Perturbation of 3D nuclear architecture, epigenomic aging and dysregulation, and cannabinoid synaptopathy reconfigures conceptualization of cannabinoid pathophysiology: part 2—Metabolome, immunome, synaptome
1Division of Psychiatry, University of Western Australia, Crawley, WA, Australia
2School of Medical and Health Sciences, Edith Cowan University, Joondalup, WA, Australia
*Correspondence: Albert Stuart Reece stuart.reece@uwa.edu.auAbstract
The second part of this paper builds upon and expands the epigenomic-aging perspective presented in Part 1 to describe the metabolomic and immunomic bases of the epigenomic-aging changes and then considers in some detail the application of these insights to neurotoxicity, neuronal epigenotoxicity, and synaptopathy. Cannabinoids are well-known to have bidirectional immunomodulatory activities on numerous parts of the immune system. Immune perturbations are well-known to impact the aging process, the epigenome, and intermediate metabolism. Cannabinoids also impact metabolism via many pathways. Metabolism directly impacts immune, genetic, and epigenetic processes. Synaptic activity, synaptic pruning, and, thus, the sculpting of neural circuits are based upon metabolic, immune, and epigenomic networks at the synapse, around the synapse, and in the cell body. Many neuropsychiatric disorders including depression, anxiety, schizophrenia, bipolar affective disorder, and autistic spectrum disorder have been linked with cannabis. Therefore, it is important to consider these features and their complex interrelationships in reaching a comprehensive understanding of cannabinoid dependence. Together these findings indicate that cannabinoid perturbations of the immunome and metabolome are important to consider alongside the well-recognized genomic and epigenomic perturbations and it is important to understand their interdependence and interconnectedness in reaching a comprehensive appreciation of the true nature of cannabinoid pathophysiology. For these reasons, a comprehensive appreciation of cannabinoid pathophysiology necessitates a coordinated multiomics investigation of cannabinoid genome-epigenome-transcriptome-metabolome-immunome, chromatin conformation, and 3D nuclear architecture which therefore form the proper mechanistic underpinning for major new and concerning epidemiological findings relating to cannabis exposure.
1. Introduction
Part 1 of this paper introduced the salience of an epigenomic and aging perspective in understanding cannabinoid pathophysiology (1). Part 2 presents the metabolomic and immunomic basis and underpinning of these changes and then considers their particular application to neurons and the synapse. Thus, this part should be read in conjunction with Part 1 and is properly considered as an extension of it and has been presented in a separate format in view of space considerations.
Much recent attention has focused on the spatial organization of the cell nucleus and the manner in which three dimensional topologically associated domains and transcription factories are epigenetically coordinated to precisely bring enhancers into close proximity with promoters to control gene expression. Many recent papers make the 3D architecture of the nucleus including chromatin conformation a major and principal focus of interest. This perspective is necessarily interactive between genome-epigenome-transcriptome. However, it is increasingly apparent that nuclear events are based squarely on events occurring in the cytoplasm, particularly as relates to metabolomic and immunomic activity which are not only permissive of nuclear changes but determinative of them. Moreover, neuronal and synaptic activity is a special case of these integrated and interactive levels of control and are clearly of particular relevance to the broad spectrum of neuropsychiatric pathologies linked with cannabis exposure, which now include several major neuropsychiatric syndromes and autistic spectrum disorder.
These features also mechanistically underpin several recent large epidemiological studies linking cannabis to the incidence of several cancers (2–12) and numerous congenital anomalies in both USA and Europe (13–19) and with aging syndromes of various types (12, 20), which were detailed further in Part 1.
3. Immunome
The immunome may be defined as the set of peptides, proteins, receptors, signaling systems, genes, and cells that together comprise the innate and adaptive immune systems (112, 113). Immune processes are highly relevant in many ways as the strongly oxidizing environment of immune and inflammatory processes can induce DNA breaks. So important is immune activity that chronic immune stimulation is now described as one of the pillars of aging (114) and the immunome has now had its own biological clock developed with which to measure organismal biological age (115). Interestingly, most of the predictive power of this clock is related to CXCL9 (115). Chief among the immune pathways is the recently described innate cytosolic nucleic acid sensing and signaling pathway cGAS-STING.
3.1. cGAS-STING
3.1.1. Basic function
cGAS-STING is a powerful cytosolic sensor of double stranded DNA (dsDNA) which has been recently described and is of broad relevance to cancer, aging, and genomic stability (116). cGAS also senses R-loops (117–121). Once cGAS binds to dsDNA, its endoplasmic reticulum-bound binding partner STING stimulates a type I interferon response (122). Oligomerization of cGAS, formation of long DNA-protein ladders, and phase separation (in a gel) lead to STING oligomerization and an augmented STING response (122–124). cGAS is activated by DNA in a length-dependent manner and is more powerfully stimulated by longer DNA segments (125). Termination of the cGAS-STING signal occurs by clathrin-associated AP-1 (126). The electron microscopic structure of cGAS-STING has been determined (122, 127–129).
3.1.2. Roles in disease
In addition to innate immunity, cGAS-STING also drives autophagy, cell survival, infection, inflammation, cancer, and senescence pathways (123). cGAS can transfer between cells and activate STING nearby in paracrine fashion (123). cGAS-STING drives aging, senescence, anti-tumor immunity, autoimmune disease, and acute and chronic heart failure inflammatory disorders such as pancreatitis, macular degeneration, alcoholic hepatitis, cancer, metastasis, myocardial infarction, sepsis, systemic lupus erythematosus, and Parkinson's disease (116, 130).
cGAS-STING has been shown to be a major driver of organ pathology in COVID-19 infections (131).
3.1.3. Cancer
cGAS-STING surveils cancer cells. Mice and human cells deficient in cGAS-STING tolerate oncogenic Ras signaling. Neither irradiation nor cancer generate the usual inflammatory responses in cGAS-STING deficient cells. cGAS-STING activity correlates with the inflammatory component of many cancers (116).
cGAS-STING surveillance of micronuclei generates an immune response triggered by genomic instability upon breakdown of the micronuclear envelope (132). Micronuclei can be generated from lagging chromosomes.
cGAS-STING drives the IL-6 dependent survival of chromosomally instable cancers through a chromosomal instability/cGAS-STING/Stat3 /RelB /NF-κB/Stat1/ILK-6/IL-6R/JNK /ASK /cell death pathway (133). This mechanism was found commonly in many cancers that express IL-6R including triple negative breast cancer (133).
cGAS-STING also drives cancer metastasis. Chromosomal missegregation forms micronuclei that trigger cGAS-STING and downstream non-canonical NF-κB signaling that triggers metastasis (134).
In pancreatic cancer cGAS-STING-induced regulatory B-cells compromise NK cells anti-tumor immune response. cGAS-STING promotes immune evasion and metastasis through induction of PD-L1 expression (135). B-cells can also suppress tumor immunity through GABA signaling (136).
