Cardio-respiratory development in bird embryos : new insights from a venerable animal model

The avian embryo is a time-honored animal model for understanding vertebrate development. A key area of extensive study using bird embryos centers on developmental phenotypic plasticity of the cardio-respiratory system and how its normal development can be affected by abiotic factors such as temperature and oxygen availability. Through the investigation of the plasticity of development, we gain a better understanding of both the regulation of the developmental process and the embryo’s capacity for self-repair. Additionally, experiments with abiotic and biotic stressors during development have helped delineate not just critical windows for avian cardio-respiratory development, but the general characteristics (e.g., timing and dose-dependence) of critical windows in all developing vertebrates. Avian embryos are useful in exploring fetal programming, in which early developmental experiences have implications (usually negative) later in life. The ability to experimentally manipulate the avian embryo without the interference of maternal behavior or physiology makes it particularly useful in future studies of fetal programming. The bird embryo is also a key participant in studies of transgenerational epigenetics, whether by egg provisioning or effects on the germline that are transmitted to the F1 generation (or beyond). Finally, the avian embryo is heavily exploited in toxicology, in which both toxicological testing of potential consumer products as well as the consequences of exposure to anthropogenic pollutants are routinely carried out in the avian embryo. The avian embryo thus proves useful on numerous experimental fronts as an animal model that is concurrently both of adequate complexity and sufficient simplicity for probing vertebrate cardio-respiratory development.


Introduction
The avian embryo has a been a model for exploring animal biology for more than two millennia, encompassing ancient Egyptians, Aristotle, Leonardo Da Vinci, and William Harvey (Stern, 2004).Legions of modern-day biologists have used the bird embryo to provide insights into numerous facets of biology, including development, evolution, physiology, morphology, and taxonomy.In fact, analysis of the PubMed data base maintained by the US National Institutes of Health (www.pubmed.org)reveals that the number of papers on the embryo of chickens alone (not including other birds) is second only to mice among the grouping of mouse, chicken, rat, cow, and fish (Figure 1).Given that nearly 60,000 papers in PubMed contain the word pairing of "chick" and "embryo", it is obviously beyond the scope of this review to consider this enormous literature on avian development.For an entry into this literature, the reader is referred to both classic and modern reviews of avian embryonic morphology (Goodrich, 1930;Hamburger and Hamilton, 1951;Romanoff, 1960;Bellairs and Osmond, 2014), physiology (Romanoff, 1960;Burggren and Keller, 1997;Burggren and Crossley, 2002;Elfwing et al., 2011;Mueller et al., 2015a), and biochemistry and cellular/molecular biology (Romanoff, 1967;Stern, 2004;Gabrielli and Accili, 2010;Streit et al., 2013;Kain et al., 2014;Hirst and Marcelle, 2015;Schneider, 2015).Rather, the focus of this paper is to examine specifically how avian embryos have contributed to our growing understanding of cardio-respiratory development, not just in birds but also in vertebrates generally.The PubMed database lists more than 5000 papers that contain the search terms "chick" plus "embryo" plus "heart", and more than 1250 papers containing the search terms "chick" plus "embryo" plus "lung".These cited statistics do not even include the considerable research on the embryonic cardiorespiratory system of bird species other than chickens that are commonly used in experiments (e.g., zebra finch, duck, emu, etc.).Thus, our strategy for limiting this review to reasonable length is to first provide a simple overview of the basic cardio-respiratory ontogeny of the avian embryo.More correctly, we point the reader to some key papers that will provide an introduction to this topic.This brief overview will then serve as a back drop for more in-depth consideration of several relatively new areas of cardio-respiratory research on avian embryos that are providing -or have the potential to provide -major new insights into developing vertebrates.These topics are the ontogeny of organ systems interactions, developmental phenotypic plasticity, critical developmental windows, fetal programming, and epigenetic inheritance and toxicology.

