Haptoglobin-Matrix Metalloproteinase 9 Complex As a Biomarker for Acute Inflammation in Cattle

Home , Fibronectin, Haptoglobin, Lactoferrin

Haptoglobin-Matrix Metalloproteinase 9 Complex as a Biomarker for Acute Inflammation in Cattle

THESIS

Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University

Charles Austin Hinds D.V.M

Graduate Program in Veterinary Clinical Sciences

The Ohio State University

2011

Master's Examination Committee:

Jeffrey Lakritz, Advisor

Andrew Niehaus

Christopher Premanandan

Paivi Rajala-Schultz

D. Michael Rings

Copyrighted by

Charles Austin Hinds

2011

Abstract

Bovine respiratory disease (BRD) is a major cause of economic loss in feedlots in the

United Sates. These losses are associated not only with morbidity and mortality, but also the expense of using antimicrobial drugs unnecessarily. One of the recurring problems is the inability to diagnose and therefore treat respiratory disease appropriately.

An Hp-MMP 9 protein complex has been identified in neutrophil granules and in the serum of cattle with acute bacterial sepsis. The purpose of this project was to evaluate the utility of an Hp-MMP 9 complex ELISA in the diagnosis of acute septic inflammation in cattle. Three experiments were performed.

The first experiment was designed to determine whether Hp-MMP 9 could be used in the prediction of BRD in calves recently admitted to feedlots. Using health, treatment and weight gain data, our aim was to determine whether Hp-MMP 9 could predict which calves would be identified with clinical respiratory disease and would require therapy in the days following sample collection. We compared serum concentrations of Hp to Hp-MMP 9 to assess how well the complex performed in these animals. Our results clearly show that serum Hp and Hp-MMP 9 complex are present in cattle admitted to a feedlot. It appears that the presence of these two analytes are independent; however, due to the inability to obtain health and treatment records, we were not able to draw any conclusions about Hp-MMP 9 as a predictor for respiratory or other diseases or reduced ADG in feedlot cattle. ii

The second experiment was designed to evaluate the utility of Hp-MMP9 complex ELISA in comparison to ELISA for total Hp or MMP 9 alone as an indicator of acute septic inflammatory disease in cattle. Animals were classified as being healthy, having acute inflammation, or having chronic inflammation. Serum Hp, MMP 9, and

Hp-MMP 9 concentrations were measured from each animal and compared to disease status. The results of this experiment demonstrated significant differences in serum Hp-

MMP 9 concentrations observed in cattle with acute septic disease compared to those animals with chronic inflammatory/metabolic disease or healthy animals. Total Hp concentrations were higher in diseased animals but not different between acute and chronic disease. Serum MMP 9 concentrations were not different between any groups.

The final experiment evaluated the serum concentrations of Hp-MMP-9 during an

LPS challenge as a surrogate of an inflammatory event. For this experiment, we administered E. coli lipopolysaccharide IV to calves and measured the Hp-MMP 9 complexes in serum over time (-24 hrs to 96 hrs) by serum ELISA. The concentration versus-time curves for Hp-MMP 9 concentration in the calves in this study, albeit, much lower than serum haptoglobin, were remarkably very similar to the concentration versus- time curves of serum Hp concentration. It is reasonable to assume that a majority of the

Hp detected by the Hp ELISA is neutrophil in origin.

These experiments indicate that Hp-MMP 9 is an indicator of acute inflammation in cattle. Further study is needed to determine the usefulness of this test in the field.

iii

Acknowledgments

I would like to thank Dr. Lakritz for his guidance in not only this project, but also in my clinical and academic endeavors throughout my residency. I would like to thank the rest of my Master’s committee Drs. Niehaus, Premanandan, Rajala-Schultz, and Rings.

I would also like to thank the entire Food and Fiber section for their support throughout my residency.

Vita

May 2000 ...... Saltillo High School, Saltillo, MS

2007...... D.V.M., Mississippi State University

2007-2008 ...... Intern, Food and Fiber Medicine and

Surgery, The Ohio State University

2007-2008 ...... Resident, Food and Fiber Medicine, The

Ohio State University

Publications

Gregory A. Bannikov, C. Austin Hinds, Paivi J. Rajala-Schultz, Chris Premanandan, D.

Michael Rings. Serum Haptoglobin-Matrix Metalloproteinase 9 (Hp-MMP 9) Complex as a Biomarker of Systemic Inflammation in Cattle. Vet Immunol Immunopathol

2011;139:41-49.

Fields of Study

Major Field: Veterinary Clinical Sciences

Table of Contents

Abstract ...... ii

Acknowledgments...... iv

Vita ...... v

List of Tables ...... vii

List of Figures ...... viii

Chapter 1: Literature Review ...... 1

Chapter 2: The Utility of Hp-MMP 9 Complex ELISA vs. Hp ELISA in Predicting

Bovine Respiratory Disease in Feedlot Cattle ...... 25

Chapter 3: Serum Hp-MMP 9 Complex in Cattle with Clinically Defined Diseases...... 33

Chapter 4: Serum Hp-MMP 9 Complex Response Following a Single Intravenous Dose of Lipopolysaccharide ...... 48

References ...... 60

Appendix A: Tables ...... 70

Appendix B: Figures ...... 79

List of Tables

Table 1. Calves grouped by number of positive Hp time-points ...... 71

Table 2. Calves grouped by number of positive Hp-MMP 9 time-points ...... 72

Table 3. Samples grouped by Hp concentration and corresponding ADG ...... 73

Table 4. Samples grouped by Hp-MMP 9 concentration and corresponding ADG ...... 74

Table 5. Disease Classification 1 Acute with Hp, MMP 9, and Hp-MMP 9 ...... 75

Table 6. Disease Classification 2, Chronic with Hp, MMP 9, Hp-MMP 9 ...... 76

Table 7. Disease Classification 3, Healthy with Hp, MMP 9, Hp-MMP 9 ...... 77

Table 8. Mean Hp, MMP 9, Hp-MMP 9 for disease classifications ...... 78

vii

List of Figures

Figure 1. Correlation of Hp, Hp-MMP 9 in all samples ...... 80

Figure 2. Correlation of Hp, Hp-MMP 9 excluding Hp=0 ...... 81

Figure 3. Hp-MMP 9 concentration when Hp=0 vs. Hp>0 ...... 82

Figure 4. ADG when Hp was positive at 0, 1, >1 time-points ...... 83

Figure 5. ADG when Hp-MMP 9 was positive at 0, 1, 2, >2 time-points ...... 84

Figure 6. ADG compared to concentration of Hp...... 85

Figure 7. ADG compared to concentration of Hp-MMP 9 ...... 86

Figure 8. Hp-MMP 9 and MMP 9 ELISA specificity ...... 87

Figure 9. Hp, MMP 9, and Hp-MMP 9 concentrations compared to diseases status ...... 88

Figure 10. Mean respiratory rates of calves during endotoxin trial ...... 89

Figure 11. Mean rectal temperature of calves during endotoxin trial ...... 90

Figure 12. Mean heart rate of calves during endotoxin trial ...... 91

Figure 13. MeanWBC of calves during endotoxin trial ...... 92

Figure 14. Mean band neutrophils of calves during endotoxin trial ...... 93

Figure 15. Mean Hp of calves during endotoxin trial ...... 94

Figure 16. Mean Hp-MMP 9 of calves during endotoxin trial ...... 95

Figure 17. Hp, Hp-MMP 9 over time ...... 96

Figure 18. Hp-MMP 9, respiratory rate over time ...... 97 viii

Chapter 1: Literature Review

Introduction

Bovine respiratory disease (BRD) is the leading cause of morbidity and mortality in feedlot cattle in the United States. Roughly 45% of death losses in cattle in 2005 were due to inflammatory disease, 60% of which was due to BRD.1 A review of the relevant literature indicates losses in average daily body weight gain (ADG) range from 0.14-0.33 kg/head/day (0.31 – 0.73 lb./head/day) in those animals experiencing 1 or more episode of BRD. Using estimates from one study, lost ADG observed in cattle with BRD resulted in up to 46 lbs/head over a 273 day feeding period.2 Furthermore, figures from the

1990s indicate that the medical expenses associated with treating respiratory disease range from $20.76 to $37.90 per animal treated.3 Economic losses are not only associated with treating diseased animals, but also with treating animals that are not diseased and not detecting and treating animals that are diseased. In one study2, 35% of all steers were treated for respiratory disease between birth and slaughter and only 72% of these steers had pulmonary lesions at slaughter. More importantly, 68% of the untreated animals had pulmonary lesions at slaughter.2 In those animals with pulmonary lesions at slaughter, significant reductions in mean daily gain were observed during the feeding period. This study suggests that the clinical diagnosis of BRD is frequently inaccurate. 1

The loss in ADG and the added expenses of feeding and treatment makes BRD an economically important disease of feedlot cattle. Early and accurate detection of inflammation is vital in reducing morbidity and mortality losses associated with BRD.

Reducing the number of animals treated with antimicrobial agents and recognizing lung inflammation earlier remains an elusive goal facing the veterinarian and producer.

Inability to identify sub-clinical BRD early and accurately has been identified as responsible for the relatively stable occurrence of BRD in cattle populations under current production systems.4 The inability to accurately identify BRD in recently arrived feedlot animals results in over-treatment of animals with clinical signs compatible with

BRD, and lack of early treatment for those animals not recognized as having BRD. Sub- optimal case definition results in over-treatment of animals, and increased costs to production; whereas, inaccurate case definition also contributes to reduced mean daily gain of feedlot animals. Delayed therapy may also be responsible for development of chronic lung lesions, which are strongly associated with reduced weight gain during the feeding period.2, 5-7 It is clear that early diagnosis based upon improved case definition is critical to reducing the impact of BRD on cattle production.

Inflammation and the Acute Phase Response

When sentinel cells (i.e., macrophages and dendritic cells) are activated by invading microbes or microbial products, these cells release molecules including major

. inflammatory cytokines (TNF-α, IL-1, IL-6), chemokines, oxidants (O2-, H2O2, OH, and

NO.), and lipids (leukotrienes and prostaglandins).8 These molecules are responsible for

2 up-regulation of inflammatory responses, including recruiting and activating other cells such as neutrophils, vasodilation, increasing vascular permeability, and stimulating production of acute phase proteins.8

The body’s overall response to inflammation is a complicated process involving gene expression, protein production, and changes in physiologic responses, which together form what is known as the acute phase response. The acute phase response begins with a change in plasma concentration of cytokines which, in turn, leads to a large number of behavioral, physiologic, biochemical, and nutritional changes such as fever, somnolence, anorexia, leukocytosis, decreased gluconeogenesis, and plasma protein alterations (acute phase proteins (APP)).9 The acute phase response is mediated in large by production of APP in the liver. In most cases of localized inflammation (e.g. the lung, gut, liver, mammary gland) disseminated (systemic) responses include upregulation of liver gene expression for APP as well as protein expression and release into the circulation. Other organs/tissues produce APP (uterus in pregnant women produces haptoglobin10); however, serum concentrations of APP produced by the liver are thought to greatly outweigh those quantities produced by non-hepatic sources.11

Acute Phase Proteins

In human medicine an APP has been defined as one whose plasma concentration changes (either positive or negative) by at least 25 percent during inflammatory disorders.9 During inflammation, increased serum concentration of positive APPs is largely dependent on hepatocyte production in response to the up-regulation of genes

3 encoding the APPs.9 This gene up-regulation is stimulated by cytokines. Interleukin 6

(IL-6) is a major stimulator of the production of most acute-phase proteins.9 There are also negative acute phase proteins, mainly albumin, whose serum concentrations decrease as the liver switches protein production towards positive APP production. Overall, this results in increased production of positive APP (alpha and beta globulins), and reduction of production of negative APP’s such as albumin. These events occur with a time course of days to weeks. The end result, elevated serum globulins with reduced serum albumin are observed in animals with chronic inflammation.

In cattle, several APP have been evaluated to determine their usefulness as biomarkers of inflammation. These include haptoglobin (Hp),12-22 α-1 proteinase inhibitor (α-1 antitrypsin),12, 13 ceruloplasmin,12, 13, 21 α-1 acid glycoprotein (AGP),13, 17, 20 serum amyloid-A (SAA),14, 16-20, 22, 23 fibrinogen,15-17 and lipopolysaccharide (LPS) binding protein (LBP).20 Many of these APP function to initiate or sustain the inflammatory response, while others may have anti-inflammatory actions.9 The function of Hp will be discussed in detail in the following text but is mainly involved in sequestration of iron. Serum amyloid A has been demonstrated to result in chemotaxis of monocytes, polymorphonuclear cells, and T cells, as well as marked inhibitory effects.24

Ceruloplasmin is a copper containing ferroxidase that oxidizes toxic ferrous iron to its non-toxic ferric form. It protects the tissue from the iron-mediated free radical injury and is involved in various antioxidant and cytoprotective activities.25 AGP also has pro- inflammatory and anti-inflammatory effects through downregulation of neutrophil and platelet function to decrease local tissue damage and stimulating pro- and anti-

4 inflammatory cytokines.26 The APP α1-acid glycoprotein binds LPS and inhibits its activity; α1-acid glycoprotein also can bind to numerous drugs and may sustain drug transport levels even with a decrease in albumin.24. In contrast, LBP is necessary for LPS to initiate its inflammatory effects.8

Studies evaluating the utility of APP in cattle with inflammatory disease

Bovine mastitis

In a study examining the utility of APP in acute mastitis, serum concentrations of

3 acute phase proteins (ceruloplasmin, α1 antitrypsin, and haptoglobin) were evaluated.

Haptoglobin is the only APP-biomarker absent in the serum of healthy cattle; during inflammation, sharp increases were associated with the development of mastitis.12 In those cows studied by Conner et al., 1986, the average duration of clinical signs (mastitis, fever, abortion, diarrhea, swollen hocks, recumbent, depression, anorexia) was 1.6 ± 2.5 days (range: 0 – 9 days), and serum concentrations of AGP, ceruloplasmin and haptoglobin were all increased compared with those values obtained from normal healthy cows. These authors also evaluated the concentrations of Hp, SAA and AGP in serum and milk from cattle with mastitis.27 Serum concentrations of haptoglobin in healthy cows in this study were less than the lower limit of detection of the test (<0.02 mg/mL).

However, cattle with mild mastitis had median serum Hp concentrations of 0.47 (range

0.02 – 1.36) mg/mL; whereas the cattle with moderate mastitis had median serum concentrations of Hp of 0.74 (range 0.02 – 1.84) mg/mL.27 Median serum concentrations of SAA in cows with mild mastitis was 13.8 µg/mL (range 5.4 – 142 µg/mL), while cows

5 with moderate mastitis had median serum concentrations of 29.9 µg/mL (range 5.9 – 141

µg/mL). A significant association between serum Hp and SAA concentrations was observed in cattle with mastitis.27 Both serum Hp and SAA were significantly elevated in mastitis when compared with normal cows.27 In addition, serum concentrations of AGP were significantly higher in cows with mastitis than they were in normal cows. Median serum concentrations of AGP in healthy cows were 0.31 mg/mL (range 0.2 – 1.04 mg/mL), whereas cows with mild (0.53 mg/mL; range 0.11 – 1.26 mg/mL) and moderate mastitis (0.54 mg/mL; range 0.3 – 2.0 mg/mL) were not significantly different from each other.27

Milk concentrations of Hp, SAA and AGP were also compared to serum concentrations of these APPs.27 Milk concentrations of Hp in normal animals were lower than the lower limit of detection of this assay (<0.02 mg/mL). However, median milk concentrations of Hp in cattle with mild mastitis (clots in milk) were 0.09 (range 0.02 –

2.15 mg/mL); whereas median milk concentrations of Hp in cows with moderate mastitis

(clots in milk and visible signs of inflammation in mammary gland) were 0.11 mg/mL range 0.02 – 1.19 mg/mL). Milk Hp concentrations of the opposite quarter (diagonal) was more often than not <0.02 mg/mL suggesting limitation of the inflammation to one quarter in cattle with mild to moderate mastitis.27 Overall, milk concentrations of Hp correlated well with serum concentrations of Hp.27

Milk concentrations of SAA > 0.2 μg/mL were present in roughly 25% of the cows without mastitis, probably representing differences in cow or sub-clinical mastitis

(although SCC were within the normal range).27 Median milk SAA concentrations in

6 cows with mild mastitis were 2.6 μg/mL (range 0.2 – 32 μg/mL), and approximately 30% of the opposite diagonal quarter SAA concentrations were > 0.2 μg/mL. Milk SAA concentrations in cows with moderate mastitis were higher than those of normal cows

(20.6 μg/mL; range 0.2 – 95 μg/mL) and 54% of the milk samples from the opposite diagonal were >0.2 μg/mL suggesting more intense inflammation than that in mild mastitis cases.27 Milk from infected quarters had significantly higher concentrations of

SAA than non-mastitic cows and were significantly greater than those of the opposite diagonal quarter. Milk SAA concentrations in cows with moderate mastitis was significantly greater than those in mild mastitis, however milk concentrations of SAA were not correlated with milk Hp, nor was it correlated with serum concentrations of

SAA.27 Somatic cell count was also not different between mild and moderate mastitis groups. The outcome of this study suggested increased serum concentrations of Hp, SAA and AGP were evidence of a systemic inflammatory response in mastitis.27 More mild inflammatory responses in the udder tend to be associated with APP responses in the udder, while more severe inflammation demonstrated inflammation present consistently in both the milk and serum. 27 Further, serum concentrations of Hp and SAA do not differentiate between mild and more severe mastitis but milk concentrations of these

APPs did.27 Acute phase protein responses differ in their ability to differentiate between acute versus chronic disease. Serum amyloid A and Hp appear to be more sensitive for acute onset mastitis, whereas AGP is useful mainly for more chronic diseases.27 Serum amyloid A performs better than Hp for milk samples and this is mainly due to endogenous lactoperoxidase interference with the Hp ELISA. While both reflect

7 inflammation and both increase over 100 fold in serum, SAA measurements appear to be more sensitive than Hp values determined from milk samples.27 This is especially true of differences in milk SAA concentrations between cows with mild versus moderate mastitis. The determination of milk SAA should be able to detect differences in mastitis severity which may be useful for the early provision of treatment of this disease.27

Of interest, the results of this study suggested the possibility that concentrations of SAA in milk reflected local production of this APP in the mammary gland.27 While on the surface this seems to be a believable interpretation of the data, more recent work has also demonstrated mammary haptoglobin production occurs during inflammation.28 The same studies of amyloid A in the mammary epithelium have not yet been performed. The presence in milk and serum of Hp or other APP produced locally may make interpretation of findings in specific disease processes more difficult and more local production in one animal than another may lead to overlapping serum or milk concentrations of this APP with a variety of clinical presentations. A biomarker specific to one organ or tissue would appear to be more reflective of disease in that organ and would be more useful in the assessment of that body system. Furthermore, local production/release of the biomarker should reflect tissue specific function and not systemic responses of the liver.

