Cross-Talk between the (Endo)Cannabinoid and Renin-Angiotensin Systems: Basic Evidence and Potential Therapeutic Significance

This review is dedicated to the cross-talk between the (endo)cannabinoid and renin angiotensin systems (RAS). Activation of AT1 receptors (AT1Rs) by angiotensin II (Ang II) can release endocannabinoids that, by acting at cannabinoid CB1 receptors (CB1Rs), modify the response to AT1R stimulation. CB1R blockade may enhance AT1R-mediated responses (mainly vasoconstrictor effects) or reduce them (mainly central nervous system-mediated effects). The final effects depend on whether stimulation of CB1Rs and AT1Rs induces opposite or the same effects. Second, CB1R blockade may diminish AT1R levels. Third, phytocannabinoids modulate angiotensin-converting enzyme-2. Additional studies are required to clarify (1) the existence of a cross-talk between the protective axis of the RAS (Ang II—AT2 receptor system or angiotensin 1-7—Mas receptor system) with components of the endocannabinoid system, (2) the influence of Ang II on constituents of the endocannabinoid system and (3) the (patho)physiological significance of AT1R-CB1R heteromerization. As a therapeutic consequence, CB1R antagonists may influence effects elicited by the activation or blockade of the RAS; phytocannabinoids may be useful as adjuvant therapy against COVID-19; single drugs acting on the (endo)cannabinoid system (cannabidiol) and the RAS (telmisartan) may show pharmacokinetic interactions since they are substrates of the same metabolizing enzyme of the transport mechanism.


Introduction
In recent years more and more publications have appeared regarding the cross-talk between the (endo)cannabinoid and renin angiotensin (RAS) systems. This issue has become even more important in the era of the coronavirus disease 2019 (COVID-19) pandemic since angiotensin-converting enzyme type 2 (ACE2), one of the key RAS components, was described as a receptor required for the entry of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) into host cells [1][2][3][4]. The primary aim of our review is to compare all publications regarding the interaction between (endo)cannabinoids and RAS and to find as many common effects and mechanisms as possible. The secondary aim is dedicated to the determination of its potential therapeutic significance.

Materials and Methods
To find the most relevant articles dealing with the interaction between angiotensin and cannabinoid systems, we performed a comprehensive search in the PubMed database and cannabinoid systems, we performed a comprehensive search in the PubMed database (closed on 20 May 2022). The following key phrases were used in the search engine: "angiotensin cannabinoid" (which yielded 94 results), "angiotensin 1-7 cannabinoid" (4 results), "ACE cannabinoid" (15 results), "ACE2 cannabinoid" (13 results), "ACE2 cannabinoid COVID" (11 results) and "cannabinoid SARS-CoV-2" (68 results). Taking approved therapeutic usage into consideration, search phrases also consisted of: "sativex", "nabilone", "dronabinol", and "epidiolex" coupled with the word "angiotensin" (a total of 13 results).
First, the titles and then the abstracts and full texts of the identified papers were analyzed, and duplicated articles or those with non-relevant content were removed from the list. Only articles in English were considered. Non-relevant studies included, for example, no described interaction between the two systems of interest or "ACE" being an acronym for unrelated definitions such as "central amygdala". Finally, 43 publications were included in this review, which can be seen systematically summarized in Tables 1-4. Additional papers listed in the References were used to provide more background on angiotensin and cannabinoid systems, as well as SARS-CoV-2. The screening process is summarized in Figure 1, illustrating the PRISMA flowchart [5]. To make reading fluent, the current article is structured into paragraphs, first describing the cannabinoid and angiotensin systems (Sections 3.1.1 and 3.1.2) and then the interaction between both systems (Sections 3.2-3.5). Interactions are being described in the following order: overall cross-talk between RAS and eCBS (Section 3.2 and Table 2), the in- . Green circles with a plus sign describe agonism at the respective receptor; red circles with a minus sign describe antagonism, inv agonism, or inhibition at the respective mechanism. Synthetic, plant-derived compounds a docannabinoids are written in black, green, and brown font, respectively; the blue font is f zyme inhibitors. Up arrows, increase; down arrows, decrease; green arrows, desired effect arrows, undesired effects. 1st gen., first-generation antagonists; 2nd gen., second-generatio tagonists; * weak affinity. ACEA, Arachidonyl-2'-chloroethylamide; CBD, cannabidiol; CB nabigerol; CBN, cannabinol; DAGs, diacylglycerols; THC, Δ 9 -tetrahydrocannabinol; THCV hydrocannabivarin; THL, tetrahydrolipstatin.

Renin-Angiotensin System
The RAS plays a significant role in the regulation of blood pressure and sodium and water homeostasis and occurs as (1) circulating hormonal system, (2) local or paracrine RAS expressed in many organs (both in the CNS and in the periphery), and (3) intracellular or intracrine RAS in several cell types [25,26]. The RAS comprises different peptides with opposing biological effects ( Figure 3). Briefly, angiotensinogen is converted into angiotensin I (Ang I) by renin and subsequently into angiotensin II (Ang II) by the angiotensin-converting enzyme (ACE). Ubiquitous actions of the main effector of the RAS, namely Ang II, are related to the activation of several signal transduction pathways, which are predominantly attributed to the Ang II type 1 receptor (AT 1 R). In addition, Ang II also acts through the Ang II type 2 receptor (AT 2 R), or is degraded by ACE2 to Ang 1-7 acting via the Mas receptor (MasR) [27,28]. All receptors are GPCRs; for details of the most important signaling pathways of AT 1 Rs, see Section 3.2.1.
As shown in Figure 3, the RAS consists of two counter-regulatory arms: the classic one (red rectangle; deleterious axis) containing ACE/Ang II/AT 1 Rs and the alternative one (green rectangles; beneficial axis) constituted by (i) AT 2 Rs and (ii) ACE2/angiotensin 1-7 and its MasRs. The effects following activation of the particular receptors are shown in Figure 3. The increase in the activity of one axis results in the decrease of the other [26,29,30]. Importantly, disturbance of the balance between the deleterious and the protective axis with the predominance of the ACE/Ang II/AT 1 R axis leads to many pathologies, including metabolic (diabetes mellitus), renal, cardiovascular (heart failure, myocardial infarction, hypertension), lung, hepatic, digestive, endocrine, neurodegenerative, hematological, reproductive, and muscular disease [4,26,[30][31][32]. Even if the classic arm plays an important and unpleasant role in many disease states of humans, it is not per se deleterious. One should consider that well-being and even survival would not be possible if this arm were missing in the presence of severe loss of fluid and sodium and/or marked hypotension.
The RAS is the target of several (classes of) drugs ( Figure 3, Table 1). Angiotensin II itself is used for the treatment of septic shock. Antagonists at the different levels of the RAS system, i.e., renin, angiotensin II and aldosterone, are used to treat essential hypertension and/or congestive heart failure. For the latter two indications, inhibitors of the degradation of angiotensin I to angiotensin II (ACE inhibitors) are important drugs.
Ang II, are related to the activation of several signal transduction pathways, wh predominantly attributed to the Ang II type 1 receptor (AT1R). In addition, Ang acts through the Ang II type 2 receptor (AT2R), or is degraded by ACE2 to Ang 1via the Mas receptor (MasR) [27,28]. All receptors are GPCRs; for details of the m portant signaling pathways of AT1Rs, see Section 3.2.1. Simplified diagram of the renin-angiotensin system (RAS) and modifying drugs tioned in this review. The RAS consists of two axes that counteract each other: a deleteriou (so-called classic; red rectangle) containing angiotensin-converting enzyme (ACE)/angiote (Ang II)/Ang II type 1 receptors (AT1Rs) and a protective one (so-called alternative; green r gles) constituted by (i) Ang II type 2 receptors (AT2Rs) and (ii) angiotensin-converting enz type 2 (ACE2)/angiotensin 1-7) and its Mas receptors (MasR). VP, vasopressin. Red circles minus signs describe antagonism, inverse agonism, or inhibition at the respective mechani arrows, increase; down arrows, decrease; green arrows, desired effects; red arrows, undesi effects. Figure 3, the RAS consists of two counter-regulatory arms: the one (red rectangle; deleterious axis) containing ACE/Ang II/AT1Rs and the alterna (green rectangles; beneficial axis) constituted by (i) AT2Rs and (ii) ACE2/angiote and its MasRs. The effects following activation of the particular receptors are sh Figure 3. The increase in the activity of one axis results in the decrease of th [26,29,30]. Importantly, disturbance of the balance between the deleterious and . Simplified diagram of the renin-angiotensin system (RAS) and modifying drugs mentioned in this review. The RAS consists of two axes that counteract each other: A deleterious one (so-called classic; red rectangle) containing angiotensin-converting enzyme (ACE)/angiotensin II (Ang II)/Ang II type 1 receptors (AT 1 Rs) and a protective one (so-called alternative; green rectangles) constituted by (i) Ang II type 2 receptors (AT 2 Rs) and (ii) angiotensin-converting enzyme type 2 (ACE2)/angiotensin 1-7) and its Mas receptors (MasR). VP, vasopressin. Red circles with minus signs describe antagonism, inverse agonism, or inhibition at the respective mechanism. Up arrows, increase; down arrows, decrease; green arrows, desired effects; red arrows, undesired effects.

