Effects of cannabis oil extract on immune response gene expression in human small airway epithelial cells (HSAEpC): implications for chronic obstructive pulmonary disease (COPD)

Cannabis sativa oil extract

The cannabis oil extract employed in these experiments was kindly provided by F. Ferri, NCM Biotech, Newport, RI, USA. This extract, a proprietary formulation covered under US Patent 9,199,960, contained approximately 95 mg/ml of total cannabinoids in ethanol. This was composed of a mixture of about 55 mg/ml of cannabidiol plus cannabidiolic acid and about 40 mg/ml of tetrahydrocannabinol plus tetrahydrocannabinolic acid. The preparation did not contain cannabinol. An analysis for terpenes or flavonoids was not conducted; the former but not the latter was likely present in the preparation. The stock solution was diluted 1:100 in ethanol and stored at 4 C. This preparation was further diluted 1:100 into cell culture media and tested for cytotoxic effects against primary HSAEpC cells in vitro. It was found that stock extract dilutions of 1:400 and higher were not cytotoxic. Ethanol (non-cytotoxic at a 0.25% [2.5 ul/ml] final concentration) was the negative control. Neither ethanol alone nor the combination of extract in ethanol showed any cytotoxic effects at a 1:400 dilution in cell culture media. Accordingly, 1:400, 1:800 and 1:1600 stock extract dilutions (corresponding to final concentrations of 1:40,000, 1:80,000 and 1:160,000, or about 2.4, 1.2 and 0.6 μg/ml of total cannabinoids in cell culture media) were selected for further testing.

Treatment of cells with cannabis oil extract

HSAEpC were seeded in 6-well cell culture plates at a density of 200,000 cells per well. After overnight adhesion, cells were treated for 24 h with media containing cannabis oil extract (1:400, 1:800 and 1:1600 dilutions). Negative control cells were treated with an equivalent volume of ethanol or phosphate buffered saline (PBS) (1:400 dilution for a 0.25% final concentration). Each treatment dilution had four replicates, whereas control conditions (ethanol and PBS) had 6 replicates.

RNA extraction and reverse transcription

After 24 h of exposure to cannabis oil extract, cells were washed with PBS, collected in Trizol (Thermo Fisher Scientific, Waltham, MA) and total mRNA was extracted following manufacturer’s instructions. Briefly, samples were incubated for 5 min at room temperature and 200 ul of chloroform (Sigma-Aldrich, Saint Louis, MO) was added to 1 ml of Trizol. The aqueous phase containing RNAs was separated by centrifugation at 12,000 g for 15 min and transferred in a new tube. RNA was subsequently precipitated by adding 500 ul of isopropanol (Sigma-Aldrich) for 10 min followed by centrifugation for 10 min at 12,000 g. Pellets containing RNAs were washed three times in 75% ethanol and resuspended in 15–40 ul of nuclease-free water.

After the yield of RNA extraction was quantified using a Nanodrop (Thermo Fisher Scientific), complementary DNA (cDNA) was synthesized from 250 ng of RNA using the reverse transcription (RT2) First Strand kit (Qiagen, Valencia, CA) following the manufacturer’s instructions.

Pathway-focused quantitative polymerase chain reaction (qPCR) assays

In order to identify respiratory-related Th2 and associated immune response genes whose expression could be altered by the cannabis oil extract treatment, RT² Profiler™ PCR Arrays were used according to the manufacturer instructions (Qiagen, catalog number PAHS 067Z, product number 330231). Briefly, this assay in 384 well plate format contains primers for 84 genes of interest (GOIs) and 4 housekeeping genes in 4 replicates. A listing and description of the Th2 cytokine/chemokine and related immune response genes in this PCR array are shown in the accompanying Additional file 1: Table S1. PCR mixes including SYBR Green dye and 0.17 ul of cDNA were added to the plate wells and the PCR reaction was performed following the manufacturer’s instructions using an Applied Biosystems® ViiA™ 7 Real-Time PCR System (Thermo Fisher Scientific).

