Nanofractionation Analytics for Comparing MALDI-MS and ESI-MS Data of Viperidae Snake Venom Toxins

Worldwide, it is estimated that there are 1.8 to 2.7 million cases of envenoming caused by snakebites. Snake venom is a complex mixture of protein toxins, lipids, small molecules, and salts, with the proteins typically responsible for causing pathology in snakebite victims. For their chemical characterization and identification, analytical methods are required. Reversed-phase liquid chromatography coupled with electrospray ionization mass spectrometry (RP-LC-ESI-MS) is a widely used technique due to its ease of use, sensitivity, and ability to be directly coupled after LC separation. This method allows for the efficient separation of complex mixtures and sensitive detection of analytes. On the other hand, matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) is also sometimes used, and though it typically requires additional sample preparation steps, it offers desirable suitability for the analysis of larger biomolecules. In this study, seven medically important viperid snake venoms were separated into their respective venom toxins and measured by ESI-MS. In parallel, using nanofractionation analytics, post-column high-resolution fractionation was used to collect the eluting toxins for further processing for MALDI-MS analysis. Our comparative results showed that the deconvoluted snake venom toxin masses were observed with good sensitivity from both ESI-MS and MALDI-MS approaches and presented overlap in the toxin masses recovered (between 25% and 57%, depending on the venom analyzed). The mass range of the toxins detected in high abundance was between 4 and 28 kDa. In total, 39 masses were found in both the ESI-MS and/or MALDI-MS analyses, with most being between 5 and 9 kDa (46%), 13 and 15 kDa (38%), and 24 and 28 kDa (13%) in size. Next to the post-column MS analyses, additional coagulation bioassaying was performed to demonstrate the parallel post-column assessment of venom activity in the workflow. Most nanofractionated venoms exhibited anticoagulant activity, with three venoms additionally exhibiting toxins with clear procoagulant activity (Bothrops asper, Crotalus atrox, and Daboia russelii) observed post-column. The results of this study highlight the complementarity of ESI-MS and MALDI-MS approaches for characterizing snake venom toxins and provide a complementary overview of defined toxin masses found in a diversity of viper snake venoms.


Introduction
Venom is a chemical weapon used by many species across the animal kingdom, with some of the venoms most medically relevant to humans found in snakes.While venomous snakes primarily use venom for prey capture, it can be deployed defensively, and it is in this context that human snakebite occurs.Every year, approximately 5.4 million people suffer from snakebite, resulting in up to 2.7 million cases of envenoming.These envenomings can cause severe long-term health implications, such as amputations or other permanent disabilities, in perhaps as many as 400,000 people a year [1].However, it is thought that an additional 138,000 victims suffer mortality as the result of the bite, and the greatest burden falls on those based in the tropics.Because of this, the World Health Organization added snakebite to their priority list of neglected tropical diseases in 2017 [2][3][4].
Snakebite pathology is the result of the hemotoxic, cytotoxic, or neurotoxic properties of snake venom toxins [5], which may cause local or systemic damage, including necrosis, hemorrhage, and respiratory paralysis [3].Many types of compounds are present in snake venom, including inorganic cations, lipids and carbohydrates.The toxic properties, however, are mainly caused by proteins and peptides, which are highly abundant in snake venom (>90% dry weight) [6].These toxins vary extensively between snakes, of which the most prevalent and medically significant are members of the families Elapidae (elapids) and Viperidae (vipers), the latter of which is divided into the Viperinae (true vipers) and Crotalinae (pit vipers) subfamilies.Within the vipers, the dominant protein toxin families are phospholipases A 2 (PLA 2 s), snake venom metalloproteinases (SVMPs), and snake venom serine proteases (SVSPs).Secondary toxin families include Kunitz peptides, L-amino acid oxidases, cysteine-rich secretory proteins, C-type lectins, disintegrins, and natriuretic peptides [7,8].
The toxin bioactivities of viper venoms have been thoroughly identified for many decades, both at the level of crude venom and purified toxins.From these studies, we know that SVSPs can affect hemostasis through their influence on platelet aggregation, blood coagulation, and fibrinolysis [9,10].In contrast, SVMPs can cause local and systemic hemorrhage via the cleavage of basement membrane proteins and coagulopathy via activation of key clotting factors (prothrombin, factor X) [11].These proteases are also associated with other pharmacological actions, such as myotoxicity and inflammation [12,13].Additionally, synovial-type group II PLA 2 s are prevalent in most vipers [14].These lipid-hydrophilizing enzymes display a variety of pharmacological functions: acidic PLA 2 s inhibit platelet aggregation, whereas basic PLA 2 s are myotoxic, neurotoxic, and/or anticoagulant in nature [15].
Advances in snake venom analysis have revolutionized our understanding of the complex compositions and biological activities of venoms [16].One significant area of progress is in venom separation techniques, where high-resolution analytical chromatographic methods such as reversed-phase liquid chromatography (RPLC), size-exclusion chromatography (SEC), and cation exchange chromatography have been used to fractionate venom components based on their physicochemical properties [17][18][19][20].Subsequently, mass spectrometry (MS) analysis is often directly (i.e., online or hyphenated) or indirectly (i.e., offline) conducted, enabling the measurement of accurate masses of individual venom proteins and peptides [21][22][23].In parallel, post-column bioassay screening approaches have been combined by, for example, nanofractionation analysis for post-column offline screening using biochemical assays to elucidate the cytotoxicity, neurotoxicity, and hemotoxicity of venom toxins [21,24,25].Other examples include venom toxin profiling for its effects on ion channels, enzymes, and cellular signaling pathways [26][27][28].And interaction with antivenom treatment, i.e., antivenomics [29].LC coupled online to MS has been described for several years [30], and MS analysis after electrospray ionization (ESI) following LC separation is the most well-known approach.
ESI is a widely used technique in MS for the analysis of complex biological samples, including snake venoms.ESI allows for the efficient ionization of venom toxins, enabling their detection and characterization [31][32][33].While LC-ESI-MS is most often used for intact toxin analysis of toxins in venoms, matrix-assisted laser desorption/ionization MS (MALDI-MS), which is also a soft ionization technique, can also be used.However, unlike LC-ESI-MS, MALDI-MS cannot be directly coupled post-column to LC, as after post-separation and fraction collection, off-line sample preparation is required prior to MALDI-MS analysis.
In MALDI-MS, the analytes of interest, in this case fractionated snake venom toxins, are mixed with a matrix compound solution and then deposited on a target MALDI plate.After analyte and matrix spotting on the MALDI plate, followed by the drying of the solvents, the MALDI plate is inserted into the mass spectrometer for analysis.The matrix consists of a small organic molecule that absorbs the laser energy of the MALDI source and facilitates the desorption and ionization of the analyte molecules upon laser irradiation enabling efficient ionization of toxins.MALDI-MS is, therefore, much more labor-intensive than LC-ESI-MS and cannot easily be fully automated.Nevertheless, instead of depending solely on one MS method for toxin identification, researchers are increasingly adopting multi-analytical approaches by combining various techniques side by side to obtain more mass spectral information from samples of interest.
In this study, we used LC coupled to nanofractionation analysis for crude venom separation, followed by high-resolution fractionation towards MALDI-MS analysis.In parallel, via a post-column split and prior to fractionation, ten percent of the eluent was directed to online ESI-MS.In addition to performing MALDI-MS on the collected fractions, they were processed for coagulation bioassaying.The analytical setup employed here focused on viper venoms, for which a comparison of measured LC-MALDI-MS and LC-ESI-MS data was made, and integrated with an additional coagulation bioassay.We show that this approach can greatly facilitate the discovery and MS-based identification of bioactive toxins from viper snake venoms.