Bovine respiratory diseases

Studies examining APPs during outbreaks of natural respiratory disease caused by a variety of agents and under experimental challenges have been reported.13-16, 22, 29 One such study evaluated the serum Hp concentrations in calves naturally infected with

Bovine Respiratory Syncytial virus (BRSV).20 This study was designed to evaluate age related changes in APP in calves. Calves were housed in group pens with automated milk replacer feeders and free access to water. Calves were sampled weekly and serum was analyzed for haptoglobin (Hp), serum amyloid A (SAA), Lipopolysaccharide binding protein (LBP) and alpha-1 acid glycoprotein (AGP). However, as the calves enrolled became ill (near the 2nd week), tracheobronchial lavage samples were evaluated for cytology, microbial growth and the presence or absence of BRSV on weeks 2 and 6 and serum was evaluated by enzyme linked immunosorbent assay (ELISA) for BRSV specific IgG. Initially, it was evident that there were two groups of calves, based upon their BRSV specific serum IgG titers. Overall, those calves with higher BRSV specific

IgG titers manifested increases in SAA, Hp and LBP that peaked at week 2 of the study and gradually declined over the remaining 4 weeks of the study. Those calves with lower serum BRSV specific IgG demonstrated 2 APP (SAA, Hp, LBP) peaks, the first occurring at week 1, with serum concentrations not appreciably different than calves with higher serum BRSV specific antibody responses, and the second peak at week 3 which was higher than the mean of the high BRSV-specific IgG titer group. Serum concentrations of AGP were not different in either group and declined over the course of the study.

Of note, this study20 also documented the presence of a variety of bacterial, viral and mycoplasmal pathogens in the tracheal-bronchial secretions.20 The presence of more than one agent was commonly observed in these calves, which makes interpretation of the APP results complicated. Furthermore, although mean serum concentrations of APP

9 presented in the results demonstrated distinct peaks which were reproducible over time, the raw data presented, indicated wide variability between individuals grouped by BRSV specific IgG titer, as well as over all groups. Since the calves were grouped according to

BRSV specific, serum IgG titers, additional information would have been useful in the interpretation of these results (e.g., colostrum source and administration; serum total protein concentration in the period immediately after birth). With the multiple BRD pathogens isolated in many animals in both groups, information relative to the quality of colostral transfer and their impact on immunologic responses would have been helpful.

The investigators assumed that the second peak of SAA and LBP as well as the high concentration of Hp on week 3 was a response to secondary bacterial infection. There are also several other possible explanations for these peaks.

In contrast Heegaard et al.22 demonstrated that experimental infection with BRSV induced increased serum Hp and SAA concentrations with peak concentrations occurring

6-7 days post infection. The investigators demonstrated that aerosol and intra-tracheal inoculation resulted in increased serum SAA and Hp which began to increase at 3-4 days post-inoculation (SAA) and 5-7 days post-inoculation (Hp). Serum Hp concentration increases were dramatic with concentrations in some experimentally infected animals approximating 1 to up to 10 mg/mL.22 The animals in this study that had the fastest Hp response and highest serum concentrations of Haptoglobin were also found to be infected with Pasteurella multocida at necropsy.22 These results demonstrate that physiologic and biochemical responses to viral disease can induce APP responses and that while serum concentrations of SAA increase more rapidly, the change in serum concentrations of Hp

10 are much greater than other APP. As determined at necropsy and culture, these changes were associated with secondary bacterial infection in some animals.

Another study evaluating acute phase proteins compared experimental infection with Pasteurella (Mannheimia) haemolytica, Ostertagia ostertagia, and endotoxin administration.13 Three calves were given IV endotoxin from E. coli O55:B5 (50 ug total dose), 4 calves were inoculated with M. haemolytica by intratracheal aerosol, and 19 parasite naïve calves were given O. ostertagia (250,000 3rd stage larvae) orally. Serum levels of Hp, seromucoid (Sm), ceruloplasmin (Cp), α-1 proteinase inhibitor (PI) were measured. The ostertagia calves did not mount a measurable APP response. Following endotoxin injection, all measured APP were significantly increased to peak at days 4-5 post LPS infusion. Serum Hp concentrations demonstrated the most significant changes, peaking at day 5 and returning to 0 by day 9. Following M. haemolytica challenge, similar results were found. Haptoglobin did not increase on day 1; however, sharp increases on day 2 leading to peak concentrations on day 3 (higher than with endotoxin injection), tapering off over the next 5 days but did not reach baseline levels by the end of the study. Ceruloplasmin did not increase after M. haemolytica challenge. This study revealed that Hp is the only APP in cattle whose serum concentrations are negligible in healthy cattle and increases the most in response to inflammatory stimuli.

Other studies evaluated serum APP responses of cattle as a means of determining chronicity of disease. One study14 categorized cattle into 4 groups: healthy, acute inflammation, subacute inflammation, and chronic inflammation. Significant differences were observed in serum Hp concentrations between the acute and chronic diseases. The

11 serum Hp concentrations in cattle with chronic inflammation were significantly greater than the acute or subacute disease categories. Serum amyloid-A concentrations were significantly greater in diseased animals than in healthy animals, but there was no significant difference between disease classifications. Another study30 examined cattle antemortem and postmortem in a commercial abattoir. Cattle were categorized by the meat inspection staff and blood obtained after the cattle were killed and bled. Once the cattle were opened, the presence of disease was recorded with the antemortem findings and these findings compared to plasma concentrations of Hp and SAA.30

Retrospectively, mean serum Hp concentrations of healthy cattle were significantly lower than that observed in diseased cattle.30 Likewise, mean SAA concentrations were significantly greater in diseased cattle than in healthy cattle.

One study31 measured Hp in a group of feedlot calves as to its usefulness in predicting subsequent clinical respiratory disease. Blood was collected at feedlot entry and 40 and 65 days on feed (DOF). The highest episode of respiratory tract disease in this population immediately followed the 40 DOF sample collection and the proportion of calves observed with clinical signs increased as Hp concentration increased however the observed differences were small (P < 0.1). At 65 DOF calves with serum Hp concentrations >10 mg/dl had higher (P<.05) rate of subsequent clinical disease than calves with lower levels. Because this wasn’t significant across all time periods, and high cutoff values were needed to make Hp a useful predictor of clinical disease, it appears that measuring Hp at one time point is not very useful for predicting clinical disease.31

After a feeding period of approximately 10 months, all steers were slaughtered. Most

12 heifers were retained for replacements. The lungs of 144 steers were available for postmortem examination for the presence or absence of gross lesions. Thirty-nine percent of calves with undetectable Hp concentrations at any 3 collection times had lung lesions, whereas 63% had lesions among calves that had Hp concentrations >10 mg/dL at

1 or more time-points.31

A follow-up study29 was performed examining haptoglobin response to clinical respiratory tract disease in the same calves. Sixty of the 366 feedlot calves were observed with clinical respiratory disease during their first 65 days on feed. Calves were randomly assigned to 2 groups: “standard antibiotic treatment” or “observe only” groups.

Serum was harvested for Hp analysis at initial and final examination. Thirty-three of the calves were followed at slaughter and lung lesions were classified as either present or absent. Serum Hp concentrations were greatly variable (0-508 mg/dL) at initial examination. Haptoglobin concentrations were not different between treatment groups at initial exam but at final exam calves that were treated had significantly lower Hp concentration than the “observe only” group. Within treatment groups, serum Hp concentrations did not predict case outcomes.29 Almost all calves that received antibiotic treatment had Hp concentrations of 0 at final examination. The mean serum concentration of Hp at initial exam was numerically higher in calves that had lung lesions at slaughter but this was not statistically significant. No correlation could be found between Hp concentrations and rectal temperature, serum protein concentration, plasma pH, or hematocrit at either initial or final examination. At final exam, positive correlation was observed between haptoglobin concentration and plasma protein (P =

0.08) and fibrinogen (P < 0.05) concentrations. This data reveals that Hp concentrations is not a predictor of case outcome or severity, but can be used as an indicator of antimicrobial effectiveness.

Another study examined the utility of haptoglobin, fibrinogen (Fb), and serum protein fractions to select treatment protocols for BRD.15 The calves were treated (or not) as necessary and then retrospectively grouped into 2 treatment groups. The first group was made up of calves that required no treatment or antibiotics alone, and the second group were calves that required antibiotics and anti-inflammatory drugs.15 These authors reported statistically significant differences in serum Hp concentrations in diseased (86.0 ± 160.0 mg/L (mean ± SD)) vs. recovered (10.1 ± 16.0 mg/L) animals.

Fibrinogen (2.6 vs. 2.1 g/L), total protein (58.0 vs. 59.1 g/L), albumin (29.9 vs. 30.6 g/L),

β-1 globulins (8.2 vs. 9.3 g/L), and γ-globulins (6.1 vs. 5.6 g/L) were also reportedly significantly different between diseased and recovered, although examination of the data presented suggests this difference is small. Other measured proteins (α-1 globulins, α-2 globulins, and β-2 globulins) were not significantly different in diseased and recovered animals. Similar results were reported between groups.15 Serum concentration of haptoglobin was significantly different between animals that didn’t require anti- inflammatories and those animals that did (26.0 ± 72.0 vs. 164.0 ± 205.0). Fibrinogen, total protein, α-1 globulin, α-2 globulin, and β-2 globulin were reportedly significantly different, however, the differences between groups was small. Other measured values were not different between groups. Haptoglobin, in this study, did seem to be a predictor

14 of severity of disease, however there was high variability among calves as evidenced by the high standard deviations (SD).

Acute phase proteins have also been evaluated to determine their response to dietary energy and starch and to determine whether APP could be used as a diagnostic or prognostic tool for BRD.16 Neither dietary energy nor starch content effected fibrinogen,

Hp, or SAA concentrations.16 The calves included in this study were retrospectively placed into 3 groups: no antibiotic treatment, 1 antibiotic treatment, and more than 1 antibiotic treatment. Fibrinogen concentrations were significantly higher in calves receiving more than one antibiotic treatment than calves receiving 0 or 1 antibiotic treatment, however regression analysis showed a lack of fit between the number of antimicrobial treatments and Fb concentration.16 There was no difference between 0 treatments and 1 treatment groups. Haptoglobin concentrations increased as the number of antimicrobial treatments increased on day 0 and 7, whereas on day 14 and 28 Hp concentration was greater for the >1 treatment group compared with the 0 and 1 treatment groups. Serum concentrations of haptoglobin were elevated on day 7 for calves never treated and returned to baseline by day 14. Calves treated for BRD had elevated Hp concentrations on day 0 and 7, but fell below baseline (day 0) concentration by day 14.

This data indicates that some animals that do not require treatment may have elevated

Hp, but strengthens the data presented by Wittum et al., that Hp may be useful for determining antimicrobial efficacy, which does have practical value.29 Serum amyloid-

A, in this study, was not statistically different among treatment groups, or between morbid vs. recovered calves.16

A study was also performed to determine whether vitamin E supplementation would affect growth rate or acute phase proteins.17 Three hundred eighty seven heifer calves were randomly assigned to groups that received 2000 IU vitamin E for 0, 7, 14, or

28 days. There was no difference in ADG between groups at day 0 or 42. There were modest differences in serum concentrations of APP between groups. Serum Hp concentrations were generally lower in calves receiving Vitamin E supplementation; however this difference was not statistically significant.17 Haptoglobin concentrations decreased between days 7 and 28 in all groups indicating that Hp concentrations are variable. Consumption of vitamin E was associated with significant decrease in SAA on day 7 for all groups receiving vitamin E compared to the control group. On day 28 SAA was significantly lower in the 7 day vitamin E supplementation group than any other group. Alpha-1 acid glycoprotein (AGP) was also significantly higher in the control group than in the groups fed vitamin E for 7 and 14 days. During the trial, 225 of the 387 calves developed clinical respiratory disease and had to be treated.17 The calves were retrospectively grouped into antibiotic treatment groups 0, 1, or >1 treatments.

Fibrinogen concentration and number of treatments were not significantly correlated for any day on which samples were collected.17 A significant correlation was detected between Hp concentration and the number of treatments on day 0 but not on day 7 or 28.

This is in contrast to the findings of Berry in 200416 in which there was a significant correlation between Hp and number of treatments at day 0 and 7.16 Serum amyloid-A and AGP were not significantly different among antibiotic treatment groups at any time point.17

Based on the information presented in this review of APPs, Hp is the most reliable and consistent APP used as an indicator of inflammation in cattle. It is also clear in reviewing relevant literature that serum Hp concentration measurements have pitfalls in that they do not accurately predict the necessity of treatment, severity of disease or case outcome.29 Serum concentrations of Hp have also been shown to be elevated in some non-inflammatory (metabolic) conditions such as abomasal displacement32 and fatty liver33, 34, as well as transportation stress18.

Haptoglobin

Function

The main function of haptoglobin is to bind free hemoglobin released as a result of inflammation. The formation of very high affinity complexes between Hp and free hemoglobin (Hb) protects tissues from the negative effects of the inflammation generated through iron and hemoglobin.35 These complexes (but not Hp or Hb alone) bind to a macrophage receptors (CD163) that serve to protect host tissue through endocytosis and subsequent intracellular degradation of Hb.35 The iron associated with Hb is conserved by reducing renal losses, allowing recycling of the iron for future Hb formation.36

Recycling of iron involves sequestration of this metal within the cells for later transport to other tissues via binding to transferrin. Furthermore degradation of hemoglobin reduces vascular changes associated with inflammation (vasospasm and hypoxemia).

Other functions of haptoglobin include sequestration of iron via binding of hemoglobin, in effect preventing microbial access to this necessary growth factor. In one

17 experimental model37, prior intra-peritoneal injection of mice with haptoglobin followed by experimental infection with an E. coli strain was capable of preventing peritonitis and septicemia suggesting a role of Hp in anti-microbial activity.37 Since iron is a necessary substrate for microbial growth, rapid removal of iron rich Hb prevents growth of E. coli through sequestration of iron. Finally, this clearance is important because heme and iron released from free Hb may participate in generation of reactive oxygen species (ROS) and promote tissue injury.36 Local production of Hp within extra-hepatic tissues was shown in the bovine mammary gland after intra-mammary LPS challenge28, whereas uterine production of Hp in human females is hypothesized to be protective to the fetus, placenta and uterus through local sequestration of iron in animals with hemochorial placentation as well as immunomodulatory functions.10 Elevated serum haptologlobin in transport stressed cattle has also been shown to reduce lymphocyte proliferation in response to concanavalin A suggesting a role for haptoglobin immmunomodulation in cattle as well.38 Lactoferrin, another neutrophil granule associated protein that binds iron, has also been shown to suppress lymphocyte blastogenesis via mechanisms involving iron sequestration.39, 40

Production

Haptoglobin is produced mainly by hepatocytes in response to systemic distribution of inflammatory mediators. Haptoglobin has also been shown to be produced and stored in leukocytes, specifically neutrophils.36 Hp is synthesized at specific time periods, based upon the targeting by timing hypothesis and is packaged into gelatinase

18 granules.41 As such, Hp is expressed and produced in the later part of neutrophil development.41 This expression immediately precedes or overlaps with production of

Matrix metalloproteinase 9. Haptoglobin is stored in neutrophil granules.36 Release of

Hp from these neutrophil granules has been experimentally induced using phorbol

36 myristate acetate (PMA) and various inflammatory stimuli.

