Abstract

Background

   Cervical cancer, one of the lethal cancers among women, is a
   challenging disease to treat. The current therapies often come with
   severe side effects and the risk of resistance development. Traditional
   herbal medicine, with its potential to offer effective and less toxic
   options, is a promising avenue. This study was undertaken to
   investigate the potential of Rhamnus prinoides (R. prinoides) root bark
   extracts in selectively inhibiting the proliferation of cervical cancer
   cells, using the HeLa cell line as an in vitro model.

Methods

   R. prinoides plant extracts were first screened at a fixed
   concentration of 200 μg/ml to determine the active extract. The
   selective anti-proliferative activity of the active extract was
   evaluated in a concentration dilution assay using the
   (3-(4,5-dimethylthiazol- 2-yl)-2,5-diphenyltetrazolium bromide) MTT
   assay on cancerous (HeLa) cells and non-cancerous (Vero) cells to
   determine the half-maximal inhibitory (IC[50]) and half-cytotoxic
   concentrations (CC[50]), respectively. Functional assays on cell
   morphology (by microscopy), cell migration (wound healing assay) and
   cell cycle (by flow cytometry) were also conducted. The active extract
   was analyzed using Gas Chromatography/Mass Spectrometry (GC/MS) to
   determine any compounds it contained. Following identification of
   possible gene targets by network pharmacology, the genes were validated
   by molecular docking and Real-Time Quantitative Polymerase Chain
   Reaction (RT-qPCR).

Results

   The ethyl acetate extract of R. prinoides (EARP), the most active
   extract, selectively inhibited the growth of cervical cancer cells,
   their migration and induced cell cycle arrest at the S phase. In silico
   analysis revealed that squalene,
   3,3a,6,6-tetramethyl-4,5,5a,7,8,9-hexahydro-1H-cyclopenta[i]indene and
   Olean-12-en-3.beta.-ol, acetate showed acceptable drug-like
   characteristics and may be partly attributed to the bioactivity
   demonstrated and the deregulation of the mRNA expression of AKT1,
   NF-κB, p53, Bax, Bcl-2, and Er-b-B2.

Conclusion

   This study, for the first time, demonstrates the anti-proliferation
   effects of EARP and forms a firm foundation for further drug
   development studies.

   Keywords: Cervical cancer, Cytotoxicity, Rhamnus prinoides, Network
   pharmacology, Anti-proliferative activity

1. Introduction

   Worldwide, the most frequent cause of fatalities is cancer, which has
   even been hailed recently as a “major public health challenge”. After
   lung, colorectal, and breast cancers, respectively, cervical cancer
   comes fourth in terms of prevalence. The bulk of new instances and
   fatalities from cervical cancer occurs in low-to-middle-income nations,
   which bear a disproportionate amount of the burden when compared to
   nations with higher income; actually, 84 % of all cervical cancer cases
   worldwide are found in Sub-Saharan Africa [[29]1]. Cervical cancer
   accounts for 12.4 % of all malignancies in Kenya, coming second in
   terms of prevalence, nevertheless, being the primary cause of
   cancer-related loss of life in the nation, accounting for 11.9 % of all
   cancer-related deaths, with breast cancer, which is first in terms of
   prevalence, accounting for 11.5 % of all cancer-related deaths. Factors
   such as inadequate health facilities, personnel and late diagnosis are
   responsible for the increased prevalence and high mortality rates
   [[30][1], [31][2], [32][3], [33][4]]. Cytoreductive surgery, radiation
   therapy, and chemotherapy are the most widely used treatment options
   for cervical cancer in Kenya. The high cost of treatment along with
   high instances of recurrence, chemo-resistance and severe side effects,
   are some of the problems associated with the chemotherapy
   [[34]5,[35]6]. Thus, searching for less harmful and more effective
   alternatives is both necessary and essential.

