Investigating an increase in Florida manatee mortalities using a proteomic approach

This proteomic survey was conducted to identify proteins that were differentially expressed in the serum of manatees affected by two distinct mortality episodes: a red tide group and an unknown mortality episode group in the IRL. These groups were compared to a control group sampled at Crystal River. The red tide group’s exposure was evidenced by the presence of the PbTx antigen, with brevetoxin values in the 4.3 to 14.4 ng/ml range. The other group did not present with clinical symptoms except for mild cold stress in some animals. Two proteomics approaches were employed, 2D-DIGE and shot gun proteomics using LC–MS/MS, which provided similar results, suggesting that several serum proteins were specifically altered in each of the manatee mortality episode groups compared to the Crystal River control group. The differentially expressed serum proteins were cautiously identified based on annotation of the manatee genome^6,7 and their amino acid sequence homologies with human serum proteins. While additional work still needs to be done to confirm that the identified manatee proteins function similarly to their human homologs, possible insight on the function of the proteins can be derived from human studies.

The two proteomics methods used, 2D-DIGE and iTRAQ LC–MS/MS are complementary and both rely on LC–MS/MS for protein identification. 2D-DIGE is a top-down approach, quantifying the differentially expressed proteins at the protein level before identifying the protein by LC–MS/MS, while the iTRAQ method is a bottom-up approach, where the whole proteome is first digested with trypsin, the generated peptides are separated by chromatography and identified and measured by mass spectrometry. Mass spectrometry has become the primary method to analyze proteomes, benefitting from genomic sequences and bioinformatics tools that can translate the sequences into predicted proteins. There are excellent reviews of proteomics methods and how they may be used across species^8,9.

In total, 19 of the 26 proteins identified using the 2D-DIGE method were also identified by iTRAQ (Supplementary Table 1) which showed that these findings were replicated using two complementary experimental methods. In the 2D-DIGE method, most of the proteins were found in multiple spots, suggesting that they were differentially modified. 2D-DIGE can separate proteins based on a single charge difference. Some of the spots contained multiple proteins so it was difficult to determine the fold change of each of the proteins in these spots. For example, protein C4A was identified in 7 different spots, likely representing multiple isoforms. We were not able to corroborate the different post-translational modifications (PTMs) with iTRAQ, as the experiment was not designed to look for PTMs, only total protein quantitation. A drawback of 2D-DIGE is that keratin introduced into the sample from reagents at the time of electrophoresis or through the multiple steps required for protein extraction is also seen in the gels^10,11,12. It is unlikely that the keratins were from the serum samples, as blood was collected directly into vacuum tubes. Because of the issue of keratin contamination, the 2D-DIGE method is considered more qualitative in its determination and thus in this study, iTRAQ data were the primary basis for quantitation.

Pathway analysis detects groups of proteins that are linked in pathways that may be related to disease processes. We used Pathway Studio using subnetwork enrichment analysis to determine disease pathways potentially in place for the red tide and IRL manatees. The Pathway Studio database is constructed from relationships detected between proteins and diseases from articles present in Pubmed but is heavily directed towards human and rodent proteomes. To be able to use this tool, we assigned human homologs to the identified manatee proteins, assuming that based on their sequence homology the proteins would function in a similar way. There are many studies that suggest this assumption has merit, for example Nonaka and Kimura have examined the evolution of the complement system and found clear indications of homology among vertebrates¹³.

The top 20 pathways for the red tide group (Table 3) and the IRL group (Table 4) show the diverse set of molecular pathways that may be affected by the exposures. Many of the same pathways appeared for both groups including thrombophilia, inflammation, wounds and injuries, acute phase reaction and amyloidosis. Thrombophilia was the most upregulated pathway for the IRL group (p-value 1.10E-19) and the second most upregulated pathway for the red tide group (p-value 4.1E-19). Thrombophilia, a condition in which blood clots occur in the absence of injury, happens when clotting factors become unbalanced. We obtained proteomics information on 12 of the proteins in this pathway, with some moving in opposing directions. The dysregulated proteins that were increased for both red tide and the IRL groups were SERPIN D1 (Serpin family member D 1), CRP (C-reactive protein), and PLAT (plasminogen activator) and the ones that were decreased in both groups, were SERPIN C1 (Serpin family member C 1), F5 (coagulation factor 5), and ALB (albumin). One protein, AGT (angiotensinogen), was upregulated in the red tide group but downregulated in the IRL. HRG (histidine rich glycoprotein), PROS1 (Protein S), C4BPA (complement component 4 binding protein alpha, and F2 (coagulation factor 2, also known as prothrombin) were downregulated in the red tide group but upregulated in the IRL group. The disparate regulation of proteins in this pathway suggests that clotting was among the pathways disrupted in the affected manatees. Red tide exposed manatees often present with hemorrhagic issues in their intestines, lungs and the brain (¹⁴), suggesting that downregulation of coagulation factors may be responsible for this clinical evaluation. Interestingly HRG was upregulated in the IRL by 1.34-fold and downregulated in the red tide group by 0.56-fold, making this protein a good biomarker to distinguish the two events.

