Vaccination of animals is a critical preventive strategy against potentially fatal or debilitating infectious diseases in animals and is an important tool to protect the public from zoonotic diseases. Vaccines provide protection against these diseases by stimulating immune responses to vaccine antigens originating from the targeted pathogen, but may also induce an immune response to other vaccine components.1–4 Unintended or adverse immune responses may develop, however, following vaccination, and concern of an adverse vaccine reaction is a major deterrent to vaccination for many people. Thus, vaccine safety is a critical concern to owners of companion animals and the veterinarians who care for these animals.
Cell cultures or culture media for vaccine components are a potential source of mammalian proteins that can stimulate an immune response. Administration of multiple vaccines at one time may increase the number of administered proteins and the risk of a vaccine reaction in dogs.5 Dogs with adverse vaccine reactions have demonstrated immunoreactivity to proteins of bovine origin (eg, from fetal calf serum).6 Bovine serum albumin is a large (66-kDa) ionized molecule that has been identified as a potentially common protein-immunogen in several canine vaccines,6 but other specific protein components of canine vaccines have not been elucidated or quantified.
Proteomic strategies have a significant potential to contribute to new vaccine development, to improve existing vaccines, and to provide a better understanding of immune responses to vaccines.7 So far, proteomic vaccine studies have been limited in general, and focus primarily on obtaining global views of the pathogen proteome and host’s immune response.8 Studies that focus on characterizing licensed vaccines, especially comparing different vaccines for the same disease, can provide useful information for improving current vaccines and allow the selection of the most efficient vaccine product.
The objective of our study was to use proteomic analysis to profile the protein components of selected, commonly used veterinary vaccines against canine distemper, leptospirosis, borreliosis, and rabies.
Materials and Methods
Vaccines
Vaccines were selected to be representative of the most common commercial parenterally administered canine vaccines against leptospirosis and canine distemper virus (L&D; Recombitek 4 Lepto and Recombitek C6 Da2P+Pv with Lepto [Boehringer Ingelheim Animal Health Inc]; ULTRA Duramune 4L, ULTRA Duramune DAP+4L, and Duramune DAP+4L [Elanco Animal Health Inc]; Nobivac Lepto 4 and Nobivac Canine 1-DAPPV+L4 [Merck Animal Health, Merck & Co, Inc]; and Vanguard L4 and Vanguard Plus 5 L4 [Zoetis Animal Health, Zoetis Services LLC]), Borrelia burgdorferi (Recombitek Lyme [Boehringer Ingelheim Animal Health Inc]; ULTRA Duramune Lyme, Duramune Lyme, and Duramune Lyme + LeptoVax 4 [Elanco Animal Health Inc]; Nobivac Lyme (Merck Animal Health, Merck & Co, Inc); and Vanguard crLyme and Lymevax [Zoetis Animal Health, Zoetis Services LLC], and rabies (Imrab-3TF and Imrab-1TF [Boehringer Ingelheim Animal Health Inc], Rabvac 3 and Rabvac 1 [Elanco Animal Health Inc], Nobivac 3 Rabies and Nobivac 1 Rabies [Merck Animal Health, Merck & Co, Inc], and Defensor 3 and Defensor 1 [Zoetis Animal Health, Zoetis Services LLC]) in the United States. A total of 10 L&D, 7 Borrelia, and 8 rabies vaccines for dogs were analyzed. Vaccines for L&D were sometimes marketed as separate lyophilized powder and liquid (ie, Leptospira bacterin) for reconstitution and mixing prior to administration. For these vaccines, the lyophilized and liquid components were analyzed separately, resulting in 15 proteomic analyses for the 10 L&D vaccines. Vaccines were obtained from a commercial veterinary supply distributor, typically in 25- or 50-dose packages from a single lot. One vial from each vaccine or component package was chosen by a randomization sequence. Selected vials were then deidentified regarding specific manufacturer or product name, product vials were randomized to sequence order for proteomic analysis, and then samples per vial were analyzed in duplicate.