   Traditionally, use of therapeutic plants, dates back a long way in
   treating and mitigating illnesses, and they hold a vast potential as
   reservoirs of potent anti-cancer agents [[36]7,[37]8]. Rhamnus
   prinoides L'Herit (R. prinoides), which belongs to the genus
   Rhamnaceae, is a plant exotic to Kenya and native to many parts of
   Eastern and Central Africa. The tree is small, attaining an average
   height of just 4 m. It is known locally as “Mukarakinga” among the
   Kikuyu community and “Olkonyil” among the Maasai community in Kenya.
   Traditionally, a variety of plant parts of R. prinoides are used, these
   include, the leaves, stembark, and root bark, to address chest issues,
   tonsils, pneumonia, as an appetizer, to manage malaria, sexually
   transmitted diseases, and to cleanse blood, depending on the region
   [[38][9], [39][10], [40][11], [41][12]]. R. prinoides extracts were
   shown to have anti-microbial qualities [[42]13], anti-oxidant
   properties [[43]14,[44]15], anti-malarial effects [[45]16], and
   anti-inflammatory action [[46]15]. To the best of our knowledge, R.
   prinoides anti-proliferation and anti-cancer activities have not been
   assessed experimentally.

   We sought to explore the potential of anti-proliferation of crude-,
   hexane-, water- and ethyl acetate extracts of R. prinoides against
   human cervical cancer cells (HeLa cells) and non-cancerous
   Cercopithecus aethiops epithelial line from African green monkey kidney
   cells (Vero cells). The results demonstrated that the ethyl acetate
   extract fraction (EARP) exhibited the most effective selective action
   towards cervical cancer. We identified the compounds in the extracts
   that can be partly attributed to the recorded bioactivity using GC/MS
   analysis. We then conducted in silico studies which suggested that
   certain compounds had promising putative binding affinities with
   various genes essential for the development and progression of cervical
   cancer. We further validated these findings through in silico docking
   studies, cellular functional assays (ability to limit cell migration
   and induce cell cycle arrest at the S phase) and by gene expression
   analysis using RT-qPCR. Notably, we found that EARP down-regulated the
   mRNA expression of AKT1, Bcl-2 NF-κB and Er-b-B2 genes while
   up-regulating the m-RNA expression of tumor suppressor protein 53 and
   Bax.

   Therefore, we showed the potential of R. prinoides ethyl acetate
   extract as an anti-cancer agent and laid a solid foundation for further
   studies to isolate the compounds responsible for activity and validate
   their activity in higher organisms.

2. Materials and methods

2.1. Plant collection and extraction

2.1.1. Plant collection

   The root barks of R. prinoides were fetched from the region of Mount
   Kenya at the coordinates 0° 07′ 15.60″ N, 37° 20′ 7.20″ E. A taxonomist
   assisted in the plant's identification, collection and a voucher
   specimen was archived in the National Museums of Kenya's East Africa
   Herbarium (Mwitari/Gathiuru/RM/001/2022). The root barks were sent to
   the Kenya Medical Research Institute's Center for Traditional Medicine
   and Drug Research for processing and extraction.

2.1.2. Plant preparation

   After the root barks were cleaned to get rid of dust and other
   particles, they were chopped into smaller bits and allowed to air dry
   at room temperature then crushed into a powder with a laboratory mill
   [[47]17].

2.1.3. Solvent extraction

   The cold solvent extraction method followed by solvent partitioning was
   used as described previously with minor adjustments [[48]18,[49]19].
   800g of the ground root bark of R. prinoides was weighed using a
   laboratory electric balance and macerated in dichloromethane: methanol
   mixture in a 1:1 ratio until submerged for 72 h with occasional shaking
   to get a crude extract. The solvent mixture was then decanted into a
   conical flask with a Whatman No.1 filter (Schleicher and Schuell
   Microscience Gmbh, Germany) and a vacuum rotary evaporator (Buchi,
   Switzerland) used to condense the extract at 55 °C. A portion of the
   crude extract was allowed to air dry. The remaining crude extract was
   re-soaked in 400 ml hexane to extract the non-polar compounds and
   concentrated in a rotary evaporator at 59 °C and allowed to dry at room
   temperature.

