LC-MS/MS-Profiling of Human Serum Unveils Significant Increase in Neuroinflammation and Carcinogenesis Following Chronic Organophosphate Exposure

Abstract

The utilization of organophosphate pesticides (OPs) has escalated in response to the growing global food demand driven by a rapidly increasing population and the environmental disruptions caused by climate change. While acute exposure leads to cholinergic poisoning, chronic OP exposure has been linked to organ dysfunction, inflammation, and carcinogenesis. Serum samples from healthy individuals (n = 11), patients with acute OP exposure (n = 12), and those with chronic OP exposure (n = 31) were analyzed to discern the differentially expressed pathways after acute and chronic OP exposure. Differential expression analysis identified 132 proteins altered in chronic exposure vs. control, 86 in acute exposure vs. control, and 124 in chronic vs. acute exposure. Pathway analysis revealed increased blood coagulation and reduced LXR/RXR activation and DCHR24 signaling in both acute and chronic exposures. Elevated levels of pro-inflammatory proteins, such as S100A8, VWF, and GPIBA, were observed, particularly in chronic exposure, highlighting significant inflammatory effects of OP exposure. These findings provide insights into the pathological mechanisms underlying chronic OP exposure and its contribution to inflammation and long-term health risks.

Graphical Abstract

Introduction

Pesticides refer to chemical agents applied to crops for protection from pests, weeds, and insects and include a wide range of insecticides, fungicides, herbicides, and rodenticides^{1, 2}. Four major categories of pesticides are used today, classified based on their chemical composition, mode of action, level of toxicity, and method of application; Carbamates, Organochlorine, Organophosphate, and Pyrethroids¹. Organophosphate-based pesticides (OPs) are molecules containing esters, amides, or thiols derived from phosphoric acid^{1, 3}. In recent years, the prevalence of OPs has increased dramatically, primarily driven by the increasing food demand due to the rapidly growing population, especially in developing countries⁴. However, with the increased use of OPs comes environmental degradation and biological toxicity⁴. Detectable levels of organophosphate residues – such as chlorpyrifos, diazinon, and malathion – have been identified in biological matrices such as human blood serum and urine^5–7.

Organophosphates primarily exert their pesticide effects by the inhibition of acetylcholinesterase (AChE) in pests; however, because OPs have the same effect on humans, avoidance of exposure is necessary^{8, 9}. AChE is primarily located in the neuromuscular junctions, as well as peripheral and central cholinergic nerve terminals¹⁰. AChE is responsible for the hydrolysis of acetylcholine at these sites into acetate and choline¹¹. Acetylcholine (ACh) plays a pivotal role in modulating several vital physiological functions, including myocardial contractions, regulation of vascular tone, gastrointestinal peristalsis, and exocrine glandular secretion^{10, 12}. Within the CNS, ACh is integral to cognitive processes such as learning, memory consolidation, and the regulation of circadian rhythm. At the neuromuscular junction, ACh facilitates the excitation of skeletal muscle fibers, initiating muscle contraction¹². Prolonged inhibition of AChE can cause a buildup of acetylcholine at such neuromuscular junctions leading to cholinergic poisoning¹³. Cholinergic poisoning at neuromuscular junctions can lead to muscle fasciculations and inflammation due to overactivation of acetylcholine receptors¹⁴. Whereas the inhibition of AChE in the central nervous system may lead to insomnia, confusion, and drowsiness^{14, 15}.. AChE is primarily located in the neuromuscular junctions, as well as peripheral and central cholinergic nerve terminals¹⁰. AChE is responsible for the hydrolysis of acetylcholine at these sites into acetate and choline¹¹. Acetylcholine (ACh) plays a pivotal role in modulating several vital physiological functions, including myocardial contractions, regulation of vascular tone, gastrointestinal peristalsis, and exocrine glandular secretion^{10, 12}. Within the CNS, ACh is integral to cognitive processes such as learning, memory consolidation, and the regulation of circadian rhythm. At the neuromuscular junction, ACh facilitates the excitation of skeletal muscle fibers, initiating muscle contraction¹². Prolonged inhibition of AChE can cause a buildup of acetylcholine at such neuromuscular junctions leading to cholinergic poisoning¹³. Cholinergic poisoning at neuromuscular junctions can lead to muscle fasciculations and inflammation due to overactivation of acetylcholine receptors¹⁴. Whereas the inhibition of AChE in the central nervous system may lead to insomnia, confusion, and drowsiness^{14, 15}.

Recent epidemiological studies have identified a correlation between organophosphate exposure and an elevated risk of neurodegenerative disorders such as Alzheimer’s Disease (AD), Parkinson’s Disease (PD), and Amyotrophic Lateral Sclerosis (ALS)^{16, 17}. Although the acute neurotoxic effects of OPs, particularly their role in cholinergic poisoning, are well-established, there remains a substantial gap in our understanding of how OPs interact with the broader human proteome. Moreover, emerging evidence suggests that OPs may also contribute to multi-organ dysfunction¹⁸. This includes hepatic fibrosis and impaired liver function¹⁹, pulmonary damage²⁰, and cardiovascular complications²¹, all of which are underexplored in current literature. The need for comprehensive studies on the systemic effects of OPs beyond their neurotoxic impact is critical for a fuller understanding of their long-term health consequences.^{16, 17}. Although the acute neurotoxic effects of OPs, particularly their role in cholinergic poisoning, are well-established, there remains a substantial gap in our understanding of how OPs interact with the broader human proteome. Moreover, emerging evidence suggests that OPs may also contribute to multi-organ dysfunction¹⁸. This includes hepatic fibrosis and impaired liver function¹⁹, pulmonary damage²⁰, and cardiovascular complications²¹, all of which are underexplored in current literature. The need for comprehensive studies on the systemic effects of OPs beyond their neurotoxic impact is critical for a fuller understanding of their long-term health consequences.

