Protein biomarkers for radiation injury and testing of medical countermeasure efficacy: promises, pitfalls, and future directions

ABSTRACT

Introduction

Radiological/nuclear accidents, hostile military activity, or terrorist strikes have the potential to expose a large number of civilians and military personnel to high doses of radiation resulting in the development of acute radiation syndrome and delayed effects of exposure. Thus, there is an urgent need for sensitive and specific assays to assess the levels of radiation exposure to individuals. Such radiation exposures are expected to alter primary cellular proteomic processes, resulting in multifaceted biological responses.

Areas covered

This article covers the application of proteomics, a promising and fast developing technology based on quantitative and qualitative measurements of protein molecules for possible rapid measurement of radiation exposure levels. Recent advancements in high-resolution chromatography, mass spectrometry, high-throughput, and bioinformatics have resulted in comprehensive (relative quantitation) and precise (absolute quantitation) approaches for the discovery and accuracy of key protein biomarkers of radiation exposure. Such proteome biomarkers might prove useful for assessing radiation exposure levels as well as for extrapolating the pharmaceutical dose of countermeasures for humans based on efficacy data generated using animal models.

Expert opinion

The field of proteomics promises to be a valuable asset in evaluating levels of radiation exposure and characterizing radiation injury biomarkers.

KEYWORDS:

• Proteomics
• radiation injury
• omic platforms
• countermeasures
• cytokines
• gamma-radiation
• mixed-field radiation

1. Introduction

Radiological and nuclear injuries as a result of natural calamities, nuclear power plant safety failures/accidents, military conflicts, or terrorist attacks will no doubt need medical interventions [1–3]. Under worst-case scenarios involving nuclear plant disasters and associated core meltdowns, there would no doubt be devastating consequences to human health and to the environment by way of large releases of radioactivity [4,5]. The events associated with the Fukushima Daiichi nuclear reactor disaster illustrate the health hazards posed by accidental radiation exposure [6]. Full and proper radiological exposure management most certainly includes but is not limited to dose assessment and the identification of the individuals requiring medical intervention.

Acute radiation syndromes (ARS) manifest as hematopoietic (H-ARS), gastrointestinal (GI-ARS), and neurovascular syndromes (NV-ARS) after exposures to high doses of radiation delivered at relatively high dose rates. Clinical reports have documented that individuals exposed to the doses of 6–10 Gy will die generally within a few weeks due to GI-ARS. While lower doses of 2–6 Gy are not lethal immediately, they compromise the integrity of the hematopoietic system and if not treated, subject the individual to risk of death within the months following exposure [7,8]. However, some exposed individuals might manifest little signs of ARS (due to lower levels of radiation exposure or perhaps partial-body exposure) but later develop delayed-types of pathologies (e.g. organ system fibrosis, progressive kidney and cardiovascular diseases, cancer, etc.). Such late arising diseases have been documented in follow-up studies of survivors of the atomic bombing of Hiroshima and Nagasaki that ended World War II, as well as the survivors of the Chernobyl nuclear accident [9–11]. These symptoms are known as late effects or delayed effects of acute radiation exposure (DEARE). Accordingly, improved analytics that allow for the identification, validation, and regulatory approval of biomarkers capable of accurately estimating the absorbed dose appears to be the forefront of the effort to better clinically manage individuals/victims of unwanted radiation exposure. Such biomarkers will be helpful for making medical decisions for the treatment of exposed individuals [12,13].

The current clinical strategy for assessing an individual’s radiation exposure is based on the observation of symptoms developed over time [8]. These methods are time-consuming and not practical for real-time, large-scale exposure scenarios. Furthermore, the latent period is a critical time in which therapeutic intervention could help save lives. Therefore, there is an urgent and unmet need to develop biochemical or molecular biomarkers that could help to develop and to implement effective medical countermeasures (MCMs) for radiation injury in heavily exposed individuals, and thus, provide a basis for treatment decisions. An important consideration of pre-clinical biomarker development is that the appearance of clinical symptoms may lag behind the underlying pathophysiological changes. The understanding of exactly when an exposed individual transitions from having curable exposure symptoms to more refractive pathologies remains elusive. Having specific biomarkers of radiation exposure and related dose estimates would perhaps allow tissue-specific injuries to be prognosticated, thus allowing for more efficient triage and clinical interventions [14,15].

Over many decades, attempts have been made to estimate the an individual’s received radiation dose using hematological, biochemical, and cytogenetic parameters [12]. Currently, lymphocyte depletion kinetics, clinical observation, and dicentric chromosome assays are used for post-exposure dose assessment. These assays are highly technical, time-consuming, and labor intensive. Thus, there is an urgent need to identify novel biomarkers that are not only robust and sensitive tofull range and conditions of radiation exposure, but also to employ techniques that are minimally invasive. A biomarker is an objective feature which can be measured and can indicate a pathological, biological, or therapeutic effect [16]. It may exhibit biological processes related to the disease or a step downstream of the early effect. Biomarkers can be grouped into diagnostic markers to triage people and prognostic markers to be used as appropriate therapeutic options. A predictive biomarker may predict a favorable effect or indicate an adverse effect of a specific strategy. A composite biomarker is comprised of several molecules combined in a particular algorithm to achieve a single output. Biomarkers can serve as early detection surrogates for injuries or diseases and can have significant impact on diagnosis and therapy. Biomarkers should be easily detectable and quantifiable and provide important information. Another type of biomarker is a proteomics biomarker, which is a biomarker that is derived from the analysis of proteins or peptides in a biological sample, such as blood, urine, saliva, or tissue. The discipline of proteomics identifies, characterizes, and measures proteins expressed under a variety of conditions. Proteomics biomarkers can reflect the molecular mechanisms and pathways involved in a disease or a response and can provide information on the functional status and dynamics of the biological system. Since the majority of biological and disease processes manifest at the proteome level, proteomics complements to the approaches of genomics and transcriptomics. In brief, there is tremendous interest in proteomics to develop new biomarkers for diagnosis, prognosis, and prediction of diseases, and to accelerate drug development [17,18]. A radiation (injury) biomarker is a type of proteomics biomarker that can be used to assess the exposure, dose, effect, or susceptibility of an individual or a population to ionizing radiation. Radiation biomarkers can help to identify and stratify individuals who have been exposed to radiation, either accidentally or intentionally, and to guide the appropriate medical intervention and treatment. There is a necessity for biomarkers of total-body irradiation (TBI) and partial-body irradiation (PBI), which induce H-ARS and GI-ARS, and also markers of longer-term (delayed and late effects) radiation-induced lesions, such as fibrosis and necrosis [12]. There are several reports of proteomes as radiation injury biomarkers [19–21].

We limit the scope of our discussion to the results of proteome analysis in various animal models using acute lethal doses of different qualities of radiation exposures. Furthermore, emphasis has been given to recent studies using the latest high-throughput proteomic platforms for identifying and validating proteins as biomarkers for radiation injury and radiation MCM effects.

2. Techniques used in proteomic studies

A number of techniques are currently being applied to identify and validate proteomic entities/proteins that might serve as useful biomarkers for radiation injury (Figure 1) [22]. Furthermore, many of these proteins are also being investigated in terms of their modulation by MCM treatment using unirradiated and irradiated animals. Such studies will be useful for converting MCM efficacious doses from animal models to humans for their regulatory approval following the US Food and Drug Administration (US FDA) Animal Rule.

Figure 1. An overview of applications and techniques used for proteomic analysis.

2.1. Conventional proteomics

There are several conventional techniques used in proteomic analysis including gel electrophoresis, mass spectrometry (MS), two-dimensional gel electrophoresis, western blotting, protein microarrays, enzyme-linked immunosorbent assay (ELISA), Luminex, and isotope labeling [23–27]. The latter techniques represent just a few examples of the conventional techniques currently used in proteomics, while many other methods are continually being developed and refined in order to improve our understanding of proteomes and their functions. These techniques are being used in the area of radiation biology for identifying and validating specific proteomes thought to be involved in the fundamental molecular biological processes of radiation injury development and expression.

2.1.1. Western blotting

Western blotting is a widely applied, exceedingly useful laboratory technique used to detect specific proteins within biological samples. It can also be used to detect changes in protein abundance and confirm protein–protein interactions. In brief, the technique involves the separation of proteins by gel electrophoresis; they are then transferred onto a membrane, and subsequently probed with a primary antibody specific to the target protein of interest [28].

2.1.2. Chromatography based techniques

Affinity, size exclusion, and ion exchange chromatography-based proteomic techniques are sets of methods used to separate, identify, and quantify proteins within given samples based on their physicochemical properties.

2.1.2.1. Affinity chromatography

This technique separates proteins based on their specific binding to a ligand immobilized on a chromatography resin [29]. The immobilized ligand can be a small molecule, protein, or nucleic acid that has a high affinity for the target molecule. There are various types of affinity chromatography; protein A chromatography is used to purify antibodies, immobilized metal ion affinity chromatography (IMAC) is used to purify proteins that have histidine tags, lectin chromatography utilizes specific lectins as the immobilized ligand, and DNA affinity chromatography utilizes single-stranded DNA as the immobilized ligand [30].

2.1.2.2. Size-exclusion chromatography

Size exclusion chromatography is a non-denaturing technique that separates proteins based on their size and shape without altering their structure or activity [31]. The stationary phase is typically composed of porous beads made from materials such as dextran, agarose, or polyacrylamide, with pore sizes ranging from 10 to 500 Å. The mobile phase is usually a buffer solution that is optimized for the separation of the target molecules [32].

2.1.2.3. Ion-exchange chromatography

This method separates proteins based on the charged groups on its surface (anionic and cationic) and later on the basis of physiological pH [33]. Ion-exchange chromatography can be performed in two modes: cation-exchange chromatography and anion-exchange chromatography. Cation-exchange chromatography separates negatively charged molecules by attracting them to positively charged stationary phase materials, while anion-exchange chromatography separates positively charged molecules by attracting them to negatively charged stationary phase materials [34].

2.1.3. ELISA

ELISA is a commonly used laboratory technique to detect and quantify specific proteins, antibodies, and other molecules in a sample [35].

