Dynamic changes of serum protein in rats with acute intoxication of Chinese cobra snake venom by proteomic analysis

Abstract To elucidate the toxic mechanism of snake venom at the protein level, proteomics technology was applied to investigate the effect of venom on circulation in the mammalian body. Temporal proteomic analysis was performed to profile the dynamic changes in the sera of Sprague–Dawley rats administered with Chinese cobra venom or saline. Using 8-plex iTRAQ analysis, 392 and 636 serum proteins were identified to be linearly upregulated or downregulated over time in the low-dose group and high-dose group, respectively. These proteins were mainly associated with the acute phase response pathway, complement system, and liver X receptor (LXR)/retinoid X receptor (RXR) and farnesoid X receptor (FXR)/RXR activation pathways. Compared with the low-dose group, the immune response and integrin pathways were inhibited in the high-dose group, although no obvious effect was observed. With consistently higher or lower expression in the high-dose group compared to the low-dose group throughout the whole process of venom poisoning, two proteins, Kininogen-1 (KNG1) and orosomucoid 1 (ORM1), which are involved in metabolism and immune response, occupied a core position in the pathway network and are considered venom dose-dependent biomarker candidates.


Introduction
Venomous snakebite is a frequently devastating albeit neglected public health threat [1,2]. There are more than 2 700 species of snakes worldwide, and about one-fifth of them are venomous and are capable of producing venom with biological activities from venom glands, favouring the capture and digestion of prey [3]. Once injected into the bloodstream, snake venom spreads out into the body through circulation and causes various local or systematic poisoned symptoms. Approximately 100 000 snakebites occur annually in China, of which 73% of victims are young and middle-aged. Approximately 5%-10% of the victims were killed by snakebite, and another 25%-30% of the victims suffer from permanent tissue damage [4]. The Chinese cobra (Naja atra) is a member of the Elapidae family and one of the 10 most venomous snakes in China [5]. The effects of the Chinese cobra bite are acute and critical, and it is commonly observed in emergency departments. The Chinese cobra can spit large amounts of venom containing both neurotoxin and haemotoxin. Highly lethal neurotoxins may result in neuromuscular inhibition, muscle paralysis and respiratory failure, while haemotoxins can destroy red blood cells, causing a breakdown or inflammation in the body. Rapid absorption of the neurotoxin and haemotoxin from the injection site into the systemic circulation indicates the fast onset of these principal toxins, which are responsible for the early systemic manifestation of envenoming [6].
In the early days, the mechanism of venom toxicity was generally studied using traditional methods, such as immunofluorescence, ELISA and enzyme activity detection [7][8][9]. However, these methods exhibit low efficiency and are more likely to miss key factors. As an emerging technology, mass-spectrometry-based proteomics has been widely employed in the study of protein profiles and proprieties in body fluids, including serum, cerebrospinal fluid, nipple leaching fluid, tears, saliva, amniotic fluid, urine and ascites [10]. Proteomic tools have been used to study the composition and crossreactivity of anti-venoms with venoms to address the pathology of snakebite envenoming [11]. Malih et al. [12] elucidated the venom proteome of Naja haje legionis using a combination of size exclusion chromatography, reverse-phase high-performance liquid chromatography (HPLC), Tricine/SDS-Page and Q-TOF tandem mass spectrometry. N. haje legionis venom contained neurotoxic activities, inducing irreversible blockade of neuromuscular transmission in both rodent and chick nerve-muscle preparations, and exhibited myotoxic and cardiotoxic activities [13]. In addition, a functional proteomic approach was used to identify geographic variations of king cobra venoms from southeast Asia and China [14]. Thus, proteomics may serve as a powerful tool to study the mechanism of venom toxicity on a global scale. Snake venoms were analyzed by "venomics", a proteomic strategy used to determine their composition [15][16][17]. Venoms of the Naja species can induce a predominantly cytotoxic pattern of envenoming that may evolve into tissue necrosis and gangrene [18]. Tanaka et al. [19] found that the venom of Micrurus nigrocinctus, a species of venomous elapine snake, could activate the complement pathway. Once activated, the complement system produces a large number of anaphylatoxin to promote vasodilation and thus accelerate the spread of other toxic proteins in circulation. An L-amino acid oxidase from Ophiophagus hannah snake venom demonstrated potent inhibitory activity on platelet aggregation induced by ADP and U46619 [20]. However, few systemic studies have been performed about serum proteins after intoxication of Chinese cobra venom. Therefore, a proteomics approach was performed to systematically study the effect of snakebite time and venom dose on serum proteins, in order to investigate the alteration of serum proteins caused by Chinese cobra venom. First, Sprague-Dawley (SD) rats were used to establish the snakebite mode, and dose tests were performed to determine the lethal dose. Next, rat sera 1, 2 and 4 h post-poisoned were collected and analyzed using proteomics to understand the process of venom entrance into the circulation. By analyzing the proteome alteration in the haematological system after venom entrance, it may provide additional insight on the effect of venom on circulation and identify potential biomarkers for intoxication of Chinese cobra venom, which may be useful in forensic toxicology.

