An omics approach to delineating the molecular mechanisms that underlie the biological effects of physical plasma

Abstract

The use of physical plasma to treat cancer is an emerging field, and interest in its applications in oncology is increasing rapidly. Physical plasma can be used directly by aiming the plasma jet onto cells or tissue, or indirectly, where a plasma-treated solution is applied. A key scientific question is the mechanism by which physical plasma achieves selective killing of cancer over normal cells. Many studies have focused on specific pathways and mechanisms, such as apoptosis and oxidative stress, and the role of redox biology. However, over the past two decades, there has been a rise in omics, the systematic analysis of entire collections of molecules in a biological entity, enabling the discovery of the so-called “unknown unknowns.” For example, transcriptomics, epigenomics, proteomics, and metabolomics have helped to uncover molecular mechanisms behind the action of physical plasma, revealing critical pathways beyond those traditionally associated with cancer treatments. This review showcases a selection of omics and then summarizes the insights gained from these studies toward understanding the biological pathways and molecular mechanisms implicated in physical plasma treatment. Omics studies have revealed how reactive species generated by plasma treatment preferentially affect several critical cellular pathways in cancer cells, resulting in epigenetic, transcriptional, and post-translational changes that promote cell death. Finally, this review considers the outlook for omics in uncovering both synergies and antagonisms with other common cancer therapies, as well as in overcoming challenges in the clinical translation of physical plasma.

I. INTRODUCTION

Physical plasma (henceforth referred to as plasma) is an emerging treatment for cancer. Plasma, sometimes referred to as “the fourth state of matter,” is an ionized gas containing free electrons and positive ions. Plasma occurs naturally in the Sun, stars, interplanetary and interstellar space, as well as in lightning, auroras, and the ionosphere that surrounds the Earth. Plasma is also widely used in neon signs, fluorescent lamps, and welding, while medical applications are growing. Plasma is typically produced under low-pressure conditions where a high voltage is applied between electrodes in a gas such as helium, argon, nitrogen, or their mixtures^1,2 (Fig. 1). This process creates a discharge containing reactive chemical species, ions, electrons, and UV and electromagnetic radiation, all of which can elicit biological effects.^1,3 A convenient type of plasma is formed at atmospheric pressure in a flowing gas, known as an atmospheric pressure plasma jet. This plasma forms a plume that can be directed to activate a liquid.

FIG. 1.

Creation of an atmospheric pressure plasma jet. A pulsed or oscillating high voltage is applied between (a) a pair of electrodes surrounding a dielectric tube carrying gas at atmospheric pressure, or (b) one electrode inside and coaxial with a dielectric tube and another outside the tube. In both cases, a plasma plume is formed.

In medical applications, plasma can be used directly, where the plasma jet is applied onto cells or tissues, or indirectly, where the plasma jet is first applied to a solution and the plasma-activated liquid (PAL) is then used for treatment (Fig. 2).^1,4 In direct plasma treatment, in addition to the reactive oxygen and nitrogen species (RONS) created by the interaction of electrons, positive ions, neutral species, and UV radiation produced by the plasma,⁵ other plasma-associated stimuli, notably the electric and magnetic fields, can affect cellular and molecular mechanisms and signaling pathways. As the general biological effects of direct plasma have been comprehensively reviewed recently,^6,7 this review will focus on the effect of RONS that are present in both direct plasma applications and PAL. RONS inactivate microbial pathogens, and the antitumor effect of direct plasma treatment has been demonstrated in several cancers in vitro including melanoma,^8,9 leukemia,^10,11 and lung cancer.¹² For example, plasma treatment has been shown to selectively induce apoptosis of tumor cells without damaging normal cells in mice with transplanted tumors.¹³ In comparison, the literature on indirect PAL treatments is less advanced but growing. While most PAL applications in vivo involve injections into tumors or tissues,^14,15 studies have shown that the oral treatment of PAL on immuno-deficient mice reduced tumor size without any lethal or other acute effects.^16,17 Considering the known side effects of radiotherapy, the absence of significant safety concerns in plasma treatment makes it a desirable form of cancer treatment.

FIG. 2.

Two approaches to using plasma in medical applications. (a) Direct plasma treatment on cancer cells in vitro or on tissues and tumors in vivo. (b) Indirect plasma treatment to form plasma activated liquid (PAL), which is then applied to cancer cells in vitro or to tissues and tumors in vivo. Created with BioRender.com.

Several recent publications^11,18–20 have described the use of omics to reveal new mechanisms and pathways underlying the effects of plasma on cancer cells. The term “omics” refers to a comprehensive or global assessment of a set of molecules in a biological entity. In some cases, omics studies have provided additional insight into the roles played by key molecules in pathways such as apoptosis and oxidative stress, previously associated with plasma-induced cancer cell death.^8,21–25 In this systematic review, Sec. II describes the technologies and the development behind the more common omics, Sec. III examines the biological effects of RONS and plasma treatment, and Sec. IV showcases how omics have helped to reveal the biological pathways and molecular mechanisms that underpin plasma treatment of cancer. Finally, Sec. V discusses the potential use of these insights and information to uncover synergies, as well as antagonisms, with other common cancer therapies, supporting the further clinical translation of plasma.

