**Cell lines**

The human CRC cell lines SW480, SW620, HT29, DLD-1, HCT115, LS174, and RKO were obtained from the American Type Culture Collection (ATCC, Manassas, VA) and cultured as previously described [7]. Briefly, cells were grown in Dulbecco's Modified Eagle's Media supplemented with 10% fetal bovine serum, 2 mM Glutamax, 10 U/mL penicillin, 10 μg/mL streptomycin, and 0.25 μg/mL Amphotericin B at 37 ◦C in 5% CO2. RKO cells were grown under the same conditions except that RPMI was used in place of Dulbecco's Modified Eagle's Media.

CRC cell lines were treated as previously described, except cells were seeded at a density of 10,000 cells/well [7] and 16 h later treated with vehicle (DMSO), CBGV, CBGB, CBG, CBGP, CBGN at 10 μM for 48 h. In all treatments, the DMSO was maintained at a constant 1%. Results in two cell lines (HCT116 and SW480) were confirmed by trypan blue staining. Cells were plated and treated as described above and after 48 h adherent and nonadherent cells were collected and stained with 0.2% trypan blue; cells were counted on a Countess 3 automated cell counter (ThermoFisher, Pittsburgh, PA, USA). For dose effect curves, cells were seeded as described above and then treated with Vehicle (DMSO, CBGV, CBGB, CBG, CBGP, or CBGN at concentrations of 333 nM, 1 μM, 3.3 μM, 10 μM, 18.56 μM, 33 μM, and 56μM. Cell viability for all experiments was measured using the MTT ((3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2Htetrazolium), Biovision; Milpitas, CA). MTT (0.5 mg/mL, 15μL) was added to each well and incubated for 2 h at 37 ◦C in 5% CO2. Formazan crystals were solubilized by adding stop solution (10% Triton X-100, 0.05% HCl in isopropanol) and vigorously pipetting the mixture. Absorbance was measured at 570 nm on a FlexStation 3 (Molecular Devices, San Jose, CA, USA). For each experiment, the cell line/treatment was measured from triplicate wells and the average was determined. Data are presented as the signal normalized to vehicle control.

### **Statistics**

All results are shown as mean ± standard deviation. Statistical significance was determined using GraphPad Prism Software (9.3.1, San Diego, CA, USA) using a one-way ANOVA with Dunnett's multiple comparison post hoc tests.

#### **3. Results**

First, to generate CBG and CBG variants, we synthesized corresponding resorcinol fragments. Several approaches have been reported in the literature for the synthesis of resorcinol derivatives [33,35]. We have developed an efficient method to generate the resorcinol fragments as shown in Scheme 1. This was accomplished in three steps involving: (1) preparation of corresponding olefins, **2**–**5** using 3,5-dimethoxybenzyl triphenylphosphonium bromide (**1**) and the corresponding aldehyde (Wittig reaction); (2) hydrogenation of the resultant E/Z-olefin mixture in Parr-hydrogenation apparatus to give compounds, **6**–**9**; and (3) deprotection of methoxy group of compounds, **6**–**9** by using BBr3 gave corresponding substituted resorcinol derivatives, **10**–**13**. All of these operations overall gave moderate yields (40–80%).

Each of these corresponding substituted resorcinol derivatives, **10**–**13**, were coupled with geraniol in the presence of catalytic amounts of *p*-toluenesulfonic acid monohydrate as shown in Scheme 2 to give the corresponding CBG analogs, **14**–**17**, in 20–30% yields [36]. The solubility of CBGV and CBG is between 25–50 mg/mL or around 80–180 mM in pure DMSO; the solubility drops by 100-fold or more when the compounds are in 25% DMSO [37].