   Furthermore, the remaining extract was soaked in a combination of water
   and ethyl acetate in a 1:1 ratio in a separating funnel to extract the
   mid-polar and the polar compounds. The mixture was left to separate
   overnight forming two distinct layers. The two layers were then
   separated using density. The ethyl acetate partition was concentrated
   in a vacuum rotary evaporator at 67 °C while the water partition was
   lyophilized in a freeze drier (Modulyo Edwards high vacuum, Crawley
   England, Britain, Serial No. 2261). The extracts were left to air dry
   and kept at −20 °C until use. The percentage yield of the crude-,
   hexane-, water- and ethyl acetate extracts was calculated as;
   % yield = [(weight of dried extract) / (weight of dried plant sample)]
   x 100

   With the use of an electric balance, 100 mg each plant extract fraction
   was measured and dissolved in 1 ml of laboratory-grade 100 % Dimethyl
   Sulphoxide (DMSO, FINAR, India) and kept at −20 °C until use as the
   stock solution.

2.2. Cell culture and anti-proliferation assay

2.2.1. Cell culture

   The American Type Culture Collection (ATCC) supplied the non-cancerous
   Vero cells (ATCC Cat# CCL-81, RRID: CVCL_0059) and cervical cancer
   HeLa-229 cells (ATCC Cat# CCL-2.1, RRID: CVCL_1276) that were used for
   the experiments. The MycoStrip™ Mycoplasma Detection Kit (Invitrogen,
   USA) was used to ascertain the purity of the cell lines. 10 % Fetal
   Bovine Serum (FBS, Sigma Aldrich, USA) was added to the Minimum
   Essential Medium (MEM, GIBCO, USA) along with 7.5 % Sodium Hydrogen
   Carbonate (LOBA Chemie, India), 1 % L-Glutamine (Sigma Aldrich, USA),
   1 % Penicillin/Streptomycin (Sigma Aldrich, USA) and 1 % HEPES buffer
   (Gold Biotechnology, USA). They were maintained at 37 °C with 5 % CO[2]
   inside an incubator with humidity (Thermo Fisher Scientific, USA).

2.2.2. Viability and proliferation assay

   Proliferation and viability of HeLa and Vero cells was determined using
   the MTT assay as described previously [[50]20]. In brief, HeLa and Vero
   cells were seeded at a density of 10,000 cells/well in 96 well plates
   (Thermo Fisher Scientific, USA) and left overnight to grow at 37 °C
   with 5 % CO[2] in an incubator with humidity. After 24 h, the used
   media was dispensed with and replenished with 100 μl of fresh media.
   The plant extract concentrations were made by diluting the initial
   stock, reconstituted with DMSO, using growth media with the final DMSO
   concentration being 0.4 %. The first screening was conducted on water-,
   crude-, hexane- and ethyl acetate extracts and also doxorubicin
   hydrochloride as a positive control at a fixed concentration of
   200 μg/ml for 48 h in HeLa cells. 0.4 % DMSO was used as a negative
   control. The most active extract was selected and the cells were
   treated to a range of concentrations of the active extract (starting
   from 240 to 0 μg/ml in HeLa and 800-0 μg/ml in Vero cells) for 48 h.
   The positive control utilized was doxorubicin hydrochloride at
   concentrations starting from 10 to 0.5 μg/ml for HeLa cells and
   100–1.5625 μg/ml for Vero cells. The media with the extract was then
   discarded and the cells washed lightly using phosphate-buffered saline
   (PBS, Sigma Aldrich, USA). Thereafter, fresh media containing MTT dye
   (Solar Bio, China) at a concentration of 5 mg/ml in PBS was put in and
   the plates incubated for 4 h at 5 % CO[2] and 37 °C. After 4 h the
   formation of formazan crystals was confirmed through the change in
   color from yellow to purple. The media was dispensed with and 100 μl of
   DMSO put into the wells and the absorbance read at 570 nm and 720 nm
   was used reference wavelength to capture non-specific absorbance in a
   microtiter reader (Thermo Fisher Scientific, USA). The experiment was
   run in triplicate at least three times. The anti-proliferation effect
   of the extract was expressed as a percentage of the ratio of the
   treated cells to those treated with the negative control and the IC[50]
   (Inhibitory concentration that kills 50 % of cells), IC[25] and
   IC[12.5]values were estimated with the Graph pad Prism 8.0 software (La
   Jola, California, USA).
   % Viability = [Absorbance of Sample/Absorbance of Control] *100

2.3. Cell phenotypic analysis

   To delve into how the EARP extract affects the morphological phenotype
   of HeLa cells, for 48 h the cells were treated with EARP at IC[50],
   IC[25] and IC[12.5] concentrations and the positive control,
   doxorubicin hydrochloride, at its IC[50] concentration. The resulting
   phenotypic changes were recorded using an EVOS™ XL Core Imaging system
   (Thermo Fisher Scientific, USA).

