Fighting the storm: novel anti- TNFα and anti-IL-6 C. sativa lines to tame cytokine storm in COVID-19

The main aspects of severe COVID-19 disease pathogenesis include the increasing hyper-induction of proinammatory cytokines, also known as ‘cytokine storm’, that precedes acute respiratory distress syndrome (ARDS) and often leads to death. COVID-19 patients often suffer from lung brosis, a serious and untreatable condition. There remains no effective treatment for these complications. Out of the cytokines, TNFα and IL-6 play crucial roles in cytokine storm pathogenesis and are likely responsible for the escalation in disease severity. These cytokines also partake in the molecular pathogenesis of brosis. Therefore, new approaches are urgently needed that can eciently and swiftly block TNFα, IL-6, and the inammatory cytokine cascade in order to curb inammation and prevent brosis, and lead to disease remission. cytokine storm and ARDS pathogenesis.


Introduction
To date, raging pandemic of COVID-19 disease caused by the SARS-CoV2 virus has affected over 4.7 million people and claimed over 310,000 lives worldwide. SARS-CoV2 has human-human transmission and spreads easily via airborne and contact routes; its R 0 is currently estimated to be 2-2.5 1 . COVID-19 has a rather broad spectrum of clinical manifestations, ranging from asymptomatic, to mild u-like disease, to pneumonia, that in some cases can further progress to acute respiratory distress syndrome (ARDS), major organ failure and death. Approximately 20% of COVID-19 cases are serious or severe, and death rate is currently estimated to be around 10%. While elderly and individuals with pre-existing conditions are among the most affected, it has recently become apparent that COVID-19 affects all age groups.
Of the cytokine milieu, TNFα and IL-6 play key roles in cytokine storm and are likely to be responsible for the escalation in disease severity [8][9][10] . TNF alpha is an in ammatory cytokine that stimulates and maintains cellular activation and migration of leukocytes to in ammatory sites. TNF acts though binding to its receptors (TNFR) that are located throughout the body. Interaction of TNF with receptors causes increased expression of other cytokines (IL-1 and IL-6) and chemokines, which, in turn, activate leukocytes, suppresses regulatory T cells, causes production of MMP proteins which degrade tissues and induces apoptosis 11 . IL-6 is another important player in the acute host response to infection whereby it promotes in ammation, immune reactions, and hematopoiesis. Long-term elevation of IL-6 levels maintains chronic in ammation and autoimmunity, making IL-6 one of the main druggable targets in autoin ammatory and autoimmune disorders 12 .
Even though TNFα and IL-6-mediated cytokine storm and ARDS has been previously well-documented in SARS, MERS, as well as in severe cases of in uenza 3,13 , there still is no effective treatment for this grievous complication. Therefore, new approaches are urgently needed that can e ciently and swiftly block TNFα, IL-6 and in ammatory cytokine cascades and thus curb in ammation and lead to disease remission.
Furthermore, COVID-19 convalescents face a long recovery and may be at risk of developing pulmonary brosis (PF), a debilitating complication that is very hard to treat 14 . Mechanisms of PF are not fully understood, albeit it has been established that in ammatory cytokines and chemokines, such as IL-1, IL-6, TNFα, C-C motif chemokines are important in its etiology 15,16 . New therapies are much needed to prevent and mitigate pulmonary brosis complications in COVID-19 patients. Since COVID-19, and especially ARDS patients are extremely weak and vulnerable, it would be crucial that novel anti-cytokine storm and anti-brosis therapies have minimal side effects.
Cannabis sativa has been proposed to modulate gene expression and in ammation and is under investigation for several potential therapeutic applications against autoin ammatory diseases and cancer. Here, we hypothesized that extracts of novel C. sativa lines may be used to modulate expression of pro-in ammatory cytokines and pathways involved in in ammation and brosis.

Results And Discussion
To analyze the anti-in ammatory effects of novel C. sativa lines, we used a well-established full thickness human 3D skin arti cial EpiDermFT TM tissue model, whereby tissues were exposed to UV to induce in ammation and then treated with extracts of seven new cannabis lines.