This study aims to address this research gap using an advanced bottom-up LC-MS/MS-based proteomics approach to analyze blood serum samples acquired from patients who were either acutely or chronically exposed to OPs. We aim to elucidate the alterations in biological mechanisms and pathways that underpin the diverse and often severe health effects associated with chronic OP exposure and to distinguish these mechanisms from those involved in acute exposure. These insights have the potential to transform the therapeutic approach to OP exposure and inform the development of a more targeted public health strategy amidst the escalating use of pesticides worldwide.

Materials and Methods

Chemicals and Reagents

HPLC-grade water (H₂O), acetonitrile (ACN), and MS-grade formic acid (FA) were sourced from Fisher Scientific (Fair Lawn, NJ, USA). Ammonium bicarbonate (ABC), dithiothreitol (DTT), and iodoacetamide (IAA) were obtained from Sigma-Aldric (St. Louis, MO, USA). The Trypsin/Lys-C mix MS-grade was supplied by Promega (Madison, WI, USA). The 14 Multiple Affinity Removal Column, along with Buffer A and Buffer B, were purchased from Agilent Technologies Inc. (Santa Clara, CA).

Study Participants

The study was conducted according to a protocol approved by the Ethics Committee of the Faculty of Medicine, Alexandria University (approval number 0107603) and in accordance with the Helsinki Declaration of 1964, and its later amendments. The study was registered on the Protocol Registration and Results System of ClinicalTrials.gov under the Identifier NCT06021262. The study aimed at examining the potential neuroinflammatory biomarkers in patients with chronic OP exposure. For recruitment, sample size was estimated guided by reported serum levels for two inflammatory markers reported to increase in Parkinson’s disease, a neuroinflammatory consequence of OP exposure²², the first being IL-6²³, while the second was the C-reactive protein²⁴. Based on the differences and observed variabilities reported for these markers^{23, 25}, a sample size analysis²⁶ for 1(control or acute OP exposure):2(Chronic OP exposure) enrolment ratio returned a 10/20 sample based on IL-6 values, and a 9/18 sample based on CRP values for an 80% power with 95% confidence. Since the present study was exploratory with an interest in changes in chronic exposure patients, patient recruitment in this group extended to an additional 50% margin to improve the detection of differences, while maintaining a reasonable recruitment time frame.

Blood samples, including control (n = 11), chronic OP exposure (n = 31), and acute exposure (n = 12) were obtained from 18–60-year-old subjects after signing an informed consent. Chronic exposure samples were collected in October 2022 from male small-holder vegetable farmers from Kafr El-Sheikh Governorate in Egypt with known repeated exposure to the OP pesticide through repeated self-application over at least the previous three cultivation seasons. Farmers were interviewed to confirm the exposure through the lack of use of personal protective gear. Acute exposure samples were collected from patients presenting to the Poison Center at the Alexandria Main University Hospital with documented OP pesticide exposure between May and October 2023. Control samples were obtained from age-matched city dwellers from Alexandria, Egypt. Five ml of whole blood were collected in tubes containing a separating gel and clot activator. The tubes were left to rest for 1 hour then centrifuged in a refrigerated centrifuge at 1,500 x g for 10 minutes, the serum was then separated into 0.5ml aliquots stored at −80°C. At the time of experiment, serum proteins were extracted from 100 μL of human blood serum by adding 200 μL Dichloromethane/ methanol (DCM/MeOH) (1:2 v/v) and vortexing for 30 s. The samples were then incubated on ice for 1 h before 75 μL of DCM and water were added and vortexed vigorously for 30 s each time. Next, the samples were centrifuged at 5000 RPM for 15 min. Three layers consisting of water soluble metabolites, proteins, and lipid fractions were formed, from which the middle layer containing the proteins was carefully collected and transferred to a new tube.

The mean age was 40 for control, 39 for acute exposure, and 43 for chronic exposure. Profenofos was the predominant OP compound, accounting for 5 (41.67%) of acute cases and 17 (54.85%) of chronic cases. Other significant OP compounds included Chlorpyrifos, Malathion, and Methamidofos. Most patients had no prior health conditions; however, among chronic exposure cases, 2 patients had diabetes and 2 had hypertension, while 1 acute exposure patient presented with chronic kidney disease. The primary route of exposure for both acute and chronic exposure were inhalation and dermal contact. Lastly, the median exposure time for chronic patients was 15 years.

Depletion of High-Abundance Proteins and Protein Assay

Prior to LC-MS/MS analysis, 14 high-abundant proteins in the serum samples were depleted to enhance the sensitivity of low-abundant proteins. These proteins include Albumin, IgG, antitrypsin, IgA, haptoglobin, transferrin, alpha-2-macroglobulin, fibrinogen, alpha-1-acid-glycoprotein, IgM, complement C3, apolipoprotein AI, apolipoprotein AII, transthyretin. These high-abundance proteins constitute the majority of the total protein concentration in serum and interfere with the detection and quantification of low-abundance proteins. Given their critical biological significance, the study of low-abundance proteins is crucial for a comprehensive proteomics analysis.

The samples were reconstituted in 200 μL of 25 mM ammonium bicarbonate (ABC) solution containing 2.5% sodium deoxycholate (SDC). The samples were then homogenized using a microtube homogenizer and subjected to sonication on ice for an hour. Then, 30 μL of both control and experimental samples were subjected to depletion using the Human 14 Multiple Affinity Removal Column by Agilent. Following depletion, the resulting fractions were concentrated using a 5K MWCO Amicon spin concentrator and then reconstituted in 50 mM ammonium bicarbonate (ABC) buffer for further analysis. The protein concentrations for the final samples were determined using the Micro BCA^™ Protein Assay Kit from Thermo Scientific, Rockford, PIL.