Bovine Haptoglobin

In the early 1990s bovine haptoglobin was characterized and found to have a molecular mass of 1000-2000 kDa, composed of two subunit peptides, an approximately

20 kDa peptide (α-chain) and an approximately 35 kDa glycopeptide (β-chain) linked by disulfide bonds.42 Both peptides are homologous to each counterpart of human Hp.

Highly polymerized Hp in serum is composed of 2-20 polymerized forms of α2β2 tetramer.42 Human haptoglobin has been shown to be synthesized as a single polypeptide chain which is then cleaved to generate its two characteristics α and β subunits.43

In cattle, increased concentrations serum Hp can be found in a variety of inflammatory diseases such as bovine respiratory disease13, 20 and mastitis12 as well as with endotoxin administration13 and inflammation caused by intramuscular injection of oil and turpentine32. On the other hand, Hp has also been found in a variety of non- inflammatory (metabolic) conditions such as abomasal displacement32 and fatty liver33, 34, as well as stress18. Decreasing Hp has been shown to be a good indicator of response to therapy, although it appears to be unrelated to case severity or need for treatment.29

Matrix Metalloproteinases

Matrix Metalloproteinases (MMPs) are a group of Zn++ endopeptidases, each having specific yet overlapping substrate activity.44 Collectively, the MMPs cleave most if not all of the constituents of extracellular matrix by cleaving peptide bonds in the helical region of the molecule.44, 45 MMPs can be categorized simplistically into 3 major functional groups, 1) interstitial collagenases (MMP-1 and -8) which cleave collagen types I, II, and III; 2) stromelysins (MMP-3, -10, and -11) which cleave laminin, and 3) gelatinases (MMP-2 and -9) which have highest activity towards collagen type IV.46 The collagenases selectively cleave collagen into the TCa and TCb fragments, and gelatinase degrades the TCa fragments as well as denatured collagen, i.e. gelatin. 45, 47 The gelatinases, gelatinase A (MMP-2, 72 kDa gelatinase) and Gelatinase B (MMP-9, 92 kDa gelatinase), are very similar in many characteristics; however MMP-2 does not appear to be expressed by human PMN leukocytes but is constitutively expressed by a number of other cell types and is present in circulating plasma.44, 48 Gelatinase B mRNA is not constitutively expressed in leukocytes in circulation, however is expressed during granulogenesis during bone marrow formation of leukocytes.49 Gelatinase B is stored preformed in neutrophil granules and is rapidly released upon activation of these cells during inflammation and clotting of blood.50 Gelatinase B expression is rapidly up- regulated by a variety of other cell types, including macrophages during the inflammatory process.

Gelatinases and other MMPs are produced by many cell types and have many functions. Cells that have been shown to express MMP-9 include osteoclasts51, cultured

20 synoviocytes induced with LPS52, fibroblasts and epithelial cells following exposure to pro-inflammatory mediators53, alveolar macrophages cultured with LPS54, and odontoclasts associated with bovine deciduous tooth resorption55. MMP-9 activity has been show in association with epithelialization56, 57, tumor metastasis and invasion58, 59, many aspects of reproduction60-63, and inflammatory disease such as rheumatoid arthritis64, 65, septic arthritis66 and pneumonia67. The only source in which MMP-9 is stored preformed in granules is the neutrophil.

MMP-9 from Neutrophils

Neutrophils are the major cell type attracted to the site of inflammation and are one of the first responders.8 Through activation of endothelial cells, neutrophils adhere to the capillary endothelium and migrate towards the site of inflammation.8

Transmigration across the vascular endothelium, interstitium, and epithelial layers requires proteolytic degradation of those barriers.68

Neutrophils have been shown to cause much of the early lung damage seen in calves with pneumonia.69 The landmark work of Slocombe et al. demonstrated that when normal and neutrophil depleted calves were inoculated intratracheally with P. haemolytica the calves with normal of neutrophil counts became hypoxemic within 2hrs of inoculation.69 These calves developed tachypnea, bradycardia, neutropenia, and lymphopenia. At necropsy (6 hrs after inoculation) lung lesions consisted of necrosis of the alveolar walls, intra-alveolar hemorrhage, and severe necrotizing bronchopneumonia, with accumulation of proteinaceous fluid in alveoli and lymphatics. The neutrophil

21 depleted calves showed no signs of respiratory disease and mild histologic changes were observed at necropsy.

Polymorphonuclear leukocytes (neutrophils) were shown to be responsible for basement membrane degradation as early as the mid 1960s.47 Two neutral metalloproteinases, collagenase and gelatinase, were isolated from human neutrophils in the early 1980s.45, 70 Both were originally thought to be localized to the neutrophil’s specific granules71, 72, however a later study confirmed the existence of gelatinase-rich granules that are lighter and mobilized more easily than specific granules.73 Gelatinase B was subsequently found to be a major factor human neutrophil migration across basement membranes.74

Several studies demonstrate that MMP-9 expression from other cellular sources depends upon up-regulation of gene expression as opposed to the release of preformed, stored enzyme. In the first study, human monocytes were stimulated with lipopolysaccharide (LPS, endotoxin) and phorbol ester (phorbol 12-myristate 13-acetate

(PMA)).75 The results revealed that MMP-9 production is subject to regulation by cytokines and other physiologic reagents and is regulated transcriptionally by mRNA production and post-transcriptionally by stabilization of the mRNA. Secretion of MMP-9 from these stimulated monocytes was not detectable until 12 hours post PMA stimulation. Another study76 examining spontaneous bovine B-cell lymphosarcoma cell lines infected with Theileria annulata, demonstrated that MMP-9 activity was induced due to increased mRNA levels. MMP-9 was not detected in the uninfected cell line, suggesting that activation of latent proteinase is not an element in the observed induction.

The most compelling evidence, however, was presented in a study77 in which endotoxin

(LPS) and other inflammatory mediators, such as TNF, IL-8, and granulocyte colony- stimulating factor were used to induce a rapid (within 20 minutes) release of gelatinase-B

(MMP-9) zymogen in whole human blood.78 The polymorphonuclear neutrophil was identified as the cell responsible for this rapid secretion, as a result of the release of preformed enzyme stored in granules.77 More than 7 hours was required to detect low levels of proMMP-9 secretion in supernatants from stimulated monocytes.77 In this same study, human volunteers exposed to LPS (IV injection) demonstrated measureable increases in serum MMP-9 concentration similar to that seen with experiments performed in whole blood ex vivo. This study also included 2 case reports in which proMMP-9 and proMMP-2 were increased in 2 patients with natural gram negative sepsis.77

A 92 kDa gelatinase (MMP-9) was found to be secreted constitutively by bovine neutrophils and the secretion can be enhanced by PMA.78 To confirm that the MMP-9 released from bovine neutrophils was due to degranulation, not from de novo synthesis, bovine neutrophils were also treated with PMA in the presence of protein synthesis inhibitors. These inhibitors did not inhibit the secretion of MMP-9 by bovine neutrophils in vitro.78

Hp-MMP-9 Complex

Purified bovine neutrophils stimulated with phorbol ester showed a rapid (10 min) release of MMP-9, confirming that preformed MMP-9 is present in bovine neutrophils.78

These MMP-9 species were characterized and found to be the expected 105 kDa pro-

MMP-9 monomer and the 210 kDa pro-MMP-9 dimer.79 In addition, several gelatinolytic bands in the range of 300-500 kDa and >500 kDa were observed. The larger

MMP-9 molecules were identified by SDS-PAGE, immunoblot, and peptide mass spectrometry to be MMP-9 complexed with alpha and beta Hp.79 The MMP-9 molecule found within the MMP-9-Hp complex is the pro-enzyme, but once activated possessed the same activity as the Hp-free MMP-9. This study also showed that MMP-9-Hp complex also retains the ability to bind hemoglobin (Hb), thus the complex retains the biochemical properties of both protein constituents.79 These MMP-9-Hp complexes were also shown to be present in sera of cows with inflammatory conditions but absent in healthy cows.79

A recent study looking at Hp, MMP-9, and Hp-MMP-9 complex in serum of animals with acute or chronic inflammation compared to normal animals demonstrated that while free Hp and free MMP-9 were not reliable indicators of acute inflammation, the Hp-MMP-9 complex is absent in healthy cattle, greatly increased in acute inflammation, and much lower in chronic inflammation.80

Chapter 2: The Utility of Hp-MMP 9 Complex ELISA vs. Hp ELISA in Predicting Bovine Respiratory Disease in Feedlot Cattle

Introduction

Bovine respiratory disease (BRD) is the leading cause of morbidity and mortality in feedlot cattle in the United States.2 In BRD, elevated serum Hp concentrations have been determined to be a reliable indicator of the response to appropriate therapy;29 however, several studies demonstrate elevated serum Hp concentrations appear to be unrelated to case severity or need for treatment.16, 29, 31

The majority of measurable serum haptoglobin, in published reports involving experimentally induced or natural diseases, is thought to be due to hepatocyte production in response to systemic distribution of inflammatory mediators. In a classic model of inflammation in cattle, intramuscular injection of turpentine results in severe inflammation locally, with concomitant increases in serum haptoglobin concentrations.81

This model has been used for the purification of bovine haptoglobin, for the production of purified protein standard, for diagnostic testing, and the preparation of haptoglobin antibodies. Unfortunately, commercially available polyclonal antiserum recognizes haptoglobin in addition to proteins bound to it (including MMP 9) since the antigen injected into rabbits or goats was a mixture of haptoglobin and haptoglobin in complex with MMP 9. Unless these antisera are affinity purified, detection of MMP 9 by

25 antibodies to this protein in the antisera, or to Hp-MMP 9 on immunoblots, IP or ELISA is observed (Lakritz unpublished data).

Haptoglobin has also been shown to be produced and stored in leukocytes, specifically neutrophils.36 Neutrophil Hp is synthesized at specific times. Based upon granule protein targeting by timing hypothesis, the expression of Hp overlaps with expression and packaging of neutrophil gelatinase.41 Since production of gelatinase granules (during the band and neutrophilic stages of these cells) overlaps with production of MMP 9, and MMP 9 forms dimmers with itself (due to lack of TIMP-1 expression), this may be one reason for the presence of Hp-MMP 9 complexes within the cell.

Release of Hp from these neutrophil granules has been experimentally induced using

PMA and various inflammatory stimuli.36

A haptoglobin matrix metalloproteinase 9 (Hp-MMP 9) complex was recently found to be released during bovine neutrophil degranulation in vitro.79 High molecular weight molecules with gelatinolytic activity were discovered and further characterized as

MMP 9 bound to Hp.79 These molecules were subsequently found to be present in serum of cattle with inflammatory disease but absent in healthy cattle.79 Further study revealed that when clinical cases were classified as healthy, acute inflammation, or chronic inflammation, Hp-MMP 9 was significantly higher in cattle with acute inflammation compared to those with chronic inflammation and was absent in serum of healthy cattle.80

Total Hp, although absent in healthy cattle, was not significantly different between acute and chronic groups.

We hypothesized that Hp-MMP 9 can be used as an indicator of early inflammation and therefore a predictor of subsequent disease in feedlot cattle. The current study was designed to determine whether Hp-MMP 9 could be used in the prediction of BRD in recent admissions to feedlots. Further, using health, treatment and weight gain data, our aim was to determine whether Hp-MMP 9 could predict which calves would be identified with clinical respiratory disease and would require therapy in the days following sample collection. We compared serum concentrations of Hp to Hp-

MMP 9 to assess how well the complex performed in these animals.

Materials and Methods

Study population – The study population consisted of 393 calves that entered the

Ohio Department of Corrections and Rehabilitation feedlots between November 7, 2008 and February 10, 2009. The calves were purchased by the feedlot owner in groups of approximately 130 animals at 3 time points. The calves were delivered to the feedlot, admitted and allowed to acclimate for 24 hours prior to processing. Processing included determination of post-arrival body weight, placement of eartags and electronic identification tags (EID), and collection of blood for harvesting serum. On admission, the average weight of calves was 755.6 ± 101.6 lbs (mean ± SD). Weights and serum were collected on 2 additional times points throughout the feeding period at 31-55 day intervals and the final sample was obtained once the calves reached >1000 lbs; calves were slaughtered at this time. Average total feeding period length for this group of animals was 136 ± 74 days.

Data collection – The study design included calf weights, with blood sample acquisition at approximately 40 day intervals. In addition to body weights and serum concentrations of Hp, Hp-MMP 9, complete medical records including diagnosis and treatment (if any) were to be collected and analyzed in terms of the weight gain, medical diagnosis and treatment history over the entire feeding period. Unfortunately, access to medical records was prohibited. All samples (3-4 per calf) collected from the first 65 calves entering the feedlot were analyzed for Hp and Hp-MMP 9 concentrations (~290 samples/out of >1,400 samples).

Serum haptoglobin determination – Serum Hp concentrations were determined using commercial bovine haptoglobin ELISA test kits (Bovine haptoglobin 96-well

ELISA, Life Diagnostics, West Chester, PA 19380) according to manufacturer’s instructions. Standard curves were prepared using purified bovine haptoglobin standard

(2.5 µg/mL) included with the kit at a concentration range from 7.8 to 250 ng/mL. Serum samples were diluted according to the kit instructions (1:2000 dilution).

Serum Hp-MMP 9 determination – The ELISA for bovine Hp-MMP 9 was modified from one developed in the laboratory, exploiting un-conjugated bovine MMP 9 monoclonal antibody (MAb) 10.1 as a capture antibody (100 µL, containing 2 µg per well in tris buffered saline (TBS) + 0.1% BSA (Bovine Serum Albumin, Fraction V,

Fisher Scientific, Thermo Fisher Scientific, Pittsburgh, PA 15275)) and affinity purified

HRP-conjugated rabbit anti-bovine haptoglobin (Immunology Consultants Laboratory,

RHPT-10P, Newberg, OR, 97132) was used at a final concentration of 0.033µg/mL in

TBS with 1mg/mL BSA as a detection antibody. This ELISA was modified to increase

28 the sensitivity of the assay by addition of a third antibody (goat anti-rabbit HRP; 1:40,000 final dilution). These concentrations of both capture and detection antibodies were chosen on basis of preliminary experiments. Bovine serum albumin, 20mg/mL in TBS or a commercial blocking buffer (Pierce Super Block Blocking buffer, Thermo Fisher

Scientific, Rockford, IL 61105) was used as a blocking agent. Further, since we did not have access to purified bovine Hp–MMP 9 complexes, we used serial dilutions of serum containing known concentrations of Hp-MMP 9 complexes to produce linear standard curves. Based upon the analysis of 6 standard curves, the assay was linear over the range of 0.45 ng/mL to 24 ng/mL. Controls included were normal bovine serum, 5% BSA in

TBS and blank wells.

Statistical analysis – All analysis were carried out using the GraphPad Prism software (Prism 5 for Windows, version 5.01). Correlations between Hp and Hp-MMP 9 were determined using a Pearson production moment correlation. The differences between 2 groups was determined using t test, whereas the differences between more than

2 groups were determined using a one-way ANOVA with Tukey’s multiple comparison post test. The differences was considered significant for p<.05.

Results

A total of 248 time-points from the 65 calves were analyzed for Hp and Hp-MMP

9. One hundred seventy-six (176, 71%) time points had detectable Hp-MMP 9 with Hp concentrations of 0. Only one time-point had a detectable Hp with Hp-MMP 9 complex concentration of 0. There no correlation between Hp and Hp-MMP 9 (r2 = 0.064) in all

29 time-points (Figure 1). A correlation was detected when Hp=0 was removed from the data set (r2 = 0.337) (Figure 2). Haptoglobin-MMP 9 complex concentrations were significantly higher at time-points where Hp > 0 as compared to time-points in which Hp

= 0 (P = 0.0065) (Figure 3).

The ADG for all (393) calves was 1.4 ± 0.8 lbs (mean ± SD) for the entire feed period. The ADG for the 65 calves in which serum was analyzed was 1.7 ± 0.5 lbs (mean

± SD). Forty (40, 61.5%) of the 65 calves had Hp concentrations of 0 at all time-points

(Hp 0). Twenty-two (22) calves had elevated Hp concentrations at 1 time point (Hp 1).

