Dynamic Changes in Rat Mesenteric Lymph Proteins Following Trauma Using Label-Free Mass Spectrometry

Abstract

Early events triggered by post-trauma/hemorrhagic shock currently represent a leading cause of morbidity and mortality in these patients. The causative agents of these events have been associated with increased neutrophil priming secondary to shock-dependent alterations of mesenteric lymph. Previous studies have suggested that unknown soluble components of the post-shock mesenteric lymph are main drivers of these events.

In the present study, we applied a label free proteomics approach to further delve into the early proteome changes of the mesenteric lymph in response to hemorrhagic shock. Time-course analyses were performed by sampling the lymph every thirty minutes post-shock up until 3h (the time window within which a climax in neutrophil priming was observed).

There are novel, transient early post-hemorrhagic shock alterations to the proteome and previously undocumented post-shock protein alterations. These results underlie the triggering of coagulation and pro-inflammatory responses secondary to trauma/hemorrhagic shock, metabolic deregulation and apoptosis, and alterations to proteases/anti-proteases homeostasis, which are suggestive of the potential implication of extracellular matrix proteases in priming neutrophil activation.

Finally, there is a likely correlation between early PSML post-shock neutrophil priming and proteomics changes, above all protease/anti-proteases impaired homeostasis (especially of serine proteases and metalloproteases).

Introduction

In the 1960’s, the development of modern resuscitation techniques led to improved immediate survival. However, late death increased as a result of the emergence of organ failure (e.g. liver, lungs, kidneys and heart). The phenomenon of organ failure was thus documented as a sequela to shock and infection (1). Despite decades of improvements, organ failure remained one leading cause of post-traumatic death after the first 24 h following injury (2). While the molecular events driving multiple organ failure (MOF) have not been completely elucidated, its etiogenesis has been progressively associated to the occurrence of intestinal ischemic/reperfusion injury, secondary to splanchnic vasoconstriction associated with hemorrhagic shock (HS) followed by fluid resuscitation (3–5). This in turn promotes the failure of the intestinal barrier, bacterial translocation, systemic sepsis and neutrophil (PMN) priming, thus resulting in the progression to organ failure (3–6).

Over the last decade, this concept has been further expanded (4–6) to indicate lymphatic fluid derived from the compromised intestine (mesenteric lymph (ML)) as a contributor to the development of early post-shock events related to increased likelihood of mortality in trauma patients. Experimental work has further suggested that post-shock mesenteric lymph (PSML) could serve as a conduit by which causative agents, that are contained in exudates from these stressed splanchnic beds, are conveyed to the systemic circulation (7). In animal models, the diversion of mesenteric lymphatics, albeit not portal blood, prior to trauma/HS (T/HS) attenuates post-shock neutrophil priming which decrease potential PMN cytotoxicity adherence, and pulmonary sequestration, as well as endothelial adhesion molecule expression and remote organ injury (3–6). Unlike portal blood, the ML is able to carry potentially “toxic” factors from the intestine that avoid any hepatic first-pass modification or detoxification, and are instead delivered directly to distant organs (especially the heart and lungs) (3–6,8,9).

In the light of this accumulating body of evidence, causative molecular contributors to early post-shock reactions and organ failure have been searched for in the PSML, by targeting the lipid fraction (8), proteins and lipoproteins (9–20). Accumulating evidence suggested a synergistic interplay of these classes of bio-molecules in mediating early post-shock reactions and organ failure (14). As part of this evidence, altered concentrations of proteins, cholesterol, triglycerides, and high-density lipoprotein (HDL) are observed in PSML (8). PSML is characterized by gelsolin depletion (18,19), which is relevant in the light of the role of this protein as an actin scavenger and lipid binding protein. Additionally, Dayal et al. (20) have demonstrated cytotoxicity in the aqueous fraction of PSML, suggesting a role for proteins in mediating inflammatory responses driving organ failure. Such observations paved the way for a series of studies on the PSML proteome in animal models (14–17) and humans (12), with the goal to characterize protein components of the lymph that may provide key insights into post-shock pathophysiology.