2.1.4. Multiplex Luminex

Luminex uses color-coded polystyrene beads (5.6 µm in diameter) that are internally dyed with different proportions of red and infrared fluorophores corresponding to a distinct spectral signature, or bead region. The surface of the bead can be coated with a different capture antibody in order to identify the antigen of interest within the test samples. The beads with bound target proteins are separated by laser excitation and quantitated. Beads coupled with different antibodies, each with a distinct fluorescence signature, are mixed, thus enabling simultaneous detection of multiple target proteins in a single well of a 96-well plate. This high-throughput system allows for rapid multiplexed analyses of panels, and proteomic biomarkers associated with select disease states. This general testing platform, therefore, is of high practical value in a clinical setting [36].

2.1.5. Meso scale discovery (MSD) platform

MSD immunoassays are based on a technology combining electrochemiluminescence and multi-array for detection of multiple proteins in a single sample. This is a sandwich-based assay consisting of a Multi-Spot microplate, where each spot is coated with unique capture antibodies. This assay works on the same principle as ELISA but uses an electrochemiluminescence signal as the detection method, and it has the capability to measure multiple protein analytes in a single well. Unlike ELISA plate readers, MSD instruments provide increased throughput with fast read times. Furthermore, MSD has a wider dynamic range, while Luminex has higher sensitivity and precision [37].

2.1.6. Aptamer-based SOMAscan assay

The oligonucleotide aptamer-based SOMAscan platform of SomaLogic is used for quantification of >1,300 proteins or more [38]. Similar to antibody-based assays, these aptamers are nucleotide sequences to bind specific antigens and, in this way, this assay platform can detect very low abundance proteins within test samples which compares favorably to other immunoassay systems. However, the large SOMAscan platform is limited in terms of providing only relative quantitative information and not absolute concentrations of target molecules of interest.

2.2. Advanced proteomic technology

Advanced proteomic technologies are a set of techniques used to study the complex and dynamic nature of the proteome. These technologies include MS, protein microarrays, high-resolution liquid chromatography, structural proteomics, and top-down proteomics. These advanced proteomic techniques have revolutionized our understanding of the proteome and have enabled the discovery of new biomarkers, drug targets, and therapeutic interventions [22].

2.2.1. Gel-based proteomic techniques

Gel-based proteomic approaches are powerful techniques used to separate, identify, and quantify proteins in complex biological samples. Three commonly used gel-based proteomic approaches are SDS-PAGE (known as two-dimensional polyacrylamide gel electrophoresis), 2D-PAGE, and 2D-DIGE (two-dimensional in gel electrophoresis). They have been widely used in a range of fields, including basic research, clinical diagnosis, and drug discovery. SDS-PAGE is a one-dimensional gel electrophoresis used to separate proteins based on their molecular weight [39]. 2D-PAGE leverages the power of SDS-PAGE by separating proteins based on both their molecular weight and isoelectric point (pI) [40]. 2D-DIGE is a variation of 2D-PAGE that allows for the direct comparison of two protein samples on the same gel [41].

2.2.2. Mass spectrometry (MS)

MS is a powerful analytical technique that can be used to analyze the mass and composition of target molecules within test samples. There are several different types of MS including discovery, targeted, and isotopic labeling. Discovery MS involves the analysis of complex mixtures of molecules, such as proteins or metabolites and the use of high-resolution mass spectrometers, such as Orbitrap or Fourier-transform ion cyclotron resonance (FT-ICR) [42]. Discovery proteomics aims to identify and quantify as many proteins as possible in a given sample, without prior knowledge or selection of the proteins of interest. Targeted MS, on the other hand, is a more focused approach that is used to measure the abundance of a predefined set of proteins of interest using selected reaction monitoring (SRM) or parallel reaction monitoring (PRM) in a sample with the use of triple quadrupole mass spectrometers, which are highly sensitive and selective instruments that can be used to quantify small molecules [43]. The newly developed Olink Proteomics technique falls under the category of targeted proteomics. It is a multiplexed immunoassay-based approach designed to measure protein biomarkers in biological samples, and it uses proximity extension assay (PEA) technology, which combines DNA oligonucleotide pairs linked to antibodies. When the target proteins bind to their respective antibody pairs, the oligonucleotides are brought into proximity, enabling their hybridization and subsequent amplification for quantification. The technique allows for the simultaneous measurement of multiple protein biomarkers in a highly sensitive and specific manner [44]. Isotopic labeling aims to compare the relative abundance of proteins between different samples or conditions using stable isotopes to label the peptides or proteins, and is a technique that is used to track the movement of atoms or molecules through metabolic pathways. Isotopes are atoms that have the same number of protons but different numbers of neutrons and can be used to label molecules so that they can be tracked using MS. There are several types of isotopic labeling including stable isotopic labeling, in which stable isotopes are used to label molecules, and radioactive isotopic labeling, in which radioactive isotopes are used to label molecules [45]. Data-independent acquisition (DIA) is a relatively recent enhancement in the field of MS-based proteomics. This innovative approach enables thorough and unbiased data collection by generating fragment ion spectra for every precursor ion within a predefined mass range. It can be seamlessly integrated into targeted proteomics workflows, allowing for the simultaneous quantification of multiple proteins of interest. DIA is a versatile MS approach that encompasses different strategies for acquiring comprehensive data. SWATH is known for its comprehensive coverage [46], PRM is specialized for targeted quantification, and MSX [25,47,48] enhances throughput by simultaneously fragmenting multiple precursor ions. These approaches diverge in how they segment the mass range into smaller windows for fragmentation and in their methods for capturing the resulting fragment ion spectra. The choice of strategy depends on the specific research goals, whether they involve comprehensive proteome analysis, precise quantification of specific proteins, or high-throughput screening. The usage of DIA in combination with stable isotope labeling is quite rare and is typically applied in specialized scenarios like single-cell proteomics [49]. However, there are some examples of its use in proteomics research. For example, a study by Wang et al. [50] used DIA with stable isotope labeling to quantify protein expression in human plasma. Its remarkable versatility, which encompasses both identification and quantification capabilities, positions DIA as an invaluable asset in the toolkit of proteomic research [51].

2.2.3. Protein microarray

Protein microarrays have proven to be useful tools for high-throughput analysis of protein–protein interactions, protein–DNA interactions, and protein-small molecule interactions. There are different types of protein microarrays, including reverse phase, analytical, and functional microarrays. Reverse phase protein microarrays (RPPAs) allow for the quantitative analysis of protein expression levels in biological samples. This technique involves immobilizing samples of lysed cells or tissues onto a microarray and probing them with a panel of specific antibodies against proteins of interest. The antibodies are then detected using fluorescent or chemiluminescent signals, and the resulting data is used to quantitatively measure protein expression levels [52]. Analytical protein microarrays are used for the detection and quantification of specific proteins or protein modifications. These microarrays are typically made by printing specific antibodies or protein-binding molecules onto a solid support, such as a glass slide or a membrane. The protein of interest is then applied to the array, and any bound protein is detected using a variety of labeling or detection techniques, such as fluorescence or chemiluminescence [53]. Functional protein microarrays are used to study protein–protein interactions or protein-small molecule interactions. These microarrays typically contain a library of purified proteins, which are printed onto solid supports. The arrays are then probed with fluorescently labeled target proteins or small molecules, and any bindings are detected using fluorescence imaging. Functional protein microarrays can be used to identify novel protein interactions or to screen large libraries of compounds for their ability to bind to specific proteins [53].

2.2.4. Quantitative proteomics

Various methods have been developed for quantitative proteomics, enabling the comparison of protein expression levels between different samples. Label-free methods and labeling methods using MS1 and MS2-based quantification are two categories of MS-based quantification methods for proteomic analysis. These include label-based approaches like stable isotope labeling with amino acids in cell culture (SILAC) or isobaric tags for relative and absolute quantitation (iTRAQ/TMT), ICAT as well as label-free methods based on spectral counting, or intensity-based quantification. Label-free methods do not use any chemical or isotopic labels to mark the peptides or proteins in different samples, but rely on the intensity or count of the peptide signals in the MS spectrum to estimate their relative abundance [54]. Labeling methods use chemical or isotopic labels to mark the peptides or proteins in different samples, which can be distinguished by their mass difference in the MS spectrum and are used to quantify their relative abundance [55]. MS1-based quantification methods use the intensity of the peptide precursor ions in the survey scan (MS1) to measure their relative abundance. MS2-based quantification methods use the intensity or count of the peptide fragment ions in the tandem mass spectrometry scan (MS2) to measure their relative abundance.

Label-Free Methods: Label-free quantification methods in proteomics rely on the relative abundance of peptides or proteins in different samples without the use of chemical or isotopic labels. These methods compare the intensities or spectral counts of peptides between samples to determine relative protein abundance. Common label-free techniques include spectral counting and various intensity-based methods that rely on quantifying ion intensities, such as spectral index or precursor ion intensity, which essentially involves extracted ion chromatograms. These techniques leverage MS’s capability to measure ions for quantification purposes. Label-free quantification is advantageous for its simplicity, cost-effectiveness, and suitability for a wide range of experimental designs [54]. Labeling methods: iTRAQ/TMT, (isobaric tags for relative and absolute quantitation/tandem mass tag). TMT and iTRAQ are indeed isobaric labeling techniques. The reporters are isobaric and have the same mass, but they differ in their fragmentation patterns and can be used to quantify peptide abundance. They label peptides with tags that have the same nominal mass but different isotopic compositions. The reporters, however, have different masses and are used for quantification [56]. iTRAQ and TMT differ in some aspects, such as the number of tags available, the structure of the tags, and the fragmentation behavior of the tags [55]. The labeled peptides are then combined and analyzed by MS, allowing for the relative quantitation of multiple samples in a single experiment [57]. ICAT is a method of quantitative proteomics that involves labeling proteins with isotopically labeled cysteine-containing tags. The labeled proteins are then digested with a protease and analyzed by MS, allowing for the quantitation of proteins between two samples [58]. SILAC is a metabolic labeling method that involves incorporating isotopically labeled amino acids into proteins during cell culture [59].

2.3. High-throughput proteomics

High-throughput proteomic technologies refer to a range of techniques that enable the analysis of large numbers of proteins in a high-throughput manner. These technologies are critical for large-scale proteomic studies, such as those aimed at establishing basic, molecular processes of disease development, as well as for drug discovery or biomarker identification. Some of the key high-throughput proteomic technologies include MS, protein microarrays, tandem affinity purification, chemical cross-linking, and quantitative proteomics.