Tryptic Digestion of Low Abundance Proteins

The equivalent of 50 μg of protein was taken from each sample and denatured at 90 °C for 15 minutes. The denatured proteins were then reduced by the addition of 2.5 μL of 200 mM dithiothreitol (DTT) and incubation at 60°C for 45 minutes. The reduced samples were then carbamidomethylated using 10 μL of 200 mM iodoacetamide (IAA) and incubated at 37°C for 45 minutes. A second addition of 2.5 μL of 200 mM DTT followed by incubation at 37°C for 30 minutes was performed to quench the excess IAA. Trypsin was added to the alkylated samples in a 1:25 enzyme-to-protein ratio and the samples were incubated at 37°C for 18 hours. Samples were then dried using a SpeedVac concentrator, reconstituted in 0.1% formic acid, and subjected to LC-MS/MS proteomics analysis.

LC-MS/MS Proteomics Analysis

The tryptic digests, once dried, were reconstituted in a solution containing 2% ACN and 0.1% FA. For analysis, 1 μL of these samples, corresponding to 1 μg of total protein, was injected into a C18 trap column (75 μm x 2 cm, 2 μm, 100 Å; from Thermo Scientific, Pittsburgh, PA) and maintained there for 10 minutes. Subsequently, the samples were transferred to an Aclaim PepMap C18 capillary column (75 μm x 15 cm, 2 μm, 100 Å; also, from Thermo Scientific, Pittsburgh, PA) via an Ultimate 3000 nanoUHPLC system (Dionex, Sunnyvale, CA). The system operated at a flow rate of 300 nL/min and a temperature of 30°C. The mobile phase A comprised an aqueous solution with 2% ACN and 0.1% FA, while mobile phase B was a mix of ACN, 2% water, and 0.1% FA. The peptide separation gradient involved initially holding 5% of mobile phase B for the first 10 minutes, followed by a linear increase to 35% between 11 and 80 minutes, then a rise to 60% at 110 minutes. From 110 to 113 minutes, mobile phase B was increased from 60% to 90%. This 90% level was maintained for 5 minutes, then reduced to 5% between 118 and 119 minutes, and finally held at 5% from 119 to 135 minutes. The nanoUHPLC system was connected to an Orbitrap QExactive HF mass spectrometer (Thermo Scientific, San Jose, CA), operating in positive ion mode. The mass spectrometer was operated in data-dependent acquisition mode with an MS1 resolution of 120,000 for MS1, covering an m/z range of 370–1800 with an AGC target of 1E6. The MS2 acquisition was performed at a resolution of 45,000, with a scan range of 200–2000 m/z and a dynamic exclusion window of 10 seconds. A summary of the proteomics workflow is presented in Figure 1.

Figure 1.

Summary of the Experimental Workflow.

Protein Identification and Quantification

Protein identification and quantification were performed using Proteome Discoverer (PD) software (Thermo Scientific) using the UniProtKB/Swiss-Prot human protein database. The peptide length was specified between 6 and 40 amino acids. Carbamidomethylation of cysteine was set as a constant modification, with acetylation of the N-terminus and oxidation of methionine were variable modifications. Identification criteria included 1% FDR for high-confidence proteins and 5% FDR for medium-abundance proteins. Only high-confidence peptides were considered. Protein precursor mass tolerance was set to 10 ppm. A total of 401 proteins were identified, comprising 386 high-confidence and 15 medium-confidence proteins. The low-confidence proteins (FDR > 0.05) were filtered out prior to further analysis. None of the 14 high-abundance proteins were among the 401 proteins used for further analysis.

Further analysis, including statistical testing, differential expression analysis via hierarchically clustered heatmaps and volcano plots, and gene set enrichment analysis were performed using R and Python scripts. Ingenuity pathway analysis (IPA) by Qiagen was used to examine functional correlations and perform further pathway analysis.

Data Analysis

Displaying high dimensional data, beyond 3-dimensions, can be extremely challenging. Principal component analysis is a dimensionality reduction operation used to reduce the dimensionality of large data sets while preserving as much information in the data as possible. PCA is particularly valuable in proteomics experiments as it allows for the visualization of high-dimensional data, where each protein represents a dimension, in either two- or three- dimensional space. Furthermore, its unsupervised nature allows for the visualization of natural clusters within the data without prior knowledge of sample classification, making it especially useful for uncovering differences between cohorts or identifying hidden patterns. Principal components (PCs) are ranked based on the amount of information they contain. Usually, the first 2 principal components are displayed in a 2D scatter plot, each representing a unique sample. PCA was performed on the entirety of the proteomics data, the resulting plot is displayed in Figure 2A.

Figure 2.

A) Principal Component Analysis (PCA) plot of the whole proteome in control, acute exposure, and chronic exposure cohorts. Volcano plot comparisons for B) chronic exposure (n = 31) vs. control (n = 11), C) acute exposure (n = 12) vs. control (n = 11), and D) chronic exposure (n = 31) vs. acute exposure (n = 12). The significance thresholds are displayed by the dashed lines (p-value < 0.05 and fold-change > 2) and color-coded points, with red signifying upregulation and green downregulation. The grey points represent proteins outside the statistical threshold.

Following the PCA, differential expression analysis was conducted to identify proteins displaying significantly altered expression pattern. Three comparisons were made, chronic OP exposure vs. control, acute OP exposure vs. control, and chronic vs. acute OP exposure using a Welch’s T-test. Proteins with p-values < 0.05 were considered to be differentially expressed. Volcano plots were constructed to display the expression patterns of the significantly altered proteins. A volcano plot is a scatter plot displaying the log2-fold chance on the x-axis and the -log10(p-value) on the y-axis, with the upregulated highly differentially expressed proteins (red) being towards the top right, and downregulated (green) towards the top left of the scatter plot.