Most of the proteomics investigations reported in the literature rely upon the application of two dimensional gel electrophoresis (2DE) and differential in-gel electrophoresis (DIGE) approaches, followed by confirmatory Western Blots, to investigate the alteration to the PSML proteome following HS in a well-established animal model of T/HS in Sprague Dawley rats (14–16). Briefly, shock is induced in male Sprague-Dawley rats by controlled hemorrhage, and the mesenteric duct is cannulated for lymph collection after 3h from HS (14–16). Findings reported by Peltz et al. (14), Zurawel et al. (15), and Fang et al. (16) are consistent in that decreased relative concentrations were observed in PSML in the levels of several components of coagulation cascades, protective protease inhibitors, serum albumin, complement C3 precursor and haptoglobin. On the other hand, products of hemolysis, glycolytic enzymes, major urinary protein and lipid carriers were found to be increased in PSML (14–16).

A different animal model involving Wistar rats was exploited by Mittal and colleagues (17) to investigate the major changes to the mesenteric lymph proteome in response to HS via 2D-HPLC separation coupled with isobaric tag for relative and absolute quantitation (iTRAQ), thus exploiting peptide labeling for relative quantification purposes. This workflow enabled the detection of overall 245 ML proteins, out of which sixty had a significant increase in their relative abundance in the HS group. A bioinformatics elaboration of these results highlighted a shock-dependent alteration of ML proteins associated to injury and metabolic responses (17). Recent proteomics investigations on the human ML (12) indicated that the human and rat ML proteomes under control conditions are surprisingly comparable, suggesting that any insight gained from the analysis of animal models could be further translated in human samples as well.

While early proteomics studies on animal models pointed to the main proteome changes to HS (pre- vs post-shock ML proteomes), to the best of our knowledge no proteomics investigation has addressed dynamic changes in response to HS so far. Here we applied a label free proteomics approach to further delve into the ML proteome changes in response to HS, while performing a time-course analysis by assaying post-HS samples every thirty minutes post-shock up until 3h (the time window within which a climax in neutrophil priming was observed). As a result, we report transient early post-HS alterations to the proteome and previously undocumented post-HS protein increases.

Materials and Methods

An extended version of this section can be found in the Supplementary File – Materials and Methods extended. The Institutional Animal Care and Use Committee approved all experiments performed on animals.

Hemorrhagic shock

Controlled hemorrhage and lymph collection were performed on six male Sprague-Dawley rats through long-established procedures (14,15), as detailed in the Supplementary File – Materials and Methods extended. After 1 hr of lymph collection, HS was induced by controlled blood loss to maintain a mean arterial pressure (MAP) of 30mmHg and sustained for 40min. Resuscitation was performed by infusing 2× shed blood volume (SBV) in normal saline (NS) over 30min, followed by ½ SBV returned over 30min, then completed with 2× SBV in NS over 60min.

Lymph Sample Preparation and Protein Digestion

Lymph samples were collected from 6 rats both pre-shock and every 30 minutes post-shock up until 3h. Upon determination of protein concentration via Bicinchoninic acid assay, samples were reduced and alkylated addition of 10 mM dithiothreitol and 55 mM iodacetamide, respectively. Trypsin digestion was performed overnight in 25 mM ammonium bicarbonate (Promega, Madison, WI).

Nano-HPLC MS/MS analyses were performed via an Agilent 1200 (Palo Alto, CA) and FT-ICR LIT (ThermoFisher LTQ-FT Ultra). Approximately 1.0 μg of each sample was loaded onto a trapping column (ZORBAX 300SB-C18, dimensions 5×0.3 mm 5 μm; Agilent Tachnologies, Santa Clara, CA) before injection and separation throughout an 85 minute gradient. Separation was achieved via a 15 cm long Synergy C18 column 4.0 μm particle size, operated at 40°C (Penomenex; Torrance, CA). MS/MS peak lists were searched against databases using an in-house Mascot™ server (Version 2.2, Matrix Science) or Peptide Prophet. For searches mass tolerances were +/− 15 ppm for MS peaks, and +/− 0.6 Da for MS/MS fragment ions.