2.4. Wound healing assay

   The effect of EARP on the limitation of HeLa cells migration was
   estimated according to the previously described method [[51]21]. HeLa
   cells were seeded in 24-well plates and left to attach overnight in a
   humidified incubator at 37 °C with 5 % CO[2]. An artificial wound was
   made by use of a sterile p200 tip once a confluent monolayer had been
   formed. After scraping the cells and washing away loose cells with PBS,
   new media containing the EARP extract at its IC[50], IC[25] and
   IC[12.5]concentrations was added to the wells. Using an ultra-fine tip
   marker pen (Sharpie, USA), a reference line was made perpendicular to
   the artificial wound. The scratch areas were captured at 0 h, 24 h and
   48 h using an EVOS™ XL Core Imaging system (Thermo Fisher Scientific,
   USA) and analyzed using Image J software ([52]imagej.nih.gov). The
   percentage wound closure was calculated using the formula:
   %Wound Closure = [(A (0) - A (t))/A (0)] × 100

   Whereby, A (0) is the area at time zero (0) and A (t) is the area after
   incubation time (t).

2.5. Cell cycle analysis

   In order to examine how EARP affects the cell cycle, HeLa cells at
   100,000 cells/ml were propagated in T-25 culture flasks using MEM
   culture media and incubated overnight at 37 °C with 5 % CO[2]. The
   cells were then exposed to EARP at IC[50], IC[25] and negative control
   and incubated for 48 h at 37 °C with CO[2] at 5 %. Thereafter, trypsin
   was used to harvest the cells after they had been incubated and rinsed
   twice with ice-cold PBS. Then cells were suspended in 70 % ethanol for
   24 h. After discarding the ethanol, the cells were centrifuged for
   2 min at 2000 rcf and re-suspended in PBS with 0.25 % Triton X-100 for
   15 min on ice. After centrifugation of the cells for 2 min at 2000 rcf,
   the supernatant was discarded. The cells were kept at 24 °C for 30 min
   in a dark place after being re-suspended in PBS with 10 μg/ml of RNase
   A and 25 μg/ml of propidium iodide. A flow cytometer (BD Canto II, BD
   Sciences) was used for recording and analysis done using FlowJo
   software (version 10.7.0), to determine the distribution of the cell
   cycle at G1, S and G2/M phases [[53]21].

2.6. Phytochemical screening of EARP

   The color based phytochemical screening tests of EARP were done
   according to standard methods [[54]14,[55]22].

2.6.1. Test for alkaloids

   Half a gram of plant extract was weighed and stirred in 2 ml of 1 %
   aqueous hydrochloric acid and heated in a boiling water bath for
   10 min. The mixture was filtered while hot and treated with
   Dragendorff's reagent. Turbidity or precipitation was used to indicate
   the presence of alkaloids.

2.6.2. Test for terpenoids

   Half a gram of the extract was defatted with hexane. The residue was
   then extracted in dichloromethane and the solution dehydrated with
   magnesium sulfate anhydride. The mixture was treated with 0.5 ml acetic
   anhydride followed by addition of 2 drops of concentrated sulphuric
   acid. A gradual appearance of green blue color was used to indicate
   presence of sterols while color change from pink to purple indicated
   the presence of triterpenes.

2.6.3. Test for saponins

   Half a gram of the plant extract was dissolved in 5 ml of distilled
   water and shaken for at least 5 min. Persistent frothing for at least
   half an hour indicated the presence of saponins. The quantity of froth
   was used to estimate quantity.