II. APPLICATIONS OF OMICS TO ANALYZE BIOLOGICAL RESPONSES FOLLOWING PLASMA TREATMENT

A. What are omics studies?

The suffix “ome” in biology is used to denote a complete set of biomolecules in a biological entity, such as a genome for the complete set of an individual's genes, a transcriptome for the total set of RNA molecules in a tissue sample, and a proteome for the complete set of proteins expressed. Over the last three decades, omics, the comprehensive analysis of the respective ome, have led to the widespread detection of genes (genomics), epigenetics (epigenomics), mRNA (transcriptomics), proteins (proteomics), metabolites (metabolomics), and microbes (microbiomics). Compared to traditional biochemical and biophysical analyses, where a selected set of genes/proteins/metabolites are targeted and investigated, omics technologies typically analyze hundreds to thousands of these entities simultaneously (Table I). As such, omics can provide a more holistic and systematic study of molecules in a cell, tissue, or organism.

TABLE I.

Omics definitions and associated technologies.

Omics	Definition	Technologies
Genomics	The study of the structure, function, and expression of genes in an organism	High-throughput DNA sequencing
Transcriptomics	The study of genes expressed within a cell or organism by analyzing the total mRNA	RNA sequencing (RNA-Seq)
		RNA microarray
		Expressed sequence tag (EST)/serial analysis of gene expression (SAGE)
Metabolomics	The study of low molecular weight compounds or metabolites, such as amino acids, fatty acids, carbohydrates, or products of cellular processes present in a system	Gas chromatography (GC), liquid chromatography (LC), or capillary electrophoresis–mass spectrometry (CE–MS)
		Nuclear magnetic resonance (NMR) spectroscopy
		Fourier transform infrared spectroscopy
Proteomics	The study of abundance, structure, modification, and interaction of proteins within a cell, tissue, or organism	Protein microarray
		GC or LC–MS
		Tandem MS (MS/MS)
		NMR spectroscopy
Epigenomics	The genome-wide study of reversible modifications of DNA and associated proteins, including but not limited to DNA methylation or histone acetylation	Chromatin immunoprecipitation sequencing (ChIP-Seq)
		Pyrosequencing
		DNA sequencing
		DNA microarray

Each omics approach primarily provides information on one set of biological molecules using one or multiple technologies (Table I) with the aid of bioinformatics and the increasing availability of open-access databases. These strategies enhance understanding of normal physiological and disease processes, as well as the effect of treatments on a system. In cancer research, omics technologies have been widely used to identify genetic risk factors, novel drug targets, and biomarkers, as well as to improve current therapies.^26,27

Following the elucidation of the DNA structure by Watson and Crick in 1953, the field of molecular biology has rapidly evolved, propelled by the development of accompanying technologies, including automated DNA sequencing in the 1980s and the more recent next-generation sequencing (NGS) from the late 1990s.²⁸ NGS provides high-throughput, deep sequencing where a genomic region is sequenced multiple times, leading to high coverage and a large volume of data. Advances in bioinformatics and the creation of gene and metabolite databases have enabled researchers to analyze the information gathered from omics studies. DNA sequencing and advanced computing technology have enabled the mapping of the complete human genome. In recent years, owing to decreasing costs along with increasing speed and accuracy of sequencing, these technologies have become more readily available for use in omics studies. However, to study the biological responses to treatments of an organism, a genomics approach is often insufficient and ineffective since environmental influences, transcriptional controls, and post-transcriptional changes can significantly impact the final phenotype without imparting genomic changes. Hence, omics approaches such as epigenomics, transcriptomics, proteomics, and metabolomics are complementary and suited to revealing new information about the molecular effects of plasma treatment. For further information, the reader is referred to recent reviews on the use of multi-omics in cancer research.^29–33

B. Epigenomics

Epigenomics is the study of the complete set of epigenetic modifications on DNA and associated proteins (Table I). DNA in a cell is tightly wound around histone proteins, resulting in a compact DNA–protein complex called chromatin.³⁴ Epigenetic modifications cause a change in gene expression without alteration of the DNA sequence itself. The main epigenetic mechanisms involve histone modifications, such as acetylation and methylation, DNA methylation at CpG sites (DNA dinucleotides where a cytosine is followed by a guanine), and recognition and binding of microRNA to the 3′-untranslated region of mRNA.³⁵ CpG sites are often found at the start of a gene with methylation of the cytosine associated with gene silencing. Disruptive and aberrant interaction between epigenetic mechanisms induced by environmental and cellular stresses (e.g., oxidative stress, UV radiation, and lifestyle factors) may result in inappropriate gene expression or silencing, leading to cancer and other epigenetic diseases.

The involvement of DNA methylation and histone modification in cancer have been well-characterized, with studies showing that many cancer cells undergo changes in chromatin structure that cause a general hypomethylation of the whole epigenome, and hypermethylation of specific oncogenes.^36,37 Since oxidative stress is known to cause an imbalance in the interaction of histones and their modifiers (HDAC1, HMT1, and HAT1) and epigenetic regulators (DNMT1, DNMT3a, and MBD4) in cancer cells,³⁶ the production of reactive oxygen species from plasma treatment may provide a mechanism to induce desirable epigenetic changes.

To investigate plasma-induced epigenetic changes, pyrosequencing and chromatin immunoprecipitation sequencing (ChIP-Seq) have been used to analyze DNA methylation and histone modification, respectively (Table I).^18,38 Pyrosequencing is a “sequencing-by-synthesis” method where the real-time addition of nucleotides is quantitatively monitored through the conversion of the released pyrophosphate to a proportional light signal.³⁹ This method involves bisulfite treatment where unmethylated cytosine residues in the DNA are converted to uracil, while the methylated cytosine residues are conserved. Hence, after bisulfite treatment, the ratio of T and C is used to determine the degree of methylation at each CpG dinucleotide in the sequence (Fig. 3).