2.4. Bioinformatics analysis

Bioinformatics plays an essential role in the analysis and interpretation of proteomic data. For full and proper analyses, massive amounts of data generated by proteomic technologies, such as MS and protein microarrays, require sophisticated computational tools and algorithms to extract meaningful information. Some key bioinformatics analysis steps in proteomic technology include the following processes:

2.4.1. Multivariate analyses

Multivariate analyses in proteomics are statistical methods that can handle and interpret data with multiple variables such as protein abundance, expression, modification, interaction, etc. Multivariate analyses can reveal complex patterns and relationships among proteins and can also reduce the dimensionality and sparsity of proteomic data, improve the accuracy and specificity of protein identification and quantification, and facilitate the biological interpretation and functional annotation of proteomic data. Multivariate techniques such as Support Vector Machines (SVM), Partial Least Squares (PLS), or Elastic Net Regression are commonly employed for biomarker identification and validation [60].

2.4.2. Database searching

Proteomic data is typically analyzed by searching against a protein database to identify proteins and their post-translational modifications (PTMs). This step involves the use of some key software tools and databases such as Mascot [61], Uniport [62], Proteome Discoverer [63], PRIDE [64], Skyline [65], or MaxQuant [63] to match experimental spectra to theoretical spectra generated from protein sequences.

2.4.3. Protein quantification

Proteomic technologies such as iTRAQ, ICAT, and SILAC generate quantitative data, which require specialized software tools to analyze. These tools typically involve normalization, statistical analysis, and visualization of the data [63].

2.4.4. Post-translational (PTM) analysis

PTMs play important roles in protein function and regulation, and proteomic technologies can be used to identify and quantify PTMs. Bioinformatics analysis of PTMs typically involves the use of specialized algorithms and databases, such as PhosphoSitePlus or UniProt, to identify modified peptides and assign specific PTMs [62,63].

2.4.5. Functional analysis

Proteomic data can be used to infer protein function and identify potential biological pathways or networks. Functional analysis typically involves the use of specialized tools, such as gene ontology (GO) analysis or pathway analysis, to identify over-represented functional categories or enriched pathways. These include PANTHER [66], Gene Ontology (GO) analysis [67], and David database [68].

2.4.6. Integration with other data types

Proteomic data can be integrated with other types of omics data, such as transcriptomic or genomic data, to provide a more comprehensive view of biological systems. Bioinformatics tools, such as ingenuity pathway analysis (IPA) or Cytoscape, can be used to integrate and visualize different types of omics data.

3. Radiation-induced altered proteomic profiles

A sizable number of studies have been reported demonstrating radiation-induced alterations of proteomes in body fluids and organs within both small and large animals subjected to different modes, levels, and qualities of ionizing radiation [12,69]. Recent studies have used state of the art omic platforms such as DIGE, protein microarray, NMR-spectroscopy, MS, and X-ray crystallography to demonstrate the proteomic changes in serum, plasma, and various tissue samples collected from irradiated and unirradiated animals (Table 1).

Table 1. Proteomic studies for radiation biomarkers using various analysis platforms and different animal models.

Species/strain	Radiation Quality	Exposure type	Dose (Gy)	Dose Rate (Gy/min)	Proteomic Platform	Sample	Time points	Study details	References
CD2F1 mice	^60Co γ	TBI	9.2	0.6	Luminex	Serum	.5, 1, 2, 4, 8, 12, 24, 48 h	G-CSF and IL-6 induction, highest level at 12 h post-irradiation	[73]
CD2F1 mice	^60Co γ	TBI	9.2 & 11	0.6	Luminex	Serum	4, 12, 24, 36 h; 6, 12 d	IL-1β, IL-6, IL-10, IL-12(p70), G-CSF, GM-CSF, KC, and TNF-α stimulated	[83]
CD2F1 mice	^60Co γ	TBI	9.2	0.6	Luminex	Serum	.5, 1, 2, 4, 8, 12, 24, 48 h	IL-1β, IL-6, IL-10, IL-12 (p70), G-CSF, GM-CSF, KC and TNF-α stimulated	[84]
CD2F1, BALB/c mice	^60Co γ, X-ray	TBI, PBI	Various	Various	ELISA, MSD, Luminex	Total protein	Various	G-CSF, GADD45α, SAA, TPO, EPO, FLT-3 L	[91,92,95,98,99]
CD2F1 mice	^60Co γ	TBI	5, 7, 8, 9, 10	0.6	ELISA	Spleen protein	1, 3, 7, 9, 13 d	IL-18 increased	[96]
CD2F1 mice	^60Co γ	TBI	0.5, 1, 3, 5, 7, 8, 9 or 10	0.6	ELISA	Serum	1, 3, 4 d	IL-18 increased	[97]
CD2F1 mice	^60Co γ	TBI	5, 7, 8, 9, 10, 12	0.6	ELISA, cytokine array	Serum, BM, spleen, thymus	4, 8, 24 h; 1, 3, 6, 9 d	IL-18 increased	[101]
B6D2F1 mice	^60Co γ	TBI	3–12	0.6 or 1.9	MSD	Plasma	1, 2, 4, 7 d	FLT-3 L, IL‐5, IL‐10, IL‐12, IL‐18, G-CSF, GM-CSF, TPO, EPO, SAA	[85]
B6D2F1 mice	Neutrons and gamma rays, two ratios of N/G	TBI	1.5–6	0.6 or 1.9	MSD	Plasma	1, 2, 4, 7 d	FLT-3 L, IL‐5, IL‐10, IL‐12, IL‐18, G-CSF, GM-CSF, TPO, EPO, SAA	[85]
C57BL/6 mice	X-ray	TBI	8, 10, 12, 14	0.812	LC-MS/MS	Upper ileum	1, 3, 6 d	A number of trends in protein changes	[108]
C57BL/6 mice	X-ray	TBI	8, 10, 12, 14	1.25	LC-MS/MS	Lung	1, 3, 6 d	Several proteomic changes in response to irradiation	[109]
CD2F1 mice	^60Co γ	TBI	7	0.6	UPLC-MS/MS	Spleen	0, 1, 2, 4, 7, 10, 14 d	Upregulation of the Wnt signaling pathway and actin-cytoskeleton linked proteins	[110]
CD2F1 mice	^60Co γ	TBI	9.2, 11	0.6	2D-DIGE	Serum, jejunum	24 h	Changes in beta-2-glycoprotein 1, alpha-1-acid glycoprotein 1, alpha-2-macroglobulin, complement C3, mannose-binding protein C, and major urinary protein 6, and several pathways	[111,112]
CBA/CaJ	^137Cs γ	TBI	3 Gy	0.72	2D-DIGE	Plasma	2, 7 d	18 proteins up-regulated on d 2; 14 proteins upregulated on d 7	[113]
C57BL6, BALB/c, C3H/HeJ, CD2F1 and outbred CD-1 mice	X-ray	TBI	1–8	3.59	ELISA	Plasma	1, 2, 3, 7 d	Strain specific differences do exist between expression levels of FL, MMP9, SAA, PTX3 and FGB	[114]
Mice	^137Cs γ	TBI	10.4	2.36	LC-MS/MS	Urine, serum	24, 72 h	47 differentially secreted proteins identified in urinary and serum exosomes.	[118]
Human CD45+ lymphocytes from NOD SCID gamma mice	X-ray	TBI	1, 2	0.88	LC-MS/MS	Human CD45+ lymphocytes	3 d	FDXR, DDB2, ACTN1 as best candidate markers	[121]
C57BL6 mice	X-ray	TBI	2, 3.5, 8	2.28	SOMAscan	Plasma	1, 3, 7 d	Several proteomes modulated – control vs irradiated and also in relation to various doses	[125]
C57BL/6 mice	X-ray	TBI/PBI	2, 8 TBI, 2, 8 PBI lung and gut; 2, 8, 16 brain	2.28	SOMAscan	Plasma, gut, lung, brain	1, 3, 7, 14 d	Several proteomes differentially regulated in various tissues and with different doses	[126]
WAG/RijCmcr rats	SARRP X-ray	TBI, PBI (LKI)	10	1.73 TBI, 1.43 PBI	LC-MS/MS	Urine	Up to 24 h	Several proteomes differentially regulated by TBI and PBI	[123]
Sprague-Dawley rat	^60Co γ	TBI	1, 3, 5	1	LC-MS/MS	Plasma	1, 3, 5, 7 d	503 proteins differentially expressed	[122]
C57BL6 mice, Sinclair minipigs, NHP, human	^60Co γ, X-ray	TBI	Various	Various	SomaScan	Plasma	Various	Identified HIST1H1C as a novel biomarker, characterized panels of proteins and functional proteomic pathways altered by radiation exposure across murine, minipig, NHP and human	[124]
Gottingen minipigs	^60Co γ	TBI	1.6, 1.78	0.6	ELISA	Plasma	3 h; 1, 2, 3 d	IL-18 increased	[86,101]
NHPs	X-ray	WTLI	11.5	0.8	MSD	Plasma, lung	Plasma – various times; Lung − 180 d	IP-10, MCP-1, IL-12, IL-15, IL-16, IL-7, IL-6, MIP-1α, TARC, IL-17, TNF-β	[107]
NHPs	^60Co γ	TBI	6.5	0.624– 0.640	MSD	Plasma	10, 20, 30, 40, 50, 60 d	FLT-3 L, TPO, EPO, IL-12 (p40), CRP, SAA, G-CSF, GM-CSF	[72]
NHPs	^60Co γ	TBI	7	0.6	ELISA	Serum	2, 4 d	IL-18 increased	[86,101]
NHPs	^60Co γ	TBI	5, 6.5, 8.5	0.55	ELISA	Urine	6 h − 7 d	IL-18 elevated	[88]
NHPs	^60Co γ	TBI	5.8, 6, 6.6, 7, 7.2, 7.5, 8, 8.5	0.6	Luminex	Serum	Various	IL-1β, IL-6, IL-8, IL-10, G-CSF, GM-CSF, TNF-α	[102,103,106]
NHPs	^60Co γ	TBI	6.75	1.1 or 1.03	Luminex	Plasma	2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24 h	IL-6 and G-CSF stimulated significantly	[105]
NHPs	^60Co γ	TBI	6.7, 7.4	0.6	High-resolution MS	Urine, plasma	Until 12 d	Large number of proteomes modulated	[127,128]
NHPs	^60Co γ	TBI	7.2	0.6	NanoUPLC-MS/MS	Serum	−4 d; 24, 36, 48, 96 h	Radiation exposure induced various protein abundance	[20]
NHPs	LINAC X-ray	PBI, 2.5% BM sparing	12	0.8	NanoUPLC-MS/MS	Jejunum, kidney, lung, plasma	Various	Modulation of a large number of proteomes	[129–132]
NHPs	^60Co γ	TBI	6, 6.5, 7, 7.5, 8, 8.5	0.6	NanoUPLC-MS/MS	Serum	1, 2, 6 d	Serpin family A member 9 with other metabolites for all doses, other proteomes for specific doses used in this study	[19]
NHPs	LINAC X-ray	PBI, 2.5% BM sparing	12	0.8	NanoUPLC-MS/MS	Heart, lymph node	Over a period of 3 weeks	Large number of proteomes modulated	[135,136]