2.6.4. Test for flavonoids

   Two hundred milligrams of the extract was dissolved in 4 ml of 50 %
   methanol. The solution was warmed and then metal magnesium added. Five
   drops of concentrated sulphuric acid was added. Development of a red
   color indicated the presence of flavonoids while orange color showed
   presence of flavones.

2.6.5. Test for tannins

   The ferric chloride test was used. Half a gram of the extract was
   dissolved in 2 ml of distilled water and filtered. Two drops of ferric
   chloride was added to the filtrate. Development of a blue-black
   precipitate indicated the presence of tannins.

2.6.6. Test for phenols (ferric chloride test)

   2 ml of distilled water was added to 1 mg of plant extract followed by
   a few drops of 10 % aqueous ferric chloride solution. Formation of blue
   or green color indicated the presence of phenols.

2.6.7. Test for glycosides

   2 mg of plant extract was dissolved in 1 ml distilled water and then
   aqueous sodium hydroxide was added. Formation of a yellow color
   indicated the presence of glycosides.

2.7. Gas Chromatography/Mass Spectrometry (GC/MS) of EARP

   The compounds contained in EARP were identified through GC/MS QP-2010SE
   instrument (Shimadzu, Kyoto, Japan) with a BPX5 capillary column with
   dimensions 30m *0.25 mm * 0.25 μm film in width with a low polarity. In
   order to reach the isothermal temperature of 280 °C and maintain it for
   15 min and 30 s, the oven was configured to start at 50 °C and stay
   that way for 1 min before progressively raising the temperature by
   10 °C per minute. The injector's temperature was retained at 200 °C and
   its flow rate at 1.08 mL per minute. Helium was used as the carrier
   gas. As AS30000 auto-sampler connected to a GC in split mode with a
   split ratio of 10:1 automatically injected 1 μl of the diluted sample
   in solvent after a 4 min delay. The source of the ions and interface
   were adjusted to 250 °C and 200 °C, respectively. In full scan mode,
   the EI mass spectra above the range of m/z 35–550 and at 70 eV were
   then fetched. The chemical components in the extract were identified
   using the mass spectrum database of the National Institute of Standards
   and Technology (NIST) [[56]23].

2.8. In silico studies

2.8.1. Drug screening

   PubChem ([57]https://pubchem.ncbi.nlm.nih.gov/) was utilized to
   generate canonical SMILES of chemical compounds identified through
   GC/MS. Their ADMET (Absorption, Distribution, Metabolism, Excretion and
   Toxicity) attributes were gotten using the Swiss ADME online database
   ([58]http://www.swissadme.ch/index.php) and the pkCSM tool
   ([59]http://structure.bioc.cam.ac.uk/pkcsm). The parameters that were
   described in these databases were used to project the drug-likeness of
   the compounds on the basis of the following criteria: Lipinski's rule
   of five (molecular weight in Daltons <500, log P o/W, hydrogen bond
   acceptors ≤10 and hydrogen bond donors ≤5), oral bioavailability;
   interaction with cytochromes; crossing the blood-brain barrier (BBB) as
   well as the topological surface area (TPSA) [[60]24,[61]25].

2.8.2. Identification of therapeutic targets

   Genes associated with cervical cancer were assembled using Gene Cards
   ([62]https://www.genecards.org/), DISGeNET
   ([63]https://www.disgenet.org/), OMIM ([64]https://www.omim.org/) and
   NCBI ([65]https://www.ncbi.nlm.nih.gov/gene) databases with the phrase
   “cervical cancer” being the key word. All the results were merged and
   the duplicates were removed. Targets from EARP were gotten from Swiss
   Target Prediction ([66]http://www.swisstargetprediction.ch/),
   Similarity Ensemble Approach (SEA) ([67]https://sea.bkslab.org/) and
   Super-PRED
   ([68]https://prediction.charite.de/subpages/target_prediction.php)
   databases. The gene ids were acquired from the Universal Protein
   Resource (Uniprot), ([69]https://www.uniprot.org/). The outputs from
   the databases was put together and duplicates were deleted. Common
   genes between the targets and the cancer genes were gotten from an
   online Venn diagram generator
   ([70]https://bioinformatics.psb.ugent.be/webtools/Venn/) [[71]26].