Vaccine sample preparation for liquid chromatography-tandem mass spectroscopy analysis
The total protein content of the vaccines was measured initially using a bicinchoninic acid (BCA) assay. After the BCA assay, 50 µg (equivalent volume) of vaccine samples were mixed with 4 volumes of cold (–20 °C) acetone, vortexed, and stored at –20 °C overnight to precipitate the proteins. The pellet was washed with 80% cold (–20 °C) acetone and dissolved in 50 μL of 8M urea. The samples were reduced with 10 mM dithiothreitol, alkylated with 20 mM iodoacetamide, and digested overnight at 37 °C with a mass spectroscopy (MS)-grade trypsin and Lys-C mix (Promega) at a minimum 1:25 enzyme-to-substrate ratio. Digestion was stopped by adding 1% formic acid. Peptides were desalted using C18 spin columns (Pierce Biotechnology), eluted with 80% acetonitrile (ACN) and 0.1% formic acid (FA), and dried at room temperature in a vacuum concentrator. The peptides were resuspended in 97% purified water, 3% ACN, and 0.1% FA at a final concentration of 0.2 µg/µL; 5 µL was used for liquid chromatography (LC) tandem mass spectroscopy (MS/MS) analysis.
Liquid chromatography-tandem mass spectroscopy
Samples were analyzed by reverse-phase LC electrospray ionization MS/MS using the Dionex UltiMate 3000 RSLC Nano System coupled to the Orbitrap Fusion Lumos Mass Spectrometer (Thermo Fisher Scientific) as described previously.9 Reverse-phase peptide separation was accomplished using a trap column (300 μm internal diameter [ID] X 5 mm) packed with 5-μm, 100-Å PepMap C18 medium and then separated on a reverse-phase column (50 cm long X 75 µm ID) packed with 2-µm, 100-Å PepMap C18 silica (Thermo Fisher Scientific). The column temperature was maintained at 50 °C to prevent high column back pressure. Mobile-phase solvent A was 0.1% FA in water; solvent B was 0.1% FA in 80% ACN. Loading buffer was 98% water, 2% ACN, and 0.1% FA. Peptides were separated by reverse phase by loading into the trap column in a loading buffer for 5 minutes at a 5-µL/min flow rate and eluted from the analytical column with a linear 82-minute linear gradient of 6.5% to 27% of buffer B, then changing to 40% of B at 90 minutes, 100% of B at 97 minutes, at which point the gradient was held for 7 minutes before reverting to 2% of B at 104 minutes. Peptides were separated from the analytical column at a flow rate of 300 nL/min. The mass spectrometer was operated in positive-ion and standard data-dependent acquisition mode with the Advanced Peak Detection function activated. The fragmentation of precursor ion was accomplished by higher energy collision dissociation at a normalized collision energy setting of 30%. The resolution of the Orbitrap mass analyzer was set to 120,000 and 15,000 at 200 m/z for precursor ion (MS1) and product ion (MS2) scans, respectively, with a maximum injection time of 50 ms for MS1 and 20 ms for MS2. The dynamic exclusion was set at 60 seconds to avoid repeated scanning of identical peptides. The charge state was set at 2 to 7, with 2 as a default charge, and a mass tolerance of 10 ppm for both high and low masses. The full-scan MS1 spectra were collected in the mass range of 375 to 1,500 m/z and MS2 in 300 to 1,250 m/z. The spray voltage was set at 2 kV, with an automatic gain control target of 4e5 for MS1 and 5e4 for MS2.
Liquid chromatography-tandem mass spectroscopy data analysis
Liquid chromatography-MS/MS data were analyzed using MaxQuant software version 1.6.3.310,11 against the Uniprot Bos taurus protein database (www.uniprot.org) concatenated with a common contaminant and a target-decoy database. Cleavage enzymes were setup as Trypsin/P and Lys-C allowing up to 2 missed cleavages. Mass error was set to 10 and 20 ppm for MS1 and MS2, respectively. Alkylation of cysteine and oxidation of methionine was set to fixed and variable modifications, respectively. The false discovery rate (FDR) threshold was set to 0.01 for both the peptides and the proteins. The peptide quantitation was conducted using “unique” plus razor peptides. The razor peptides are the nonunique peptides assigned to protein-protein group with most other peptides. After MaxQuant processing, contaminants and decoy hits were removed, and proteins with no MS/MS count and zero label-free quantitation (LFQ) intensity were also removed. Downstream data analysis and data visualization were performed using JMP version 14 (SAS Institute Inc) and JASP version 0.14.1 (JASP Team).