Results and Discussion
This study investigated the combined use of LC-ESI-MS and LC-MALDI-MS to identify similarities and complementarities in data acquired for viper venom research.Parallel acquired and integrated coagulation bioassaying results were included to demonstrate simultaneous bioassaying capability, and the corresponding MS data collected facilitated the identification of certain coagulopathic toxins.To acquire clear comparisons between the two ionization approaches, Extracted Ion Currents (EICs) were extracted from the ESI-MS data, while the signal intensity of each significant mass in each mass spectrum measured from each MALDI-MS-processed well was plotted on a y-axis versus the retention time of fractionation on the x-axis to create so-called MALDI-MS chromatograms for each toxin.Figure 1A shows a typical MALDI-MS spectrum obtained for nanofractionated toxins from D. russelii venom at the retention time fraction of 20.0 min.The toxin ions detected presented as singly (M + H) + and doubly charged (M + 2H) 2+ ions (which were used for toxin mass deconvolution).Figure 1B shows zoomed-in toxin-ion peaks of the toxin with m/z of 6822 and 13,644, which were the doubly and singly charged ions, respectively, of the toxin, with a mass of 13,644 Da, based on manual deconvolution.
Throughout this paper, toxin masses detected from both the LC-ESI-MS and LC-MALDI-MS results generated in this study are named "matched" results.Matched results were defined as having mass differences of ±10 Da.Masses only found in the LC-ESI-MS data or the LC-MALDI-MS data are referred to as "unmatched" masses.The "matched" and "unmatched" mass ions of the nanofractionated toxins will be discussed, and correlations will be made with the measured coagulation activity bioassay data to highlight parallel biochemical profiling within the workflow.In total, seven viper venoms were analyzed, and the results for each were addressed in the order of Bitis arietans, Bothrops jararaca, Bothrops asper, Crotalus atrox, Daboia russelii, Echis carinatus, and Echis ocellatus.
The results of the B. arietans venom analysis are shown in Figure 2, and we found 16 distinct masses from the ESI-MS measurement.For each charge state envelope representing a toxin's accurate mass, the mass-to-charge (m/z) value belonging to the highest abundant charge state was selected for plotting the respective toxin's EIC.Most toxin masses included for comparison were around 6-29 kDa.For the so-called MALDI-MS chromatograms (Figure 2C), we plotted 14 clearly observed masses retrieved from the fractionated toxins (the masses calculated from ion peaks were generated by the Flexanalysis software 3.4).Among those, four masses were matched with the masses of the ESI-MS results, as defined as having mass differences lower than ±10 Da (Table 1).Three doubly charged m/z (6965, 7121, and 15,027 Da) values were deconvoluted with the accurate masses of 13,930, 14,247, and 30,072 Da.For the bioassay results, only one negative peak representing anticoagulant activity was observed at a retention time of around 13.5 min.The accurate masses corresponding to this biological activity were deconvoluted from both MS methods as 6976.09Da (6966 Da in MALDI) and 13,213.51Da (13,209 Da in MALDI).The latter of these masses, which is in the range known for PLA 2 toxins, was likely responsible for the bioactivity observed.Throughout this paper, toxin masses detected from both the LC-ESI-MS and LC-MALDI-MS results generated in this study are named "matched" results.Matched results were defined as having mass differences of ±10 Da.Masses only found in the LC-ESI-MS data or the LC-MALDI-MS data are referred to as "unmatched" masses.The "matched" and "unmatched" mass ions of the nanofractionated toxins will be discussed, and correlations will be made with the measured coagulation activity bioassay data to highlight parallel biochemical profiling within the workflow.In total, seven viper venoms were analyzed, and the results for each were addressed in the order of Bitis arietans, Bothrops jararaca, Bothrops asper, Crotalus atrox, Daboia russelii, Echis carinatus, and Echis ocellatus.
The results of the B. arietans venom analysis are shown in Figure 2, and we found 16 distinct masses from the ESI-MS measurement.For each charge state envelope representing a toxin's accurate mass, the mass-to-charge (m/z) value belonging to the highest abundant charge state was selected for plotting the respective toxin's EIC.Most toxin masses included for comparison were around 6-29 kDa.For the so-called MALDI-MS chromatograms (Figure 2C), we plotted 14 clearly observed masses retrieved from the fractionated toxins (the masses calculated from ion peaks were generated by the Flexanalysis software 3.4).Among those, four masses were matched with the masses of the ESI-MS results, as defined as having mass differences lower than ±10 Da (Table 1).Three doubly charged m/z (6965, 7121, and 15,027 Da) values were deconvoluted with the accurate masses of 13,930, 14,247, and 30,072 Da.For the bioassay results, only one negative peak representing anticoagulant activity was observed at a retention time of around 13.5 min.The accurate masses corresponding to this biological activity were deconvoluted from both MS methods as 6976.09Da (6966 Da in MALDI) and 13,213.51Da   Figure 3 shows the results of the analysis of B. jararaca venom.In Figure 3D, 22 distinct masses found from the ESI-MS measurement of this venom are presented with the highest intensity m/z value of the related charge states plotted.Most toxin masses included for comparison were around 6-27 kDa.For the so-called MALDI-MS chromatograms (Figure 3C), we plotted 20 clearly observed masses.Among those, six masses were matched with masses from the ESI-MS (Table 1).Three doubly charged m/z values (6901, 7119, and 24,520) were deconvoluted to the accurate masses of 13,806, 14,240, and 48,925 Da, respectively.For the bioassay results, no significant anticoagulant or procoagulant peaks were observed.Based on previous studies on the same venom [34], it is likely that the procoagulant proteases present in this venom denatured during the chromatographic separation or that the venom concentration used was insufficient to yield a strong signal in the bioassay.