Quantitative data analysis was performed either via Progenesis LC-MS v 2.0 (Nonlinear Dynamics, Durham, NC) or Scaffold (v4.3.2., Proteome Software) (21). Changes greater than 1.5 fold in protein abundance and p-value < 0.05 (ANOVA) were considered statistically significant. Further statistical elaborations were performed through the software GENE-E (v. 3.0.200 – Broad Institute, Inc.) to plot heat maps and perform hierarchical clustering analyses (one minus Pearson correlation), or the Excel macro Multibase2014 for Principal Component Analyses.

Functional annotation for biological functions and cell compartments were performed either with Scaffold or David v. 6.7 (David Bioinformatics services).

Metalloproteinase activity assay

Gelatin zymography of metalloproteinase activity was performed as described by Kleiner and Stetler-Stevenson, as detailed in the Supplementary File – Materials and Methods extended (22).

Isolation of neutrophils and oxygen consumption

Neutrophil separation and oxygen consumption assays were performed as previously reported (23).

Results and Discussions

Neutrophil priming

PSML is responsible for PMNl priming (24). In the present study, PMN priming assays highlighted a post-HS time-dependent increase in neutrophil priming upon incubation with the hydrophilic (protein) fraction of the PSML. PMN priming reached a climax within 2 or 3h after HS (Post-6) (Figure 1), suggesting a correlation within dynamic increases in the levels of positive protein regulators of priming, or decreased levels of negative ones.

Figure 1.

Neutrophil priming as determined by assaying the maximum oxygen consumption rate in post-shock samples at the indicated time points, as in Johnson et al. (23). A climax was reached in between the second and the third hour following trauma/hemorrhagic shock.

Time course analyses

Six individual rats were used for lymph collection in the pre- and post-shock states and mesenteric lymph was assayed by label free quantitative liquid chromatography coupled online with tandem mass spectrometry (LC-MS/MS) either pre- (Pre) or 3h post-HS (Post-6). Intermediate time point assays were performed in a separate experiment at 30 min (Post-1), 1h and 30min (Post-3), 2h (Post-4), 2h and 30min (Post-5) from T/HS. Overall, we detected 284 proteins, out of which 37 proteins showed a statistically significant (p<0.05) quantitative variation (fold-change > 1.5) (Supplementary Table 1 – highlighted in green). Appreciable fold-change increases (>1.5) were observed for a further 238 proteins (Supplementary Table 1 – highlighted in orange), although statistical significance was not reached for these protein changes, owing to the limited set of biological replicates assayed in the present study. Quantitative trends for the observed dynamic changes are reported in the form of heat maps in Supplementary Figure 2. Following hierarchical clustering analyses (1-Pearson correlation), the list of observed 284 proteins was divided into four main clusters in Figure 2, including:

1. proteins showing a trend towards increase secondary to HS (the great majority, in agreement with previous reports (13–17) – Figure 2.A);

2. proteins showing transient decrease immediately after HS (Post-1 30 min time-point – Figure 2.B – upper and middle cluster, respectively);

3. proteins showing late alterations, such as increases in Post-1 or Post-2 (30 min or 1h after HS – Figure 2.C – upper clusters);

4. Proteins following a trend towards decrease in response to T/HS (Figure 2.C – lower cluster).

Figure 2.

Hierarchical clustering analyses of quantitative dynamic changes of pre and post-shock mesenteric lymph proteins, either increasing (A), transiently decreasing after 30 minutes from trauma/hemorrhagic shock (Post-1 – B), increasing after 30 minutes or 1h and half (Post-3) from trauma/hemorrhagic shock (C, upper and middle cluster, respectively) or decreasing progressively in the assayed samples (C, lower cluster). Results from Progenesis LC-MS have been exported into Mascot Daemon for protein identification and exported in GENE-E for statistical analysis and clustering. Samples were collected from four animals either pre-shock (Pre), or every thirty minutes post-trauma/hemorrhagic shock (Post-1 to Post-6, except for the 1h time point). Values increase progressively from blue to red in the color scale, as indicated in the legends on top of each cluster. Proteins are enlisted with their relative UniProt ID names, though the _RAT taxonomy specification for Rattus norvegicus has been deleted as to improve the clarity of the figure. Further details are reported in Supplementary Table 1 (protein list) and 2 (GO term enrichment).