Seven (7) calves had elevated Hp concentrations at >1 time point (Hp >1) (Table 1).

There was no significant difference in ADG between groups based on number of positive

Hp time-points (P = 0.0691) (Figure 4). Only 2 calves had no detectable Hp-MMP 9 complex at all time points (Complex 0), 4 calves had elevated Hp-MMP 9 complex at 1 time-point (Complex 1), seven calves had elevated Hp-MMP 9 complex at 2 time points

(Complex 2), and 51 calves had elevated complex at >2 time-points (Complex >2)(Table

2). There was no significant difference in ADG between groups based on number of positive Hp-MMP 9 time-points (P = 0.6031) (Figure 5).

Individual sample collections were also grouped based on Hp concentration and

Hp-MMP 9 concentration and the ADG for the period following sample collection was compared. Samples were first grouped based on Hp concentration. The number of samples in each cutoff group and mean and SD are provided in Table 3. There was a significant difference (p<.05) between in ADG following Hp concentrations that were

>100 mcg/ml compared to the other groups; however, there were only 3 samples in this

30 group (Figure 6) of which one of the samples was severely negative (-3.53 lbs). The number of samples in each cutoff group and mean and SD for the Hp-MMP 9 grouping is presented in Table 4. There were no significant differences in ADG following collections between groups based on Hp-MMP 9 concentrations (Figure 7).

Discussion

The major goal of this study was to compare Hp-MMP 9 data with disease and treatment records and to determine whether the presence of this marker was associated with reduced ADG or development of BRD. Unfortunately no health or treatment data was provided to us at the completion of this study.

Our data suggests that serum Hp-MMP 9 complex is present independent of Hp produced by the liver. This was evidenced by the large number of time-points that had elevated Hp-MMP 9 complex with no detectable free Hp. Figures 1 and 2 indicate that there was some correlation between Hp and Hp-MMP 9 when time-points that have Hp concentrations of 0 were removed. Figure 3 shows that Hp-MMP 9 concentration is significantly higher at time-points in which Hp was >0. Only one time-point was positive for Hp-MMP 9 and had a concentration of 0 on the Hp ELISA. Since the Hp ELISA detects total serum Hp, which includes Hp that has MMP 9 bound to it, we cannot state how much of the serum Hp determined by the ELISA was actually free Hp. The instances where Hp-MMP 9 was positive but the Hp was negative was likely because the

Hp in the complex at these points was below the limit of detection of the Hp ELISA.

Since original estimates suggested a ratio of 1 MMP 9 to 13 alpha and 13 beta

Haptoglobin molecules present within these complexes, and there are more than 6 different Hp-MMP 9 complexes (by molecular mass as determined sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE)) it is difficult to estimate proportionally how much Hp is present within Hp-MMP 9 complexes.79

9 complexes since all samples evaluated were from serum and many had Hp-MMP 9 complex concentrations of 0. Determination of serum concentrations of MMP 9 would be helpful in this regard. In a recent study, serum samples from apparently healthy cattle were analyzed for serum MMP 9 and Hp-MMP 9 concentrations. In all of these animals, serum concentrations of Hp-MMP 9 were 0; whereas serum concentrations of MMP 9 were similar to that observed in the diseased animals.80

In conclusion, our results clearly show that serum Hp and Hp-MMP 9 complex are present in cattle admitted to a feedlot. It appears that the presence of these two analytes are independent; however, due to the inability to obtain health and treatment records, we were not able to draw any conclusions about Hp-MMP 9 as a predictor for respiratory or other diseases or reduced ADG in feedlot cattle.

Chapter 3: Serum Hp-MMP 9 Complex in Cattle with Clinically Defined Diseases

Introduction

Bovine respiratory disease (BRD) and other acute diseases continue to have a major impact on livestock productivity in the USA. Recent NAHMS studies suggest

BRD is a leading cause of illness and death in US feedlots.1 Under field conditions, the use of acute phase proteins (APP) to detect animals requiring treatment and animals developing lung lesions has demonstrated some utility. For example, haptoglobin (Hp) responses to inflammation in cattle have been evaluated in acute bronchopneumonia15, 29,

34, 42, acute rumen acidosis82, coliform mastitis83, 84, hepatic lipidosis85, and transport stress38, 86. Serum concentrations of Hp in acutely ill cattle increase (>100-fold) reaching maximum concentrations between 48 and 96 h.27, 28, 34, 42, 87 However, serum Hp is less useful as an indicator of morbidity than an indicator of clinical resolution of disease in calves with BRD.15, 16 The observed modest increases of serum Hp in cows with chronic illnesses, that do not necessarily have clinically apparent signs of inflammation, diminishes this diagnostic test’s value as an indicator of inflammation.33, 88 In previous work, we identified covalent, heteromeric complexes of Hp and matrix metalloproteinase

9 (MMP 9) in neutrophil granules and the serum of cattle with acute septic inflammation of the abdomen or thorax.79 In most of these cases, polymicrobial sepsis was evident.79

In contrast to free Hp, which is produced mainly by the liver during inflammation, serum 33

Hp-MMP 9 complexes are believed to be released from neutrophils. As such, these complexes should represent systemic neutrophil activation.79

We hypothesize that Hp-MMP 9 complexes, produced by neutrophils in vitro and found in acute phase sera, have specific functional significance that differs from un- complexed forms of free Hp and MMP 9 alone and as a consequence, serum concentrations of Hp-MMP 9 might serve as an independent indicator of clinically important events occurring during acute inflammation.79 Most likely, these events precede those associated with liver production of Hp in response to mediators released by cells such as the macrophage.54 We believe serum Hp-MMP 9 complex concentrations reflect systemic activation of neutrophils. The objective of the present study was to evaluate the utility of Hp-MMP9 complex ELISA in comparison to ELISA for total Hp or

MMP 9 alone as indicators of acute septic inflammatory disease in cattle.

Materials and methods

Animals – Animals enrolled in the study (n = 35) were those admitted to The

Ohio State University VTH for evaluation of a variety of disorders (22) or sampled in a local dairy (13). Animals were either bled during routine clinicopathologic work-up or as part of a herd infectious disease survey. Serum was harvested in routine fashion by centrifugation after coagulation and frozen at −20 ◦C until analyzed. Animal age

(3.5±1.7 years; range 0.4–8 years), breed (Holstein = 20, Jersey = 8, Hereford = 2,

Guernsey = 1, Angus = 3; Brown Swiss = 1), sex (M= 3, F = 32) were obtained from patient and/or herd records (Tables 5, 6 and 7). Cattle were grouped into acute septic

34 disease (n = 15), chronic disease (n = 10) and normal animals (n = 10) based upon clinical signs, history and the results of clinico-pathologic and pathologic evaluation in those animals that died or were euthanized. As these animals were examined by hospital veterinarians, they were classified as to their disease category prior to the analysis of the serum proteins by ELISA. The individual analyzing the samples by ELISA was not aware of the individual animal’s identity or disease classification and was only given frozen serum samples with the animal’s laboratory designation. The animals selected for inclusion were done so under the guidance of The Ohio State University VTH Hospital

Executive Committee with approval of The Ohio State University Institutional Animal

Care and Use Committee.

Purification of MMP 9 molecular species – Purification of bovine MMP 9 monomer, dimer and Hp-MMP 9 complexes was performed as described previously.79 In brief, neutrophils from fresh cow blood were stimulated with phorbol myristate acetate

(PMA) and conditioned media was subjected to reactive red 120 agarose (Sigma) chromatography, gelatin agarose affinity chromatography and Ultragel AcA 34 gel- filtration. Fractions were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS PAGE), pooled, concentrated and stored at −20 °C.

Monoclonal antibody production – A mixture of affinity purified MMP 9 monomer and dimer produced by bovine neutrophils was used for immunization of

BALB/C mice. Mice were immunized twice with 100 µg of purified MMP 9 over a 7- day interval. Four weeks later, mice were boosted with 100 µg of MMP 9 and after 4 days, sacrificed. Harvested splenocytes were fused to the B cell myeloma SP2/0 cells

35 and plated in 96-well plates. Hybridomas were selected in the presence of 100 µM hypoxanthine, 0.4 µM aminopterin, 16 µM thymidine following established methods.

Clones that secreted antibodies with specificity for MMP 9 as tested by ELISA were further cloned by limiting dilution twice in order to ensure clonality.89 Clone 10.1 was selected for positive reactivity and cellular amplification in a bioreactor was performed as described (Celline, CL1000, Sartorius Stedim, North America, Inc., Bohemia, NY

11716). The reactivity of the hybridoma supernatant was tested using Western immunoblot of purified monomer and dimer under reducing and non-reducing conditions and with enzymatically activated MMP 9 monomer and dimer. In addition, whole cell lysates from bovine neutrophils and from peripheral blood leukocytes were immunoblotted with the monoclonal antibody (MAb). Immunoglobulin class (IgG1) of the MAb was determined using a commercial MAb isotyping kit (Pierce Rapid ELISA

Mouse mAB isotyping kit, Thermo Fisher Scientific, Rockford, IL 61105).

IgG from the hybridoma media was purified with Protein G Agarose (Pierce

Protein G Agarose, Thermo Fisher Scientific, Rockford, IL 61105) and stored at a concentration of 1mg/mL at 4 ◦C.

The anti-bovine MMP 9 monoclonal antibody (Clone 10.1) reacted positively in

Western blot with bovine MMP 9 monomer and dimer whether reduced or not and after enzymatic activation. The monoclonal antibody does not react with MMP 2 or other proteins in any of the immunoblots (Bannikov, 2009, unpublished).

Horse radish peroxidase (HRP) conjugation of 10.1 antibody – Approximately

1mg of affinity purified monoclonal antibody 10.1 was conjugated to HRP with the

Roche HRP protein labeling kit (Peroxidase Labeling Kit, Roche Diagnostics

Corporation, Indianapolis,IN 46250) in accordance with the manufacturer’s instructions.

The resulting antibody HRP conjugate was purified by gel-filtration on Ultragel AcA 34

(Sigma U8878 – Ultrogel AcA 34, Sigma-Aldrich Co., St. Louis, MO 63178) and stored at 4 °C at a concentration 60 µg/mL.

Serum Hp ELISA – Total serum Hp concentrations were determined using commercial Bovine haptoglobin ELISA test kits (Bovine haptoglobin 96-well ELISA,

Life Diagnostics, West Chester, PA 19380) according to manufacturer’s instructions.

Standard curves were prepared using purified bovine haptoglobin standard (2.5 µg/mL) included with the kit at a concentration range from 7.8 to 250 ng/mL. Serum samples were diluted according to the kit instructions (1:2000 dilution) or used at higher concentration for samples containing low concentration of Hp and were run in duplicate.

Controls included were normal bovine serum, 5% bovine serum albumin (BSA) in tris buffered saline (TBS) and blank wells.

Serum MMP 9 ELISA – The ELISA for bovine MMP 9 was developed in the laboratory, exploiting un-conjugated bovine MMP 9 MAb 10.1 as a capture antibody

(100 µL, containing 1 µg per well in TBS) and HRP-conjugated 10.1 antibody (0.3

µg/mL in TBS, 1mg/mL BSA) for detection of MMP 9. These concentrations were chosen on the basis of preliminary experiments. After capture antibody binding, the plates were washed 3 times for 5min with TBS and the wells blocked with bovine serum albumin (Bovine Serum Albumin, Fraction V, Fisher Scientific, Thermo Fisher

Scientific, Pittsburgh, PA 15275) at a concentration of 20mg/mL in TBS or with Super

Block Blocking Buffer (Pierce Super Block Blocking buffer, Thermo Fisher Scientific,

Rockford, IL 61105). After washing the blocked wells, and addition of serially diluted serum samples or known concentrations of purified bovine MMP 9 monomer (60 min at

22 °C), the plates were again washed 3 times for 5min with TBS and HRP-conjugated

MAb (10.1) added to each well at a final dilution of 1:200. The excess HRP conjugate was removed by washing after which 100 µL of 3,3',5,5'-tetramethybezidine (TMB) was added to each well for determination of HRP activity. The wells were incubated and colorimetric development followed over time using a microtiter plate reader (λ = 630 nm). When color developed was maximal in the highest concentration standards, the reaction was stopped by addition of 1N hydrochloric acid. The color development was then determined using λ = 450 nm. Quantitation of MMP 9 protein was performed using purified bovine MMP 9 protein as a calibration curve covering a range of concentrations from 39 to 2500 ng/mL. Samples registering <39 ng/mL were considered to be zero.

Serum samples from sick and healthy animals were analyzed by serially diluting of each sample with TBS containing 1 mg/mL BSA in duplicate. Once each set of samples were analyzed, the absorbance value from the dilution which lay within the linear range of the standard curve was used to quantitate MMP 9 concentrations.

Controls included were normal bovine serum, 5% BSA in TBS, and blank wells.

Serum Hp-MMP 9 complex ELISA – Purified MAb 10.1 was used as a capture antibody as described above (100 µl, containing 1 µg per well). However, the detection antibody chosen was an affinity purified HRP-conjugated rabbit anti-bovine haptoglobin

(Immunology Consultants Laboratory, RHPT-10P, Newberg, OR, 97132) used at a final

38 concentration of 0.033 µg/mL in TBS with 1mg/mL BSA. These concentrations of both capture and detection antibodies were chosen on basis of preliminary experiments

(unpublished). Bovine serum albumin (20mg/mL in TBS) or a commercial blocking buffer (Pierce Super Block Blocking buffer, Thermo Fisher Scientific, Rockford, IL

61105) was used as a blocking agent. Purified bovine Hp-MMP 9 complexes were used to create standard curves using concentrations of 40–5000 ng/mL. Samples registering lower than 40 ng/mL were considered to be 0 ng/mL. Once each set of samples were analyzed, the absorbance value from the dilution which lay within the linear range of the standard curve was used to quantitate Hp-MMP 9 concentrations. For analysis sera were serially diluted with PBS, containing 1mg/mL BSA and run in duplicate. Controls included were normal bovine serum, 5% BSA in TBS, and blank wells.

Specificity of serum Hp-MMP 9 ELISA – Affinity purified bovine Hp, MMP 9 monomer and Hp-MMP 9 complexes obtained as described above were used. In plates where the capture antibody was the anti-bovine MMP 9 monoclonal antibody (Clone

10.1), increasing concentrations of Hp, MMP 9 and Hp-MMP 9 complexes (over the range of 4.8–5000 ng/mL) were added to separate wells and allowed to bind. After washing, HRP-conjugated secondary antibodies were added as follows: (1) To wells where Hp was added, anti-bovine Hp HRP conjugate or anti-bovine MMP 9 HRP conjugate was added. (2) To wells where affinity purified MMP 9 was added, anti- bovine Hp HRP conjugate or anti-MMP 9 monoclonal antibody HRP conjugate was added. (3) Finally, to wells where Hp-MMP 9 complexes were added, anti-bovine Hp

HRP conjugate or anti-bovine MMP 9 HRP conjugate was added. After addition of TMB

39 substrate, quantitation of antibody binding by determination of absorbance of each well was determined by the plate reader at 450 nm.

Statistical Analysis – Animals were categorized based upon the findings of clinical examination, laboratory or other diagnostics and postmortem examination and designated as acute septic disease = 1; chronic or metabolic disease = 2; or normal = 3.

ELISA assays were performed by one of the authors (GB) without prior knowledge of the disease classification. Individual animal data for each of the 3 ELISA assays was entered into a spreadsheet spreadsheet by animal ID (animal # 1–35). Data from animals in each of the classifications, and for each of the 3 ELISA assays were then described statistically

(mean, standard deviation, minimum, maximum and median). Since the data were not normally distributed (zero’s in normal and chronically ill animals for haptoglobin and haptoglobin-MMP 9 complex), nonparametric methods (Kruskal–Wallis ANOVA on ranks) were used to compare results from each of the 3 ELISA in each of the 3 groups.

Further examination of the data included pair-wise comparisons of the 3 ELISA assays using the Wilcoxon rank-sum test. To account for multiple (n=3) pair-wise comparisons, a Bonferoni adjustment was made and p-value < 0.0167 (=0.05/3) was considered statistically significant.

Results

Analytical sensitivity of both MMP9 ELISA and Hp-MMP 9 complex ELISA used in this study was 39 ng/mL. Analytical specificity of these assays is corroborated by the fact that irrelevant molecules (MMP 9 and Hp for Hp-MMP 9 complex ELISA and

Hp-MMP 9 complex for MMP 9 ELISA) were not detectable up to concentrations 5000 ng/mL (Figure 8).

The history and the results of clinico-pathologic and pathologic evaluation of animals used in this study together with corresponding values of Hp, MMP 9 and Hp-

MMP 9 complex in their sera are presented in Tables 5, 6, and 7.