2.8.3. Generation of the protein-protein interaction (PPI) network

   The similar genes were loaded into the STRING 12.0 database
   ([72]https://string-db.org/), where the minimum threshold was set at
   0.4. Cytoscape software (version 3.10.2) was utilized to analyze the
   topology based on the degree of centrality with the Cytohubba plug-in
   being used to generate the top 30 genes based on five centralities, MCC
   (Maximal Clique Centrality), ECC (Eccentricity), BN(Betweenness), DG
   (Degree) and CN (Closeness).

2.8.4. Gene ontology (GO) and Kyoto Encyclopedia of genes and Genomes (KEGG)
enrichment analysis

   The online enrichment tool Shiny GO version 0.80
   ([73]http://bioinformatics.sdstate.edu/go/) was utilized to conduct
   these studies [[74]27]. The parameters set were; species was “Human”,
   false discovery rate (FDR) cut-off was 0.05 and the number of pathways
   displayed were up to 20 [[75]22].

2.9. Molecular docking

   3D Structured Data File (SDF) format of the compounds selected after
   ADMET parameters were analyzed were gotten from PubChem database. The
   RCSB protein database ([76]https://www.rcsb.org/) was used to get the
   structures of the proteins of the top 30 genes. The filters applied
   were; the species was Human, the experimental method was X-ray
   Diffraction, the resolution was between 0.5 and 3.5 Å and the structure
   determination methodology was experimental. Using Discovery Studio
   2021, the targeted proteins were curated whereby the Gasteiger charges
   were added, water molecules were removed, all co-crystallized ligands
   removed and polar hydrogen were added. The 3D SDF compounds (ligands)
   were edited and transformed into the “pdbqt” format with the aid of the
   inbuilt Open Babel software in Pyrx program
   (([77]https://sourceforge.net/projects/pyrx/). The embedded Autodock
   Vina software in Pyrx program was then used to dock the prepared
   ligands to the prepared proteins. The size of the protein was used to
   centralize the three-dimensional grid box. Discovery studio 2021
   ([78]https://discover.3ds.com/discovery-studio-visualizer-download) was
   used to generate the 3D protein-ligand complexes with the lowest
   binding energies and their 2D interaction diagrams [[79]28].

2.10. Survival analysis

   Some of the genes with the lowest binding affinities were assessed with
   the aid of the GEPIA database ([80]http://gepia.cancer-pku.cn) to
   ascertain the correlation between their expression and overall patient
   survival and overall disease free survival [[81]29].

2.11. RNA isolation and quantification using RT-qPCR

   To validate the docked proteins as well as assess the effect of the
   EARP extract on key oncogenic genes, RT-qPCR was used. HeLa cells were
   seeded in T-75 culture flasks and upon attaining a confluence of >70 %,
   they were treated with the EARP extract at IC[50] concentration for
   48 h. After treatment, the flasks were rinsed with PBS thrice and RNA
   was obtained using the Pure Link™ RNA Mini Kit (Thermo Fisher
   Scientific, USA) extraction kit. A NanoDrop, ND-2000 spectrophotometer
   (Thermo Scientific, USA), was employed in the quantification of the
   extracted RNA. The RNA was preserved at −80 °C until use. 1 μg of RNA
   was converted to cDNA using the Firescript® RT cDNA synthesis kit
   (Solis Biodyne, Estonia). The cDNA was stored at −20 °C until use. The
   genes used were purchased from Macrogen (Macrogen, South Korea),
   ([82]Table 1). The HOT Firepol® Evagreen® qPCR Mix plus (ROX) (Solis
   Biodyne, Estonia) was utilized for quantitative RT-qPCR analysis using
   the Quant Studio™ 5 RT-qPCR system (Applied Biosystems, USA). A blank
   containing the master mix and the primers without the cDNA was used.
   Analysis of the melt curve verified that there was no non-specific
   binding. The targeted genes' mRNA levels were quantified using the 2
   ^−ΔΔ Ct method. GAPDH and β-actin controls were the internal references