For Bothrops asper (Figure 4), we extracted 13 distinct masses from ESI-MS (Figure 4D), and14 masses were deconvoluted from MALDI-MS and plotted (Figure 4C).Of these, eight masses were matched from the two MS methods (Table 1).Two doubly charged m/z-values (6977 and 6964) were deconvoluted to the masses of 13,956 and 13,931 Da, respectively.As for the results of the plasma coagulation assay, procoagulant activity was observed as positive peaks at retention times of 17.5 and 20.5 min, corresponding to the masses of 13,885.40 and 24,837.02Da in ESI-MS, respectively (only 24,828 Da eluted at 20.5 min was found for the MALDI-MS results).The significant anticoagulant peak observed at a retention time of 17.0 min correlated with the mass of 13,956.45Da (13,956 in MALDI).
For the results of Crotalus atrox venom (Figure 5), 12 distinct masses were deconvoluted from ESI-MS (Figure 5D) and 22 masses were deconvoluted from MALDI-MS (Figure 5C).From there, six masses were found to be shared across both MS methods (Table 1).Two doubly charged m/z values (13,435 and 12,329) were deconvoluted to masses of 26,829 and 24,660 Da, respectively.For bioassaying, both procoagulant and anticoagulant activity were observed for the nanofractionated C. atrox venom.Two notable negative peaks were observed showing anticoagulant effects at retention times of 12.0 and 14.For Bothrops asper (Figure 4), we extracted 13 distinct masses from ESI-MS (Figure 4D), and14 masses were deconvoluted from MALDI-MS and plotted (Figure 4C).Of these, eight masses were matched from the two MS methods (Table 1).Two doubly charged m/zvalues (6977 and 6964) were deconvoluted to the masses of 13,956 and 13,931 Da, respectively.As for the results of the plasma coagulation assay, procoagulant activity was observed as positive peaks at retention times of 17.5 and 20.5 min, corresponding to the masses of 13,885.40 and 24,837.02Da in ESI-MS, respectively (only 24,828 Da eluted at 20.5 min was found for the MALDI-MS results).The significant anticoagulant peak observed at a retention time of 17.0 min correlated with the mass of 13,956.45Da (13,956 in MALDI).For the results of Crotalus atrox venom (Figure 5), 12 distinct masses were deconvoluted from ESI-MS (Figure 5D) and 22 masses were deconvoluted from MALDI-MS (Figure 5C).From there, six masses were found to be shared across both MS methods (Table 1).Two doubly charged m/z values (13,435 and 12,329) were deconvoluted to masses of 26,829 and 24,660 Da, respectively.For bioassaying, both procoagulant and anticoagulant activity were observed for the nanofractionated C. atrox venom.Two notable negative peaks were observed showing anticoagulant effects at retention times of 12.0 and 14.5 min and had corresponding ESI-MS masses of 14,210.65 Da (14,217 Da in MALDI-MS) and 26,835.81Da (26,829 Da in MALDI-MS), respectively.Procoagulant bioactivity peaks were observed at retention times of 17.0 and 19.0 min, which corresponded to the ESI-MS-derived accurate masses of 13,587.89Da (13,586 Da in MALDI-MS) and 13,584.87Da (13,584 in MALDI-MS), respectively.For the results of Daboia russelii venom (Figure 6), 11 distinct masses were deconvoluted from the ESI-MS data (Figure 6D), and 15 were deconvoluted from MALDI-MS (Figure 6C).Of these, six masses that matched were found by both MS methods (Table 1).As for the plasma coagulation assay results, both procoagulant and anticoagulant activity were observed (Figure 6B).Three notable negative anticoagulant peaks were found at retention times of 16.5, 23.0 and 25.0 min, where their corresponding masses by ESI-MS were 13,648.61Da (13,644 Da in MALDI-MS), 13,635.12Da (13,633 Da in MALDI-MS), and 27,300.12Da (27,292 Da in MALDI), respectively.In addition, a procoagulant venom peak was observed at a retention time of 24.0-26.0min, which corresponded to the accurate masses of 27,300.12Da determined by ESI-MS and 27,292 Da determined by MALDI-MS.
For Echis carinatus venom (Figure 7), we deconvoluted 13 and 10 distinct, clearly observed masses from the results of the ESI-MS and MALDI-MS experiments, respectively (Figure 7C,D).Among those, five masses were matched across both approaches (Table 1).Only one anticoagulant peak was observed (Figure 7B) at a retention time of 10.0 min, which corresponded to two accurate masses of 13,793.29 Da (13,798 Da in MALDI-MS) and 13,593.18Da (13,590 Da in MALDI).These toxins seem mostly likely to be PLA 2 toxins, given the detected mass.While no procoagulant venom activity was detected in the bioassay, it is likely that the procoagulant proteases present in this venom denatured during chromatographic separation.For the results of Daboia russelii venom (Figure 6), 11 distinct masses were deconvoluted from the ESI-MS data (Figure 6D), and 15 were deconvoluted from MALDI-MS (Figure 6C).Of these, six masses that matched were found by both MS methods (Table 1).As for the plasma coagulation assay results, both procoagulant and anticoagulant activity were observed (Figure 6B).Three notable negative