Morphometric examination of the equine adult and foal lung

MORPHOMETRIC EXAMINATION OF THE EQUINE ADULT AND FOAL LUNG A Thesis Submitted to the College of Graduate Studies and Research In Partial Fulfillment of the Requirements for the Degree of Masters of Science In the Department of Veterinary Biomedical Sciences University of Saskatchewan Saskatoon By LAURA JOHNSON Copyright Laura Johnson, August, 2013. All Rights Reserved. PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a Postgraduate degree from the University of Saskatchewan, I agree that the Libraries of this University may make it freely available for inspection. I further agree that permission for copying of this thesis in any manner, in whole or in part, for scholarly purposes may be granted by the professor or professors who supervised my thesis work or, in their absence, by the Head of the Department or the Dean of the College in which my thesis work was done. It is understood that any copying or publication or use of this thesis or parts thereof for financial gain shall not be allowed without my written permission. It is also understood that due recognition shall be given to me and to the University of Saskatchewan in any scholarly use which may be made of any material in my thesis/dissertation. Requests for permission to copy or to make other uses of materials in this thesis in whole or part should be addressed to: Head of the Department of Veterinary Biomedical Sciences University of Saskatchewan Saskatoon, Saskatchewan (S7N 5B4) Canada i ABSTRACT To fully understand the mechanisms of lower airway inflammation associated with many equine diseases such as heaves or Rhodococcus equi infection, which are age specific, we must first identify baseline “normal” structural characteristics of the horse lung. To develop a detailed understanding of the morphology of the horse lung, stereological methods were adapted and applied to the lungs from healthy adult horses (n=4) and one day (n=5) and 30 day (n=5) old foals. The left lung from each animal was fixed in situ and was then removed from the body cavity and remained in fixative overnight before beginning an unbiased sampling procedure. The tissue samples were fixed in plastic and paraffin blocks for stereological evaluation and immunohistochemistry, respectively. The lung was characterised into parenchyma and nonparenchyma, where median parenchymal density (alveolar airspace, ductal airspace and tissue) was 81.0% in one day old foals, 84.4% in 30 day old foals and 93.7% in adult lungs. The median volume density of alveolar airspace per lung was 45.9% in one day old, 55.5% in 30 day and 66.9% in adult horse lungs. Ductal airspace and alveolar tissue volume density was unchanged between the age groups. The median alveolar surface area (m2) seemed to increase with age, from about 205.3m2, 258.2m2 and 629.9m2 in one day old foals, 30 day old foals, and adults, respectively. While the median alveolar surface density decreased with age, the mean linear intercept increased with age. Alveolar surface area was consistently greater than endothelial surface area (m2) within each lung, however the ratio between alveolar and endothelial surface density remains unchanged with age. The median endothelium surface area was 106.2m2 in one day, 147.5m2 in 30 day and 430m2 in adult lungs. The data show that the foal is born with a functionally developed lung and its basic architecture changes with age. Foal lung development and remodelling postnatally is a result of alveolar expansion paralleled with angiogenesis. ii ACKNOWLEDGMENTS I am thankful for the opportunity to study under the supervision of Dr. Baljit Singh. As I reflect on my experience here at the University of Saskatchewan I recognize how supportive, encouraging and brilliant Dr. Singh has been. It has been a pleasure to learn from someone so passionate about research, learning and teaching. He has been incredibly supportive to me over these two years and for that, I cannot thank him enough. I have learned more than I thought possible not only with respect to my research but also to life in general. And I am thankful to have been given such an opportunity to grow. I would also like to express thanks to my advisory committee members, Dr. Matthias Ochs, Dr. Hugh Townsend and Dr. Jaswant Singh for their inspirational mentoring. I am thankful for the opportunity from Dr. Baljit Singh to learn from Dr. Ochs in Hannover, Germany. The hospitality and accommodation of everyone at the Hannover Medical School will not be forgotten. The generous help and efforts of the students, faculty and support team there was exceptional. I especially want to thank and acknowledge Jan Philipp Schneider for his time and efforts in teaching me the stereology techniques and for continuing to be an advisor. He went above and beyond in welcoming me to the school and Hannover. The team here at the WCVM also deserves much gratitude and thanks. To begin, Dr. Julia Montgomery deserves sincere thanks as she was more than generous with her time in mentoring me and helping with every aspect of my project. I acknowledge Dr. Fernando Marques, for his support, guidance and discussions in determining suitable adult horses for the study, Dr. Stacy Anderson for her assistance with the clinical work, Tara Bocking for being an exceptional and enthusiastic summer student and of course all my lab mates for lending a hand. The collaborative nature of everyone involved with this project, from the WCVM’s Animal Care facility staff, and Prairie Diagnostic Services, to the use of the Olympus microscope system in WCVM’s Pathology Department, was wonderful. I would also like to acknowledge the kind support of the Ryan Dubé Equine Research Fund and the WCVM’s graduate enhancement scholarship. Finally, I want to thank my parents, family, boyfriend and friends for their never-ending support and inspiration. I am lucky to have such a support group who is there for me through every venture. iii TABLE OF CONTENTS PERMISSION TO USE …………………………………………….……………………….......i ABSTRACT ………………………………………………………………………….…..…......ii AKNOWLEDGMENTS …………………………………………………………………….....iii TABLE OF CONTENTS …………………………………………………………………........iv LIST OF TABLES ………………………………………………………..................................vi LIST OF FIGURES …………………………………………………………………………....vii LIST OF ABBREVIATIONS………………………………………………...………………..ix SECTION 1: INTRODUCTION……………………………………………………………….1 SECTION 2: REVIEW OF LITERATURE…………………………………………………....3 2.1 Lung Development 3 2.1.1 Prenatal Development 3 2.1.2 Postnatal Lung Maturation and Growth 5 2.2 Immunological Relevance 8 2.3 Morphometry Theory 10 SECTION 3: HYPOTHESIS AND OBJECTIVES …………………………………………...13 SECTION 4: MATERIALS AND METHODS …………………………………………….…14 4.1 Pilot Studies 4.1.1 Mouse Trial 4.1.2 Cow Lung Trial 4.2Animals 4.3 Fixation 4.4 Tissue Sampling 4.5Tissue Embedding 4.6 Stereological Evaluation 14 14 15 15 17 18 18 19 SECTION 5: STATISTICAL ANALYSIS ………………………………………………...….26 SECTION 6: RESULTS ……………………………………………………………………….27 6.1 Histology 6.2 Stereological Evaluation of Parenchymal Characteristics 6.3Stereological Evaluation of Surface Area Characteristics 6.4 Immunohistochemistry for von Willebrand Factor and Macrophages iv 27 27 28 30 SECTION 7: DISCUSSION …………………………………………………………………...41 7.1 Methods Development 7.2 Lung Stereology of Parenchyma 7.3 Stereological Evaluations of Surface Area 7.4 Molecular Mechanisms of Lung Development 7.5 Study Limitations 41 41 45 48 51 SECTION 8: CONCLUSION AND FUTURE DIRECTIONS ………………………….…….52 LIST OF REFERENCES ……………………………………………………….........................53 APPENDIX I ………………………………………………………...........................................62 APPENDIX II ………………………………………………………………………………….65 v LIST OF TABLES Table A. Lung developmental stages of horse, human and rat in relation to gestational stage and morphology postnatally. Table B. Species comparisons of basic lung morphology data. Table C. Species comparison of lung alveolar surface area density and total area per kg of bodyweight. Table D. Reference values for tracheal wash and bronchoalveolar lavage cytology inclusion criteria for adult horses. Table E. Bronchoalveolar lavage and tracheal wash cytology results from adult horses included in the study. vi LIST OF FIGURES Figure 4.1. Micrograph from a Foal Lung Indicating Regions of Non-Parenchyma. Figure 4.2. Example of the Point Counting Grid. Figure 4.3. Example of the Line Counting Grid. Figure 6.1. Representative lung histology images from Adult Horse, 30 day old Foal and 1 day old Foal Figure 6.2. Volume density of parenchymal characteristics (%) between adults, 30 day and 1 day old foals. Figure 6.3. Relative total lung volume of parenchyma, alveolar airspace, ductal airspace and tissue between adults, 30 day and 1 day old foals. Figure 6.4. Alveolar and endothelial surface area and density counts from adults, 30 day and 1 day old foals. Figure 6.5. Alveolar and endothelial surface density represented as a ratio Figure 6.6. Surface area (m2) per 500mL standardized volume. Figure 6.7. Surface area estimates per kilogram of bodyweight. (A) Alveolar surface area (m2) per body weight (kg). Figure 6.8. Alveolar measurements. Figure 6.9. Immunohistochemistry with anti-macrophage antibody MAC-387. Figure 6.10. Immunohistochemistry for von Willebrand Factor in foal and adult lungs. Figure A. Adult horse hoisted into upright position with right lung clamped off and fixative instilled into the left lung. Figure B. The Right lung has been removed from the chest cavity after fixative has been instilled. The left lung is in the process of extraction. vii Figure C. Image Series of Slicing Adult Equine Left Lung. This images show the large size of the equine lung (60cm in length) shown on a tray ( 65cmX45cm in dimension). Figure D. Slicing Adult Equine Left Lung. Lungs were measured for total longest length measurement and slab thickness was determined based on producing 10-12 equal thickness slabs. Slabs were laid with the cut face upwards all in the same direction. Figure E. Slicing Adult Equine Left Lung. Lung volume was estimated using a point counting grid for the Cavalieri Method. The points on the grid were adjusted to produce 100-200 point counts per lung. Figure F. Slicing Adult Equine Left Lung. Figure G. Slicing Adult Equine Left Lung. Figure H. Cascade design of Lung Tissue Analysis. viii LIST OF ABBREVIATIONS ANOVA Analysis of Variance BAL Bronchioalveolar Lavage COPD Chronic Obstructive Pulmonary Disease DC Dendritic Cell EM Electron Microscopy IHC Immunohistochemistry LM Light Microscopy Lm Mean linear Intercept PIMs Pulmonary Intravascular Macrophage RAO Recurrent Airway Obstruction SURS Systematic Uniform Random Sampling Sv(alv/par) Surface density of alveoli per parenchyma S(alv/lung) Total Alveolar surface area per lung Sv(endo/par) Surface density of endothelium per parenchyma S(endo/lung) Total Endothelial surface area per lung (t) Arithmetic Mean Thickness TW Tracheal Wash Vv(al air/lung) Volume density of alveolar airspace per lung Vv(al air/par) Volume density of alveolar airspace per parenchyma ix Vv(al duct/lung) Volume density of alveolar ductal airspace per lung Vv(al duct/par) Volume density of alveolar ductal airspace per parenchyma Vv(tissue/lung) Volume density of alveolar tissue per lung Vv(tissue/par) Volume density of alveolar tissue per parenchyma Vv(par/lung) Volume density of parenchyma per lung Vv(nonpar/lung) Volume density of non-parenchyma per lung Vv(par/ref) Volume density of parenchyma per reference space Vv(par/par+fine non-par) Volume density of parenchyma per reference space of parenchyma and fine non parenchyma (microscopic images) Vv(par+fine non-par/par+fine no-par+coarse non-par) Volume density of parenchyma and fine non-parenchyma per reference space of parenchyma, fine non-parenchyma and coarse non-parenchma (macroscopic slabs) VEGF Vascular Endothelial Growth Factor x SECTION 1: INTRODUCTION The ultimate limiting factor of racehorse success is the respiratory system, as respiratory disease can reduce horse performance by impacting gas exchange and airway resistance (Hodgson and Rose. 1994). According to a risk-screening questionnaire conducted on 110 horses in Calgary, Alberta, 83% indicated mild to moderate lower airway inflammation based on abnormal bronchiolar lavage (BAL) cytology (Wasko et al. 2011). This high incidence was seen in a random population of horses and ponies of various ages that were involved in performance, pleasure or racing in Alberta (Wasko et al. 2011). The cumulative estimate of the number of horses across Alberta involved in these activities is 41% or a total of 124,774 horses (Evans. 2011). Therefore, the data indicate that a large number of horses are potentially affected by respiratory disease and further assistance to the horse community is needed in identifying, managing and preventing lower airway inflammation. The incidence of respiratory disease found by Wasko and colleagues (Wasko et al. 2011) is comparable to a study of 303 randomly selected Swiss horses in 1991(Bracher et al. 1991). In the Swiss study, a minimum of 88% of horses were affected by subclinical, mild or moderate chronic obstructive pulmonary disease (COPD), currently known as recurrent airway obstruction (RAO), a lower airway inflammatory disease (Bracher et al. 1991). This high percentage of COPD affected horses showed no obvious signs of respiratory disease to their owners (Bracher et al. 1991). Despite significant advances in understanding of equine respiratory biology, there is an obvious need for understanding equine inflammatory airway conditions to improve the treatments and preventative medicine techniques for the horse industry globally. To fully understand the mechanisms of lower airway inflammation associated with many equine diseases such as heaves or Rhodococcus equi infection, which is age specific, we must first identify cellular and structural characteristics of the normal horse lung. Defining the structure of the normal and abnormal equine lung, inherently will improve our understanding of the disease process. Due to the complexity of the cellular interactions comprising lung physiology, there are innate differences between the neonate and adult lung. Since the mechanisms underlying this susceptibility are unclear, there may be some inherent differences in the complement of various cells in the lung of the horse compared to the foal. 1 To effectively study the structure and cellular components of the lung, stereology must be used. Stereology is considered the gold standard in evaluating the lung as it encompasses systematic random sampling and evaluation of the lung (Hsia et al. 2010). This technique allows for the comparisons of structural and cellular differences between samples in order to speculate their implications on physiology. This study develops the use of stereology in our lab to characterise the lung of healthy adult horses and foals to define the basic cellular morphology and propose its impact on immune function. I hypothesize that the normal foal lung will have less structurally differentiated lung morphology in comparison to the adult horse. Therefore, I quantified basic lung morphology by estimating the volume density of lung parenchyma and non-parenchyma structures, alveolar ductal airspace, alveolar airspace, alveolar tissue, and alveoli epithelial and endothelial surface area. Through improved understanding of the basic structures of the normal horse lung using a standardized approach there will be a stronger basis of comparison for respiratory diseases, such as RAO and Rhodococcus equi pneumonia in foals as there will be robust baseline “normal” values. 2 SECTION 2: REVIEW OF LITERATURE 2.1 Lung Development There are three stages/models of prenatal lung development: glandular, cannalicular and saccular stages. Postnatal lung development involves the alveolar stage, microvascular maturation and the stage of late alveolarization (Burri. 2006). Species differences in lung morphology at birth may be due to the precocial or altricial nature of new born animals (Burri. 2006). For example, new born rats are considered relatively altricial as they are very immature at birth (Burri. 2006)while foals would be considered less altricial and more precocial. 