Network analysis of proteins
Network analysis was performed using STRING version 11.5 (http://string-db.org/)12 with the following parameters: full string network, confidence as the meaning of edges, experiments and databases as active interaction sources, highest confidence level (0.9) with no additional interactors beyond the query proteins, and hiding disconnected nodes in the network. For comparisons, 1-year and 3-year rabies vaccines were separated.
Statistical methods
Protein numbers and quantities were compared between vaccine groups by repeated-measures ANOVA to account for correlation between duplicates and allow for multiple group comparisons in proteomics data.13 To reduce the effect of multiple comparisons resulting from the number of proteins identified, an FDR with a P value of < 0.01 was considered statistically significant. Data from duplicates were averaged for presentation, including those in tables and figures. Summary statistics for protein quantification and identification, when presented, are shown as median (range) because of small sample size.
Results
Protein identification
Total protein concentration ranged from 2.46 to 16.07 µg/µL for L&D vaccines, 0.06 to 0.75 µg/µL for Borrelia vaccines, and 0.90 to 2.98 µg/µL for rabies vaccines based on BCA assay. Within L&D vaccines, the greater protein concentrations were associated with the lyophilized components, and the least with the liquid components.
Liquid chromatography-MS/MS analysis identified 550 mammalian proteins or protein groups in the L&D, Borrelia, and rabies vaccines sampled when searched against a Bos taurus database. The 3-year rabies vaccines contained the largest number of identified proteins, with a median count of 243 (range, 184 to 339) proteins, followed by 1-year rabies with a median of 193 (range, 169–350) proteins. The Borrelia and L&D vaccines had the lowest number of proteins: medians 23 (range, 0 to 45) and 11 (range, 1 to 75) proteins, respectively. When samples were separated by their presentation (in solution, lyophilized, or reconstitution gel), no significant differences in protein numbers were observed. Similarly, comparison of protein molecular weight did not indicate any significant differences among the different vaccine targets. (A list of all the identified peptide sequences and their mapped proteins and protein groups with relative abundances among different vaccines are shown in Supplementary Table S1.)
Proteins identified in each vaccine target were ranked by their normalized abundances determined from their LFQ intensities. Table 1 shows the top 20 most abundant proteins separated by target and are classified as those involved in protein-binding activity modulators, metabolite interconversion enzymes, transfer-carrier proteins, cytoskeletal proteins, defense-immunity proteins, calcium-binding proteins, and extracellular matrix proteins. There was good overlap of the top 20 proteins among rabies, Borrelia, and L&D groups.
Top 20 most abundant proteins by vaccine target in canine vaccines.
| Rabies | Molecular weight (Da) | Borrelia | Molecular weight (Da) | Leptospirosis-distemper | Molecular weight (Da) |
|---|---|---|---|---|---|
| Serum albumin | 69,323 | Serum albumin | 69,323 | Serum albumin | 69,323 |
| Alpha-2-HS-glycoprotein | 38,418 | Alpha-2-macroglobulin | 167,570 | Collagen alpha-1(I) chain | 138,940 |
| Collagen alpha-1(I) chain | 138,940 | Apolipoprotein A-I preproprotein | 30,276 | Alpha-2-HS-glycoprotein | 38,418 |
| Alpha-1-antiproteinase | 46,103 | Vitamin D binding protein | 53,327 | Catenin beta-1 | 69,449 |
| Alpha-fetoprotein | 67,861 | A6QM09_BOVIN uncharacterized protein | 24,732 | Vitamin D binding protein | 53,327 |
| Alpha-2-macroglobulin | 167,570 | Complement C3 | 187,250 | Alpha-1-antiproteinase | 46,103 |
| Actin, cytoplasmic 2 | 41,792 | Serpin A3-1 | 46,236 | Alpha-2-macroglobulin | 167,570 |
| Embryo-specific fibronectin 1 transcript variant | 262,420 | Alpha-2-HS-glycoprotein | 38,418 | Alpha-fetoprotein | 67,861 |
| Alpha-1-acid glycoprotein | 21,440 | Alpha-1-antiproteinase | 46,103 | Apolipoprotein A-I preproprotein | 30,276 |
| Complement C3 | 187,250 | Adiponectin D | 26,133 | Actin, cytoplasmic 2 | 41,792 |