FIG. 3.

Schematic showing the key steps of pyrosequencing. Bisulfite treatment converts unmethylated cytosine to uracil and conserves methylated cytosine (C^m). During PCR amplification, the addition of nucleotides release pyrophosphate (PPi), which is converted to a proportional signal in the presence of sulfurylase and luciferase. By pyrosequencing analysis, any unmethylated cytosine, C, is measured as the relative content of T at the CpG site, whereas methylated cytosine (C^m) is measured as the relative content of C at the CpG site. Adapted from Ref. 40 (Pyrosequencing).

ChIP-Seq is another approach that combines chromatin immunoprecipitation and high-throughput DNA sequencing techniques to enable the identification of the binding sites of DNA-associated proteins. In this method, proteins such as histones and transcription factors are first covalently linked to their relevant DNA substrates, and then, the DNA–protein complexes are precipitated from solution through protein-specific antibodies⁴¹ (Fig. 4). The DNA is then extracted and sequenced, creating a high-resolution map of the binding sites for any given protein. Both pyrosequencing and ChIP-Seq provide high resolution and accurate information on epigenetic changes caused by cancer or its treatments. For more information, the reader is directed toward recent reviews regarding epigenomic techniques and their applications, particularly in cancer.^31,42–44

FIG. 4.

Schematic showing the key steps of ChIP-Seq. DNA binding proteins are covalently cross-linked to their relevant DNA substrate, and the DNA sequence is broken into fragments. Using protein-specific antibodies, the DNA–protein complexes are immunoprecipitated and then purified. The purified DNA is sequenced, and the reads are mapped to a reference genome. The resulting peaks correspond to motifs, or the DNA binding sites. Reprinted with permission from BioRender.com, see https://app.biorender.com/biorender-templates/t-5f20a148563a0600ad9d9f87-chip-sequencing for “ChIP Sequencing (2020)” (accessed October 21, 2021). Copyright 2020 Biorender.⁴⁵

C. Transcriptomics

Transcriptomics refers to the study of the transcriptome (Table I) mainly through the interrogation of messenger RNA (mRNA) in a sample. mRNA is a type of RNA that is transcribed from the DNA for protein synthesis. Transcriptomic analysis of gene expression allows detection of biologically significant coordinated trends that may not have been observed in more targeted assays of selected single genes. Two common techniques are used in transcriptomics: microarrays, which quantify a set of target sequences, and RNA sequencing (RNA-Seq), which analyses all mRNA sequences through high-throughput sequencing. In this section, we briefly review techniques that have been used in most, if not all, transcriptomic studies of plasma treatment of cancers. Recently, there has been growing popularity in the use of single-cell sequencing and spatial transcriptomics in cancer research. For further information, the reader is referred to excellent reviews on these techniques.^46–51

RNA microarray technology is a hybridization-based technique that involves the binding of complementary sequences to DNA fragments, arranged in rows and columns, on a chip (Fig. 5). DNA mutations or changes in gene expression can be identified through the comparison of the genomic hybridization between the different loci on the chip and complementary DNA (cDNA). In cancer, microarrays are used to classify tumors and detect cancer biomarkers, genes associated with chemoresistance, drug discovery, mutations, and single-nucleotide polymorphisms.⁵²

FIG. 5.

Schematic showing the key steps of RNA microarray technology. Total RNA is extracted from an experimental sample (e.g., cancer cells) and a reference sample (e.g., healthy cells). The two mRNA samples are then converted into complementary DNA (cDNA) through reverse transcription with each sample tagged with a different colored fluorescent probe, e.g., green for reference and red for experimental sample. The two cDNA samples are mixed and allowed to bind on a microarray, which contains thousands of probes where each probe corresponds to a known gene. Hybridization to a particular probe indicates the expression of that gene in the sample. Adapted from Ref. 53 (DNA Microarray).

While RNA microarrays are less expensive than RNA-Seq studies, the latter provides a more comprehensive differential gene expression analysis compared to microarrays, due to higher sensitivity, specificity, and range of expressed genes either at high or low levels^54–56 (Table II). RNA-Seq involves the quantification of RNA through next-generation sequencing (NGS, Fig. 6). Bioinformatics analysis of the more comprehensive RNA-Seq data will enable the identification of novel transcripts, alternative splicing forms, and allelic expression, thus providing insights into the pathways or genes affected by different treatments or diseases.

TABLE II.

Comparison between microarrays and RNA-Seq. Adapted from Ref. 57.

Feature	Microarray	RNA-seq
Principle	Hybridization	Cloning and sequencing
Required amount of RNA	High	Low
Resolution	Several to ∼100 base pairs	Single base
Sensitivity	Low	High
Dynamic range	Low	High
Target analysis	Known transcripts only	Novel transcripts identified
Identify alternative splicing forms	Limited	Yes
Identify allelic expression	Limited	Yes
Reproducibility	Yes	Yes
Cost	Medium	High, due to sample preparation and sequencing

FIG. 6.

Schematic showing the key steps of RNA-Seq. Total RNA is extracted and fragmented, followed by conversion into cDNA through reverse transcription. Adapters containing platform-specific sequences are added to enable the library fragments to bind to the sequencer. Libraries hybridize on the flow cells; then, DNA clusters are amplified producing millions of copies of ssDNA. Complementary modified nucleotides with a fluorescent tag and reversible terminator bind to the DNA fragments, and the fluorescence is recorded. The DNA reads (or clusters) are aligned by comparing overlapping regions, resulting in an assembled sequence. Adapted from Ref. 58 [Next Generation Sequencing (Illumina)].