Global gene expression pro ling revealed that several new extracts strongly down-regulated expression of interleukins, pro-in ammatory cytokines, C-C motif chemokines and C-X-C subfamily cytokines involved in ADRS and other autoin ammatory conditions (padj<0.05) ( Fig. 1 and Table 1).
We further explored the effects of C. sativa extracts on the levels of IL-6 protein using western immunoblotting, and found that all extracts, except #15, downregulated IL-6 ( Fig. 2). Interestingly, application of extract #12 down-regulated IL-6 on the protein level, but not on the level of the transcript. This is an interesting nding that may suggest the presence of post-transcriptional regulation of IL-6 expression via small interfering RNAs and the potential effects of cannabis extracts on these processes. All extracts down-regulated an important in ammation marker, COX-2 (Fig. 2).
Along with two key regulators of cytokine storm -TNFα and IL-6, C. sativa extracts also affected the levels of other key pro-in ammatory interleukins -IL-1, IL-17, IL-23 ( Fig. 1, Table 1). IL-1 family of interleukins is important in innate in ammation and autoimmunity 17 . IL-1α was shown to be constitutively present in numerous epithelial and mesenchymal cell types of healthy individuals, whereas IL-1β is mainly induced under disease conditions 17 . Both pro-in ammatory interleukins are upregulated in numerous in ammatory and autoin ammatory diseases and are important druggable targets. Recent studies show that levels of IL-1 were strongly elevated in individuals with COVID-19, and IL-1 levels correlated with disease severety 18 . Here, we noted that extracts #4 and #8 down-regulated both IL-1α and IL-1β ( Fig. 1, Table 1).
While on the one hand, the IL-17 family confers protection from a variety of extracellular pathogens and was shown to drive leukocyte in ltration to facilitate clearance of infectious pathogens, aberrant IL-17 signaling can lead to excess in ammation and tissue damage and brosis 22 , and has been implicated in ARDS, cystic brosis, and pulmonary brosis and other pathological conditions (reviewed in 22 ).
In addition, extracts #4, #8, #13 and #14 signi cantly down-regulated the expression of NFKB2 gene. NF-κB pathway has been often referred to as a prototypical proin ammatory signaling pathway. NF-κB is usually upregulated by IL-1 and TNFα, and play important roles in the expression of other proin ammatory genes 25 .
Further analysis revealed that extracts #4, #8, #13 and #14 down-regulated CCL2, also known as MCP-1 ( Fig. 1 and Table 1). Along with in ammation and ARDS, CCL2 expression is an important hallmark of brosis, and CCL2 has been explored as a potential druggable anti-brotic target 26 . In previous studies, CCL2 was shown to promote broblast differentiation and facilitate their recruitment to the alveolar space, thus leading to excessive collagen deposition 27 . Besides, CCL2 promoted broblast survival and stimulated IL-6 production 28 .
Importantly, along with CCL2, IL-1, IL-6 and TNFα also regulate brosis 26 , and their down-regulation may be viewed as a potential anti-brotic effect. Together with IL-1, IL-6 and TNFα genes, novel cannabis extracts regulated the expression of various other genes involved in brosis, including pulmonary brosis (PF) ( Table 1). Among those were metalloproteinases (MMPs), key proteases involved in ECM remodelling 26 . MMP1, MMP2, MMP7, and MMP9 were previously reported to be upregulated in PF. In our study, several extracts down-regulated MMPs (Table1).
Extracts #4, #6, #8, #14 and #13 also down-regulated WNT2 and WNT5a. WNT signaling alterations have been linked to pathogenesis of a variety of diseases and conditions, including pulmonary brosis 29,30 . Previous studies have shown that in vivo inhibition of WNT-5A attenuated tissue destruction, improved lung function and restoration of alveolar epithelial cell markers expression in two animal models of COPD 29,31 . Furthermore, extracts also affected the levels of iCAM1 and iCAM5 genes. Levels of iCAM1 were shown to be elevated in sera of PF patients 32 , and recent studies showed that iCAM-1 inhibition reduced exacerbations of lung in ammation 33 .