| Vitamin D binding protein | 53,327 | Apolipoprotein A-II | 11,202 | Alpha-1-acid glycoprotein | 21,440 |
| Apolipoprotein A-I preproprotein | 30,276 | Antithrombin-III | 60,068 | A6QM09_BOVIN uncharacterized protein | 24,732 |
| Fetuin-B | 42,663 | Alpha-1B-glycoprotein | 53,553 | Fetuin-B | 42,663 |
| Heat shock cognate 71-kDa protein | 71,240 | CD5 molecule like | 50,211 | Actin, alpha cardiac muscle 1 | 42,019 |
| Vimentin | 53,727 | Transthyretin | 15,727 | Vimentin | 53,727 |
| Endoplasmic reticulum | 72,399 | Apolipoprotein A-IV | 42,989 | Transthyretin | 15,727 |
| A6QM09_BOVIN uncharacterized protein | 24,732 | Beta-2-glycoprotein 1 | 38,245 | Creatine kinase M-type | 42,988 |
| Alpha-1B-glycoprotein | 53,553 | Apolipoprotein C-III | 10,692 | Histone H4 | 14,987 |
| Pyruvate kinase | 57,948 | Inter-alpha-trypsin inhibitor heavy chain H4 | 101,510 | Histone H2A | 14,105 |
| Annexin A2 | 38,612 | Alpha-2-antiplasmin | 54,710 | CD5 molecule-like | 50,211 |
There were 36 proteins identified as common to the 4 vaccine groups (Figure 1). The rabies vaccines had the largest number of proteins in common (> 300 proteins), with an additional 34 uniquely identified proteins in 1-year and 45 in 3-year products. Borrelia and L&D had 16 and 22 uniquely identified proteins, respectively. There were more common proteins identified between L&D and rabies than between Borrelia and rabies. The uniquely identified proteins by vaccine target were divided by class, resulting in 16 protein classes, 9 of which were uniquely attributed to specific vaccine targets. L&D vaccines had proteins uniquely classified as extracellular matrix proteins, nucleic acid metabolism, and transporters. In rabies vaccines, chaperones, chromatin-chromatin binding, scaffold-adaptors, and transfer-carrier proteins were identified only in 3-year labeled products, and transmembrane signal receptor only in 1-year products. In Borrelia vaccines, the intercellular signal molecule class was the only protein class uniquely identified (Figure 1).
A—Venn diagram illustrating the distribution of identified proteins across canine rabies, Borrelia, and leptospirosis-distemper (L&D) vaccines. B—Dot plot representing the class association of the uniquely identified proteins by vaccine target; the size and color of the dots represents the percentage of total.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Venn diagram illustrating the distribution of identified proteins across canine rabies, Borrelia, and leptospirosis-distemper (L&D) vaccines. B—Dot plot representing the class association of the uniquely identified proteins by vaccine target; the size and color of the dots represents the percentage of total.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Venn diagram illustrating the distribution of identified proteins across canine rabies, Borrelia, and leptospirosis-distemper (L&D) vaccines. B—Dot plot representing the class association of the uniquely identified proteins by vaccine target; the size and color of the dots represents the percentage of total.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Venn diagram illustrating the distribution of identified proteins across canine rabies, Borrelia, and leptospirosis-distemper (L&D) vaccines. B—Dot plot representing the class association of the uniquely identified proteins by vaccine target; the size and color of the dots represents the percentage of total.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Venn diagram illustrating the distribution of identified proteins across canine rabies, Borrelia, and leptospirosis-distemper (L&D) vaccines. B—Dot plot representing the class association of the uniquely identified proteins by vaccine target; the size and color of the dots represents the percentage of total.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
Although a considerable number of proteins was identified, serum albumin was the most abundant protein across all analyzed vaccine products, averaging 57% of the total protein content in rabies vaccines, 67% in Borrelia vaccines, and 83% in L&D vaccines. The top 40 most abundant proteins are shown (Figure 2); most of the proteins have relative abundances < 10%.