D. Metabolomics

Metabolomics is the study of the complete set of metabolites in a cell, tissue, organ, or organism. In metabolomics, metabolites are usually defined to include any biological molecule of less than 1500 Da, not just intermediate molecules from cellular metabolism. Metabolomic data provide information on molecular alterations caused by genetic, epigenetic, proteomic, and environmental factors.⁵⁹ Thus, metabolomics is more sensitive to the state of the cells or tissues, and their molecular phenotype than the other omics. Analysis of the metabolite profile can be used to identify biomarkers that can help improve early diagnosis of a disease or predict treatment response and survival.^60,61

The two most common techniques used in metabolomics are mass spectrometry (MS), due to its high selectivity and sensitivity,⁶² and nuclear magnetic resonance (NMR) spectroscopy, for its superior reproducibility and resolution. MS differentiates molecules based on their molecular mass and fragments, whereas NMR spectroscopy reveals direct linkages between atoms and their spatial proximity from their normalized frequency, relative intensity, and signal patterns. The choice of technique depends on the amount and complexity of sample available, and the purpose of the metabolomic study, as MS- and NMR-based techniques have vastly different sensitivities and provide different types and detail of data.

Due to their higher sensitivity, MS-based techniques tend to be more suitable than NMR-based techniques for studies on the effects of plasma treatment. Some metabolites of interest in plasma-treated samples are likely to be present in below micromolar concentrations. Therefore, MS is often coupled with separation techniques that increase resolution and sensitivity, enabling analysis of secondary metabolites for which the detection limit is even lower. Metabolomic studies following plasma treatment to date^10,11,20 have utilized gas chromatography (GC) or capillary electrophoresis (CE) separation techniques with time-of-flight MS (TOF–MS) (Fig. 7). For further information on the applications of metabolomics in cancer research and clinical oncology, the reader is referred to the recent reviews in the field.^{29,31,63–70}

FIG. 7.

Schematic showing different types of separation technique followed by mass spectrometry. The compounds in the sample are separated and prepared for MS analysis by either (a) capillary electrophoresis, based on their electrophoretic mobility in an applied high voltage environment and subsequent vaporization through electrospray ionization (ESI), or (b) gas chromatography (GC), where the compounds separated are usually organic, gaseous, or volatile. In mass spectrometry (c), the compounds are ionized and separated based on their mass-to-charge ratio (m/z). Compounds are distinguished both qualitatively and quantitatively based on their m/z and abundance. Adapted from Ref. 71 (Mass Spectrometer Principle).

TOF-MS involves the acceleration of ions through the same potential difference using an electric field. These ions then move through a “flight” tube to a detector where their mass is determined from the time taken to reach the detector. This method enables the detection of signals from polar and nonpolar metabolites within a broad molecular weight range.

In GC–TOF–MS, samples are vaporized, and compounds separated using a capillary column containing a stationary inert carrier gas. Compounds advance through and elute from the column at different times depending on their polarity and boiling point. After elution, the compounds are ionized and fragmented by the MS, with the fragmented ions separated and detected based on their different m/z ratios. GC–TOF–MS is especially useful for separating and detecting volatile organic compounds, including reactive oxygen species produced by a plasma.

In CE–MS, through the application of high voltage to a sample, electro-osmotic flow is produced within the CE capillary, which then separates ions based on their electrophoretic mobility (Fig. 8). These compounds are then identified by MS. The main advantages of CE–MS are its fast separation, robustness, and simple separating principle, with high reproducibility and capacity to analyze heterogeneous samples with interfering compounds.⁷² However, the methodology behind CE–MS is currently less developed than for GC–MS and other MS-based techniques. GC- and CE–MS are highly suitable for studying the effects of plasma treatment as both techniques are sensitive to small variations in the concentration of redox and apoptosis-related metabolites such as glutathione.

FIG. 8.

Schematic showing how molecules are separated in capillary electrophoresis. The walls of the separation capillary consist of negatively charged fused silica. The positive buffer running through the capillary is attracted to the capillary walls, neutralizing the walls, and enabling the compounds in the sample to flow evenly in the middle section of the tube. Under the high electric field between the anode and cathode, compounds migrate at a rate dependent on their charge, viscosity, and size. Created with BioRender.com.

E. Proteomics

Proteins are the molecules that carry out most of the work in a biological system following the process of translation from mRNA. Therefore, the perturbed epigenomic and transcriptomic changes in cancerous cells often have downstream effects on protein abundance, localization, and function. Proteomics is the comprehensive study of proteins in a sample, and can provide information about their location, abundance, post-translational modifications, and their interactions with other proteins. Thus, proteomics studies can reveal individual or groups of proteins that deviate from normalcy under disease states or following treatments. Proteomic approaches have proved informative in cancer research since cancer cells are known to be more sensitive to proteasome inhibition due to increased cell proliferation and loss of protein translation control.⁷³ Studies can also be conducted to assess protein changes following treatment. For example, comparing the proteome before and after plasma treatment may reveal the individual proteins supporting cancer cell death and/or trends in groups of proteins with similar function, implicating the protein networks at play.