DAVID bioinformatics analysis summary

Treatment of HSAEpC cells with cannabis oil extract markedly altered expression of 37 immune response and related genes out of 84 tested after 24 h of treatment. Some of these genes are known in the literature to be associated with respiratory diseases such as COPD. Bioinformatics analysis using the DAVID program was helpful in elucidating groups of genes with related biological activities. For example, 11 GOIs were part of the same KEGG pathway for cytokine and cytokine receptor interactions. This pathway is important in the early recognition and response phases of both innate and adaptive immunity (Arakelyan et al. 2017; Gardy et al. 2009). Furthermore, out of 26 respiratory disease-related GOIs, there were 10 GOIs specifically related to COPD. This bioinformatics analysis was coupled with literature evaluations in order to highlight the most relevant of these genes. The logical starting point for this was the eight GOIs that were up-regulated at two or three test dilutions (Tables ​(Tables11 and ​and4),4), along with the six GOIs that were down-regulated at two test dilutions (Table ​(Table22).

Up-regulated GOIs

One prominent up-regulated GOI is IL4, which encodes interleukin IL-4, a known Th2 cytokine (Barczyk et al. 2006; Borish and Steinke 2003; May and Fung 2015; Röcken et al. 1996). There is already a body of research showing that increased IL-4 production can exacerbate eosinophilia and allergen-induced asthma, at least in those instances where the disease is Th2-driven. Conversely, IL-4 could modulate COPD pathophysiology in cases where COPD can be identified as a Th1-mediated inflammatory disease (Barnes 2016; Cornwell et al. 2010; Cosio et al. 2009). In one patient study, COPD severity was correlated with lower levels of IL-4, while higher IL-4 levels contributed to wound repair in the lung epithelium (Perotin et al. 2014). IL-4 also is a signaling molecule in the JAK-STAT signaling pathway (David et al. 2001). By binding to its cell surface receptor, a cytokine such as IL-4 causes receptor dimerization, mediating a signaling process involving Janus kinase phosphorylation and the subsequent phosphorylation and dimerization of STAT proteins. The activated STAT proteins in turn induce the transcription and expression of a variety of genes involved with promoting or maintaining immune responses (Yew-Booth et al. 2015). IL13RA2, which encodes IL-13 receptor alpha 2, also was up-regulated. This receptor has more complicated regulatory activities that can involve IL-4, IL-13 and their related receptors (Andrews et al. 2013, 2009; Chomarat and Banchereau 1998). In fact, while IL4 and IL13RA2 were up-regulated, IL4R (encoding IL-4 receptor) and IL13RA1 (encoding IL-13 receptor alpha 1) were downregulated at the 1:400 sample dilution. These results suggest that, despite cannabis oil extract containing a mixture of cannabinoid compounds, it was selectively regulating Th2 cytokine and cytokine receptor gene expression. The activation of the JAK-STAT signaling pathway via IL-4 or other upregulated GOIs may also promote lung epithelial cell migration and repair (Bansal et al. 2012; Crosby and Waters 2010; Kida et al. 2008). Additionally, IL-13 receptor alpha 2 has been found to inhibit the detrimental tissue remodeling induced by IL-13 in murine models of both lung inflammation and eosinophilic esophagitis, a chronic inflammation of the esophagus (Zheng et al. 2008; Zuo et al. 2010).

Both IL4 and PPARG, which encodes peroxisome proliferator-activated receptor gamma (PPAR-gamma), were identified through pathway analysis as the two GOIs whose gene products act as negative regulators of acute inflammation. PPAR-gamma is a nuclear hormone receptor with known anti-inflammatory activities (Nencioni et al. 2003; Ricote et al. 1998). There is a substantial body of literature showing that PPAR-gamma (which was up-regulated by all three cannabis oil extract dilutions) can have beneficial effects in the therapy of COPD, and more specifically, that airway epithelial cells are affected by PPAR-gamma (Belvisi and Mitchell 2009; Morissette et al. 2015; Solleti et al. 2015). PPAR-gamma was capable of both reducing cigarette smoke-induced inflammation and protecting against Haemophilus influenzae-induced symptom exacerbations in a mouse model of COPD (Morissette et al. 2015; Solleti et al. 2015).