Dot plot representation of the 40 more abundant proteins separated by canine vaccine target. Color is the mean abundance of each protein in percentages over the total protein abundance of each vaccine dose. ALB = Albumin.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
Dot plot representation of the 40 more abundant proteins separated by canine vaccine target. Color is the mean abundance of each protein in percentages over the total protein abundance of each vaccine dose. ALB = Albumin.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
Dot plot representation of the 40 more abundant proteins separated by canine vaccine target. Color is the mean abundance of each protein in percentages over the total protein abundance of each vaccine dose. ALB = Albumin.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
To observe differences between individual vaccines within vaccine groups in low-abundance proteins, serum albumin was excluded, and the individual vaccine doses were analyzed separately by vaccine target. Other proteins, next in abundance after serum albumin, included alpha-2-HS-glycoprotein (38 kDa), alpha-2-macroglobulin (168 kDa), and collagen alpha-1(l) chain (139 kDa). The largest protein identified in the 20 most abundant proteins was a fibronectin 1 transcript variant (262 kDa), followed by complement C3 (187 kDa). Among the rabies vaccines, alpha-2-HS-glycoprotein, collagen alpha-1(I) chain, and alpha-1-antiproteinase were the most abundant proteins after serum albumin (Figure 3). The proportion of the proteins varied significantly by rabies vaccine maker (P < .01). However, a small variation was observed by period of coverage; the only proteins significantly different between rabies vaccines with 1- and 3-year protection were 14-3-3 protein zeta/delta, 26S proteasome non-ATPase regulatory subunit 2, complement component C7, heat shock protein family A (Hsp70) member 4, and T-complex protein 1 subunit alpha. Analysis of L&D vaccines separated by brand showed significant differences for 15 proteins (P < .01) In Borrelia vaccines, only embryo-specific fibronectin 1 transcript variant was significantly different.
A—Canine rabies vaccine dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (R = Rabies), duration of coverage (1 year or 3 years), and manufacturer (deidentified to A, B, C, or D). B—Canine leptospirosis-distemper (L&D) dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (LD = Leptospirosis-distemper) and deidentified lyophilized or liquid components (A to N).
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Canine rabies vaccine dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (R = Rabies), duration of coverage (1 year or 3 years), and manufacturer (deidentified to A, B, C, or D). B—Canine leptospirosis-distemper (L&D) dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (LD = Leptospirosis-distemper) and deidentified lyophilized or liquid components (A to N).
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Canine rabies vaccine dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (R = Rabies), duration of coverage (1 year or 3 years), and manufacturer (deidentified to A, B, C, or D). B—Canine leptospirosis-distemper (L&D) dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (LD = Leptospirosis-distemper) and deidentified lyophilized or liquid components (A to N).
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Canine rabies vaccine dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (R = Rabies), duration of coverage (1 year or 3 years), and manufacturer (deidentified to A, B, C, or D). B—Canine leptospirosis-distemper (L&D) dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (LD = Leptospirosis-distemper) and deidentified lyophilized or liquid components (A to N).
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—Canine rabies vaccine dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (R = Rabies), duration of coverage (1 year or 3 years), and manufacturer (deidentified to A, B, C, or D). B—Canine leptospirosis-distemper (L&D) dose protein composition (measured as a percentage) without serum albumin. Left-margin vaccine labels indicate type (LD = Leptospirosis-distemper) and deidentified lyophilized or liquid components (A to N).
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
Protein network and functional enrichment of identified proteins in canine vaccines
String network analysis was constructed using the 316 proteins that were common in the rabies vaccines to visualize the interaction between the detected proteins (Figure 4). The edge’s thickness indicated the confidence of the proteins’ interaction, and disconnected protein nodes were excluded. The resulting 153 nodes produced 363 edges, with enriched networks for specific pathways such as ribosome, involved in protein translation; proteasome, key regulator of cell cycle and differentiation as well as antigen processing for immune responses; or glycolysis, central to the production of ATP, nicotinamide adenine dinucleotide, and precursor metabolites.