Such a workflow allowed us to monitor early dynamic changes following HS, which is relevant in that trauma is classified as occurring in two separate phases, also referred to as the “ebb” phase and the “flow” phase (24). While the “flow” phase occurs later after compensation of the state of trauma-dependent HS, the “ebb” phase is transient (24), in that it is initiated within minutes after trauma and persists for several hours after the initial insult in humans. The main features of this phase are characterized by a decline in body temperature and metabolic depression, accompanied by reduced oxygen consumption, a biological strategy that is aimed at reducing post-traumatic energy depletion (24). However, it is now recognized that the kinetics of responses to trauma may not be schematized as two independent entities such as “ebb” and “flow”. Time course events following T/HS might be rather characterized by a complex interplay of altered factors, which contribute to profound inflammatory and immunologic dysfunction. These events, especially those arising early after T/HS, result in the activation of the innate immune system, the complement system, and the release of pro-inflammatory cytokines (24).

Despite the consolidated relevance of post-T/HS kinetics, proteomics studies have so far addressed only the main changes between pre- vs post-HS (3h). To the best of the Authors’ knowledge, no study has been published so far documenting dynamic alterations to the proteome during the earliest stages after T/HS.

While the great majority of proteins followed a trend towards increase (208 out of 284 – Figure 2.A), thirty four proteins decreased transiently thirty minutes after T/HS (Figure 2.B). Sixteen proteins were found to be increased in abundance and then decrease again soon after 30 min from T/HS (Post-1 – Figure 2.C – upper cluster). Eleven proteins clustered together as they were transiently increasing in Post-3 samples (1h and 30 min after HS – Figure 2.C, middle cluster). General trends towards progressive decrease were observed for sixteen proteins as well (Figure 2.C – lower cluster).

Proteins involved in coagulation responses

The main biomedical endpoints in the management of shock imply the restoration of adequate tissue perfusion and oxygen balance, and the compensation of T/HS-dependent compromised cellular metabolism (24). It is largely expected that the immediate biological reactions to HS would be represented by the activation of coagulation cascades. In this view, depletion of coagulation factors is an expected feature in post-HS, in that the activation of the clotting cascade would promote proteolysis. In parallel, resuscitation strategies (in the present study, 2xSB in NS) would result in the dilution of coagulation factor components (25). Consistently, transient decreases of proteins involved in coagulation cascades was observed in Post-1 samples (Figure 2.B), including heparin cofactor 2, T-kininogen 1, and Plasma kallikrein. Fibrinogen exists as a heterohexamer, assembled through disulfide bonds and composed of 2 sets of 3 non-identical chains: alpha, beta, and gamma. All three chains were found to decrease in Post-1 samples (Figure 2.B).

Analogously, progressive decreases in the levels of coagulation factor X (FA10 – Figure 2.C – lower cluster) were observed in response to T/HS. Our findings are consistent with previous 2DE (16) and DIGE-based (15) reports on the T/HS-dependent alterations to the rat mesenteric lymph proteome have indicated a post-shock decrease in the levels of these proteins.

Kallikrein deregulation (Figure 2.C – Supplementary Table 2) is relevant within the framework of clotting cascades, in that this protein is involved in coagulation events (factor XII activation). Its post-HS depletion is consistent with previous reports indicating a T/HS-triggered depletion of proteases involved in coagulation cascades (15).

However, post-HS increases in coagulation factor IX have been reported as well by Mittal et al. (17), and might stem from the accumulation of activated factor IX fragments. Consistently, we observed increases in the levels of coagulation factor IX, together with progressive accumulation of coagulation factor XII (Figure 2.A). Such increases might correlate with the increased abundance of other proteins that promote wound healing responses (GO:0042060 – platelet factor 4, prothrombin and fibronectin), or either act as negative regulators of coagulation (gelsolin, vitamin K-dependent protein S (Pros1), plasma protease C1 inhibitor (SerpinG1) – Supplementary Tables 1 and 2).

Accumulation of Pros1 (Figure 2.C – upper cluster) and decreases in plasminogen (PLMN – Figure 2.C, middle cluster) are suggestive of a fine tuning of coagulation cascades. Both Pros1 and PLMN are anticoagulants that either act as cofactors to activated protein C in the degradation of coagulation factors Va and VIIIa, or contribute to attenuating coagulation and stimulating fibrinolysis, respectively.