⁶⁰Co γ: cobalt-60 gamma rays; ¹³⁷Cs γ: cesium-137 gamma rays; BM: bone marrow; CRP: C-reactive protein; d: day; ELISA: enzyme-linked immunosorbent assay; EPO: erythropoietin; FLT-3 L: Flt-3 ligand; G-CSF: granulocyte colony-stimulating factor; GADD45α: growth arrest of targeted cells and temporally related DNA damage-α; GM-CSF: granulocyte macrophage colony-stimulating factor; Gy: gray; IL: interleukin; KC: keratinocyte-derived chemokine; LINAC: linear accelerator; LKI: local kidney irradiation; MCP: macrophage chemotactic protein; MIP: macrophage inflammatory protein; MSD: Mesoscale Discovery; NHPs: nonhuman primates; PBI: partial-body irradiation; SAA: serum amyloid A; SARRP: small animal radiation research platform; TARC: thymus- and activation-regulated chemokine; TBI: total-body irradiation; TNF: tumor necrosis factor; TPO: thrombopoietin; WTLI: whole thoracic lung irradiation.

Since there are a large number of studies, only limited studies have been discussed above. NHPs used in all experiments are rhesus.

Studies presented in Tables 2–5 list various MCMs, and these studies also have a separate group without MCMs as a control demonstrating the effects of irradiation. Irradiated animal samples have been compared with samples derived from unirradiated animals to investigate the effects of radiation exposure.

Table 2. Studies for cytokines/growth factors using MCMs in mice using various analysis platforms.

MCMs	Species & strain	Radiation Quality	Exposure type	Dose (Gy)	Dose Rate (Gy/min)	Cytokine Platform	Samples	Time points	Study details	References
Entolimod	NIH-Swiss mice	Unirradiated mice	N/A	N/A	N/A	LINCOplex, Luminex	Plasma	Before injection and .5, 1, 2, 4, 8 and 24 h after injection	Entolimod stimulated production of G-CSF, KC, IL-8, TNF, IP-10	[141]
Entolimod	ICR and C57BL/6J mice	^137Cs γ (also unirradiated mice)	TBI	9	1.4–2.4	Luminex, ELISA	Plasma	Pre-dose: .5, 1, 2, 4, 8, 24 h post-dose in unirradiated mice; .25, .5, 1, 2, 4, 8, and 24 h after drug injection in irradiated mice	Entolimod stimulated production of G-CSF, IL-6 in irradiated as well as unirradiated mice	[104,105]
CBLB613	CD2F1 mice	^60Co γ	TBI	3 and 7	0.6	Luminex	Serum	.25, .5, 1, 2, 4, 8, 12, 24 h post-drug; .5, 1, 2, 4, 8, 12, 24, 48 h post-irradiation	CBLB613 stimulated production of IL-6, KC, G-CSF, IL-1β, IL-10, IL-12, GM-CSF, and TNF-α in irradiated mice	[84]
CBLB612	BALB/cJ, C57BL/6J, B6.SJL-Ptprca Pepcb/BoyJ mice	Unirradiated mice	N/A	N/A	N/A	Luminex	Serum	0, 1, 2, 4, 8, 24, and 48 h post-drug injection	CBLB612 stimulated production of IL-6, KC, G-CSF, IL-1β, IL-10, IL-12, SCF, GM-CSF, and TNF-α in irradiated mice	[82]
BIO 300	CD2F1 mice	^60Co γ	TBI	7	0.6	Luminex, cytokine array	Serum,spleen, bone marrow	4 h after drug injection and 4 h and 24 h post-TBI	Genistein stimulated production of IL-6 and GM-CSF in irradiated and unirradiated mice	[151]
5-AED	C3H/HeN mice	^60Co γ	TBI	7.5	0.6	Luminex, QRT-PCR, ELISA, cytokine array	Serum, spleen, bone marrow	4 h after injection and 4 h and 24 h post-irradiation	Large number of cytokines, chemokines and growth factors (protein and message) were stimulated by 5-AED in serum and tissues	[143,144]
GT3	CD2F1 mice	^60Co γ	TBI	7, 9.2, 11	0.6	Luminex, cytokine array	Serum	6 h post-drug to 2 d post-TBI	GT3 treatment significantly increased G-CSF, IL-1α, IL-1β, IL-6, IL-12p70, IL-17, MIP-1α and KC levels	[103,137,156–158]
TS	CD2F1 mice	^60Co γ	TBI	3, 7, 9.2	0.6	Luminex, QRT-PCR	Serum, bone marrow	2, 4, 8, 12, 24, 36, 48 h after TS admin for Luminex; 12, 24 h after TS injection for QRT-PCR	TS modulated the expression of antioxidant enzymes and several cytokines and growth factors. G-CSF stimulation was notable	[139,140,165–168]
DT3	CD2F1 mice	^60Co γ	TBI	9.2	0.6	Luminex	Serum	24 h post-injection, 6 h post-TBI	DT3 stimulated several cytokines in irradiated or unirradiated mice, and G-CSF induction was notable	[36]
BBT-059	CD2F1 mice	^60Co γ	TBI	7	0.6	ELISA	Serum	0, 2 h, 1, 2, 3, 7, 15, 30 d	BBT-059 inhibited irradiation-induced TPO, EPO and FLT-3 L	[160]
JNJ-26366821	CD2F1	^60Co γ	TBI	8.0	0.6	ELISA	Serum	30 d	JNJ-26366821 reduced irradiation-induced FLT-3 L in mice at 30 d post-irradiation	[162]
PLX-R18 Cells	C57BL/6 mice	^60Co γ	TBI	8.0	0.6	Luminex, ELISA	Serum	3, 9, 14, 21, 30 d	PLX-R18 attenuated EPO, FLT-3 L, SAA and irradiation-induced inflammatory cytokines/chemokines/growth factors: G-CSF, MIP-1α, MIP-1β, IL-2, IL-6 and MCP-1	[164]
Captopril	C57BL/6 mice	^60Co γ	TBI	7.5	0.6	ELISA, QRT-PCR	Plasma	2, 7, 10 d	Captopril showed different effects on EPO and TPO production in irradiated and unirradiated mice depending on treatment schedule and time course	[169]

5-AED: 5-androstenediol; d: day; ⁶⁰Co γ: cobalt-60 gamma rays; ¹³⁷Cs γ: cesium-137 gamma rays; BBT-059: PEGylated interleukin-11; DT3: delta-tocotrienol; ELISA: enzyme-linked immunosorbent assay; EPO: erythropoietin; FLT-3 L: Flt-3 ligand; G-CSF: granulocyte colony-stimulating factor; GM-CSF: granulocyte macrophage colony-stimulating factor; GT3: gamma-tocotrienol; Gy: gray; h: hour; IL: interleukin; IP-10: interferon-inducible protein 10; KC: keratinocyte-derived chemokine; MCM: medical countermeasure; MCP-1: monocyte chemoattractant protein-1; MIP: macrophage inflammatory protein; NHPs: nonhuman primates; PLX-R18: 3D-expanded placenta-derived stromal cell product; QRT-PCR: quantitative reverse transcription polymerase chain reaction; SAA: serum amyloid A; SCF: stem cell factor; TBI: total-body irradiation; TNF: tumor necrosis factor; TPO: thrombopoietin; TS: tocopherol succinate.

Table 3. Studies for cytokines/growth factors using MCMs in minipigs, canines, and NHPs using various analysis platforms.

MCMs	Species & strain	Radiation Quality	Exposure type	Dose (Gy)	Dose Rate (Gy/min)	Cytokine Platform	Samples	Time points	Study details	References
Captopril	Göttingen minipig	^60Co γ	TBI	1.79	0.485–0.502	ELISA	Serum, bone marrow	−5 to 35 d	Captopril reduced radiation-induced EPO and GM-CSF, and suppressed radiation-induced SAA	[146]
Entolimod	Beagle canines	Unirradiated	N/A	N/A	N/A	MSD Electrochemiluminescence MULTI-SPOT 964 Spot	Plasma	Pre-dose; .5, 1, 2, 4, 8, 24, 48, 120 h post-dose	Entolimod stimulated production of IL-6 in unirradiated canines	[104]
Entolimod	Rhesus NHP	^60Co γ (also unirradiated NHPs)	TBI	5.0, 5.3 (midline)	0.6 Gy/min, ~0.8–1.0	Luminex, ELISA	Plasma	Pre-dose; 1, 2, 4, 8, 24, 48 h after treatment in unirradiated and irradiated NHPs	Entolimod stimulated production of G-CSF and IL-6 in irradiated as well as unirradiated NHPs	[104]
Entolimod	Rhesus NHP	^60Co γ	TBI	6.75 (6.2 midline)	~0.8–1.1	Luminex, Fluorokine MAP, Beadlyte Multi-Cytokine Flex assays	Plasma	0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24 h	Entolimod stimulated production of G-CSF and IL-6 in irradiated NHPs	[105]
GT3	Rhesus NHP	^60Co γ	TBI	5.8, 6.5, 7.2	0.6	Luminex	Serum	Pre drug; pre-rad; .25, .5, 1, 2, 4, 8, 12 h, 1, 2, 3, 4, 5, 6, 7, 10, 14, 20 d after drug; 4, 8, 12 h, 1, 2, 3, 6, 8, 10, 12, 21, 28, 34, 42, 60 d after irradiation	GT3 stimulated production of IL-6, IL-8, IL-10, G-CSF, GM-CSF, IL-1β, and TNFα in irradiated as well as unirradiated NHPs	[103]
AEOL 10,150	Rhesus NHP	LINAC X-ray	WTLI	11.5	0.8	Meso Scale Discovery	Plasma, lung tissue	d 0, week 2, 4, 8	Plasma IP-10, MCP-1, IL-12, IL-15, IL-16, IL-7 and IL-6 were changed due to WTLI. Plasma IP-10 and MDC were changed due to drug. Lung tissue showed changes in several cytokines	[107]

⁶⁰Co γ: cobalt-60 gamma rays; d: day; ELISA: enzyme-linked immunosorbent assay; EPO: erythropoietin; G-CSF: granulocyte colony-stimulating factor; GM-CSF: granulocyte macrophage colony-stimulating factor; GT3: gamma-tocotrienol; Gy: gray; h: hour; IL-1: interleukin; IP-10: interferon-inducible protein 10; LINAC: linear accelerator; MAP: multianalyte profiling; MCM: medical countermeasure; MCP-1: monocyte chemoattractant protein-1; MDC: macrophage-derived chemokine; MSD: Meso scale discovery; NHP: nonhuman primates; SAA: serum amyloid A; TBI: total-body irradiation; TNF-α: tumor necrosis factor – alpha; WTLI: whole thorax lung irradiation.