Instruments and reagents
Chinese cobra venom was obtained from a rainbow snake farm in Yingtan, China. All reagents used in this study were purchased from Sigma (St. Louis, MO, USA) unless specified otherwise. Isobaric tags for the relative and absolute quantification kit was purchased from AB Sciex (Framingham, MA, USA).
Acetonitrile hypergrade was obtained from Merck (Darmstadt, Germany). The ProteoMiner protein enrichment kit was purchased from BIO-RAD (Hercules, CA, USA). The 2D Clean-up kit and Plus One 2-D Quant kit were purchased from GE Healthcare (Uppsala, Sweden).
Rat models of snakebite SD rats (450-600 g) of SPF grade were obtained from Shanghai SLAC Laboratory Animal Co., Ltd. Lyophilized venom powder from Chinese cobra were dissolved in saline in a concentration series and intramuscularly injected into the rat leg, respectively. The lethal dose was selected as the amount that results in death 4 h post-injection, which was 0.04 mg venom per 100 g rat weight.
Rats were divided into three groups according to the dosage as follows: the high-dose group (lethal dose, 0.4 mg/g, n ¼ 12), low-dose group (half the lethal dose, 0.2 mg/g, n ¼ 8) and control group injected with 0.2 mL saline (n ¼ 6). Sera were collected from the rat's orbit at 1, 2 and 4 h post-injection. After 4 h of injection, all of the surviving rats were sacrificed by cervical dislocation.

Enrichment of low-abundance serum proteins and iTRAQ labelling
Low-abundance serum proteins were concentrated using the ProteoMiner enrichment kit according to the manufacturer's instructions and further purified using the 2D Clean-up kit. Precipitated proteins were re-dissolved in the solution provided in the iTRAQ kit and the concentration was determined using the PlusOne 2-D Quant kit. Next, 100 mg protein were used for iTRAQ labelling proteomics analysis. Protein digestion and labelling were performed according to the manufacturer's instructions. Briefly, after reduction and alkylation, proteins were digested for 16 h at 37 C using sequencing grade modified trypsin (W protein :W enzyme ¼ 50:1). Digested peptides from each sample were then labelled with different iTRAQ tags and incubated at room temperature for 1 h. Labelled peptides were then mixed, dried under vacuum and stored at À20 C for subsequent analysis.
Peptide fractions obtained from SCX were reconstituted in 50 mL buffer A (50% ACN, 0.1% formic acid) and separated on a ZORBAX 300SB-C18 column (0.1 mm Â 15 mm, 5 mm, 300 Å). Reverse phase highperformance liquid chromatography was performed using the Shimadzu LC20-AD system. Peptides were loaded onto a trap column at 60 mL/min and eluted with a 90-min gradient of 5%-35% solvent B (95% ACN, 0.1% formic acid) at 300 nL/min. MS data were acquired using the data-dependent proteomics method on the Q-Exactive mass spectrometer (ThermoFisher Scientific, San Jose, CA, USA) equipped with a HESI-II Spray source. Nitrogen was used as both the sheath gas and the auxiliary gas at 35 bar and 10 bar, respectively. HESI source parameters were set as capillary temperature 320 C, spray voltage of 3 800 V and heater temperature 300 C. In each cycle, a full scan was acquired with a mass range of m/z 400-1 800. The most intense precursor ions were selected and subsequently fragmented by higher energy collisional dissociation (HCD).

Data analysis
The raw MS data from data-dependent acquisition (DDA) were analyzed by Peaks Studio (Thermo Fisher) for protein identification and quantification. K-means clustering and Pearson correlation coefficient analysis were performed to identify proteins with linearly increased or decreased expression over time in each of the poisoned groups (low and high dose). These proteins were further analyzed using the Ingenuity Pathway Analysis platform (IPA, QIAGEN, Redwood City, CA, USA; www.qiagen.com/ingenuity), including downstream effect analysis, pathway analysis, regulator effect analysis and protein-protein interaction analysis, in order to investigate the corresponding biological response over the time course. Subsequently, proteins in each poisoned group were further compared with the control group, and those with a ratio less than 0.7 or greater than 1.33 were defined as significant proteins [21].