Furthermore, hierarchically clustered heatmaps were constructed for each comparison to identify clusters of significantly altered proteins displaying similar changes in expression. Following comparative statistical analysis, gene ontology (GO), hallmark, and panther gene set enrichment analysis were conducted for the statistically significant proteins across each comparison. The analysis aimed to discern the altered biological processes, molecular functions, and cellular pathways underlying organophosphate exposure. The clusterProfiler package was employed along with Hallmark, Panther, and the Genome-wide annotation for humans for the gene set enrichment analysis. Protein-protein interaction (PPI) plots were constructed to identify connections between differentially expressed proteins. PPI plots, also known as string analysis, are used to analyze connections between a group of proteins. Proteins are represented as nodes (circles) and the connections between them as edges. Proteins that are either expressed, colocalized, or involved in similar functions tend to cluster together with bolder connections

Parallel Reaction Monitoring Validation

Finally, parallel reaction monitoring (PRM) was conducted to validate the expression of 14 differentially expressed proteins, including VWF, DNAH9, ACTG1, ITIH2, ITIH3, MBL2, GPX3, APOL1, IGF2, AGT, APOA4, S100A8, AGT, and APOE. These proteins were specifically selected due to their involvement in key pathways discussed throughout this article. Only 26 out the 31 chronic exposure samples could be used for PRM analysis due to limitations in sample quantity. The LC-MS/MS parameters were identical to those of the full scan, except for the addition of the inclusion list. A retention time window ± 5 minutes (totaling 10 minutes) around the peak elution time was used for each peptide to accommodate minor shifts in the elution time. Peptide identification was performed using Proteome Discoverer (Thermo Scientific) and subsequent quantification was performed using Skyline (MacCoss Lab Software) using the fragments provided by Proteome Discoverer. The quantified data was log2 transformed and a Welch’s t-test was utilized for differential expression analysis, with a p-value < 0.05 signifying differential expression.

Results

The 2D PCA plot represented in Figure 2A demonstrates a distinct separation among the control, chronic exposure, and acute exposure groups, highlighting substantial proteomic changes following organophosphate exposure. The ellipsoids represent a 98% confidence interval for each sample group, calculated based on a normal distribution. It is important to note that gender-specific principal component analysis was not performed due to the chronic exposure cohort consisting solely of male patients. For detailed comparisons, PCA plots for individual group analyses are provided in Figure S1.

Differential Expression Analysis via Volcano Plots and Hierarchically Clustered Heatmaps

The differential expression analysis identified 132 differentially expressed proteins in the chronic exposure versus control comparison, 86 in the acute exposure versus control comparison, and 124 in the chronic versus acute exposure comparison. The detailed results of this differential expression analysis are presented in Table 1. The volcano plots in Figure 2B–2D display the expression of significantly altered proteins. Two major clusters were identified in the hierarchically clustered heatmaps for each comparison, belonging to upregulated and downregulated proteins. A subset cluster of upregulated proteins involved in blood coagulation was identified after chronic exposure, including kininogen (KNG1), protein C (PROC), and von Willebrand factor (VWF). The heatmaps for the significantly altered proteins in each comparison are displayed in Figure S2–S4.

Table 1.

Summary of Differential Expression Analysis using Welch’s T-Test for pairwise comparison

Comparison	Differentially Expressed1	Upregulated2	Downregulated2
Chronic vs. Control	132	61	71
Acute vs. Control	86	37	49
Chronic vs. Acute	124	69	55

¹Proteins with p-value < 0.05 are classified as differentially expressed across each comparison

²Differentially expressed proteins with log2(fold-change) > 0 are considered upregulated and log2(fold-change) < 0 are considered downregulated

Comparative Analysis of Differentially Expressed Proteins

Intersection analysis was performed to identify common differentially expressed proteins across different comparisons. As shown in Figure 3A, 31 proteins were significantly altered and common to all three comparisons. The corresponding heatmap and pathway involvements are displayed in Figure 3B, 3C, and 3D respectively. Additionally, 28 proteins were uniquely altered in the chronic exposure vs. control comparison, 5 were unique to acute exposure vs. control, and 20 were unique to chronic vs. acute exposure. The representative barplots for the 31 proteins commonly altered in all three groups are provided in Figure S5.

Figure 3.

Comparison analysis of differentially expressed proteins represented by the A) Venn diagram comparing common significant proteins between the three comparisons, B) Hierarchically clustered heatmap for the 31 proteins commonly differentially expressed between the three cohorts. The Sankey plot displaying Hallmark gene annotations for common differentially expressed proteins in C) chronic exposure vs. control and D) acute exposure vs. control, only processes involving two or more proteins are displayed.

Notably, A major downregulation of apolipoproteins was observed in both acute and chronic exposure, with the decrease being more pronounced in acute exposure. Apolipoprotein E (APOE), C1 (APOC1), C3 (APOC3), and A4 (APOA4) were significantly downregulated in both chronic and acute exposure. The downregulation in acute exposure was substantial enough to also be significant when comparing chronic vs. acute exposure.

A significant disruption in the serpin family of proteins was observed, with specific members showing differential expressions in response to chronic and acute organophosphate exposure. Specifically, SERPINA4, SERPINA6, SERPINA7, SERPINF1, SERPINF2 were downregulated in patients with chronic exposure, while SERPINA10 and SERPIND1 were upregulated. In cases of acute exposure, SERPINA6 and SERPINC1 were downregulated, whereas SERPINA7 and SERPINA10 were upregulated. The serpin proteins are known to play critical roles in processes such as blood coagulation, inflammation, and complement activation, indicating significant dysregulation in these biological pathways.

Additionally, a similar pattern of disruption was observed among apolipoproteins, with APOA4, APOC1, APOC3, and APOL1 clustering together, as illustrated in the heatmap in Figure 3B. This clustering further suggests that organophosphate exposure leads to widespread proteomic alterations that may have profound implications for lipid metabolism and related physiological processes. The corresponding boxplots for these proteins are available in Figure S5, providing a detailed view of their expression levels across different exposure conditions.

Gene Set Enrichment and Protein-Protein Interaction Analysis

Hallmark and panther gene set enrichment revealed major alterations in the blood coagulation pathway in chronic and acute exposure cohorts. Out of the significant proteins for chronic exposure vs. control and acute exposure vs. control, there were 26 and 16 proteins involved in the coagulation pathway respectively. Panther analysis revealed disruptions in the nicotinic acetylcholine receptor signaling pathway after chronic and acute exposure, with the effect being more pronounced in chronic exposure, suggesting a worsening effect of organophosphates on acetylcholinesterase over prolonged exposure.