One more important pro-brotic protein is CXCL12, and its down-regulation was shown to dampen brocyte recruitment and collagen deposition 34 . In our study, extracts #6 and #13, along with downregulation of numerous pro-in ammatory cytokines, upregulated CXCL12. The role of CXCL12 upregulation in PF still needs to be fully established, but, based on the current knowledge, CXCL12 upregulation can be viewed as a potential PF contributor, and thus its upregulation may negate the potential bene ts of cytokine down-regulation by extracts #6 and #13.
Having seen cannabis extract-induced changes in pro-in ammatory and pro-brotic genes, we further conducted an in-depth analysis of the effects of the extracts on global signalome using Pathview Bioconductor platform. We found that extracts # 4, #8, #14 signi cantly down-regulated cytokinecytokine receptor interaction pathway, rheumatoid arthritis pathway, chemokine signalling, Toll-like receptor signalling, JAK-STAT signalling and other pathways involved in in ammation, immunity and autoimmunity, as well as tissue remodeling and brosis. Contrarily, extract #12 upregulated these pathways (Table 2, Fig. 3).
Overall, our study revealed that cannabis extracts exerted different effects on the 3D tissue in ammation model -some profoundly down-regulated pro-in ammatory cytokines and pro-brotic molecules, some affected only several key cytokines, some did not cause any signi cant changes at all (extract #15), while extract #12 promoted expression of pro-in ammatory genes. This is a very important nding that shows that cannabis is non-generic, and each C. sativa line has to be thoroughly evaluated for its medicinal properties.
Taken together, our results suggest that out of 7 studied extracts of novel C. sativa lines three were most effective down-regulating pro-in ammatory pathways and key cytokines implicated in the cytokine storm and ARDS in COVID-19. Extracts #4, #8 and #14 were the most effective, causing profound and concerted down-regulation of TNFα, IL-6 and CCL2.
Pronounced inhibition of TNFα and IL-6 is the most important nding, as these molecules are currently considered to be the key actionable targets in COVID-19 cytokine storm and ARDS. Anti-cytokine therapies are thought to be important for prevention of COVID-19 pneumonia 35 , as currently there is a race to develop novel anti-cytokine storm regimens. To that effect several anti-cytokine therapies have been proposed and are now in clinical trials. These include anti-IL-6 receptor antibody tocilizumab 9,10,36 , colchicine, an agent that can potentially in uence levels of IL-6 and other cytokines 37 , chroloquine 13 , metronidazole 38 , and statins 39 , as well as melatonin as an antiin ammatory adjuvant therapy 4 . Chloroquine has some immunomodulatory effects, potentially suppressing the production and release of TNF-α and IL-6 13 . Colchicine has been shown to effectively suppress interleukin IL-1b, IL-18 and IL-6 in patients with acute coronary syndrome 40,41 and is now being trialed in COVID-19 ARDS, albeit it also has very signi cant side effects 37 .
Numerous rheumatological drugs are now being evaluated for therapeutic potential to tame COVID-19 pneumonia, ARDS, and prevent further complications such as PF 18 . Suppression of pro-in ammatory IL-1 family members and IL-6 have been shown to have a therapeutic effect in many in ammatory diseases, including viral infections, and has been explored as a potential therapeutic avenue in COVID-19 42 . Tocilizumab, a humanized monoclonal antibody against the interleukin-6 receptor, is showing some promise, albeit it carries a hefty price tag and a lot of side effects 10,43 .
TNFα not only is the main cytokine storm driver, it also was shown to mediates the transition from pulmonary in ammation to brosis 44 . Surprisingly, up to now, no TNFα inhibitors have been trialed for COVID-19. The recent expert commentary in Lancet stated that "trials of anti-tumour necrosis factor therapy for COVID-19 are urgently needed" 45 . While potentially effective, anti-TNFα and anti-IL-6 biologics are very expensive and cause an array of side effects, including malignancies.