2.1.1 Prenatal Lung Development Generally, neonatal lung morphogenesis begins after primary lung budding initiates the development of the bronchial tree (McMurtry. 2002), which is termed the glandular stage (Burri. 2006; McGorum et al. 2006). After the initial branching and conductive airway formation, the lung enters the canalicular phase for gas-exchange tissue development, followed by the saccular stage for parenchymal growth, connective tissue thinning and surfactant production (Burri. 2006; McMurtry. 2002; Copland and Post. 2004). There are inter-species differences in the time frames of lung development and these are summarized in Appendix I, Table A. During the 330 day equine gestation, the canalicular stage of the fetus begins around 190-210 days of gestation with the formation of acinar structure (McGorum et al. 2006). During the saccular stage, thick septal walls of parenchymal airspaces create thinner channels and saccules observed in newborn lung (Burri. 2006) and surfactant protein expression begins during this phase (McMurtry. 2002). The saccular stage begins around 260 days during the beginning of alveolar type II epithelial cell differentiation (Barnard et al. 1982; Paradis. 2006). Surfactant is important in reducing surface tension within the lung during alveolar development to prevent alveolar collapse (McGorum et al. 2006) and a critical component of maintaining healthy, clean and dry lungs of new born foals (Paradis. 2006; Ochs. 2006). In foals, there is some controversy around the time of production of significant quantities of surfactant. By 260 days or 80% of the gestation period, alveolar type II pneumocytes are differentiated and begin synthesizing and storing surfactant (Barnard et al. 1982; Paradis. 2006). Pattle and colleagues 3 suggest surfactant is not fully developed in the foal until after 300 days of gestation or even until after foaling, however traces of surfactant were found at 150 days of gestation (Pattle et al. 1975). Similarly, Arvidson et al. showed phospholipids in lung tissue and dipalmitoyl lecithin in amniotic fluid between 100-150 days of gestation (Arvidson et al. 1975). Through histological evaluations, the fetal horse lung appears to mature in utero showing a single layer of capillaries and mature connective tissue trabeculae (Barnard et al. 1982). Bronchiolar epithelium was incomplete at 300 days of gestation, but by 320 days bronchiolar epithelium, alveoli and vasculature resembled the mature adult tissue (Barnard et al. 1982). However there are no quantitative data to validate these observations. By 320 days of equine gestation, the lung has numerous alveoli, greater in number than respiratory bronchioles, and relatively mature surfactant system (Barnard et al. 1982; Paradis. 2006; Jobe and Ikegami. 2001). Alveolar type II epithelial cells, total volume of lamellar bodies and total alveolar surface area was quantified in normal horse lung tissue samples, and appear to be inversely correlated to respiratory rate (Wirkes et al. 2010). This would suggest that foals have fewer alveolar type II epithelial cells, lamellar bodies and alveolar surface area than the adult horse. However, this issue requires further investigation on the rate of post-natal cellular development. Bulk of bronchial tree formations in humans seem to occur during a relatively short period between the 10th and 14th weeks of gestation followed by some branching of the bronchial tree until week 16 (Bucher and Reid. 1961). At this point, epithelial cells begin to line the bronchial tree and development of alveoli begins at the blunt end of the bronchial tree (Bucher and Reid. 1961). As well, the last generations of the bronchial tree begin to form the respiratory bronchioles through loss of cuboidal epithelium along with capillary ingrowth (Bucher and Reid. 1961). Bronchial cartilage development continues after bronchial branching is complete and until 25 weeks of gestation (Bucher and Reid. 1961). Sacculation in human fetal lung begins during the last stage of pregnancy and continues during postnatal development (Massaro and Massaro. 2002). During this time, a central primary connective tissue layer houses a capillary network which forms projections into airspaces, creating secondary septa and providing the framework for mature alveoli (Burri. 2006). Understanding prenatal lung development is important for clinicians to develop appropriate management strategies for preterm infants (Copland and Post. 2004). Premature births present 4 large problems for neonatologists as one of the major complications are lung immaturity because 75% of premature infant mortality and disability is attributed to immature lung phenotypes (Copland and Post. 2004). Recently, it has been shown that stimulation of NF-kβ signaling in murine fetal lung macrophages disrupts airway branching between the cannalicular and saccular stages (Blackwell et al. 2011). Thus suggesting, the lung is particularly susceptible to injury and infection prior to the saccular stage of lung development (Blackwell et al. 2011). 2.1.2 Postnatal Lung Maturation and Growth Postnatal lung development and maturation has been described previously (Burri. 2006) and is well characterized in the rat, human and mouse lung. The lung at birth is morphologically different between species. Alveolarization and microvascular maturation are largely postnatal events and species dependant in terms of postnatal age and length of time taken for development. The alveolar stage is characterized by a double capillary network with primary and secondary septa, that mature in the “microvascular maturation stage” to form a single more efficient capillary network, which is present in adults (Burri. 2006). Maturation of the double capillary network is due to preferential growth of fusions between the capillary layers extensively transforming the alveolar walls in a relatively short time frame (Burri. 2006). The altricial nature of some species, which have an immature microvascular network in the lung at birth, is an indicator for the maturity of the lung at birth (Burri. 2006). The gas exchange surface area, in humans and rats, increases by about 20-fold during the alveolarization process between birth and adulthood (Burri. 2006). It is thought that alveolarization in foals continues after birth (Paradis. 2006) but it is unclear at what age it ceases. In humans, alveolarization begins prior to birth and can last up to 1-2 years postnatally while the microvasculature maturation begins prior to alveolarization and continues along-side until about 3 years of age (Burri. 2006). In infants with severe chronic lung disease, secondary septation and microvascular maturation can become stunted (Thibeault et al. 2004). In rats, the lung is less mature than in humans at birth, yet alveoli formation occurs within 10 days of birth (Burri. 2006). The major step for improving gas exchange capacity is bulk alveolarization and the expansion of air spaces with microvascular maturation (Bolle et al. 2008). Bulk alveolarization refers to the massive appearance of alveoli within the lung parenchyma (Burri. 2006). The 5 mechanisms of development of gas exchange surface area in rat lung has been characterized into three postnatal events; expansion, replication and subdivision (Blanco. 1995). Development of the gas-exchange surface area of the rat lung likely follows a model where replication (saccule formation) occurs postnatally up to adulthood and alveoli are formed through saccule septation (expansion) to a prescribed volume (Blanco. 1995). Microvascular maturation also occurs postnatally, and is characterized by the formation of a single layered network allowing for improved gas exchange in an adult rat (Bolle et al. 2008). In mice, microvascular maturation and bulk alveolarization occur postnatally (Mund et al. 2008). Similar to human lung development, it is unclear when alveolarization ceases (Burri. 2006; Mund et al. 2008). Lung remodelling occurs and is obvious in mice between 24 hours, 5 and 7 days. The primary saccules which are large and simple at 24 hours are subdivided to form new alveoli, ducts and sacs by day 5 and 7 (Thurlbeck. 1975). Recently, Mund and colleagues (2008) observed new septa formation until postnatal day 36 and alveolar surface area continued to increase to adult levels after alveolarization was complete due to the growth process (Mund et al. 2008). The stage of late alveolarization represents the theory of compensatory alveolar growth late in postnatal human lung development (Burri. 2006). Due to the plasticity of lung tissue, septation could be occurring in adults due to the presence of the prerequisites for alveolar formation (Burri. 2006). The occurrence of late alveolarization in humans is controversial due to the unanswered question: when does alveolar formation cease? Various external environmental and genetic influences on lung development and growth have clouded the understanding of alveolar formation (Burri. 2006). The onset, rate and cessation of alveolar formation are suggested to be controlled by a poorly understood mechanism of postnatal oxygen uptake (Massaro and Massaro. 2002). In rats, it was shown that alveoli formation and surface area can be impacted by underfeeding or by dexamethasone treatment (Massaro et al. 1985). Normal lung growth in humans appears to be in two phases. During the first phase, the lung has more airspace due to lung volume growth being greater than parenchymal septa, while reducing interstitial tissue (Burri. 2006). Here, the lung becomes more functional. In the 2nd phase, lung compartments grow proportionately with body weight and lung volume (Burri. 2006; Zeltner et al. 1987). 6 There are various studies on the morphology of the lung in different species (Wirkes et al. 2010; Thurlbeck. 1975; Gehr and Erni. 1980; Gehr et al. 1978; Lakritz et al. 1995; Constantinopol et al. 1989; Thurlbeck and Angus. 1975). The use of different methodologies in these studies makes it challenging to compare the data from various species. In the horse, only a handful of studies have used stereology to characterize the lung and all have used slightly differing methods of fixation, embedding and analysis protocols. In Appendix I, Table B shows a summarized table of morphological features of the lung among multiple species. In the horse, parenchyma has been reported to comprise between 73% (Constantinopol et al. 1989) and 86% (Gehr and Erni. 1980) of the total lung. Within the parenchyma, alveolar airspace represents as low as 60 % (Lakritz et al. 1995) up to 82.9% (Gehr and Erni. 1980), while parenchyma tissue from 17.1% (Gehr and Erni. 1980) up to 39.8% (Lakritz et al. 1995) of the total lung. In humans, 25% of the lung at birth is tissue, while at about 8-10 years of age 10% of the lung is tissue and the remainder, 90%, parenchyma (Thurlbeck. 1975; Thurlbeck and Angus. 1975). Efficient gas-exchange requires adequate surface area of the total alveolar structure (McGorum et al. 2006). Estimated surface area density provides difficult comparisons due to the type of analysis, light microscopy or electron microscopy and relative magnifications used during analysis. Electron microscopic studies on human lung estimated total lung surface area to be 75% more than that determined with light microscopy (Gehr et al. 1978). This estimation could be applied to various reports of equine alveolar surface area shown in Appendix I, Table C. With the use of an electron microscope, the estimated alveolar surface area was found to be 2,457m2 in a 510kg horse (Gehr and Erni. 1980) and compared to 1577 ± 677m2 in an average horse of 447 ± 62 kg (Wirkes et al. 2010; Constantinopol et al. 1989). Differences in quantity of total alveolar structure could be due to differences in horse breeds, fitness level and methodology used in the studies. For example, standardbred professionally-trained horses (Wirkes et al. 2010; Constantinopol et al. 1989) may have a different level of physiological fitness compared to halfbred geldings with spinal injuries (Wirkes et al. 2010; Gehr and Erni. 1980). The methodologies are fairly similar, except some differences in the fixation of the lungs, which may cause differences in lung volume (Gehr and Erni. 1980). There is a lack of morphometric data focussed on lung development in the horse; however, there is a comprehensive morphometric set of data on terminal bronchiole duct ending (TBDE) 7 density in prenatal, postnatal and mature normal horse lung (Beech et al. 2001). Horses, unlike some other mammals, lack respiratory bronchioles and the terminal bronchiole opens directly to the alveolar duct (McGorum et al. 2006; Tyler et al. 1971). The study by Beech et al., (2001) suggests postnatal development of TBDE as well as increase in gas exchange surface area per TBDE in Thoroughbred horses, may increase pulmonary capacity and athletic performance of race horses (Beech et al. 2001). One drawback to this study was the use of formalin fixation and paraffin-embedding of samples, which could lead to differential tissue shrinkage resulting in misinterpretation of actual lung morphology (Beech et al. 2001). They do argue however that the paraffin sections were floated in warm water which could have reduced some of the shrinkage effects. Another drawback of this study is the small number of foals at certain age groups used for both ponies and Thoroughbred breeds. This study provided the first morphometric data on foal lungs, which provides for good comparisons to the current study however low numbers result in low statistical comparisons. In contrast to the Thoroughbred horse, human TBDE number is complete before birth and did not increase significantly between 0 and 66 weeks of age (Beech et al. 2000). 2.2 Immunological Relevance The structure of the lung and its basic morphology are important in understanding the developmental biology of the lung and diseases specific to this organ. It is believed that more precise definition of the basic structures of the lung is needed to understand its relevance to innate and adaptive immunity. It is essential to understanding innate immune responses in the equine airway because many pathogenic triggers result in acute inflammatory diseases of the lung (Tizard. 2000). Complement, mononuclear phagocytic cells, and cytokines are key components of innate immunity (McGorum et al. 2006), whereas monocytes will later become inflammatory macrophages (Tizard. 2000). Cattle, sheep, pigs, goat and horses have resident mononuclear phagocytes, which are called pulmonary intravascular macrophages (PIMs) (Winkler and Cheville. 1985; Winkler and Cheville. 1987; Rybicka et al. 1974; Longworth et al. 1992; Atwal and Saldanha. 1985; Atwal et al. 1992; Warner. 1996); however, these are conditionally absent in human and rat lung (Aharonson-Raz and Singh. 2010). Some differentiated mononuclear cells were present in new born and 3 day old porcine lungs closely apposed to the endothelium; 8 although, by 7 days these monocytes were differentiated into PIMs similar to porcine lung at 30 and 60 days of age (Winkler and Cheville. 1985). Similarly, within 2 weeks of postnatal development, lambs show a dramatic increase in PIMs (Longworth et al. 1992). Currently, there is a lack of understanding on whether PIMs are present in the foal (Aharonson-Raz and Singh. 