The gene with the highest fold changes at all three sample dilutions was IL1RL1, which encodes IL-1 receptor-like 1 (also known as ST2). This protein is present in lung mucosa, including airway epithelial cells, and has a critical role in the Th2-mediated immune response (Coyle et al. 1999; Traister et al. 2015; Zhao and Zhao 2015). The observed levels of IL1RL1 up-regulation by all three cannabis oil extract dilutions may make this gene a suitable biomarker for further in vitro and in vivo studies.

CCL22, CCL24 and CCL26 were three chemokine genes up-regulated by cannabis oil extract. They encode chemokine (C-C motif) ligands 22, 24 and 26, respectively. While within the limits of the qPCR detection threshold, their basal expressions in PBS and ethanol controls were low, especially for CCL24, which led to a high fold change. CCL22 was the only chemokine gene of the three that was up-regulated at both the 1:400 and 1:800 dilutions. Chemokine (C-C motif) ligand 22 has been described as both a Th2-attracting and homeostatic chemokine (Ying et al. 2008; Zlotnik and Yoshie 2012). Macrophage CCL22 gene expression in COPD has been the subject of several studies. CCL22 up-regulation has been found in sputum macrophages from COPD patients (Frankenberger et al. 2011). Eapen et al. observed a reduction in M2 (Th2) macrophages and an increase in M1 (Th1, pro-inflammatory) macrophages in the small airway walls of smokers and COPD patients (Eapen et al. 2017). However, they also reported that CCL22, along with Th2-related genes IL4, IL10 and IL13 were up-regulated in bronchiolar lavage fluid from COPD patients. In the same study, mRNA levels of CCL22 was increased in the lungs of mice chronically exposed to cigarette smoke relative to normal controls. The authors noted that these chemokines and cytokines were characteristic of a Th2 (or for macrophages, M2) immune response.

Another GOI up-regulated by cannabis oil extract was TSLP, which encodes thymic stromal lymphopoietin. TSLP protein is known to be expressed in lung epithelial cells, and is associated with the Th2 immune response, including the downstream production of Th2 cytokines and chemokines such as IL-4 and chemokine (C-C motif) ligand 22 (Fang et al. 2010; He and Geha 2010; Redhu and Gounni 2012). Also, TSLP protein, in conjunction with Th2 cytokine IL-13, was capable of increasing airway epithelial cell proliferation as well as cell repair after mechanical injury (Semlali et al. 2010).

Two up-regulated genes, CSF2 and IL17A, were identified in pathway analysis as positive regulators of IL-23 production. IL-23 is a pro-inflammatory cytokine involved in the cell-mediated immune response (Oppmann et al. 2000). This cytokine has been described as heterodimeric, composed of two subunits, IL-23A and IL-12B. Neither IL23 itself or IL23A were part of the gene array, but the IL12B gene, encoding interleukin 12B (natural killer cell stimulatory factor 2, cytotoxic lymphocyte maturation factor 2, p40) was up-regulated at the 1:400 dilution. IL17A (alternatively, IL17) encodes cytokine IL-17A, which, like IL-23, plays a role in inflammation (Chen and Kolls 2017). Increased production of IL-17A can occur in both asthma and COPD, and may contribute to COPD pathophysiology (Alcorn et al. 2010; Chang et al. 2014). However, IL-17A also functions in host defense against microbial pathogens (Aujla et al. 2007; Tsai et al. 2013). Perhaps a more interesting finding was that certain T-cell populations in asthma patients can produce both IL-17A and IL-4, and such cells appear to be a combination of Th17 (an immune response involving IL-17 and separate from Th1 and Th2) and Th2 (Cosmi et al. 2010).

CSF2 (alternatively, GMCSF), the gene with the second highest fold change after IL1RL1, encodes colony stimulating factor 2 (granulocyte-macrophage), or GM-CSF. This protein, a member of the hematopoetins family of cytokines, functions in white blood cell proliferation, migration and differentiation processes (Francisco-Cruz et al. 2014; Metcalf 2008). While GM-CSF was overexpressed in submucosal and airway smooth muscle cells of both asthma and COPD patients, elevated GM-CSF expression was linked to disease severity only in asthma (Saha et al. 2009). Furthermore, the production of GM-CSF in airways epithelial cells can be induced by IL-17A, possibly with respect to protection from microbial infections (Ponce-Gallegos et al. 2017).