Graphic representation of the distribution of the ELISA data in the three health status classifications are portrayed in Figure 9. Few healthy animals (disease classification 3) had measurable serum Hp concentrations (8/10 were “0”), whereas serum Hp was high in acute septic (disease classification 1) and chronic inflammation/metabolic disease (disease classification 2) groups (Tables 5 and 6, Figure

9). There was considerable overlap between Hp concentrations in serum of acute and chronic inflammation/ metabolic disease animals (Figure 9).

Median and 25th and 75th percentiles for serum haptoglobin concentrations for acutely septic animals, animals with chronic or metabolic disease and for healthy animals are presented in Table 8. The serum haptoglobin concentrations in animals with acute inflammation (median 415 µg/mL; 25th–75th percentile 302, 680) and chronic inflammation/metabolic disease (median 178 µg/mL; 25th–75th percentile 131, 567) were not significantly different from each other (p = 0.1655). However, serum Hp concentrations in acute septic diseases and chronic inflammation/metabolic diseases were significantly greater than those observed in healthy cows (0 µg/mL; 25th–75th percentile

0, 0; p < 0.0001, p = 0.005, respectively).

All normal and 8 of 10 animals with chronic inflammation/ metabolic disease had no measurable serum Hp-MMP 9 complexes. Two animals with chronic inflammation/ metabolic disease had elevated serum Hp-MMP 9 concentrations (Table 6, Figure 9).

One such animal was diagnosed with thymic lymphoma and the other possibly had acute disease that was not detected clinically. In acute septic animals, 14/15 animals possessed high serum concentrations of Hp-MMP 9 complexes (Table 5, Figure 9).

Median and 25th and 75th percentiles for serum Hp-MMP 9 complexes in animals with acute septic disease (median 1014 ng/mL; 25th, 75th percentiles: 477, 1392 ng/mL) were statistically greater than those observed in animals with chronic inflammation/metabolic disease (median 0 ng/mL; 25th, 75th percentiles: 0, 0 ng/mL; p <

0.0031) and in normal animals (median 0 ng/mL; 25th, 75th percentiles: 0,0 ng/mL; p <

0.0001) (Table 8). The differences in serum concentrations of Hp-MMP 9 complexes between normal or chronically diseased animals were not statistically significant (p =

0.1468).

The range of serum MMP 9 concentrations was much wider among animals with acute septic inflammation than among healthy animals and animals with chronic inflammation/metabolic disease (Table 8). There was, however, considerable overlap between the diseased animals (Table 7, Figure 9). Median and 25th and 75th percentiles for serum free MMP 9 in animals with acute septic (median 2745 ng/mL; 25th, 75th percentiles: 1115, 3871 ng/mL), and with chronic/metabolic disease (median 2233 ng/mL; 25th, 75th percentiles: 1625, 2910 ng/mL) were significantly greater than those measured in healthy cattle (median 330 ng/mL; 25th, 75th percentiles: 153, 573 ng/mL; (p

= 0.0011 and p = 0.0019, respectively) (Table 8). However, serum concentrations of

MMP 9 in animals included in the acute disease and chronic/metabolic disease categories were not significantly different from each other (p = 0.5417) (Table 8).

Discussion

This study compared the diagnostic utility of serum concentrations of matrix metalloproteinase 9 (MMP 9), haptoglobin (Hp) and haptoglobin-matrix metalloproteinase 9 (Hp-MMP 9) complexes in cattle with acute septic disease or chronic inflammatory/metabolic disease to normal animals. This analysis was prompted by our previous finding that Hp-MMP 9 in conditioned media from phorbol ester stimulated bovine neutrophils in vitro and the observation of the presence of Hp-MMP 9 in serum of septic cattle but not in serum of chronically inflamed or healthy cows.79 The design of this study also included measurement of serum concentrations of two individual components of Hp-MMP 9 complex: Hp, which is a major acute phase protein in cattle, and MMP 9, which concentrates in serum of cattle undergoing inflammation have not yet been investigated.

In order to conduct the present study, we developed a monoclonal antibody to bovine MMP 9 and an ELISA assay that specifically recognizes either MMP 9 or Hp-

MMP 9 complexes. As determined experimentally, the Hp-MMP 9 ELISA does not recognize free serum MMP 9 or free serum Hp as individual molecules (Figure 8).

Furthermore, we confirmed that the MMP 9 ELISA does not recognize Hp-MMP 9 complexes in serum (Figure 8). Finally, serum Hp-MMP 9 complexes in cattle with

43 acute septic disease do not distort data for Hp–ELISA because these complexes are not recognized by the Hp ELISA, and the serum concentrations of Hp-MMP 9 complexes are present in concentrations that are 3 orders of magnitude lower than the concentrations of

Hp.

We previously showed that MMP 9 is stored pre-formed in granules in peripheral blood neutrophils.79 This was in agreement with previously published studies.78, 90, 91 In the current study, MMP9 was clearly elevated in acutely or chronically ill animals in comparison to clinically normal cows (Tables 5–7, Figure 9). On the other hand, our data demonstrated that serum concentrations of MMP 9 in chronic inflammatory/metabolic disease and acute septic disease of cattle are nearly identical; suggesting that serum MMP

9 is not suitable for differentiation of acute and chronic inflammatory or metabolic diseases (Figure 9). At this point, we also cannot recommend the use of serum MMP 9 concentrations as an acute phase protein since detection of this protein could lead to a false-positive diagnosis due to substantial variation of serum MMP9 concentrations even in healthy cattle. Further study in additional animals is warranted, especially in light of our finding that serum MMP 9 in healthy animals appears to be common. As others have suggested, the lack of clear-cut boundaries between concentrations of MMP 9 in all three groups of animals studied may be related to dual roles of MMP 9 in inflammation favoring either pro-inflammatory or anti inflammatory activity.90-92

Although Hp is recognized as an indicator of acute inflammation in cattle, moderate increases of serum Hp have been described for cows with hepatic lipidosis, despite having no clinically apparent signs of inflammation.33, 34 Further, substantial

44 variation in Hp concentrations in milk has been observed in cows with chronic, sub- clinical mastitis.88 We have confirmed results from previous studies demonstrating that

Hp concentration in sera of healthy cows is negligible (approximately 20 ng/mL or lower) but can increase in acute inflammation to values as high as 950 µg/mL.

On the other hand, a majority of the cows with chronic inflammation/metabolic diseases in our study also have considerable concentration of Hp in their sera. Because of the lack of statistical difference between Hp concentrations in chronic inflammatory/metabolic disorders and acute septic disease, we propose that serum Hp concentrations, by themselves, are a poor discriminator between acute and chronic inflammation in cattle.

Data from our current study demonstrated a significant diagnostic advantage of the Hp-MMP 9 ELISA over the Hp and MMP 9 ELISA assays. There are significant differences in serum Hp-MMP 9 concentrations observed in cattle with acute septic disease compared to those animals with chronic inflammatory/metabolic disease or healthy animals. Our data suggest that the Hp-MMP 9 assay is specific for acute, septic diseases. Of the 15 animals identified clinically as having acute inflammation, 14 animals had high serum concentrations of Hp-MMP 9. Only 1 of these 15 animals had serum concentrations of Hp-MMP 9 complexes <39 ng/mL (interpreted as 0). This cow

(Table 5, # 8) had recently calved and developed Klebsiella spp. mastitis. Serum concentrations of Hp and MMP 9 were within the ranges of those observed in the serum of both disease category 2 and 3 animals. This animal had been treated with Predef2XTM

(isoflupredone acetate) for 3 days. We suggest the lack of serum Hp-MMP 9 complexes

45 may be related to previous corticosteroid treatment since corticosteroids have profound effects on inflammation in vitro and in vivo.93

The specificity of Hp-MMP 9 complex ELISA for acute septic condition is further corroborated by its detection of the complex in only 2 of 12 chronically ill animals. One of these two cases (Table 6, #16) was diagnosed with thymic lymphoma, without any pathologically described necrosis or bacterial infection. The presence of the Hp-MMP 9 complexes in cattle with neoplasia should be investigated. A second Hp-MMP 9 positive animal was placed into disease classification 2 based upon clinical findings (Table 6, #

19). Although it did possess serum Hp-MMP 9 complexes, the lack of diagnostic testing performed on this particular animal prior to discharge precludes an explanation of the presence of these serum Hp-MMP 9 complexes.

Another favorable feature of the Hp-MMP 9 ELISA was the narrow range of concentrations in the sera of acutely septic animals: median concentrations of Hp-MMP 9 in animals classified clinically as acute septic were 1014 ng/mL (with 25th, 75th percentiles ranging from 586–1373 ng/mL). Median values for the chronic inflammation/metabolic disease cases were 0 ng/mL. The relatively narrow range of serum Hp-MMP 9 complex concentrations in acutely septic animals enhances its diagnostic utility.

The biological rationale for the specificity of an ELISA based assay for serum

Hp-MMP 9 complexes likely consists of the uniqueness of its cellular origin (at present, the only known source is degranulating neutrophils).79 In contrast, free serum Hp and free MMP 9 are produced and secreted by a number of cellular sources in response to a

46 variety of challenges.94, 95 This is supported by our data in which 8 of 10 cattlw having either chronic inflammatory or metabolic disease possessed elevated free Hp and free MP

9 levels but lacked measurable quantities of the Hp-MMP 9 complexes.

In conclusion, when compared to free Hp or MMP 9, serum concentrations of Hp-MMP 9 appear to be a more reliable indicator of acute septic inflammation in cattle. We propose that application of the Hp-MMP 9 ELISA may be beneficial for the diagnosis of early events in acute septic conditions in the bovine.

Chapter 4: Serum Hp-MMP 9 Complex Response Following a Single Intravenous Dose of Lipopolysaccharide

Introduction

The body’s response to inflammation is a multi-faceted process involving gene expression and protein production, and changes in physiologic processes, which form the acute phase response. The acute phase response begins with a change in plasma concentration of cytokines which, in turn, lead to a large number of behavioral, physiologic, biochemical, and nutritional changes such as fever, somnolence, anorexia, leukocytosis, decreased gluconeogenesis, and plasma protein alterations (acute phase proteins).9 The acute phase response is mediated in part by production of acute phase proteins (APP) in the liver. In most cases of localized inflammation (e.g. the lung, gut, liver, mammary gland) disseminated (systemic) responses include upregulation of liver gene expression for APP as well as protein expression and release into the circulation.

Haptoglobin is a major acute phase protein in cattle and has been found in a variety of inflammatory diseases such as bovine respiratory disease13, 20 and mastitis12 as well as with endotoxin administration13 and inflammation caused by intramuscular injection of oil and turpentine.32 In addition to increased liver production of haptoglobin during inflammatory stimuli, haptoglobin is also produced by extrahepatic tissues including the gravid uterus10 and produced and stored in leukocytes, specifically neutrophils.36 Neutrophil Hp is synthesized at specific times, which provides the basis 48 for targeting specific proteins to the various intracellular granules. The expression of Hp overlaps with expression and packaging of neutrophil gelatinase (MMP 9).41 Release of

Hp from these neutrophil granules has been experimentally induced using phorbol myristate acetate (PMA) and variety of other inflammatory stimuli.36

A haptoglobin matrix metalloproteinase 9 (Hp-MMP 9) complex was recently described within the supernatant media after bovine neutrophil activation in vitro.79 High molecular mass polymers of haptoglobin with gelatinolytic activity were further characterized as MMP 9 bound to Hp.79 These molecules were later found to be measurable in the serum of cattle with inflammatory disease but absent from healthy cattle.79 Further study revealed that when clinical cases were classified as healthy, acute inflammation, or chronic inflammation, Hp-MMP 9 was significantly elevated in cattle with acute inflammation in comparison to those with chronic inflammation and was absent in serum of healthy cattle.80 Serum total Hp, although absent in healthy cattle, was not significantly different between cattle within the acute or chronic groups.80

We proposed that in vivo neutrophil activation is associated with rapid increases in serum Hp-MMP 9 complex concentrations. Intravenous endotoxin injection has been shown to produce physiologic and biochemical alterations in cattle including increases in

APP (Hp,13, 96 seromucoid,13 ceruloplasmin,13 α-1 proteinase inhibitor,13 and SAA96), decreased feed intake,97, 98 increased rectal temperature,23, 96, 97 increased respiratory rate,23, 96, 98 increased cytokines (TNFα,23, 96-98 IL-1β,23 IL-6,23and IFN-γ23) increased cortisol,23, 97 and increased insulin.97, 98 We used this model to monitor the response of

Hp-MMP 9 and Hp during acute inflammation.

Materials and Methods

Animals and Sample Collection – The following experimental protocol was approved by The Ohio State University, Laboratory Animal Care and Use Committee.

The study group consisted of 9 healthy Jersey bull calves between 65-82 days of age

(average weight – 74.5 ± 13.1 kg; range – 58.1-100 kg). Calves were housed in groups of

2 animals in climate controlled stalls. Physical exams were performed daily throughout their stay in the hospital. The body weights of these animals were monitored periodically. The calves were allowed to acclimate to their stalls by feeding free-choice hay and grain for 3-10 days before initiation of each experiment. Twelve hours prior to initiation of experiment, calves were fitted with indwelling jugular venous catheters

(Angiocath, 16GA, Becton Dickinson, Sandy, Utah) after clipping the hair and aseptic cleansing the skin using 1% iodine scrub and 70% isopropyl alcohol. Blood was collected from the catheter for CBC and serum which was centrifuged at 2750 RPM for

20 minutes within 12 hours of collection. The serum was removed and stored at -18° C until analysis. A 30 cm extension line with infusion port was attached and the catheter was held in place with elastic tape (Elastikon). The catheter was flushed with heparinized saline.

At time (T) = 0 hours, a physical exam was performed and blood was collected for a CBC and for serum collection which was stored for later analysis.

Lipopolysaccharide (Escherichia coli O111:B4; L2630 Sigma-Aldrich, St. Louis, MO)

2.5µg/kg diluted in 5 ml of autologous serum was administered via the IV catheter.

Physical exams and blood collection (as previously described) were performed at T = 0.5,

1, 1.5, 2, 3, 4, 6, 8, 12, 16, 24, 36, 48, 72, 96 hr post-LPS infusion. After each blood collection, equal volume of normal saline (0.9% NaCl for injection) was administered to maintain blood volume. IV catheters were removed after the last collection time point.

ELISA for Hp – A commercially available bovine Hp ELISA kit was used to determine Hp concentrations on the serum from the calves as well as the supernatants and elutions from the bovine MMP-9 affinity purification. The analysis was conducted according to the manufacturer’s instructions (Life Diagnostics, West Chester, PA).

Specifically, all serum samples from calves were diluted 1:2,000 in sample buffer prior to aliquoting 100 uL to each well for Hp analysis.

Elisa for Hp-MMP 9 – The ELISA for bovine Hp-MMP 9 was developed in the laboratory, exploiting un-conjugated bovine MMP 9 MAb 10.1 as a capture antibody

(100 µL, containing 2 µg per well in TBS + 0.1% BSA [Bovine Serum Albumin, Fraction

V, Fisher Scientific, Thermo Fisher Scientific, Pittsburgh, PA 15275]), rabbit anti-bovine

Hp as a secondary anti-body (100 µL, containing 0.1 µg per well in TBS + 0.1% BSA) and horse radish peroxidase (HRP)-conjugated goat anti-rabbit IgG (100 µL, containing

0.0025 µg per well in TBS + 0.1% BSA) as a detection antibody. These concentrations were chosen on basis of preliminary experiments.80 Capture antibody was allowed to bind to uncoated 96 well ELISA plates at 4C overnight. After capture antibody binding, the plates were blocked with TBS containing 2% bovine serum albumin, at 4 C for 120 minutes. After blocking the wells, calf serum samples and serum containing known concentrations of Hp-MMP 9 at a standard dilution were added (120 min at 21°C). The plates were again washed 4 times with TBS and rabbit anti-bovine Hp was added to the

51 wells (60 min at 21°C). Washing was repeated as previously described and the HRP- conjugated goat anti-rabbit IgG was added to each well. The excess HRP conjugate was removed by washing, after which 100 µL of TMB was added to each well for detection of bound Hp-MMP 9. The wells were incubated for 20 minutes and the reaction stopped by addition of 100 uL of 1N hydrochloric acid. The concentrations of Hp-MMP 9 were determined using the linear portion of the equation of the line described by absorbance of the calibrators at 450 nm and the known concentration of these calibrators.

Statistical analysis – Data from the LPS challenge study (temperature, pulse, respiration, WBC count, differential count, serum Hp concentrations, and serum Hp-

MMP 9 complex concentrations) were tabulated by time point and evaluated graphically in a commercial spreadsheet (Microsoft Excel 2007). After visual comparison of the data, the changes over time in different physiological and blood parameters were examined using PROC MIXED in SAS v.9.2 (SAS Inst. Inc, Cary, NC). Compound symmetry covariance structure was used to account for the correlated data structure of the repeated measures from individual calves over time. All measurements at different time- points were compared with the baseline value at T=0. Significant differences were considered when p<0.05.