Table 4. Proteomic studies in murine models using radiation MCMs for ARS and various platforms for analysis.

MCM	Mouse Strain	Radiation Quality	Exposure type	Dose (Gy)	Dose Rate (Gy/min)	Proteomic Platform	Sample	Time points	Study details	References
GT3	CD2F1	^60Co γ	TBI	11	0.6	2D-DIGE/MS	Serum	24 h	A change in expression of 18 proteins was observed in response to TBI, and these changes were reversed by GT3 prophylaxis	[112]
GT3	CD2F1	^60Co γ	TBI	11	0.6	2D-DIGE/MS	Jejunum	24 h	A total of 13 protein expression changes by TBI were found to be corrected by GT3 prophylaxis	[111]
GT3	CD2F1	^60Co γ	TBI	11	0.6	UPLC-MS/MS	Spleen	0, 1, 2, 4, 7, 10, 14 d	GT3 attenuated irradiation-induced perturbations in Wnt signaling pathway	[110]
TS	CD2F1	^60Co γ	TBI	7	0.6	2D-DIGE/MS	Jejunum	24 h	Similar to GT3, 13 protein expression changes by TBI were reversed by GT3 prophylaxis	[111]
Melatonin	C57BL/6	^137Cs γ	PBI (abdominal)	13	0.88	TMT-2D-LC-MS/MS	Small intestines	3 d	In addition to other signaling pathways, p53 and DNA repair pathways were enriched by melatonin treatment in mice receiving abdominal irradiation	[175]
G-003 M	C57BL/6	^60Co γ	TBI	9.0	1.08	2DE-MS	Jejunum	1, 24, 72, 120, 216, 360, 840 h	G-003 M prophylaxis decreased total number of differential proteins, mediated protection to cytoskeleton, and modulated stress, apoptosis and inflammatory proteins in irradiated mice	[174]

2D-DIGE: two-dimensional differential in-gel electrophoresis; 2D-LC: two-dimensional liquid chromatography; 2DE: two-dimensional gel electrophoresis; d: day; ⁶⁰Co γ: cobalt-60 gamma rays; ¹³⁷Cs γ: cesium-137 gamma rays; ARS: acute radiation syndrome; DNA: deoxyribonucleic acid; GT3: gamma-tocotrienol; Gy: gray; MCM: medical countermeasure; h: hour; MS: mass spectrometry; p53: tumor protein P53; PBI: partial-body irradiation; TBI: total-body irradiation; TMT: tandem mass tag; TS: tocopherol succinate; and UPLC: ultra-performance liquid chromatography.

Table 5. Proteomic studies in an NHP model using radiation MCMs for ARS and the NanoUPLC-MS/MS platform.

MCM	Radiation Quality	Exposure type	Dose (Gy)	Dose Rate (Gy/min)	Proteomic Platform	Samples	Time points	Study details	References
BIO 300	NHPs were not irradiated	N/A	N/A	N/A	NanoUPLC-MS/MS	Serum	.25, .5, 1, 2, 4, 8, 16 h	Minor alterations in protein levels observed 2 h after BIO 300 administration (Cmax time) either by im or po route	[21]
Ex-Rad	^60Co gamma	TBI	7.2	0.6	NanoUPLC-MS/MS	Serum	−4 d, 24, 36, 48, 96 h	Ex-Rad mitigated radiation exposure-induced perturbations in protein abundance and restored protein homeostasis	[20]

⁶⁰Co γ: cobalt-60 gamma rays; ARS: acute radiation syndrome; C_max: maximum drug concentration in serum; Gy: gray; h: hour; im: intramuscular; MCM: medical countermeasure; MS: mass spectrometry; NanoUPLC: nanoflow ultra-performance liquid chromatography; NHP: nonhuman primates; po: oral administration; TBI: total-body irradiation.

3.1. Induction of growth factors and other cytokines by radiation exposure

Studies from several institutions have reported induction of a large number of cytokines and growth factors in response to radiation exposure in different animal models with or without countermeasure use [70,71]. Though the majority of studies have used western blot, chromatography, ELISA, and multiplex Luminex for measuring radiation-induced responses of cytokines and growth factors, there are also reports using the MSD MULTI-ARRAY 96-well plate HT platform [70,72].

3.1.1. Animal models of radiation injury and related proteome responses

Small animal murine model(s): In an earlier study, it has been shown that exposure to ionizing radiation induced significant levels of G-CSF and interleukin-6 (IL-6) in mice receiving 9.2 Gy ⁶⁰Co gamma-radiation. Maximal levels of G-CSF were observed in peripheral blood of mice 8 h post-irradiation, and IL-6 levels were highest at 12 h post-irradiation [73]. Further, it was shown that the administration of specific antibodies to G-CSF significantly enhanced mortality in acutely irradiated mice. These reported findings clearly suggest that acute radiation exposure induces G-CSF and IL-6, and that neutralizing antibodies to G-CSF aggravates the injury caused by ionizing radiation to select vital organ systems of the body such as the hematopoietic and GI organ systems. This study also suggests that the irradiation-induced G-CSF may play an important role in injury recovery as well. In this context, it is important to note that recombinant products of G-CSF (Neupogen and Neulasta) have already been approved by the US FDA for mitigating acute, potentially lethal radiation injuries [74–77]. Two additional related agents, granulocyte-macrophage colony stimulating factor (GM-CSF, Leukine) and Nplate (thrombopoietin (TPO) analog), have also been approved by the US FDA for comparable radiation injuries [78–81]. Furthermore, studies have demonstrated induction of other cytokine- or inflammatory-like agents (e.g. C-reactive proteins) in various murine models of radiation exposure injuries under a variety of experimental conditions [12,82–95]. For example, the immunological effector molecule, IL-18, has been shown to be induced (modulated) in rodents in response to total body-gamma ray exposures as evidenced by elevated levels of IL-18 in both blood and urine samples [88,96,97].

While the majority of these studies have involved total-body exposures using gamma rays (⁶⁰Co gamma rays), the effects of partial-body exposures have employed mainly small animal research platforms (SARRP) or linear accelerators (LINAC) for the experimental irradiations. One research group, for example, has conducted several studies using CD2F1 and BALB/c male mice using both types of irradiation platforms (i.e. ⁶⁰Co gamma and SARRP). These studies have reported not only on the modulation of G-CSF following irradiation, but also on the growth arrest of targeted cells and temporally related DNA damage-α (GADD45α), serum amyloid A (SAA), TPO, erythropoietin (EPO), and FMS-related tyrosine kinase 3 ligand (FLT-3 L) responses [91,92,95,98,99]. Interestingly, under equivalent radiation exposures/doses, ⁶⁰Co gamma (megavoltage) appears less injurious than SARRP X-ray (orthovoltage) [100].

The latter observations appear to suggest the influence of the ‘quality’ of ionizing radiation being employed in these studies. To this point, there are a few studies that have used mixed-field radiation (neutrons and gamma fields generated by a TRIGA reactor). One mixed-field study was performed over a broad dose range (1.5–6 Gy), with dose rates of either 0.6 or 1.9 Gy/min, and two different proportions of neutrons and gammas (either 67% neutrons plus 33% gammas or 30% neutrons plus 70% gammas) using B6D2F1 mice. Blood was collected 1, 2, 4, and 7 d after TBI [86]. Irradiation-induced proteins were measured using the MSD’s high-throughput multi-array platform and ELISA. Results demonstrated dose- and time-dependent changes in FLT-3 L, IL-5, IL-10, IL-12, IL-18, G-CSF, GM-CSF, TPO, EPO, SAA, lipopolysaccharide binding protein (LPSBP), surface plasma neutrophil (CD45) and lymphocyte (CD27) markers, ratio of CD45 to CD27, and procalcitonin (PCT). It is important to note that regarding the above comments concerning ‘radiation quality’ effects, higher levels of IL-18, G-CSF and GM-CSF, SAA, and PCT were observed in animals irradiated with 67% neutrons + 33% gamma compared to those irradiated with 30% neutrons + 70% gamma. Comparisons were also made with another study conducted with TBI using gamma-rays (⁶⁰Co gamma ray source) under similar conditions [85]. Results from a ⁶⁰Co gamma study demonstrated significant dose- and time-dependent changes with all cytokines listed above. No differences were observed in cytokines between different dose-rate groups. Furthermore, no differences were observed between two sexes. Again, as suggested earlier, radiation-quality comparisons demonstrated that equivalent doses of gamma rays and mixed-field radiation do not produce equivalent biological effects. H-ARS occurs at lower doses of mixed-field radiation, accompanied by increased mortality. Cytokines in the ⁶⁰Cogamma study with 3 Gy were equivalent to the mixed field study with 1.5 Gy.