Given the potential of cannabinoids to evoke cancer cell death, we next investigated the effects of the CBG side-chain variants on cancer cell viability at 10 μM for 48 h. Following the 48-h timepoint, cell viability was assessed using the MTT assay. Our results show that CBGV, across all cell lines, demonstrated the greatest reduction in cell growth (Figure 1A–G). Additionally, we found that the 4-carbon variant, CBGB, decreased cancer cell viability, although not to the same extent as CBGV. Moreover, we found that the effects of CBG were dependent upon the cell lines tested, as CBG only reduced colorectal cancer cell viability in 4 of the cell lines tested (HT-29, DLD-1, LS174, and RKO, Figure 1C,D,F,G). Even in cell lines sensitive to CBG, CBGV remained more efficacious, except in DLD-1 cells where the CBG and CBGV had a similar effect on cell viability (Figure 1D). The longer 7 and 9 carbon variants (CBGP and CBGN) did not significantly influence cell viability in any of the cell lines tested. The results of the MTT assay were confirmed in HCT116 and SW480 cell lines by trypan blue staining (data not shown).

**Figure 1.** Effect of CBG side-chain variants on colorectal cancer cell viability. Colorectal cancer cell lines were treated with 10 μM of cannabinoid for 48 h and then viability was measured using the MTT assay. (**A**) SW480, (**B**) SW620, (**C**) HT29, (**D**) DLD-1, (**E**) HCT116, (**F**) LS174, and (**G**) RKO. *n* = 5 experiments per group. \* *p* ≤ 0.05, \*\* *p* ≤ 0.01, \*\*\* *p* ≤ 0.005 vs. vehicle.

To better examine the impact of these compounds on colorectal cancer cell growth, we performed dose effect curves. Consistent with our data at 10 μM, we found that CBGV had the lowest IC50 value across all cell lines, except DLD-1 cells (Figure 2A–G, Table 1). We also found that only in DLD-1 cells was the IC50 value of CBG similar to that found with CBGV (Figure 2D, Table 1). The highest IC50 values were found for the molecules with the longer side chains (Figure 2, Table 1). In general, the IC50 values for CBGB and CBG were found to be between those observed for CBGV and the larger chain molecules (CBGP and CBGN) (Figure 2, Table 1).

**Figure 2.** Dose response curves of CBG side-chain variants on colorectal cancer cell viability. Colorectal cancer cell lines were treated with varying doses of cannabinoid for 48 h and then viability was measured using the MTT assay. Representative dose response curves are shown for (**A**) SW480, (**B**) SW620, (**C**) HT29, (**D**) DLD-1, (**E**) HCT116, (**F**) LS174, and (**G**) RKO cell lines.

Next, we investigated the anti-nociceptive properties of CBG variants in a model of CIPN. Previously, we have demonstrated, using the von Frey test, that CBG (10 mg/kg i.p.) was effective at significantly reducing mechanical hypersensitivity in a preclinical model of CIPN [15]. Using this model, we investigated and compared the anti-nociceptive properties of the CBG side chain variants, CBGV, CBGB, CBG, CBGP, and CBGN. Neuropathic male mice were treated with vehicle control, a CBG variant, or the positive control indomethacin each at 10 mg/kg i.p. Mice then underwent von Frey testing of the hind-paw, 1 h following injections, to measure the force required to elicit a paw withdrawal response. Our results show that all variants were equally as effective as CBG and the positive control, indomethacin, in reversing CIPN (F(6,63) = 9.56, *p* < 0.0001; one-way ANOVA with Tukey's post-test). (Figure 3).


**Table 1.** IC50 values for CBG variants in Colorectal Cancer Cell Lines.

IC50 values are from 3–4 independent dose effect curves and are presented as mean ± standard deviation, concentrations are in μM. \* *p* ≤ 0.05, \*\* *p* ≤ 0.01, \*\*\* *p* ≤ 0.005, \*\*\*\* *p* ≤ 0.001 vs. CBG.

**Figure 3.** Effect of CBG side-chain variants on mechanical sensitivity. Neuropathic male mice were treated with 10 mg/kg of vehicle, side-chain variant of CBG, or indomethacin as a positive control and mechanical sensitivity was assessed using von Frey filaments. *n* = 10 mice per group, \*\*\*\* *p* ≤ 0.0001.