Results

Physically, all calves exhibited signs of severe depression and respiratory embarrassment in less than 1 hour after endotoxin administration. The most reliable indicator of clinical illness was respiratory rate which increased significantly within 0.5

52 hours of LPS infusion (38 ± 17 breaths per minute (BPM) at time (T)=0 to 98 ± 37 BPM at T=30 minutes) and remained elevated through 4 hours (p<0.05; Figure 10).

Respiratory rate was at T=6 hours was not significantly different from that at T=0 and remained so through T=96 hours (p>0.05). Heart rate and rectal temperature didn’t change significantly during the study (Figures 11 and 12). Decreased ruminations/rumen stasis was also noted in all calves within the first hour and began to return between T=4 hours and T=8 hours.

Complete blood counts were performed at each sample collection throughout the study. Total white blood cell count (Mean ± SD; WBC) significantly (p<0.0001) decreased from 8.9 ± 2.88 x 109/L at T=0 to 4.43 ± 1.24 at T=0.5 hours, reached its nadir

(1.12 ± 0.53) at T=4 hrs, and became not significantly different from baseline at 16 hours

(p<0.05). Total WBC then became significantly higher than baseline at T=24 hrs (11.43

± 2.40; p<0.001), returned to close to baseline value at T=36 hrs and remained constant thereafter (Figure 13).

Total blood lymphocyte number made up the largest proportion of white cell count in pre-challenge samples and decrease after endotoxin injection. Lymphocyte numbers changed from 6.07 ± 1.73 x 109/L at T=0 to 0.92 ± 0.39 at T=4 hours and remained significantly lower than T=0 between 0.5 and 16 hours post LPS challenge

(p<0.003; Figure 13). Segmented neutrophils decreased and subsequently increased in the same time-frame as total WBC (Figure 13). These changes in segmented neutrophils was significantly different from T=0 between 0.5 hour and 36 hours (p<0.02). Band neutrophils were not present in peripheral blood until 6 hours post LPS and were

53 significantly different from baseline between 8 and 24 hours (P<0.05 compared to T=0).

All but one calf (calf # 105) developed a degenerative left-shift after LPS infusion. This appearance of band neutrophils occurred at approximately 6 hours, peaked at 16 hours

(0.20 ± 0.16 x 109/L), and then decreased back to 0 in all calves by 72 hours (Figure 14).

Cortisol was measured in 2 calves to determine whether the lymphopenia was associated with increase serum cortisol. Serum cortisol increased in these calves at the same time as the lymphocyte count decreased.

Serum haptoglobin concentrations were less than the lower limit of quantitation in all but 2 calves before LPS injection. All calves had Hp concentrations below 10 µg/ml, which was considered normal for cattle. Serum haptoglobin concentrations began to increase at approximately 6 hours and peaked at approximately 36 hours (400 ± 77

μg/mL; Figure 15) before returning to baseline by 96 hours. Two calves (Calf #108 and

#109) had Hp concentrations which did not decline by 96 hrs post-LPS (Figure 17).

Serum Hp concentrations were significantly higher than baseline between 12 and 96 hours post LPS infusion (p<0.009). The mean serum Hp concentration at the end of the trial was skewed by these 2 animals in which Hp concentration remained constant through the end of the trial.

Serum Haptoglobin-matrix metalloproteinase 9 (Hp-MMP 9) complex concentrations were not detectable in any calves pre-challenge. Increases in serum Hp-

MMP 9 were observed at approximately 12 hours and peaked at 36 hours (0.68 ± 0.50 ng/ml) before decreasing to baseline by the end of the study (Figure 16). Serum concentrations of Hp-MMP 9 were significantly higher than baseline between 16 and 48

54 hours post-LPS infusion (p<0.001). One calf (Calf 109) had Hp-MMP 9 concentrations that continued to increase throughout the end of the study (Figure 17). This was also one of the calves that had sustained Hp throughout the end of the study. This calf also had a second increase in respiratory rate that appears to correspond with Hp-MMP 9 increase

(Figure 18).

Discussion

The physical exam findings of the calves were normal prior to LPS challenge and after the challenge demonstrated expected changes in respiratory rate, and gastrointestinal motility as described previously.98 The primary clinical signs noted were associated with transient respiratory difficulty characterized by increased rate and effort of breathing. As these animals became depressed and recumbent, the respiratory signs were exacerbated.

While recumbent, the calves were depressed; some open mouth breathing with tongue extended and had no palpable or auscultable GI motility (rumen stasis). Subjectively, the calves became markedly depressed post-LPS infusion progressing from eating and drinking to recumbent and anorexic with respiratory difficulty. These clinical variables normalized within hours of LPS infusion.

The precipitous decrease in lymphocyte count was considered to be associated with a marked stress response; and in 2 calves, serum cortisol measurements confirmed marked increases that were associated with peripheral lymphopenia. Since the stress leukogram is characterized by a mature neutrophilia and concurrent lymphopenia, we assumed that elevated serum cortisol caused a rapid peripheral lymphopenia. In this

55 study, it was assumed that the neutropenia occurred due to margination of these cells to the vascular endothelium as previously described.99-101 A nearly identical peripheral leukocyte response was observed in dairy cattle after experimental E. coli mastitis; however, serum cortisol concentrations were not measured.101 In in vitro studies, dexamethasone has been shown to increase hepatocyte APP gene expression.102, 103 This suggests that cortisol production during inflammation may be necessary for the acute phase response.

The Hp concentration in these calves at baseline (T=0) was <10 ug/mL in all cases, which has been reported in numerous prior studies.13, 29, 31, 81 In the majority of calves studied, serum haptoglobin concentrations increased (beginning at 8 hours), peaked (36 hrs), and returned to baseline very quickly (96 hours). A previous study evaluating Hp responses to LPS infusion reported increases which occurred by 24 hours, peak concentrations achieved by 96 hours and return to baseline at 216 hrs.13 In that same study a similar response was noted with experimental infection with P. haemolytica

(M. haemolytica).13 A much more delayed response was noted in animals after experimental infection with viable bovine respiratory syncytial virus (BRSV).22 Prior studies also demonstrated higher peak concentrations of serum haptoglobin (800 μg/ml hemoglobin binding capacity)13 than those in the calves of the current study; however, since these studies used hemoglobin binding assay12, 13, 15, 20, 81 to quantitate serum haptoglobin, we are unsure as to how closely these two assays compare (hemoglobin binding assay vs. ELISA).

Serum concentrations of Hp-MMP 9 were determined by an ELISA based assay which incorporates a monoclonal antibody to bovine MMP 9 as the capture reagent, and rabbit anti-bovine haptoglobin polyclonal antibody as the detection reagent. This combination has been shown to be specific for the Hp-MMP 9 complex originating from the bovine neutrophil in vitro.79 Neither unbound Hp nor MMP 9 was shown to be detectable by this combination of antibodies, nor did excess concentrations of these purified proteins interfere with the ability of this assay to detect the combination.80

Measuring MMP 9 (Gelatinase B) in serum has been shown to be unreliable due to a number of factors including activation of neutrophils occurring during ex vivo blood clotting process.50 We believe that ex vivo clotting is not associated with neutrophil release of Hp-MMP 9 since all samples evaluated were from serum, and Hp-MMP 9 complex concentrations consistently occurred at the same time points in each calf.

Likewise, in a recent study, serum samples from apparently healthy cattle were analyzed for serum MMP 9 and Hp-MMP 9 concentrations. In all of these animals, serum concentrations of Hp-MMP 9 were 0; whereas serum concentrations of MMP 9 were similar to that observed in the diseased animals.80

In cattle with acute and septic inflammation, serum concentrations of Hp-MMP 9 complexes were within a relatively narrow range of approximately 800-1500 ng/mL.

Furthermore, normal, healthy, lactating cows, and animals with chronic or metabolic diseases (LDA, hepatic lipidosis) did not have measurable concentrations in most cases.

This suggests that serum concentrations of a novel neutrophil activation marker may be useful in the study of acute, septic or toxemic inflammation in cattle. This is significant

57 for several reasons. First, this protein biomarker is synthesized and stored pre-formed in neutrophil granules. Pre-produced protein is available inside the cell for immediate release after activation. Second, this marker is indicative of neutrophil activation and therefore is a marker of one of the earliest events in inflammation. This rapid onset process occurs within minutes to hours and results in substantial tissue injury in experimental models such as M. hemolytica pneumonia.69 Slocombe et al. (1985) demonstrated a critical role for neutrophils in the pathogenesis of inflammation associated with bovine respiratory disease. This and more recent studies suggest that neutrophils play key roles in the pathogenesis of lung injury associated with BRD.100, 104,

105 A marker of neutrophil activation may therefore provide early information of impending toxemia/septicemia in sub-clinically affected animals. Third, diseases characterized by rapid neutrophil accumulation would be expected to possess Hp-MMP 9 complexes in the milk or lung fluids since neutrophilic infiltrates will be composed of neutrophils in varying states of activation. In BAL fluids from calves after experimental, intra-bronchial LPS infusion, Hp-MMP 9 complexes are observed zymographically

(Lakritz, et al unpublished).

Finally, serum concentrations of Hp-MMP 9 complexes in animals undergoing systemic inflammation (in this study, LPS induced inflammation) are much lower than those observed in clinically septic or toxemic animals.80 This suggests that serum concentrations of Hp-MMP 9 complexes may represent a graded response to inflammation which would also provide information as to severity.

The concentration versus-time curves for Hp-MMP 9 concentration in the calves in this study, albeit, much lower than serum haptoglobin, were remarkably very similar to the concentration versus-time curves of serum Hp concentration (Figure 17). The commercial haptoglobin ELISA measures total serum haptoglobin, which includes haptoglobin produced by the liver and extra-hepatic sources, some of which may originate from the neutrophil, as well as free Hp and Hp that is bound to other molecules

(MMP 9). It is reasonable to assume that due to the quick increase and decrease in Hp and the close association of the Hp and Hp-MMP 9 time curves, that a majority of the Hp detected by the Hp ELISA is neutrophil in origin (bound to MMP 9). If not all Hp in the neutrophil is bound to MMP 9, Hp-MMP 9 is still a specific marker for neutrophil activation and measuring this complex will indicate this specific event in early inflammation.

References

1. National Agricultural Statistics Service. Cattle and Calf Death Losses. 2006;Mt An 12 (5-06).

2. Wittum TE, Woollen NE, Perino LJ, Littledike ET. Relationships among Treatment for Respiratory Tract Disease, Pulmonary Lesions Evident at Slaughter, and Rate of Weight Gain in Feedlot Cattle. J Am Vet Med Assoc 1996;209:814-818.

3. Smith RA. Impact of Disease on Feedlot Performance: A Review. J Anim Sci 1998;76:272-274.

4. Gorden PJ, Plummer P. Control, Management, and Prevention of Bovine Respiratory Disease in Dairy Calves and Cows. Vet Clin North Am Food Anim Pract 2010;26:243- 259.

5. Thompson PN, Stone A, Schultheiss WA. Use of Treatment Records and Lung Lesion Scoring to Estimate the Effect of Respiratory Disease on Growth during Early and Late Finishing Periods in South African Feedlot Cattle. J Anim Sci 2006;84:488-498.

6. Gardner BA, Dolezal HG, Bryant LK, Owens FN, Smith RA. Health of Finishing Steers: Effects on Performance, Carcass Traits, and Meat Tenderness. J Anim Sci 1999;77:3168-3175.

7. Lalman D, Ward CE. Effects of Preconditioning on Health, Performance and Prices of Weaned Calves. Conf Proced AABP 2005;38:44-50.

8. Tizard IR. Veterinary immunology : an introduction, 8th ed. St. Louis, Mo.: Saunders Elsevier; 2009:574.

9. Gabay C, Kushner I. Acute-Phase Proteins and Other Systemic Responses to Inflammation. N Engl J Med 1999;340:448-454.

10. Sharpe-Timms KL, Nabli H, Zimmer RL, Birt JA, Davis JW. Inflammatory Cytokines Differentially Up-Regulate Human Endometrial Haptoglobin Production in Women with Endometriosis. Hum Reprod 2010;25:1241-1250.

11. Cooray R, Waller KP, Venge P. Haptoglobin Comprises about 10% of Granule Protein Extracted from Bovine Granulocytes Isolated from Healthy Cattle. Vet Immunol Immunopathol 2007;119:310-315.

12. Conner JG, Eckersall PD, Doherty M, Douglas TA. Acute Phase Response and Mastitis in the Cow. Res Vet Sci 1986;41:126-128.

13. Conner JG, Eckersall PD, Wiseman A, Bain RK, Douglas TA. Acute Phase Response in Calves Following Infection with Pasteurella Haemolytica, Ostertagia Ostertagi and Endotoxin Administration. Res Vet Sci 1989;47:203-207.

14. Alsemgeest SP, Kalsbeek HC, Wensing T, et al. Concentrations of Serum Amyloid-A (SAA) and Haptoglobin (HP) as Parameters of Inflammatory Diseases in Cattle. Vet Q 1994;16:21-23.

15. Humblet MF, Coghe J, Lekeux P, Godeau JM. Acute Phase Proteins Assessment for an Early Selection of Treatments in Growing Calves Suffering from Bronchopneumonia Under Field Conditions. Res Vet Sci 2004;77:41-47.

16. Berry BA, Confer AW, Krehbiel CR, et al. Effects of Dietary Energy and Starch Concentrations for Newly Received Feedlot Calves: II. Acute-Phase Protein Response. J Anim Sci 2004;82:845-850.

17. Carter JN, Meredith GL, Montelongo M, et al. Relationship of Vitamin E Supplementation and Antimicrobial Treatment with Acute-Phase Protein Responses in Cattle Affected by Naturally Acquired Respiratory Tract Disease. Am J Vet Res 2002;63:1111-1117.

18. Lomborg SR, Nielsen LR, Heegaard PM, Jacobsen S. Acute Phase Proteins in Cattle After Exposure to Complex Stress. Vet Res Commun 2008;32:575-582.

19. Angen O, Thomsen J, Larsen LE, et al. Respiratory Disease in Calves: Microbiological Investigations on Trans-Tracheally Aspirated Bronchoalveolar Fluid and Acute Phase Protein Response. Vet Microbiol 2009;137:165-171.

20. Orro T, Pohjanvirta T, Rikula U, et al. Acute Phase Protein Changes in Calves during an Outbreak of Respiratory Disease Caused by Bovine Respiratory Syncytial Virus. Comp Immunol Microbiol Infect Dis 2009.

21. Carroll JA, Arthington JD, Chase CC,Jr. Early Weaning Alters the Acute-Phase Reaction to an Endotoxin Challenge in Beef Calves. J Anim Sci 2009;87:4167-4172.

22. Heegaard PM, Godson DL, Toussaint MJ, et al. The Acute Phase Response of Haptoglobin and Serum Amyloid A (SAA) in Cattle Undergoing Experimental Infection with Bovine Respiratory Syncytial Virus. Vet Immunol Immunopathol 2000;77:151-159.

23. Carroll JA, Reuter RR, Chase CC,Jr, et al. Profile of the Bovine Acute-Phase Response Following an Intravenous Bolus-Dose Lipopolysaccharide Challenge. Innate Immun 2009;15:81-89.

24. Cray C, Zaias J, Altman NH. Acute Phase Response in Animals: A Review. Comp Med 2009;59:517-526.

25. Tabrizi AD, Batavani RA, Rezaei SA, Ahmadi M. Fibrinogen and Ceruloplasmin in Plasma and Milk from Dairy Cows with Subclinical and Clinical Mastitis. Pak J Biol Sci 2008;11:571-576.

26. Hochepied T, Berger FG, Baumann H, Libert C. Alpha(1)-Acid Glycoprotein: An Acute Phase Protein with Inflammatory and Immunomodulating Properties. Cytokine Growth Factor Rev 2003;14:25-34.

27. Eckersall PD, Young FJ, McComb C, et al. Acute Phase Proteins in Serum and Milk from Dairy Cows with Clinical Mastitis. Vet Rec 2001;148:35-41.

28. Hiss S, Mielenz M, Bruckmaier RM, Sauerwein H. Haptoglobin Concentrations in Blood and Milk After Endotoxin Challenge and Quantification of Mammary Hp mRNA Expression. J Dairy Sci 2004;87:3778-3784.

29. Wittum TE, Young CR, Stanker LH, et al. Haptoglobin Response to Clinical Respiratory Tract Disease in Feedlot Cattle. Am J Vet Res 1996;57:646-649.

30. Tourlomoussis P, Eckersall PD, Waterson MM, Buncic S. A Comparison of Acute Phase Protein Measurements and Meat Inspection Findings in Cattle. Foodborne Pathog Dis 2004;1:281-290.