On the other hand, anti-TNFα and anti-IL-6 cannabis extracts that are generally regarded as safe (GRAS) modalities can be a useful addition to the current anti-in ammatory regimens to treat COVID-19, as well as various rheumatological diseases and conditions such as rheumatoid arthritis (RA), psoriasis and psoriatic arthritis, osteoarthritis, bromyalgia and others. Indeed, lines targeting TNFα, IL-6, IL-1β and causing concerted and signi cant down-regulation of the rheumatoid arthritis pathway, pending thorough veri cation and clinical validation, may present a novel and promising natural resource for RA treatments and management of other TNFα, IL-6, IL-1β-mediated diseases.

Conclusions
Overall, we are the rst to show that application of C. sativa extracts profoundly decreases the level of pro-in ammatory cytokines in human 3D tissues. Still, our study has several pitfalls. Here, we used human 3D full-thickness skin model to analyze the effects of cannabis extracts on in ammation and brosis. While it would be important to replicate the data in an airway epithelial and alveolar tissue models, our data can be used as a roadmap for the future analysis. Moreover, key fundamental mechanisms of in ammation and brosis are similar in various tissues, and key roles of TNFα, IL-6 and other interleukins, chemokines, and MMPS have been well-established in an array of broproliferative diseases 15 . Pending further validation in lung tissue models, our novel extracts need to be studied in a clinical trial aimed to prevent or mitigate COVID-19 pneumonia and ARDS. To do so, the extracts have to be administered early upon positive diagnosis has been made to allow su cient time for modulation of cytokine levels.
Most importantly, out of 7 selected extracts, only 3 performed best, one had no effects at all, and one exerted effects that may in turn be deleterious, signifying that cannabis is not generic and cultivar selection must be based on thorough pre-clinical studies. Furthermore, the current study was developed to analyze the effects of medical cannabis applications rather than smoking.
In the future, anti-TNFα and anti-IL-6 extracts need to be analyzed for their potential to mitigate in ammation in rheumatoid arthritis, ankylosing spondylitis, and other rheumatologic conditions, especially given the fact that extracts profoundly downregulate the RA pathway and target TNFα and IL-6. Also, the effects of novel extracts also need to be analyzed for their potential to combat 'in ammaging'the in ammatory underpinning of aging and frailty 46 .

Materials And Methods
Plant growth, extract preparation: All cannabis plants were grown in the licensed facility at the University of Lethbridge (license number LIC-62AHHG0R77-2019). C. sativa lines #4, #6, #8, #12, #13, #14, #15 were used for the experiments. Four plants per line were grown at 22°C 18 h light 6 h dark for 4 weeks and then transferred to the chambers with 12 h light/ 12 h dark regime to promote owering. Plants were grown to maturity and owers were harvested and dried. Flower samples from four plants per variety were combined and used for extraction. Three grams of the powdered plant tissue per each line were used for extraction. Plant material was placed inside a 250 mL Erlenmeyer ask, 100 mL of Ethyl Acetate was poured into each ask. The asks were covered with tin foil and incubated overnight in the dark at 21C with continuous shaking at 120 rpm. Extracts were ltered, concentrated using a rotary vacuum evaporator and transferred to a tared 3-dram vial. The leftover solvent was evaporated to dryness in an oven overnight at 50 o C to eliminate the solvent completely. Levels of cannabinoids was analysed using Agilent Technologies 1200 Series HPLC system. The extract stocks were prepared from the crude extracts whereby 3-6 mg of crude extract were dissolved in DMSO (Dimethyl sulfoxide anhydrous, Life Technologies) to reach 60 mg/mL nal concentration and stored at -20 o C. Appropriate cell culture media (RPMI + 10% FBS or EMEM + 10% FBS) were used to dilute the 60 mg/mL stock to make working medium containing 0.01 mg/ml. Extracts were sterilized using 0.22 µm lter.