Results and discussion
Behavioral changes of rats after the injection of Chinese cobra venom Immediately after injection with a lethal dose of Chinese cobra venom, the rats' activities were attenuated and their injected limb appeared to be paralyzed. The injected limb swelled 30 min later. One hour post-injection, the rats' lips became purple, and their eyes appeared to be dark and some of the rats had rales. Two hours later, the rats began to twitch and die. Four hours after injection, the rats that were still alive were sacrificed and the injected sites and organs were harvested for observation. A bruise was observed in the muscle and skin at the injected site. The lungs were darker than the control group, while the other organs were normal.
After injection with a low dose of Chinese cobra venom, the poisoned symptoms were similar as those described above, which are declined physical activities, paralyzed limbs, purple lips, dark eyes and rales. However, all of the poisoned rats still survived 4 h after injection. Similarly, a bruise was also observed in the muscle at the injected site and the lungs also appeared darker than the control group, while the other organs looked normal.

Time course study of serum proteome profiles after venom injection
To investigate the effect of venom on serum proteins, iTRAQ was employed to compare the serum proteome change between the control group and poisoned rats (high-dose group and low-dose group) 1, 2 and 4 h post-venom injection. A total of 1 565 proteins were identified, and 367 of these were venom proteins. Serum proteins linearly upregulated or downregulated over time could serve as potential biomarkers to indicate the effect of venom on circulation. In this study, proteins persistently increased or decreased over the observed time points (0, 1, 2 and 4 h) and with absolute number of Pearson correlation coefficient over 0.8, P-value < 0.05 were taken as candidates ( Figure 1). There were 392 dysregulated proteins in the low-dose group and 636 in the high-dose group. Interestingly, the expressions of most candidates were linearly decreased in either the high-or low-dose group, indicating that venom entrance could result in severe damage to the rat's circulation.

Significantly regulated proteins in serum after venom entrance
The expression of serum proteins in the dosage group was further compared with the control group and proteins with a ratio of less than 0.7 or greater than 1.33 were determined as proteins with significantly altered expression (Tables 1 and 2).
Most of the significantly altered serum proteins were identified in both the high-and low-dose groups, with a similar upregulation or downregulation profile. For example, the consistent upregulation of haemoglobin indicated that cobra venom caused damage to red cells and such an observation was consistent with the bruise observed at the injected sites. The upregulation of Tropon indicated a heart problem after injection. Pannexin-2 predominantly exists as transmembrane channels and is primarily expressed in the nervous system, predominantly in the brain and spinal cord [22]. The persistent increased expression of Pannein-2 in rat serum  The expression of serum proteins in the dosage group was further compared with the control group and proteins with a ratio of less than 0.7 or greater than 1.33 were determined as proteins with significantly altered expression. MW: molecular weight.
suggested that cobra venom could destroy neurons and release Pannexin-2 into the blood. Tubulin is a major component of microtubules and predominantly exists in the cytoplasm and cytoskeleton [23]. Downregulation of tubulin post-venom injection indicated that cobra venom could destroy tubulin protein, subsequently disrupting cells. Reports have shown that matrix metalloproteinase is upregulated after the destruction of tubulin, and thus promotes the digestion of collagen and proteoglycan and increases cell apoptosis in the early stage. Taken together, these findings indicate neural and hematic toxicities in cobra venom. Some significantly altered proteins, such as anionic trypsin, 60S ribosomal protein L6, adenylate kinase isoenzyme 1, coagulation factor VII, insulinlike growth factor 2 mRNA-binding protein 1, calpain small subunit 1, T-complex protein 1 subunit β, ATP synthase subunit β, troponin and nucleotide exchange factor SIL1, were unique for the low-dose group ( Table 1). The upregulation of anionic trypsin suggested that Chinese cobra venom could cause damage to the pancreas, as anionic trypsin was also upregulated in the serum of patients with pancreatic carcinoma [24]. 60S ribosomal protein L6, a ribosomal protein, was downregulated in rat serum after venom injection. This protein could be decreased during neuronal differentiation [25]. Chinese cobra venom contains nerve growth factor (NGF), which is involved primarily in the growth, proliferation and survival of neurons. Thus, this finding suggested that neuronal differentiation in rats could be affected upon venom entrance. Adenylate kinase isoenzyme 1 is an adenylate kinase, which is upregulated during neuronal differentiation [26]. Its increased expression could be associated with the NGF in venom. Calreticulin is a Ca 2þ -dependent protein that is widely distributed in eukaryotic cells. The expression of serum proteins in the dosage group was further compared with the control group and proteins with a ratio of less than 0.7 or greater than 1.33 were determined as proteins with significantly altered expression. MW: molecular weight.
Once calreticulin binds with Ca 2þ , it is activated and can trigger an array of kinases and phosphatases, thereby regulating the metabolism. As an important protein involved in signalling transduction, calreticulin exerts multiple biological functions, such as inflammation, metabolism and cell apoptosis, in which many critical enzymes are involved [27][28][29]. In this study, cobra venom injection resulted in the downregulation of calreticulin, which reduces the binding of Ca 2þ ions and renders calreticulin inactive. Thus, various biochemical reactions in rats were hindered, which further resulted in the multiple poisoned symptoms. Some proteins that could be associated with the toxicity mechanism of venom were found in the high-dose venom group (Table 2). Retinol-binding protein 4 (RBP4), a low molecular protein, belongs to the lipocalin family and is the specific carrier for retinol in blood from the liver to the peripheral tissues. This protein can specifically bind to retinal pigment epithelial cells and provide retinol. RBP4 is widely distributed in blood, urine and other body fluids. Due to its low molecular weight and short half-life, RBP4 is important for clinical diagnosis and prognosis of liver and renal disease. Several studies [30][31][32] have shown that its serum concentration could be decreased during liver injury or suffered with dysfunction of renal tubular reabsorption. In our study, RBP4 was linearly downregulated over time, indicating that cobra venom could cause acute injury to rat liver and renal systems. The acute injury of the renal system by the venom has been confirmed in previous studies [33,34], and there has been no literature referring to the hepatotoxicity of venom. Because venom has shown its power in the treatment of hepatic fibrosis, it is possible to affect live function, and further study is needed. RBP4 may be the potential biomarker of cobra venom toxicosis. In the high-dose group, we found some significantly altered proteins associated with mitochondria, such as electron transfer flavoprotein, acetyl-CoA acetyltransferase, NADH dehydrogenase [ubiquinone] flavoprotein 2, 39S ribosomal protein L46 and trifunctional enzyme. These results implied that cobra venom could damage the function of mitochondria. Mitochondria are the primary energy-providing organelles, and its damage may result in various abnormal symptoms [35,36].