2010). The example of recruitment of PIMs during early post-natal life of pigs further underscores a need to use well established methods of stereology to better define the cellular architecture of the lung. Foals are particularly susceptible to infection of Rhodococcus equi a gram-positive, opportunistic pathogen, primarily affecting foals between 1 –3 weeks of age (Knottenbelt et al. 2004). It is thought that infection is dependent on individual hosts rather than environment or farm-specific factors (McGorum et al. 2006; Beech et al. 2001). Therefore, it is important to study differences in the morphology of the lung in various age groups of foals to fully understand the underlying molecular and cellular mechanisms of pulmonary diseases. Typically it is thought that infection occurs within the first several days of life (Horowitz et al. 2001) as lesions have been seen in foals between 10-17 days of age (Horowitz et al. 2001; Cohen et al. 2013), and clinical signs are commonly seen between 1-3 months of age ((McGorum et al. 2006; Horowitz et al. 2001). Adult horses are not affected by the pathogen as they may harbour R. equi in the colon and subsequently found in their excretions (Knottenbelt et al. 2004). Dust inhalation from feces of infected adult horses may be the cause of infection; therefore, removal of feces in paddocks is important in limiting R. equi spread (Knottenbelt et al. 2004). Recently, air concentration of virulent strains of R. equi was significantly associated with the development of pneumonia in foals within 2 weeks of birth (Cohen et al. 2013). In foals, R. equi can cause enteric infection or bronchopneumonia through abscessation of colonic lymphatic tissue or by destroying alveolar macrophages, respectively (Knottenbelt et al. 2004; Johnson et al. 1980). Cellular and humoral immunity, specifically T-helper 1 lymphocytes (Giguere et al. 1999), are required to protect against this pathogen, which are immature in young foals (Knottenbelt et al. 2004). R. equi is similar in its pathogenicity to Mycobacterium tuberculosis in humans where the pathogen invades and survives in alveolar macrophages resulting in granulomatous lesions (McGorum et al. 2006). Therefore, characterising normal foal lung morphology is an attempt to have a better understanding of the normal lung and its ability to ward off infection. 9 Another important immune cell is the dendritic cell. Maturational differences in dendritic cell’s (Merant et al. 2009) may contribute to the increased susceptibility of foals infected with R. equi compared to adult horses (Heller et al. 2010). The survival rate of foals infected with R. equi is 72% and is influenced by the severity of the infection (Ainsworth et al. 1997). It is controversial whether foals that survive will have successful racing careers based on the data that 62% of the infected horses had less than average national earnings (Ainsworth et al. 1997) indicating rhodococcal pneumonia can cause a negative impact on athletic performance of the horse (Barnard et al. 1982; Ainsworth et al. 1997; Bernard et al. 1991). Overall, the current knowledge of lung development and morphometry in the foals and adult horses has several gaps. The aim of this study is to identify and quantify various lung structures to understand the neonate lung and potential immune response in the foal lung compared to adult horse through characterizing lung morphology. In order to effectively study and quantify lung structures, stereology must be employed. 2.3 Morphometry Theory Stereology uses quantification of 2-D structures to statistically define irregular 3-D structural properties through image techniques, while morphometry is the measurement and practical application of stereology (Hsia et al. 2010). Lung stereology provides a method of describing and quantifying the natural lung composition to evaluate architecture, development and disease (Ochs. 2006). Stereology can be used to quantify volume, surface area, length and particle number through various methods of counting using different test systems (Hsia et al. 2010). It provides efficient methods for representing biological systems through appropriate fixation, sampling and processing techniques depending on the desired form of analysis (Ochs. 2006). Stereological methods have been critically evaluated for most efficacious designs recently in the Official Statement from the American Thoracic Society/European Respiratory Society (Hsia et al. 2010). The three essential components of stereology are proper fixation, unbiased organ sampling and processing. Proper, even fixation is critical in maintaining tissue integrity. The goal in fixing the lung is to maintain its shape similar to when it is in the body cavity. Unbiased sampling provides a systematic procedure to obtain tissue samples representative of the organ as a whole in an 10 objective manner. Finally, the processing component evaluates the fixed tissues using specific point or line grids to estimate desired characteristics and produce real numbers representative of the entire organ. Fixation can be performed through instillation of chemicals into the airways (Ochs. 2006) through a cannulated trachea and provides a dependable method of fixing the lung for stereology purposes (Knust et al. 2009). The chemical composition of the fixatives is important and depends on whether the tissues will be analysed with light or transmission electron microscope (Ochs. 2006). The preferred fixative for studies requiring EM is glutaraldehyde as the cell structure is rapidly stabilized to provide a better structural integrity and visualization. However, glutaraldehyde fixation renders the tissue un-suitable for immunohistochemistry (IHC) (Hsia et al. 2010). Therefore, the fixative of choice for studies using IHC is 0.1% glutaraldehyde and 4% paraformaldehyde inorder to best fix the tissue while allowing antibody binding for IHC. Stereology limits assumptions and bias through objective selection of tissue blocks and image fields in order to equally represent the entire sample (Hsia et al. 2010). Accurate sampling provides each tissue section the equal chance of selection for analysis and is ensured through low variability of repetitive independent sampling (Hsia et al. 2010). Estimating the total lung volume is an essential component of the sampling procedure which can be determined using the ‘Cavalieri Method’. This method estimates the total lung volume through quantifying 2D sections (Hsia et al. 2010) . An alternative method of estimating the volume of the lung is through fluid displacement. However, the ‘Cavalieri Method’ provides a more accurate representation of total lung volume as it represents tissue that will be analyzed in subsequent stereological methods as it is determined after tissue fixation, prior to sampling (Ochs. 2006). Sampling is composed of two procedures in which initial tissue sampling is followed by use of a test system to represent the organ from which the tissue was taken (Ochs. 2006). Systematic Uniform Random Sampling (SURS) is an unbiased simple sampling method, to ensure each unit of lung had equal probability of selection (Hsia et al. 2010; Schneider and Ochs. 2013). There are four different test systems; test points, lines, planes and volumes (Ochs. 2006). One fundamental relationship in stereology is the relationship between the dimension of the test system and the dimension of the structure parameter, which must total at least three (Ochs. 11 2006). For example, “test lines (1-D) ‘feel’ surface area (2-D)” (Ochs. 2006). Therefore, we use points and lines to estimate volume and surface area, respectively. Finally, processing transforms the ratio-generated estimates of various organ characteristics into absolute values through a multilevel sampling process (Ochs. 2006). Typically in light microscopy, tissues are embedded in paraffin, causing some tissue shrinkage (Ochs. 2006). However, processing accounts for various types of shrinkage of the entire tissue or random sections within the tissue (Ochs. 2006). Therefore, plastic (glycol methacrylate) embedding and performing osmication and tissue staining in uranyl acetate is recommended to reduce shrinkage (Ochs. 2006). The use of plastics has limitations, since tissue embedding in plastic prevents the use of antibodies during analysis. As a result, when embedding with paraffin, shrinkage must be accounted for during calculations (Knust et al. 2009). 12 SECTION 3: HYPOTHESIS AND OBJECTIVES I hypothesize the foal lung will have a less structurally differentiated lung morphology in comparison to the adult horse. In order to effectively study the lung architecture, such as parenchymal components and surface area, stereology must be employed. Therefore, the first objective was to establish a stereology protocol, new to our laboratory, to evaluate the horse lung. The second objective was to use this method to analyze the normal horse and foal lung to produce a robust set of quantitative data on the architecture of the “normal” foal and adult horse lung. These data allow us to speculate on lung development in the healthy animal and create a better basis of comparison for diseased lungs. 13 SECTION 4: MATERIALS AND METHODS 4.1 Pilot Studies I conducted two pilot studies to test the method previously established in mice (Knust et al. 2009). The procedure was practiced in two lungs obtained from mice euthanized at the Western College of Veterinary Medicine and one cow lung obtained fresh from an abattoir. The objective of conducting pilot studies was to go through mechanics of handling, slicing, fixing and embedding tissues from lungs from small and large animal species. All protocols and experimental procedures were approved by the University Committee on Animal Care and Supply (UCACS) and the Animal Research Ethics Board (AREB). 4.1.1 Mouse trial The mice (N=2) were anaesthetized with xylazine and ketamine and the trachea was exteriorized. Following thoracotomy, the pulmonary vein was clamped in one mouse while in the other it was left unclamped to more closely resemble the scenario in the horse lung. About 5ml of fixative (0.1% GA and 4% PFA) poured into the lungs intratracheally. Because of a leak in the connection between the trachea and the intratracheal tube, one of the lungs did not inflate completely. After in situ fixation for 15minutes, the lung was removed and immersed in fixative in a beaker overnight (12 hours) at 4oC. Both lungs were removed in the morning and one was submerged in agar after volume estimation by volume displacement. The agar was used to determine if it would ease the cutting of the lungs. We placed the lung on a grid which was used to slice the lung into pieces of equal thickness. Because the lung wiggled during cutting, it was difficult to make slices of equal thickness. We did not use a fractionator design for the sampling rather the lungs were sliced into equal distances and laid out to be counted with the Cavalieri principle. Because of an oversight the lungs were not laid cut face up and therefore the point counting method was not accurate as lung surfaces were miscounted. The slices were divided into two parts and the bottom part was embedded in plastic and the top in paraffin. The mouse pilot study was useful to practice plastic embedding and identify any major flaws in the procedure, which could be adapted and modified as needed for application to the horse lung. 14 4.1.2 Cow Lung trial To relate the mouse pilot study experience to a large animal species, we obtained one cow lung fresh from a slaughter house. Since the lung had already deflated, we wanted to add some air to the lung so the fixative could flow in more easily. I inserted and secured a tube in the trachea for pouring the fixative into the lungs without any pressure. I did not find any leakage of the fixative. The lung appeared to re-inflate slightly as we let fixative flow via gravity into the lung. After overnight fixation, we poured agar (4%) over the entire lung to make the surface firm to assist with the sampling. I found that lungs were not evenly fixed most likely because the lung was deflated prior to instillation of the fixative. The uneven fixation and extensive connective tissue found in the bovine lung made slicing difficult. 4.2 Animals Lungs were collected from 4 adult horses and 11 foals (five foals were one day old and six were 30 days old) following euthanasia. The adult horses, mixed quarter horse type breed between 315 years of age, were examined for respiratory disease and classified as “respiratory healthy” prior to their inclusion in the study. I realized very quickly that it is difficult to find a horse with healthy lungs from the general equine population and that there is a lack of precise criteria for classifying a horse as healthy based on pulmonary evaluation. This created challenges and we tested ten horses over a period of eight months for their respiratory health to finally find 4 adult horses for inclusion in our study. Adult horses underwent a complete physical examination including a re-breathing exam, complete blood count, broncho-alveolar lavage (BAL), and tracheal wash (TW) to confirm the healthy state of their airways. Cut off values for the BAL and tracheal wash were set according to established parameters (Koblinger et al. 2011) and the values are included in Appendix I1, Table D. Initially, thoracic radiographs were performed on each horse prior to the TW/BAL and pending the radiographic results, the TW/BAL was performed. We performed the TW/BAL only if the radiographs were “normal”. The radiographs were analyzed by a veterinarian, who is a Diplomate of the American College of Veterinary Radiologists. Dusty climactic conditions and poor quality hay were thought to be major contributors in the difficulty to find the required five “healthy” adult horses. To overcome this challenge we transported the four most “normal” horses to Goodale Farm, the University research farm located outside of the city, and kept on good pasture isolated from other livestock 15 for 5 weeks and re-examined with a BAL and TW (no radiographs taken). Cytology results found three of the horses to be healthy and eligible to participate in the study, Appendix II Table E. The BAL cytologies were evaluated by the same board certified veterinary pathologist. To ensure that my MSc program was finished in a reasonable time the committee decided to have a sample size of four instead of five adult horses. The final group of horses (n=4) ranged in age from 3-16 years of age with a bodyweight between 356-598kg. There were five foals of one day of age and five foals of 30 days of age used in the study. We obtained eleven pregnant mares, draft cross types, from a local breeder and kept them in the Animal Care Unit of the Western College of Veterinary Medicine for a period of two months. Foal checks were performed daily by our research team for three weeks after mare arrival and twice daily a week before expected due dates until all foals were born. All foals were born overnight and were observed for signs of maturity/immaturity (McGorum et al. 2006) (domed forehead, joint laxity, entropian and suckle reflex) within nine hours of foaling by a veterinarian. All foals except for one appeared normal and healthy. All of the foals and dams were housed together, and all foaling occurred within a one month period. One day old foals (n=5) weighed between 58-68kg and 30 day old foals (n=5) weighed between 115-140kg. Foals were assigned to one day or 30 day old groups based on foaling dates, where the first five foals born were assigned to the 30 day old group and the remainder to the one day old group. Prior to euthanasia, mares and foals were sorted using a chute system for safety, which required three students and at least one veterinarian. The mares were given a 0.03mg/kg intramuscular injection of acepromazine to minimize stress of removing the foal. A physical examination including a complete blood count (CBC) was performed and evaluated by a large animal internist. One foal was excluded from the study upon maturity and CBC evaluation due to markers of systemic inflammation. Serum was collected (whole blood spun at 2000 rpm, 20 minutes, 4oC) and stored at -80oC. After euthanasia via pentobarbital overdose, the left lung was fixed in situ within 45 minutes of death and the right lung was sampled and stored at -80oC. The right lung of one foal was fixed and sampled while the left lung was kept for frozen tissue analysis because of barbiturate injection into the heart after euthanasia which could have potentially nicked the left lung altering stereology calculations. 