Large animal/minipig models: These porcine models have also been used sparingly for select radiobiological studies, including a cytokine study of the induction pattern of IL-18 following acute irradiation [101]. Gottingen minipigs (males) were subjected to 1.6 or 1.78 Gy of ⁶⁰Co gamma rays. Prior to or following the exposures, blood samples were collected periodically (hours to days), and serum samples were prepared and analyzed by ELISA. Modulatory patterns, specifically elevated blood levels of IL-18, were noted at various time points relative to the radiation exposure [101].

Large animal/rhesus nonhuman primate models: Nonhuman primates (NHPs) have been extensively used of late to explore the types and patterns of injury-induction, along with the associated proteomic responses under a range of doses of radiation exposure conditions. The use of these NHP models has provided clear demonstration of specific induction patterns of select cytokines of interest (e.g. G-CSF, IL-6, IL-8, IL-18, etc.) that have been implicated in acute injury and/or repair processes within the hematopoietic and GI systems over specific post-exposure periods [102–106]. These cytokine assays were conducted using mainly multiplex Luminex testing platforms. Of note, the kinetic induction patterns of these cytokines appeared quite distinct. In yet another study of NHPs acutely exposed to a range of gamma rays (1.0–8.5 Gy TBI at 0.55 Gy/min), a number of plasma proteins of interest (e.g. CRP, SAA, IL-6, and FLT-3 L) were analyzed over a short-time course following exposure (i.e. at 0.24, 1, 2, 3, 4, and 7 days post-irradiation) and were found to be significantly altered. This observation gave rise to the working hypotheses that such dynamic, quantitative changes of these blood plasma proteins may function as prognostic indicators of the time course and severity of ARS [90]. Further, the combination of response patterns of multiple proteins may provide still greater accuracy for early, acute radiation injury assessment than would any single protein response pattern.

Results of subsequent studies utilizing NHPs tend to support our ‘working hypothesis’ (i.e. induction patterns of select plasma proteins are sufficiently distinct as to provide a prognostic indicator of acute radiation injury). Specifically, these follow-up investigations showed that cytokine induction patterns generally followed similar trends over a range of radiation doses, apart from a few instances (e.g. deviant IL-7 and MCP-1 response patterns), where certain radiation dose groups deviated from the other radiation dose groups [102].

In another supporting study that compared the induction of a large number of cytokine- or cytokine-like plasma proteins (FLT-3 L, CD14, TPO, EPO, IL-12 p40, AMY2A, CRP, SAA, G-CSF, GM-CSF, PCT, LBP, CD177, CD20, and CD27) in NHPs under both full and minimal supportive care, a consistent, temporal modulation of many of these blood plasma proteins was observed [72].

Ancillary cytokine studies have utilized NHPs and have focused on primary, irradiation-induced pulmonary injuries and subsequent repair responses. In one study, the levels of some 24 cytokines were measured in serially collected plasma samples (as well as the terminal sampling of lung tissues) of NHPs subjected to intense (11.5 Gy) doses of PBI (thoracic). Of the cytokines assayed, those that were shown to be significantly altered as a result of whole-thoracic-lung irradiation (WTLI) include seven plasma cytokines (interferon gamma-induced protein 10 (IP-10), MCP-1, IL-12, IL-15, IL-16, IL-7 and IL-6) along with seven of the assayed cytokines (i.e. IP-10, MCP-1, MIP-1α, TARC (thymus- and activation-regulated chemokine), IL-17, TNF-β and IL-6) in lung tissue. Of note was the finding that the terminal plasma concentrations of IP-10, TARC, IL-12, IL-15 and IL-6 were significantly correlated with their levels in the lung tissue [107].

3.2. Proteomic analysis using other high-throughput platforms

During the last 10 years, there have been numerous reports using different animal models, particularly murine and NHPs, and various proteomic platforms (e.g. LC-MS and 2D-DIGE) to study the radiation exposure-induced proteomic changes in serum/plasma or other tissue samples (Figures 2 and 3). In general, all of these studies provide foundational support for both the identification and validation of biomarkers for radiation injury-related studies.

Figure 2. Schematic of two-dimensional differential in-gel electrophoresis and animal study, sample preparation, and subsequent data analysis.

Using a C57BL/6J mouse model of radiation injury, proteomic analyses were carried out using LC-MS/MS of GI (upper ileum) tissues collected from animals during early periods (1, 3, 6 d) following acute TBI (8, 10, 12, 14 Gy) [108]. Time- and dose-dependent protein changes were revealed with select upstream, regulatory proteins of key canonical pathways being altered by the acute exposures [108]. Using the same strain of mice subjected to WTLI, and with proteomic analyses focused on the early, post-exposure (1, 3, 6 d) responses of pulmonary tissues, several response trends were observed in the proteome alterations, including protein changes greater than 10-fold, proteins consistently up- or down-regulated at all time points and radiation dose, and time and dose-dependent proteomic changes [109].

Still further advanced proteomic studies using small rodent mouse models of radiation injury have been reported, including a report demonstrating marked alterations of proteomic profiles (with splenic tissues) seen under steady state to those profiles observed during early post-exposure. For example, the upregulation of the Wnt signaling pathway and actin-cytoskeleton linked proteins in mediating radiation injury responses in spleens of CD2F1 mice subjected to TBI was revealed by MS, reverse transcription-polymerase chain reaction (RT-PCR), and western blot analyses of sampled tissues [110]. Comparable studies have been conducted and have reported major shifts of proteomic profiles of select organs/tissues of acutely irradiated animals. For example, 2D-DIGE followed by MS and IPA demonstrated a change in the expression of proteins in jejunum and serum samples [111,112]. IPA revealed a network of associated proteins including β2-glycoprotein 1, α1-acid glycoprotein 1, α2-macroglobulin, complement C3, mannose-binding protein C, and major urinary protein 6 that are involved in cellular movement, immune cell trafficking, and inflammatory response. These could be translated as biomarkers of radiation injury. A study with CBA/CaJ mice and¹³⁷Cs TBI at a dose of 0–3 Gy using plasma samples and 2D-DIGE/MS at 2 and 7 d post-irradiation was conducted [113]. Although different proteins (mostly involved in inflammatory responses) were detected in exposed mice, alterations in expression-levels of clusterin, gelsolin, kininogen, and alpha-2-HS-glycoproteins were found at both times. Still another study compared the expression of five proteomic biomarkers (FLT-3 L, MMP9 (matrix metalloproteinase 9), SAA, pentraxin 3 (PTX3), and fibrinogen (FGB)) across females of five murine strains (C57BL6, BALB/c, C3H/HeJ, CD2F1, and CD-1) to test a multivariate dose prediction model based on a single C57BL/6 strain [114]. Significant differences were found between irradiation-induced expression of FLT-3 L, MMP9, SAA, PTX3, and FGB among the five strains. This study suggests that strain-specific differences exist for biomarker expression and the use of multiple biomarkers for dose prediction strengthens the predictive accuracy of a model. There are additional studies using murine models that serve to further strengthen the predictive features of our working model [115–120].

A study to identify candidate radiation-responsive protein biomarkers in human lymphocytes in vivo using humanized NOD SCID gamma (Hu-NSG) mouse model has been reported. Human CD45⁺ cells were collected from the Hu-NSG mouse spleen 3 d after X-ray exposure (0–2 Gy, 88 cGy/min) and changes were analyzed by MS. A total of 46 proteins were differentially expressed in response to X-ray exposure [121]. The combination of FDXR, ACTN1, and DDB2 showed the lowest mean absolute errors (≤0.13 Gy) and highest coefficients of determination (R² = 0.96).

A Sprague Dawley rat model of radiation injuries that arise relatively early (1–7 d) following acute, but relatively low ionizing radiation exposures (⁶⁰Co gamma rays, 1–5 Gy TBI) has demonstrated significant shifts using advanced analytics (i.e. tandem mass tag labeling, in combination with LC-MS/MS) in the blood concentrations of a select number of plasma proteins, namely alpha-2-macroglobulin, chromogranin-A, and glutathione peroxidase [122]. In another related report using comparable analytics [123], significant shifts of urinary proteins were noted following either TBI or partial-body (LKI - localized kidney irradiation) exposures of large rodents (WAG/RIJICmer rats). Notably, urinary protein profiles changed markedly, but selectively following the two exposure protocols (TBI and LKI). Following TBI, fetuin-B, beta-glucuronidase, tissue kallikrein, chondroitin sulfate proteoglycan NG2, and vitamin D-dependent calcium binding protein were detected, while following LKI exposures, RNA binding protein 19, major urinary protein-1, WD repeat and FYVE domain containing protein 3, sorting nexin-8, neuron navigator, Dapper homolog 3, ankycorbin, and aquaporin. Urine protein (Up) and creatinine (Uc) ratios (Up/Uc) and urinary albumin abundance decreased in both TBI and LKI groups. As such, these reported observations suggest that the urine proteome may be useful in determining organ-specific (kidney) injury.

Minipigs have been added recently to the array of mammalian species currently being used to evaluate the role played by the proteome in manifesting acute radiation injuries using advanced analytics [124–126]. A recently applied, novel aptamer-based proteomic assay has been used to investigate changes to the blood proteome after TBI across four different mammalian species including mice, minipigs, NHPs, and humans. Plasma samples, both unirradiated and irradiated either in situ or in vitro, were collected and assayed using the aptamer-based SomaLogic SomaScan assay to evaluate changes in the expression of 1,310 protein analytes. Detectable values were obtained for all 1,310 proteins, and the HIST1H1C protein was shown to be radiation responsive in humans, NHPs, and murine species.

There are several studies using rhesus macaque NHPs and TBI or PBI models for studying proteomic changes in response to irradiation. Urine and plasma samples from a study with NHPs exposed to high, lethal doses of gamma irradiation (6.7 and 7.4 Gy doses delivered at relatively high exposure dose rates, 0.6 Gy/min) have been analyzed using high-resolution MS analysis [127,128]. A total of 2,346 and 663 proteins were identified in urine and plasma, respectively, and out of these proteins, 414 were found to be in common and significant. A total of 33 proteins from plasma and 46 from urine were found to be significantly altered in abundance. Furthermore, 10 proteins were found to significantly differentiate between both dose and time points in plasma and urine from the same animals [127,128].