Tissue models and treatments
Tissue models: EpiDermFT TM tissues were purchased from Mattek Life Sciences (Ashland, MA), equilibrated and cultured according to manufacturer's instructions. Three tissues were used per extract.
EpiDermFT recreates normal skin tissue structure with differentiated dermis and epidermis. It consists of human-derived epidermal keratinocytes and dermal broblasts that are mitotically and metabolically active. The tissues were cultured according to the manufacturer's protocol, using an air-liquid interface tissue culture technique. To induce in ammation, tissues were exposed to UV. The extracts or vehicle (DMSO) were dissolved in media and applied to the media surrounding the tissues (n=3 for each condition). Tissues were incubated with extracts for 24 hours and ash frozen for RNA and protein analysis.
Gene expression analysis RNA extraction: Three tissues per group were used for the analysis of gene expression pro les. RNA was extracted from tissues using TRIzol® Reagent (Invitrogen, Carlsbad, CA), further puri ed using an RNAesy kit (Qiagen), and quanti ed using Nanodrop2000c (ThermoScienti c). Afterwards, RNA integrity and concentration were established using 2100 BioAnalyzer (Agilent).
Library construction and sequencing: In all cases, the sequencing libraries were prepared using NEBNext Ultra II mRNA library preparation kit for Illumina (NEB) following the manufacturer's instructions. The samples were processed by the same technician at the same time to avoid the introduction of technical batch effects. The cDNA fragment libraries were sequenced using NextSeq500 sequencing analyzer (Illumina). The samples were balanced evenly across the lanes of the sequencing owcell.
Bioinformatics analysis: Base-calling and demultiplexing were done with Illumina CASAVA v.1.9 bioinformatics pipeline. The base qualities were examined using FastQC v.0.11.8. The adapters and lowquality bases were trimmed using Trim Galore! v.0.6.4 https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/ . Trimmed reads were mapped to the human genome version GRCh37 using HISAT2 version 2.0.5 47 . Counts of reads mapping to the gene as a meta-feature were obtained using featureCounts v.1.6.1 48 taking to account the directionality of the sequencing libraries. Counts of reads mapping to features were loaded into R v.3.6.1 and normalized using DESeq2 v.1.24.0 Bioconductor package as described in the manual 49 . The differences between all experimental groups were examined using the likelihood ratio test (LRT) test implemented in DESeq2. The reduced model included the intercept and the full model was the experimental group (Cannabis extracts and controls). Multiple comparisons adjustment of p-values was done using Benjamini-Hochberg procedure 50 . Speci c comparisons between groups were extracted using results() function with contrast argument speci ed. Genes with adjusted p-values below 0.05 were considered signi cant.

Western blot analysis
After treatment with cannabis extracts for the indicated time, whole cellular lysates of 3D tissues were prepared in radioimmunoprecipitation assay buffer using 2.0 mm ZR BashingBead beads (Zymo Research). Proteins (30-100 μg per sample) were electrophoresed in 10% sodium dodecyl sulfate polyacrylamide gel and electrophoretically transferred to polyvinylidene di uoride membranes (Amersham Hybond TM -P, GE Healthcare) at 4 o C for 1.5 h. The blots were incubated for 1 h with 5% nonfat dry milk to block nonspeci c binding sites and subsequently incubated at 4 o C overnight with 1:1000 dilution of polyclonal antibody against IL-6 and COX-2 (Abcam). Immunoreactivity was detected using a peroxidase-conjugated antibody and visualized with the ECL Plus Western Blotting Detection System (GE Healthcare). The blots were stripped before reprobing with antibody against actin (Santa Cruz Biotechnology). Quanti cation of Western blot bands was performed using ImageJ in duplicate. Declarations 300. Tables   Table 1. Effects of novel C. sativa extracts on the expression of inflammation and fibrosis-related genes. Data are shown as log2 fold changes as compared to induced tissues. All changes shown here are statistically significant, p adj <0.05, ANOVA-like analysis and pair-wise comparison.   Effects of novel C. sativa extracts on the levels of IL-6 and COX2 in human EpiDerm FT tissues.