anticoagulant peaks were found at retention times of 16.5, 23.0 and 25.0 min, where their corresponding masses by ESI-MS were 13,648.61Da (13,644 Da in MALDI-MS), 13,635.12Da (13,633 Da in MALDI-MS), and 27,300.12Da (27,292 Da in MALDI), respectively.In addition, a procoagulant venom peak was observed at a retention time of 24.0-26.0min, which corresponded to the accurate masses of 27,300.12Da determined by ESI-MS and 27,292 Da determined by MALDI-MS.The final venom tested was that of Echis ocellatus (Figure 8), for which we extracted six distinct masses from ESI-MS (Figure 8D), and ten masses were deconvoluted from MALDI-MS (Figure 8C).From there, three masses were matched within both MS methods (Table 1).As for the plasma coagulation assay results (Figure 8B), anticoagulant activity was seen as a negative peak at a retention time of 9.3 min corresponding to a mass of 13,815.24Da in ESI-MS (13,819 Da in MALDI-MS), which again likely represents a PLA 2 toxin.As with E. carinatus, no procoagulant venom activity was observed, which, given the nature of this procoagulant venom, is almost certainly due to proteases denaturing during the chromatographic separation step.For Echis carinatus venom (Figure 7), we deconvoluted 13 and 10 distinct, clearly observed masses from the results of the ESI-MS and MALDI-MS experiments, respectively (Figure 7C,D).Among those, five masses were matched across both approaches (Table 1).Only one anticoagulant peak was observed (Figure 7B) at a retention time of 10.0 min, which corresponded to two accurate masses of 13,793.29 Da (13,798 Da in MALDI-MS) and 13,593.18Da (13,590 Da in MALDI).These toxins seem mostly likely to be PLA2 toxins, given the detected mass.While no procoagulant venom activity was detected in the bioassay, it is likely that the procoagulant proteases present in this venom denatured during chromatographic separation.The venom of viper snakes represents complex mixtures of toxins that exhibit diverse structures and bioactivities.With a better understanding of the toxins in venoms, their variability, and envenoming mechanisms, new pathology-and/or target-specific snakebite treatments may be developed [35].Moreover, it may give insights into snake evolution [36] and might even result in the discovery of new (bio)pharmaceuticals [37,38].The final venom tested was that of Echis ocellatus (Figure 8), for which we extracted six distinct masses from ESI-MS (Figure 8D), and ten masses were deconvoluted from MALDI-MS (Figure 8C).From there, three masses were matched within both MS methods (Table 1).As for the plasma coagulation assay results (Figure 8B), anticoagulant activity was seen as a negative peak at a retention time of 9.3 min corresponding to a mass of 13,815.24Da in ESI-MS (13,819 Da in MALDI-MS), which again likely represents a PLA2 toxin.As with E. carinatus, no procoagulant venom activity was observed, which, given the nature of this procoagulant venom, is almost certainly due to proteases denaturing during the chromatographic separation step.The venom of viper snakes represents complex mixtures of toxins that exhibit diverse structures and bioactivities.With a better understanding of the toxins in venoms, their bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Discussion
In this study, we presented a comparison of analytical results by investigating data obtained from several viper venoms analyzed by both ESI-MS and MALDI-MS approaches.In addition, post-column plasma coagulation screening was included in this study to assess bioassay integration.This study involved results obtained from seven geographically diverse and medically important viper species, specifically B. arietans, B. jararaca, B. asper, C. atrox, D. russelii, E. carinatus, and E. ocellatus.For the analysis using MALDI-MS, we permitted a 10 Da cut-off limit to determine whether deconvoluted masses matched with corresponding accurate masses retrieved from the ESI-MS data.Similar mass errors for MALDI-TOF-MS, as used in our study, were also used in previous studies (shown in Table 2).In general, the deconvoluted masses from both ESI-MS and MALDI-MS presented an overlap in masses found, ranging from between 25% and 57% of toxins detected, depending on the venom analyzed (Table 1).When looking at the matching masses found for both methods, the focus was on clearly observed ions in the MS spectra.In contrast to ESI-MS, MALDI-MS required labor-intensive manual effort during both the sample preparation and data processing phases.In total, 39 good-intensity-observed matching masses were found across the resulting datasets.Most matching masses were found to be around 5-9 kDa (28.2%; calculated by the number of deconvoluted masses from 5 to 9 kDa in MALDI-MS/the total matched masses in MALDI-MS × 100%) and 13-15 kDa (38.4%) kDa.Five good-intensity toxins were found with masses around 24-28 kDa.These masses were 24,837.02Da (24,828 Da in MALDI-MS) and 27,854.79Da (27,844 Da in MALDI-MS) for B. asper venom, 26,835.81Da (26,829 Da in MALDI-MS) and 24,650.4Da (24,660 in MALDI-MS) for C. atrox venom, and 27,300.12Da (27,292 Da in MALDI-MS) for D. russelii venom.While considering the complex composition and significant venom variations in Viperidae species, for certain mass ranges, it is difficult to define the toxin family only by its molecular mass.For full identification of all toxins in each venom analyzed, additional proteomics studies have to be conducted.These studies represent standalone elaborate studies and were out of the scope for the current study in which a new analytical approach for venom research was presented.For the venoms included in our study, venom