Examples of Cross-Talk between the (Endo) Cannabinoid and Renin-Angiotensin Systems
Details of particular publications regarding the cross-talk between the (endo)cannabinoid and renin-angiotensin systems are shown in Table 2. We have concentrated mainly on changes in components of both systems but not on physiological or pathophysiological effects induced by activation of the individual receptors. For a description of all compounds mentioned in our review, see Figures 2 and 3.

Angiotensin II Causes Transactivation of Cannabinoid-Mediated Effects
The AT 1 Rs for Ang II interact with various G proteins, including G q/11 [53,65,66]. Activation of G q/11 protein-coupled receptors and the subsequent stimulation of phospholipase C leads to Ca 2+ signal generation and indirectly causes rapid biosynthesis of eCBs (mainly 2-AG) by DAGL activation (Figure 4). As mentioned above, 2-AG exerts its effects mainly via CB 1 Rs and CB 2 Rs [8]. This reasoning suggests that Ang II might cause transactivation of cannabinoid-mediated effects.  [53,65,66]. Activation of Gq/11 protein-coupled receptors and the subsequent stimulation of phospholipase C leads to Ca 2+ signal generation and indirectly causes rapid biosynthesis of eCBs (mainly 2-AG) by DAGL activation (Figure 4). As mentioned above, 2-AG exerts its effects mainly via CB1Rs and CB2Rs [8]. This reasoning suggests that Ang II might cause transactivation of cannabinoid-mediated effects.