The disruption was primarily evident from the dysregulation of key proteins such as MYH1, MYH7B, MYH8, MYH6, ACTBL2, ACTG1, and BCHE for both chronic and acute exposure. Butyryl choline esterase is involved in the scavenging of organophosphate radicals before they act upon AChE in the central nervous system^{27, 28}. Unlike AChE, loss of function of BCHE is not associated with any toxicity in humans, hence upregulation of BCHE in both acute and chronic patients suggests a compensatory mechanism from the human body against organophosphates. Furthermore, under high concentrations of acetylcholine, as would be the case upon suppression of AChE, BCHE becomes highly efficient at the hydrolysis of acetylcholine^{29, 30}.

Additionally, the p53 and Alzheimer’s disease-presenilin pathways were uniquely dysregulated in chronic OP exposure, indicating potential neurodegenerative and carcinogenic effects associated with prolonged exposure. Lastly, the complement system was observed to be disrupted in both acute and chronic exposure samples. In total, 14 proteins associated with the complement pathway were dysregulated in patients with chronic exposure, compared to 10 proteins in those with acute exposure, suggesting a more pronounced impact on the immune system with prolonged exposure.

PPI analysis for the significantly altered proteins in chronic OP exposure vs. control revealed a major cluster of proteins involved in lipid metabolism, complement activation, and coagulation. Furthermore, another small cluster of proteins involved in muscle contraction was also observed. These clusters were also observed in acute exposure vs. control. The PPI plots are displayed in Figures 4E and 4F.

Figure 4.

Hallmark gene set enrichment for A) chronic exposure vs. control and C) acute exposure vs. control. The x-axis represents the -log10(p-value), with higher values indicating higher significance. Panther gene set enrichment for B) chronic exposure vs. control and D) acute exposure vs. control, similarly with the x-axis representing -log10(p-value). Protein-protein interaction plots for E) chronic exposure vs. control and F) acute exposure vs. control with major clusters being highlighted in the red-dashed circles.

Ingenuity Pathway Analysis

While gene set enrichment analysis can identify disrupted pathways, they are essentially hypergeometric and may not predict whether a pathway is activated or inhibited. Thus, Ingenuity Pathway Analysis (IPA) software (Qiagen) was used for directional pathway analysis. A summary of the statistically significant (p-value < 0.05) are provided in supplementary figure S6 and S7 for chronic exposure vs. control and acute exposure vs. control, respectively.

For chronic exposure patients, LXR/RXR activation and DHCR24 signaling pathways were the most downregulated, while the formation of fibrin clots and serotonin receptor signaling pathways were upregulated as displayed in Figure 5A. Furthermore, adhesion of blood platelets, aggregation of blood cells, and coagulation were all functions that were upregulated suggesting a significant increase in the coagulation and clotting pathways (Figure 5A). Function of muscle was downregulated, suggesting muscle damage effect of chronic organophosphate exposure. Multiple proteins such as VWF, SERPIND1, S100A8, GP1BA, APOE, APOB, and FGA were found to be involved in more than one of these pathways, as displayed in Figure 5A and 5C. Upstream analysis for chronic samples predicted inhibition of key transcription factors such as HNF4A, CEBPA, TP53, HTT, and RET. A dramatic increase in the serotonin signaling pathway was observed in chronic exposure. Activation of serotonin receptor signaling has previously been reported in acute OP exposure^31–33, here we observe that the increase was more prominent in chronic exposure patients. Serotonin signaling was elevated in acute exposure patients but was not statistically significant. Lastly, the inhibition of the SUMO family of proteins, including SUMO1, SUMO2, and SUMO3 along with PRDX1 and NFE2L2 was also predicted, as represented in Figure 5B.

Figure 5.

Ingenuity pathway analysis for differentially expressed proteins. A) Canonical pathways and disease and function enrichment for chronic exposure vs. control. B) Predicted downregulation of SUMO family, PRDX1, and NFE2L2 in chronic exposure vs. control. C) Canonical pathways and disease and function enrichment for acute exposure vs. control. D) Predicted downregulation of the TP53 gene after organophosphate exposure. Predicted downregulation is represented in blue, predicted upregulation in orange, observed downregulation in green, and observed upregulation in red.

LXR/RXR activation and DCHR24 signaling pathways were similarly inhibited in acute exposure patients, along with the activation of the S100 family signaling pathway (Figure 5C). The disease and function analysis revealed an increase in the formation of blood clots, inflammation of the lungs, and cerebrovascular dysfunction along with a decrease in the concentration of fatty acids as shown in Figure 5C. Lastly, TP53 was predicted to be significantly downregulated after both acute and chronic OP exposure, however a larger number of proteins that predicted the downregulation of TP53 were altered under chronic exposure, suggesting a bigger change (Figure 5D).

PRM Validation Results

A total of 86 peptides from the 14 proteins chosen for PRM quantitation were identified and quantified, with majority of the peptides displaying a similar trend as observed in the main study. Boxplots for six representative peptides are displayed in Figure 6, with the further boxplots being displayed in Supplementary Figure S9. Additionally, the abundance data for each peptide has been provided in Supplementary Table S4 along with their gene names and Uniprot accession IDs.

Figure 6.

Boxplot displaying the PRM abundance of 6 representative peptides (control in grey, acute exposure in yellow, and chronic exposure in red). The majority of the peptides were found to follow a similar trend in PRM and full-scan MS.