proteomics studies have been performed and published.These proteomics studies comprehensively analyzed one or several venoms with, as a core focus, the elucidation of the venom's toxin composition.For instance, Wagstaff et al. characterized the venom proteome and the venom gland transcriptome of Echis ocellatus to provide a detailed compositional analysis for designing novel toxin-specific immunotherapies [39].Alape-Girón et al. performed a comparative proteomic characterization of the geographic variation in Bothrops asper venom from Central America and northern South America [40].Calvete et al. used venomics approaches to gain an insight into the overall Crotalus atrox venom proteome [41].Sharma et al. unveiled the complexities of Daboia russelii venom in their venomics study and identified 63 different proteins belonging to 12 families [42].Dingwoke et al. analyzed the venom of Bitis arietans by a venomics method, which resulted in the identification of 79 toxins belonging to 14 families [43].Zelanis et al. studied the sex-dependent variation in the Bothrops jararaca proteome and highlighted the role of glycosylation as a key PTM (post-translational modification) contributing to variability through a combination of SDS-PAGE and 2-DE and through shotgun proteomic analysis [44].Patra et al. carried out a venomics study to analyze the Echis carinatus proteome and found large differences when using different databases for processing their proteomics MS data [45].Our focus was on developing a preliminary methodology for the chromatographic comparisons and alignments of several analytical datasets for screening Viperid venoms, which can broaden the analytical toolbox for toxicovenomics research.We did not perform any MS/MS (MS2) experiments or bottom-up proteomics experiments and only acquired toxin masses next to toxin bioactivities.Adding proteomics data to our study would deviate from our focus and would make this study way too lengthy.As such, it is not possible to correlate our data with previously published proteomics studies.Nevertheless, including proteomics data in future studies would be a valuable advancement of the here-presented research, in which the proteomics data allows for actual (bioactive) toxin identification, i.e., bottom-up, top-down, Edman degradation, or a combination of these [46].As for the coagulation bioactivity data, the venoms of most species, except B. jararaca, were found to have procoagulant and/or anticoagulant activity.While anticoagulant effects were observed in all six remaining venoms, clear procoagulant effects were only observed in the nanofractionated venoms of B. asper, C. atrox, and D. russelii.Given that metalloproteinase and serine protease enzymes have been demonstrated to have thrombin-like activity, platelet aggregation activity, and coagulation factor activating activities and are typically found in viper venoms [10,13,47], it was anticipated that procoagulant toxins would be observed in all the samples tested.However, for the venoms of E. carinatus and E. ocellatus, only anticoagulant effects were seen at our tested venom concentration of 1 mg/mL.Our previous studies investigating the bioactivity of these two venoms, such as the study by Xie et al. [34], previously detected procoagulant effects, though higher venom concentrations of up to 5 mg/mL were used.It seems likely that the lower venom concentrations applied here, coupled with these larger proteins at least partially denaturing during venom separation by RP-LC, are responsible for the disparities in our results.The presented analytical approach has a number of current limitations.The methodology gives non-quantitative toxin data.The MS data are only valuable for assessing toxin accurate masses and cannot be used for quantitative purposes since toxin sensitivity is different for each toxin and is highly dependent on many factors, among others a toxin's ionization efficiency (which, in turn, is dependent on its structure).Based on LC-UV chromatogram peak areas, we, however, can make a crude overview of relative toxin abundances at the level of low, medium, or high abundance in the venom analyzed.The analytical methodology was narrowed down to a limited number of toxin masses to be correlated to the bioactivity peaks observed.However, for chromatographic separation, insufficient separation of toxins in several cases resulted in close co-elution, thereby making it hard to pinpoint what exact toxin caused a bioactivity.Also, toxin denaturation can happen during separation.For the ESI-MS data, when using a standard ESI-Q-TOF-MS coupled to RP-HPLC separations, charge-state distributions change to higher numbers upon increasing toxin masses.Among other effects, this results in sensitivity decreases due to toxin mass increases.For the analysis of MALDI-MS, limitations include the fact that mass resolution was limited to 10 Da, and in cases where no two charge stages were found for a toxin, it was difficult to deconvolute the m/z values found from the MS data.Producing the plotted MALDI-MS chromatograms is a manual, time-consuming, and laborious task.Future progress in this regard can include scripts to make it easier to process this data at a higher throughput.Although the focus was on the toxins detected with good signal abundance, not many higher-mass-range toxins were found to have a good signal abundance.Some were found in the MALDI-MS data, but with very low abundance.The Supplementary Materials PowerPoint document "Superimposed MALDI-MS spectra of all toxin fractions analyzed for each venom.pptx"gives the MALDI-MS spectral data of all the fractionated toxins analyzed.