16 4.3 Fixation All stereological methods were based on those recently recommended (Hsia et al. 2010). The foals and adults were positioned in a supine position with their rump on the ground and neck extended upwards, allowing the inflated lung to hang in the body cavity (Appendix II, Figure A, B). The trachea was exteriorized and the right front leg was removed to gain access to the rib cage. The first three ribs were cut and removed allowing access to the chest cavity without damaging the pleural membrane. Fascia was removed to allow access to the tracheal bifurcation. The right lung was tied off with a string and clamped with hemostats to ensure no fixative could leak into the right lung. About 5 inches up from the tracheal bifurcation, a small slit, just large enough to fit the tube, was made in the trachea. The tube (30cm length) was advanced gently into the tracheal bifurcation into the left lung until it would not advance any farther and fixative flowed under gravity. Fixative (0.1% gluataldehyde, 4% paraformaldehyde in 0.2M HEPES buffer) was poured very slowly into the tubing, about 100mL at a time to minimize the bubbles. After 500mL of fixative, for the one day old foals, was poured into the lung, the left lung was felt by hand to ensure it had inflated. At this time the pleural membrane in the chest cavity was broken. Another 300mL was added to ensure the cranial lobe was filled and fixed. I found that occasionally the cranial lobe was not as evenly fixed as the rest of the lung. While the right lung was removed from the body cavity over a period of 10 minutes, the left lung was allowed to fix in situ. The right lung was weighed and samples were collected in a random fashion to be snap frozen and stored in liquid nitrogen until transferring to -80oC freezer. After removing the right lung, the trachea and left lung was removed carefully to prevent any leakage of the fixative. The left lung was stored in a bag full of fixative which was floating in a bucket of water overnight at 4oC to help reduce any lung collapse, distortion or compression (Hsia et al. 2010; Fehrenbach and Ochs. 1998). Next, the tissues were sampled using systematic uniform random sampling (SURS), processed and embedded for light microscopy (LM). One third of the tissues collected will be processed for analysis of LM using glycol methacrylate (plastic), as well as paraffin embedding for immunohistochemistry (IHC). The remainder of the fixed tissues will be later processed for EM. 17 4.4 Tissue Sampling The foal and adult lungs left in the fixative over night at 4oC were measured with a ruler for length determination and approximate height (at the highest point). Due to the large size of the adult lungs, they were cut in approximate half (cranial and caudal regions) to then employ the sampling protocol on each region reducing sampling error. After measuring lung length, lungs were cut into serial sections of equal distance apart to make 10-12 slabs. Each slab was laid down in the same direction with the cut face upwards. The cavalieri method was used for volume estimation with a 2cm, 3cm or 4cm grid overlaid on each slab for point counting to obtain 100200 point counts on the cut surface. Volume estimation of the adult lungs was comprised of the Cavalieri estimate of both caudal and cranial regions of the lung. A series of images in Appendix II, Figures C-G, illustrate the sampling procedure. Pictures of each lung slab were supposed to be taken in order to estimate non-parenchyma airspace, however, this was not performed consistently for each slab of the lung due to some oversight during the rush of the procedures. A fractionator design was employed based on the size of each lung to result in 30 tissue blocks, 10 for each of paraffin, plastic and electron microscopy protocols. The fractionator design was chosen to best count various cell types in the lung. Once the rigorous sampling protocol was complete, tissues were processed immediately for paraffin and plastic embedding. The EM blocks were stored in sodium cacodylate buffer at 4oC for later processing. 4.5 Tissue Embedding Glycolmethacrylate (plastic) was used to embed tissues and a series of uranyl acetate was used to en block stain prior to osmication, dehydration and embedding in plastic (Appendix II, Protocol I). This helped to prevent differential shrinkage, as previously reported (Ochs. 2006). Shrinkage is an important aspect of morphometric studies as paraffin tissues are widely used for immunohistochemistry while unpredictable tissue shrinkage is common (Ochs. 2006). Tissue shrinkage can be classified as global or differential depending on the uniformity of shrinkage (Ochs. 2006). Differential shrinkage, which can occur even with the use of plastic embedding, can be prevented through the use of osmication and staining with uranyl acetate (Ochs. 2006). The tissue blocks were also embedded in paraffin for other tissue analysis. The paraffin embedding was performed with a series of dehydrations and xylene followed by paraffin 18 infiltration and is described in Appendix II, Protocol II. Paraffin blocks were sliced using a microtome at 5µm thickness and mounted on slides coated with poly-L-lysine. These sections were then used for immunohistochemistry antibody standardization with von Willebrand factor (vWf) and an anti-macrophage antibody (Mac-387). 4.6 Stereological Evaluation Tissue blocks embedded in plastic (Technovit7100) were sectioned with a microtome at 1.5µm, mounted and stained with Toluidine blue. Slides were scanned with the Aperio Image (Aperio, ScanScope CS2) scanner at 40X magnification into the computer and at 60X with oil immersion using Olympus VS110 (ASW 2.4 Software). Sections were then evaluated using systematic uniform random sampling (SURS) techniques with ImageScope and evaluated with the Stepanizer (Tschanz et al. 2011). A cascade type system was used to estimate volume density of lung characteristics (Appendix II, Figure H) similar to the cascade sampling procedure according to Cruz-Orive and Weibel (1981) (Hsia et al. 2010; Cruz-Orive and Weibel. 1981; Zeltner and Burri. 1987). To determine the volume density of parenchyma and non-parenchyma, I evaluated macroscopic and microscopic (4X magnification) images. Unfortunately, as indicated earlier limited photos were taken during the Cavalieri volume estimation, leaving 7 evaluable photos of one adult lung and 11 photos of two foal lungs. During the Cavalieri estimation, very large primary bronchioles were omitted from the estimate when a point fell in the middle of the airway. This means that the volume estimation of the lung is slightly inaccurate as it does not include every component of the lung. Upon evaluation of the limited Cavalieri photos, this omission of very large primary bronchiolar airways accounts for about 1-2% of the lung. Based on this small fraction omitted and through careful consideration, all calculations on the density of characteristics of the lung do not include the very large primary bronchioles. The volume density estimations and estimations of absolute volume are therefore slightly skewed as they do not account for these very large airways. Due to this limitation, values reported as “per lung” do not account for the very large primary bronchioles, rather the reference space is the parenchyma, fine non-parenchyma and coarse non –parenchyma. 19 When estimating parenchyma on macroscopic photos of the lung slabs, points fell either on parenchyma and fine non-parenchyma or on coarse non-parenchyma. The fine non-parenchyma was defined as vessels or airways, which are not directly involved in gas exchange and would fit into the 1mm3 tissue block. The coarse non-parenchyma was defined as vessels or airways, which are not directly involved in gas exchange and would not completely fit into the 1mm3 tissue block. During microscopic evaluations of the parenchyma and non-parenchyma tissues, the fine non-parenchyma included the outer edge of the coarse non-parenchyma structures as this was hard to determine on macroscopic photos. Parenchyma was defined by the areas of the lung directly involved in gas exchange; alveoli, alveolar ducts, alveolar tissue and small capillaries within the alveolar septum were included in this count. All the counts were performed by the same observer to minimize the inter-observers variation. Examples of microscopic evaluations are shown below and they include large vessels, airways and capillary beds with distinguishable surrounding tissue not involved in gas exchange, Figure 4.1. Point counting grids on microscopic sections at 4X magnification were used to estimate volume density of parenchyma (Vv(par/lung) and non-parenchyma (Vv(nonpar/lung) (Schneider and Ochs. 2013), and seen in Figure 4.2. Equation (4.1) was with the raw data obtained from the Stepanizer and then applied to equation (4.2) to estimate volume density of parenchyma. Vv(nonpar/ref) = ΣP nonpar/ Σ P ref (4.1) Vv(par/ref)= 1 - Vv(nonpar/ref) (4.2) In order to combine the microscopic and macroscopic evaluations of the lung, to create a “total” estimate of parenchyma and non-parenchyma densities, the following calculations were made. Parenchyma was calculated by multiplying the volume densities from both microscopic and macroscopic evaluations, equation (4.3). Vv(par/lung)= Vv(par/par+fine non-par) * Vv(par+fine non-par/par+fine nonpar+coarse non-par) (4.3) Vv(nonpar/lung)= 1 – Vv(par/lung) (4.4) 20 Parenchyma, subdivided into alveolar airspace, alveolar ductal airspace and alveolar tissue, required 10X magnification for analysis. Equation (4.5), (4.6) and (4.7) are applied to obtain volume density estimates. Vv(al air/lung) = Vv(al air/ par) • Vv(par/lung) (4.5) Vv(al duct/lung) = Vv(al duct/par) • Vv(par/lung) (4.6) Vv(al tissue/lung) = Vv(al tissue/par) • Vv(par/lung) (4.7) Surface density of alveolar surface area (Sv(alv/par) was estimated at 60X objective magnification with oil immersion using line intersection counting techniques (Figure 4.3) and practical recommendations by Schneider and Ochs, 2013 (Schneider and Ochs. 2013). Equation (4.8) was used to calculate surface density of alveolar surface area and equation (4.9) was applied to calculate total surface area per left lung. Sv(alv/par)= 2 • (Σ I alv)/(l(p) • Σ P par) (4.8) S alv,lung = Sv(alv/par) * Vv(par/lung) * V lung (4.9) Endothelial surface area density was estimated using equation (4.10) and total endothelial surface area per lung was estimated using (4.11). Slides were scanned at 60X with oil immersion inorder to identify endothelium. Sv(endo/par) = 2 • (Σ I endo)/(l(p) • Σ P par) (4.10) Sendo, lung = Sv(endo/par) * Vv(par/lung) * V lung (4.11) Mean linear intercept (Lm), shown in equation (4.12), estimates the volume to surface ratio of the lung airspace (Hsia et al. 2010). This was calculated using the volume of alveolar airspace and ductal airspace, and the surface area estimate per lung. The Lm measures the “mean free distance” of the acinar complex and the gas exchangeable surfaces (Hsia et al. 2010). Lm = 4 * (Vair+duct)/Sa (4.12) The arithmetic mean thickness (t) of the alveolar septum was measured using equation (4.13). It is the ratio between alveolar tissue density and alveolar surface density (Hsia et al. 2010). 21 t = 2 * Vv(tissue,lung)/Sv(alv, lung) (4.13) In order to determine the optimal point counting grid densities and to obtain sufficient fields of view on scanned tissue samples for each step of sampling, multiple densities and step lengths were tested. Ultimately, 100 fields of view per animal (from 5 different blocks of tissue) were used to evaluate the parenchyma and non-parenchyma volume densities. There were 100 fields of view sampled per tissue section for volume density of each of the parenchyma parameters, alveolar surface area and endothelial surface density estimates. At least 100-200 point counting events were obtained for each characteristic of interest following recommended sampling guidelines (Schneider and Ochs. 2013). Maintaining the same level of precision and accuracy among all samples was important and taken into account. The artificial edge effect was avoided by omitting photos sampled when the edge was through the middle of the counting field of view. Any damaged tissue or section was also omitted, as well as tissue where we were not certain of the counting characteristics. 22 Figure 4.1. Micrograph from a Foal Lung Indicating Regions of Non-Parenchyma. The red arrows indicate vessels or bronchioles which were counted as non-parenchyma during the volume density estimations. The remaining alveolar airspace, ducts and tissue were considered parenchyma. Original magnification, 4x. Magnification bar 600µm. 23 Figure 4.2. Example of the point counting grid. This grid was used to determine the volume density of parenchymal characteristics. Original magnification, 4x. Magnification bar 600µm. 24 Figure 4.3. Example of a Line Counting Grid. This type of grid was used for alveolar and endothelial surface area using a line counting grid. Original magnification, 10x and magnification bar is 300µm. 25 SECTION 5: STATISTICAL ANALYSIS Statistical analysis was performed using statistical software (Prism 6 for Windows, Version 6.02, GraphPad Software Inc.) to compare lung values for adult horses, 30 day and 1 day old foals. Because the majority of the data were not normally distributed and was with unequal variances (Bartlett and Brown-Forsythe tests), the data were rank transformed prior to analysis. One-way ANOVA was performed on the ranked data and Tukey’s multiple comparison post-hoc test was used to detect differences among the means of the ranks for each age group. Results are reported using medians and inter-quartile ranges. P-values less than 0.05 were considered significant. 26 SECTION 6: RESULTS I used a customized cascade design to systematically analyse the lung tissues with Aperio Image software and the STEPanizer (Tschanz et al. 2011). The lungs were evaluated for the volume density of gas exchangeable region (parenchyma, Vv(par,lung)) compared to non-gas exchangeable regions (non-parenchyma, Vv(nonpar,lung)). Next, the parenchyma region was estimated as volume density of alveolar airspace (Vv(air,lung)), ductal airspace (Vv(duct,lung)) and alveolar tissue (Vv(tissue,lung)) using point counting grids. Surface density of alveolar surface area (Sv(alv,par) and endothelial surface area (Sv(endo,par) was estimated using intersection counting. Estimated total surface area of alveoli and endothelium were also calculated on a per lung basis. All graphs indicate the median value for each set of data with a horizontal line. 6.1 Histology Representative histology images at 4X magnification with the LM show the difference in the lung between the three age groups. On simple observation, relative size of alveoli is smaller in foals compared to adults (Figure 6.1A-F). As well, lamellar bodies are present in one day old, 30 day old and adult horses (Figure 6.1G-I). 6.2 Stereological Evaluation of Parenchymal Characteristics Total lung volume was estimated on the left lung using the Cavalieri Principle and by multiplying this reference left lung volume by 2. The median adult lung volume (13.9L; 9.3 to 15.5 L) was significantly greater than both the lung volume of the 30 day old foals (4.0 L; 3.7 to 4.4L; P≤0.001) and the one day old foals (2.6L; 1.70 to 3.2L; P≤0.001). The 30 day old foals had a greater reference lung volume than one day old foals (P≤0.001). All parenchymal volume density characteristics are shown in Figure 6.2. Volume density of alveolar airspace per lung was found to be statistically significant between the three age groups. The median alveolar airspace density in adult horses (66.9%; 63.5 to 77.6%) was significantly greater than the 30 day old foals (55.5%; 47.8 to 56.2%; P≤0.05) and one day old foals (45.9%; 40.7 to 48.8%; P≤0.01). The 30 day old foals had a greater volume density of alveolar airspace than one day old foals (P≤0.05). 