Accumulation of gelsolin (Figure 2.A, Supplementary Table 1) is an interesting, albeit unexpected finding, in that gelsolin is known to decrease in mesenteric lymph after 3h from shock (24,25). Gelsolin is a plasma protein that participates in the depolymerization of actin filaments thereby preventing capillary plug formation following tissue injury. It also binds proinflammatory lipids such as lysophosphatidic acid, sphingosine 1-phosphate and phosphoinositides, thus mediating inflammatory responses (19). It is worthwhile to stress that previous papers indicating a post-HS decrease in the levels of gelsolin only relied upon gel-based approaches (either 2DE, DIGE and WB (19)), while HPLC-MS/MS based assays (either coupled to iTRAQ quantitation (19) or label free approaches, such as in the present study) tend to highlight a post-HS increase of this protein. While bearing in mind possible platform-dependent biases on gelsolin quantitation, from a biological standpoint gelsolin increase might underlie systemic responses to trauma implying increases in anti-inflammatory mediators.

Proteases/Anti-protease balance

Transient responses were characterized by increases in the levels of specific proteases (such as trypsin – Figure 2.C) and decreases in a group of proteins involved in enzyme inhibitory activities, such as heparin cofactor 2 (SerpinD1), murinoglobulin-1, angiotensinogen, T-kininogen 1 and 2, complement components C3 and C4, serine protease inhibitor A3K and A3M (SerpinA3K and SerpinA3M), alpha-1-inhibitor 3, Fetuin-B, alpha-1-microglobulin (Figure 2.B – Supplementary Table 2: GO:0004866). These proteins (especially Serpins) play a key role in inhibiting the activity of serine proteases (such as thrombin) (36). Transient down-regulation of the proteins listed above might underlie a temporary deregulation of the protease/anti-protease poise in order to promote the former, and thus clotting cascades.

These progressive decreases in the levels of anti-proteases were counteracted by time-dependent increases in other proteins of the same functional group (GO:0004857 – enzyme inhibitor activity – Figure 2.A, Supplementary Table 1 and 2). Time-dependent increases were observed for several serine-protease inhibitors (Serpins), including: Alpha-1-antiproteinase (SerpinA1), Serine protease inhibitor A3N (SerpinA3), phosphatidylethanolamine-binding protein 1, serine protease inhibitor Kazal-type 3, Serine protease inhibitor A3L (SerpinA3L), corticosteroid-binding globulin (SerpinA6), plasma protease C1 inhibitor (SerpinG1).

Other proteases inhibitors were found to increase as well, including general anti-proteases (alpha-2-macroglobulin), cysteine protease inhibitors (cystatin-B and C), and lipocalins (major urinary protein (MUP). Consistently, decreased levels of proteases or proteolysis-related proteins (GO:0006508) were observed as well in post-HS time-dependent fashion, including complement components (factor D and I), C-reactive protein, haptoglobin, beta-2-glycoprotein 1, plasminogen coagulation factor X, and ubiquitin-conjugating enzyme E2 N.

Hypercoagulability after major trauma is a major source of morbidity and mortality, the onset of which is most prevalent within the first 24h from injury and influenced by gender (more likely in women (27)). The present evidence suggests that the dynamic accumulation of a wide series of protease inhibitors, especially Serpins, might promote an unbalanced ratio of proteases/anti-proteases (decreasing/increasing or increasing/decreasing as pointed out above, respectively). This would in turn affect coagulation cascades at different stages post-HS/trauma. Trauma-dependent alterations in the levels of MUP, a pheromone-carrying protein of the lipocalin family, have been previously associated with gender-dependent responses to HS (14). MUP alterations might also be tied to the lipocalin system-dependent priming of neutrophil activation, which would in turn promote an increase in inflammatory responses (16). Moreover, increases in SerpinG1 could result in the promotion of complement activation, fibrinolysis and the generation of kinins, while preventing hypercoagulation (28).