Conclusions
In this study, we provided a chromatographic comparison of analytical results by investigating data obtained from several viper venoms analyzed by both ESI-MS and MALDI-MS approaches.In addition, post-column plasma coagulation screening was included in this study to assess bioassay integration.To finalize, the present work gave chromatographic analytical alignments that showed promising toxin identification results for viper venom analysis by combining both MS methods, which enriched the analytical toolbox for snake toxicovenomics research.

Venoms
The venoms of Bitis arietans (captive bred, Nigeria), Bothrops jararaca (captive bred, Brazil), Bothrops asper (Central American lancehead, Costa Rica), Crotalus atrox (Western diamondback rattlesnake, USA), Daboia russelii (Russell's viper, Sri Lanka), Echis carinatus (Indian saw-scaled viper, India), and Echis ocellatus (West African carpet viper, Nigeria) were stored as freeze-dried powders at 4 • C until use.These lyophilized venoms were provided by the Centre for Snakebite Research and Interventions at the Liverpool School of Tropical Medicine, UK, either from wild-caught snakes maintained in the herpetarium or from the historical venom stocks held in this facility.Crude venom solutions (1 ± 0.1 mg/mL) were prepared in MQ water as stock solution aliquots and then stored at −80 • C until use.

Nanofractionation Analytics
Liquid chromatography separation, nanofractionation, and parallel MS analysis were performed by a Shimadzu HPLC system.A Shimadzu SIL-20AC autosampler was set at a 50 µL injection volume.The Shimadzu LC-20AB pump was set at a total flow rate of 0.5 mL/min.The Shimadzu SPD-20A UV detector was set at 220 nm for UV detection.The samples were separated on a Waters XBridge Peptide BEH C18 column (4.6 × 100 mm; 5 µm; 300 Å) with an XBridge Peptide BEH C18 Guard Column (130 Å, 5 µm, 10 mm × 10 mm).The separation was carried out at 30 • C in a Shimadzu CTO-10AC column oven.Mobile phase A was composed of 98% H 2 O, 2% ACN, and 0.1% TFA, while mobile phase B was 98% ACN, 2% H 2 O, and 0.1% TFA.The gradient started at 0% mobile phase B, after which the mobile phase B concentration increased to 13% in the first 5 min and then was linearly increased to 50% within a time period of 30 min, followed by a linear increase from 50% to 90% B within 3 min and then a 7 min isocratic elution at 90% B. Next, mobile phase B was decreased linearly to 0% in 1 min, and the column was then equilibrated for 5 min at 0% B. All settings of the system were controlled by Shimadzu Lab Solutions software 1.0.After separation and UV detection, the elute was split into a 1:9 ratio, of which the 90% fraction was transferred to a FractioMate TM nanofractionator (VU, Amsterdam, The Netherlands) controlled by FractioMator software (V1.0,VU, Amsterdam, The Netherlands).Fractions were collected at a resolution of 6 s/well onto transparent 384-well plates (F-bottom, rounded square well, polystyrene, without lid, clear, non-sterile; Greiner Bio One, Alphen aan den Rijn, The Netherlands).After fragmentation, the well plates were frozen and then vacuum-centrifuged to dryness overnight using an RVC 2-33 CD plus maxi concentrator (Salm en Kipp) with a rotation speed of 1500 rpm, pressure of 1.0 mbar, and temperature of 30 • C.After solvent evaporation, the dried plates were stored at −80 • C prior to further analysis.The smaller fraction was sent to UV detection, which was followed by mass spectrometry detection (or the effluent was sent to the waste after UV detection).Each nanofractionation analytics workflow procedure for analyzing each venom was performed in duplicate.