Among the manatees in the red tide group, inflammation was ranked 3rd (p-value < 6 E−19) and wounds and injuries, 6th (p-value < 8E−18), while for the IRL group, wounds and injuries was ranked 2nd (p-value < 1.3 E−19) and inflammation, 4th (p-value < 2 E−18). Given the importance of these pathways related to acute immunological reaction, their graphical interpretations (including their component entities) are shown in Fig. 3. There is significant overlap in proteins that are associated with inflammation and wounds and injuries in the two groups, Fig. 3 A,B. The combined pathways were built first on the data that were obtained from the red tide group (Fig. 3A). Using the same set of entities, we overlaid the data from the IRL group (Fig. 3B). In these pathways, red indicates proteins that are increased, while blue indicates proteins that are decreased, and gray denotes proteins that were unchanged. The intensity of the color is proportional to the degree of increase or decrease. While many of the proteins are altered in the same direction in both the red tide and IRL groups, some of the proteins are altered in opposite directions, for example ITIH2 (inter-alpha (globulin) inhibitor H2), GSN (gelsolin), and KNG1 (kininogen), among others, suggesting that the exposures were quite distinct. Interestingly, red tide exposure has been correlated to immune dysfunction in manatees, since isolated lymphocytes from manatees exposed to red tide are less able to proliferate when exposed to a mitogen such as Concanavalin A (ConA) or phytohemagglutinin (PHA)¹⁵.

Proteins that comprise the acute phase reaction pathway showed increased concentrations in both the red tide and IRL manatees (Fig. 3C,D). The pathway was built on the data obtained from the red tide group (Fig. 3C) and overlaid with the IRL group (Fig. 3D). This pathway includes the complement cascade, which is an integral part of innate immunity¹⁶, which is highly conserved among vertebrates¹³. Comparisons to the well-studied system in humans provides hypotheses as to what might be happening in manatees. The complement pathway in mammals is comprised of over 50 different proteins that act together to facilitate inflammation and immune response, promote phagocytosis, promote antibody/antigen binding, and attack the membrane of foreign invaders¹⁷. The complement cascade is most often activated by antibody-antigen binding which instigates the activation of complement 3 (C3)¹⁸. The most common initiating event is C1q, which binds to several molecules such as C-reactive protein and undergoes a conformation change¹⁶. Both C3 and C4 and their molecular subcomponents are early actors in the complement cascade that activate later proteins in the pathway, such as C5-C9 activation, which can cause self-inflicted damage to the cell or attack foreign microbial agents¹⁶.

In the red tide manatees, the complement cascade protein complement C1s subcomponent (FC 1.33), complement C1q subcomponent subunit B (FC 1.46), complement C3 (FC 1.42), and complement C4-A isoform 1 (FC 1.83) were all upregulated. Western blot, used as an orthogonal validation of C4-A upregulation in the red tide group, showed it to be increased by 9.8-fold. iTRAQ is known to underestimate the actual increase in proteins due to suppression from the reactive chemistry of the isobaric tags¹⁹. Since the rabbit anti-human C4-A antibody cross reacts with the manatee homolog, this could be used to create an assay. C4 is not very specific for red tide, as other insults may change its expression. However, it may be useful if included in a panel of blood proteins.

In the IRL manatees, the C4-A (FC 1.25) showed a small increase in expression over controls in the iTRAQ experiment and the increase was confirmed by the Western blot, however the change was not significant compared to controls. Other members of the complement cascade, which also appeared among the proteins, were at most increased by only 10% over controls and these also were not significant. Thus, the complement cascade may not represent a major pathway for IRL manatees.

Interestingly, while many of the same entities were altered in the Acute Phase Reaction Pathway, the pattern of alteration was different for the two groups. Some of the entities were regulated in the same direction but some in reverse direction. Two examples for regulation in reverse direction are alpha-2-macroglobulin (A2M) and complement component 4 binding protein (C4BPA), where A2M was upregulated in manatees exposed to red tide and downregulated in manatees from the IRL and C4BPA was downregulated in the red tide group but upregulated in the IRL group.

The amyloidosis pathway is also important to consider for manatees exposed to red tide and the IRL mortality episode (Fig. 3E,F). Previous studies have shown a relationship between serum amyloid protein (SAA) with an increased prevalence of inflammation^20,21. SAA is an acute phase protein and is one of the few existing biomarkers used to diagnose pre-existing illness and inflammation in manatees. It is routinely measured during manatee health assessments^21,22 and thus for the current study, we have SAA values for the manatees sampled during health assessment in Brevard County and Crystal River. This protein was not measured for the red tide manatees that were recovering at Zoo Florida. All of the SAA values were in the normal range (< 50 mg/ml)²¹, except for one manatee from Crystal River, which was slightly out of range at 60 mg/ml. SAA values > 80 mg/ml to 1200 mg/ml are considered truly elevated and are likely to be present in disease states²¹. SAA was not identified in either the 2D-DIGE or iTRAQ experiments in this study. Instead, other acute phase proteins were evaluated. These included C-reactive protein (CRP), which was highly upregulated in the red tide group (Fig. 3E) and slightly upregulated in the IRL group (Fig. 3F), and serum amyloid P-component (APCS), which was downregulated in the IRL group (AR 0.81) (Fig. 3F) and to a lesser extent in the red tide group (AR 0.96) (Fig. 3E). CRP has not been previously evaluated in manatees because available commercial mammalian antibodies do not apparently cross react²¹.