Proteomics studies are usually conducted using protein microarrays or MS-based techniques. Analogous to RNA microarrays, protein microarrays involve the immobilization of purified proteins onto a glass slide containing a library of antibodies, aptamers, or specific antibodies to the proteins of interest. However, since protein microarrays do not provide full coverage of the proteome, the high-throughput and sensitivity of MS-based techniques like those described in Sec. II D are often more appropriate, with the added advantage of relatively low costs. However, due to the larger size of proteins compared with metabolites, liquid chromatography (LC)-MS is more common than GS- and CE–MS in proteomics. Following a proteomics study, the proteins of interest identified will typically need to be subsequently validated with more specific experimental analysis (such as those mentioned in Fig. 9). For a more comprehensive discussion, the reader is directed toward the following recent reviews on proteomics.^{50,70,74–76}

FIG. 9.

Schematic depicting key steps in a proteomic workflow. Cancer cells, both untreated (control) and treated with direct (pictured) or indirect cold atmospheric plasma, are collected and lysed to obtain the protein samples. The samples are analyzed by combined chromatography and MS-based techniques to reveal their identity and abundance, identifying individual or groups of proteins that deviate from the control. In favorable cases, protein modifications can also be detected. Following this, perturbed proteins and their networks can be further studied using additional experimental techniques to validate the proteomic findings and to unveil their importance. Created with Biorender.com.

III. BIOLOGICAL EFFECTS OF PLASMA TREATMENT

A. The molecular mechanisms and roles of reactive species

The effects of plasma treatment both in vitro and in vivo are well documented.^{9,21,22,25,77–79} Oxygen has two unpaired electrons in separate orbitals in the outer electron shell, making it susceptible to reduction by plasma. ROS include free oxygen radicals such as superoxide (O₂^•−), hydroxyl (^•OH), peroxyl (RO₂^•), and alkoxyl (RO^•), as well as non-radicals such as hypochlorous acid (HOCl), ozone (O₃), dioxygen (O₂), and hydrogen peroxide (H₂O₂). Meanwhile, RNS are generated when nitric oxide (NO) and other NO-containing species react with ROS. Here, we refer to both classes of reactive species collectively as RONS.

In mammals, RONS form as a natural by-product of metabolism, for example, a major source originates from processes that cause the uncoupling of electron transport with the mitochondria.⁸⁰ RONS play a diverse range of cellular roles, including cellular proliferation, differentiation and migration, iron hemostasis, and inflammation (Table III),^80,81 and contribute to key biological processes including growth, tissue repair, regulation of vascular tone, and angiogenesis⁸² (Fig. 10). In a healthy cell, the generation of RONS is tightly controlled by multiple enzyme systems such as endoplasmic reticulum-bound and cytoplasmic enzymes. RONS propagation is terminated by enzymatic and nonenzymatic antioxidant systems (AOS), which cause low intracellular RONS concentrations.⁸² These systems allow reversible oxidative/nitrosative modifications of redox-sensitive residues in regulatory proteins. These modifications often act as redox switches in cell signaling pathways and transcription factors. Therefore, a delicate cellular redox homeostasis has to be maintained between RONS production and AOS activity.^80,81

TABLE III.

RONS regulation of pathways. Adapted from Ref. 81. Nrf2: nuclear factor erythroid 2-related factor 2; Ref1: redox-factor 1; ATM: ataxia-telangiectasia mutated; IRP: iron regulatory protein; ASK1: apoptosis signal-regulated kinase 1; PI3K: PI3 kinase; PTP: protein tyrosine phosphatase; Shc: Src homology 2 domain containing.

Cellular pathways	Enzymes involved	Main species involved	References
Antioxidant, anti-inflammatory response	Nrf2	mtROS (H2O2, O2•−)	83–85
	Ref1
DNA damage	ATM	H2O2	83, 86
Iron hemostasis	IRP	NO•, ONNO−, H2O2, O2•−	83, 84, 87
Cellular proliferation, survival, differentiation	ASK1, PI3K, PTP, Shc	H2O2	83, 84, 86

FIG. 10.

Schematic showing the pathways affected by RONS. The production of ROS and RNS regulates several pathways, including iron hemostasis, signal transduction, protein modifications, DNA damage, cell proliferation, survival and differentiation, apoptosis, antioxidant, anti-inflammatory response, and cellular redox balance and metabolism. Among the RONS, hydrogen peroxide (H₂O₂), superoxide (O₂^•−), nitric oxide (NO^•), and peroxynitrite (ONOO⁻) are the main species influencing these pathways. Created with BioRender.com.

B. Plasma-induced protein modifications

Elevated RONS cause damage to biomolecules such as proteins, lipids, and nucleic acids, resulting in misfolding, aberrant interactions, and aggregation, potentially causing cell death and apoptosis.⁸⁰ Most proteins contain cysteine and methionine residues, which are sensitive to the local redox environment, as well as aromatic residues that can react with RONS. Plasma treatment has been reported to cause two main types of protein modifications: oxidative modifications^88–92 and structural changes.⁹³ Plasma-induced protein oxidation can further be categorized into hydroxylation, nitration, dehydrogenation, and dimerization.⁹⁴ RONS generated by plasma hydroxylate and nitrate the benzene ring of phenylalanine and tyrosine side chain, where the H groups are replaced by OH groups and eventually by NO₂ groups. Similarly, sulfur-containing carbon-chain amino acids can be easily modified by RONS,⁹⁴ where the free sulfhydryl groups are oxidized and sulfonated.