A—String network analysis of proteins identified in rabies vaccines. B—Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of proteins for rabies vaccines. The y-axis indicates the pathway names and the x-axis indicates the enrichment factor for each pathway. The bubble size indicates the number of proteins in each pathway. The color bar indicates the corrected P value (false discovery rate [FDR]) of the pathways. The blue color represents higher values and the red represents lower values. ECM = Extracellular matrix. TGF = Transforming growth factor.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—String network analysis of proteins identified in rabies vaccines. B—Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of proteins for rabies vaccines. The y-axis indicates the pathway names and the x-axis indicates the enrichment factor for each pathway. The bubble size indicates the number of proteins in each pathway. The color bar indicates the corrected P value (false discovery rate [FDR]) of the pathways. The blue color represents higher values and the red represents lower values. ECM = Extracellular matrix. TGF = Transforming growth factor.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—String network analysis of proteins identified in rabies vaccines. B—Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of proteins for rabies vaccines. The y-axis indicates the pathway names and the x-axis indicates the enrichment factor for each pathway. The bubble size indicates the number of proteins in each pathway. The color bar indicates the corrected P value (false discovery rate [FDR]) of the pathways. The blue color represents higher values and the red represents lower values. ECM = Extracellular matrix. TGF = Transforming growth factor.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—String network analysis of proteins identified in rabies vaccines. B—Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of proteins for rabies vaccines. The y-axis indicates the pathway names and the x-axis indicates the enrichment factor for each pathway. The bubble size indicates the number of proteins in each pathway. The color bar indicates the corrected P value (false discovery rate [FDR]) of the pathways. The blue color represents higher values and the red represents lower values. ECM = Extracellular matrix. TGF = Transforming growth factor.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
A—String network analysis of proteins identified in rabies vaccines. B—Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis of proteins for rabies vaccines. The y-axis indicates the pathway names and the x-axis indicates the enrichment factor for each pathway. The bubble size indicates the number of proteins in each pathway. The color bar indicates the corrected P value (false discovery rate [FDR]) of the pathways. The blue color represents higher values and the red represents lower values. ECM = Extracellular matrix. TGF = Transforming growth factor.
Citation: American Journal of Veterinary Research 84, 3; 10.2460/ajvr.22.11.0192
The low number of uniquely identified proteins in L&D and Borrelia vaccines limited the results of the network and enrichment analysis in these vaccines. String analysis for the 22 proteins identified on the L&D vaccine yielded 14 nodes and 5 edges, with no network or pathway enrichment. For Borrelia vaccine proteins, the string network analysis had 10 nodes and 4 edges, which were annotated with Uniprot keywords as secreted proteins but were not enriched in any pathways.
Discussion
Advances in proteomics have provided us a new way to assess the quality of vaccines. The majority of vaccine proteomic studies so far have focused on identifying new antigens with increased speed and sensitivity,7 and proteomic approaches are being used in many aspects of new vaccine development.8 However, vaccines currently available were mostly developed through empirical methods, which did not use more sensitive modern genomics and proteomics approaches for quality control and quality assurance. Identifying the protein composition of these currently available commercial vaccines using modern MS-based proteomics can be beneficial to determine their quality and safety, and to develop new assays to improve future productions. In this investigation, we used label-free proteomics to identify the mammalian protein components of veterinary vaccines qualitatively and quantitatively against L&D, rabies, and borreliosis.
We focused on mammalian proteins that can be present in vaccines because of the production processes used to generate the vaccine viruses and bacteria. The peptides from the LC-MS/MS analysis were screened against a bovine protein database based on the assumptions that bovine serum is used in cell culture and that there is a high degree of conservation in protein sequences among different cell types used to generate viruses. The potential association of more proteins within a vaccine and postvaccination adverse local or systemic responses requires further elucidation. One-year or 3-year duration of immunity rabies vaccines had the greatest number of identifiable proteins, with approximately 10 to 20X more proteins compared to canine vaccines against Borrelia burgdorferi or Leptospira infection. The much greater quantity of proteins in rabies vaccines may play a role in reported localized reactions and ischemic dermatopathy after rabies vaccination.14,15 Rabies vaccines have also been noted to have a potential in initiating adverse systemic responses when administered in combination with other vaccines.16,17 When reidentified by manufacturer and vaccine type, excluding recombinant vaccines, the estimated total protein concentration for vaccines of the same target generally varied between manufacturers by less than 2-fold. This difference between biological products is markedly reduced from published differences of > 10-fold in some canine vaccines analyzed > 15 years ago.6 It is unknown whether this reduction in variability between vaccines of the same target can be appreciated clinically in the safety records of vaccines by different manufacturers. Large-scale safety studies allowing such comparisons are lacking in veterinary medicine.