3.1.4. Aging
During senescence, reduced lamin B synthesis leads to partial breakdown of the nuclear envelope and blebs of nuclear membrane break off carrying chromosomes. These cytoplasmic chromosomes stimulate cGAS-STING (116). Similar nuclear blebs have also been observed from cannabis exposure (53, 55, 60).
Chronic immunostimulation is a prominent hallmark of aging, particularly the elaboration of the well-characterized Senescence Associated Secretory Phenotype (SASP) made up of growth factors and cytokines (114, 137). cGAS-STING recognizes cytosolic DNA fragments in senescence which generates the SASP and drives senescence with autocrine and paracrine effects (138). This was shown experimentally using irradiation and oncogene induction of senescence pathways (138).
cGAS-STING was shown to be essential for senescence induction. When it was inactivated senescence did not occur and mouse fibroblasts became oncogenically transformed more easily (139). cGAS-STING was required for the induction of senescence by irradiation and cytotoxic therapy. DNA damage produces cytoplasmic DNA which is sensed by cGAS. DDR is a key event in senescence induction. Lung cancers with low cGAS-STING expression have a worse prognosis (139).
YAP/TAZ activity in stromal cells controls cGAS-STING activity and this prevents aging (140).
3.2. Cannabinoid immune actions
Immune cells produce endocannabinoids (eCB) and also have receptors and metabolic machinery to transduce and extinguish eCB signals. This implies that eCB's signal through autocrine and paracrine routes to immune cells and their neighbors (141).
Toll-like receptors (TLR) are one of the major receptors of innate immunity and respond directly to a wide variety of exogenous DAMPs and PAMPs. Cannabinoids are generally suppressive of TLR signaling albeit there are many exceptions to this (141). Inflammatory signaling can either stimulate or inhibit ECS signaling via CB1R and CB2R signaling or by changing the levels of activity of eCB metabolic enzymes (141).
When THC was administered to homogenized cultures of rat telencephalic brain it was metabolized by mixed cultures of glia along with neurons (142). GABAergic neurons were most sensitive to a single application of 1 μM THC. After repeated applications of 1–2 μM THC markers of GABAergic, cholinergic and astrocytic damage were greatly elevated (142). IL-6 release was also elevated 4-fold after a single application of THC documenting important pro-inflammatory activity for this major cannabinoid (142).
Many other immune actions of cannabinoids are described which have been reviewed (143–156). Both immunosuppressive (143–145, 157, 158) and immunostimulatory (142, 159–161) actions have been described. Cannabidiol appears to be mainly immunosuppressive in its actions (156, 162). CB1R activation is often immunostimulatory (159–161). CB2R activation is frequently immunosuppressive (152, 163–165).
The immunosuppressive activities of cannabinoids are of clinical interest and importance. The immunosuppressive activities of cannabinoids have been shown to be relevant to different human disorders where heightened immune reactivity is problematic and complicate situations where immune compromise may be cause for concern. Experimental studies have explored the application of cannabidiol to skin graft rejection (166), to protect against the dietary and immune dysregulation induced by a high-fat diet (167), to protect against hypothalamic microgliosis and astrogliosis induced by a high-fat diet (168), to protect against experimentally induced oral mucositis induced by chemotherapy (169), to modulate TLR4 co-receptor signaling and thereby improve morphine mediated analgesia (170), and potentially as an adjunctive application in COVID-19-related cytokine storm (171). Contrariwise, cannabinoids have been found to exacerbate experimental sepsis (172) and to suppress anti-tumor immune responses by inhibiting T-cell Jak/Stat signaling (173).
Human studies have explored the use of cannabinoids in multiple sclerosis (156, 174–180). In general terms, cannabinoids have been found to reduce muscle spasticity (176–181). Cannabidiol has been shown to reduce leukocyte recruitment to plaque inflammatory lesions during peak disease activity (182). However, sedation, psychotomimetic, and gastrointestinal symptoms are often problematic side effects of cannabinoids, particularly in patients who are cannabinoid naïve (174, 176–180).
3.3. Long-lasting epigenetic effects in immune stem cells
Whilst there are numerous published examples of the epigenomic effects of cannabinoids on immune cells, detailed consideration of a few key examples illustrates the importance of this feature. One of the long-standing mysteries of adaptive immunity has been the way in which long-term memories of past antigenic exposure can be retained despite the fact that most immunocyte effector cells have only short half-lives. This mystery was elucidated recently by the demonstration that the memories are retained epigenetically in the epigenome of the long-lived immune stem cell population (183, 184). Epigenetic immunological memory storage has been demonstrated in macrophages of the bone marrow, tissue resident macrophages and microglia, monocytes, haemopoietic progenitors, natural killer cells, and innate lymphoid cells (183). Moreover, similar nuclear epigenetic storage of past cell experiences including toxic exposures has been demonstrated on cells of the skin, lung, intestine, pancreas, muscle cells, sperm, endothelial cells, nasal epithelia, Schwann cells, and neurons (183). Moreover, these epigenetic memories have been shown to affect the subsequent risk of diseases such as cancer and Alzheimer's disease. Further, encounters with stress (including the nutrient stress of famine) and environmental toxicants can be passed epigenetically from mother to child in utero (183). A broad spectrum of epigenetic machinery is implicated in encoding such memories (183).
One implication of the epigenomic recording of inflammatory memories has been demonstrated in haemopoietic stem cells which become exhausted after recording such encounters (185). This is the mechanistic basis of many of the hematological changes observed in aging including anemias, pancytopaenias, Age-Related Clonal Hematopoiesis (ARCH), leukemogenesis, fatty replacement of the bone marrow, and bone marrow hypocellularity. This study was performed over 1 year in mice, which is equivalent to about 25 years in humans. Thus, these past immune encounters were shown to drive hematological aging (185).
Through blood bank studies, ARCH has recently been shown to be an important risk factor for myocardial infarction, cancer, and all-cause mortality (186). Somatic mutations of DNMT3a are very common in ARCH, although ARCH does not arise for several decades after these are acquired. This has suggested to some investigators that an environmental or inflammatory factor may contribute to oncogenic progression. By studying a mouse line of DNMT3a knockout cells, it was shown that either chronic infection, or just an injection of interferon-γ by itself, was sufficient to cause hypomethylation, reduce cellular differentiation, and lower stress-induced apoptosis rates, which together account for the dominance of the DNMT3a mutant clones amongst myeloid populations during infections (186). Human DNMT3a mutant hemopoietic stem cells exhibit similar defective interferon-γ induced differentiation. These studies demonstrate the manner in which interferon-γ signaling during chronic infection can drive DNMT3a loss of function in ARCH (186).