27 There was no difference (P=0.2161) in the median ductal airspace density of adult horses (11.8%; 6.9 to 18.7%), 30 day old foals (19.6%; 13.2 to 21.4%) and one day old foals (21.7%; 13.4 to 25.3%). There was also no difference (P=0.3071) in the median alveolar tissue volume density (%) of adult horses (11.9%; 8.1 to 15.9%) 30 day old foals (13.6%; 13.1 to 17.0%) and one day old foals (19.9%; 9.72 to 22.6%). Volume density of parenchyma per lung (%) was significantly different between the three age groups. The median parenchyma per lung of adult horses was greater (93.7%; 93.3 to 94.2%) than 30 day foals (84.4%; 83.4 to 87.4%; P≤0.01) and one day old foals (81.0%; 80.1 to 82.7%; P≤0.001). There was a greater volume density of parenchyma in the 30 day old foals compared to the one day old foals (P≤0.01). Absolute volume of parenchymal characteristics was also calculated based on the reference lung volume (Figure 6.3). The estimated median volume of alveolar airspace per lung for adult horses (4.8L; 3.1 to 5.5L) is greater than 30 day old foals (1.1L; 0.68 to 1.3L; P≤0.01) and one day old foals (0.55L; 0.41 to 0.68L; P≤0.001). The 30 day old foals had a greater volume of airspace compared to one day old foals (P≤0.001). The median volume of alveolar ductal airspace in adult horses (0.87L; 0.34 to 1.3L) was greater than one day old foals (0.26L; 0.15 to 0.39L; P≤0.05). There was no difference between adult horses and 30 day old foals (0.26L; 0.12 to 0.39L; P≥0.05), nor between 30 day old foals and one day old foals. The median volume of alveolar tissue per lung in adult horses (0.71L; 0.53 to 0.95L) was greater than 30 day old foals (0.28L; 0.25 to 0.36L; P≤0.05) and one day old foals (0.20L; 0.11 to 0.29L; P≤0.001). There was no difference between the foal age groups. Total volume of parenchyma per lung in adult horses (6.53L; 4.35 to 7.27L) was greater than 30 day old foals (1.69L; 1.56 to 1.90L; P≤0.01) and one day old foals (1.10L; 0.68 to 1.28L; P≤0.001). The 30 day old foals had more parenchyma per lung than one day old foals (P≤0.001). 6.3 Stereological Evaluations of Surface Area Characteristics Total alveolar surface area (m2) per lung was significantly different between the three age groups. The median alveolar surface area (m2) for the adult horses (629.9m2; 452.6 to 936.7m2) 28 was greater than 30 day old foals (258.2m2; 220.4 to 280.2m2; P≤0.01) and one day old foals (205.3m2; 97.2 to 215.3m2; P≤0.001). The 30 day old foals had a greater alveolar surface area compared to the one day old foals (P ≤0.05). All surface area and density graphs are found in Figure 6.4. Total median endothelial surface area (m2) per lung was significantly greater in adult horses (430m2; 288.3 to 675.6m2) than 30 day old foals (147.5m2; 112.3 to 182.8m2; P≤0.01) and one day old foals (106.2m2; 68.1 to 169.3m2; P≤0.01). There was no statistical difference detected between the endothelial surface area in both the groups of foals. The median alveolar surface density per lung for the adult horses (0.053/µm; 0.047 to 0.066/µm) was less than was less than the one day old foals (0.081/µm; 0.069 to 0.091/µm; P≤0.01) but no difference was detected between 30 day old foals (0.075/µm; 0.067 to 0.078/µm). There was no difference between the two groups of foals. The median endothelial surface density per lung was not different (P=0.3183) between the adult horse (0.038/µm; 0.029 to 0.046/µm), one day old foals (0.066/µm; 0.045 to 0.078/µm) or 30 day old foals (0.044/µm; 0.036 to 0.048/µm). Similarly, no difference was seen between the two age groups of foals. A ratio was calculated between the alveolar and endothelial surface density per lung and compared between the three age groups (Figure 6.5). The median alveolar to endothelial surface density ratio was not different (P=0.4047) between the adult horses (1.46; 1.2 to 1.8), 30 day old foals (1.5; 1.5 to 2.1) and one day old foals (1.39; 1.2 to 1.9). Total alveolar surface area per lung was standardized to a set volume size of 500mL (Figure 6.6). The median alveolar surface area per 500mL in adult horses (1.26m2/500mL; 0.91 to 1.9 m2/500mL) was greater than 30 day old foals (0.52 m2/500mL; 0.44 to 0.56 m2/500mL; P≤0.01) and one day old foals (0.41 m2/500mL; 0.20 to 0.43 m2/500mL; P≤ 0.001). The 30 day old foal alveolar surface area per 500mL was also greater than the one day old foals (P≤0.05). Total endothelial surface area per lung was standardized to a set volume size of 500mL. The median endothelial surface area per 500mL in adult horses (0.86 m2 /500mL; 0.58 to 1.35 m2 /500mL) was greater than 30 day old foals (0.30 m2 /500mL; 0.22 to 0.37 m2 /500mL; P≤0.05) 29 and one day old foals (0.21 m2 /500mL; 0.14 to 0.34 m2 /500mL; P≤0.01). There was no difference detected between the foals. The median alveolar surface area per kg of bodyweight in adults (Figure 6.7) in adults (1.32 m2/kg; 1.16 to 1.64 m2/kg) was less than 30 day old foals (1.91 m2/kg; 1.80 to 2.14 m2/kg; P≤0.05) and one day old foals (3.51 m2/kg; 1.64 to 3.78 m2/kg; P≤0.05). There was no difference between the 30 day and one day old foals. Median endothelial surface area (m2) per kg of bodyweight was not different (P=0.072) between adult horses (0.96 m2/kg; 0.69 to 1.21 m2/kg), one day old foals (1.80 m2/kg; 1.14 to 3.00 m2/kg) or 30 day old foals (1.16 m2/kg; 0.94 to 1.32). There was also no difference between the 30 day old foals and one day old foals. The median alveolar air-blood thickness (t) in adult horses (4.51µm; 3.22 to 6.56µm) was not different (P=0.9093) than 30 day old foals (4.96µm; 3.92 to 5.50µm) or one day old foals (5.04µm; 3.61 to 6.65µm) (Figure 6.8). The median value of the mean linear intercept (Lm) for the adult horses (33.1µm; 26.3 to 37.9µm) was greater than one day old foals (18.2µm; 16.4 to 24.8µm; P≤0.05) but not greater than the 30 day old foals (22.5µm; 20.89 to 24.6µm). There was no difference between the 30 day old foals and one day old foals (Figure 6.8). 6.4 Immunohistochemistry for von Willebrand Factor and Macrophages I standardized two antibodies, von Willebrand factor (vwf) and Macrophage marker (Mac-387) in adult horse, 30 day old foal and 1 day old foal lung tissue embedded in paraffin. vWf was stained at 1:500 and Mac-387 at 1:50 for optimal staining (Figures 6.10, 6.9, respectively) . 30 31 Figure 6.1. Representative lung histology images from Adult Horse, 30 day old Foal and 1 day old Foal. A, D, G, 1 day old Foal, B, E, H, 30 day old foal, C, F, I, Adult Horse lung. A,B, C are at original magnification of 4x. D, E, F are at 40x original magnification with oil and G, H, I are at 60x original magnification with oil. Letters indicate lung structures, “a” represents alveoli, “b” represents bronchiole, “d” represents ductal space, and “v” represent vessels. Arrows indicate alveolar epithelium and lightning bolt indicates lamellar bodies. Magnification bar for image A,B,C is 200µm, and for D,E,F,G,H,I is 20 µm. Figure 6.2. Volume density of parenchymal characteristics (%) between adults, 30 day and 1 day old foals. (A) Volume density of alveolar airspace per lung. (B) Volume density of ductal airspace per lung. (C) Volume density of alveolar tissue per lung. (D) Volume density of total parenchyma per lung. Letters “a” and “b” indicate statistical significance between groups (P≤0.05). Median values represented by horizontal line in graphs. 32 Figure 6.3. Relative total lung volume of parenchyma, alveolar airspace, ductal airspace and tissue between adults, 30 day and 1 day old foals. (A) Total volume of parenchyma per lung (mL). (B) Total alveolar airspace volume per lung (mL). (C) Total ductal airspace volume per lung (mL). (D) Total tissue volume per lung (mL). Letters “a”, “b”, and “c” indicate significance between the groups (P≤0.05). Median values represented by horizontal line in graphs. 33 Figure 6.4. Alveolar and endothelial surface area and density counts from adults, 30 day and 1 day old foals. (A) Surface area of alveoli (m2). (B) Surface area of endothelium (m2). (C) Surface density of alveoli (per µm). (D) Surface density of endothelium (per µm). Letters “a”, “b”, “c” indicate statistical significance between the groups (P≤0.05). Median values represented by horizontal line in graphs. 34 Figure 6.5. Alveolar and endothelial surface density represented as a ratio. No statistical significance found between the age groups (P≥0.05). Median values represented by horizontal line in graphs. 35 Figure 6.6. Surface area (m2) per 500mL standardized volume. (A) Alveolar surface area standardized to 500mL volume. (B) Endothelial surface area standardized to 500mL volume. Letters “a”, “b”, “c” indicate statistical significance between groups (P≤0.05). Median values represented by horizontal line in graphs. 36 Figure 6.7. Surface area estimates per kilogram of bodyweight. (A) Alveolar surface area (m2) per body weight (kg). (B) Endothelial surface area (m2) per body weight (kg). Letters “a”, “b”, “c” indicate statistical significance between groups (P≤0.05). Median values represented by horizontal line in graphs. 37 Figure 6.8. Alveolar measurements. (A)Thickness of alveolar septum (t). (B) Mean linear intercept (Lm). Letters “a”, “b”, “c” indicate statistical significance (P≤0.05). Median values represented by horizontal line in graphs. 38 A B C D F E G Figure 6.9. Immunohistochemistry with anti-macrophage antibody MAC-387. While only a few septal macrophages (single arrows) are labeled in one day old foal lung (A-B), many more positive cells are observed in lung sections from 30 day old foal (C-D) and adult horse lungs (EF). Higher magnification views show septal location of MAC-387-positive cells. Larger boxed area in E is magnified in G to show location of MAC387 positive cells within the endothelium. Magnification bar is 50 µm. 39 > A > B C > D E > > F G Figure 6.10. Immunohistochemistry for von Willebrand Factor in foal and adult lungs. Omission of primary antibody resulted in absence of staining a lung section from one day old foal (A), while inclusion of vWF antibody caused staining (arrows) in vascular endothelium but not in bronchiolar epithelium (double arrows) and alveolar septa (arrowheads) in a lung section from one day old foal (B-C). Lung sections from 30 day old foal (D and magnified view of boxed area in E) show staining vascular endothelium (single arrows) but not in airway epithelium (double arrows). Alveolar septa (arrowheads) continue to be either negative or weakly positive. Adult horse lung (F-G) showed staining in vascular endothelium (single arrows) and alveolar septa (arrowheads) but not in airway epithelium (double arrows). Magnification bar is 50 µm. 40 SECTION 7: DISCUSSION I report fundamental data on the basic morphometric characteristics of the equine lung. Within the study are quantitative descriptions of the lung comparing one day and 30 day old foals as well as the adult horse. These data are important because the basic features such as lung capacity, endothelial and alveolar epithelial surface area have implications for pulmonary physiology. 7.1 Methods Development It was difficult to find “normal” healthy adult horses based on BAL, TW and radiographic assessments. Cut-off values for BAL and TW cytology was based on published reference values (see Appendix I). Endoscopic exams were repeated on 4 horses after changing their environment from outdoor paddocks to pasture, which found cytology values within our inclusion criteria range. Repeating the BAL and TW procedures were performed after a 7 day break from the previous endoscopic exam to allow for appropriate recovery time. Although it has been argued that the BAL can be repeated with only one day in between (Tee et al. 2012). The main interest in this study was to evaluate normal equine lung, which meant we must eliminate concern of lower airway inflammation to characterize the “normal” or “healthy” lung morphology. Therefore, both the TW and BAL were performed. It has been suggested that upper (tracheal) and lower (lung) airway endoscopy cytology results are independent, in terms of inflammation (Koblinger et al. 2011) however, we chose to perform both the BAL and TW to rule out the risk of a compromised airway integrity. 7.2 Lung Stereology of Parenchyma I determined the reference lung volume using the Cavalieri principle instead of volume displacement because of the large size of the horse lungs (Hsia et al. 2010; Yan et al. 2003). The Cavalieri method is performed by slicing the lung into equally thick slabs, placing the slabs with the cut face upwards and overlaying a point counting grid. This grid has points of known distance apart, and through accounting for the area per point and the slab thickness an estimate of the organ volume can be determined. The use of the immersion method (volume displacement) compared to the Cavalieri method can cause inflation of the volume of the lung especially in large lungs (Yan et al. 2003). It has also been recommended to use the Cavalieri principle to estimate the volume of fixed lungs as it represents tissue that will be analyzed in subsequent 41 stereological methods (Ochs. 2006). In previous studies, the estimated volume of the horse lung using volume displacement was estimated to be 46.5±3.5L in three Standardbred harness trained horses, 447±36kg (Constantinopol et al. 1989) and 37.65L in two half-bred geldings, about 5 years old weighing 510kg (Gehr and Erni. 1980). My experiments yielded a total reference lung volume with a median of 13.9L (9.3 to 15.5L) in the adult horses, which is lower than those reported in previous studies. The reasons for the differences in lung volume in my study and the others are not clear, but could be attributed to the differences in methods, as I used the Cavalieri Method as opposed to volume displacement for lung volume estimation. As well, the volume estimation used in my study was slightly skewed due to the omission of very large nonparenchymal airways resulting in the total lung volume being at least 1-2% less than an accurate estimation. Lung volume was reported previously in Thoroughbred and pony foals and adult horses (Beech et al. 2001). Left lung volume was estimated with the Cavalieri Principle and reported only on a graph and appears to be between 400mL and 1L for newborn foals, while adult left lung volume appears between 2 and 5L (Beech et al. 2001). It is also noted that the lung volume does not change after 50 weeks of age (Beech et al. 2001). However, the adult lung volume Beech et al., (2001) reports is dramatically lower than reported previously (Constantinopol et al. 1989); (Gehr and Erni. 1980) and also by me. While the foal lung volumes are similar to those I report assuming a total lung volume based on similar volumes for left and right lungs. The lung was characterised by volume density of parenchyma, which was subdivided into alveolar airspace, ductal airspace and alveolar tissue volume densities. These characteristics describe the morphology of the lung and therefore differences detected between the age groups represent change within the lung. The volume density of parenchyma in the adult lung is thought to be between 85-90% while the remaining 10-15% of the lung is the tissue (Weibel et al. 2007). This was generally the case between the three age groups of horses where all three groups were different and adult horses had a greater density of lung parenchyma (median 93.7%) than 30 day foals (median 84.4%) and one day old foals (median 81.0%) (Figure 6.2D). Previously, Gehr (1980) (Gehr and Erni. 1980) reported the parenchymal volume density of the adult equine lung (n=2) to be 87.2% and 85.2%. These values are smaller compared to the values I report, which could be due to the difference in methods as the volume density of parenchyma was estimated 42 prior to tissue sampling in the study by Gehr (1980). In our study, the availability of only a few macroscopic photos of the lung slabs probably led to an overestimation of parenchymal density fraction by 1-2%. In another morphometric study of standard-bred horse and steer lung, the mean parenchymal volume density was 73% and 61%, respectively (Constantinopol et al. 1989). While the lung fixation was done in situ similar to our methods, the lungs however were allowed to collapse prior to fixative instillation (Constantinopol et al. 1989), which may not have fully inflated the lung leading to a perceived lower parenchymal volume density. To my knowledge, the volume density of lung parenchyma has not been examined or compared between foals of different ages. However, there are data to show that volume density of lung parenchyma in children increases with age (Thibeault et al. 2004; Zeltner and Burri. 