Discussion

While the acute neurotoxic effects of OP exposure, primarily cholinergic poisoning, are well-documented and understood, emerging evidence suggests that chronic exposure to OPs can lead to significant and often overlooked organ dysfunction^19–21. Gulf War illness (GWI), affecting 30–44% of 1990–1991 Gulf War veterans, is primarily linked to chronic organophosphate exposure and remains a major health concern despite decades of research and significant healthcare investments^{34, 35, 36}. This study aimed to identify the biochemical pathways and protein interactions that contribute to the disorders observed after chronic OP exposure, beyond the well-known acute neurotoxic effects. Findings revealed disrupted lipid metabolism, increased inflammation via dysregulated liver X receptor (LXR) pathways, reduced SUMOylation, and upregulation of pro-inflammatory proteins such as S100A8 and Von Willebrand Factor (VWF). These changes suggest a persistent inflammatory state contributing to organ dysfunction and neuroinflammation-related cognitive disorders. Serum proteins play essential roles in transport, immune response, inflammation regulation, and blood coagulation, making them critical targets for studying the effects of organophosphate exposure. Furthermore, serum proteins provide a minimally invasive diagnostic tool for detecting organophosphate exposure, offering a practical alternative to tissue biopsies.

The liver X receptors (LXRs) are type II nuclear receptors part of the nuclear hormone receptor superfamily responsible for gene regulation via controlling cholesterol and lipid metabolism³⁷. LXRs are accountable for raising lipoprotein concentration, and the downregulation of various apolipoproteins such as APOE, APOC3, APOC1, and APOL were majorly responsible for the prediction of the inhibition of LXR/RXR pathway³⁸. LXRs possess an anti-inflammatory effect by repressing the expression of various pro-inflammatory proteins^{37, 39}. LXRs are activated by various oxysterol ligands and form heterodimers with Retinoid X Receptors (RXRs) upon ligand binding⁴⁰. The LXR/RXR heterodimer promotes the transcription of APOE (downregulated), ABCA1, ABCG, and various other proteins involved in lipid metabolism^{39, 41}.The liver X receptors (LXRs) are type II nuclear receptors part of the nuclear hormone receptor superfamily responsible for gene regulation via controlling cholesterol and lipid metabolism³⁷. LXRs are accountable for raising lipoprotein concentration, and the downregulation of various apolipoproteins such as APOE, APOC3, APOC1, and APOL were majorly responsible for the prediction of the inhibition of LXR/RXR pathway³⁸. LXRs possess an anti-inflammatory effect by repressing the expression of various pro-inflammatory proteins^{37, 39}. LXRs are activated by various oxysterol ligands and form heterodimers with Retinoid X Receptors (RXRs) upon ligand binding⁴⁰. The LXR/RXR heterodimer promotes the transcription of APOE (downregulated), ABCA1, ABCG, and various other proteins involved in lipid metabolism⁴¹. Furthermore, SUMOylated LXRs also interact with the corepressor complex at the promoter of NFkB, preventing the clearance of these corepressors and thus inhibiting the expression of these pro-inflammatory genes (Figure 5). Inhibition of the LXR/RXR activation pathway suggests an increase in the transcription of pro-inflammatory proteins and an overall increase in inflammation^{40, 42, 43}. Furthermore, the anti-inflammatory effects of synthetic LXRs have previously been demonstrated^{39, 44–46}.

Notably, a decrease in SUMOylation was also observed in patients with chronic exposure, as indicated by the predicted inhibition of the SUMO family of proteins shown in Figure 5B. SUMOylation, much like ubiquitination, is a post-translational modification involving the attachment of a small ubiquitin-like modifier (SUMO) proteins to lysine residues in target proteins⁴⁷. This modification plays a critical role in the regulation of transcription, cell cycle progression, DNA repair, and subcellular localization^{47, 48}. Additionally, SUMOylation inhibits NF-κB pathway activity by modifying IκBα, with a reduction in SUMOylation suggesting an increase in the release of pro-inflammatory mediators such as IL-6^49–52. Furthermore, SUMOylation is essential for the function of the survival of motor neuron (SMN) complex, with deficiencies in SMN linked to spinal muscular atrophy⁵³. These findings highlight the multifaceted impact of SUMOylation on inflammation and cellular function, suggesting that disrupted SUMOylation may contribute significantly to the pathology observed in chronic OP exposure. Notably, a decrease in SUMOylation was also observed in patients with chronic exposure, as indicated by the predicted inhibition of the SUMO family of proteins shown in Figure 5B. SUMOylation, much like ubiquitination, is a post-translational modification involving the attachment of a small ubiquitin-like modifier (SUMO) proteins to lysine residues in target proteins⁴⁷. This modification plays a critical role in the regulation of transcription, cell cycle progression, DNA repair, and subcellular localization^{47, 48}. Additionally, SUMOylation inhibits NF-κB pathway activity by modifying IκBα, with a reduction in SUMOylation suggesting an increase in the release of pro-inflammatory mediators such as IL-6^49–52. Furthermore, SUMOylation is essential for the function of the survival of motor neuron (SMN) complex, with deficiencies in SMN linked to spinal muscular atrophy⁵³. These findings highlight the multifaceted impact of SUMOylation on inflammation and cellular function, suggesting that disrupted SUMOylation may contribute significantly to the pathology observed in chronic OP exposure.

S100 calcium-binding protein A8 (S100A8) was notably upregulated after chronic and acute exposure, with the effect being more prominent in chronic exposure patients. S100A8 plays an essential role in inflammation via the recruitment of neutrophils and has been observed to be upregulated in various inflammatory diseases such as gout, diabetes, and obesity^{54, 55}. Von Willebrand Factor (VWF), another protein involved in the progression of inflammation via neutrophil recruitment, was also observed to be significantly upregulated in both acute and chronic exposure patients⁵⁶.