31. Young CR, Wittum TE, Stanker LH, et al. Serum Haptoglobin Concentrations in a Population of Feedlot Cattle. Am J Vet Res 1996;57:138-141.

32. Eckersall PD, Conner JG. Bovine and Canine Acute Phase Proteins. Vet Res Commun 1988;12:169-178.

33. Nakagawa H, Yamamoto O, Oikawa S, et al. Detection of Serum Haptoglobin by Enzyme-Linked Immunosorbent Assay in Cows with Fatty Liver. Res Vet Sci 1997;62:137-141.

34. Katoh N, Nakagawa H. Detection of Haptoglobin in the High-Density Lipoprotein and the very High-Density Lipoprotein Fractions from Sera of Calves with Experimental Pneumonia and Cows with Naturally Occurring Fatty Liver. J Vet Med Sci 1999;61:119- 124.

35. Kristiansen M, Graversen JH, Jacobsen C, et al. Identification of the Haemoglobin Scavenger Receptor Nature 2001;409:198-201.

36. Theilgaard-Monch K, Jacobsen LC, Nielsen MJ, et al. Haptoglobin is Synthesized during Granulocyte Differentiation, Stored in Specific Granules, and Released by Neutrophils in Response to Activation. Blood 2006;108:353-361.

37. Eaton JW, Brandt P, Mahoney JR, Lee JT,Jr. Haptoglobin: A Natural Bacteriostat. Science 1982;215:691-693.

38. Murata H, Miyamoto T. Bovine Haptoglobin as a Possible Immunomodulator in the Sera of Transported Calves. Br Vet J 1993;149:277-283.

39. Hutchens TW, Lönnerdal B. Lactoferrin : interactions and biological functions, Totowa, N.J.: Humana Press; 1997:408.

40. Dawes ME, Tyler JW, Marsh AE, et al. In Vitro Effects of Lactoferrin on Lipopolysaccharide-Induced Proliferation, Gene Expression, and Prostanoid Production by Bovine Peripheral Blood Mononuclear Cells. Am J Vet Res 2008;69:1164-1170.

41. Le Cabec V, Cowland JB, Calafat J, Borregaard N. Targeting of Proteins to Granule Subsets is Determined by Timing and Not by Sorting: The Specific Granule Protein NGAL is Localized to Azurophil Granules when Expressed in HL-60 Cells. Proc Natl Acad Sci U S A 1996;93:6454-6457.

42. Morimatsu M, Syuto B, Shimada N, et al. Isolation and Characterization of Bovine Haptoglobin from Acute Phase Sera. J Biol Chem 1991;266:11833-11837.

43. Raugei G, Bensi G, Colantuoni V, et al. Sequence of Human Haptoglobin cDNA: Evidence that the Alpha and Beta Subunits are Coded by the Same mRNA. Nucleic Acids Res 1983;11:5811-5819.

44. Birkedal-Hansen H, Moore WG, Bodden MK, et al. Matrix Metalloproteinases: A Review. Crit Rev Oral Biol Med 1993;4:197-250.

45. Murphy G, Bretz U, Baggiolini M, Reynolds JJ. The Latent Collagenase and Gelatinase of Human Polymorphonuclear Neutrophil Leucocytes. Biochem J 1980;192:517-525.

46. Goetzl EJ, Banda MJ, Leppert D. Matrix Metalloproteinases in Immunity. J Immunol 1996;156:1-4.

47. Lazarus GS, Daniels JR, Lian J, Burleigh MC. Role of Granulocyte Collagenase in Collagen Degradation. Am J Pathol 1972;68:565-578.

48. Johansson S, Smedsrod B. Identification of a Plasma Gelatinase in Preparations of Fibronectin. J Biol Chem 1986;261:4363-4366.

49. Borregaard N, Sehested M, Nielsen BS, Sengelov H, Kjeldsen L. Biosynthesis of Granule Proteins in Normal Human Bone Marrow Cells. Gelatinase is a Marker of Terminal Neutrophil Differentiation. Blood 1995;85:812-817.

50. Zucker S, Hymowitz M, Conner C, et al. Measurement of Matrix Metalloproteinases and Tissue Inhibitors of Metalloproteinases in Blood and Tissues. Clinical and Experimental Applications. Ann N Y Acad Sci 1999;878:212-227.

51. Vu TH, Shipley JM, Bergers G, et al. MMP-9/gelatinase B is a Key Regulator of Growth Plate Angiogenesis and Apoptosis of Hypertrophic Chondrocytes. Cell 1998;93:411-422.

52. Mentzel K, Brauer R. Matrix Metalloproteinases, IL-6, and Nitric Oxide in Rat Antigen-Induced Arthritis. Clin Exp Rheumatol 1998;16:269-276.

53. Warner RL, Bhagavathula N, Nerusu KC, et al. Matrix Metalloproteinases in Acute Inflammation: Induction of MMP-3 and MMP-9 in Fibroblasts and Epithelial Cells Following Exposure to Pro-Inflammatory Mediators in Vitro. Exp Mol Pathol 2004;76:189-195.

54. Lakritz J, Marsh AE, Cockrell M, Smith MF, Tyler JW. Characterization of Gelatinases in Bronchoalveolar Lavage Fluid and Gelatinases Produced by Alveolar Macrophages Isolated from Healthy Calves. Am J Vet Res 2004;65:163-172.

55. Linsuwanont B, Takagi Y, Ohya K, Shimokawa H. Expression of Matrix Metalloproteinase-9 mRNA and Protein during Deciduous Tooth Resorption in Bovine Odontoclasts. Bone 2002;31:472-478.

56. Agren MS. Gelatinase Activity during Wound Healing. Br J Dermatol 1994;131:634- 640.

57. Agren MS, Mirastschijski U, Karlsmark T, Saarialho-Kere UK. Topical Synthetic Inhibitor of Matrix Metalloproteinases Delays Epidermal Regeneration of Human Wounds. Exp Dermatol 2001;10:337-348.

58. Bjorklund M, Koivunen E. Gelatinase-Mediated Migration and Invasion of Cancer Cells. Biochim Biophys Acta 2005;1755:37-69.

59. Xu D, McKee CM, Cao Y, et al. Matrix Metalloproteinase-9 Regulates Tumor Cell Invasion through Cleavage of Protease Nexin-1. Cancer Res 2010.

60. Tsang PC, Poff JP, Boulton EP, Condon WA. Four-Day-Old Bovine Corpus Luteum: Progesterone Production and Identification of Matrix Metalloproteinase Activity in Vitro. Biol Reprod 1995;53:1160-1168.

61. Smith MF, McIntush EW, Ricke WA, Kojima FN, Smith GW. Regulation of Ovarian Extracellular Matrix Remodelling by Metalloproteinases and their Tissue Inhibitors: Effects on Follicular Development, Ovulation and Luteal Function. J Reprod Fertil Suppl 1999;54:367-381.

62. Goldberg MJ, Moses MA, Tsang PC. Identification of Matrix Metalloproteinases and Metalloproteinase Inhibitors in Bovine Corpora Lutea and their Variation during the Estrous Cycle. J Anim Sci 1996;74:849-857.

63. Walter I, Boos A. Matrix Metalloproteinases (MMP-2 and MMP-9) and Tissue Inhibitor-2 of Matrix Metalloproteinases (TIMP-2) in the Placenta and Interplacental Uterine Wall in Normal Cows and in Cattle with Retention of Fetal Membranes. Placenta 2001;22:473-483.

64. Kruze D, Wojtecka E. Activation of Leucocyte Collagenase Proenzyme by Rheumatoid Synovial Fluid. Biochim Biophys Acta 1972;285:436-446.

65. Van den Steen PE, Proost P, Grillet B, et al. Cleavage of Denatured Natural Collagen Type II by Neutrophil Gelatinase B Reveals Enzyme Specificity, Post-Translational Modifications in the Substrate, and the Formation of Remnant Epitopes in Rheumatoid Arthritis. FASEB J 2002;16:379-389.

66. Arican M, Coughlan AR, Clegg PD, Carter SD. Matrix Metalloproteinases 2 and 9 Activity in Bovine Synovial Fluids. J Vet Med A Physiol Pathol Clin Med 2000;47:449- 456.

67. Simonen-Jokinen TL, Eskelinen UM, Hartel HM, et al. Gelatinolytic Matrix Metalloproteinases-2 and -9 in Tracheobronchial Lavage Fluid obtained from Calves with Concurrent Infections of Pasteurella Multocida and Mycoplasma Bovirhinis. Am J Vet Res 2005;66:2101-2106.

68. Corry DB, Rishi K, Kanellis J, et al. Decreased Allergic Lung Inflammatory Cell Egression and Increased Susceptibility to Asphyxiation in MMP2-Deficiency. Nat Immunol 2002;3:347-353. 65

69. Slocombe RF, Malark J, Ingersoll R, Derksen FJ, Robinson NE. Importance of Neutrophils in the Pathogenesis of Acute Pneumonic Pasteurellosis in Calves. Am J Vet Res 1985;46:2253-2258.

70. Murphy G, Reynolds JJ, Bretz U, Baggiolini M. Partial Purification of Collagenase and Gelatinase from Human Polymorphonuclear Leucocytes. Analysis of their Actions on Soluble and Insoluble Collagens. Biochem J 1982;203:209-221.

71. Murphy G, Reynolds JJ, Bretz U, Baggiolini M. Collagenase is a Component of the Specific Granules of Human Neutrophil Leucocytes. Biochem J 1977;162:195-197.

72. Hibbs MS, Bainton DF. Human Neutrophil Gelatinase is a Component of Specific Granules. J Clin Invest 1989;84:1395-1402.

73. Kjeldsen L, Bjerrum OW, Askaa J, Borregaard N. Subcellular Localization and Release of Human Neutrophil Gelatinase, Confirming the Existence of Separate Gelatinase-Containing Granules. Biochem J 1992;287 ( Pt 2):603-610.

74. Delclaux C, Delacourt C, D'Ortho MP, et al. Role of Gelatinase B and Elastase in Human Polymorphonuclear Neutrophil Migration Across Basement Membrane. Am J Respir Cell Mol Biol 1996;14:288-295.

75. Saarialho-Kere UK, Welgus HG, Parks WC. Distinct Mechanisms Regulate Interstitial Collagenase and 92-kDa Gelatinase Expression in Human Monocytic-Like Cells Exposed to Bacterial Endotoxin. J Biol Chem 1993;268:17354-17361.

76. Baylis HA, Megson A, Hall R. Infection with Theileria Annulata Induces Expression of Matrix Metalloproteinase 9 and Transcription Factor AP-1 in Bovine Leucocytes. Mol Biochem Parasitol 1995;69:211-222.

77. Pugin J, Widmer MC, Kossodo S, et al. Human Neutrophils Secrete Gelatinase B in Vitro and in Vivo in Response to Endotoxin and Proinflammatory Mediators. Am J Respir Cell Mol Biol 1999;20:458-464.

78. Li X, Zhao X, Ma S. Secretion of 92 kDa Gelatinase (MMP-9) by Bovine Neutrophils. Vet Immunol Immunopathol 1999;67:247-258.

79. Bannikov GA, Mattoon JS, Abrahamsen EJ, et al. Biochemical and Enzymatic Characterization of Purified Covalent Complexes of Matrix Metalloproteinase-9 and Haptoglobin Released by Bovine Granulocytes in Vitro. Am J Vet Res 2007;68:995- 1004.

80. Bannikov GA, Hinds CA, Rajala-Schultz PJ, et al. Serum Haptoglobin-Matrix Metalloproteinase 9 (Hp-MMP 9) Complex as a Biomarker of Systemic Inflammation in Cattle. Vet Immunol Immunopathol 2011;139:41-49.

81. Conner JG, Eckersall PD, Wiseman A, Aitchison TC, Douglas TA. Bovine Acute Phase Response Following Turpentine Injection. Res Vet Sci 1988;44:82-88.

82. Gozho GN, Plaizier JC, Krause DO, Kennedy AD, Wittenberg KM. Subacute Ruminal Acidosis Induces Ruminal Lipopolysaccharide Endotoxin Release and Triggers an Inflammatory Response. J Dairy Sci 2005;88:1399-1403.

83. Nielsen BH, Jacobsen S, Andersen PH, Niewold TA, Heegaard PM. Acute Phase Protein Concentrations in Serum and Milk from Healthy Cows, Cows with Clinical Mastitis and Cows with Extramammary Inflammatory Conditions. Vet Rec 2004;154:361-365.

84. Ohtsuka H, Kudo K, Mori K, et al. Acute Phase Response in Naturally Occurring Coliform Mastitis. J Vet Med Sci 2001;63:675-678.

85. Stengarde L, Traven M, Emanuelson U, et al. Metabolic Profiles in Five High- Producing Swedish Dairy Herds with a History of Abomasal Displacement and Ketosis. Acta Vet Scand 2008;50:31.

86. Arthington JD, Eichert SD, Kunkle WE, Martin FG. Effect of Transportation and Commingling on the Acute-Phase Protein Response, Growth, and Feed Intake of Newly Weaned Beef Calves. J Anim Sci 2003;81:1120-1125.

87. Larsen K, Macleod D, Nihlberg K, et al. Specific Haptoglobin Expression in Bronchoalveolar Lavage during Differentiation of Circulating Fibroblast Progenitor Cells in Mild Asthma. J Proteome Res 2006;5:1479-1483.

88. Gronlund U, Hallen Sandgren C, Persson Waller K. Haptoglobin and Serum Amyloid A in Milk from Dairy Cows with Chronic Sub-Clinical Mastitis. Vet Res 2005;36:191- 198.

89. Kohler G, Milstein C. Continuous Cultures of Fused Cells Secreting Antibody of Predefined Specificity. Nature 1975;256:495-497.

90. Kolaczkowska E, Arnold B, Opdenakker G. Gelatinase B/MMP-9 as an Inflammatory Marker Enzyme in Mouse Zymosan Peritonitis: Comparison of Phase-Specific and Cell- Specific Production by Mast Cells, Macrophages and Neutrophils. Immunobiology 2008;213:109-124.

91. Peake NJ, Foster HE, Khawaja K, Cawston TE, Rowan AD. Assessment of the Clinical Significance of Gelatinase Activity in Patients with Juvenile Idiopathic Arthritis using Quantitative Protein Substrate Zymography. Ann Rheum Dis 2006;65:501-507.

92. Le NT, Xue M, Castelnoble LA, Jackson CJ. The Dual Personalities of Matrix Metalloproteinases in Inflammation. Front Biosci 2007;12:1475-1487.

93. Cohn LA. The Influence of Corticosteroids on Host Defense Mechanisms. J Vet Intern Med 1991;5:95-104.

94. D'Armiento J, Dalal SS, Chada K. Tissue, Temporal and Inducible Expression Pattern of Haptoglobin in Mice. Gene 1997;195:19-27.

95. Van den Steen PE, Dubois B, Nelissen I, et al. Biochemistry and Molecular Biology of Gelatinase B Or Matrix Metalloproteinase-9 (MMP-9). Crit Rev Biochem Mol Biol 2002;37:375-536.

96. Kahl S, Elsasser TH. Exogenous Testosterone Modulates Tumor Necrosis Factor- Alpha and Acute Phase Protein Responses to Repeated Endotoxin Challenge in Steers. Domest Anim Endocrinol 2006;31:301-311.

97. Steiger M, Senn M, Altreuther G, et al. Effect of a Prolonged Low-Dose Lipopolysaccharide Infusion on Feed Intake and Metabolism in Heifers. J Anim Sci 1999;77:2523-2532.

98. Waldron MR, Nishida T, Nonnecke BJ, Overton TR. Effect of Lipopolysaccharide on Indices of Peripheral and Hepatic Metabolism in Lactating Cows. J Dairy Sci 2003;86:3447-3459.

99. Zarbock A, Ley K. Mechanisms and Consequences of Neutrophil Interaction with the Endothelium. Am J Pathol 2008;172:1-7.

100. Radi ZA, Caverly JM, Dixon RA, Brogden KA, Ackermann MR. Effects of the Synthetic Selectin Inhibitor TBC1269 on Tissue Damage during Acute Mannheimia Haemolytica-Induced Pneumonia in Neonatal Calves. Am J Vet Res 2001;62:17-22.

101. Bannerman DD, Kauf AC, Paape MJ, Springer HR, Goff JP. Comparison of Holstein and Jersey Innate Immune Responses to Escherichia Coli Intramammary Infection. J Dairy Sci 2008;91:2225-2235.

102. Baumann H, Jahreis GP. Regulation of Mouse Haptoglobin Synthesis. J Cell Biol 1983;97:728-736.

103. Baumann H, Jahreis GP, Gaines KC. Synthesis and Regulation of Acute Phase Plasma Proteins in Primary Cultures of Mouse Hepatocytes. J Cell Biol 1983;97:866-876.