1987). In mice, it can be assumed that the parenchymal density of the lung grows isometrically with age (Mund et al. 2008). Overall, my data show an increase in the parenchymal volume density in the lung with age similar to other species. The alveolar airspace volume density increased with age, from one day old foals to adult horses (Figure 6.2). In previous studies, alveolar airspace in the adult equine lung was about 71%, which I estimate from their report of airspace density of the parenchymal fraction (Gehr and Erni. 1980) and approximately 68% in normal Thoroughbred race horses in training (Lakritz et al. 1995). In my study, the median value of alveolar airspace in the lung of the adult horses was 66.9%, 55.5% in 30 day old foals and 45.9% in 1 day old foals (Figure 6.2 A). In adult human lung, 86.5% of the lung was airspace, including both ductal and alveolar airspace (Gehr et al. 1978). The volume density of airspaces in children at 1 month of age (72-77%) increased by 6 months of age (85-89%) (Zeltner and Burri. 1987). Finally, in newborn piglets 82% of parenchyma was alveolar airspace, which remained unchanged at 30 days of age (Winkler and Cheville. 1987). It is difficult to compare airspace volume densities between the studies due to a difference in the definitions of the reported values, lack of separation between ductal and alveolar airspace, and even the methods used to arrive at the densities. I separated the alveolar airspace and ductal airspace since we were interested in characterizing the developing lung. The adult equine airspace density estimate reported by both Gehr and Erni (1980) and Lakrtiz et al., (1995) combines alveolar and ductal airspace and are quite close to my estimates. Perhaps their definition of alveolar airspace represents the alveoli, yet the definition of “airspace” is not specified exactly in either report. The human airspace density (Gehr et al. 1978) estimate is 43 slightly larger which could be attributed to the inclusion of both the alveolar airspace and ductal airspace volume density in this value. In Fisher 344 male rats, the volume density of alveolar airspace rapidly increased from 1 month to 3 months of age, while ductal airspace volume density decreased (Mizuuchi et al. 1994). However, in the current study, there was no difference detected in the ductal airspace volume density (P=0.2162) between the adult horses, 30 day old foals and one day old foals. As well, there was no difference detected in the alveolar tissue volume density (P=0.3071) between the adult horses, 30 day old foals and one day old foals. However there appears to be a trend for both duct and tissue where the density is lowest in adult horses, then 30 day old foals and 1 day old foals. These results are less than parenchymal tissue volume densities reported by Lakrtiz (1995) (Lakritz et al. 1995) of adult Thoroughbred race horses, where the volume density of parenchymal tissue was approximately 32%. The lungs were fixed via perfusion and tissue blocks were embedded in paraffin (Lakritz et al. 1995). Perhaps the difference in athletic ability of the horses, type of fixation and embedding could be attributed to the differences seen in parenchymal tissue volume density in the lung. The lack of difference in my results is interesting and could be due to an error, although not very likely, in uneven fixation of the lungs causing variability in the data. However, my results could also be indicative of the foal lung as having a full functional imprint at birth as the ductal and tissue densities do not appear to change. In a study of eight adult human lungs, it was found that 7.8% of lung parenchyma is tissue (Gehr et al. 1978). In adult horses reported previously (n=2), 8.1% and 8.8% of the lung parenchyma is tissue (Gehr and Erni. 1980). In piglets, the volume density of parenchymal tissue at birth was 11%, which remained unchanged at 30 day old pigs (11.7%) (Winkler and Cheville. 1987). In the current study, 11.9% was the median value (n=4) of total parenchymal tissue density in the adult horse lung. There was no difference in volume density detected between the adult horse and foals, however the trend showed alveolar tissue volume density decreasing with age. Similarly, in children parenchymal tissue volume density decreased with age (Thurlbeck. 1975; Thurlbeck and Angus. 1975; Zeltner and Burri. 1987). The absolute volume estimates of the parenchymal characteristics take into account the reference lung volume, which expectedly discovers differences between the foals and adult horses simply 44 due to the size of the lung. The density counts are independent of the lung volume and therefore may be more representative of the differences in the lung between the age groups. 7.3 Stereological Evaluations of Surface Area All of the morphometric analysis was done with the light microscope (LM), and the values are estimated to be about 75 % lower than those obtained with an electron microscope (EM) (Gehr et al. 1978). While it is known that the lower resolution of LM leads to underestimation of surface area compared to EM (Hsia et al. 2010), I am unable to find indication as to which estimate is more accurate. Therefore the values reported previously of alveolar surface area in the horse lung, 2,457m2 (Gehr and Erni. 1980) are likely inflated compared to the data reported in my study due to the differences in microscope magnification. In the human lung, the LM showed alveolar surface area to be 82m2 while the EM study yielded an area of 143m2(Gehr et al. 1978). Based on my data, the alveolar surface area estimate for an adult horse is about 8 times greater than the human lung (Gehr et al. 1978). The alveolar surface area estimates we report for the adult horses and foals show an age related difference, even when the surface area estimate was standardized to a 500mL volume. Not only is the total surface area (m2) increased with age from a median of 205.3m2 at one day of age to 258.2m2 at 30 days of age to 629.9m2 as adult horses (Figure 6.4A). But the alveolar surface density decreases from one day old foals (0.081/µm) to adult horses (0.053/µm) (Figure 6.4C). These data show that in the foal lung alveolar epithelium is more dense than the adult horse, possibly due to alveoli expanding or from entire lung growth. Our results of alveolar surface area are much greater than those previously reported in ponies and thoroughbred horses which range from 20-60m2 in the left lung of one day old foals and 130180m2 per left lung of adult horses (Beech et al. 2001). Yet, both studies show an increase in alveolar surface area (m2) with post natal age (Beech et al. 2001). Their data consisted of left lung measurements from 20 thoroughbred foals, 0-8months of age, 4 yearlings and 4 adult horses (Beech et al. 2001). Perhaps the difference in alveolar surface area in my data compared to Beech et al., (2001) is due to the difference in total lung volume which total surface area is dependent upon. Overall, the increase in alveolar surface area with age in the horse is similar to those reported in the rat (Bolle et al. 2008) and child (Thibeault et al. 2004). The alveolar surface area per kg bodyweight is greatest in the one-day old foals, likely because of basic body size. Where, adult horses (356-530kg) weigh more than foals (580140kg). 45 Similarly, gas-exchangeable surface area per kg bodyweight decreased with age in Thoroughbred foals (Beech et al. 2001). In an earlier study, there were neither multiple animals per age group nor statistics performed on their data (Beech et al. 2001). However, the trend in their data on gasexchangeable surface area per bodyweight is slightly less (between approximately 0.25-2.5m2/kg BW) than results from my experiments which yielded a median of 1.32m2/kg in adults, 1.91m2/kg in 30 day old foals and 3.51m2/kg in one day old foals. Perhaps this indicates that the body and lung subsequently grow at unequal amounts, or isometrically after a certain point of postnatal age which was also noted in the parenchymal characteristics of the mouse lung (Mund et al. 2008). The total endothelial surface area per lung, similar to the alveolar surface area, also increased from the foals to the adult horses. Similar data on the increase in endothelial surface have been reported in lungs of infant children (Zeltner et al. 1987). However, the endothelial surface density was unchanged between the various treatment groups in my study as was observed in newborn and 30 day old pigs (Winkler and Cheville. 1987). Perhaps higher magnification or use of the EM would be a more effective and precise method to determine endothelial surface area as occasionally fields of view were omitted due to low quality resolution and being unable to differentiate endothelium. Endothelial surface area is less than total alveolar surface area (m2) in both horses, as reported here, and humans as reported earlier (Gehr et al. 1978) and in the dog lung, where alveolar surface area was found to be 90m2 with a capillary surface of 72m2 (Siegwart et al. 1971) . Interestingly, analysis of mice lungs show that alveolar surface area is 82.2cm2 while capillary surface area is 124cm2(Knust et al. 2009), and reasons for this finding are unknown. To further understand the relationship between the alveolar and endothelial surface areas, I examined their ratios and found no difference between the various groups of animals in my study. The lack of differences in the ratios may be because of simultaneous expansion of the alveolar and endothelial surfaces for optimization of gas exchange across the septal barrier (Burri. 2006; Warburton et al. 2005) I did not find a change in the thickness of the alveolar septum between the foals and adult horses. This is in contrast to the rat lung where alveolar septal thickness decreases with age, from day 7 (13.4µm), 14 (8.1µm), 21 (5.4µm) and 90 (6.4µm) (Bolle et al. 2008). One of the reasons for the 46 differences between the horse and the rat lungs could be relative immaturity of the rat lung at birth. Alveoli formation in the rat lung begins postnatally (Bolle et al. 2008) rather than being born with somewhat developed alveoli as I found in the foal lung. The estimation of mean linear intercept is an estimate of the volume to surface ratio of acinar airspaces or the distance between gas exchangeable surfaces (Hsia et al. 2010). Thurlbeck (1975) suggested that lung development occurs in two phases where in the first phase alveolar dimensions remain constant during alveolar multiplication and interalveolar distance (Lm) remains the same, but in the second phase the alveoli increase in dimension and should be reflected by an increase in Lm (Thurlbeck. 1975; Thurlbeck and Angus. 1975). My data show an increase in the Lm between one-day old foals and adult horses to suggest an expansion of alveoli. There was no difference between the 30 day old foals and any other age group. Perhaps more animals per group would show significance. Interestingly, the alveolar surface density decreases with age while Lm increases, suggesting larger alveoli with age. Taken with the decrease of alveolar surface area per kg body weight data with age, might suggest lung growth in terms of alveolar expansion at a different rate than body growth. Further studies should be performed to determine when lung development has reached the adult capacity and complete functionality. Overall, the volume density of lung parenchyma and alveolar airspace increases from foals to adult horses, while alveolar surface density decreases and total alveolar and endothelial surface areas increase. The mean linear intercept increases with age and the ratio between alveolar and endothelial surface density remains constant. The evidence from this quantitative descriptive study indicate that foals appear to be born with a functionally developed lung seen by the presence of alveoli and microvasculature, supported by their more precocial nature (Burri. 2006) and physical ability to run within hours of birth (Paradis. 2006). Perhaps the alveoli continue to develop postnatally, through expansion and possibly secondary septation, resembling the second phase of lung development, alveolar expansion, seen in humans (Thurlbeck. 1975). It has been thought that alveolarization likely continues in foals after birth (Paradis. 2006), however a study of basic histology (Barnard et al. 1982) describes the fetal foal lung to resemble the adult in their vascular and trabecular pattern. Nevertheless, few studies on lung development in the foal have been performed making it unclear whether postnatally there is an increase in number or capacity 47 of functional units (Beech et al. 2001). The current results suggest the alveolar surface area is changing possibly due to an expansion in size, as Lm increases, (Thurlbeck. 1975) while simultaneously angiogenesis is occurring through this phase of alveolarization to maintain normal lung development (Warburton et al. 2005; Jakkula et al. 2000). The histology images (Figure 6.1) showing alveolar size support this through simple observation but further analysis on the number and size of alveoli are required for a more thorough stereological evaluation. These differences detected in parenchymal characteristics and gas exchangeable epithelium and endothelium in the lung could have functional implications. The changing proportions of the lung parenchyma indicate there is a greater gas exchangeable capacity with age. As well, there is an increase in alveolar epithelial and endothelial surface areas with age. Presumably, lung function would then increase as the ability for gas exchange is improved. The relative amount of gas exchange and lung capacity is increasing to adulthood. It is understandable that the lung is changing in its relative proportions with growth, which leads to the question if there is an optimal time for exercise in horses and what implications of disease could result from exercise before lung development is complete. Perhaps microvasculature remodelling could have an impact on exercise induced pulmonary hemorrhage in race horses. It has been suggested that remodeling of pulmonary veins could be a precursor or related to the development of exerciseinduced pulmonary hemorrhage in horses (Williams et al. 2013). 7.4 Mechanisms of Lung Development It is difficult to speculate on the cellular and molecular mechanisms of lung development in the foal as it appears they are born with a more functionally developed lung than other reported species and we don't have any data yet. However, with the descriptive data in hand and tissues processed for cellular and molecular analyses, I believe this is the stage to undertake some discussion of the potential mechanisms regulating the post-natal lung development in the horse. The mouse (Mund et al. 2008), rat (Blanco. 1995; Mizuuchi et al. 1994) and Quakko Wallaby (Makanya et al. 2001) appear to have the least developed lungs at birth. The human newborn lung lacks defined alveoli, presenting thick saccules containing a double capillary network (Burri. 1984), which later transform into a single capillary layer (Zeltner et al. 1987). However, in the newborn foal lung, functional alveoli appear to be present which is more similar to the neonatal porcine lung (Winkler and Cheville. 1987). 48 While developing an idea of foal postnatal lung development, it is helpful to refer to established mechanisms in the human. Human lung development appears to be biphasic, where the first phase is characterized by shifts in the proportion of parenchymal characteristics, while the second phase is a growth of these components (Zeltner and Burri. 1987). The increase in surface area is likely due to an increase in the complexity of the air-blood interface and maybe due to septal growth resulting in lengthening of the septa and deepening of alveoli (Zeltner et al. 1987). The aeration process in human infant lung is accompanied by a proportional decrease of interstitial tissue, while airspace volume increases at the same time as capillary volume increases (Burri. 2006). The human infant is born with a specified mass of interstitial cells which are dispersed in the septa during the growth process (Zeltner and Burri. 1987). Perhaps this occurs in foals, where the functional structures are present at birth and they become distributed for alveolar expansion postnatally seen through an increase in alveolar surface area, decrease in alveolar surface density and increase in Lm. The alveolar epithelium is lined by alveolar type I and II pneumocytes (Tyler and Julian. 1991), which have been developed during the canalicular stage of prenatal lung development (Burri. 