Ontogeny of cardio-respiratory morphology, physiology, biochemistry, and celluar/molecular biology
As introduced above, the avian embryo has been extensively studied to provide insight into the ontogeny of the respiratory and cardiovascular systems and their regulation in vertebrates.What are the unusual features of the avian embryo that have made this such a popular animal model for cardio-respiratory studies?The answer to this question is the same for many areas of avian embryo investigation, and hinges on three key characteristics.First, the circulation of the avian embryonic circulation is readily accessible, even in the early formative periods, allowing both sampling of blood and injection of agonists/antagonist.Second, the embryo in the egg is not subjected to direct maternal influences (as is, for example, the mammalian embryo or fetus in the uterus).This allows the experimenter to modulate the embryonic environment during growth (e.g., inducing hypoxia at controlled levels).When these two attributes are combined with the relatively large size of the embryos of some avian species, it becomes apparent that avian embryos have allowed invasive experimental manipulations (Burggren et al., 2004;Kowalski et al., 2013;Shell et al., 2016).Such invasive procedures have primarily focused on the embryo of the chicken but also the embryos of other birds such as the emu (Dromaius novaehollandiae) and ostrich (Struthio camelus).Importantly, such experimental procedures simply are difficult if not impossible in the early developmental stages of mice or rats or other popular animal models.
Experimental approaches to study the avian embryonic cardio-respiratory system have involved the topical application of drugs (agonists or antagonists) directly onto the heart and central vessels of the very early embryo (Borkowf and Kolesari, 1986;Rosenbruch et al., 1993;Branum et al., 2013).As the embryo and its circulation grow (and in particular as the vessels of the chorioallantoic membrane become accessible), drug injection directly into the blood vessels of the chorioallantoic membrane can be through a "keyhole" in the shell, and in some cases through an indwelling catheter, e.g., Altimiras and Crossley (2000), Burggren and Crossley (2002), Crossley et al. (2003a), Crossley et al. (2003b), Crossley and Altimiras (2000), Lindgren et al. (2011), Mueller et al. (2013a), Mueller et al. (2014a), andTintu et al. (2009).
An additional approach to understanding embryonic hemodynamics and the ontogeny of their control involves the mechanical rather than pharmaceutical modification of cardiovascular function.For example, arterial outflow tract restriction has proven useful in understanding the hemodynamics of blood flow through the heart, and how they may influence cardiac development (Burggren and Keller, 1997;Broekhuizen et al., 1999;Lucitti et al., 2005;Lucitti et al., 2006;Stekelenburg-de Vos et al., 2007).
The discussion above has only scratched the surface in describing the "established" approaches applied towards "conventional" lines of inquiry on the ontogeny of the avian cardio-respiratory physiology and its control.In the limited remaining space in this review, let us now consider some of the more recent, less conventional uses of the avian embryo model in cardio-respiratory biology.

Ontogeny of interactions of system components and organ systems
Effective gas exchange in vertebrates -mature or immature -depends upon adequate ventilation of gas exchange organs and their perfusion with appropriate levels of blood pumped by the heart.Thus, a pivotal point in the organogenesis, growth, and maturation of the avian embryo (or, indeed, the mammalian fetus) is the regulatory control over the heart, vessels, and lungs geared towards the onset of air breathing.Such control systems involve motor neurons releasing neurotransmitters, endocrine organs releasing circulating hormones, and cellular receptors on effector organs that are stimulated by these neurotransmitters and hormones.At first glance, it might seem logical that all necessary elements of a physiological regulatory system would develop concurrently, such that the embryonic animal is largely or fully functional upon hatching/parturition. Do all the components of a regulatory system develop functionality concurrently or do they develop in a sequence?In fact, numerous studies have revealed a temporal mismatch of the onset of functionality of the various regulatory components.For example, the neural and hormonal control of the heart in perinatal birds involves a complex interplay between vagal tone, sympathetic tone, locally acting hormones (e.g., nitric oxide) and circulating hormones such as catecholamines and angiotensin II (Altimiras and Crossley, 2000;Crossley and Altimiras, 2000;Burggren and Crossley, 2002;Crossley et al., 2003a;Crossley et al., 2003b;Chiba et al., 2004;Mueller et al., 2013a;Iversen et al., 2014;Mueller et al., 2014a).Both muscarinic and adrenergic receptors are present in the first few days of normal embryonic development in the chicken, potentially enabling both sympathetic and parasympathetic neural control.Yet, early signs of sympathetic innervation appear several days after the presence of receptors capable of being stimulated by them.Full vagal innervation does not appear until a few days before hatching (Burggren and Crossley, 2002).Thus, most of normal embryonic development of the chicken is characterized by adrenergic cardiac control, with little or no involvement of vagal innervation (Crossley and Altimiras, 2000;Burggren and Crossley, 2002).Similar temporal displacement of the onset of numerous components of the renin-angiotensin system and its control of the cardiovascular system is evident from analysis of the literature of ANG II and its effects in chicken embryos (Brand et al., 2007;Crossley et al., 2010;Mueller et al., 2013a;Mueller et al., 2014a).