3.4. Immune effects on stem cells
Immune activity has long been known to suppress stem cell regeneration. In part this can be a useful activity because it helps to maintain stem cells in quiescence and thus protects their genomic material from replication-associated damage. When DNA breaks were introduced into immune cells in mice, changes of accelerated aging were noted in the immune system and also prominently in the liver, lung, and kidney (187). Alterations included increased oxidative levels and hydoxynonenal, oxidized glutathione, and oxidative DNA mutations (187). Thus, immune aging was shown to be causal for systemic aging. Immune activity works partly by inducing oxidative stress which includes the production of DNA breakages and also by induction of the SASP.
Aging and senescent skeletal muscle stem cell niches are known to become inflamed, and this suppresses satellite stem cell regenerative capacity (188). Similarly, obesity also triggers a systemic inflammatory state especially in the abdominal fat. In obese mice, PPARγ is downregulated and its support restores TH2 differentiation of CD4 T-cells and abrogates highly inflammatory TH17 T-cell differentiation which is seen in obese rather than lean mice (189). Interestingly in this study the PPARγ receptor was noted to heterodimerize with the RXRα receptor.
3.5. Transposable elements
Repeat elements comprise 54% of the human genome and transposable elements (TEs) make up 46% of the total genome length (190). TEs may be either class I retrotransposons, which encode a reverse transcriptase, or class II DNA transposons, which do not. Retrotransposons may be short interspersed repeat elements (SINEs), long interspersed repeat elements (LINEs), or long terminal repeat/endogenous retroviruses (LTR/ERV). Epigenetic remodeling in the cancer landscape can lead to mobilization of the TEs which are also ligands of innate immunity particularly of RNA via TLR3, RIG1, and MDA5 receptors, of DNA via TLR3/7/8, and cGAS (190). cGAS also binds DNA:RNA hybrids as may commonly occur during transcription (190). Low levels of DNA methylation derepress retroelements which then begin to jump through the genome and activate innate immune signaling whilst doing so (190). It has been shown that this process is particularly strongly activated in cancer genome atlas samples from stomach, bladder, liver, and head and neck squamous cancers. However, as mentioned, these mobile elements are immunoreactive and this immunogenicity can be re-purposed to therapeutic advantage for the development of cancer therapeutics (190). These TEs can also be used as a source of adjuvant antigens for CAR T-cell development for personalized cancer treatments. As cannabis is known to hypomethylate the genome and induce DNA breaks, these activities are highly relevant to cannabis medicine.
Class I transposable genomic elements are retrotransposons and include endogenous retroviruses (ERVs) and long interspersed repeat elements (LINE-1) which encode a reverse transcriptase and are thus auto-transcribing and mobilize across the genome through an RNA intermediate (191). One such LINE1 element in mice, known as Lx9c11, controls a Lx9c11-RegoS non-coding RNA which in turn controls hyperstimulated immune responses to viral infection and rescues infective lethality (191). In this case this transposable element is functional and suppresses immune reactivity.
3.6. Cancer immunogenomics
The immune system can either promote or suppress tumourigenesis (192). In either case it is clearly a major force impacting and sculpting tumor development, including selecting for those clones which are able to evade antitumor immunity (192). Chronic inflammation is a well-known precursor to tumor development in many tissue beds (10, 20, 193).
It was shown in pancreatic carcinogenesis that recurrent pancreatic inflammation raises IL-6 released from macrophages, which drives the expression of early growth factor 1 (EGF1) in acinar cells, which is the master transcription factor for neocarcinogenesis and widespread field change (194). The effect of these epigenetic rearrangements is to reduce zymogen secretion with subsequent bouts of pancreatitis and thus the degree of tissue damage of subsequent inflammatory bouts. Induction of oncogenic Kras expression similarly reduced the tissue injury caused by pancreatic inflammation, a finding which suggests that oncogenes might be actively selected due to their anti-inflammatory effect (194). Both EGF and Kras were identified in the epigenomic cannabis screen of Schrott et al. (86).
Other researchers compared gene expression from tumors in mice from two genetic backgrounds with and without a mature immune system. Surprisingly they found that the tumors from the mice in which the immune system was intact were enriched in tumor suppressor genes which was interpreted as indicating that suppression of these genes was important to assist with immune evasion (195). One pathway by which this occurred was CCL2 secretion which attracted immunosuppressive M2 macrophages. CCL2 was positively identified in the epigenomic cannabis screen of Schrott et al. (86).
Age-Related Clonal Hemopoiesis (ARCH) has been described above. It has also been shown that similar effects with clonal dominance happen in nearly all tissue beds (196). It is known to occur often through epigenomic changes that may silence one of the DNA oxidoreductase demethylases TET2/3 or DNMT3A (197). The rate of cancer development in ARCH is greatly elevated and this particularly relates to leukemias, which are 12-times elevated. Of these, acute myeloid leukemia (AML) is a lethal disease carrying a very high mortality rate. Disordered innate immune signaling is common in the development of leukemia and pre-leukemias. TRAF6 is a downstream adapter known for transducing signals from the TNF receptor. It was shown that TRAF6 co-mutation along with TET2 dysfunction caused a highly aggressive and transplantable AML-like disorder in mice (197). TRAF6 was found to function as an E3 ubiquitin ligase which attaches a ubiquitin molecule to K148 of Myc and thereby competes with its activating acetylation at this residue which induces constitutive Myc oncogene activity. TRAF6K148Ub does not affect Myc protein stability. This becomes important as Myc is known to interact with histone acetyltransferases and induce its own acetylation in numerous cancers (197). TRAF6 reduction has been identified in some human AML cases and these cases have a worse prognosis than others. In some cases this is due to increased methylation of the TRAF6 promoter (197). Leukemic blasts and TET2/3 deficient cells are proinflammatory and this may account for the higher incidence of atherosclerotic plaque and myocardial infarction observed in ARCH patients. Thus, ARCH is both fed by and feeds inflammation (196). The effects of inflammation on Myc expression may also be more general (196). Of relevance to cannabinoid medicine, TRAF6 was positively identified in the human sperm cannabis exposure epigenomic screen from Schrott et al. (86).
It is known that many infiltrating immunocytes into cancers are functionally downregulated and exhausted. The cause for this immune exhaustion is not well-understood and appears to be multifactorial in origin. A systematic screen of T-cells in mouse and human tumor models found that two remodeling complexes in the SWI/SNF ATP-dependent chromatin remodeling families INO80 and BAF were involved in exhaustion induction (198). The Arid1a complex in the BAF family was particularly implicated in maintaining the exhausted phenotype. Arid1a depletion increased the chromatin accessibility of active T-cell genes and improved anti-tumor immunity (198). Importantly, one of the activities of the INO80 complex is to resolve R-loops (DNA:RNA hybrids) which would otherwise stall DNA polymerase and thereby allow uninterrupted proliferation of cancer cells (199). When R-loops occur in the cytoplasm they are highly immunogenic via cGAS-STING (200). There were 210 DMRs identified in the epigenomic screen of Schrott for Arid1, many hits for members of the BAF complex, and five hits for the INO80 complex (86).