Figure 4.
Influence of CB1 receptor (CB1R) antagonists on the effects elicited by activation of AT1 receptors (AT1R). Stimulation of AT1Rs leads to Ca 2+ -signal generation and rapid biosynthesis of 2arachidonoylglycerol (2-AG) from diacylglycerols (DAGs) by diacylglycerol lipase (DAGL) activation. If CB1R and AT1R activation lead to the same effect, CB1R antagonism will inhibit the AT1Rmediated effect, i.e., will decrease blood pressure (BP) and heart rate (HR), gastro-protection, or the chorda tympani response to sweet compounds (upper part of the figure). In the case of opposite effects of CB1R and AT1R activation, CB1R blockade will enhance the AT1R-mediated effects, i.e., its positive inotropic action, its facilitatory effect on miniature excitatory postsynaptic currents (mEPSCs) or its vasoconstrictor effect (lower part of the figure). For the respective literature, see Table 2 and Section 3.2.1.
The following facts determined by Turu et al. [33] on Chinese hamster ovary (CHO) cells co-expressing CB1Rs and AT1Rs clearly confirm the above working hypothesis. Firstly, Ang II increased levels of 2-AG (measured by mass spectrometry) but not of AEA. Secondly, Ang II elicited Go protein activation (determined by detecting the dissociation of G protein subunits with bioluminescence resonance energy transfer; BRET) that was blocked both by the CB1R antagonist AM251 and the DAGL inhibitor tetrahydrolipstatin (THL); AM251 and THL by themselves inhibited basal CB1R activity. The pioneering observations by Turu et al. [33] have been confirmed and extended by the same group on HEK293 and COS7 cells from kidney and fibroblasts, respectively [34]. Thus, Ang II caused CB1R-dependent activation not only in an autocrine but also in a paracrine manner [34], i.e., when AT1Rs and CB1Rs are localized on the same or separate cells, respectively.
The latter interaction obtained on cell culture has been verified in blood vessels and other body systems. First of all, Szekeres et al. [36] showed in vitro that Ang II enhanced the calcium signal and the 2-AG level in rat aortic vascular smooth muscle cells (VSMCs). Both responses to Ang II were reduced by the DAGL inhibitor THL and enhanced by the MAGL inhibitor JZL184 leading to inhibition of 2-AG synthesis and degradation and finally to a decrease and increase in 2-AG levels, respectively. The results obtained on cell culture were further confirmed on isolated vessels examined in the organ bath, including the wire myograph for studies on small resistance arteries. As shown in Table 2, Ang II activation. If CB 1 R and AT 1 R activation lead to the same effect, CB 1 R antagonism will inhibit the AT 1 R-mediated effect, i.e., will decrease blood pressure (BP) and heart rate (HR), gastro-protection, or the chorda tympani response to sweet compounds (upper part of the figure). In the case of opposite effects of CB 1 R and AT 1 R activation, CB 1 R blockade will enhance the AT 1 R-mediated effects, i.e., its positive inotropic action, its facilitatory effect on miniature excitatory postsynaptic currents (mEPSCs) or its vasoconstrictor effect (lower part of the figure). For the respective literature, see Table 2 and Section 3.2.1.
The following facts determined by Turu et al. [33] on Chinese hamster ovary (CHO) cells co-expressing CB 1 Rs and AT 1 Rs clearly confirm the above working hypothesis. Firstly, Ang II increased levels of 2-AG (measured by mass spectrometry) but not of AEA. Secondly, Ang II elicited G o protein activation (determined by detecting the dissociation of G protein subunits with bioluminescence resonance energy transfer; BRET) that was blocked both by the CB 1 R antagonist AM251 and the DAGL inhibitor tetrahydrolipstatin (THL); AM251 and THL by themselves inhibited basal CB 1 R activity. The pioneering observations by Turu et al. [33] have been confirmed and extended by the same group on HEK293 and COS7 cells from kidney and fibroblasts, respectively [34]. Thus, Ang II caused CB 1 R-dependent activation not only in an autocrine but also in a paracrine manner [34], i.e., when AT 1 Rs and CB 1 Rs are localized on the same or separate cells, respectively.
The latter interaction obtained on cell culture has been verified in blood vessels and other body systems. First of all, Szekeres et al. [36] showed in vitro that Ang II enhanced the calcium signal and the 2-AG level in rat aortic vascular smooth muscle cells (VSMCs). Both responses to Ang II were reduced by the DAGL inhibitor THL and enhanced by the MAGL inhibitor JZL184 leading to inhibition of 2-AG synthesis and degradation and finally to a decrease and increase in 2-AG levels, respectively. The results obtained on cell culture were further confirmed on isolated vessels examined in the organ bath, including the wire myograph for studies on small resistance arteries. As shown in Table 2, Ang II caused a concentration-dependent vasoconstriction of rat and/or mice aortic rings [36], skeletal muscle arterioles and saphenous arteries [37], and intramural coronary resistance arterioles [38]. All vasoconstrictory responses to Ang II were enhanced by antagonists of CB 1 Rs (the inverse agonists rimonabant and AM251 or the neutral antagonist O2050) and by the DAGL inhibitor THL via reduction of the CB 1 R-mediated vasodilatation or 2-AG levels, respectively. On the other hand, the MAGL inhibitor JZL184, increasing 2-AG levels, reduced the above effects of Ang II. Moreover, the CB 1 R antagonists had an amplificatory influence only in wild-type but not in CB 1 R knockout mice [36,37]. The presence of CB 1 Rs in the particular vessels or aortic VSMC has been confirmed with immunohistochemistry or using the Western blot technique [36][37][38] or functionally by studying the vasodilatory effect of the CB 1 R agonist WIN55212-2 with or without CB 1 R blockade [36][37][38]. Moreover, WIN55212-2-induced vasodilatation was not observed in CB 1 −/− mice [36,37]. An interaction between Ang II and CB 1 Rs resembling that in systemic arteries (see above) was also observed in rat pulmonary arteries, in which CB 1 Rs are present in smooth muscle, as determined using immunohistochemical staining [39]. Thus, the Ang II-induced concentration-dependent contraction of pulmonary arteries was amplified by AM251 and the DAGL inhibitor RHC8026 and diminished by inhibition of 2-AG but not AEA degradation by JZL184 and URB597, respectively. Unfortunately, the Ang II-evoked response in human pulmonary arteries was very weak and did not allow for examining its potential interaction with CB 1 -Rs [39].
In uterine arteries isolated from control and TgA rats (model that mimics many features of preeclampsia), the vasoconstrictor responses to Ang II (which were stronger in TgA than in control animals) were decreased by inhibitors of FAAH (URB597) and/or MAGL (JZL184) but not by the CB 1 R antagonist rimonabant [40]. Authors suggest that eCBs released in response to Ang II reduced the Ang II-induced contraction in a CB 1 Rindependent manner. Indeed, the non-selective COX inhibitor indomethacin reduced the maximal contraction and sensitivity to Ang II in arteries from TgA rats, suggesting that eCBs (both AEA and 2-AG) released by Ang II were quickly converted to vasoactive eicosanoids via COX, yielding vasodilator prostaglandins and/or prostacyclin.
In summary, the above in vitro experiments on isolated arteries clearly show the existence of protective negative feedback in response to the vasoconstrictory effect of Ang II. Thus, Ang II acting via AT 1 Rs not only contracts particular arteries but also stimulates a DAGL-dependent release of eCBs (mainly 2-AG) from vascular endothelium or smooth muscles that cause vasodilatation in a CB 1 R-dependent ( Figure 4) or independent manner, thereby diminishing the initial response to Ang II.
An interaction between AT 1 Rs and eCBs is also suggested to be present in the heart ([42]; Table 1, Figure 4). In rat isolated Langendorff-perfused hearts Ang II reduced coronary flow (CF) and caused a moderate negative inotropic effect, although it is known to exert a positive inotropic effect in other heart preparations (e.g., on rat cardiomyocytes in the study by Rajagopal et al. [67]). In Langendorff-perfused hearts, the activation of CB 1 Rs (the presence of which was confirmed by immunohistochemistry) by WIN55212-2 and 2-AG increased CF and decreased cardiac contractility in a manner sensitive to the CB 1 R antagonist O2050. However, in contrast to previous experiments on rat coronary arteries [38], O2050 and the DAGL inhibitor orlistat did not amplify but reduced both cardiac effects of Ang II. Miklós et al. [42] argue that the final response to Ang II in the heart is determined by an interplay between coronary arteries and cardiomyocyte contractility, and the degree of CF results primarily from the cardiac oxygen demand and not only from the direct vascular effects of vasoactive agents. Authors suggest that CB 1 R blockade by O2050 prevented the CB 1 R-induced negative inotropic effect elicited by 2-AG released under Ang II stimulation. Ang II then increases cardiac contractility (otherwise masked by flow deprivation) and, subsequently, the overall oxygen demand and the CF.
The question arises whether the protective negative feedback of Ang II-eCBs-CB 1 Rs observed in vitro is relevant also under in vivo conditions. The answer is "yes", as shown in Table 1. First of all, Szekeres et al. [36] showed that, as in experiments on isolated arteries, the increase in blood pressure (BP) elicited by intravenous (i.v.) infusion of Ang II was enhanced by the CB 1 R antagonist O2050 in anesthetized CB 1 +/+ R wild type as opposed to CB 1 −/− R knockout mice. A similar mechanism may occur in experiments performed on conscious rats with hypertension induced by infusion of Ang II [19] or Ang II plus vasopressin [46]. Thus, the CB 1 R antagonist AM251 enhanced and the FAAH inhibitor URB597 reduced BP [19], and URB597 enhanced the fall in BP elicited by AEA [46]. Similarly, in spontaneously hypertensive (SHR) rats (in which the RAS is overactivated [68]) but not in their normotensive Wistar Kyoto (WKY) controls, the CB 1 R antagonist rimonabant and the FAAH inhibitor URB597 enhanced and reduced BP, respectively [19]. The blockade of the hypotensive effect of AEA and WIN55212-2 by AM251 confirms the involvement of CB 1 Rs in the hypotensive action of both compounds [46].
Interestingly, an opposite effect of CB 1 R blockade was obtained in transgenic (mRen2)27 rats, i.e., a monogenetic model of Ang II-dependent hypertension in which the mouse Ren2 renin gene was transfected into the genome of the Sprague-Dawley (SD) rat. (mRen2)27 rats have the phenotype of chronic hypertension with markedly impaired baroreflex control over heart rate (HR) and increased body weight. Unlike in SHR or hypertension models induced by Ang II [19] or Ang II with vasopressin infusion [46], acute or chronic administration of rimonabant decreased BP and body weight in (mRen2)27 rats but not in their normotensive SD controls [47]. One can exclude that the reduction of body weight is responsible for the fall in BP since it resulted not only from chronic but also from acute administration of the CB 1 R antagonist. Moreover, chronic rimonabant administration improved sympathetic and parasympathetic baroreflex sensitivity for HR control but failed to change plasma Ang II and Ang 1-7 levels or ACE activity. The authors concluded that an upregulated endocannabinoid system contributes to hypertension and impaired autonomic function in this Ang II-dependent model and that systemic CB 1 R blockade may be an effective therapy in this type of hypertension [47]. In a subsequent study, Schaich et al. [48] confirmed their conclusion showing enhanced levels of 2-AG (but not AEA) in the solitary tract nucleus (NTS) of (mRen2)27 rats in comparison to SD or ASrAOGEN rats (the latter have a low level of glial angiotensinogen). Microinjection of rimonabant into the NTS dose-dependently enhanced baroreceptor sensitivity in (mRen2)27 but decreased it in ASrAOGEN animals. Thus, upregulation of the brain endocannabinoid system may be responsible for the impairment of baroreceptor sensitivity in this Ang II-dependent model of hypertension.
In this context, one should keep in mind that Ang II administered centrally and peripherally increases BP [69], whereas cannabinoids can elicit different cardiovascular responses (including hypotension and hypertension) [70,71]. Interactions between the (endo)cannabinoid and RAS systems may not only occur in vessels but also in the CNS. The paraventricular nucleus of the hypothalamus (PVN) is one of the crucial brain regions involved in the regulation of cardiovascular functions, in which the balance between sympathoexcitatory glutamatergic and sympathoinhibitory GABAergic transmission is responsible for the final integration of the sympathetic outflow [72]. Gyombolai et al. [53] were the first to show that microinjection of the CB 1 R antagonist AM251 into the PVN of anesthetized rats reduced the increase in BP elicited by application of Ang II to the same nucleus, although the underlying mechanism has not been clarified. We found that the cannabinoid receptor agonist CP55940 given into the PVN of rats anesthetized with urethane decreased BP but in the presence of AM251 given i.v. produced a clear pressor response, sensitive to AM251 given into the PVN [54]. The AT 1 R antagonist losartan given i.v. inhibited the depressor response to CP55940 only but reversed its pressor response to a fall in BP. In our most recent study, performed on conscious SHR and their normotensive (WKY) controls in which all compounds were microinjected into the PVN, the pressor responses to Ang II and to CP55940 (both of which were stronger in SHR than in WKY) were inhibited by antagonists of AT 1 and AT 2 receptors, losartan and PD123319, respectively. On the other hand, AM251 diminished the increase in BP induced by Ang II [55]. The pressor effects of Ang II and CP55940 given into the PVN result mainly from the activation of facilitatory AT 1 Rs on glutamatergic [73,74] and inhibitory CB 1 Rs on GABAergic [54,55] neurons, respectively ( Figure 5). In addition, Ang II may lead to the release of eCBs that, via CB 1 Rs, reduce the GABAergic tone and eventually lead to increased activity of the glutamatergic neurons associated with a pressor effect. On the other hand, AM251 blocking presynaptic CB 1 Rs on GABAergic neurons may increase the GABAergic inhibitory influence on the sympathoexcitatory glutamatergic tone stimulated by Ang II and, in this way, reduces its pressor response ( Figure 5). The potential mechanism responsible for the interaction of AT 2 Rs (which, unlike the AT 1 Rs, are not coupled to G q/11 proteins; [75]) with CB 1 Rs remains to be established. AT 2 Rs are known mainly to counteract the pressor effect of AT 1 R stimulation by enhancing GABAergic neurotransmission (AT 2 Rs are located predominantly on GABAergic neurons; [76] The pressor effects of Ang II and CP55940 given into the PVN result mainly from the activation of facilitatory AT1Rs on glutamatergic [73,74] and inhibitory CB1Rs on GABAergic [54,55] neurons, respectively ( Figure 5). In addition, Ang II may lead to the release of eCBs that, via CB1Rs, reduce the GABAergic tone and eventually lead to increased activity of the glutamatergic neurons associated with a pressor effect. On the other hand, AM251 blocking presynaptic CB1Rs on GABAergic neurons may increase the GABAergic inhibitory influence on the sympathoexcitatory glutamatergic tone stimulated by Ang II and, in this way, reduces its pressor response ( Figure 5). The potential mechanism responsible for the interaction of AT2Rs (which, unlike the AT1Rs, are not coupled to Gq/11 proteins; [75]) with CB1Rs remains to be established. AT2Rs are known mainly to counteract the pressor effect of AT1R stimulation by enhancing GABAergic neurotransmission (AT2Rs are located predominantly on GABAergic neurons; [76]).