A schematic of how normal regulatory control system development can be viewed during normal development is shown in Figure 2. In Panel A, all elements of a regulatory system mature simultaneously, allowing full physiological regulation by a comprehensive control system to appear concurrently at the end of embryonic development.In Panel B, the development of motor neurons affecting a cardiovascular response lags behind formation of receptors on cells of the heart and vasculature, and the appearance of circulating hormones lags even further.The net effect is that the comprehensive regulatory system is not fully mature at the end of embryonic development.Panel C presents the reverse of Panel B, with the development of circulating hormones lagging the appearance of motor neurons.However, the effect in C is the same as in B -the lack of a fully developed comprehensive regulatory system.The concept of concurrent vs. sequential development of regulatory systems can be extended beyond regulatory controls of a single organ system, such as the cardiovascular system, to include multiple organ systems.Embryonic development is characterized not just by the progressive growth of organ systems in isolation, but also by the progressive maturation of the interactions between developing systems.For example, in most vertebrates, the kidney influences cardiovascular function through the renin-angiotensin system, but the influence is not just one way.Hormones secreted from heart tissue such as atrial naturietic peptide (ANP) also influence kidney function.Yet, very poorly understood is the ontogeny of the regulatory interactions of these two systems, as opposed to the ontogeny of each system by itself (Burggren et al., 2014).Making matters more complex, the heart begins to function much earlier in embryonic development than do the various stages (pro-, meso-, and metanephros) of the embryonic kidneys.Thus, the heart potentially develops control over the developing kidneys before the kidneys can exert control over the already well-developed cardiovascular system (Burggren et al., 2014).This hypothesis remains to be fully tested but, if true, indicates another example where there is failure of temporal symmorphosis in the development of the avian embryo.
Noteworthy is that the discussion above considers "normal" timing of developmental processes.Yet, the timing of onset or components of a developing regulatory system can also be somewhat plastic as a consequence of environmental stressors.This developmental plasticity in the timing of appearance of the components of a regulatory system (or indeed, of any developmental process) affected by an altered environment (broadly defined) will now be discussed in the general framework of developmental phenotypic plasticity.

Developmental phenotypic plasticity and critical windows of the cardiorespiratory system
The study of avian embryos is showing us how the embryonic environment modulates the development of these systems -that is, to what extent avian embryos are plastic in their growth and maturation.To fully appreciate this, we now consider the intertwined concepts of developmental phenotypic plasticity and critical windows.