Tumor-infiltrating fibroblasts have also been shown to induce immune exhaustion (201). Irradiation is known to provoke an inflammatory response. In a murine model of rectal cancer, irradiation caused IL-1α release which both polarized fibroblasts toward an inflammatory phenotype and triggered DNA damage thereby predisposing them to P53-mediated senescence which in turn caused chemoradiotherapy resistance and disease progression (201). Blockade of IL-1α, prevention of fibroblast senescence, or senolytic therapy (ablation of senescent cells) rendered the tumors radiosensitive. The main antagonist of IL-1α is IL1RA (IL-1 receptor antagonist). Patients with a lower IL-1RA level had a worse prognosis (201). Thus, this pathway shows both the importance of immune-fibroblast interactions and defines an important therapeutic target for future work. There were eight DMRs identified in the Schrott epigenomic screen for cannabis exposure (86).
Similar findings were made in a mouse model of pancreatic cancer (202). Healthy fibroblasts could be transformed into immunosuppressive cancer-associated LRRC15+ myofibroblasts under the influence of TGFBR2 signaling (202). These cancer-associated fibroblasts (CAFs) suppressed the cytolytic activity of infiltrating CD8 T-cells. Deletion of these CAFs restored tumor sensitivity to checkpoint PD-L1 inhibition (202). There were 152 DMRs identified in the Schrott screen for TGFBRs (86).
It has also been shown that mouse haemopoietic stem cells (which are also lymphopoietic stem cells) present antigens to CD4 T-cells on class II MHC antigens as a means of quality surveillance. Cells that do not pass these checks are induced to differentiate and thereby eliminated from the stem cell pool (203, 204). Similar pathways operate in human haemopoietic stem cells. These findings demonstrate bidirectional signaling between HSCs and surveilling immunocytes and this process is also applied—and perturbed—in cancer (204).
3.7. Other immune actions
Microglia are known to play an important role in sculpting away unused dendrites and synapses from neurons (205–209) and also in controlling the growth of myelin sheaths laid down by oligodendroglial cells and their progenitors (210). Lung injury including smoking and lung infections are known to exacerbate autoimmune central nervous system disease (211). Using a rat model of experimental allergic encephalitis (EAE), researchers were able to show that prior tracheal insufflation of neomycin completely blocked the effects of subsequent intrathecal vaccination with myelin basic protein and the development of EAE (211). However, if the neomycin was not given EAE predictably developed. Neomycin shifted the pulmonary microbiota toward lipopolysaccharide releasing phyla (211). Changing the flora with polymyxin B removed the lipopolysaccharide releasing microflora and exacerbated EAE. These lung microbiota changes were associated with an alteration of brain macrophage from type I interferon releasing to type II interferon priming. These results demonstrated the existence of a lung-brain axis in the same way as a gut-brain axis has been demonstrated (211).
4. Synapses—Activity dependent synaptic plasticity
4.1. Cannabinoid psychological epidemiology
Documentation relating to an association between poor physical and mental health, somnolence, and unemployability of hashish devotees has long been a major issue in traditional societies and together constitute the reasons cannabis was out of favor to varying degrees in Islamic communities from the 9th to 18th centuries (212). Cannabis was first noted to be linked with insanity in western medical literature as long ago as 1930 (212). More recently, cannabis has been linked with adverse mental health outcomes in diverse psychiatric illnesses including an amotivational state (213–218), anxiety (219–225), depression (226–238), bipolar disorder (239–246), schizophrenia (235, 247–260), and suicidality (222, 230, 240, 243, 261–267). Whilst in previous decades the nature of these associations was somewhat controversial, these matters are now settled in the medical literature. Cannabis is also associated with a dependency and withdrawal syndrome which is more common in daily smokers, those who commence regular use in the early teenage years, and those who use cannabis products with higher THC concentrations where its incidence may rise to 50% of regular users (35, 218). Most recently, cannabis has been linked with mass homicide attacks, especially in the USA (230, 268–276). As has been astutely observed, “cannabis consumption and mental illness in adolescents and young adults are increasing in the United States” (277).
Pediatric autism spectrum disorder (ASD) has also been shown to be growing exponentially across the USA since 2000 (278–280). Increasing evidence supports a link with rising cannabis use, availability, and potency (281–289) and indeed in formal space-time and quantitative causal analyses cannabis has been shown to be the primary driver of this modern autism renaissance (290). Although cannabis or cannabidiol is frequently advocated as potential treatments for ASD, formal clinical trials have only produced mixed and conflicting results (291, 292).
Given that the fundamental units of computation in the brain are variously described as the neuron, the neuronal epigenome, subcellular organellar networks, the synapse, the dendrite, local cortical micronetworks, midbrain internuclear signaling, the machinery of the synaptic boutons, post-synaptic densities, and their associated astroglia, and since cannabinoid signaling disturbs all of these to varying degrees, it is appropriate to consider these issues in the present context to advance understanding of cannabinoid neurological synaptopathies.
4.2. Classical cannabinoid actions
The brain has 100 billion neurons and 1015 synapses a number which far surpasses the number of nodes in artificial intelligence networks (293, 294). However, this number can be multiplied by the many astrocytes which are the most numerous cells in the brain and also participate actively in tripartite synaptic trafficking in many ways (295, 296). Dendritic spines are believed to be a key site of much brain computation, learning, and memory, and are involved in essentially all brain functions and numerous neuropsychiatric disorders including depression, schizophrenia, and autism (293, 297, 298). Dendritic spines comprise 70% of the synapses in the cortex and occur with a frequency of 10 per micrometer (293).
Endocannabinoid (eCB)-dependent synaptic plasticity is highly dependent on synaptic activity and traffic. This is referred to as Hebbian learning (299). Synaptic plasticity can be induced by many stimulation protocols, especially those which involve repetitive short latency firing or that are associated with dopamine transients (293, 300, 301). Many neuronal subcompartments are known to be controlled by activity including the dendritic arbors, synaptic spines, axon initial segments, and presynaptic boutons (302). Endocannabinoid (eCB) induction of long-term potentiation has been identified in multiple brain regions including the hippocampus, striatum, amygdala, nucleus accumbens, nucleus of the solitary tract, ventral tegmental area, cerebellum, and the prefrontal, somatosensory, visual, and insular cortices, and spinal cord (303, 304). Synaptic potentiation is normally tightly controlled in both space and time and happens over a range of 0.5–1 micron and over seconds (293). eCBs are also involved in synaptic scaling and metaplasticity which scale the plasticity of the whole system generally (303).
Cannabinoids have long been known to mediate depolarization-induced suppression of inhibition (DSI) at inhibitory synapses (which is excitatory) (305, 306) and have also been shown to induce depolarization-induced excitation (DSE) at excitatory synapses (which is inhibitory) (307). Short- and long-term potentiation and depression of synaptic transmission have now been demonstrated (STE, STL, LTD, and LTP) (305, 308, 309). LTP mediated via CB1R is induced primarily post-synaptically and is demonstrated by increased receptor numbers and spine size, spine head area, and spine volume. It requires extended eCB stimulation (303). It can be induced by many stimulation protocols but is especially sensitive to induction by dopaminergic stimuli (300, 303). LTP is dependent on protein synthesis (300, 310). eCBs induce modifications of the active postsynaptic density zone matrix (303).