Acute phase proteins and immune responses

From a proteomics standpoint, early responses to T/HS were characterized by transient increases in the levels of proteins involved in complement activation (C4b-binding protein alpha chain and complement component C9 – Figure 2.C – upper cluster), and progressive increases of others (Complement C1qA, C1qB and C1s subcomponent, – Figure 2.A). Activation of the complement cascade plays a key role in promoting adverse immune consequences in response trauma/hemorrhagic shock and resuscitation (26). However, additional components of the inflammatory responses (C-reactive protein) or complement activation system (KEGG pathway: rno04610, including Complement C3, C4 and C6) were either progressively or transiently decreased soon after HS, to be then increased in abundance again to pre-shock values, or even higher in a time-dependent fashion following T/HS (Figure 2.B–C, Supplementary Table 1 and 2).

At the same time, negative regulators of the complement system increased progressively upon T/HS (Plasma protease C1 inhibitor, – Figure 2.A). These observations testify the complexity of the complement/coagulation system. It is worth noting that such post-shock dynamic changes have been so far under-investigated since the relative abundances of most complement components did not show significant fold/change variations in pre- vs post-shock samples (17). Transient downregulation of proteins involved in complement cascades and acute phase responses were also accompanied by transient depressions in the levels of immunoglobulins – Ig gamma-2A chain C region and Ig gamma-1 chain C region, or immunoglobulin domain-containing protein Class I histocompatibility antigen Non-RT1.A alpha-1 chain (Figure 2.B – IPR003597: Immunoglobulin C1-set – Supplementary Table 2).

Similar time-dependent increases in PMN priming (Figure 1) were accompanied by the progressive accumulation of a wide series of proteins more or less directly involved in acute inflammatory responses (GO:0002526 – Figure 2.A, Supplementary Table 2). In this group we could detect increases in the levels of lysozyme C-1 and alpha-1-acid glycoprotein, an acute-phase response protein synthesized in the liver under hormonal and cytokine control, whose expression increases up to 50-fold upon inflammation (29). Analogously, increases in platelet factor 4 were observed, a protein that might promote chemotaxis for neutrophils and monocytes (29). Increases in macrophage migration inhibitory factor were observed as well, a pro-inflammatory cytokine involved in the innate immune response to bacterial pathogens that regulates the function of macrophages in host defense (29). Progressive accumulation of beta-2-microglobulin, a MHC I component (Figure 2.A), might indirectly support an increased antigen presentation capacity to the immune system.

The extracellular matrix and metalloproteinase 2

In the present study, we observed a progressive increase in the levels of certain extracellular matrix-related components. These proteins lie at the crossroad between clotting cascades and immune system activation. Their progressive increase in response to HS was consistent with the trends observed for neutrophil priming responses (Figure 1–3). Proteins involved in this group included 72 kDa type IV collagenase (matrix metalloproteinase 2 – MMP2). MMP2 (Figure 3.A), in particular, MMP2 is involved in diverse pathways, such as remodeling of the vasculature, angiogenesis, tissue repair and responses to neutrophil activation (30). Quantitative analyses revealed a clear pattern towards increase for each detected peptide of this protein (Figure 3.B). Zymography of time-course samples (Post-HS time-points vs Pre-shock) highlighted a widespread increase in the activity of MMPs, especially the 62kDa MMP2 (22,31), as a general response to HS (Figure 4).

Figure 3.

In A, sequence coverage of the 72kDa collagenase protein, matrix metalloproteinase 2 (MMP2). Relative quantitative trend for each detected peptide are shown as heat maps in B, together with Z-score normalized graphs of the quantitative trends and a detail of the 3D plot from the Progenesis LC-MS report for the peptide identified as FPFLFNGR.

Figure 4.

Zymography of matrix metalloproteinase (MMP) activities as determined by assaying mesenteric lymph samples either pre-shock, or at 1h, 2h or 3h from trauma/hemorrhagic shock. Activities of several MMPs were increased in response to trauma/hemorrhagic shock, including MMP2, MMP7, MMP9 and MMP13.

Of note, several extra-cellular matrix protein substrates to MMPs were found to increase in a time-course fashion following T/HS, including (GO:0005539) glycosaminoglycan binding proteins of the extracellular matrix fraction (GO:0031012), such as biglycan, lumican, extracellular matrix protein 1, SPARC-like protein 1, galectin-1, collagen alpha-2(I) chain and fibronectin.