#### **4. Discussion**

Recently, a variant of CBD and THC has been identified in which the 5-carbon side chain is two carbons longer, these molecules were termed CBDP and THCP [22]. The authors went on to show that THCP binds to cannabinoid receptors with a higher affinity and was more effective at reducing pain than THC. In contrast, we did not see any impact of side-chain length on the ability of CBG to reduce pain in a mouse model of CIPN. However, we did find a significant anti-nociceptive effect of CBG and all CBG variants in a model of CIPN.

Our data indicate that side-chain length plays a role in the ability of CBG to reduce colorectal cancer cell growth in vitro. We found that, molecules with shorter side chains are more efficacious at reducing cell growth compared to longer side chains. Our findings with CBG are in contrast with previous work on CBDV and THCV which did not find a significant difference between these compounds and the more common 5-carbon variant (CBD and THC) [38–40]. Furthermore, neither the 4 or 7 carbon variants of CBD was found to have any greater impact on breast cancer cell growth than the 5-carbon molecule [41]. This could be due to the unique nature of CBG, which has been found to be an agonist of α2-adrenogeric receptors [42] and peroxisome proliferator-activated receptors (PPAR) α and γ [43–46], and the activation of these receptors has previously been reported to inhibit colorectal cancer cell growth.

It has been shown that side chain length can influence receptor binding for cannabinoids. For example, THCP has been shown to bind to CB1 receptors with a higher affinity than THC [22]. In contrast, THCV acts as an antagonist of the CB1 receptor, the opposite activity of THC, THCB, and THCP [24,47]. It is known that CBG binds with differing affinities and activities to a variety of receptors compared to THC and CBD [25]. One possible explanation for the differences we observe between the pain assay and cytotoxicity effects of these molecules is that different receptors mediate the analgesic and cytotoxic effects of CBG. Alternatively, the varying side chain lengths may create differing pharmacokinetics in vivo, thus normalizing their effects. Further studies on the binding of these novel molecules at known CBG receptors and additional studies on the mechanism that leads to cytotoxicity may provide useful insights into the mechanism by which CBGV and CBGB are slightly more cytotoxic but not more analgesic. These studies would also provide novel insights into how CBG interacts with known receptors.

#### **5. Conclusions**

We have developed a unique and adaptable process for generating cannabinoids with varying side chain lengths. Several recent reports have identified variants of CBD and THC with 3, 4, 6, and 7 carbon side chains; however, such side chain variants also likely exist for other cannabinoids such as cannabichromene (CBC) and must exist for CBG since this is the precursor molecule for the other cannabinoids. Surprisingly, we did not observe any effect of side-chain length regarding the ability to reduce neuropathic pain, which is in contrast to the data regarding THCP. However, we did find that CBG variants, such as CBG, produced significant anti-nociceptive effects in a murine model of CIPN. Importantly, we found that shorter side-chain variants of CBG were better able to reduce colorectal-cancer cell viability compared to longer-side chains.

**Author Contributions:** Conceptualization, D.D., K.E.V., N.M.G., S.A. and W.M.R.-K.; formal analysis, D.E.S., N.M.G., W.M.R.-K. and D.D.; investigation, D.E.S., W.M.R.-K., D.P.M. and D.D.; resources, K.E.V.; data curation, D.D. and W.M.R.-K.; writing—original draft preparation, D.E.S., D.D. and W.M.R.-K.; writing—review and editing, D.D., K.E.V., N.M.G. and W.M.R.-K.; visualization, W.M.R.- K.; project administration, W.M.R.-K.; funding acquisition, K.E.V. and N.M.G. All authors have read and agreed to the published version of the manuscript.

**Funding:** This project was supported by a NARSAD Young Investigator Award (27364; NMG) and by the Pennsylvania Department of Health using Tobacco CURE Funds (NMG). KEV (and the Penn State College of Medicine) is the recipient of research support from PA Options for Wellness (a state-approved medical marijuana clinical registrant). The funding sources were not involved in: study design, providing any experimental materials, data collection, analysis and interpretation; writing of the report; or the decision to submit the article for publication.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors would like to acknowledge members of the state-approved medical marijuana academic clinical research center at Penn State for insights and comments on the data and study design, along with the Drug Discovery, Development, and Delivery (D4) core and Organic Synthesis core for assistance with the preparation of the compounds and viability testing.

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