Cannabinoids are also known to play a key modulatory role on brain development during the neonatal and postnatal critical periods when GABAergic synapses are sculpting the excitatory and inhibitory brain circuits (307). Projections neurons from striasomes in the striatum form baskets around clusters of neurons in the ventral aspect of the Substantia nigra pars compacta; these structures are known as “striosome-dendron bouquets” (311). CB1Rs are necessary for the proper formation of these bouquets (311). Actin and spectrin form a regular repeating structure on axons (312, 313) which has been shown by super-resolution microscopy to form the framework into which CB1Rs are fitted (314–317). CB1Rs also play a critical role in the stabilization of nascent immature spines (318). All of these GABAergic changes are modulated by eCBs and act on somatostatin- and parvalbumin- positive GABAergic interneurons and also astrocytes (307). eCBs therefore play a critical role in the excitatory/inhibitory balance in both health and disease during development and maturity (307).
The activity of many receptor types drives eCB release and activation including mGluR1, mGluR5, M1/M3 muscarinic, 5HT2 serotoninergic, cholecystokinin (CCK1), orexin (OX1) and oxytocin (OT1), D2R dopaminergic, and α2-adrenergic receptors (319). Reward behavior has also been shown to control the strength of the hippocampus-nucleus accumbens synapses and the strength of these synapses is reduced by stress (320).
Human cannabis use disorder has been shown to reduce brain glucose oxygen consumption by PET scanning in cerebellum, orbitofrontal, and prefrontal cortices and basal ganglia (321). This has been linked with deficits in striatal dopamine release in cannabis dependence (322, 323). Downregulation of brain CB1Rs also occurred to a degree related to the years of cannabis smoking, which reversed after 4 weeks of monitored abstinence (324). This is the molecular representation of cannabinoid tolerance at the nanoscale. Thinning of cortical gray and white matter has also been demonstrated in cannabis dependence (325). Damage to white matter was shown innervating the posterior cingulate and parietal cortex, the basal ganglia, and the temporal cortex. Gray matter in the precuneus was thinned in a causal and dose-response manner (325). Regions with higher MAGL expression (the catabolic enzyme for 2AG) in human postmortem brains had more gray matter damage.
An fMRI study of cannabis use disorder showed increased connectivity between subcortical nuclei in the ventral striatum (housing the nucleus accumbens), the midbrain (accommodating the ventral tegmental area and substantia nigra), and lateral thalamus and brainstem (325). The level of suppression was noted to be related to the years of cannabis use and the degree of negative emotionality, depersonalization, and social alienation and perceived persecution experienced by patients (325). Thus, the heightened subcortical connectivity was believed to generate the negative emotionality which the reduced cortical connectivity was unable to ameliorate (325). It was also shown that cannabis use disorder patients had more recruitment of cerebral cortex to complete neurally demanding tasks together with less discrimination between cognitive and emotional processing (326). Disrupted thalamocortical connectivity in cannabis use disorder was documented in a further study from this group (327). Suppressed nucleus accumbens activity was seen in cannabis use disorder together with depressed corticostriatal and thalamocortical connectivity (327).
Significant structural effects on the brain from cannabis have been found in a meta-analysis of MRI studies and include relative atrophy of the hippocampus and medial and lateral orbitofrontal cortices, which have been found with effect sizes measured as standardized mean differences of 0.14, 0.30, and 0.19, respectively (328).
The clear convergence in this data from the world leading group at Brookhaven National Laboratory's state of the art scanners between negative emotionality, psychological persecutory complexes, social alienation and estrangement, impaired cortical control of subcortical hedonic drives, and confusion between emotional and cognitive processes and difficulty performing neural computation tasks and the above described predisposition to self-directed and other-directed acts of violence is both noteworthy and of serious concern. Data strongly suggest a causal pathway to irrational, ill-considered, and confused psychological states and behaviors.
4.3. Non-classical endocannabinoid actions
Whilst eCB release is usually considered to be phasic and eCBs are described as being synthesized on demand, administration of eCB antagonists and genetic studies clearly demonstrates that there is also tonic eCB tone (305, 319). As well as classically described coupling of CB1Rs and CB2Rs to Gi proteins, non-classical coupling with GS and G0 G-proteins and β-arrestins is also documented (305). CB2Rs have also been identified on endoplasmic reticulum (305). When a newer and more sensitive marker for labeling astrocytes was employed, it was shown that 12% of the cerebral CB1R staining occurred on astrocytes (329). Astrocytes also display CB1Rs and eCB release onto astrocytes causes increased release of the gliotransmitters glutamate, D-serine, and adenosine, causing increased LTP (296, 305, 330).
4.4. Spine types
Synaptic spines can be of three types: filipodia, small, and large (293, 302). There is a tight relationship between spine structure and function with LTP increasing and LTD decreasing spine size (302). At the nanoscale, spines fluctuate in size over minutes, hours, and days, with the largest spines being the most stable (331). Filipodia are small thin spines which are usually transient with only 3% lasting for 1 day (293, 297). Spines are rapidly responsive to stimulation within 1 min (293). If they are contacted by an axon, they can become a small spine. Increasing evidence implicates pathology at spines in many diseases including anxiety, depression, autism, intellectual disability, schizophrenia, and bipolar disorder (302). LTP leads to an increase in the size, number, and stability of spines, actin polymerization, an increase in the number of AMPARs and their surface trafficking, and the size of the post-synaptic density matrix (302). NMDA glutamate receptors allow ingress of extracellular calcium and via calcicalmodulin kinase II (CAMK2) are key to the orchestration of the changes of synaptic growth and plasticity (293, 302). Compound synapses also occur of two to six synapses and are more stable than single synapses (293). Large synapses are more stable as they are further from the pruning boundary (293).
Structural studies show an increase in the number and density of spines in the frontal, temporal, and parietal lobes in Layer II in autism and a reduction in schizophrenia (302). GWAS studies of neuropsychiatric disorders most prominently show a reduction in postsynaptic glutamatergic signaling and also changes in cytoskeletal organization, chromatin modifiers, and transcription regulation (302).