Figure 5.
Potential mechanisms of the cross-talk between AT1 receptors (AT1Rs) and cannabinoid type 1 receptors (CB1Rs) in the paraventricular nucleus of the hypothalamus. AT1R activation increases blood pressure and heart rate due to a direct and indirectly mediated increase in glutamate (Glu) release. The indirect effect involves an inhibitory γ-aminobutyric acid (GABA) interneuron. In detail, AT1R activation increases the release of endocannabinoids (mainly 2-arachidonoyl-glycerol) acting at presynaptic cannabinoid CB1 receptors (CB1Rs), activation of which decreases the inhibitory influence of GABA on sympathoexcitatory glutamatergic neurons. On the other hand, the CB1R antagonist AM251 blocking presynaptic CB1Rs increases the inhibitory influence of GABA on glutamatergic neurons excited by AT1R activation, resulting in a decreased Ang II-induced pressor response. Facilitatory influences are shown by green arrows and plus signs, whereas inhibitory effects are represented by red bars and minus signs.
Interestingly, the mechanism related to the Ang II-induced transactivation of the cannabinoid-mediated effects is not restricted to the cardiovascular system. Thus, AM251 enhanced the Ang II-induced miniature excitatory postsynaptic currents (mEPSCs) in supraoptic magnocellular neurons from the supraoptic nucleus (SON), which is crucial for body fluid homeostasis (Table 2, Figure 4). The results suggest that eCBs released by Ang II modulate the excitatory synaptic inputs via presynaptic inhibitory CB1Rs [56]. The authors based their conclusion on the results of their previous publication, in which they proved that CP55940 reduced the frequency of spontaneous mEPSCs in magnocellular neurosecretory neurones of the SON in a manner sensitive to AM251 [77]. On the other hand, there may be an alternative explanation for the interaction. Thus, CB1Rs are precoupled receptors, and an inverse agonist like AM251 is expected to lead to an increase in Figure 5. Potential mechanisms of the cross-talk between AT 1 receptors (AT 1 Rs) and cannabinoid type 1 receptors (CB 1 Rs) in the paraventricular nucleus of the hypothalamus. AT 1 R activation increases blood pressure and heart rate due to a direct and indirectly mediated increase in glutamate (Glu) release. The indirect effect involves an inhibitory γ-aminobutyric acid (GABA) interneuron. In detail, AT 1 R activation increases the release of endocannabinoids (mainly 2-arachidonoyl-glycerol) acting at presynaptic cannabinoid CB 1 receptors (CB 1 Rs), activation of which decreases the inhibitory influence of GABA on sympathoexcitatory glutamatergic neurons. On the other hand, the CB 1 R antagonist AM251 blocking presynaptic CB 1 Rs increases the inhibitory influence of GABA on glutamatergic neurons excited by AT 1 R activation, resulting in a decreased Ang II-induced pressor response. Facilitatory influences are shown by green arrows and plus signs, whereas inhibitory effects are represented by red bars and minus signs.
Interestingly, the mechanism related to the Ang II-induced transactivation of the cannabinoid-mediated effects is not restricted to the cardiovascular system. Thus, AM251 enhanced the Ang II-induced miniature excitatory postsynaptic currents (mEPSCs) in supraoptic magnocellular neurons from the supraoptic nucleus (SON), which is crucial for body fluid homeostasis (Table 2, Figure 4). The results suggest that eCBs released by Ang II modulate the excitatory synaptic inputs via presynaptic inhibitory CB 1 Rs [56]. The authors based their conclusion on the results of their previous publication, in which they proved that CP55940 reduced the frequency of spontaneous mEPSCs in magnocellular neurosecretory neurones of the SON in a manner sensitive to AM251 [77]. On the other hand, there may be an alternative explanation for the interaction. Thus, CB 1 Rs are precoupled receptors, and an inverse agonist like AM251 is expected to lead to an increase in mEPSCs even in the absence of eCBs. Thus, Ang II and AM251 may increase mEPSCs via two independent mechanisms, i.e., via agonism at facilitatory AT 1 Rs and inverse agonism at inhibitory CB 1 Rs, respectively.
Gyires et al. [57] showed that the gastroprotective effect of Ang II given intracerebroventricularly (i.c.v.) against ethanol-induced gastric lesions in conscious rats was reduced not only by the AT 1 R antagonist candesartan but also by AM251 and the 2-AG synthesis inhibitor THL (all compounds were injected i.c.v.). Moreover, the beneficial influence of Ang II was absent in CB 1 −/− knockout mice. The authors suggest that Ang II caused gastroprotection via stimulation of central AT 1 Rs that led to the release of eCBs in a DAGL-dependent manner and the subsequent activation of CB 1 Rs (Figure 4). CB 1 Rs, like AT 1 Rs, are known to exert beneficial effects on gastrointestinal function [57].
Shigemura et al. [58] proved that intraperitoneal (i.p.) administration of Ang II decreased the response of the gustatory nerve to NaCl both in the wild type and in CB 1 −/− knockout mice. By contrast, the genetic deletion of CB 1 Rs prevented the enhancement of the response to sweeteners induced by Ang II. Authors suggested that activation of AT 1 Rs by Ang II stimulates 2-AG release, which may act as an autocrine enhancer for CB 1 Rs on sweet taste cells (Table 2, Figure 4). The CB 1 R-dependent enhancement of gustatory nerve responses to sweeteners induced by AEA or 2-AG had previously been demonstrated in mice by Yoshida et al. [78].
The authors of all the above publications argue that Ang II causes the release of eCBs (mainly 2-AG). Unfortunately, as shown in Table 2, this interesting hypothesis has been confirmed directly by determining eCB levels in cell culture and vessels only [33,36]. In the other studies it is based mainly on indirect proof using inhibitors of eCB synthesis and/or degradation and CB 1 R antagonists.