In yet another study employing an advanced analytic procedure of NanoUPLC-MS/MS global molecular profiling of radiation-induced proteomic changes in blood samples collected from acutely irradiated NHPs (7.2 Gy), it was found that a total of 63, 69, 48, and 53 proteins were significantly dysregulated at 24, 36, 48, and 96 h post-irradiation, respectively [20]. This study also revealed alterations of a host of type-specific proteins (e.g. proteins related to hemostasis, innate immunity, protein metabolism, etc.), as well as signs of generalized dysregulation of basic cell/tissue networks.

Advanced proteomic analytics have been applied to studies investigating the pathophysiology of radiation injury within NHPs following PBI. As an example, one such study involved high dose PBI (12 Gy, 6 MV LINAC-derived photons at 0.8 Gy min) but with 2.5% marrow sparing. Proteome analyses (using LC-MS/MS) of various blood and tissue samples collected over a 3-week period were conducted using LC-MS/MS [129–132]. A number of trends were observed, including pronounced qualitative/quantitative changes as well as changes that were consistently up- or down-regulated at all time points and dose levels investigated.

In jejunum, 3,245 unique proteins were quantified out of more 3,700 identified proteins. A total of 289 quantified proteins demonstrated consistent and significant responses across at least three time points, with 263 up- and 26 down-regulated [132]. In the kidney, 3,432 unique proteins were identified. A total of 265 proteins demonstrated significant responses across at least three time points, of which 230 were up- and 35 were down-regulated [130]. In the lung, 3,101 unique proteins were identified and a total of 252 showed consistent response, of which 215 were up- and 37 were down-regulated [131]. Investigation with plasma samples suggested that CRP, SAA1, and CASS4 (Cas scaffolding protein family member 4) are the most promising proteins as biomarkers for radiation injury [129]. In brief, significant pathway and biological process alterations were observed after PBI.

3.3. Multi-omics approach using NHP model

While several studies have reported putative metabolite, protein, or miRNA-based biomarkers for predicting radiation exposure, few have used a multi-omics approach in large animal models for reproducibility and possible clinical translation [12,19,133–135]. Such multi-omics studies have used TBI as well as PBI models.

NHPs with different doses of TBI (⁶⁰Co gamma ray) have been studied to delineate a multi-omics (proteomics and metabolomics) serum probability index. For example, in a recent report, blood plasma (serum) samples were collected from both male and female NHPs prior to acute irradiation during the initial period following exposure (1, 2, and 6 d post-irradiation) and were analyzed for select metabolites, lipids, and proteins of interest using MS analytics [102]. A logistic regression model was used to develop a four-analyte panel that serves to identify unirradiated and irradiated animals within a given population with a high degree of predictability [19]. Remarkably, the panel appeared to be highly efficacious in identifying radiation-exposed animals across all six radiation exposure levels (6 to 8.5 Gy) during early periods of testing (e.g. 24 h to 6 d post-exposure). The panel of putative biomarkers consisted of Acetyl-carnitine, PC (16:0/22:6), Serpin Family A9, and Suberylglycine [19]. The panel’s protein biomarkers were validated with western blots of serum samples. Additional studies using other multi-analyte panels for specific dose assessments have also been reported, providing additional evidence of the utility of the ‘multi-omics’ approach. In sum, this study demonstrated the potential usefulness of a combinatorial molecular characterization. In particular, the use of multi-analyte panels of potential biomarkers for the development of minimally invasive assays for radiation dose assessments.

In addition to the above-mentioned studies that have focused on proteomic changes with TBI, there are additional, but related reports that have examined the effects of PBI. These reports certainly include but are not limited to attempts to better characterize the pathogenesis of select types of radiation-induced injuries using multi-omics approaches (proteomics and metabolomics) [135,136]. For example, the proteome and metabolome of left and right heart ventricles were quantitatively profiled in NHPs following 12 Gy PBI and with 2.5% bone marrow sparing over a time period of 3 weeks [135]. Global proteome profiling identified >2,200 unique proteins, with 220 and 286 in the left and right ventricles, respectively. Targeted metabolomics analyzed a total of 229 metabolites, showing significant responses compared to baseline levels. Furthermore, pathway analyses using metabolomic and proteomic data demonstrated dysregulation of energy metabolism, inflammation, and myocardial functions, along with dysregulation of the retinoid homeostasis pathway. Related PBI-induced deregulatory effects (proteome and metabolome) within mesenteric lymph nodes of NHPs were also noted using multi-omic analytics [136].

4. Proteome biomarkers and MCMs

There have been extended efforts to make use of cytokine-induction patterns following acute irradiation as potential biomarkers of radiation injuries, as well as bioindicators of the effectiveness of given agents designed to thwart such injuries. Several recent studies have utilized advanced, high-throughput analytics for these proteomic/MCM-based studies. This is an important area of MCM research, as the US FDA requires efficacy-based biomarkers for any given MCM to be approved under the Animal Rule. Further, such biomarkers are considered critical for drug dose conversions from animals to humans. Studies presented in Tables 2–5 list various MCMs, and these studies have separate groups without an MCM as a control demonstrating the effects of irradiation. Irradiated animal samples have been compared with samples derived from unirradiated animals to investigate the effects of radiation exposure.

4.1. Induction of G-CSF and other cytokines by MCMs and their role in radioprotection

Related to this research effort, a large number of MCMs have been shown to influence induction of cytokines and growth factors in unirradiated and irradiated murine and NHP models. Few studies have been conducted in minipig and canine models as well. For the sake of brevity, we highlight below the major findings of various reports that have used principally either small rodents, canines or NHP models of acute radiation injuries and related MCM radioprotective or mitigative actions.

Using a murine model, a number of potential useful MCMs have been extensively assessed for their radioprotective/mitigative/therapeutic actions and have shown that these radiation injury countering effects are mediated, at least in part, by their capacity to induce high blood and tissues levels of protective G-CSF. By blocking G-CSF induction (via infusion of specific antibody to G-CSF), the cytokine’s reparative effects of acute radiation injury can be effectively ablated. These MCMs are gamma-tocotrienol (GT3) [137,138], delta-tocotrienol (DT3) [36], tocopherol succinate (TS) [139,140], entolimod (CBLB502) [104,105,141], and 5-androstenediol (5-AED) [142–144]. Again, all these agents have been shown to have the capacity to induce (during early post-exposure periods) high levels of G-CSF in a dose- and time-dependent manner.

A large number of MCMs have been used for studying cytokines/growth factors and related molecules using irradiated and/or unirradiated murine [145] (Table 2), minipig [146], and canine [104] models (Table 3). Most notably these MCMs include agents such as entolimod (a polypeptide derived from the Salmonella filament protein flagellin and ligand for toll-like receptor 5 (TLR-5, CBLB502)) [104,105,141,147,148], CBLB613 (naturally occurring mycoplasma-derived lipopeptide ligand for TLR-2/6) [84,149], CBLB612 (another synthetic lipopeptide that mimicking the structure of naturally occurring mycoplasma bacterial lipopeptide) [82], BIO 300 (genistein) [150,151], 5-androstenediol (5-AED) [143,144], gamma-tocotrienol (GT3) [103,137,138,152–158], PEGylated IL-11 (BBT-059, a long acting PEGylated IL-11 analog created using site-specific PEGylation technology and is modified with a single branched 40 kDa-PEG at a cysteine residue (PEG-*179C) added to the C-terminus of the protein) [159,160], thrombopoietin mimetic (JNJ-26366821, a TPO mimetic with no sequence homology to endogenous TPO) [161,162], PLX-R18 Cells (3D-expanded placenta-derived stromal cell product) [163,164], tocopherol succinate (TS) [70,140,165–168], delta-tocotrienol (DT3) [36], captopril [146,169], and AEOL 10,150 (a superoxide dismutase) [107,170].

NHPs have been deemed by select investigators/regulatory agencies as ‘the gold standard’ of animal models for drug development, specifically for drugs being developed following the FDA Animal Rule [171]. Several promising MCMs under development have been used for cytokine assessment using the NHP model. These MCMs include but are not limited to the following agents: entolimod [104,105,141,147], gamma tocotrienol [103], and AEOL 10,150 [107,170]. The highlights of these studies are listed in Table 2.

4.2. Proteomic analysis with MCMs using advanced high-throughput platforms

There are several studies in murine and NHP models using LC-MS and 2D-DIGE and various MCMs to identify proteomes that can be used as biomarkers for efficacy assessments and drug dose conversions while developing new MCMs following the FDA’s Animal Rule guidelines (Figures 2 and 3) [172].

Figure 3. Experimental design for biomarker discovery of radiation exposure and MCM treatments using nanoUPLC-MS/MS.

A recent report, for example, used 2D-DIGE/MS as the major analytic tool to demonstrate the modulating/mitigating effect(s) of one such MCM, GT3, on acute radiation-induced shifts within proteomes of both hematopoietic and gastrointestinal systems of acutely irradiated mice [89]. Specifically, these analyses demonstrated a change in the expression of 18 proteins in response to TBI in blood plasma (serum) samples of the irradiated animals. Further, these changes were reversed with prophylactic treatment of GT3 [112]. Significant expression changes in β2-glycoprotein 1, α1-acid glycoprotein 1, α2-macroglobulin, complement C3, mannose-binding protein C, and major urinary protein 6 were of particular interest after TBI, and were reversed with GT3 treatment. In another study with jejunum samples with the same mice strain, jejunal lysate from irradiated mice was compared to that of GT3-treated/irradiated mice, and 65 proteins were differentially expressed [111].

There is a report for the analysis of protein expression in spleen tissue of male CD2F1 mice treated with GT3 with 7 Gy ⁶⁰Co gamma TBI (Table 4). Additional MCM studies have employed advanced analytics to evaluate proteomic shifts induced by acute irradiation or by the administration of a given MCM (e.g. TS [111,167,173], G-003 M [174], melatonin [175], Ex-Rad [20,176–180], BIO 300 [21,150,181–186], etc.) under test. The major findings of these limited studies are listed in Tables 4 and 5.