104. Caverly JM, Radi ZA, Andreasen CB, et al. Comparison of Bronchoalveolar Lavage Fluid obtained from Mannheimia Haemolytica-Inoculated Calves with and without Prior Treatment with the Selectin Inhibitor TBC1269. Am J Vet Res 2001;62:665-672.

105. Weiss DJ, Bauer MC, Whiteley LO, Maheswaran SK, Ames TR. Changes in Blood and Bronchoalveolar Lavage Fluid Components in Calves with Experimentally Induced Pneumonic Pasteurellosis. Am J Vet Res 1991;52:337-344.

Appendix A: Tables

Number ADG of (lbs ± SD) Calves Hp 0 40 1.7 ± 0.4 Hp 1 18 1.9 ± 0.5 Hp >1 7 1.5 ± 0.7 Table 1. Calves grouped by number of positive Hp time-points Number of calves and mean ADG for each Hp group based on number of time-points that Hp was positive. Hp 0 group were calves that had 0 detectable Hp at all time-points, Hp 1 had detectable Hp at exactly 1 time-point, and Hp >1 had detectable Hp concentrations at 2 or more time-points. No statistical difference was noted between groups.

Number ADG of (lbs ± SD) Calves Complex 0 2 1.9 ± 0.5 Complex 1 4 2.0 ± 0.5 Complex 2 7 1.6 ± 0.5 Complex >2 51 1.7 ± 0.5 Table 2. Calves grouped by number of positive Hp-MMP 9 time-points Number of calves and mean ADG for each Hp-MMP 9 complex group based on number of time-points that Hp-MMP 9 was positive. Complex 0 group were calves that had 0 detectable Hp-MMP 9 complex at all time-points, Complex 1 had detectable Hp-MMP 9 complex at exactly 1 time-point, Complex 2 had detectable Hp-MMP 9 complex at exactly 2 time-points, and Complex >2 had detectable Hp-MMP 9 complex concentrations at 3 or more time-points. No statistical differences was noted between groups

Number of ADG Samples (lbs ± SD) Hp=0 mcg/ml 157 1.60 ± 1.13 Hp=1-10 mcg/ml 6 1.86 ± 0.91 Hp=10-100 mcg/ml 10 1.18 ± 1.82 Hp>100 mcg/ml 3 -0.33 ± 2.86 Table 3. Samples grouped by Hp concentration and corresponding ADG Individual samples grouped based on Hp concentration and corresponding ADG for the groups. Only Hp>100mcg/ml was significant different from the rest of the samples. There were only 3 samples that were in this group, therefore the power of this analysis is lacking.

Number of ADG Samples (lbs ± SD) Hp-MMP 9 = 0 28 1.75 ± 1.33 Hp-MMP 9 = 1-10 mcg/ml 50 1.66 ± 0.80 Hp-MMP 9 = 10-20 mcg/ml 55 1.46 ± 1.45 Hp-MMP 9 = 20-30 mcg/ml 20 1.63 ± 1.00 Hp-MMP 9 = 30-50 mcg/ml 21 1.45 ± 1.41 Hp-MMP 9 > 50 9 0.9 ± 1.50 Table 4. Samples grouped by Hp-MMP 9 concentration and corresponding ADG Individual samples grouped based on Hp-MMP 9 concentration and corresponding ADG for the groups. No statistical difference between was noted between groups.

HP- Disease HP MMP9 Animal # Age (yr) Breed Sex Diagnosis MMP9 Outcome Classification (ug/ml) (ng/ml) (ng/ml) 1 2 G F Peritonitis-Uterus/Gut Acute 544 1392 1115 Euthanized

2 3 BS F Peritonitis-Gut Acute 308 212 3139 Euthanized Fibrinosuppurative bronchopneumonia, 3 4 Hol F Acute 945 2240 2745 Euthanized endometritis 4 2 J F Peritonitis-Gut Acute 415 947 2334 Released

5 5 Hol F Peritonitis-Gut Acute 401 1315 1767 Euthanized

6 1 A M Peritonitis-Gut Acute 687 1496 312 Euthanized

7 4 Hol F Fibrinous bronchopneumonia Acute 772 1493 3871 Euthanized

8 4 J F Coliform mastitis, metabolic Acute 217 0 1056 Released Peritonitis-Mammary vein 9 6 Hol F Acute 403 399 4972 Released thrombophlebitis 10 3 Hol F Peritonitis- ruptured Abscess Acute 463 913 646 Euthanized

11 3 Hol F Peritonitis-Uterus Acute 456 1187 9433 Euthanized

12 2 Hol F Peritonitis Acute 681 1014 3122 Euthanized

13 3 Hol F Toxic mastitis Acute 5 477 6456 Euthanized Septic caval thrombophlebitis/Hepatic Acute on 14 4 Her F abscess/Fibrinosuppurative 301 1039 1181 Euthanized chronic bronchopneumonia Endocarditis/Chronic Acute on 15 4 Hol F 4 979 3742 Euthanized bronchopneumonia, Chronic Table 5. Disease Classification 1 Acute with Hp, MMP 9, and Hp-MMP 9

HP- Disease HP MMP9 Animal # Age (yr) Breed Sex Diagnosis MMP9 Outcome Classification (ug/ml) (ng/ml) (ng/ml)

16 4 Hol F Peritonitis-Surgical Chronic 603 0 2129 Euthanized

17 3 Her M Thymic lymphoma Chronic 40 673 2336 Euthanized Chronic suppurative 18 4 Hol F Chronic 0 0 1881 Released bronchopneumonia 19 4 Hol F Metabolic, Surgical inflammation Chronic 689 2885 2910 Released Abomasal 20 4 Hol M lymphoma/Perforation/Omental Chronic 131 0 1358 Euthanized bursitis 21 0.4 A F Chronic pneumonia Chronic 146 0 1625 Euthanized

22 0.7 A F Chronic pneumonia Chronic 168 0 3240 Released

23 4 Hol F Mastitis Chronic 567 0 895 Released

24 4 Hol F Metritis Chronic 320 0 2531 Released

25 8 J F Metritis/Retained Placenta Chronic 187 0 3078 Released Table 6. Disease Classification 2, Chronic with Hp, MMP 9, Hp-MMP 9

HP- Disease HP MMP9 Animal # Age (yr) Breed Sex Diagnosis MMP9 Outcome Classification (ug/ml) (ng/ml) (ng/ml) 26 4 Hol F Healthy Normal 0.0 0 153 Released

27 3 Hol F Healthy Normal 20 0 440 Released

28 3 Hol F Healthy Normal 0. 0 100 Released

29 3 Hol F Healthy Normal 0 0 480 Released

30 3 Hol F Healthy Normal 0 0 143 Released

31 7 J F Healthy Normal 0 0 220 Released

32 6 J F Healthy Normal 0 0 2870 Released

33 5 J F Healthy Normal 20 0 573 Released

34 6 J F Healthy Normal 0 0 153 Released

35 4 J F Healthy Normal 0 0 1606 Released Table 7. Disease Classification 3, Healthy with Hp, MMP 9, Hp-MMP 9

Haptoglobin (µg/ml) Hp-MMP 9 (ng/ml) MMP 9 (ng/ml)

Disease Classification Median 25th-75 Percentile Median 25th-75 Percentile Median 25th-75 Percentile

1 Acute septic 415 302, 680 1014 477, 1392 2745 1115, 3871

2 Chronic, metabolic 178 131, 567 0a 0, 0 2233 1625, 2910

3 Normal 0a 0, 0 0a 0, 0 330a 153, 573 Table 8. Mean Hp, MMP 9, Hp-MMP 9 for disease classifications Median and the 25th and 75th percentiles for serum haptoglobin, Hp–MMP 9 complex and MMP 9 in 35 animals classified as acute septic disease (1), chronic inflammation/metabolic disease (2), and healthy cattle (3). Data are presented as the median value and 25th and 75th percentile values. Serum concentrations of Hp, Hp–MMP 9 and MMP 9 were compared between disease classification groups using the Kruskal–Wallis One way ANOVA on ranks with pair-wise comparisons using Wilcoxon rank-sum test with Bonferroni adjustment. a Concentrations with a different superscript (e.g., one with and one without a superscript) within a column are statistically significantly different from each other at P < 0.003

Appendix B: Figures

Hp, Hp-MMP 9 250000

200000

150000

100000

Hp Hp ng/ml 50000

0 20 40 60 80 100 -50000 Hp-MMP 9 ng/ml

Figure 1. Correlation of Hp, Hp-MMP 9 in all samples Correlation of Hp and Hp-MMP 9 in all samples tested. Dotted lines indicate 95% CI (r2 = 0.064).

Hp, Hp-MMP 9 without Hp=0 250000

200000

150000

100000

Hp Hp ng/ml 50000

0 20 40 60 80 -50000 Hp-MMP 9 ng/ml

Figure 2. Correlation of Hp, Hp-MMP 9 excluding Hp=0 Correlation of Hp and Hp-MMP 9 in all samples excluding time-points in which Hp = 0. Dotted lines indicate 95% CI (r2 = 0.337).

Complex if Hp=0 vs Hp>0 50

Hp-MMP9 ng/ml Hp-MMP9 10

Hp=0 Hp>0

Figure 3. Hp-MMP 9 concentration when Hp=0 vs. Hp>0 Hp-MMP 9 complex concentration at time-points in which Hp=0 vs Hp>0. Significant difference (P = 0.0065). Graph represents mean concentration. Error bars represent SD.

Hp groups 3

ADG (lbs) ADG 1

Hp 0 Hp 1 Hp >1 Figure 4. ADG when Hp was positive at 0, 1, >1 time-points Hp 0 group (n=40) were calves that had 0 detectable Hp at all time-points, Hp 1 (n=18) had detectable Hp at exactly 1 time-point, and Hp >1 (n=7) had detectable Hp concentrations at 2 or more time-points. Graph represents mean concentration. Error bars indicate SD.

Complex groups 3

ADG (lbs) ADG 1

Complex 0 Complex 1 Complex 2 Complex >2 Figure 5. ADG when Hp-MMP 9 was positive at 0, 1, 2, >2 time-points Complex 0 group (n=2) were calves that had 0 detectable Hp-MMP 9 complex at all time-points, Complex 1 (n=4) had detectable Hp-MMP 9 complex at exactly 1 time- point, Complex 2 (n=7) had detectable Hp-MMP 9 complex at exactly 2 time-points, and Complex >2 (n=51) had detectable Hp-MMP 9 complex concentrations at 3 or more time-points. Error bars indicate SD.

Hp vs ADG Time-points 4

ADG (lbs) ADG Hp=0 -2

-4 Hp=1-10 mcg/ml Hp>100 mcg/ml Hp=10-100 mcg/ml

Figure 6. ADG compared to concentration of Hp Hp concentrations compared to ADG during the period directly following sampling. Only Hp>100mcg/ml was significant different from the rest of the samples. There were only 3 samples that were in this group, therefore the power of this analysis is lacking. Hp= 0, n= 157; Hp=1-10, n=6; Hp=10-100, n=10; Hp>100, n=3.

Hp-MMP 9 vs ADG Time-points 4

2 ADG (lbs) ADG 1

Hp-MMP 9=0 Hp-MMP 9>50

Hp-MMP 9=1-10 ng/ml Hp-MMP 9=10-20Hp-MMP ng/ml 9=20-30Hp-MMP ng/ml 9=30-50 ng/ml Figure 7. ADG compared to concentration of Hp-MMP 9 Hp-MMP 9 concentrations compared to ADG during the period directly following sampling. No significant differences could be found between groups. Hp-MMP 9 = 0, n=28; Hp-MMP 9 = 1-10, n=50; Hp-MMP 9 = 10-20, n=55; Hp-MMP 9 = 20-30, n=20; Hp-MMP 9 = 30-50, n=21; Hp-MMP 9 >50, n=9.

Figure 8. Hp-MMP 9 and MMP 9 ELISA specificity Graphical representation of ELISA data demonstrating the specificity of the MMP 9 and Hp–MMP 9 ELISA assays for free MMP 9 and Hp–MMP 9 complexes. The graph above is representative of results of 3 independent experiments. The line with solid black circles represents absorbance for increasing concentrations of Hp–MMP 9 captured on MMP 9 monoclonal Ab and detected with anti-Hp HRP conjugate (Hp–MMP 9 standard curve). The green line with open circles represents wells where affinity purified Hp was added to the anti-MMP 9 coated wells followed by anti-Hp HRP conjugate. No HRP-antibody binding was detected. When affinity purified MMP 9 was added to wells (red line; filled, downward triangles) and the anti-Hp HRP conjugate was added, no antibody binding was detected. Finally, when Hp–MMP 9 complexes were added to wells, and HRP-conjugated anti-MMP 9 (Clone 10.1) was added to wells (simulation of MMP 9 ELISA), no HRP- conjugated antibody binding was detected (blue line, filled, upward triangles).

Serum Haptoglobin (mg/mL) Serum Hp-MMP 9 complex (ng/mL)

1000 3500 a b 3000  800 a a 2500

g/mL  600 2000

400 1500

1000 200 b

complex (ng/mL)

b Serum Hp-MMP 9 500

Serum Haptoglobin ( 0 0

1 2 3 1 2 3 Disease status Disease Status

Serum MMP 9 (ng/mL)

10000 a

8000

6000 a 4000 c

2000

Serum MMP 9 (ng/mL) 0

1 2 3 Disease status

Figure 9. Hp, MMP 9, and Hp-MMP 9 concentrations compared to diseases status Box – plot of data for serum haptoglobin in cattle by disease classification. Pairwise, statistical comparisons of serum concentrations of A). haptoglobin, B). haptoglobin- MMP 9 complexes and C). MMP 9 in the 35 bovine patients evaluated in this study. Statistical differences are indicated by differing letters above disease category boxes. Disease category 1 = acute inflammation (n=15), 2 = chronic inflammation (n=10), 3 = healthy (n=10).

Respiratory Rate 150

100

50 Breaths/minute

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 -24.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 Time (hr)

Figure 10. Mean respiratory rates of calves during endotoxin trial Mean respiratory rate of all calves during endotoxin challenge. Error bars indicate SD.

Temperature 105

104

F)  103

102 Temperature ( Temperature 101

100

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 -24.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 Time (hr)

Figure 11. Mean rectal temperature of calves during endotoxin trial Mean rectal temperature for all calves during trial. Error bars indicate SD.

Heart Rate 140

120

100

Beats/minute 80

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 -24.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 Time (hr)

Figure 12. Mean heart rate of calves during endotoxin trial Mean heart rate for all calves during trial. Error bars indicate SD.

WBC WBC (total) 15 Lymphocytes Segmented Neutrophils

WBC X 10^9/L X WBC 0

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 -24.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 -5 Time (hr)

Figure 13. MeanWBC of calves during endotoxin trial Mean WBC for all calves during trial. Error bars indicate SD.

Band Neutrophil 0.4

0.3

0.2

0.1 WBC X 10^9/L X WBC 0.0

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 -0.1-24.0 Time (hr)

Figure 14. Mean band neutrophils of calves during endotoxin trial Mean band neutrophil count for all calves during trial. Error bars indicate SD.

Haptoglobin 600

400

200 Hp mcg/mlHp 0

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 -24.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 -200 Time (hr)

Figure 15. Mean Hp of calves during endotoxin trial Mean Hp for all calves during trial. Error bars indicate SD.

Hp-MMP 9 1.5

1.0

0.5

0.0 Hp-MMP 9 ng/mlHp-MMP

0.0 0.5 1.0 1.5 2.0 3.0 4.0 6.0 8.0 -24.0 12.0 16.0 24.0 36.0 48.0 72.0 96.0 -0.5 Time (hr)

Figure 16. Mean Hp-MMP 9 of calves during endotoxin trial Mean Hp-MMP 9 concentrations for all calves during trial. Error bars indicate SD.

Hp-MMP9 and haptoglobin over time (in relation to LPS-infusion at time=0), by calf

101 102 104

500

1.5

400

300

200

100

0 0

105 106 107

500

1.5

400

300

200

100

0 Hp-MMP9

108 109 110

500

1.5

400

300

200

100

0 0

-20 0 20 40 60 80 100 -20 0 20 40 60 80 100 -20 0 20 40 60 80 100 Time Hp-MMP9 Haptogl

Figure 17. Hp, Hp-MMP 9 over time Hp-MMP9 and haptoglobin over time (in relation to LPS infusion, time=0) in individual calves 96

Hp-MMP9 and respiratory rate over time (in relation to LPS-infusion at time=0), by calf

101 102 104

120

1.5

100

0 20

105 106 107

120

1.5

100

20 Hp-MMP9

108 109 110

120

1.5

100

0 20

-20 0 20 40 60 80 100 -20 0 20 40 60 80 100 -20 0 20 40 60 80 100 Time Hp-MMP9 Resp

Figure 18. Hp-MMP 9, respiratory rate over time Hp-MMP9 and respiratory rate over time (in relation to LPS-infusion, time=0) in individual calves 97