Other Forms of Interaction between Angiotensin II and Cannabinoid-Mediated Effects
Numerous authors [43,47,48,53,54,63,64] argue that the interaction between CB 1 and AT 1 receptors they found is based on AT 1 R-CB 1 R heteromerization. However, this has so far been proven only in the excellent study by Rozenfeld et al. [35], in which coimmunoprecipitation, resonance energy transfer assays, and receptor-and heteromer-selective antibodies were used. The authors demonstrated that G αq -coupled AT 1 Rs and G αi -coupled-CB 1 Rs form receptor heteromers in cells in which both receptors occur naturally (hepatic stellate cells [HSCs] isolated from ethanol-administered rats), and naturally (CB 1 ) and recombinantly (AT 1 ) (cultured Neuro 2A cells) ( Table 2). The following results on cultured Neuro 2A cells co-expressing CB 1 and AT 1 receptors clearly confirm the existence of AT 1 R-CB 1 R heteromerization: (1) Ang II-induced an increase in ERK1/2 phosphorylation via G αi (for CB 1 Rs) instead of G αq (for AT 1 Rs), which was reduced by decreasing CB 1 R levels by siRNA, and by rimonabant or THL; (2) the CB 1 R agonist HU210 stimulated ERK1/2 phosphorylation only in the presence of a very low non-signaling concentration of Ang II; (3) an increase in intracellular Ca 2+ levels is induced by activation of AT 1 but not CB 1 receptors, yet requires the presence of CB 1 Rs since it was attenuated by siRNA targeting CB 1 Rs; (4) the expression of AT 1 Rs shifted the localization of CB 1 Rs from an intracellular compartment to the plasma membrane. The physiological relevance of AT 1 R-CB 1 R heteromerization has been demonstrated in HSCs isolated from ethanol-treated rats and their controls. Firstly, the levels of CB 1 Rs, AT 1 Rs, and AT 1 R-CB 1 heteromers were upregulated in ethanol-treated HSCs compared to controls. Secondly, the CB 1 R antagonist rimonabant prevented Ang II-mediated mitogenic signaling and profibrogenic gene expression. The authors concluded that enhanced CB 1 R expression in activated HSCs might affect AT 1 R properties and contribute to the profibrogenic effect of Ang II.
Attenuation of the AT 1 R-mediated effect by CB 1 R blockade is likely for the postganglionic sympathetic neurons, and AT 1 R-CB 1 R heteromerization might explain this phenomenon. In general, sympathetic neurons are equipped with presynaptic inhibitory G i/o protein-and facilitatory G q and G s protein-coupled receptors. Although G i/o and G s protein-coupled receptors in the mouse atrium act independently [79], this does not hold for the interplay between the G i/o and G q protein-coupled receptors [79,80]. Thus, if the presynaptic inhibitory α 2 -adrenoceptor (an example of a G i/o protein-coupled receptor) is blocked by an α-adrenoceptor antagonist, the AT 1 R-mediated facilitation of noradrenaline release from the sympathetic neurons is impaired [79]. Facilitation can be restored if an agonist of another G i/o protein-coupled receptor is added, e.g., a δ-opioid receptor agonist or neuropeptide Y (acting via presynaptic Y 2 receptors). The same occurs if the presynaptic bradykinin (B 2 ) receptor is considered instead of the AT 1 R [79]. In harmony with these findings, AT 1 R (and B 2 receptor)-mediated facilitation of noradrenaline release is impaired in α 2 -adrenoceptor knockout mice [80]. Unfortunately, the interaction between the presynaptic AT 1 R and CB 1 R has not been examined so far.
In a few studies, antagonists of CB 1 Rs caused down-regulation of AT 1 Rs and subsequently also the pathological effects elicited by Ang II. Such changes have been observed in vitro and in vivo following acute and chronic administration of CB 1 R antagonists. Thus, in aortic VSMCs, rimonabant and AM251 (but not the CB 2 R antagonist SR144528) reduced the oxidative stress induced by Ang II. Moreover, blockade by CB 1 Rs or their stimulation by CP55940 led to down-and up-regulation of AT 1 Rs, respectively. The level of CB 1 Rs was not affected by rimonabant in VSMCs ( [41]; Table 1). Chronic administration of rimonabant for 7 weeks to apolipoprotein E-deficient (ApoE −/− ) mice on a cholesterol-rich diet did not prevent atherosclerotic plaque development, collagen content, and macrophage infiltration but, like in the above in vitro experiments, reduced oxidative stress and improved aortic endothelium-dependent vasodilation associated with the down-regulation of AT 1 Rs [41].
Similarly, application of the CB 1 R antagonist rimonabant or AM281 for 11 weeks or genetic deletion of CB 1 Rs in mice reduced AT 1 and CB 1 receptor levels in the left cardiac ventricle and improved cardiac dysfunction in streptozotocin-induced diabetes ([43] Table 2). Moreover, treating obese Zucker rats with rimonabant for 12 months reduced the pressor response to acute i.v. injection of Ang II in comparison to vehicle-treated control animals. The CB 1 R blockade also improved renal function and metabolic profile and increased the survival of obese Zucker rats ( [49], Table 2). The authors suggested that rimonabant might reduce the vascular AT 1 R expression. However, an indirect mechanism related to the decrease in cholesterol level should also be considered.
The paper by Jourdan et al. ( [63]; Table 2) refers to the interaction between AT 1 and CB 1 receptors in kidneys. In an in vitro study on human podocytes, the peripheral CB 1 R antagonist JD5037 and the AT 1 R antagonist losartan diminished the Ang II-induced enhancement of AEA and 2-AG levels and the increased expression of both CB 1 and AT 1 receptors. These results are in line with in vivo observations on Zucker diabetic fatty rats with nephropathy, which had higher plasma levels of Ang II and aldosterone as well as up-regulated AT 1 and CB 1 receptors in the renal cortex in comparison with control lean rats [63]. Chronic administration of JD5037 for 3 months or losartan for 4 weeks prevented/attenuated the development of nephropathy and decreased the expression of AT 1 and CB 1 receptors in the renal cortex. Authors suggest that activation of renal CB 1 Rs aggravates the deleterious effects of RAS activity. Indeed, in rats with streptozotocininduced diabetes, dual 14-week therapy with the peripheral CB 1 R antagonist AM6545 and the ACE inhibitor perindopril was more effective than monotherapies. Thus, single antagonists produced various beneficial effects but only combined treatment very markedly reduced albuminuria ( [64]; for details, see Table 2). Importantly, diabetic nephropathy was associated with higher glomerular levels of CB 1 and AT 1 receptors. Both mono-and combined therapy failed to modify glomerular CB 1 R expression but normalized AT 1 R levels. Altogether, the studies by Jourdan et al. [63] and Barutta et al. [64] suggest that both RAS inhibition and CB 1 receptor blockade have a beneficial influence on the locally increased responsiveness to Ang II.
The studies by Mattace Raso et al. [50,51] and Franco-Vadillo et al. [44] document a beneficial and CB 1 R-independent influence of two cannabinoids on the harmful effects of Ang II (for details, see Table 1). A decrease in AT 1 R densities was found in the heart [44], carotid and mesenteric arteries [51], and kidney [50] in response to chronic administration of CBD for 10 days to rats [44] or PEA for 5 weeks to hypertensive SHR and their normotensive WKY counterparts [50,51]. CBD or PEA treatment also reduced detrimental changes in cardiac [44], vascular and renal [50,51] tissues, and BP in SHR [50,51]. CBD and PEA show an overlap in their biochemical effects [81]. CBD is the major nonpsychoactive cannabinoid constituent of marijuana and has a low affinity for CB 1 and CB 2 receptors (Figure 2; [82]). PEA belongs to so-called endocannabinoid-like compounds with similar chemical structures to eCBs but is devoid of affinity for any CB receptor [8]. Apart from regulating the endocannabinoid system, CBD has a complex pharmacodynamic profile, including PPARγ [82]. The primary mechanism of action of PEA is its direct activation of PPARα receptors [81]. PPARγ agonists are known to decrease BP in humans, possibly through the suppression of the RAS, including the inhibition of AT 1 R expression [83]. Similarly, activation of PPARα decreases the Ang II-mediated rise in BP [84]. Thus, PPARs may represent the intermediate link in the interaction between CBD, PEA, and RAS, although this hypothesis needs to be proven by further studies. Of course, other explanations of the influence of CBD and PEA on AT 1 R densities are possible, especially when considering the pleiotropic action of CBD.
Beneficial effects against Ang II-induced fibrosis and inflammation in mouse heart, aorta, lung, kidney, and skin were obtained after chronic administration of EHP-101, an oral lipidic formulation of the novel non-psychotropic cannabidiol aminoquinone VCE-004.8, a dual agonist of CB 2 R and PPARγ receptors ( [45]; Table 2). Both properties may contribute to a functional antagonism against the detrimental consequence of AT 1 R stimulation.
An opposite cross-talk, i.e., the ability of Ang II to reduce CB receptor densities, has been identified in brainstem and cerebellar astrocytes that play critical roles in hypertension and attention-deficit hyperactivity disorder, respectively, and in which activation of astroglial AT 1 Rs and CB 1 Rs has been determined to increase and to neutralize the inflammatory state, respectively ( [59,85]; for details, see Table 2). Authors compared the effects of Ang II on CB 1 Rs in astrocytes isolated from hypertensive SHR and normotensive WKY rats [59]; they found that basal expression of CB 1 Rs in brainstem astrocytes was lower and that in cerebellar astrocytes was higher in SHR than in WKY rats. Ang II infusion further decreased (brainstem) and increased (cerebellum) CB 1 R expression via AT 1 Rs (brainstem) and both AT 1 Rs and AT 2 Rs (cerebellum). In further studies, Haspula and Clark [60] showed that stimulation of CB 1 Rs by their potent agonist ACEA led to inhibition of AT 1 R-mediated MAPK activation. On the other hand, Ang II caused phosphorylation and, probably, inactivation of CB 1 Rs. Moreover, ACEA caused the neutralization of Ang II-mediated changes in IL-10 and IL-1β gene expression in astrocytes from both brain regions of SHR and WKY [61].
Interesting results were obtained by Shi et al. [52], who found that Ang II decreased the activity of the AEA transporter in cultured lymphocytes from WKY in a manner sensitive to the AT 1 R antagonist losartan; this in vitro observation was confirmed under in vivo conditions in SHR and WKY (Table 2). Plasma Ang II and AEA levels were higher in SHR than in WKY. The higher AEA level probably resulted from the lower activity of the AEA transporter in SHR. Chronic administration of losartan not only dose-dependently reduced BP but also increased AEA transporter activity and diminished plasma AEA levels. The authors concluded that Ang II plays a critical role in mediating the decrease in AEA transporter activity in SHRs, probably via AT 1 Rs [52].
Importantly, there are more and more examples that cannabinoids modulate ACE levels or activity. Wang et al. [62] were probably the first to show that the CB 2 R agonist JWH133 increased ACE levels in Neuro-2a cells in a manner sensitive to the CB 2 R antagonist AM630. ACE levels were also lower in hippocampus slices isolated from knockout CB 2 −/− mice in comparison to the wild-type animals ( Table 2). Further examples of the influence of cannabinoids on ACE and, in particular, on ACE2 are given in part 3.4.