Blood coagulation, also referred to as blood clotting, is a crucial physiological process that prevents excessive blood loss upon injury^{57, 58}. Coagulation is essential for maintaining the integrity of the circulatory system upon injury; however, constant upregulation of the coagulation cascade can lead to serious medical problems such as thrombosis⁵⁹, myocardial infarction⁶⁰, and ischemic stroke⁶¹. Furthermore, vascular abnormalities and alterations in the coagulation system have also been associated with mild cognitive impairment (MCI) and Alzheimer’s disease (AD)⁶². A significant increase in blood coagulation was observed in both acute and chronic exposure cohorts, as shown in Figures 4A – 4D, 5A, and 5C. Thrombospondin-1 (THBS1) is a component of the platelet-α-granules released during platelet activation and is known to accelerate platelet aggregation. THBS1 was found to be significantly increased in chronic exposure patients, and while there was an overall increase in acute exposure patients, the increase was not statistically significant. Glycoprotein Ib-IX-V-complex, composed of glycoprotein Ib-beta (GPIBB), glycoprotein Ib-alpha (GPIBA), glycoprotein IX, and GPV, acts as a receptor for VWF on the platelet surface. The binding of VWF to the GP Ib-IX-V complex facilitates platelet adhesion to vascular subendothelium to increase blood coagulation^{63, 64}.Blood coagulation, also referred to as blood clotting, is a crucial physiological process that prevents excessive blood loss upon injury^{57, 58}. Coagulation is essential for maintaining the integrity of the circulatory system upon injury; however, constant upregulation of the coagulation cascade can lead to serious medical problems such as thrombosis⁵⁹, myocardial infarction⁶⁰, and ischemic stroke⁶¹. Furthermore, vascular abnormalities and alterations in the coagulation system have also been associated with mild cognitive impairment (MCI) and Alzheimer’s disease (AD)⁶². A significant increase in blood coagulation was observed in both acute and chronic exposure cohorts, as shown in Figures 4A – 4D, 5A, and 5B. Thrombospondin-1 (THBS1) is a component of the platelet-α-granules released during platelet activation and is known to accelerate platelet aggregation. THBS1 was found to be significantly increased in chronic exposure patients, and while there was an overall increase in acute exposure patients, the increase was not statistically significant. Glycoprotein Ib-IX-V-complex, composed of glycoprotein Ib-beta (GPIBB), glycoprotein Ib-alpha (GPIBA), glycoprotein IX, and GPV, acts as a receptor for VWF on the platelet surface. The binding of VWF to the GP Ib-IX-V complex facilitates platelet adhesion to vascular subendothelium to increase blood coagulation^{63, 64}. The increase in expression of GPIBA, which contains the binding site for VWF, concurrent with the increase of VWF in chronic exposure patients suggests a systemic increase in blood coagulation.

3ß-hydroxysterol-Δ24-reductase (DHCR24) is involved in the final step of cholesterol synthesis and catalyzes the reduction of the delta-24 double bond of sterol intermediates⁶⁵. DCHR24 is known to protect cells from oxidative stress by reducing caspase 3 activity along with promoting resistance towards Alzheimer’s disease-associated neurodegeneration^{66, 67}. The strong inhibition of DHCR24 signaling (Supplementary figure S8) after organophosphate exposure suggests a strong neurotoxic effect of OPs primarily caused by cholesterol deficiency and desmosterol accumulation⁶⁷ in the brain.

Additionally, neuroinflammation and vascular damage, often caused by aberrant coagulation, have been implicated in neurodegenerative disorders such as vascular cognitive impairment and dementia^68–70. Moreover, NFkB is responsible for inducing the expression of various pro-inflammatory genes such as interleukin-6 (IL-6) as displayed in Figure 7. Therefore, downregulation of the LXR/RXR activation pathway may indirectly predict the upregulation of IL-6, which has been associated with Alzheimer’s disease and general cognitive decline^{69, 71}. The coordinated regulation of DHCR24 and LXR/RXR activation pathways plays a critical role in attenuating inflammatory responses as evidenced by the substantial overlap in protein networks involved in both pathways. Disruptions in LXR/RXR signaling and DHCR24 function, along with the upregulation of pivotal coagulation-related proteins such as VWF, GPIBA, and S100A8 and downregulation of apolipoproteins, underscore a mechanistic connection between chronic organophosphate exposure and the pathogenesis of neurodegenerative disorders.

Figure 7.

Predicted downregulation of LXR/RXR activation and its impact on inflammatory mediators and immune response. Predicted downregulation is represented in blue, predicted upregulation in orange, observed downregulation in green, and observed upregulation in red.

The increase in inflammation upon chronic exposure was further substantiated by the downregulation of inter-alpha-trypsin inhibitor heavy chains H2 (ITIH2) and H3 (ITIH3). These proteins are members of the inter-alpha-trypsin inhibitors (ITIs) family which are made up of several heavy chains (ITIH1, ITIH2, ITIH3, and ITIH4) and a light chain, bikunin⁷². The ITI family is integral to numerous physiological processes, including tumor suppression, wound healing, and the regulation of inflammation⁷². Both ITIH2 and ITIH3 are postulated to have anti-inflammatory properties and a decrease in concentration has been associated with various cancers^{73, 74}.The increase in inflammation upon chronic exposure was further substantiated by the downregulation of inter-alpha-trypsin inhibitor heavy chains H2 (ITIH2) and H3 (ITIH3). These proteins are members of the inter-alpha-trypsin inhibitors (ITIs) family which are made up of several heavy chains (ITIH1, ITIH2, ITIH3, and ITIH4) and a light chain, bikunin⁷². The ITI family is integral to numerous physiological processes, including tumor suppression, wound healing, and the regulation of inflammation⁷². Both ITIH2 and ITIH3 are postulated to have anti-inflammatory properties and a decrease in concentration has been associated with various cancers^{73, 74}. The concurrent downregulation of ITIH2 and ITIH3 in response to chronic OP exposure not only corroborates the heightened inflammatory response but also offers new insights into the potential relationship between prolonged organophosphate exposure and an increase risk of cancer development. The increase in inflammation and the inhibition of the LXR/RXR activation and DHCR24 signaling pathways were confirmed through the validation of protein expression using parallel-reaction monitoring. This analysis targeted key proteins involved in these pathways, including apolipoproteins (APOA4, APOE, APOC3, and APOL1), inter-alpha-trypsin inhibitors (ITIH2 and ITIH3), von Willebrand factor (VWF), glutathione peroxidase 3 (GPX3), protein S100-A8 (S100A8), actin (ACTG1), and angiotensinogen (AGT).