1984). The alveolar epithelial type II cells play a role in lung repair where they can spread, migrate, proliferate and differentiate into type I cells after lung injury to repair the surface (Ghosh et al. 2013; Crosby and Waters. 2010; Desai et al. 2008). Some of the mechanisms of epithelial cell repair after lung injury are described in a review recently (Crosby and Waters. 2010), however the mechanism of cell differentiation during alveolar expansion and postnatal lung development and injury are somewhat unclear (Morrisey et al. 2013). During the prenatal canalicular stage of lung development, the cuboidal epithelium and capillaries come in close contact to form the first thin alveolar septum regions (Burri. 1984). The epithelium simultaneously flattens with the close apposition with the capillaries and the alveolar epithelium differentiates into the cuboidal epithelial stem cell (type II cell) and subsequently squamous type I cells (Makanya et al. 2001; Burri. 1984). Type II pneumocytes and lamellar body size was evaluated by Schmeidl (2007) during postnatal rat lung development (Schmiedl et al. 2007). Interestingly, the size of the type II cells was smaller during alveolarization than before or as a mature adult (Schmiedl et al. 2007). The size of the lamellar bodies, however, increased during alveolarization (Schmiedl et al. 2007). Alveolar type II cell hyperplasia and hypertrophy are two 49 methods to increase the lamellar body volume per unit of alveolar surface area (Miller and Hook. 1990). In rat lungs, gas exchange surface area improves through a transition from bulk alveolarization to an airspace expansion phase, which includes microvascular maturation and reconstruction of the septa (Bolle et al. 2008). During this phase of maturation, apoptosis is necessary to reduce the number of fibroblasts and alveolar type II cells (Schittny et al. 1998) to assist in the reduction of interstitial tissue for septal thinning during microvascular maturation (Makanya et al. 2001). Through TUNEL assay, for type I and II epithelial cells, programmed cell death could be analyzed (Schittny et al. 1998) in the foal lung similar to the procedure in rat lungs to determine if there is a change in the proportion of pneumocytes postnatally. Likely there is some cellular differentiation and proliferation occurring in the foal lung postnatally with the impact of mechanical forces and other growth factors. It has been suggested that one factor could be insulin-like growth factor-1 (IGF-1) activating the Wnt-Frizzled pathway to regulate alveolar type II cell differentiation into type I cells after lung injury (Ghosh et al. 2013). It is also known that mechanical forces on the lung tissue play a role in cellular proliferation in the developing lung and after lung injury, as reviewed by Liu et al., (1999) (Liu et al. 1995). Cell culture experiments of isolated epithelial cells showed an increased DNA synthesis after strain (Liu et al. 1995). Further, it also important to distinguish if the number of alveoli are increasing postnatally through additional stereology based studies or whether a variety of factors are leading to the expansion of the alveoli. Again, while airspace volume increases, capillary volume must also resemble the same pattern (Burri. 2006) as normal lung development relies also on microvascular maturation (Thibeault et al. 2004). The volume density of parenchymal vessels and epithelial airspace with an air-blood barrier, and capillary loading (number of air-blood barriers per surface area) have been used to describe microvascular maturation in human infants (Thibeault et al. 2004). Angiogenesis and vasculogenesis are precisely regulated and required mechanisms in normal lung development (Brown et al. 1995; Zeng et al. 1998). One of the critical regulating factors of these processes is the vascular endothelial growth factor (VEGF) (Warburton et al. 2005). VEGF has been studied in fetal and preterm infants without lung disease and with bronchopulmonary dysplasia (BPD), where VEGF is important for normal microvascular development (Lassus et al. 50 2001). Immunohistochemical staining of VEGF was found in the bronchial epithelium and alveolar macrophages (Lassus et al. 2001). VEGF also maintains alveolar structure (Zeng et al. 1998). The integrin, αvβ3, also plays a role in endothelial cell spreading and motility (Leavesley et al. 1993; Tang et al. 1996). The endothelial cell responds to the extracellular matrix through these cell adhesion receptor integrins for mediated cell migration during development, healing or angiogenesis (Leavesley et al. 1993). As well pericytes, located in the basement membrane of capillaries, also contribute to angiogenesis and regulation of endothelial cell proliferation (Ribatti et al. 2011). Alveolar macrophages interact with epithelial and endothelial cells and can produce growth factors (Liu et al. 1995). Intravascular macrophages were detected in new born pigs and morphometric characterization showed an increase in absolute volume and to cover more capillary surface in 30 day old pigs (Winkler and Cheville. 1987). Polarized M2 macrophages appear to play a role in lung development as they are found in sites of branching morphogenesis and are increasing in number during alveolarization compared to M1 macrophages, as assessed through qPCR (Jones et al. 2013). I have performed preliminary staining for von Willebrand factor (vwf, 1:500) and macrophage antibody (Mac-387, 1:50) on lung tissues collected in my experiments (Figures 6.10 and 6.9). The role of macrophages as a source of growth factors could be critical in influencing the development of blood vessels and alveoli as they were found in all three age groups. Furthermore, quantification of the macrophages especially those in the vascular compartment could help us in better understanding the innate immune system in the developing lung. In addition to the vascular and alveolar quantification in the lung, the expression of angiogenic molecules such as integrin αvβ3 should also be explored. 7.5 Study Limitations As previously mentioned, the omission of the very large primary bronchioles during the Cavalieri volume estimation and lack of photos taken of the lung slabs during the volume estimation are not ideal. Our best estimations show these missing counts are about 1-2% of the lung. In future studies, these steps will be included. Another limitation of the study is the evaluation of only left lungs (except for one right foal lung). This is assuming homogeneity between right and left lungs. And finally, a larger study size, especially of the adult horse group, would help to improve our statistical significance, however this is always a limitation with large animal studies. 51 SECTION 8: CONCLUSION AND FUTURE DIRECTIONS Through this descriptive and quantitative study of the adult and foal lung morphology there are apparent differences in the newborn lung compared to the adult horse. The hypothesis that foals have a less structurally differentiated lung composition in comparison to the adult horse was confirmed based on lung stereology data describing the parenchymal components and surface areas of alveolar epithelium and endothelium changing from foal to adult horses. The newborn foal lung appears functional at birth, yet development must occur postnatally as revealed through the changes in lung morphology. The differences seen in basic lung structure contribute to understanding lung function between various age groups and for later understanding disease pathophysiology. The objectives of this study were to establish a technique to use stereology to evaluate the equine lung and to then use the technique in healthy horses and foals to describe normal lung parameters. I now have baseline morphometric data of the healthy “normal” foal and adult horse lung and further studies are required to depict the molecular mechanisms of this development. 52 LIST OF REFERENCES 1. Aharonson-Raz K and Singh B. Pulmonary intravascular macrophages and endotoxininduced pulmonary pathophysiology in horses. Can.J.Vet.Res. 74: 1: 45-49, 2010. 2. Ainsworth DM, Yeagar AE, Eicker SW, Erb HE, and Davidow E,. Athletic performance of Horses previously infected with R.equi Pneumonia as Foals. AAEP Proceedings 43: 81-82, 1997. 3. Arvidson G, Astedt B, Ekelund L and Rossdale PD. Surfactant studies in the fetal and neonatal foal. J.Reprod.Fertil.Suppl. (23): 23: 663-665, 1975. 4. Backstrom E, Hogmalm A, Lappalainen U and Bry K. Developmental stage is a major determinant of lung injury in a murine model of bronchopulmonary dysplasia. Pediatr.Res. 69: 4: 312-318, 2011. 5. Barnard K, Leadon DP and Silver IA. Some aspects of tissue maturation in fetal and perinatal foals. J.Reprod.Fertil.Suppl. 32: 589-595, 1982. 6. Beech DJ, Sibbons PD, Howard CV and van Velzen D. Terminal bronchiolar duct ending number does not increase post-natally in normal infants. Early Hum.Dev. 59: 3: 193-200, 2000. 7. Beech DJ, Sibbons PD, Rossdale PD, Ousey JC, Holdstock NB, Chavatte P and Ansari T. Organogenesis of lung and kidney in Thoroughbreds and ponies. Equine Vet.J. 33: 5: 438-445, 2001. 8. Bernard WV,DVM, Dugan J,BS, Pierce S,DVM and Gardiner,I.,DVM,PhD. The Influence of Foal Pneumonia on Future Racing Performance. Proc Am Assoc. Equine Pract. 1718, 1991. 9. Blackwell TS, Hipps AN, Yamamoto Y, Han W, Barham WJ, Ostrowski MC, Yull FE and Prince LS. NF-kappaB signaling in fetal lung macrophages disrupts airway morphogenesis. J.Immunol. 187: 5: 2740-2747, 2011. 10. Blanco LN. Mechanisms for the generation of gas-exchange surface area in rat lung. Am.J.Physiol. 269: 5 Pt 1: L698-708, 1995. 53 11. Bolle I, Eder G, Takenaka S, Ganguly K, Karrasch S, Zeller C, Neuner M, Kreyling WG, Tsuda A and Schulz H. Postnatal lung function in the developing rat. J.Appl.Physiol. 104: 4: 1167-1176, 2008. 12. Bracher V, von Fellenberg R, Winder CN, Gruenig G, Hermann M and Kraehenmann A. An investigation of the incidence of chronic obstructive pulmonary disease (COPD) in random populations of Swiss horses. Equine Vet.J. 23: 2: 136-141, 1991. 13. Brown LF, Yeo KT, Berse B, Morgentaler A, Dvorak HF and Rosen S. Vascular permeability factor (vascular endothelial growth factor) is strongly expressed in the normal male genital tract and is present in substantial quantities in semen. J.Urol. 154: 2 Pt 1: 576-579, 1995. 14. BUCHER U and REID L. Development of the intrasegmental bronchial tree: the pattern of branching and development of cartilage at various stages of intra-uterine life. Thorax 16: 207218, 1961. 15. Burri PH. Structural aspects of postnatal lung development - alveolar formation and growth. Biol.Neonate 89: 4: 313-322, 2006. 16. Burri PH. Fetal and postnatal development of the lung. Annu.Rev.Physiol. 46: 617-628, 1984. 17. Cohen ND, Chaffin MK, Kuskie KR, Syndergaard MK, Blodgett GP and Takai S. Association of perinatal exposure to airborne Rhodococcus equi with risk of pneumonia caused by R equi in foals. Am.J.Vet.Res. 74: 1: 102-109, 2013. 18. Constantinopol M, Jones JH, Weibel ER, Taylor CR, Lindholm A and Karas RH. Oxygen transport during exercise in large mammals. II. Oxygen uptake by the pulmonary gas exchanger. J.Appl.Physiol. 67: 2: 871-878, 1989. 19. Copland I and Post M. Lung development and fetal lung growth. Paediatr.Respir.Rev. 5 Suppl A: S259-64, 2004. 20. Crosby LM and Waters CM. Epithelial repair mechanisms in the lung. Am.J.Physiol.Lung Cell.Mol.Physiol. 298: 6: L715-31, 2010. 54 21. Cruz-Orive LM and Weibel ER. Sampling designs for stereology. J.Microsc. 122: Pt 3: 235-257, 1981. 22. Desai LP, Chapman KE and Waters CM. Mechanical stretch decreases migration of alveolar epithelial cells through mechanisms involving Rac1 and Tiam1. Am.J.Physiol.Lung Cell.Mol.Physiol. 295: 5: L958-65, 2008. 23. Evans V. 2010 Canadian Equine Industry Profile Study Ottawa, Canada: Equine Canada, 2011. 24. Fehrenbach H and Ochs M. Studying lung ultrastructure. In: Methods in Pulmonary Research, edited by Uhlig S and Taylor AE. Basel: Birkhauser, 1998, chapt. 17, p. 429. 25. Gehr P, Bachofen M and Weibel ER. The normal human lung: ultrastructure and morphometric estimation of diffusion capacity. Respir.Physiol. 32: 2: 121-140, 1978. 26. Gehr P and Erni H. Morphometric estimation of pulmonary diffusion capacity in two horse lungs. Respir.Physiol. 41: 2: 199-210, 1980. 27. Ghosh MC, Gorantla VK, Makena PS, Luellen CL, Sinclair SE, Schwingshackl A and Waters CM. Insulin Like Growth Factor-1 Stimulates Differentiation of ATII Cells to ATI-Like Cells Through Activation of Wnt5a. Am.J.Physiol.Lung Cell.Mol.Physiol. 2013. 28. Giguere S, Wilkie BN and Prescott JF. Modulation of cytokine response of pneumonic foals by virulent Rhodococcus equi. Infect.Immun. 67: 10: 5041-5047, 1999. 29. Heller MC, Jackson KA and Watson JL. Identification of immunologically relevant genes in mare and foal dendritic cells responding to infection by Rhodococcus equi. Vet.Immunol.Immunopathol. 136: 1-2: 144-150, 2010. 30. Hodgson DR, Christley RM, Wood JLN, Reid SWJ and Hodgson JL. Relationship between coughing and airway inflammation in young racehorses. Proceedings of a Workshop on Inflammatory Airway Disease: Defining the Syndrome 9: 16, 2002. 31. Hodgson DR and Rose RJ. Principles and practice of equine sports medicine; The athletic Horse. Philadephia, PA: W.B. Saunders Company, 1994. 55 32. Horowitz ML, Cohen ND, Takai S, Becu T, Chaffin MK, Chu KK, Magdesian KG and Martens RJ. Application of Sartwell's model (lognormal distribution of incubation periods) to age at onset and age at death of foals with Rhodococcus equi pneumonia as evidence of perinatal infection. J.Vet.Intern.Med. 15: 3: 171-175, 2001. 33. Hsia CC, Hyde DM, Ochs M, Weibel ER and ATS/ERS Joint Task Force on Quantitative Assessment of Lung Structure. An official research policy statement of the American Thoracic Society/European Respiratory Society: standards for quantitative assessment of lung structure. Am.J.Respir.Crit.Care Med. 181: 4: 394-418, 2010. 34. Jakkula M, Le Cras TD, Gebb S, Hirth KP, Tuder RM, Voelkel NF and Abman SH. Inhibition of angiogenesis decreases alveolarization in the developing rat lung. Am.J.Physiol.Lung Cell.Mol.Physiol. 279: 3: L600-7, 2000. 35. Jobe AH and Ikegami M. Antenatal infection/inflammation and postnatal lung maturation and injury. Respir.Res. 2: 1: 27-32, 2001. 36. Johnson JA, Prestcott JF and Markham RJF. Experimental studies on the pathogenesis of Corynebacterium equi infection in foals. J Comp Med 44: 280-282, 1980. 37. Jones CV, Williams TM, Walker KA, Dickinson H, Sakkal S, Rumballe BA, Little MH, Jenkin G and Ricardo SD. M2 macrophage polarisation is associated with alveolar formation during postnatal lung development. Respir.Res. 14: 41-9921-14-41, 2013. 38. Knottenbelt DC, Holdstock N and Madigan JE. Equine Neonatology Medicine and Surgery. China: Elsevier Science Limited, 2004. 39. Knust J, Ochs M, Gundersen HJ and Nyengaard JR. Stereological estimates of alveolar number and size and capillary length and surface area in mice lungs. Anat.Rec.(Hoboken) 292: 1: 113-122, 2009. 40. Koblinger K, Nicol J, McDonald K, Wasko A, Logie N, Weiss M and Leguillette R. Endoscopic assessment of airway inflammation in horses. J.Vet.Intern.Med. 25: 5: 1118-1126, 2011. 56 41. Lakritz J, Wisner ER, Finucane T, O'Brien TR, Tyler WS, Pascoe JR and Plopper CG. Morphologic and morphometric characterization of lung collagen content in clinically normal adult thoroughbreds in race training. Am.J.Vet.Res. 56: 1: 11-18, 1995. 42. Lassus P, Turanlahti M, Heikkila P, Andersson LC, Nupponen I, Sarnesto A and Andersson S. Pulmonary vascular endothelial growth factor and Flt-1 in fetuses, in acute and chronic lung disease, and in persistent pulmonary hypertension of the newborn. Am.J.Respir.Crit.Care Med. 164: 10 Pt 1: 1981-1987, 2001. 43. Leavesley DI, Schwartz MA, Rosenfeld M and Cheresh DA. Integrin beta 1- and beta 3mediated endothelial cell migration is triggered through distinct signaling mechanisms. J.Cell Biol. 121: 1: 163-170, 1993. 44. Leuenberger A, Gazdhar A, Herrmann G, Ochs M, Geiser T and Knudsen L. Cellspecific expression of human HGF by alveolar type II cells induces remodeling of septal wall tissue in the lung: a morphometric study. J.Appl.Physiol. 113: 5: 799-807, 2012. 