Biological alteration of serum after venom entrance
These candidates were further examined using IPA proteome software to investigate the corresponding biological responses over time, including downstream effect analysis, pathway analysis, regulator effect analysis and protein-protein interaction analysis. First, downstream effect analysis was performed and those with the most significantly affected functions are displayed ( Figure 2). Most of these functions were much more dysregulated in response to the high-dosage group compared to the low-dosage group, indicating its more severe toxicity to circulation. Subsequently, enriched pathway analysis was performed and a comparison was made between the two poisoned groups ( Figure 3). As shown in Figure 4, the prothrombin activation pathway was clearly inhibited in the high-dose group, indicating that a higher dose of venom caused severe damage to the rats. However, such a response appeared slightly ambiguous in the low-dose group. As shown in Figure 5, the acute phase response pathway was significantly inhibited in the high-dose group over time, indicating that the acute response system was destroyed and, thus, the rats failed to effectively fight against the venom invasion. Other pathways, such as the complement system, liver X  receptor (LXR)/retinoid X receptor (RXR) and farnesoid X receptor (FXR)/RXR activation pathways associated with the immune response and integrin pathway, were also obviously inhibited. In addition, enzymes involved in glycolysis and some metabolic pathways involved in the synthesis of nitric oxide and oxygen were also significantly inhibited in the high-dose group. In contrast, no such effects were observed in the low-dose group.
The regulatory effects feature in IPA empowers generation of a hypothesis for how a phenotype, function or disease is regulated by particular dysregulated proteins by activated or inhibited upstream transcription factors. Consequently, compared to the low-dose venom, we found that most of the upstream transcription factors in the high-dose group were more associated with the dysregulation of cell migration, adhesion and differentiation, and their major effects on biological functions were the inhibition of development, differentiation and migration.
As shown in Figure 6, IPA network analysis of significant protein interactions also differed between the high-and low-dose groups. In the low-dose group, the three major networks were cell communication, cell organization and cell death. However, the high-dose group was mainly involved in metabolism disorder, protein synthesis blockade and molecular transport blockade.