RONS can also lead to protein degradation and structural changes. The initial cleavage of peptide bonds are prompted by the attack of plasma-induced hydroxyl radical (^•OH) on the αC⁹². The intermediates are further attacked by RONS, creating two peptide fragments. This mechanism is supported by the decrease in protein concentration over time with plasma treatment.^92,93 Peng et al. reported the complete destruction of protein structures after minutes.⁹³ Furthermore, the total content of α-helix in proteins decreased while β-sheet increased, suggesting protein secondary structures are sensitive to plasma.^88,93

C. Imbalance of reactive species in cancer

The cancer microenvironment is often hypoxic, or low in oxygen, requiring cancer cells to adapt their metabolism to satisfy the growing demand for energy and nutrients for proliferation and survival.⁸² Hypoxia causes the imbalance between RONS generation and AOS activity, leading to an elevated production of RONS in cancer cells. Indeed, elevated RONS and protein oxidation adducts, accompanied by decreased AOS activity, have been discovered in several human cancer types in vitro including gastric, breast, prostate, and hepatic cancers.^95–98 In addition, RONS-induced stress changes the activity of transcription factors and signal transduction pathways, leading to protein degradation and cytoskeletal disorganization, which further contribute to the higher level of proteotoxic stress in cancer cells in comparison to normal cells.⁸²

D. How plasma-induced RONS preferentially kill cancer cells

The high concentration of RONS and other reactive species generated by plasma induces DNA damage and cell death.^24,99 In particular, the toxicity of plasma treatment had been attributed to long-lived reactive species H₂O₂ and NO₂^-, with H₂O₂ as the central agent of plasma-induced oxidative stress.¹⁰⁰ The synergy between H₂O₂ and NO₂^-, in combination with the acidic pH, drives plasma-induced toxicity in tumor cells but not in normal cells as NO₂^- production amplifies H₂O₂ toxicity.^100,101 RONS production trigger mitochondrial dysfunction followed by MAPK signaling, thus inducing an apoptotic signal cascade.¹⁰² Following treatment with both direct plasma and PAL, increases in the concentration of proteins involved in the apoptosis signaling pathway, such as caspase-3, caspase-7,^15,103 and caspase-9,²² have been reported.

Although plasma treatment causes apoptosis in cancer cells, a high therapeutic ratio is needed for effective cancer treatments. The therapeutic ratio of a treatment is defined as the reduction in survival of cancer cells relative to normal cells. Given that cancer cells typically have a higher baseline of redox reactions and higher concentration of intracellular RONS than normal cells,^1,104 it is more difficult to maintain sufficient antioxidant levels in cancer cells. A common hypothesis for selective apoptosis of cancer cells is that the increased concentration of intracellular RONS prevents antioxidant genes from functioning properly.^1,104 This RONS sensitivity contributes to the selectivity and anti-neoplastic effect of plasma treatment. It is likely that this effect is further amplified by the impaired DNA repair mechanisms and decreased levels of signaling molecules supporting survival in cancer cells.^103,105 Many cell-based studies show that direct plasma and PAL selectively induce cancer cell death and have little to no effect on normal cells.^23,103,106 The strong anti-cancer capacity of PAL has been reported for glioblastoma cells,^99,103 bladder cancer cells,¹⁰⁷ and lung carcinoma cells¹⁰⁸ in vitro. These promising in vitro results suggest a large therapeutic ratio for plasma treatment. However, more work is needed to understand the cancer cell selectivity of plasma treatment, especially the exact mechanisms and their potential interplay.

IV. HOW OMICS HAVE INFORMED THE MOLECULAR MECHANISMS BEHIND BIOLOGICAL RESPONSES TO PLASMA TREATMENT

There has been a surge in publications reporting omics studies of cancer cells following plasma treatments, including identification of critical pathways, ranging from epigenetic to transcription and post-translational changes (Table IV).

TABLE IV.

Pathways implicated in plasma treatment of cancer cells revealed by omics studies.

Omics type	Omics technology	Pathways revealed to be implicated in plasma treatment of cancer cells	Cancer cell type	Year	Reference
Epigenomics	Pyrosequencing	Cellular movement, connective tissue development and function, tissue development	Breast	2015	38
		Cell-to-cell signaling and interaction, cell death and survival, cellular development
	Chromatin immunoprecipitation sequencing	Cellular compromise, DNA replication, recombination, repair, and cell cycle	Breast	2017	18
Transcriptomics	RNA microarray	MAPK and p53 signaling pathway (cell cycle control); MEKK, GADD, FOS, and JUN gene expression (stress signaling)	Lung	2015	12
		AP-1 and Nrf2 signaling (oxidative stress regulator); Notch signaling (cell development)	Prostate	2020	19
	RNA sequencing	p53 (cell cycle control) and hypoxia pathways	Oral	2017	109
		MAPK signaling	Skin	2020	110
		Upregulation of ATF4 and CHOP (antioxidant stress response and apoptosis)	Head and neck	2021	111
		EGR1/GADD45α stress signaling activation	Thyroid	2021	112
Metabolomics	Gas chromatography–mass spectrometry (GC–MS)	Beta-alanine metabolism, propanoate metabolism, and linoleic acid metabolism	Multiple myeloma	2018	113
		d-glutamine and d-glutamate metabolism	Acute myeloid leukemia	2018	10
		Alanine, aspartate, and glutamate metabolism	Acute myeloid leukemia	2021	11
	Capillary electrophoresis–mass spectrometry (CE–MS)	Glycolysis pathway and pentose phosphate pathway	Glioblastoma	2019	114
		Impaired regulation of lipid and amino acid metabolism	Glioblastoma	2020	20
Proteomics	Nano-liquid chromatography–mass spectrometry (NLC–MS)	Cell adhesion, response to stress and infection, and extracellular matrix receptor interaction	Colorectal carcinoma and skin melanoma	2017	73