Bovine serum albumin (molecular weight, 66 kDa) was the most commonly identified protein in all vaccines except recombinant Lyme vaccines. This most likely reflected use of bovine serum (eg, fetal calf serum) in cell cultures, with serum albumin remaining the dominant excipient protein in vaccines after manufacturing. The next abundant proteins—α-2-HS-glycoprotein, α-2-macroglobulin, and collagen α-1(l) chain—were similar in molecular mass or larger. These mammalian proteins may induce an immune response when injected with vaccine formulations; however, the presence of such antibodies does not necessarily result in clinical disease. Dogs that experienced postvaccination immediate-type hypersensitivity reactions had increased immunoglobulin (Ig) E reactivity to fetal calf serum.6 The incidence of postvaccination immediate-type hypersensitivity reactions in dogs also increases as more vaccines, and more total vaccine volume, are administered during an office visit.5 However, in previous studies,18,19 we identified IgG and IgE antibodies against bovine serum albumin, fibronectin, and thyroglobulin in laboratory-bred Beagles after vaccination, yet all dogs remained healthy.
The most abundant proteins in string network analysis are classified as those involved in protein-binding activity modulators, metabolite interconversion enzymes, transfer-carrier proteins, cytoskeletal proteins, defense-immunity proteins, calcium-binding proteins, and extracellular matrix proteins. The number and diversity in these findings reflected the large number of proteins in rabies vaccines specifically, compared to the other vaccine groups. The analyzed rabies vaccines were all inactivated vaccines in which the rabies virus was grown in cell cultures (eg, hamster or feline cells) and then chemically inactivated. The detected networks of ribosomes, proteasomes, chaperones, and others are likely derived from the cell lines used in the manufacturing process. It is unknown whether these will be found in proteomic analysis of other inactivated vaccines.
Our study was conducted, and vaccines purchased through intramural funds, without knowledge by the vaccine manufacturers. The source of the vaccine vials that was analyzed was, therefore, from a public commercial vendor, as a practicing veterinarian would procure. Although vials of each vaccine were only from a single lot for each vaccine, they should (but may not) be representative of similarly labeled products from the same manufacturer. Our study evaluated only a single serial of each manufacturer’s vaccine. The extraneous proteins in each serial is reflective of how much antigen is added, which is dependent on the antigen input, and this may vary from lot to lot. Also, only a single serial for a single lot was evaluated, and the total level of proteins present may vary from serial to serial for an individual manufacturer. The goal of our study was to assess the quantity and variety of protein content by vaccine type, and was not designed to be a quality control comparison between lots. This study also did not assess vaccines for all commonly administered canine pathogens. Most notably, vaccines against specific canine respiratory pathogens (eg, canine influenza virus and Bordetella bronchiseptica) were not analyzed.
Our study reports proteomic analysis of selected companion animal vaccines, providing insights into the identification and diversity of bovine proteins found in these vaccines. These insights may serve as a foundation for future comparison studies of other vaccines, or as a springboard for investigations into the etiology of protein triggers of adverse vaccine reactions. Future studies focusing on combining proteomic analysis of the protein composition of vaccines with the protein changes that occur in response to infection and vaccination will provide a comprehensive view and will guide vaccine development with greater efficacy.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org
Acknowledgments
No external funding was used in this study.
GEM has served on the speaker bureaus of the vaccine manufacturers.
This research was presented in part as an abstract at the 2020 American College of Veterinary Internal Medicine Forum.
Proteomics experiments including sample preparation, data collection, and data analysis were performed at the Purdue Proteomics Facility. The Orbitrap Fusion Lumos mass spectrometer used for LC-MS analysis was purchased through funding from the Office of the Executive Vice President for Research and Partnership, Purdue University.
All the raw LC-MS files are made publicly available through the MassIVE data repository (massive.ucsd.edu/) under submission ID MSV000090844.
References
- 1.↑
Galassie AC, Link AJ. Proteomic contributions to our understanding of vaccine and immune responses. Proteomics Clin Appl. 2015;9(11–12):972–989. doi:10.1002/prca.201500054
- 2.
Brisse M, Vrba SM, Kirk N, Liang Y, Ly H. Emerging concepts and technologies in vaccine development. Front Immunol. 2020;11:583077.