MALDI-MS Measurements
First, 10 µL MQ was added to the wells of the well plates with fractionated and dried venom toxins.In order to dissolve the venom toxins in MQ, the plates were gently stirred on an automatic stirrer for 10 min at room temperature.Based on the LC-UV data and the retention times of the UV detectable peaks, the wells of interest (i.e., those wells that contained collected fractions that showed UV absorbance upon elution) were selected for MALDI-MS.
For this, 1 µL portions were taken from these wells and deposited on a 384-spot MALDI plate (MTP 384 target ground steel BC, Bruker, Germany).Then, 1 µL of the matrix solution was added to the deposited 1 µL portion fractions, followed by manual aspiration and dispensing for three times for mixing purposes.For the matrix solution, 10 mg/mL of α-Cyano-4-hydroxycinnamic acid (CHCA) was used in a TA50 solution of 50% ACN and 50% MQ and 0.1% TFA.An analysis of the prepared ground steel MALDI plate was performed using an ultrafleXtreme MALDI-TOF/TOF mass spectrometer (Bruker, Germany).Prior to analysis, the MS was calibrated using a peptide mixture containing bradykinin (1-7), angiotensin I, angiotensin II, substance p, bombesin, renin substrate, ACTH 1-17, ACTH 18-39, and somatostatin in a m/z range from 5000 to 55,000.The MS parameters set were as follows: source 1, 20 kV; source 2, 18.9 kV; lens voltage, 6.5 kV; laser repetition rate, 2 kHz.A total of 15,000 shots for each sample were accumulated, with the laser power gradually increasing from 60 to 100%.FlexControl (Bruker, Germany) was used for instrument control and raw data collection.FlexAnalysis 3.4 (Bruker, Germany) and Microsoft Excel were used to process the data.The ions found in the MALDI-MS spectra were used to construct chromatogram-like graphs by plotting each ion found with significantly high intensity for each 6 s fraction, i.e., its S/N value (on the y-axis), versus the fraction's retention time (on the x-axis).As the toxins were all eluted in a subsequent series of adjacent fractions, the chromatographic representations for each toxin found from the MALDI-MS data were the output.These chromatographic representations were compared with the EICs from the ESI-MS analyses.MALDI-MS usually yields singly and doubly positively charged ions.If only one ion was found for a toxin in MALDI-MS, the parallel acquired ESI-MS data were investigated to compare data, and from there the charge state of the ion found in MALDI-MS was assigned.

