the fold increase for samples where the LM was not detectable (n.d.). Glutathione-conjugated LM including cys-LT as well as peptide-conjugates in tissue regeneration (CTRs) of SPM could not be determined by this method. Representative MS spectra and chromatograms of RvD5 from CBD-treated M2-MDMs in comparison to a commercial RvD5 standard are shown Figure S7.

Determination of intracellular Ca²⁺ levels

M2- MDM (2 × 10⁶ cells/mL) were pre-stained with Fura-2/AM (1 μM; Thermo Fisher Scientific) for 30 at 37 °C in the dark, respectively. Cells were resuspended in modified Krebs-HEPES buffer (135 mM NaCl, 5 mM KCl, 1 mM MgSO₄ × 7 H₂O, 0.4 mM KH₂PO₄, 5.5 mM glucose, 20 mM HEPES and 0.1% BSA), which was adjusted to pH 7.4. CaCl₂ (1 mM) was added and M2-like MDM (2.5 × 10⁵ cells/ 200 μL) were seeded into 96-well plates, and then CBD (10 μM), positive controls (ionomycin, 2 μM; Cayman Chemical/Biomol GmbH) or vehicle (1% DMSO) were added, and the signal was monitored in a thermally (37 °C) controlled NOVOstar microplate reader (BMG Labtechnologies GmbH); emission at 510 nm, excitation at 340 nm (Ca²⁺-bound Fura-2) and 380 nm (free Fura-2). After cell lysis with triton X-100 (1%), the maximal fluorescence signals were monitored (= 100%) and after chelating Ca²⁺ with 20 mM EDTA, the minimal fluorescence signals (= 0%) were recorded. The ratio of the emission signal at 340 nm excitation and the emission signal at 380 nm excitation was used to determine intracellular Ca²⁺ levels.

Docking of CBD into human 15-LOX-1 and 5-LOX

CBD was downloaded from pubchem⁷⁵ and restraints were generated with eLBOW (electronic Ligand Building and Optimization Workbench)⁷⁶ inside the Phenix suite.⁷⁷ The AlphaFold2 (AF2)⁷⁸ model for human 15-LOX-1 was downloaded from the EMBL repository for folded structures.⁷⁹ As shown previously for docking AKBA to AF2 model of 15-LOX-1,¹² we switched the rotamers of K126, E130, and E134 to the 3^rd most populated states to allow for more accessibility to the interdomain cleft. We used the model of Stable-5-LOX bound to AKBA (PDB ID: 6NCF)¹³ for docking CBD to human 5-LOX. We utilized AutoDock Vina version 1.1.2 (https://doi.org/10.1002/jcc.21334) via UCSF Chimera (https://doi.org/10.1002/jcc.20084) for docking and visualization of the molecules. Search volumes including the whole protein and interdomain cleft between the PLAT and catalytic body were used to locate potential binding sites for CBD to 15-LOX-1 and 5-LOX. Docking was performed in triplicate with different starting search volumes to validate the reproducibility of the positions. Three allosteric pockets were identified for potential binding poses of CBD to either 5-LOX or 15-LOX-1 with two of the locations found in the interdomain cleft. The reproducibility of the binding poses from the different docking sessions and from related human LOXs support the validity of these docking positions.

Subcellular localization of 5-LOX and 15-LOX-1 by immunofluorescence microscopy

M0_M-CSF MDM (0.5 ×10⁶ cells) were seeded onto glass coverslips in 12-well plates and polarized to M2- MDM for 48 h. Cells were then washed and stimulated in PBS pH 7.4 containing CaCl₂ (1 mM) with vehicle (DMSO; 0.1%), CBD (10 μM) or SACM (1%; as positive control)¹⁹ for 90 min at 37 °C and 5% CO₂. Cells were subsequently fixed using 4% paraformaldehyde solution in PBS. Permeabilization was achieved by adding acetone (3 min; 4 °C) and then triton X-100 (0.25% solution; 10 min; room temperature). Following permeabilization, samples were blocked with normal goat serum (10%, 50062Z, Thermo Fisher Scientific). Samples were incubated with mouse monoclonal anti-15-LOX-1 antibody, 1:200 (ab119774, Abcam, Cambridge, UK) and rabbit anti-5-LOX antibody, 1:100 (1550 AK6, provided by Dr. Olof Rådmark, Karolinska Institutet, Stockholm, Sweden) overnight at 4 °C. 5-LOX and 15-LOX-1 were stained with the fluorophore-labeled secondary antibodies; Alexa Fluor 488 goat anti-rabbit IgG (H + L), 1:500 (A11034, Thermo Fisher Scientific) and Alexa Fluor 488 goat anti-rabbit IgG (H + L), 1:500 (A11034, Thermo Fisher Scientific) and Alexa Fluor 555 goat anti-mouse IgG (H + L); 1:500 (A21424, Thermo Fisher Scientific). Nuclear DNA was stained with ProLong Diamond Antifade Mountant with DAPI (P36971, Thermo Fisher Scientific). Samples were analyzed using a Zeiss Axiovert 200 M microscope, and a Plan Neofluar ×40/1.30 Oil (DIC III) objective (Carl Zeiss, Jena, Germany). An AxioCam MR camera (Carl Zeiss) was used for image acquisition.