- 3.
Leitner WW, Ying H, Restifo NP. DNA and RNA-based vaccines: principles, progress and prospects. Vaccine. 1999;18(9–10):765–777. doi:10.1016/S0264-410X(99)00271-6
- 4.↑
Coffman RL, Sher A, Seder RA. Vaccine adjuvants: putting innate immunity to work. Immunity. 2010;33(4):492–503. doi:10.1016/j.immuni.2010.10.002
- 5.↑
Moore GE, Guptill LF, Ward MP, et al. Adverse events diagnosed within three days of vaccine administration in dogs. J Am Vet Med Assoc. 2005;227(7):1102–1108. doi:10.2460/javma.2005.227.1102
- 6.↑
Ohmori K, Masuda K, Maeda S, et al. IgE reactivity to vaccine components in dogs that developed immediate-type allergic reactions after vaccination. Vet Immunol Immunopathol. 2005;104(3–4):249–256. doi:10.1016/j.vetimm.2004.12.003
- 7.↑
Adamczyk-Poplawska M, Markowicz S, Jagusztyn-Krynicka EK. Proteomics for development of vaccine. J Proteomics. 2011;74(12):2596–2616. doi:10.1016/j.jprot.2011.01.019
- 8.↑
Aslam B, Basit M, Nisar MA, Khurshid M, Rasool MH. Proteomics: technologies and their applications. J Chromatogr Sci. 2017;55(2):182–196. doi:10.1093/chromsci/bmw167
- 9.↑
Barabas AJ, Aryal UK, Gaskill BN. Proteome characterization of used nesting material and potential protein sources from group housed male mice, Mus musculus. Sci Rep. 2019;9(1):17524. doi:10.1038/s41598-019-53903-x
- 10.↑
Cox J, Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteome-wide protein quantification. Nat Biotechnol. 2008;26(12):1367–1372. doi:10.1038/nbt.1511
- 11.↑
Cox J, Neuhauser N, Michalski A, Scheltema RA, Olsen JV, Mann M. Andromeda: a peptide search engine integrated into the MaxQuant environment. J Proteome Res. 2011;10(4):1794–1805. doi:10.1021/pr101065j
- 12.↑
Szklarczyk D, Franceschini A, Kuhn M, et al. The STRING database in 2011: functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res. 2011;39(Database issue):D561–D568. doi:10.1093/nar/gkq973
- 13.↑
Oberg AL, Mahoney DW, Eckel-Passow JE, et al. Statistical analysis of relative labeled mass spectrometry data from complex samples using ANOVA. J Proteome Res. 2008;7(1):225–233. doi:10.1021/pr700734f
- 14.↑
Hendrick MJ, Dunagan CA. Focal necrotizing granulomatous panniculitis associated with subcutaneous injection of rabies vaccine in cats and dogs: 10 cases (1988–1989). J Am Vet Med Assoc. 1991;198(2):304–305.
- 15.↑
Vitale CB, Gross TL, Magro CM. Vaccine-induced ischemic dermatopathy in the dog. Vet Dermatol. 1999;10(2):131–142. doi:10.1046/j.1365-3164.1999.00131.x
- 16.↑
Moore GE, HogenEsch H. Adverse vaccinal events in dogs and cats. Vet Clin North Am Small Anim Pract. 2010;40(3):393–407. doi:10.1016/j.cvsm.2010.02.002
- 17.↑
Starr RM. Reaction rate in cats vaccinated with a new controlled-titer feline panleukopenia-rhinotracheitis-calicivirus-Chlamydia psittaci vaccine. Cornell Vet. 1993;83(4):311–323.
- 18.↑
HogenEsch H, Dunham AD, Scott-Moncrieff C, Glickman LT, DeBoer DJ. Effect of vaccination on serum concentrations of total and antigen-specific immunoglobulin E in dogs. Am J Vet Res. 2002;63(4):611–616. doi:10.2460/ajvr.2002.63.611
- 19.↑
Scott-Moncrieff JC, Azcona-Olivera J, Glickman NW, Glickman LT, HogenEsch H. Evaluation of antithyroglobulin antibodies after routine vaccination in pet and research dogs. J Am Vet Med Assoc. 2002;221(4):515–521. doi:10.2460/javma.2002.221.515