Also key to the organization of the postsynpatic bouton are the scaffolding proteins including shank and homer; actin which forms the basic structural protein of the spine, including the “knob” and the neck of the bouton, arrests passing ribosomes toward the bouton, and directs microtubules into it to carry their mitochondria as cargo to supply energy (331); and PSD95/DLG4, which forms the plate-like area subjacent to the synaptic cleft into which AMPARs and NMDARs are inserted and bound (302, 331). Cofilin (CFL1 gene) is a key molecule with severs actin and induces the formation of branches and new actin polymers (331). PICK1 and oligophrenin (OPHN1 gene) are similarly key anchoring and actin controlling proteins (331). Calcium channels such as CACNA1C are clearly key since many of the changes described are calcium-induced, as are the small Ras GTPase family including Ras and RAC1 which induce spine formation, enlargement, maturation, and stabilization, and RhoA and RAP which have the opposite effect (302). RhoA and RAC1 also control the actin cytoskeleton. The Ras family is also a well-known oncogene (302). Neurexin and neuroligin are key transsynaptic scaffolding proteins (331). Shank proteins are highly implicated in autism spectrum disorder (ASD), which highlights ASD in part as a synaptopathy (332). Genes active in neurogenesis, chromatin modification, and synaptic function have also been implicated in autism (333).
It is therefore of interest to observe that many of these key receptors were positively identified in the epigenomic cannabis screen of Schrott et al. (86). including AMPARs (GRIA 132 DMR hits), NMDARs (GRIN 26 hits), metabotropic glutamate receptors (GRM 122 hits), GABA receptors (GABR 143 hits), dopamine receptors (DRD 17 hits), orexin/hypocretin receptors (HCRT 1 hit), μ-opioid (ORPM, 5 hits), and δ-opioid (ORPD, 5 hits). Of the scaffolding and framework molecules, the following were identified: actin (207 hits), tubulin (106 hits), cofilin (CFL1, one hit), PICK1 (one hit), oligophrenin (one hit), DLG/PSD95 (37 hits), SHANK (6 hits), HOMER (2 hits), calcium channel CACNA1 (46 hits), CAMK2 (5 hits), Ras (146 hits), RhoA (1 hit), RAP (438 hits), neurexin (27 hits), and neuroligin (10 hits) (86). When genes involved in depression were intersected with genes modified by cannabis, multiple genes were found in common in mouse, rat, and human studies (334). The autism candidate gene DLGAP2 was found to be altered in rat and human sperm after cannabis exposure (282).
4.5. Synaptoenergetics
Activities at the presynaptic terminal are generally heavily energy dependent. Terminal depolarization by an action potential causes calcium to flow into the terminal and the exocytosis of neurotransmitters from presynaptic vesicles (294, 335). Restoration of ionic gradients, synaptic cargo transport, synapse assembly and maintenance, pumping out presynaptic calcium, and synaptic vesicle refiling and recycling all require ATP. Indeed, it has been estimated that the recycling of a single glutamate synaptic vesicle requires 20,000 ATP molecules and that maintenance of the terminal at a steady state requires the presence of 1,000,000 ATP molecules (294).
Whilst the human brain is only 2% of body weight, it has been estimated to consume 20% of the energy budget. Glucose may be a major source of brain energy and 55% of the brain's glucose is consumed at axon terminals (294). Running the Na+/K+ and Ca2+ pumps which maintain the ionic gradients is a major energy requirement.
ATP is generated at the nanoscale in cellular microdomains. On demand glycolytic energy can be released from glycolytic enzymes mounted on the plasmalemma, within axonal transport cargos, and on synaptic vesicles to supply motor ATPases and ion transport ATPases with fast on-board refueling. Energy is also supplied from astrocytes by the lactate shuttle. Both mitochondrial oxidative phosphorylation and glycolysis can be upregulated in periods of intense activity (294).
A large calcium influx occurs into the presynaptic terminal with depolarization. This is taken up partly by the mitochondrial voltage-dependent anion selective channels and calcium uniporter of the inner and outer membranes and this has the effect of powerfully upregulating mitochondrial metabolism by stimulating its many calcium sensitive enzymes (294). Calcium also increases the surface area of cristae and cytochrome oxidase activity and induces a proteomic adjustment in favor of increased energy production. The anti-apoptotic protein BAD also stimulates mitochondrial biomass, enhances energy production, and reduces the leak from the F1F0-ATPase, thus enhancing the efficiency of energy production (294).
Mitochondrial biogenesis is a key mechanism by which mitochondria undergo fission and fusion in order to increase energy delivery to the presynapse. Mitochondria are trafficked along dendrites and axons and are typically carried by a kinesin at one end and a dynein at the other, which move the mitochondria toward the plus and minus ends of the microtubule, respectively (294). Mitochondria trafficking along microtubules are arrested by local high calcium levels. Short-range movements of mitochondria within the spine are conducted by myosin motors (294).
Neuronal synaptic terminals possess a master energy sensor called AMP kinase (AMPK) which drives many of the changes to increase energy production for periods of high energy requirements (294). AMPK has multiple presynaptic activities including adapting glycolysis and mitochondrial respiration, sustaining LTP and high frequency stimulation, facilitating mitochondrial transport and presynaptic distribution, and driving the accumulation of mitochondria within the axonal compartment. Importantly, this is the same key molecule that is well-known to sense energy and is a key controller of lifespan in lower organisms (72, 336–348).
Thus, mitochondria play a dual role in the presynaptic terminal by virtue of their interrelated roles for energy supply and calcium buffering. Failure of synaptoenergetics has been shown to play a key role in diseases such as Alzheimer's and Parkinson's disease and in motor neuron disease (294).
Two of the key mitochondrial fusion proteins, OPA1 and MFN1, were identified in the cannabis epigenomic screen of Schrott with three and one DMR identified, respectively (86). There were 217 DMRs in the Schrott screen for kinesins, 16 for dyneins, and four for dynactin, the regulatory partner of dyneins (86). AMPK was identified in the Schrott screen with five DMR hits (86).
Endoplasmic reticulum (ER) also play an important role in presynaptic calcium control as they are able to actively acquire calcium from the cytoplasm and hold the largest cellular stores of this cation (349). Moreover, ER are electrically excitable and can be triggered by action potentials (349). ER are also a major site of protein synthesis as they house ribosomes. Lysosomes also concentrate calcium, and the release of the calcium from just one lysosome has been shown to elevate cytosolic calcium by two orders of magnitude (349).
4.6. Brain-derived neurotrophic factor—Extracellular vesicles
Extracellular vesicles (EVs) are known to originate in intracellular endosomes.
Brain EVs promote communication between neurons and non-neuronal cells, prune excitatory synapses, and modulate inhibitory synapses the transmission of viruses, and prion protein aggregates in preclinical Alzheimer's disease models (350, 351).
They are also useful for critical events such as during rapid maturation of the neuronal proteome such as in dendrite maturation and growth cone steering (350).
BDNF modulated both eCB LTD and LTP in the hippocampus, neocortex, ventral tegmental area, and striatum via its high affinity TrkB receptor. BDNF stimulates 2-AG release and CB1R activation and DSI causing increased glutamatergic LTP in the neocortex (304). BDNF release leads to the release of neuronal vesicles from the post-synaptic terminal carrying miRNAs (miR-132-5p, miR-218-5p, and miR-690) which induce clustering of the synaptic vesicle in the presynaptic terminal, increased calcium transients in the post-synaptic terminal, regulate the abundance of synaptogenic transcripts, increase BDNF-dependent dendrite outgrowth, increase synapse density, and increase neuronal network bursting and synchronized activity and connectivity (350).