While the primary risks associated with organophosphate exposure are cholinergic poisoning and inflammation, recent epidemiologic studies have linked OPs to an increased risk of developing cancer^{75, 76}. We observed a similar effect with the predicted downregulation of tumor protein P53 (TP53) after OP exposure as displayed in Figure 5D. TP53 dubbed the “guardian of the genome”, is the most well-known tumor suppressor gene and is in charge of various cell responses such as apoptosis, cell repair, cell survival, and transcriptional regulation^77–80.

Mannose-binding lectin 2 (MBL2) is associated with innate immunity and decreased MBL2 amounts have been associated with an increased risk of hepatocellular carcinoma and predict a poor outcome⁸¹. Glutathione peroxidase 3 (GPX3), also downregulated after OP exposure, is known to have tumor-suppressor properties⁸². Reactive oxygen species (ROS) such as superoxide radicals can promote carcinogenesis by causing DNA damage and promoting cell proliferation^{83, 84}. GPX3 prevents cancer growth via scavenging of ROS and its downregulation has been associated with the development of breast, ovarian, colon, and gastric cancers^{83, 85–88}. Lastly, Heat shock 70 kDa protein 8 (HSPA8), responsible for molecular chaperoning and protein folding was observed to be upregulated after OP exposure. Overexpression of HSPA8 has been observed in colorectal, liver, and breast cancer^89–92. The predicted downregulation of TP53 coupled with aberrant expression of key proteins such as GPX3, MBL2, and HSPA8 illustrate a pro-tumorigenesis effect of organophosphates.

Exposure to OP compounds, including chlorpyrifos⁹³, profenofos⁹⁴, and malathion⁹⁵, have previously been associated with an increase in systemic inflammation and carcinogenesis through generation of reactive oxygen species. However, thus far the underlying biochemical mechanisms have remained understudied. This study highlights the alterations in serum proteins that may directly contribute to the observed effects. These findings not only enhance our understanding of the biochemical alterations caused by organophosphate exposure, but also highlight potential therapeutic targets and diagnostic tools. Moreover, key pathways such as LXR/RXR activation and DHCR24 signaling may serve as possible therapeutic targets for treatment of OP exposure related symptoms. Although this study represents a beginning into the exploration of proteomic changes associated with OP exposure, future research is clearly warranted. While the findings of this study present potential protein markers that may prove valuable for diagnostic and therapeutic purposes, subsequent investigations involving a larger cohort is necessary to confirm and validate these findings. Moreover, because this study primarily focused on overall protein expression, it will be essential to investigate post-translational modifications – such as glycosylation, phosphorylation, ubiquitination, and SUMOylation – that may also contribute to the observed increase in inflammation and carcinogenesis. Future research direction may involve utilizing animal models to investigate proteomic changes associated with specific exposure routes, providing insights into how various routes of exposure may uniquely impact the proteome within a controlled environment. Additionally, animal model studies would allow for the study of tissue specific changes associated with chronic OP exposure.

Conclusion

Extensive bottom-up proteomic analysis of human serum samples from patients exposed to organophosphate (OP) pesticides revealed significant alterations in biological pathways critical to inflammation, coagulation, neurodegeneration, and cancer. Differential expression analysis identified key proteins, including apolipoproteins, serpin family members, and VWF that were especially dysregulated in chronic exposure, suggesting a progressive increase in inflammation, carcinogenesis, and liver dysfunction. Ingenuity pathway analysis revealed the downregulation of critical pathways such as LXR/RXR activation and DCHR24 signaling, indicating potential pro-inflammatory and neurotoxic effects of OPs. The predicted inhibition of key tumor suppressor proteins like TP53, along with the observed dysregulation of proteins involved in protection from oxidative stress and innate immunity, such as GPX3 and MBL2, suggests a possible carcinogenic risk associated with chronic OP exposure. These findings provide valuable insights into the molecular mechanisms underlying various neurocognitive and organ dysfunction symptoms associated with chronic OP exposure and emphasize the need for further research for the development of possible therapeutics and interventions to mitigate the adverse effects of OP poisoning.

Supplementary Material

The Supporting Information is available free of charge at *******.

Organophosphate_Proteomics_SI_V5: Principal component analysis for individual comparisons. A) Chronic exposure vs. control, B) acute exposure vs. control, C) chronic vs. acute exposure. The ellipsoids represent 99% confidence interval; Heatmaps with hierarchical clustering for chronic exposure vs. control. Upregulation is represented in red and downregulation is represented in blue. The top annotations represent the grouping with control being grey and chronic exposure dark red; Heatmaps with hierarchical clustering for acute exposure vs. control. Upregulation is represented in red and downregulation is represented in blue. The top annotations represent the grouping with control being grey and acute exposure light red; Heatmaps with hierarchical clustering for chronic exposure vs. acute exposure. Upregulation is represented in red and downregulation is represented in blue. The top annotations represent the grouping with control being dark red and acute exposure light red; Boxplots for proteins differentially expressed between all three comparisons (chronic vs. control, acute vs. control, and chronic vs. acute). *, pvalue < 0.05, **, pvalue < 0.01, ***, pvalue < 0.001; Canonical Pathways involvements of significant proteins for chronic exposure vs. control. Orange represents upregulation and blue represents downregulation. DHCR24 signaling pathway as predicted in chronic OP exposure vs. control. The downregulation of the pathway leads to an increase in cognitive impairment and the activation of the NFkB inflammatory pathway. Blue indicates predicted inhibition and orange indicates predicted activation. Boxplots displaying the abundance of peptides targeted for PRM analysis.

Data Availability Statement

The raw mass spectrometry data from this study is publicly available on data dryad via the following link: http://datadryad.org/stash/share/MEFoRXnxYPFRVnXk-ek6BLPeWjb9-Axp4UDLPNC0xAI