MMPs and their specific physiological inhibitors, tissue inhibitors of metalloproteinases (TIMPs), are thought to play an essential role in tissue repair, cell death and morphogenesis (22,31). Previous studies discovered unexpected up-regulation of genes coding for multiple MMP/TIMP family members in the lung and spleen of Sprague-Dawley rats exposed to HS and resuscitation (the same model as in the present study) (32). In particular, transcriptional levels of MMP-2, MMP-9, MMP-7, MMP-14, MMP-16, and TIMP-1, TIMP-2 were significantly affected in the spleen, while TIMP-3 and TIMP-1 in lungs, in response to different resuscitation strategies (32).

Indirect evidence about the involvement of MMPs and TIMPs in promoting post-HS injury to different organs has been produced in the last few years (31,32). Presented results strengthen the case in that they directly indicate MMPs as a likely “smoking gun” in mediating PSML-triggered neutrophil priming. This is relevant in that neutrophil priming is often associated with poor outcomes in trauma patients (23). These results provide the theoretical rationale for therapeutic treatments envisaging MMP inhibition as a strategy to promote hepatic integrity and prevent lung injury in response to T/HS (33,34).

Pre- vs Post-HS broader analysis confirms overall increased complexity of the PSML proteome

The results summarized above are further confirmed by pre- vs post-shock proteomics analyses of mesenteric lymph via HPLC-MS/MS, as summarized in Supplementary Figure 1.A and Supplementary Table 3. Of note, the validity of the analytical approach adopted in the present study is evident from the results reported in Supplementary Table 3. This Table highlights the extensive coverage of findings reported in the literature with this dataset, in the same animal model with different analytical approaches (either 2DE, DIGE or iTRAQ (14–17) – highlighted in different colors in Supplementary Table 3). At the same time, from Supplementary Table 3 it emerges that the present study expands current knowledge, by documenting with previously unreported statistically significant (p<0.05 ANOVA) protein mediators of T/HS responses.

PSML was characterized by 65 unique proteins and 91 shared features varying in abundance (Figure 2.A, Supplementary Table 3). However, quantitative analysis based upon label-free quantification (emPAI values) of each detected protein in all six biological replicates per group (Pre-1 to 6 vs Post-1 to 6) clearly showed general increased in abundance of all these proteins in the Post-HS group (Figure 5).

Figure 5.

Scaffold report of proteome changes in pre-shock vs post-shock (3h) mesenteric lymph as determined via HPLC-MS/MS label free peptidomics approaches on six biological replicates per group. The quantitative emPAI changes of the 91 proteins identified both in pre- and post-shock lymph are graphed via the software GENE-E (values increasing from blue to red).

GO term enrichment for biological functions (Supplementary Figure 1.B) or cell compartment localization (Figure 5.D) further confirmed previous classification of increasing proteins derived from time-course analyses (Supplementary Table 2) and provide a solid compendium to existing literature in the field (14–17). Principal Component Analysis (PCA) indicated that pre-shock and early post-shock proteomics changes (post-shock 1) result in a distinct protein profile in comparison to late PSML (Supplementary Figure 3).

Finally, we compared proteins unique to or increasing in PSML in comparison (Supplementary Table 3) to similar datasets from recent studies detailing proteomics profiles of human PSML and plasma (35). In particular, we performed a direct comparison of proteins listed in Supplementary Table 3 of the present study and Table 2 from Dzieciatkowska et al. (35) (including proteins unique to or increasing in abundance in human PSML in comparison to post-T/HS human plasma – highlighted in violet). As a result, we could individuate the following protein classes and matching proteins: 1.) extracellular matrix (ECM) and other structural proteins (SPARC-like protein 1, EGF-containing fibulin-like extracellular matrix protein 1, cofilin-1, actin cytoplasmic 2, lumican, gelsolin, transgelin); 2.) oxidative stress-related proteins (peroxiredoxin 1, extracellular superoxide dismutase [Cu-Zn], protein DJ-1); 3.) metabolic enzymes and transporters, that might also display non-metabolic functions in PSML (moonlighting proteins (35)), including triosephosphate isomerase, cytoplasmic malate dehydrogenase, lactate dehydrogenase A, apolipoprotein A-IV, phosphatidylethanolamine-binding protein 1; 4.) complement components (complement factor D); 5.) signaling and chaperone proteins (14-3-3 epsilon and zeta, carbonic anhydrase 1, peptidyl-prolyl cis-trans isomerase A).