Dose-dependent effect of venom on serum proteins
In this study, two different venom dosages were applied: high-dose group (0.4 mg/g) and low-dose group (0.2 mg/g). At 1, 2 and 4 h post-venom injection, rat sera were collected, and proteome changes were compared over time. To further investigate the dose-dependent proteins, bioinformatics analysis was further performed on proteins with the same expression profiles; that is, their expression in the high-dose group should be consistently higher or lower than that in the low-dose group throughout the whole three time points. Furthermore, a 1.5-fold change was set as the threshold, and, thus, 11 significant proteins were obtained (Table 3).
Interestingly, all of the 11 significant proteins are relatively associated with cancer, which could also trigger the same host immune response as venom. In addition to cancer, these proteins are associated with some other biological functions, such as Figure 5. The acute phase response signalling pathway in the high-dose group (this pathway could help the host to defend itself against an infection and trauma via acute immunological regulation). In the high-dose group, most of the proteins in this pathway were downregulated (green) or predicted to be inhibited (blue), indicating that such a response was inhibited by a high dose of venom entering. molecular transport, cell movement, cell adhesion and accumulation. Particularly, epidermal growth factor (EGF) is a growth factor that stimulates cell growth, proliferation, and differentiation by binding to its receptor EGFR. The dysregulation of EGF is associated with the development and progression of some cancers and responsible for hypomagnesemia [37,38]. Transmembrane protein SLC26A5 (solute carrier anion transporter family 26, member 5) is the motor protein of the outer hair cells of the mammalian cochlea. This protein could mediate the exchange of chloride across the plasma membrane [39], and its gene mutation could lead to sensorineural hearing loss [40]. BIN2 (bridging integrator 2) is associated with cell migration and invasion, and is involved in the formation of the pseudopod body and inhibition of phagocytosis [41]. Zinc finger protein 451 (ZNF 451) is a transcriptional cofactor that binds to specific DNA and RNA to modulate transcription. It is also the coactivator of the steroid receptor. TAGLN2 (transgelin-2) is one of the earliest markers of differentiated smooth muscle and is also a tumour growth inhibitor. Kininogen-1 (KNG1) is the precursor protein of high-molecular-weight kininogen (HMWK) and low-molecular-weight kininogen (LMWK), in which HMWK is essential for blood coagulation and assembly of the kallikrein-kinin system. Orosomucoid 1 (ORM1) is the serum acute phase protein and is involved in specific immune response. The pathways that these significant proteins are involved in are displayed in Figure 7.
Two metabolic pathways mediated by nuclear receptor, LXR\RXR activation and FXR\RXR activation, were both inhibited in the higher dose situation. LXR, forming a heterodimer with its partner Trend 1 represents proteins that are highly expressed in the high-dose group and À1 represents proteins highly expressed in the low-dose group. Figure 7. Pathways of the dose-dependent proteins. Ellipse icon shows the pathway names and the node colour displays the expression profile of the proteins. Red represents the increased expression of proteins with higher dose venom injection and green is the opposite.
RXR, is an important regulator of inflammation, fatty acid, cholesterol and bile acid metabolism. The FXR, in coordination with RXR, is involved in modulating lipoprotein, fatty acid and glucose metabolism via the regulation of bile acid levels. These results indicate that higher doses of venom could cause severe disorders of protein, cholesterol, fatty acid and glucose metabolism and thus resulted in death of rats 4 h post-venom injection.
Apart from the metabolism pathways, acute phase response, the host systematic defense response triggered by local inflammation was also inhibited when higher doses of venom were used. Acute phase response is a physiological process occurring soon after the onset of infection, trauma, tumour growth and immunological disorders, in which the plasma concentration of inflammatory cytokines and acute phase proteins is increased to maintain the host homeostasis. Such a defense response was activated in the early stage of snakebite, while it was inhibited over time or when higher doses of venom were injected.
Clathrin-mediated endocytosis signalling is used for the trafficking of nutrients, hormones and other signalling molecules from the plasma to intracellular regions. Actin cytoskeleton signalling is essential in some dynamic processes, such as cell movement, axon guidance cytokinesis and phagocytosis. The disorder of these two pathways indicated that higher dose venom would cause more severe damage to host material transmission and motor ability.
The networks of significant proteins are primarily associated with lipid and small molecular metabolism (Figure 8), in which KNG1 and ORM1, the proteins involved in metabolism and immune response, occupied a core position. As a result, these two proteins could serve as venom dose-dependent biomarker candidates.

Conclusion
This study, for the first time, systematically investigated the effect of Chinese cobra snakebite time and venom dose on serum proteins using proteomics. We found that serum proteomes differed between different dosages of snake venom. Chinese cobra venom contains complex zootoxins, including haemotoxin and neurotoxin. In our study, there were 47 and 28 significantly changed proteins in the high-dose group and low-dose group, respectively. Some of these proteins were persistently upregulated or downregulated over the time course and could be the first choice for Chinese cobra snakebite identification. KNG1 and ORM1 were probably the potential biomarkers for the intoxication of Chinese cobra venom.