Typology and Potential Therapeutic Significance of the Cross-Talk between (Endo) Cannabinoids and the Renin-Angiotensin System
A detailed analysis of the results shown in Table 2, based on 33 papers, reveals that the cross-talk between the (endo) cannabinoid and renin-angiotensin systems mainly occurs in the following two ways (a. and b.).
Depending on the model (Table 2 and Figure 4), CB 1 R blockade or genetic deletion of CB 1 Rs may enhance or reduce the effects elicited by AT 1 R activation. The final result of the cross-talk between AT 1 and CB 1 receptors depends on whether CB 1 R activation induces an effect in the same or the opposite direction compared to AT 1 R stimulation. In the case of opposite effects mediated via CB 1 and AT 1 receptors, CB 1 R antagonists will increase the AT 1 R-mediated effects. This holds for some blood vessels of rodents [36][37][38][39], the rat heart [42], and mEPSCs in magnocellular neurosecretory cells from the supraoptic nucleus of mice [56]. On the other hand, CB 1 R antagonists will reduce the AT 1 R-mediated effects if CB 1 and AT 1 receptor activation elicit responses in the same direction (Table 2; Figure 4). This holds for the pressor effect of Ang II after its application into the PVN of the rat [53][54][55], the gastroprotective influence of Ang II given i.c.v. to rodents [57], and the responses to sweeteners in mice [58] (Table 2; Figure 4). b. A decrease in AT 1 R expression due to CB 1 R antagonism has been shown less frequently. Thus, acute application of CB 1 R antagonists reduced AT 1 R levels in rodent VSMCs [41] and human podocytes [63], and their chronic administration diminished the AT 1 R expression in the aorta of ApoE −/− mice [41], mouse heart [43], and kidney [63,64] isolated from diabetic mice and/or rats (Table 2; Figure 4). A decrease in AT 1 Rs was also observed in response to chronic treatment of rats with cannabidiol (heart [44]) and PEA (kidney, arteries [44,51]).
So far, the practical relevance of the cross-talk between ECS and RAS for humans cannot be appreciated. In this respect, CB 1 R antagonists are interesting and are discussed as future drug strategies for the treatment of diabetes [86,87], chronic kidney disease [88], fatty liver disease [89], alcohol use disorder [86], and COVID-19 [90]. The CB 1 R antagonist rimonabant was used as an anti-obesity drug [15,86,91] but withdrawn from the market in 2008 because of severe central side effects [86]. Second-generation CB 1 R antagonists (e.g., AM6545, JD5037), peripherally restricted, may offer advantages over rimonabant. In this context, detailed studies are necessary to clarify whether they increase vascular and cardiac harmful AT 1 R-mediated responses or, by contrast, reduce the detrimental responses to AT 1 R stimulation or expression of AT 1 Rs. One should also keep in mind that FAAH and MAGL inhibitors, which might become future drugs for the treatment of pain, traumatic brain injury, multiple sclerosis, Parkinson's disease, anxiety-related disorders, and depression [8], may also modify the responses induced by AT 1 R activation (Table 2).