This list of proteins is relevant in that it highlights recurring proteins in rat (present study) and human PSML (45). At the same time, in the present study we concluded that MMPs and ECM remodeling might play a key role in mediating neutrophil priming and adverse reactions promoted by PSML. Consistently, these proteins have been documented to be unique or increased in abundance in PSML albeit not post-T/HS human plasma (35), strengthening the translational potential of the findings reported here with respect to adverse events related to post-T/HS.

Conclusions

In the present study, there are novel dynamic changes of the soluble protein components in PSML. The reported data expanded upon previous observations regarding HS triggering the impairment of serine proteases/anti-proteases involved in clotting responses, by reporting a dynamic early decrease in the levels of the latter, followed by a late progressive increase in the levels of Serpins. This decrease in serine proteases with progressive increases in the levels and activity (as determined through zymography assays) of extra-cellular matrix proteins and MMPs, indicating a role for the impaired protease/anti-protease poise in mediating post-HS responses.

Overall, this study complements existing literature suggesting that early post-shock dynamic changes to the mesenteric lymph proteome are as biologically informative and distinct from late ones (Supplementary Figure 3). We also demonstrated the advantages of investigating the same biological issue through alternative analytical platforms, as to complement the current body of knowledge on this topic. By comparing the present results with the literature about proteome specificity of post-T/HS human fluids, PSML and plasma (35), we highlighted recurring features that might relate the present findings with adverse events associated to PSML-mediated neutrophil priming and thus adverse events secondary to T/HS in patients. In conclusion, early post-HS alterations in the levels and activities of MMPs might represent an interesting candidate for future targeted studies about the role of mesenteric lymph in evoking inflammatory responses, such as neutrophil priming, paving the way for timely therapeutic interventions to cope with/prevent early untoward consequences of T/HS.

Supplementary Material

SDC Fig. 1

Supplementary Figure 1 Hierarchical clustering analyses of quantitative dynamic changes of pre and post-shock mesenteric lymph proteins. Results from Progenesis LC-MS have been exported into Mascot Daemon for protein identification and exported in GENE-E for statistical analysis and clustering. Samples were collected from four animals either pre-shock (Pre), or every thirty minutes post-trauma/hemorrhagic shock (Post-1 to Post-6, except for the 1h time point). Values increase progressively from blue to red in the color scale, as indicated in the legends on top of each cluster. Proteins are enlisted with their relative UniProt ID names, though the _RAT taxonomy specification for Rattus norvegicus has been deleted as to improve the clarity of the figure. Further details are reported in Supplementary Table 1 (protein list) and 2 (GO term enrichment).

SDC Fig. 2

Supplementary Figure 2 Scaffold report of proteome changes in pre-shock vs post-shock (3h) mesenteric lymph as determined via HPLC-MS/MS label free peptidomics approaches on six biological replicates per group. In A, a Venn diagram indicates the shared features (91) and differential post-shock unique (65) proteins in mesenteric lymph. In B and C, GO term enrichment for biological functions and cell compartment for the 256 statistically differential features (fold-change ≥ 1.5, p-value ≤ 0.05 ANOVA – further details are reported in Supplementary Table 3).

SDC Fig. 3

Supplementary Figure 3 Principal Component analysis of proteomics data reveals a clear cut discrimination between pre-shock and early post-shock mesenteric lymph proteome profiles in comparison to late post-shock samples.

List of abbreviations

2DE

Two dimensional gel electrophoresis

DB

Database

DIGE

Differential in-gel electrophoresis

ESI

Electrospray ionization

HS

Hemorrhagic shock

iTRAQ

Isobaric tag for relative and absolute quantitation

ML

Mesenteric lymph

nLC

Nano-flow liquid chromatography

LTQ

Linear ion trap mass spectrometer

MOF

Multiple Organ Failure

MS

Mass spectrometry

MS/MS

Tandem mass spectrometry

PSML

Post-shock mesenteric lymph

RP

Reversed Phase

FT-ICR

Fourier Transformed Ion Cyclotron Resonance mass spectrometer