BDNF was necessary and sufficient for the formation of excitatory synapse formation in the hippocampus. EVs induce synchronous neuronal network firing (350).
Of relevance to cannabinoid medicine, THC has been shown to epigenomically downregulate BDNF mRNA and protein expression in the ventral dentate gyrus (352). BDNF was identified by three DMRs in the Schrott cannabis epigenomic database and the high affinity BDNF receptor TrkB (gene NRTK2) was identified by 33 DMRs in this database (86).
4.7. Epigenetic mechanisms
From the above discussion, it is clear that whilst topographically the synapse is considered the unit of computation, changes that are localized to the synapse must clearly be coordinated with events in the cell. As mentioned, the apparatus of the synapse involves over 500 proteins which must be synthesized locally or imported and there is a high demand for local energy. Therefore, ribosomes and mitochondria are re-routed as they travel along dendritic microtubules and imported into larger spines (293). It has been convincingly demonstrated that memory formation (involving spine enlargement, consolidation, and maturation) requires protein synthesis and frequently occurs overnight (293, 310). Spine enlargement and maturation involves several signaling cascades including changed distribution of glutamate receptors and synaptic adhesion proteins (18).
From these observations, it becomes clear that epigenomic controls of protein expression must be coordinated with events at the synapse for the whole system to work cohesively. That is, memories which are encoded epigenomically must be coordinated and synergize with local synaptically encoded memories (18). Thus, disorders of DNA methylation, histone methylation, histone acetylation, CBP, CREB1, and HDAC5 have all been linked with impairments of learning and memory (18).
Signaling from the synapse to the nucleus can occur via activation-induced CREB-regulated transcriptional coactivator 1 (CRTC1). Strong training also induces the brain-specific FGF1B, which is required for CA3-CA1 learning in the hippocampus (18). FGF1B transcription in the nucleus is induced by CRTC1, which displaces the transcriptional co-repressor HDAC3 nuclear receptor corepressor (NCOR) complex, leaving phosphorylated CREB-CBP to bind in its place, thereby activating FGF1B transcription (18). Strong training-induced continued CRTC1 expression displaced CBP and drove the insertion of the HAT KAT5, which induces sustained FGF1B expression. KAT5 substitution was required for hippocampal synaptic plasticity and memory enhancement (18). Afadin is another protein that shuttles from the synapse to the nucleus and induces epigenetic change, which in this case is H2S10 which opens heterochromatin to allow a more permissive chromatin state for gene transcription (18). From remarks made above, it is also likely that extrasynaptic vesicles also traffic information into the dendrite and neuronal soma.
It is noted that, in the Schrott cannabis epigenomic screen, there were two hits for CRTC1 and six hits for NCOR (86).
Signaling from the nucleus to the synapse occurs by virtue of control of AMPAR synthesis during critical periods, stress, and drug exposure. Stress-induced cortisol release which increased HDAC2 occupancy of the G9a promoter and impacts on E3 ubiquitin ligase leads to increased ubiquitinoylation of AMPARs and their down regulation. Similarly, an epigenetic cascade of histone acetylation and H3K9 trimethylation can control the alternate splicing of neurexin-1 which controls its affinity for post-synaptic binding partners and thus synaptic activity and remodeling (18).
In this way detailed, complex, and profound coordination between the epigenomic machinery of the nucleus and the structural machinery of the synapse is orchestrated.
One elegant example of these changes is with PRKCZ. PRKCZ has been well-characterized as being involved in long-term potentiation and synaptic strengthening (353). PRKCZ abnormalities have been implicated in alcoholism, depression, and Alzheimer's disease. DNA methylation of the internal intronic promoter carrying a CREB -binding site of the human brain-specific PRKCZ gene was noted to be controlled by DNA methylation (353). Methylation of this site caused reduced CREB binding and downregulation of PRKCZ transcription (353). CREB has been implicated in both short- and long-term synaptic potentiation and adult neurogenesis and controls cassettes of genes that are involved in these processes (353). DNA methylation prevents CREB docking at the site (353). Similar control mechanisms were found in three other genes involved in long-term potentiation and neuronal differentiation from stem cells (LRRTM2, NEUROD2, and FAM163B) (353). PRKCZ recruits AMPARs to the post-synaptic density and LRRTM2 stabilizes them there.
PRKCZ was identified by two DMRs in the Schrott cannabis epigenomic screen (86).
4.8. Clinical implications
It is thus apparent that cannabinoids can modulate brain synaptic signaling at many levels including neuronal, neuronal mitochondria, neuronal and astroglial epigenomic, subcellular trafficking, and pre- and post- synaptic levels. and modulate neurotransmission in several subtle and classical directions including impacting both long-term synaptic depression and potentiation. Therefore, these various and interrelated mechanisms form the neurobiological substrate for the now documented widely diverse clinical syndromic phenomenology of adult and inheritable pediatric neurotoxicity described in the opening paragraphs.
With this widespread disruption of neural activity at many levels, it becomes clear how cannabis exposure might be related to so many diverse neuropsychiatric syndromes including depression (227, 229, 230), anxiety (224, 225, 230), schizophrenia (354–358), bipolar affective disorder (239, 240, 245), and autistic spectrum disorder (278, 279, 290, 359).
The putatively causal pathway from confused thinking, heightened subcortical hedonic drive, impaired cortical control of subcortical activity, and underlying highly negative emotional state with advanced social alienation and persecutory perseverative thinking is of particular concern in terms of contemporary issues with cannabis-related violence, suicides, and homicides. In that this pathway has been well-described in reports from the Brookhaven National Laboratories an internationally pathfinding group in the brain scanning of neuropsychiatric disorders (323, 325–328, 360–363), this evidence can only be considered state of the art.
It is also of great interest that AMPK, which is one of the best established molecules with an important role in aging medicine (364–366), is also a key and central regulator of synaptoenergetics at the presynapse (294).
A real concern has been expressed that overwhelming this delicate synaptic and epigenomic machinery of DNA methylation, histone modifications, and microRNAs with exogenous phytocannabinoids will disrupt the delicate spatiotemporally defined patterns of eCB-controlled systems which maintain normal brain function and development (367). Given that mitochondria play such a key role in synaptoenergetics, it is clear that the general disruption of mitochondrial metabolism will necessarily perturb both neuronal and synaptic activity and epigenomic regulation. Given that synaptic dysregulation is a key component of most neuropsychiatric disorders, these disruptions are likely to have far reaching and protean manifestations (293, 297, 298). In view of the fact that both the brain and mitochondria also control systemic aging and disease (364, 365) these impacts are likely to ramify beyond even psychiatry to the general domain of whole-body health.