Interaction between Angiotensin 1-7 and (Endo) Cannabinoid-Mediated Effects
So far, only two publications show that not only Ang II but also Ang 1-7 seems to modulate cannabinoid-related effects (Table 3). In the first, Brosnihan et al. [92] showed that the infusion of Ang 1-7 altered the endocannabinoid system in the decidualized uterus of pseudopregnant rats. A Low AEA level is preferred for fertilization, implantation, decidualization, and placentation, whereas a high AEA level causes embryotoxicity. Cannabinoid CB 1 R activation impaired decidualization. The role of 2-AG and CB 2 Rs is still not established in early pregnancy [92,93]. Chronic infusion of Ang 1-7 for 5 days caused an up-regulation of CB 1 R, CB 2 R, and MAGL mRNA in the decidualized uterine horn. However, as decidualization seems to be inhibited by activation of CB 1 Rs, a higher expression level may result in implantation failure, spontaneous miscarriage, or preeclampsia [93].
In the second publication, we identified the cross-talk between MasRs for Ang 1-7 and CB 1 Rs in the PVN of conscious hypertensive SHR and their normotensive controls, WKY ( [55]; Table 3). Microinjection of Ang 1-7 or the CB 1 R agonist CP55940 into the PVN produced pressor responses, which were stronger in SHR than in WKY. (Note that CP55940 was microinjected after i.v. administration of the CB 1 R antagonist AM251 to unmask the pressor effect of CB 1 R activation in the PVN; [94]). The increase in BP induced by Ang 1-7 was not only prevented by the MasR antagonist A-779, but surprisingly reversed into a hypotensive response with AM251. On the other hand, the pressor response to CP55940 was reduced not only by AM251 injected into the PVN [94] but also by A-779. Additionally, A-779 decreased BP by itself, but only after the previous i.v. administration of AM251, suggesting a CNS-based interaction between eCBs acting on CB 1 Rs with the endogenous pressor tone elicited by Ang 1-7.
In summary, two in vivo studies on rats suggest that a cross-talk between Ang 1-7 (and its MasRs) and the ECS appears to exist. So far, there are no results/suggestions regarding the mechanism (s) underlying this cross-talk and its potential therapeutic consequence (s).