45. Liu M, Xu J, Souza P, Tanswell B, Tanswell AK and Post M. The effect of mechanical strain on fetal rat lung cell proliferation: comparison of two- and three-dimensional culture systems. In Vitro Cell.Dev.Biol.Anim. 31: 11: 858-866, 1995. 46. Longworth KE, Westgate AM, Grady MK, Westcott JY and Staub NC. Development of pulmonary intravascular macrophage function in newborn lambs. J.Appl.Physiol. 73: 6: 26082615, 1992. 47. Makanya AN, Sparrow MP, Warui CN, Mwangi DK and Burri PH. Morphological analysis of the postnatally developing marsupial lung: The quokka wallaby. Anat.Rec. 262: 3: 253-265, 2001. 48. Massaro D and Massaro GD. Invited Review: pulmonary alveoli: formation, the "call for oxygen," and other regulators. Am.J.Physiol.Lung Cell.Mol.Physiol. 282: 3: L345-58, 2002. 49. Massaro D, Teich N, Maxwell S, Massaro GD and Whitney P. Postnatal development of alveoli. Regulation and evidence for a critical period in rats. J.Clin.Invest. 76: 4: 1297-1305, 1985. 57 50. McGorum BC, Robinson NE, Dixon PM and Shumacher J. Equine Respiratory Medicine and Surgery. Philadelphia, PA: Saunders Elsevier, 2006, p. 705. 51. McMurtry IF. Introduction: pre- and postnatal lung development, maturation, and plasticity. American Journal of Physiology - Lung Cellular and Molecular Physiology 282: 3: L341-L344, 2002. 52. Merant C, Breathnach CC, Kohler K, Rashid C, Van Meter P and Horohov DW. Young foal and adult horse monocyte-derived dendritic cells differ by their degree of phenotypic maturity. Vet.Immunol.Immunopathol. 131: 1-2: 1-8, 2009. 53. Miller BE and Hook GE. Hypertrophy and hyperplasia of alveolar type II cells in response to silica and other pulmonary toxicants. Environ.Health Perspect. 85: 15-23, 1990. 54. Mizuuchi T, Kida K and Fujino Y. Morphological studies of growth and aging in the lungs of Fischer 344 male rats. Exp.Gerontol. 29: 5: 553-567, 1994. 55. Morrisey EE, Cardoso WV, Lane RH, Rabinovitch M, Abman SH, Ai X, Albertine KH, Bland RD, Chapman HA, Checkley W, Epstein JA, Kintner CR, Kumar M, Minoo P, Mariani TJ, McDonald DM, Mukouyama YS, Prince LS, Reese J, Rossant J, Shi W, Sun X, Werb Z, Whitsett JA, Gail D, Blaisdell CJ and Lin QS. Molecular determinants of lung development. Ann.Am.Thorac.Soc. 10: 2: S12-6, 2013. 56. Mund SI, Stampanoni M and Schittny JC. Developmental alveolarization of the mouse lung. Dev.Dyn. 237: 8: 2108-2116, 2008. 57. Ochs M. A brief update on lung stereology. J.Microsc. 222: Pt 3: 188-200, 2006. 58. Paradis MR. Equine Neonatal Medicine A Case-Based Approach. Philadelphia, PA: Saunders Elsevier, 2006, p. 304. 59. Pattle RE, Rossdale PD, Schock C and Creasey JM. The development of the lung and its surfactant in the foal and in other species. J.Reprod.Fertil.Suppl. (23): 23: 651-657, 1975. 60. Ribatti D, Nico B and Crivellato E. The role of pericytes in angiogenesis. Int.J.Dev.Biol. 55: 3: 261-268, 2011. 58 61. Schittny JC, Djonov V, Fine A and Burri PH. Programmed cell death contributes to postnatal lung development. Am.J.Respir.Cell Mol.Biol. 18: 6: 786-793, 1998. 62. Schmiedl A, Vieten G, Muhlfeld C and Bernhard W. Distribution of intracellular and secreted surfactant during postnatal rat lung development. Pediatr.Pulmonol. 42: 6: 548-562, 2007. 63. Schneider JP and Ochs M. Stereology of the lung. Methods Cell Biol. 113: 257-294, 2013. 64. Siegwart B, Gehr P, Gil J and Weibel ER. Morphometric estimation of pulmonary diffusion capacity. IV. The normal dog lung. Respir.Physiol. 13: 2: 141-159, 1971. 65. Tang DG, Diglio CA and Honn KV. Tyrosine Phosphorylation of a ~30 kD Protein Precedes avb3 Integrin-signaled Endothelial Cell Spreading and Motility on Matrix Proteins. Pathol.Oncol.Res. 2: 1-2: 21-29, 1996. 66. Tee SY, Dart AJ, MacDonald MH, Perkins NR, Horadagoda N and Jeffcott LB. Effects of collecting serial tracheal aspirate and bronchoalveolar lavage samples on the cytological findings of subsequent fluid samples in healthy Standardbred horses. Aust.Vet.J. 90: 7: 247-251, 2012. 67. Thibeault DW, Mabry SM, Norberg M, Truog WE and Ekekezie II. Lung microvascular adaptation in infants with chronic lung disease. Biol.Neonate 85: 4: 273-282, 2004. 68. Thurlbeck WM. Lung growth and alveolar multiplication. Pathobiol.Annu. 5: 1-34, 1975. 69. Thurlbeck WM and Angus GE. Growth and aging of the normal human lung. Chest 67: 2 Suppl: 3S-6S, 1975. 70. Tizard IR. Veterinary Immunology An Introduction. Philadelphia, PA: W.B. Saunders Company, 2000, p. 482. 71. Tschanz SA, Burri PH and Weibel ER. A simple tool for stereological assessment of digital images: the STEPanizer. J.Microsc. 243: 1: 47-59, 2011. 72. Tyler WS, Gillespie JR and Nowell JA. Modern functional morphology of the equine lung. Equine Vet.J. 3: 3: 84-94, 1971. 59 73. Tyler WS and Julian MD. Gross and Subgross Anatomy of Lungs, Pleura, Connective Tissue Septa, Distal Airways, and Structural Units. In: , edited by Parent RA. Boca Raton: CRC Press, 1991, p. 37-48. 74. Warburton D, Bellusci S, De Langhe S, Del Moral PM, Fleury V, Mailleux A, Tefft D, Unbekandt M, Wang K and Shi W. Molecular mechanisms of early lung specification and branching morphogenesis. Pediatr.Res. 57: 5 Pt 2: 26R-37R, 2005. 75. Wasko AJ, Barkema HW, Nicol J, Fernandez N, Logie N and Leguillette R. Evaluation of a risk-screening questionnaire to detect equine lung inflammation: results of a large field study. Equine Vet.J. 43: 2: 145-152, 2011. 76. Weibel ER, Hsia CC and Ochs M. How much is there really? Why stereology is essential in lung morphometry. J.Appl.Physiol. 102: 1: 459-467, 2007. 77. Williams KJ, Robinson NE, Defeijter-Rupp H, Millerick-May M, Stack A, Hauptman J and Derksen FJ. Distribution of venous remodeling in exercise-induced pulmonary hemorrhage of horses follows reported blood flow distribution in the equine lung. J.Appl.Physiol. 114: 7: 869-878, 2013. 78. Winkler GC and Cheville NF. Postnatal colonization of porcine lung capillaries by intravascular macrophages: an ultrastructural, morphometric analysis. Microvasc.Res. 33: 2: 224232, 1987. 79. Winkler GC and Cheville NF. Monocytic origin and postnatal mitosis of intravascular macrophages in the porcine lung. J.Leukoc.Biol. 38: 4: 471-480, 1985. 80. Wirkes A, Jung K, Ochs M and Muhlfeld C. Allometry of the mammalian intracellular pulmonary surfactant system. J.Appl.Physiol. 109: 6: 1662-1669, 2010. 81. Yamada T, Suzuki E, Gejyo F and Ushiki T. Developmental changes in the structure of the rat fetal lung, with special reference to the airway smooth muscle and vasculature. Arch.Histol.Cytol. 65: 1: 55-69, 2002. 60 82. Yan X, Polo Carbayo JJ, Weibel ER and Hsia CC. Variation of lung volume after fixation when measured by immersion or Cavalieri method. Am.J.Physiol.Lung Cell.Mol.Physiol. 284: 1: L242-5, 2003. 83. Zeltner TB and Burri PH. The postnatal development and growth of the human lung. II. Morphology. Respir.Physiol. 67: 3: 269-282, 1987. 84. Zeltner TB, Caduff JH, Gehr P, Pfenninger J and Burri PH. The postnatal development and growth of the human lung. I. Morphometry. Respir.Physiol. 67: 3: 247-267, 1987. 85. Zeng X, Wert SE, Federici R, Peters KG and Whitsett JA. VEGF enhances pulmonary vasculogenesis and disrupts lung morphogenesis in vivo. Dev.Dyn. 211: 3: 215-227, 1998. 61 APPENDIX I Table A. Lung developmental stages of horse, human and rat in relation to gestational stage and morphology postnatally. Species Developmental Stages Gestational Stage Horse Glandular Continues until about 200 days postnatally (Pattle et al. 1975). Transition at 190-201days (McGorum et al. 2006; Pattle et al. 1975) Unclear but potentially after 260 days, when alveolar type II cells are differentiating (Barnard et al. 1982; Paradis. 2006) 16-26 wks(Burri. 2006) (add Copland 2004) 24-38 wks(Burri. 2006) (Copland 2004) Cannalicular Stage Saccular Stage Human Cannalicular Stage Saccular Stage Mouse Cannalicular Stage Saccular Stage 16-18 days prenatal 17.5 days prenatal -day 5 postnatal (Backstrom et al. 2011) Rat Cannalicular Stage Saccular Stage 18.5 days- birth (Yamada et al. 2002) Postnatally 62 Lung morphology at Birth Alveolarization continues (Paradis. 2006; Beech et al. 2001) until a specific age is unclear Alveolar stages from 36 weeks to 1-2 years. (copland 2004) Microvascular maturation from birth to 2-3 years(Burri. 2006) Microvascular Maturation postnatally (Mund et al. 2008) Alveolar stage continues from day 5-28 postnatal (Backstrom et al. 2011) Sacculation commences (day 4-13), not yet alveolar stage (Blanco. 1995) Table B. Species comparisons of basic lung morphology data. Various authors and study designs. Species and Reference Lung Volume and Fixation method BW (kg) Age Vv(Par, lung) Vv(air, lung) SA (alveoli m2) and SA density SA (endo m2) and SA density Adult horse 46.5±3.5L with volume displacement 447+/36kg ? 73±1% ? 2.62m2/kg BW, 2.55m2/kg BW, 253±16.2/cm 245± 10.8/cm 1.55±0.075m2/kg BW 1.38±0.026m2/ kg BW 290±9.9/cm 256±14.6/cm (n=3) (Constantinopol et al. 1989) In situ fixation Steer (n=3) 25.5±1.01L with volume displacement (Constantinopol et al. 1989) ? 61±2% ? In situ fixation Equine adult (n=2) 37.65L via fluid displacement (Gehr and Erni. 1980) In situ fixation Foals and Adults (total n=29) Lung volume estimate via Cavalieri’s Principle Fixed by immersion (Gehr et al. 1978; Beech et al. 2001) 474±12kg 510kg 5 years 86% 26.9L or 70.8% 2,457m2 or 756/cm (0.0756/um) 1663 m2 or 525/cm (0.0525/um) TB and ponies, various body weights 0-8months (n=20), ? ? 20-60m2 at <1 week old ? 1-2 years (n=5), 60m2 at 3-4 weeks 100-200m2 after 51 weeks 3-18 years (n=4) 74kg BW ? 86.5% Airspace 126±12m2 , 4300mL (Gehr et al. 1978) In situ fixation Wirkes 2010 46±6.3L Volume displacement 446±61kg ? V(par,lung) estimated by V(lung) x0.85 ? 1577+/- 677m2 total, SA decreased with body mass ? ? ? 24-60 months ? ? 0.028-0.033/um (no total SA because no reference lung volume) ? 213 ±31 mL (n=3)(Wirkes et al. 2010; Constantinopol et al. 1989) Adult TB Race Horses (n=7) (Lakritz et al. 1995) Assume 90% Parenchyma 143±12m2 with EM Human (n=8) 63 Table C. Species comparison of lung alveolar surface area density and total area per kg of bodyweight. Magnification is based on light microscope (LM) or electron microscope (EM) as magnification is relative for estimating surface area. Author Species Surface Density of Alveoli Total Surface Area of Alveoli per kg BW Magnification (LM or EM) (Constantinopol et al. 1989) Steer, n=3 0.029 +/- 0.0099/µm 2.62 +/- 0.104m2/kg BW EM (Constantinopol et al. 1989) Horse, n=3 0.0253 +/0.00162/µm 1.55+/- 0.075m2/kg BW *significance EM (Gehr and Erni. 1980) Horse, n=2 0.0756/µm (EM) 4.817m2/kg BW 10,000x EM (Beech et al. 2001) Horse/foal, n=29 Not available 0.25-2.5 m2/kg BW ? (Lakritz et al. 1995) Horse, n=7 0.028-0.033/µm (LM) Not available LM (? x) (Gehr et al. 1978) Human, n=8 0.037 +/- 0.003/µm 1.93m2/kg BW LM and EM (Wirkes et al. 2010) Various, horse (n=3) from Gher Not available 1577+/-677m2 (per how many kg??) 40x magnification (Knust et al. 2009) Adult mouse Not available 82.2cm2, capillary SA 124cm2 BW was 20.6g LM (Leuenberger et al. 2012) Adult male rat Not available 2.09 x103 cm2 for 240-280g BW EM 64 APPENDIX II Table D. Reference values for tracheal wash and bronchoalveolar lavage cytology inclusion criteria for adult horses. Cells Tracheal Wash (Hodgson et al. 2002) BAL (Koblinger et al. 2011) Neutrophils (%) ≤ 20 <10 Eosinophils (%) -- ≤1 Mast Cells (%) -- <2 Table E. Bronchoalveolar lavage and tracheal wash cytology results from adult horses included in the study. Cells Neutrophils Eosinophils Macrophages Lymphocytes Adult Horse 1 BAL 0.5% 3% 73.5% 23% Adult Horse 1 TW 11% - 85% 1.5% Adult Horse 2 BAL 1% - 79% 19.5% Adult Horse 2 TW 8% - 83.5% 8.5% Adult Horse 3 BAL 1% 0.5% 56% 42.5% Adult Horse 3 TW 23.5% 0.5% 60.5% 14% Adult Horse 4 BAL - 1% 88% 10.5% Adult Horse 4 TW 51.5% 1% 40% 7.5% 65 Figure A. Adult horse hoisted into upright position with right lung clamped off and fixative instilled into the left lung. 66 Figure B. The Right lung has been removed from the chest cavity after fixative has been instilled. The left lung is in the process of extraction. 67 Protocol 1. Plastic Embedding. Prepare HEPES, prepare cacodylate buffer, 1% OsO4 in 0.1M sodium cacodylate - Preparation of solutions : o 100mL of technovit o 1g hardener I (1bag) = infiltration 1-12 hours or longer depending on type of tissue and thickness o Takes about 2 hours to mix well - Embedding o 15ml preparation solution o 1ml hardener II = curing will take approx. 2 hours (consider pot life) o Only prepare right before you need it.. it will solidify in the beaker otherwise 1) Rinse samples with 0.2M HEPES, 2X6min 2) Rinse with 0.1M sodium cacodylate buffer 4 X 6 min a. Sodium cacodylate is in our lab in the chemicals cabinet 3) Post fix in 1% OsO4 in 0.1M sodium cacodylate for 2 hours 4) Rinse with 0.1M sodium cacodylate 4 x 5 min 5) Bloc-stain win half-saturated aqueous uranyl acetate overnight at 4C 6) Rinse with bidistilled water 2 x 5 min 7) Dehydrate specimens with 70% (2hr, RT), 96% (2hr, RT) and 100% (1h, RT) ethanol (this step can be done in the processor machine, program 10 “unlabelled”) 8) Transfer specimens into 1:1 mix of 100% ethanol and technovit 7100 for 2 hours (RT) a. Enough solution that the lungs are submerged 9) Infiltrate specimens with infiltration medium (GMA/technovit 7100 + hardener I) overnight (RT) a. 1g hardener (=1bag) dissolved into 100ml base liquid (Technovit 7100)  make this before hand because it takes about 1-2 hours to mix well b. Vacuum and agitation helpful – but not necessary c. At 4C solution is stable for 4 weeks 10) Prepare embedding medium (next day) a. 1ml hardener is added (via pipette) to 15ml of prep solution (infiltration medium) b. ONLY PREPARE ENOUGH FOR BOTTOM OF MOULDS 11) Prepare embedding moulds by polymerizing a thin layer of embedding medium (infiltration medium plus hardener II) about 1-3ml, for 1 hr at 400C a. This will form a solid layer at the bottom of the moulds so the tissues are not exposed to air 12) Prepare more embedding medium to fill moulds and block holders. 13) Transfer specimens onto polymerized layer and fill mold with fresh embedding medium 14) Polymerization is performed for 1-2 hours at 400C and subsequently overnight at RT before microtome holder is fixed 15) After polymerization, blocks are to be put into sealable glass containers for storage. 16) Analysis at LM level after sectioning via glass knives and mounting on slides. 68 Protocol 2. Paraffin Embedding Protocol - 70% EtOH 45 minutes 70% EtOH 45 minutes 70% EtOH 45 minutes 80% EtOH, 45 minutes 95% EtOH, 45 minutes 100% EtOH, 45 minutes 100% EtOH, 45 minutes 100% EtOH, 45 minutes Clearant (zylene) 45 minutes Clearant (zylene) 45 minutes Paraffin 30 minutes Paraffin 30 minutes Paraffin 30 minutes 69 Figure C. Image Series of Slicing Adult Equine Left Lung. This images show the large size of the equine lung (60cm in length) shown on a tray (65cmX45cm in dimension). Figure D. Slicing Adult Equine Left Lung. Lungs were measured for total longest length measurement and slab thickness was determined based on producing 10-12 equal thickness slabs. Slabs were laid with the cut face upwards all in the same direction. 70 Figure E. Slicing Adult Equine Left Lung. Lung volume was estimated using a point counting grid for the Cavalieri Method. The points on the grid were adjusted to produce 100-200 point counts per lung. Figure F. Slicing Adult Equine Left Lung. Lung slabs were laid out and the fractionator design was employed to result in a systematically random sample of slabs, to then be cut into slices of equal thickness. These slices were then laid out in order of size selecting every nth piece to result in 5 slices for further processing. 71 Figure G. Slicing Adult Equine Left Lung. The five remaining slices were then cut into thirds and assigned to electron microscopy, plastic embedding and paraffin embedding. Subsampling of these blocks was required to fit into the embedding moulds, following a fractionator design. Figure H. Cascade design of Lung Tissue Analysis. This flow chart indicates the stereological approach to evaluating the lung using the cascade design. Starting with 100% of the lung under macroscopic evaluations, determining parenchyma and non-parenchymal tissues. The next step under microscopic analysis evaluates the fine parenchyma tissues and estimates alveolar surface area. 72