5. Conclusion
This wide-ranging overview considered in Parts 1 and 2 of this review has considered the manner in which interrelated, interlocking, and interdependent deleterious changes have been shown to be induced by cannabinoids in multiple mechanistic layers including the genome, epigenome, metabolome, and immunome. Changes in synaptic plasticity illustrate and highlight these metabolic, proteomic, and epigenomic alterations. It is emphasized that these changes do not occur in isolation but in a coordinated cross-platform manner. A striking feature of the data is its cross-disciplinary concordance so that, for example, CB1R downregulation occurring in cannabis dependence and first demonstrated by autoradiography in brain slices has now been confirmed by super-resolution microscopy and has been shown to be reflected in blunted hedonic drive in the midbrain and ventral tegmental area and in inhibited cortical connectivity from the frontal lobes. Synaptic alterations are necessarily supported by metabolic and epigenomic alterations and synergize with altered dopaminergic drive, all of which impact on learning, emotionality, behavior, and memory. The evidence for altered and confused behavior including persecutory and emotively driven thinking from the Brookhaven scans is relevant to modern considerations of cannabis-associated self- and other- directed violence.
There is an impressive synergy between the altered immunome, metabolome, and epigenome that fits with the altered gene expression seen in clinical syndromes such as diverse cancers and numerous congenital anomalies identified in modern large-scale epidemiological studies (2–20, 69, 368–385) and the genomic, epigenomic, metabolomic, and immunomic changes of aging (386–391) which together reflect altered and disrupted epigenomic-metabolomic regulation. Clearly, the well-established inhibition of mitochondrial oxidation is compounded by the immunostimulatory CB1R-mediated cannabinoid actions and together they have greatly amplified downstream effects.
Thus, it is important to hold a proper appreciation of the epigenomically mediated alterations of three-dimensional genomic architecture, gene expression, and epitranscriptome in mind in order to foster a comprehensive appreciation of multifaceted and diverse cannabinoid pathophysiology. Moreover, given the rapid growth of the cannabis industry worldwide, there is a pressing and urgent need to complete single cell three dimensional chromosomal conformation capture (low input Hi-C methods) and associated mass spectrometry histone modification, DNA methylation, whole genome, whole epigenome, enhancer, superenhancer, superanchor, and m6A epitranscriptome studies in many brain areas, and in the heart, liver, respiratory tract, immune, muscle, testicular, and ovarian tissues in a timely manner.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Glossary
NF-κB, Nuclear Factor kappaB, the key master transcription factor of the immune system; ROS, Reactive Oxygen Species; also known as oxygen radical; includes hydrogen peroxide, superoxide anion, and the hydroxyl radical; HIF1α, Hypoxia Inducible Factor 1; FGF21, Fibroblast Growth Factor 21; GDF15, Growth Differentiation Factor 15; SIRT1, Silencer of Information Transfer 1; AMPK, Adenosine Monophosphate Kinase; TFS1, Stress-activated transcription factor ATFS-1; ATFS-1, Activating Transcription Factor Associated With Stress-1; GPR, G-Protein Receptor; HDAC, Histone Deacetylase; HMGB1, High Molecular Weight Group Box protein 1; ARG1, Arginase 1—catalyzes the hydrolysis of arginine to ornithine and urea; CCL5, C-C Motif Chemokine Ligand 5; SLC16A, Solute Carrier Family 16, Member 1 (Monocarboxylic Acid Transporter 1); NFAT, Nuclear Factor of Activated T Cells; VEGF, Vascular Endothelial Growth Factor; Kla, K is the abbreviation for lysine and la stands for lactate; ACAT1, Acetyl-CoA Aceteyltransferase 1; CXCL9, C-X-C Motif Chemokine Ligand 9; cGAS-STING, cyclic AMP—cyclic GMP—Stimulator of INterferon Gamma (STING); R-loops are DNA, RNA hybrids and can arise during DNA transcription; Stat, Signal Transducer and Activator of Transcription; RelB, RELB Proto-Oncogene, NF-KB Subunit; NF-κB, Nuclear Factor kappa B; JNK, c-Jun—N-terminal kinase; ASK, Apoptosis Signal Regulating Kinase 1/Mitogen-Activated Protein Kinase Kinase Kinase 5/MAPKKK5/MAP3K5; ABA, Gamma Amino-Butyric Acid; DAMP, Damage-Associated Molecular Pattern; PAMPm, Pattern-Associated Molecular Pattern; ARCH, Age Related Clonal Haematopoiesis; SASP, Senescence-Associated Secretory Phenotype; LR, Toll-Like Receptor; RIG1, Retinoic Acid Inducible Gene 1; MDA5, Melanoma Differentiation-Associated Protein 5; cGAS, cyclic GMP – cyclic AMP Synthase; CAR T-cells, Chimeric Antigen Receptor T-cell lymphocytes; CCL2, C-C Motif Chemokine ligand/MCP-1, Macrophage Chemotactic Protein 1; TRAF6, TNF Receptor-Associated Factor 6; SWI/SNF, SWItch/Sucrose Non-Fermentable chromatin remodeling complexes; INO80, INO80 Complex ATPase Subunit; BAF complex, the mammalian equivalent of SWI/SNF in yeast; Includes BANF1, BAF Nuclear Assembly Factor 1; ARID1a, AT-Rich Interaction Domain 1A; LRRC15, Leucine-Rich Repeat Containing 15; TGFBR2, Transforming Growth Factor Beta Receptor 2; mGlu1, Glutamate Metabotropic Receptor 1/GRM1; MAGL, MonoAcyl Glycerol Lipase the enzyme which metabolizes 2-arachidonyl glycerol; AMPAR, α-Amino-3-hydroxy-5-Methyl-4-isoxazole Propionic Acid glutamate Receptor; PSD95, Post-Synaptic Density 95; DLG4, Discs Large MAGUK Scaffold Protein 4; PICK1, Protein Interacting With PRKCA 1; CACNA1C, Calcium Voltage-Gated Channel Subunit Alpha1 C; Ras, KRAS Proto-Oncogene, GTPase; RAC1, Rac Family Small GTPase 1; RhoA, Ras Homolog Family Member A; RAP - RAP, Member of RAS Oncogene Family; AMPK, AMP kinase/PRKAA1, Protein Kinase AMP-Activated Catalytic Subunit Alpha 1; OPA1, OPA1 Mitochondrial Dynamin Like GTPase / Optic Atrophy 1 (Autosomal Dominant); MFN1, Mitofusin 1; BDNF, Brain-Derived Neurotrophic Factor; CBP, CREB Binding Protein; HDAC5, Histone DeACetylase 5; FGF1B, Fibroblast Growth Factor 1B; PRKCZ, Protein Kinase C Zeta; CREB1, cAMP Response Element Binding Protein 1; LRRTM2, Leucine Rich Repeat Transmembrane Neuronal 2; NEUROD2, Neuronal Differentiation 2; FAM163B, Family With Sequence Similarity 163 Member B.