In the current study, other proteins classified as regulators of amyloidosis²³ were also altered including albumin (ALB)²³, alpha-2-macroglobulin (A2M)²⁴, apolipoprotein A1 (APOA1)²⁵, fibrinogen alpha chain (FGA)²⁶, gelsolin (GSN)^27,28, and serpin peptidase inhibitor C1 (SERPINC1)²⁹, among others. Amyloidosis occurs when an abnormally high amount of amyloid protein begins to accumulate and form fibrils and deposits, which can disrupt normal tissue architecture^30,31. Fibrils are bundles of misfolded proteins that form deposits³². In Alzheimer’s disease, the amyloid β precursor protein (APP) accumulates and forms a plaque in the brain³³.

The amyloid protein transthyretin³⁴ was decreased in the red tide (FC 0.54) and IRL (FC 0.70) groups. Transthyretin transports thyroxine (T4) and triiodothyronine (T3) thyroid hormone and retinol so a decrease in expression may result in decreased thyroid function or an increase in amyloidosis³⁵. Two other proteins of note were vibronectin and gelsolin. Vibronectin is a component of the extracellular matrix in higher animals³⁶ and it has been associated with amyloid plaques in Alzheimer’s disease³⁷. Gelsolin regulates actin binding in a calcium dependent process and has been found in blood vessels and tissue basement membranes of patients with wide-spread systemic amyloidosis^27,31. Both of these proteins were over expressed in the red tide group (vitronectin, FC 1.57: and gelsolin, FC 1.41). The IRL manatees only had increased expression of vitronectin (FC 1.25).

Clearly the manatees that were confirmed to have been exposed to red tide, exhibited by the brevetoxin values in their blood, were sicker than the manatee samples collected in the IRL. The intensity of expression of the differentially expressed proteins support this observation. Some of the manatees sampled in the IRL had clinical symptoms of mild cold exposure but were also in the vicinity of the unknown mortality episode in the IRL and they may have had subclinical effects. It is likely that the changes in the serum proteome reflected both of these stressors.

This proteomic analysis identified additional serum proteins that may serve as biomarkers of disease after further evaluation and validation. The proteins with the top five-fold changes for the red tide group were ceruloplasmin-like (FC 2.32), pyruvate kinase isozymes M1/M2 (FC 2.29), angiotensinogen (FC 2.08), complement C4-A (FC 1.83), and C-reactive protein (CRP) (FC 1.68). Ceruloplasmin’s main function is in iron and copper homeostasis and it is increased when iron levels in the brain are elevated due to oxidative stress³⁸. For the IRL group, the proteins with the top five highest average fold change included kininogen-1 isoform 1 (FC 1.38), protein AMBP (FC 1.38), histidine-rich glycoprotein (FC 1.34), properdin (FC 1.30), and complement C4-A (FC 1.25).

The most under expressed proteins in the red tide manatee group included transthyretin (FC 0.54), kininogen-1 isoform 1 (FC 0.54), histidine-rich glycoprotein (FC 0.56), inter-alpha-trypsin inhibitor heavy chain H1 (FC 0.57), and fibronectin isoform 3 (FC 0.57). In the IRL group, transthyretin (FC 0.70) was the most decreased protein. While many of the proteins identified through this survey serve various cellular functions, several key biological roles including inflammatory and immune response, the complement cascade activity, acute-phase response, amyloid accumulation, and iron and copper homeostasis, were increased in the samples. Ceruloplasmin (CP) was the most increased protein in the red tide group and its association with oxidative stress and neurodegenerative effects may reflect direct effects of red tide in the brains of affected manatees^38,39, which includes disorientation, seizing and inability to surface, resulting in drowning¹⁴.

Manatees sampled in the IRL, during the unknown mortality event, also presented with elevated levels in specific serum proteins, which may point to a compromised immune system. Since many undiagnosed manatee mortalities occur every year, developing enhanced diagnosis criteria in the form of biomarkers will help researchers improve their diagnosis and treatment protocols. Proteins related to amyloid formation and the complement cascade may serve as potential markers of disease after careful validation. We recommend that these potential biomarkers receive further validation with a larger group of manatees. CRP was previously a candidate for evaluation for manatees, but in the absence of a commercial antibody was abandoned²¹. With manatee-specific antibodies, it might be possible to develop specific assays for each of these proteins.

Source: Ecology - nature.com