A. Epigenetic changes

Epigenetic modifications are known to be an underlying cause of cancer. Almost 50% of genes causing hereditary forms of cancer are dormant, inactivated by promoter hypermethylation and epigenetic silencing.³⁷ An epigenomics study by Park et al.³⁸ examined the effect of plasma treatment on MCF-7 and MDA-MB-231 breast cancer cells. Global methylation changes were monitored by examining the repetitive elements LINE1 and Alu. Pyrosequencing analysis revealed that plasma treatment induces hypomethylation at a CpG site on Alu in the estrogen-negative MDA-MB-231 cells but not on LINE1 or other cell lines, indicating that plasma acts in a cell type-specific and CpG site-specific manner. Further microarray and ingenuity pathway analysis (IPA) of the genome-wide methylation changes revealed the cell-to-cell signaling and interaction, cell death and survival, and cellular development pathways as the most significant networks impacted by plasma treatment because of changes in methylation status. Similarly, another study of plasma-induced histone modification through ChIP-Seq in MCF-7 breast cancer cells¹⁸ indicated the cellular compromise, DNA replication, recombination, repair, and cell cycle pathway as the top network showing altered H3K4me3 levels following plasma treatment. In this network, oncogenes, such as MCM8 and HELLS, were hypomethylated, while tumor suppressor genes, such as HELQ and MPC1, were hypermethylated. Despite the minimal plasma-mediated histone modification observed at a global level, local histone dysregulation of cancer-related genes (e.g., SULF1, PRPS1, and MLF2) inhibit cancer cell proliferation. Both studies demonstrate that the effects of plasma can be partly attributed to its epigenetic modifications of key cancer-relevant molecules in apoptosis and tumorigenesis.

B. Gene expression alterations

A transcriptomic study by Yang et al.¹¹⁰ investigated the impact of PAL on TE354T basal cell carcinoma and HaCaT keratinocyte cell lines. Differential gene expression analysis using RNA-Seq showed that following PAL treatment, the TNF and MAPK signaling pathways were significantly upregulated. These pathways are the two major pathways in mammalian apoptosis. The extrinsic death receptor pathway is activated by the binding of the tumor necrosis factor (TNF)-family to the death receptor. Meanwhile, the intrinsic mitochondrial death pathway is stimulated by the oxidative stress initiated by the activation of the mitogen-activated protein kinase (MAPK)-group cascades.⁸ These findings support existing knowledge about the induction of apoptosis by plasma treatment where Ishak et al.⁸ have shown that ROS production is mediated by the TNF pathway, leading to apoptosis in melanoma cells. Similarly, Xiang et al.¹¹⁵ demonstrated that increased ROS production in triple-negative breast cancer cells inhibits the initially hyper-activated MAPK/JNK pathway in cancer cells, causing reduced cell proliferation.

While omics approaches can be used to verify existing knowledge, their high-throughput nature can also provide new insights. For example, Notch, one of the genes found to be hypomethylated by the epigenomic analysis of Park et al.,³⁸ was discovered to be involved in a potential resistance mechanism by progenitor cells following plasma treatment.¹⁹ A transcriptomic investigation by Packer et al.¹⁹ involving direct plasma treatment of primary prostate cells found that plasma treatment activated Notch signaling in progenitor cells. Notch signaling plays a vital role in prostate cancer metastasis as it is implicated in cell migration and invasion, apoptosis resistance, cell proliferation, hypoxia, and determining the cell fate.¹¹⁶ Plasma-induced activation of Notch signaling in prostate progenitor cells maintains the self-renewal capabilities of prostate cancer stem cells, enabling them to resist plasma treatment. The addition of a Notch inhibitor has been shown to reduce Notch signaling, thereby causing progenitor cell differentiation, leading to reduced resistance to plasma treatment and increased cell death. Hence, the results from this transcriptomic study suggest ways to increase the effectiveness of plasma treatment by providing novel insights into the existence of cell type-specific resistance mechanisms.

C. Metabolomic changes

Apart from Notch signaling, Packer et al.¹⁹ also revealed the activation of the Nrf2 signaling pathway by plasma treatment. The phosphatidylinositol 3-kinase/AKT/mechanistic target of rapamycin (PI3K/AKT/mTOR) pathway activates nuclear factor-like 2 (Nrf2), a redox-sensitive transcription factor that regulates redox by reducing glutathione disulfide (GSSG) to glutathione (GSH).¹¹⁷ Analysis of the metabolomic profiles of U251SP glioblastoma cells by Ishikawa et al.²⁰ shows that cells, following PAL treatment, are in a reductive state as emphasized by the high intracellular glutathione redox ratio (GSH/GSSG). PCA analysis of CE–MS results revealed GSH and GSSG to be among the metabolites with the highest absolute factor loading values in Principal Components 1 and 2 (PC1 and PC2), respectively. Therefore, both the transcriptomic and metabolomic results demonstrate that the mechanism behind plasma-regulated cell death can be attributed to dysregulation of redox homeostasis in cancer cells, leading to significant metabolomic changes.