Influence of Cannabinoids on ACE2 Activity and Other Parameters Relevant in the Fight against COVID-19
The severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) infection was rapidly spreading worldwide and led to over a hundred million detected infections and more than 6,200,000 deaths by 26 May 2022 (https://covid19.who.int/ (accessed on 5 May 2022)). Problems with vaccination of part of the population and mutations of the virus mean that SARS-CoV-2 infections may continue for many more years [26]; new promising therapies against COVID-19 are constantly being proposed. Recently, many studies have been published about the relationship between the RAS and the pathophysiology and severity of COVID-19. A key role is played by ACE2, which, however, represents a "double-edged sword" [26]. On the one hand, ACE2 is a SARS-CoV-2 entry receptor, and increased ACE2 expression may facilitate virus entry into cells [1][2][3]. On the other hand, viral binding reduces ACE2 levels at the cell surface, and the shift of the RAS balance towards the deleterious over the protective axis (increase in the ACE/ACE2 ratio) may lead to a progression of disease severity ( Figure 3) [95,96]. High ACE/ACE2 ratios occur in males, geriatric, and smoking patients, and many pathologies (especially cardiovascular, pulmonary, and renal diseases and obesity) are recognized as comorbidities that may aggravate the COVID-19 infection. Low ACE/ACE2 ratios are found in women, physically trained individuals, and patients well-treated with ACE inhibitors. The beneficial properties of ACE2 can be achieved either by inhibiting the ACE/Ang II/AT 1 R axis or by promoting the ACE2/Ang 1-7/MasR axis as a result of reducing the ACE/ACE2 ratio and/or by directly stimulating RAS anti-inflammatory components such as soluble ACE2, Ang 1-7 analogs, MasR, or AT 2 R agonists ( Figure 3).
Cannabinoids have been proposed as potential therapies or adjuvant drugs against the SARS-CoV-2 infection from the beginning of the COVID-19 pandemic because of their anti-inflammatory and immunomodulatory activities [4,90,97,98].
Two previous publications, which appeared before the onset of the COVID-19 pandemic, showed that modifying the (endo)cannabinoid system may modulate ACE activity/level (Table 2). Thus, chronic administration of the endocannabinoid-like compound PEA reduced ACE levels in the aorta and kidney of SHR, which were enhanced compared to the normotensive WKY [50]. Furthermore, the CB 2 R agonist JWH133 increased the ACE level in N2a cells in a manner sensitive to the CB 2 R antagonist AM630; the ACE level in hippocampus slices from CB 2 −/− mice was lower than in their wild-type littermates [62]. Unfortunately, the promising observations have not been confirmed in conscious (mRen2)27 hypertensive rats (Table 2) and in human iPSC-cardiomyocytes infected with SARS-CoV-2 (Table 4). Thus, although chronic administration of rimonabant reduced BP and HR in hypertensive rats, the level of ACE was not affected [47]. Moreover, although the potent CB 1 /CB 2 R agonist WIN55212-2 inhibited cytotoxicity and the release of proinflammatory cytokines, it failed to affect ACE2 levels [99]. Concerning the latter study (and other studies listed in Table 4), it is of interest that tocilizumab, an antibody acting against the receptor of the proinflammatory cytokine interleukin-6, represents an important therapeutic strategy for the treatment of severe forms of SARS-CoV-2 [100].   Research dedicated to a potential therapeutic role of cannabinoids in the fight against COVID-19 significantly increased from the beginning of the pandemic onward. As shown in Table 4, the results are related mainly to changes in ACE2 activity/level under treatment of various phytocannabinoids. Thus, in an in silico docking study, both CBD and THC revealed a moderate inhibitory effect against human ACE2 and a strong (CBD) and moderate (THC) inhibitory effect on the main virus protease [101]. In another in silico study, CBD and cannabivarin were the most potent among 8 compounds derived from the cannabis plant to inhibit ACE2 [102]. Moreover, CBD decreased ACE2 expression in human 3D tissue models of oral, airway, and intestinal tissues (that serve as routes for the invasion of SARS-CoV-2) [103] and in human cell lines ( [31,105]; Table 4).
Cannabis contains cannabinoids, flavonoids, diterpenes, triterpenes, and lignans [107], which may show a synergistic, so-called "entourage" effect with CBD, i.e., the whole plant extract can have a higher potency and can be more beneficial than the individual compounds [31,108,109]. This putative entourage effect suggested previously for the therapeutic activity of CBD in COVID-19 infection (reviewed in [4]) does not appear to be mediated by direct CB 1 R or CB 2 R activation [110]. As shown in Table 3, high-CBD cannabis extracts downregulated ACE2 expression in 3D models of the oral cavity, intestinal, and lung epithelia [103]. Likewise, high-CBD/low-THC cannabis extracts [31], CBD-enriched Cannabis sativa Arbel strain extract [104], or industrial hemp oil containing terpenes and cannabidiolic acid (CBDA) and cannabigerolic acid (CBGA) [106,107] reduced ACE expression in cell lines [31,104,106]. Moreover, CBD-enriched extracts prevented infection of human epithelial cells by a pseudovirus expressing the SARS-CoV-2 spike protein and the entry of the SARS-CoV-2 alpha variant B.1.1.7 and the beta variant B.1.351 into cells at concentrations that are clinically achievable [107].
It should be highlighted that, as shown in Table 4, cannabis extracts, including various phytocannabinoids and terpenes, not only down-regulate ACE2 but can also suppress transmembrane serine protease 2 (TMPRSS2; another virus entry site into the cytoplasm of host cells) [31,103,106], main viral protease [101], neuropilin 1 (NRP1) [102], or inflammationpromoting agents such as IL-6 or IL-8 [31,102,104], CCL2 [104] or cyclooxygenase-2 [31]. However, one should keep in mind that CBD itself is characterized by a bell-shaped doseresponse curve associated with a narrow therapeutic window, which makes its effective clinical use difficult. Accordingly, the phytocannabinoid formulations (e.g., CBD with CBG and THCV) may show better and safer activity because such treatment can prevent the bell-shaped dose-response typical for CBD [104].
To summarize the 9 papers in Table 4, cannabinoids in plant extracts may inhibit SARS-CoV-2 via different mechanisms, potentially leading to enhanced effectiveness compared to the individual compounds [4,31,107]. Unfortunately, proinflammatory effects may occur as well [31,103,104]. Thus, their potential therapeutic use in the fight against COVID-19 (e.g., as adjunct therapy) must be approached with caution and requires further research.

Pharmacokinetic Interactions between Drugs Acting on the (Endo)Cannabinoid and Renin-Angiotensin Systems
This review is dedicated to interactions between the (endo)cannabinoid and reninangiotensin systems at the cellular level. However, if one considers the combined administration of drugs acting on either system, possible pharmacokinetic interactions should also be considered, i.e., one drug may interfere with the metabolism or the transport of another. Unexpectedly, this type of interaction escaped our PubMed-based search. In detail, THC, and, in particular, CBD, is known to interact with drug-metabolizing enzymes like CYP2C8/9 in the liver [111][112][113][114][115]. Moreover, they may affect transport proteins like the bile salt export pump important for losartan and telmisartan excretion. Third, they may, like ACE inhibitors, increase transaminase activity, thereby enhancing their side effects [114]. It is of interest that a wide range of over-the-counter marijuana and CBD-based products is now available, frequently characterized by an unknown CBD content (e.g., [4]). Thus, one should consider the possibility of unexpected pharmacokinetic interactions with RAS-based drugs against hypertension or heart failure.
In summary, strictly speaking, the interactions described in this section do not represent crosstalk between the ECS and the RAS but rather an interplay between individual drugs acting on either system. This interplay can be explained by the fact that the involved drugs, accidentally, share chemical properties, with the consequence that they are common substrates of metabolizing enzymes or transport mechanisms. The practical relevance is high. The concentration or half-time of one drug may be increased by the other, or side effects, e.g., on the liver, may be exaggerated.

Conclusions
During the past 15 years, there has been a noteworthy progression in the number of publications regarding the cross-talk between the (endo)cannabinoid and renin-angiotensin systems. One important mechanism is that AT 1 R activation by Ang II leads to the release of eCBs (mainly 2-AG), which, by acting at CB 1 Rs, modify the response to AT 1 R stimulation. The interplay, which has been precisely determined in the case of various arteries, should also be examined in detail in other organs, especially in the heart (considered in a single paper only). Another major mechanism is that CB 1 R antagonists diminish AT 1 R levels. A third mechanism is that phytocannabinoids influence ACE2. Moreover, the following interesting observations shown so far only in very few publications require further studies, i.e., (1) the existence of a potential cross-talk between the protective axis of the RAS (Ang 1-7 or AT 2 Rs) with components of the (endo)cannabinoid system, (2) the influence of Ang II on components of the endocannabinoid system like the levels of CB 1 Rs, CB 2 Rs or the AEA transporter and (3) the (patho)physiological significance of AT 1 R-CB 1 R heteromerization. Beyond the specific cross-talk between the RAS and the ECS, pharmacokinetic interactions have to be considered. They occur because single drugs acting on the RAS (telmisartan) or ECS (cannabidiol) accidentally share metabolizing enzymes or transport mechanisms.
The cross-talk may have consequences for drug therapy. (1) Thus, CB 1 R antagonists, the therapeutic significance of which is suggested for various diseases, may enhance (mainly vasoconstriction) or reduce (primarily effects mediated by the CNS) responses elicited by AT 1 Rs. The final reaction depends on whether stimulation of AT 1 Rs and CB 1 Rs induces opposite or the same effects, and potentially beneficial though also detrimental consequences may occur. (2) Despite initially promising results indicating beneficial effects of phytocannabinoids against COVID-19 (especially as adjuvant therapy), additional studies are required to determine their potential therapeutic role and the underlying mechanism (s) in this disease. (3) Pharmacokinetic interactions of drugs acting on the RAS and ECS may lead to an increase in the concentration and half-time of one of the drugs, and to exaggerated side effects, e.g., in the liver.

Conflicts of Interest:
The authors declare no conflict of interest.