Developmental phenotypic plasticity and critical windows
Phenotypic plasticity is usually interpreted as the interactions of an organism's Genotype with its Environment, so-called Gene × Environment or G × E interactions (Burggren and Reyna, 2011;Hutchings, 2011;Grishkevich and Yanai, 2013;Crews et al., 2014;Danchin and Pocheville, 2014;Keller, 2014;Wada and Sewall, 2014).By extension, developmental phenotypic plasticity -of morphology, physiology, or both -represents the modification of phenotype of a developing organism, which may or may not be fully carried in to the mature, adult form (Spicer and Burggren, 2003;West-Eberhard, 2003;Reeves and Gozal, 2005;Bjorklund et al., 2007;Spicer and Rundle, 2007;Burggren and Reyna, 2011;Domyan and Sun, 2011;Rymer and Pillay, 2012).Put differently, such plasticity results from interactions of an organism with its environment during the course of its development, or the less explored "G × E × D".
How environment affects the cardio-respiratory phenotype of the developing bird egg has been explored using environmental factors (stressors) such as temperature, oxygen, carbon dioxide, and various toxicants, as will be explored below (see section "Fetal programming of the avian cardio-respiratory system").Key interrelated questions that have emerged include When in development are the most susceptible or vulnerable periods for environmental stressors?and To what extent can the developing organ systems mitigate the impact of these stressors?Answering these questions requires considering the concept of critical windows in development.
The "critical window" in development (sometimes described as "susceptible period" in psychology and other fields) is that period of time during development when an organism's phenotype is most vulnerable for the effects of an environmental stressor or toxicant (Burggren, 1998;Nijland et al., 2008;Loepke, 2010;Burggren and Reyna, 2011;Burggren and Mueller, 2015).A tragic but illustrative example was the use of the drug thalidomide as an antinauseant to treat morning sickness in pregnant women.Unknown before it began being prescribed to pregnant women, thalidomide is also a powerful teratogen that disrupts human limb development beginning with week three and continuing through week five, then diminishing afterwards.Thus, the period of morning sickness, for which thalidomide was originally prescribed, unfortunately coincided with the critical window for limb development in the human fetus.The prescription of this drug for morning sickness consequently resulted in the birth of thousands of children in Europe and Canada with deformed or absent limbs before the correlation was discovered and thalidomide hurriedly pulled from the market (Franks et al., 2004;Knobloch and Rüther, 2008).This tragedy not only illustrates the concept of the critical window, but also the specificity of critical windows; that is, there are specific critical windows during development for specific organ systems.Thus, while limb development as mentioned above is in the human fetus most susceptible to teratogens in weeks 3-5, as evident from the thalidomide experience, the human fetal teeth, palate, and urogenital systems, for example are most sensitive from weeks 7-9.
A key question derives from the metaphor of the "window" -What defines the "edges" of the windows, are they sharp or fuzzy, and can they be moved by environmental factors?Typically, critical windows are described as distinct periods, that is, windows with sharp edges, e.g., "Day X through Day Y" of development (Figure 2A).Sometimes, in a slightly more nuanced approach, they are presented as "primary effects on Days X to X n with lesser secondary effects beyond Days X n+1 ".In either approach, however, there is a prevailing two-dimensional view of the critical window, with time and effect as the only two variables.Recently, we have advocated a more realistic "3D" approach that takes into account not only time and effect, but also dose of the environmental stressor (Burggren and Mueller, 2015;Mueller et al., 2016).In this approach, the injection of stressor dose adds a third dimension that can potentially alter our characterization of the critical window for a stressor (Figure 3).For example, while high doses of a stressor produce easily recognizable phenotypic changes, lower doses may produce modified phenotypes that are less likely to be noticed but are still biologically significant.Thus, the perception and reality of the extent of the critical window in development depends, in fact, on the doseresponse relationships between stressor and phenotypic characteristics.