Down-regulated GOIs

Pathway analysis of the six genes of interest that were significantly down-regulated by the 1:400 and 1:800 dilutions of cannabis oil extract (CLCA1, CMA1, EPX, LTB4R, MAF and PMCH) did not reveal common biological functions. That said, two of the down-regulated GOIs, CLCA1 and CMA1, appear to be highly relevant to the treatment of COPD with respect to mucus overproduction or hypersecretion. CLCA1 encodes chloride channel accessory 1, whose expression or up-regulation has been related to mucus overproduction as well as airway hyperreactivity in COPD (Hauber et al. 2005; Iwashita et al. 2012; Sala-Rabanal et al. 2015; Wang et al. 2007). The feasibility of inhibiting mucin production genes as a viable approach to treating COPD has been postulated (Ha and Rogers 2016). Accordingly, the use of cannabis oil extract to inhibit the expression of mucin production genes such as CLCA1 could help in controlling airway mucus hypersecretion. CMA1 encodes mast cell chymase 1, a serine proteinase. Chymase can stimulate mucin secretion from lung epithelial cells, and is known to contribute to the mucus hypersecretion process in patients with COPD (He and Zheng 2004). Also, mast cell chymase and other serine proteases are known to promote inflammation through degradation of the extracellular matrix (de Garavilla et al. 2005). An inhibitor of these proteases had anti-inflammatory effects that could be related to the treatment of COPD. The down-regulation of CMA1 would therefore potentially decrease both mucus hypersecretion and inflammation.

Two other GOIs down-regulated by cannabis oil extract that may have relevance to COPD were LTB4R (alternatively, BLT1 or LTB4R1), which encodes leukotriene B4 receptor, and EPX, which encodes eosinophil peroxidase. Leukotriene B4 is an arachidonic acid-derived, neutrophil recruiting, pro-inflammatory lipid molecule, and both this leukotriene and its receptor have a role in the pathophysiology of COPD (Dong et al. 2016; Marian et al. 2006; Pace et al. 2013). The targeting of the B4 receptor has been considered as a viable approach to the treatment of COPD (Grönke et al. 2008; Hicks et al. 2010). Accordingly, the down-regulation of LTB4R by cannabis oil extract appears to be yet another indication of its anti-inflammatory potential. Interestingly, PPAR-gamma, which was up-regulated by cannabis oil extract, can also counteract inflammation mediated by leukotriene B4 in COPD (Yin et al. 2014). While eosinophil peroxidase does not appear to have been investigated as target for the treatment of COPD, it has been identified as a biomarker of the disease (Nair et al. 2013; Yang et al. 2017).

The remaining two genes down-regulated at the 1:400 and 1:800 test dilutions were MAF, encoding V-maf musculoaponeurotic fibrosarcoma oncogene homolog (avian), and PMCH, encoding Pro-melanin-concentrating hormone. Neither of these genes appears to have an obvious role in COPD, based on a literature search. It is interesting that a substantial number of the down-regulated GOIs encode either enzymes or proteins with functions other than as immune signaling cytokines or chemokines. There were also a number of GOIs that were up- or down-regulated at a single dilution that may prove relevant to the treatment of COPD by cannabis oil extract. Perhaps the most notable of these genes was MMP9, which encodes matrix metallopeptidase 9 (gelatinase B, 92 kDa gelatinase, 92 kDa type IV collagenase). This extracellular matrix-degrading metallopeptidase functions in a variety of normal physiological processes. However, expression and activation of MMP-9 has been associated with the pathophysiology of COPD (Grzela et al. 2016; Navratilova et al. 2016; Papakonstantinou et al. 2015). Interestingly, Zhou et al. reported that the concentration of MMP-9 was increased while that of PPAR-gamma was decreased in sputum samples from COPD patients, adversely affecting respiratory function (Zhou et al. 2017). This research indicated that PPAR-gamma had a protective effect in preventing an imbalance between MMP-9 and TIMP-1, a metalloproteinase inhibitor.

The authors wish to thank Christoph Schorl, Brown University, for the use of the real-time PCR machine at Brown University’s Center for Genomics and Proteomics Core Facility.