D. Proteomic changes

Several studies have employed MS-based approaches to examine the proteome post plasma treatment in leukemia,^118,119 melanoma, and colorectal cancer cell lines,⁷³ identifying common protein networks affected by plasma treatment. In summary, plasma treatment was found to heavily regulate proteins involved in antioxidant activity, protein degradation, signaling pathways, cell growth, proliferation, and apoptosis. Specifically, several studies^73,118,119 identified increases in post-treatment protein abundance of MAPK, nuclear factor erythroid 2-related factors (e.g., Nrf2) and peroxiredoxins, important proteins that regulate protein phosphorylation, cellular resistance to oxidation, and peroxide levels, respectively, in agreement with gene expression studies.

Given the building evidence implicating protein phosphorylation events in response to cancer development and plasma treatment, it is perhaps unsurprising that emerging proteomic studies have also focused on characterizing phosphorylation events proteome-wide, otherwise known as kinomics/phospho-peptidomics.¹²⁰ For example, changes in phosphorylation state in response to indirect plasma treatment of osteosarcoma and pancreatic cancer cell lines have been reported, and they revealed system-wide hyper- and hypo-phosphorylated peptides, which (in)directly modulate autophagy, apoptosis, and ferroptosis signaling pathways.^120,121

V. DISCUSSION, CONCLUSION, AND FUTURE DIRECTIONS

Omics approaches have provided many new insights into the mechanisms behind the action of plasma treatment. Analysis of the epigenetic, transcriptomic, and metabolomic profiles of cells has revealed how the increased RONS production by plasma treatment activates cellular pathways such as MAPK, TNF, Notch, and Nrf2 signaling, causing changes in cellular metabolic processes, thus leading to cell death. The molecular mechanisms behind plasma action are beginning to be uncovered, and omics approaches have the potential to fill gaps in our understanding.

To date, most omics studies have been confined to a single or closely related set of technique(s). Multi-omics studies or meta-analysis, integrating different omics approaches will offer a more complete understanding of the mechanism of plasma treatment. For example, epigenomic studies have demonstrated that the effects of plasma can be partially attributed to its epigenetic modifications of histones via methylation, acetylation, and phosphorylation.^38,122 However, disagreements remain on whether plasma action occurs via direct damage to tumor cell DNA,¹²³ or via downregulating proliferative pathways in tumor cells thereby triggering apoptosis. It is likely that further analysis of proteins and transcripts involved in DNA repair using proteomics, transcriptomics, and epigenomics will enable understanding of the contribution of plasma to DNA damage.¹²² Similarly, some studies have suggested that the effect of plasma may be due to apoptosis mediated through a mitochondria-dependent pathway,^111,124,125 while other studies have raised the possibility of necrosis.²¹ Using a combination of proteomics, transcriptomics, and metabolomics or a meta-analysis to correlate the enzymes, metabolites and free radicals involved in plasma action will be helpful in resolving these issues and may reveal the interactions and interplay between different molecular layers following plasma treatment. To facilitate multi-omics studies, the adoption of more standardized protocols in plasma medicine, along with the exploration and integration of databases using a system biology approach will be valuable, as shown in the mixOmics project (http://mixomics.org/) and in several recently developed web applications.^126,127 Another exciting development is the use of machine learning approaches for multi-omics analysis, which are rapidly gaining popularity¹²⁸ and the adoption of multi-omics to single-cell analysis,¹²⁹ which is now possible.

Most research studies published to date on plasma action are in vitro studies using cell lines. Since the effects of plasma treatment are often cell type-specific, and whole organism studies are better suited to address aspects of cancer growth and progression such as angiogenesis,¹³⁰ metastasis, and relapse, much more information can be gained by using omics technologies to study responses to plasma treatment in animal models and clinical trials. Omics can help detect potential toxicities of plasma treatment and provide a tool for effective and safe dosages to be established.¹³¹ The integration of increasingly common personal omics data will help researchers tease out the interplay of the genome, epigenome, transcriptome, proteome, and metabolome in the human body following plasma treatments. Indeed, multi-omics approaches have already been used to identify biomarkers for radiation resistance in tumors,¹³² and these findings will help to inform whether plasma therapy can complement or enhance common cancer treatment modalities. Omics studies on biofluids, such as urine and blood, may provide a non- or minimally invasive method to evaluate the effectiveness of the use of plasma in clinical settings.¹³³ In summary, a multi-omics approach to plasma research has much to offer in the development of plasma treatment of cancer, from delineating the molecular mechanisms of action to predicting the likely response for individual patients.

AUTHOR DECLARATIONS

Conflict of Interest

The authors have no conflicts to disclose.

Author Contributions

Lou Irish Sta Ana Gonzales: Conceptualization (supporting); Data curation (lead); Formal analysis (lead); Investigation (lead); Visualization (lead); Writing – original draft (lead); Writing – review & editing (lead). Jessica Wenjie Qiao: Formal analysis (supporting); Investigation (supporting); Writing – original draft (supporting). Aston W. Buffier: Formal analysis (equal); Investigation (equal); Writing – review & editing (equal). Linda J. Rogers: Supervision (equal); Writing – original draft (supporting); Writing – review & editing (supporting). Natalka Suchowerska: Conceptualization (supporting); Funding acquisition (supporting); Resources (supporting); Supervision (equal); Writing – original draft (supporting). David R. McKenzie: Conceptualization (supporting); Funding acquisition (supporting); Resources (supporting); Supervision (equal); Writing – original draft (supporting); Writing – review & editing (supporting). Ann Kwan: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Project administration (equal); Resources (equal); Supervision (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal).

DATA AVAILABILITY

Data sharing is not applicable to this article as no new data were created or analyzed in this study.