The following question arises: "What is the role of bird embryos in studies of developmental phenotypic plasticity and critical windows cardio-respiratory systems?"Bird embryos have great promise as animal models in this area of experimentation for several reasons.First, as alluded to above, the bird egg physically is a closed, independent system, and as such the environment surrounding the embryo is much more tractable to experimental manipulation.Second, substances can be injected directly into the egg and the circulation of the embryo without mitigation by maternal influences, as would happen in placental mammals.Third, the distinct morphological stages of bird development have been highly characterized (Hamburger and Hamilton, 1951;Bellairs and Osmond, 2014), which allows subtle but significant phenotypic modifications (at least of morphology) to be readily detected.Fourth, the incubation period of some of the commonly used avian models (chicken, quail) is in the range of 20-25 days, which allows for the completion of critical windows experiments within a reasonable time frame (contrast this with the gestation period of ~114 days in swine and ~152 days in sheep, two commonly used large mammal models).Fifth, the cardio-respiratory morphology and physiology of the avian embryo is relatively well characterized (Xu and Mortola, 1989;Crossley and Altimiras, 2000;Burggren and Crossley, 2002;Crossley et al., 2003a;Crossley et al., 2003b;Maina, 2003, Maina, 2006;Ferner and Mortola, 2009; Figure 3 -Depictions of critical windows.A) Conventional linear view of critical windows, where a stressor results in a modified developmental trajectory.B) A more complex three-dimensional view of critical windows, where the degree and timing of the modification of the phenotype is dependent upon the stressor dose.After Burggren and Mueller (2015).
R. Bras.Zootec., 45(11):709-728, 2016 Mortola, 2009;Maina, 2012;Lewallen and Burggren, 2015;Mueller et al., 2015a).Finally, and in some ways among the most important of reasons, bird eggs are typically very inexpensive both to purchase and to maintain, compared with rodent models, for example.This ready availability provides for high throughput experiments with large n values.This feature is crucial, as large numbers of observations are often a requisite to achieve the level of granularity for defining critical windows (Burggren and Reyna, 2011).As we consider a more refined 3D approach of delineating critical windows in a dose-response context, access to a tractable, well understood model that has economic advantages (inexpensive and easily available) will likely accelerate the use of the avian embryo as a model for studying cardio-respiratory ontogeny.

Hypoxia, hypercapnia, and avian critical windows
Because cardiovascular and respiratory systems have specific critical periods, specific probes for these systems are often required to define and explore their critical windows.The respiratory gases oxygen and carbon dioxide have been primary tools for delineating critical windows for development of the cardiovascular and respiratory system of developing birds.
Hypoxia.Oxygen level is effective in modifying phenotype because the avian embryo directly depends upon diffusion across the bird egg shell for respiratory gas exchange -see Mortola (2009), Mueller et al. (2015a), and Rahn et al. (1987) for reviews of the extensive literature on this subject.Thus, changes in environmental oxygen levels, which can be readily induced at controlled levels in developing embryos by precisely regulating incubator gases, will be reflected in internal oxygen levels within the egg and embryo.Numerous studies have used environmental hypoxia (typically chronic exposure to 12-15% O 2 ) to reveal phenotypic developmental plasticity in both the cardiovascular and respiratory systems of the developing chicken embryo.The results reveal a complex and somewhat variable pattern, depending upon oxygen exposure used, period in incubation of exposure, and even the specific study using similar experimental protocols.Organs directly associated with the uptake and delivery of oxygen may be affected.Thus, the chorioallantoic membrane, the gas exchange organ of the embryo through most of the incubation period, shows proliferation in both size and vascularization as a result of incubation in chronic hypoxia (Adair et al., 1987;Stock and Metcalfe, 1987;Corona and Warburton, 2000;Chan and Burggren, 2005;Acosta and Hernandez, 2012; Druyan and Levi, 2012), although hypoxic incubation has also been reported to decrease the size of the chorioallantoic membrane (CAM) (Burton and Palmer, 1992).The increase in the mass and size of the CAM allow the embryo to survive the low oxygen conditions, but the increased growth in this structure means the relocation of nutrients from the embryo with direct consequences for growth (Nowak-Sliwinska et al., 2010).