Online (LC-)ESI-MS Measurements
After the flow split, the 10% split of the LC elute was sent to MS detection (Impact II QTOF, Bruker Daltonics, Germany) with electrospray ionization (ESI) in positive-ion mode at a mass range between 800 and 5000 m/z.Bruker Compass software 3.4 was used for instrument control and data analysis.For MS data processing, the total-ion current (TIC) was plotted from the recorded MS data.Extracted-ion currents (EICs) were extracted from the TIC by plotting the EIC of the most abundant charge state of each toxin.The ESI source parameters were the capillary voltage of 3.5 kV, source temperature of 200 • C, nebulizer at 0.8 Bar, and dry gas flow of 6.0 L/min.In-source collision-induced dissociation (CID) was set at 200 eV, and 1 average spectrum was stored per second.
After the deconvolution of the masses of the toxins from the MS data, for all the venom toxins analyzed for masses falling within the tentative mass range of PLA 2 toxins, from Uniprot, the amino acid sequences corresponding to PLA 2 s in the species analyzed were taken from which exact masses were recalculated, taking the cysteine bridges in PLA 2 s into account with the help of Chemdraw.Next, the accurate masses acquired from the MS data could be compared with the calculated candidate exact masses from the Uniprot information.

Delay Time Calculation between LC-UV, LC-MS, and LC-Nano Fractionation
A coagulation bioassay was performed to calculate the delay between the UV data, the MS data, and the fractions collected by the FractioMate.The anticoagulant argatroban was injected, separated by the LC system, and fractionated in a 384-well plate following vacuum centrifugation to dryness.This was followed by a coagulation assay (see the next section below) on the well plate.Where argatroban was eluted, a decreased signal was absorbed in the coagulation assay.By correlating the retention times of the argatroban peak in LC-UV, LC-MS, and the LC-bioassay, the delay times could be determined and then corrected for data analysis.The concentration of argatroban injected was 1 mg/mL with an injection volume of 20 µL.A typical result from the argatroban control analyses is given in the Supplementary Materials "Argatroban (anticoagulant compound)" analyzed by nanofractionation analytics.pptx.

Coagulation Bioassay
For bioassaying, an aliquoted tube with plasma was first thawed in a warm water bath to room temperature.The bioassay protocol was performed based on previous research [23,51,52].A CaCl 2 solution (20 mM, 20 µL per well) was pipetted onto a vacuumcentrifuge-dried 384-well plate with nanofractionated toxins using a Multidrop 384 Reagent Dispenser (Thermo Fisher Scientific, Ermelo, The Netherlands).After 3 min, plasma (20 µL per well) was also pipetted into the wells.The plate was then measured with a Varioskan TM Flash Multimode Reader (Thermo Fisher Scientific, Ermelo, The Netherlands) for kinetic absorbance measurement at 595 nm at 25 • C over 80 min.The resulting coagulation activity data measured in each well were processed into two coagulation activities, which were procoagulation (0-15 min slope) and anticoagulation (80 min endpoint).Procoagulant or anticoagulant responses were plotted on the y-axis versus the retention time of fractionation on the x-axis, resulting in two coagulation chromatograms.Positive peaks in the procoagulant chromatograms resulted from procoagulant toxins, and negative peaks in the anticoagulation chromatograms represented anticoagulant toxins.

Toxins 2024 , 24 Figure 1 .
Figure 1.MALDI-MS spectrum obtained for a nanofractionated toxin fraction.(A) shows a typical MALDI-MS spectrum obtained for nanofractionated toxins of D. russelii venom at the retention time fraction of 20.0 min.The toxin ions with the mass of 13,644 Da presented themselves as singly (M + H) + and doubly charged (M + 2H) 2+ ions (which were used for toxin mass deconvolution) with m/z values of 13,644 and 6822 based on manual deconvolution.(B) shows zoomed-in toxin-ion peaks with a m/z of 40,859.

Figure 1 . 24 Figure 2 .
Figure 1.MALDI-MS spectrum obtained for a nanofractionated toxin fraction.(A) shows a typical MALDI-MS spectrum obtained for nanofractionated toxins of D. russelii venom at the retention time fraction of 20.0 min.The toxin ions with the mass of 13,644 Da presented themselves as singly (M + H) + and doubly charged (M + 2H) 2+ ions (which were used for toxin mass deconvolution) with m/z values of 13,644 and 6822 based on manual deconvolution.(B) shows zoomed-in toxin-ion peaks with a m/z of 40,859.Toxins 2024, 16, x FOR PEER REVIEW 10 of 24

Figure 3 .
Figure 3. Integrated Bothrops jararaca venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 3 .
Figure 3. Integrated Bothrops jararaca venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 4 .
Figure 4. Integrated Bothrops asper venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 4 .
Figure 4. Integrated Bothrops asper venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 5 .
Figure 5. Integrated Crotalus atrox venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 5 .
Figure 5. Integrated Crotalus atrox venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 6 .
Figure 6.Integrated Daboia russelii venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 6 .
Figure 6.Integrated Daboia russelii venom results: superimposed results of LC-UV data, LC-(ESI)-MS data, and nanofractionation MALDI-MS data plotted chromatographically and nanofractionation coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 7 .
Figure 7. Integrated Echis carinatus venom results: superimposed results of LC-UV, LC-(ESI)-MS, and nanofractionated MALDI-MS data plotted chromatographically and nanofractionated coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.

Figure 7 . 24 Figure 8 .
Figure 7. Integrated Echis carinatus venom results: superimposed results of LC-UV, LC-(ESI)-MS, and nanofractionated MALDI-MS data plotted chromatographically and nanofractionated coagulation bioassay data plotted chromatographically. (A) LC-UV trace of separated snake venom at 220 nm (orange) and 254 nm (blue).(B) Coagulation bioactivity chromatograms representing anticoagulation (upper trace) and procoagulation activity (lower trace).(C) Chromatographically plotted MALDI-MS data of the identified toxins in the wells with nanofractionated toxins.For each toxin identified in different wells, the measured intensity from the MALDI data was plotted on the y-axis versus the retention time of fractionation on the x-axis.As all toxins were eluted over a series of subsequent wells, so-called MALDI-MS chromatograms of each toxin were the result.(D) Extracted Ion Currents (EICs) from the LC-ESI-MS data.Toxins 2024, 16, x FOR PEER REVIEW 16 of 24

Table 1 .
Comparison , the S/N signal extracted from the Bruker software 3.4) of the mass found in LC-MALDI-MS, and the percentage of the number of overlapped masses in both LC-ESI-MS and LC-MALDI-MS.

Table 2 .
Mass difference acceptance when analyzed with either only MALDI-MS or both ESI-MS and MALDI-MS.