5. Challenges and limitations

Proteins, along with carbohydrates, lipids, and nucleic acids serve as key, fundamentally important structural and functional elements of all cells and tissues. These vitally important molecules often change dynamically over time, while undergoing selective modifications during basic life processes, such as during cell reproduction, differentiation, maturation, aging, etc. As noted here in this review, changing patterns of protein expression often accompany a variety of externally driven micro-environmental events (e.g. acute ionizing radiation exposures) or progression of given diseases, or in response to various drug treatments. Hence, it is abundantly clear that measurements of such proteomic shifts or changes, induced or otherwise, within cells, body fluids, or tissues are of fundamental importance for biomedicine. More specifically, and in terms of the primary subject matter under discussion here, it would seem prudent, if not critical to extend and perhaps expand the research effort to identify, develop, and validate reliable and sensitive biomarkers for acute injury elicited by ionizing radiation. As mentioned above, these putative, proteome-based biomarkers are urgently needed to improve radiation dose assessments and MCM drug dose conversions from animal models to humans. These drug dose conversions present a number of substantial analytical and clinical challenges that need to be met and overcome prior to making use of any new, putatively predictive proteome-based biomarker of MCM efficacy from animal models to clinical use. For example, challenging variables such as sample collection, processing, and storage conditions of collected samples prior to analysis may influence downstream proteomic analyses and biomarker detection. These variables need to be accounted for in any/all work related to biomarker identification and validation. The selection of proteome analysis platforms and data processing as well as tools used for computational analysis are critical for discrimination between samples obtained from the radiation exposed and unexposed study subjects. Similarly, the above factors are also important for discrimination between samples collected from the MCM-treated and untreated study subjects. The hope here is that future studies will identify and validate specific proteomes that serve as predictive/prognostic biomarkers for specific types and conditions of ionizing radiation exposures and related injuries stemming from exposures.

Though there are advancements in the area of proteomic biomarkers, the clinical translation has proved challenging. Some of such obstructions include assays failing to meet clinical precision or the absence of comprehensive evidence that changes in these biomarkers lead to crucial health results. Although there is a vast amount of raw data concerning noted changes within the proteome following ionizing radiation exposures, the sorting, analyses and assembly of such data in order to develop useful algorithms of acute ionizing radiation-related injuries largely remains unaccomplished. In this regard, the field of bioinformatics is progressing rapidly, especially for the analyses of proteomic data and perhaps for the advent of new and improved biomedical tools to assess ionizing radiation exposures, related injuries, and the safety and efficacy of given MCMs under development.

6. Conclusion

New and improved methods for the assessment of populations at risk following unwanted exposures (i.e. accidental or intentional) to ionizing radiation is currently deemed a high priority issue by federal/state emergency response planners responsible for developing and promulgating medical contingency plans of any/all mass casualty situations. The effort to identify and to develop accurate, sensitive, and reliable biomarkers of ionizing radiation-associated exposures and/or consequent injuries are integral to this research effort to improve the effectiveness of current medical emergency plans for any given nuclear/radiological exposure contingency. As illustrated here in this brief review, there appears to be substantial utility in pursuing the use of advanced, radiation-related proteomics as a primary vehicle in attempting to achieve the overall goal of developing radiation/injury-specific biomarkers with improved prognostic, diagnostic, or predictive features.

The reality is that the tools we have available for either radiation exposure estimates or for injury assessments are not only labor intensive but time-consuming and often unreliable. For example, the utility of dose estimates based on physical dosimetry is limited. By contrast, current biodosimetric procedures in general are reasonably sensitive and accurate, but often are technically difficult to perform and inordinately constrained. Due to space limitation, we restricted the focus of this article to acute radiation effects and related issues. Unfortunately, we have not included more expanded descriptions of other types of radiobiological studies involving different radiation qualities and varying doses and dose-rates, e.g. extraterrestrial/space-related research or low-dose and/or protracted exposure-related studies [187–191]. Several studies have demonstrated proteomic changes compared with samples collected from pre-exposure or sham irradiated versus post-irradiated animals in patterns that are dependent on radiation quality, dose, and time post-exposure [12,21].

Cytokines/growth factor-associated proteomes are most frequently detected and validated post-irradiation in animal models of ARS. Proteomics is a powerful approach to investigate the changes in these essential, growth regulatory proteins prior to the onset of clinical symptoms of radiation injury. Improvements in the techniques of proteome identification, associated databases, and bioinformatics have facilitated the development of proteome panels for FDA qualification and for the potential clinical use of the technology in the future. The fully FDA-qualified biomarker(s) of radiation exposure and injury are currently needed; however, there are no such biomarkers available. Further, at this moment, the recent proteomic developments highlight a number of research needs in order to move the field forward: first, there is a need to standardized basic protocols for proteome investigation; second, to validate putatively useful, recently identified targeted biomarkers; and third, to conduct studies with samples from both genders to rule out gender bias as well as additional studies with different age cohorts.

The plasma samples would be the samples of choice for measuring proteomes to determine the effects of radiation exposure. One of the challenges with plasma samples is the high abundance of proteins. More than 95% of the serum proteome is of ∼20 high abundance proteins. These proteins interfere with the detection of low abundance proteins.

7. Expert opinion

With time, proteomics has developed into a diverse scientific discipline of quantitative precision and sensitivity measurements of basic biological events. Advanced proteomic platforms now allow for a high-throughput of samples in addition to a wide variety of studies ranging from single-cell analyses to large populations of individuals [192,193]. New technological advancements may constitute significant steps toward moving proteomics into specialized clinical diagnostic applications [194]. While protein identification based on biochemical techniques is qualitative, MS-based protein identification is highly computational by nature. MS is one of the more powerful technologies currently available to biomedical researchers for analyses of the proteome, with the combined analytics of LC-MS/MS lying at the center of current approaches being applied in the science of proteomics. Accordingly, well-established proteomic laboratories are generally equipped with LC-MS/MS systems that support various flow regimes. In regard to the latter, it is expected that in due course, micro-flow LC-MS/MS will replace nano-LC-MS/MS for large scale proteome analyses. It is also envisioned that ‘sub-proteome’ analyses will become more common, as will the analytics of relatively high abundance post-translational modifications (e.g. phosphorylation) of proteins become more amendable and accessible. The analysis of protein PTM is challenging. For PTMs such as phosphorylation and glycosylation, newer chromatographic techniques with higher resolving powers have been shown to be beneficial in the separation and subsequent analyses of such protein PTMs. The resolving capability of LC separations usually increases with flow rate but sensitivity of the measurement trends in the opposite direction [195]. In brief, micro-flow LC-MS/MS can be useful for select types of proteomic analyses, e.g. body fluids, glycoproteomes and phosphoproteomic profiling of cell lines, etc. Isomeric phosphopeptides tend to be better resolved using micro-flow LC-MS/MS rather than with nano-flow LC-MS/MS, but with a significant loss in sensitivity.

Earlier studies using the NHP model have demonstrated that multi-omics approaches can be effective in identifying putative, early arising markers of radiation exposure and related injury in blood plasma/serum [19,110,196–210]. These omic-based NHP studies designed to assess the pathophysiological effects of acute irradiation included the use of advanced, state-of-the-art MS technology. Additional animal (rodent)-based work has been reported using comparable discovery designed approaches. However, as informative as these studies might be, they are limited in value by the fact that they do not address certain response variables, most notably the influence of gender. Further, there are no follow-up studies using large animal models in which the observed change(s) of interest, i.e. putative biomarkers, have been validated relative to the major endpoints of interest, namely survival, structure/function of primary radiosensitive organs, as well as the capacity of the biomarker to differentiate exposed from unexposed individuals within the ‘at risk’ population. Biomarkers tied to measures of effectiveness of most MCMs currently under study remain insufficiently validated as well. Still further, newly proposed NHP work to validate multi-omics biomarker panels in compliance with FDA guidelines remains to be conducted and/or reported. In brief, there is presently an acute shortage of effective and reliable biomarkers available to help in the identification of individuals at risk of developing acute radiation-induced illness. Hence, there is a critical need to develop and to have available (i.e. in the emergency medical kit for nuclear/radiological exposure contingencies) well-validated biomarkers of radiation exposure and related consequent injury. Regardless of how daunting this research and development may appear; the application of high-throughput techniques will most certainly aid and abet the identification and validation of reliable and efficient biomarkers for both detection and clinical management of at-risk populations in case of radiological/nuclear emergencies.

It is important to validate biomarkers in humans. The biomarker research community may launch clinical studies on cohorts of patients exposed to therapeutic doses of radiation, such as malignancies, inflammatory diseases, or other clinical conditions to test the biofluids (serum or urine samples) for markers identified in preclinical studies. Such studies with clinical doses of radiation may not be optimal for ARS situations but such studies will increase the confidence for success in this area. Finally, on a positive note, we point out that recent developments in the area of bioinformatics are highly encouraging, making the identification and validation of the radiation exposure targeted proteome quite feasible.

Article highlights

• Biomarkers are important for assessing radiation exposure doses and also for medical countermeasure (MCM) dose conversions from animals to humans for regulatory approval following the US FDA Animal Rule.

• Research and development of such MCMs includes identification of radiation injury biomarkers.

• Fully FDA-qualified biomarker(s) of acute radiation injury is needed, and there is no such biomarker at this moment.

• Recent studies using irradiated small and large animal models suggest that proteomic changes within biofluids may prove to be useful for the identification and validation of specific biomarkers for both radiation exposure and consequent injury.

• High-throughput techniques have identified large numbers of potential biomarkers within the proteome that show promise as biomarkers for radiation exposure.

• Some proteome biomarkers for MCMs have been demonstrated across various species including humans and appear promising in terms of identifying MCM-specific biomarkers.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer disclosures

Peer reviewers on this manuscript have received an honorarium from Expert Review of Proteomics for their review work but have no other relevant financial relationships to disclose.

Author contributions

VK Singh and TM Seed performed literature searches, drafted the manuscript, revised, and finalized for publication. M Srivastava assisted in the collection and writing of methodology section.

Acknowledgments

The opinions or assertions contained herein are the private views of the authors and are not necessarily those of the Uniformed Services University of the Health Sciences, or the Department of Defense. The mention of specific therapeutic agents does not constitute endorsement by the U.S. Department of Defense, and trade names are used only for the purpose of clarification. We are thankful to Ms. Alana Carpenter for editing the manuscript and Mr. Stephen Y. Wise for literature searches. We apologize to those having contributed substantially to the topics discussed herein that we were unable to cite because of space constraints.

Additional information

Funding

This work was supported by funding from Armed Forces Radiobiology Research Institute/Uniformed Services University of the Health Sciences grant # [AFR‐B4‐10978 and 11615] awarded to VKS.