Determination of protein expression by SDS-PAGE and Western blot

M0_GM-CSF and M0_M-CSF MDM were treated with DMSO (0.1%) or CBD (3 μM) for 15 min prior to addition of respective polarizing agents for 24 h (M1-MDM) or 48 h (M2-MDM) at 37 °C and 5% CO₂ and lysed on ice with Seaman lysis buffer containing protease inhibitors (1 mM phenylmethanesulfonylfluoride, 60 μg/mL soybean trypsin inhibitor, 10 μg/mL leupeptin). Then, lysates were centrifuged (15,000 rpm, 5 min, 4 °C), cell supernatants were collected, and the protein concentration was determined by DC-protein assay kit (Bio-Rad Laboratories GmbH, Munich, Germany). After addition of 4×SDS loading buffer (50 mM Tris-HCl, pH 6.8, 2% (w/v) SDS, 10% (v/v) glycerol, 1% (v/v) β-mercaptoethanol, 12.5 mM EDTA, 0.02% (w/v) bromophenol blue), the lysates were heated at 95 °C for 5 min. Equal aliquots were separated on 10% (for 15-LOX-1, cPLA_2α and COX-2), 16% (for COX-1) SDS-PAGE gels and then blotted onto nitrocellulose membranes (Amersham Protran Supported 0.45 μm nitrocellulose, GE Healthcare, Freiburg, Germany).

The membranes were incubated with the following primary antibodies: rabbit polyclonal anticPLA2α,1:1000 (2832S; Cell Signaling Technology); mouse monoclonal anti-15-LOX-1, 1:1000 (ab119774; Abcam, Cambridge, UK); rabbit polyclonal anti-COX-1, 1:1000 (4841S; Cell Signaling Technology); rabbit monoclonal anti-COX-2, 1:1000 (12282S; Cell Signaling Technology); mouse monoclonal anti-β-actin, 1:1000 (3700S; Cell Signaling). Immunoreactive bands were stained with IRDye 800CW Goat anti-Mouse IgG (H + L), 1:10,000 (926–32210, LI-COR Biosciences, Lincoln, NE), IRDye 800CW Goat anti-Rabbit IgG (H + L), 1:15,000 (926 32211, LI-COR Biosciences) and/or IRDye 680LT Goat anti-Mouse IgG (H + L), 1:40,000 (926–68020, LI-COR Biosciences), and visualized by an Odyssey infrared imager (LI-COR Biosciences, Lincoln, NE, USA). Data from densitometric analysis were background corrected.

Determination of cytokine production

M0_GM-CSF and M0_M-CSF MDM were pre-treated with CBD (3 μM) or vehicle (DMSO; 0.1%) for 15 min before addition of LPS and IFNγ (for M1-MDM) or IL-4 (for M2-MDM) for polarization. After 48 h, supernatants of the cultures were collected by centrifugation (2000× g, 4 °C, 10 min). The cytokines IL-1β, IL-10 and TNF-α were analyzed by in-house–made ELISA kits (R&D Systems, Bio-Techne, Abingdon, UK).

Quantification of macrophage surface marker expression by flow cytometry

M0_M-CSF MDM were pretreated with CBD (3 μM) or vehicle (DMSO; 0.1%) for 15 min before addition of IL-4 (20 ng/mL) for polarization towards M2-MDM. After 48 h cells were detached and stained in PBS pH 7.4 containing 0.5% BSA, 2 mM EDTA and 0.1% sodium azide by Zombie Aqua^™ Fixable Viability Kit (Biolegend, San Diego, CA, USA) for 5 min at 4 °C to determine cell viability. Non-specific antibody binding was blocked by using mouse serum (10 min at 4 °C) prior to staining by the following fluorochrome-labelled antibodies (20 min, 4 °C): FITC anti-human CD14 (clone M5E2, #555397, BD Biosciences, San Jose, CA, USA), APC-H7 anti-human CD80 (clone L307.4, #561134, BD Biosciences), PE-Cy7 anti-human CD54 (clone HA58, #353115, Biolegend), PE anti-human CD163 (clone GHI/61, #556018, BD Biosciences), APC anti-human CD206 (clone 19.2, #550889, BD Biosciences) to determine M1- and M2-MDM surface marker expression using LSRFortessa^™ cell analyzer (BD Biosciences), and data were analyzed using FlowJo X Software (BD Biosciences).

QUANTIFICATION AND STATISTICAL ANALYSIS

Results are expressed as means ± standard error of the mean (SEM) of n observations, where n represents the number of experiments with separate donors, performed on different days, as indicated. For the in vivo studies, n represents the number of animals. Neither were the sample sizes pre-determined by statistical methods, samples blinded, nor data confirmed as normally distributed. Analyses of data were conducted using GraphPad Prism 8 software (San Diego, CA). Two-tailed t test was used for comparison of two groups. For multiple comparison, one-way analysis of variance (ANOVA) with Dunnett′s or Tukey’s or post hoc tests were applied as indicated. The criterion for statistical significance is p < 0.05.

Supplementary Material

Supplemental information includes seven figures, three tables, and four supplemental items and can be found with this article online.

Highlights:

• Anti-inflammatory CBD from cannabis induces a lipid mediator class switch

• CBD elicits 15-lipoxgenase product formation in resting human macrophages

• CBD seemingly activates 15-lipoxygenase-1 via an allosteric site

• In zymosan-induced murine peritonitis CBD increases SPM production in vivo