Both pulmonary surface area and structure are also modified to lesser or greater extents by chronic hypoxia incubation (Xu and Mortola, 1989;Azzam and Mortola, 2007;Zhang and Burggren, 2012;Lewallen and Burggren, 2015).Hypoxia also has varying degrees of effect on the morphology and performance of the heart and systemic vasculature, ranging from no effect to profound disturbance (Asson-Batres et al., 1989;Richards et al., 1991;Burton and Palmer, 1992;Altimiras and Phu, 2000;le Noble et al., 2000;Dzialowski et al., 2002;Rouwet et al., 2002;Villamor et al., 2002;Ruijtenbeek et al., 2003;Villamor et al., 2004;Chan and Burggren, 2005;Zhang and Burggren, 2012).Corresponding with these cardiorespiratory phenotypic modifications are changes in overall metabolic rate, typically measured as oxygen consumption.Thus, for example, the embryos of the domestic chicken show a lower rate of basal oxygen consumption when incubated in 15% hypoxia during the middle period of incubation (days 6-12), but not during early (days 1-6) or late (days 12-18) incubation (Dzialowski et al., 2002).More recent studies have suggested that chronic hypoxic incubation throughout avian development leads to a lowered oxygen consumption without the creation of an oxygen debt (Mortola et al., 2012).
Given that the organs exchanging and supplying oxygen and nutrients are often affected by hypoxic incubation, it is not surprising that the structures beyond just these organs and organ systems (e.g., brain, eye, liver, stomach, beak, toes, and other structures) are also potentially affected.Yet again there are conflicts in the literature regarding the magnitude and even presence of the effects (Chan and Burggren, 2005;Azzam and Mortola, 2007;Zhang and Burggren, 2012).What might be the source of this variation in the morphological and physiological responses, and even the apparent timing of the critical windows, of avian embryos in response to hypoxic incubation?Certainly, there are specific physiologies and morphologies attached to specific strains of chickens, with often profound differences in magnitude and developmental timing of the response (Sato et al., 2006;Everaert et al., 2008b;Druyan, 2010;Ho et al., 2011;Crossley and Altimiras, 2012;Burggren et al., 2015b;Buzala et al., 2015;Huth and Archer, 2015).All too often, the specifics of the strain of chicken are not provided in studies, leaving open the fact that strain-specific differences are contributing to the apparent variability between published studies.Other examples of possible sources of variation in the physiological literature -e.g., prandial state, photoperiod, and parental history -have recently been reviewed (Burggren, 2014).
Hypercapnia.Hypercapnia has been widely used as a stressor to probe the onset of acid-base balance; discussion of this extensive literature is beyond the scope of this review, and the reader is referred to the following studies for an entry into the literature (Bruggeman et al., 2007;Everaert et al., 2008a;Everaert et al., 2011;Tazawa et al., 2012;Burggren et al., 2012;Mueller et al., 2013b;Andrewartha et al., 2014;Mueller et al., 2014b;Burggren et al., 2015b;Mueller et al., 2015a).Like hypoxia, hypercapnia has differential levels in mortality/hatchability at different points in development (Taylor and Kreutziger, 1966;Taylor and Kreutziger, 1969;Everaert et al., 2007).As just a few examples, hypercapnic tolerance is high during D9-D12 of incubation in chicken embryos (Taylor and Kreutziger, 1969).Elevated CO 2 during the last half of development affects time to hatching but not a number of other key parameters at either hatching or day seven after hatching, including body mass, relative growth, or hatching success, suggesting a "remarkable tolerance" to high CO 2 levels during the second half of development (Everaert et al., 2007).However, hypercapnic effects specifically on the cardio-respiratory system have not been specifically examined.
Extrapolation of data on hypoxic and hypercapnic exposure to an actual determination of cardio-respiratory critical windows is problematic.In part, this arises because low levels of CO 2 are beneficial to incubation, while there is no parallel advantage to mild hypoxia, and more severe hypoxia has long been understood to be detrimental (Grabowski and Paar, 1958;Mueller et al., 2015a;Riddle, 1924).Moreover, a wide range of different exposure protocols and O 2 and CO 2 exposure levels have been used.Not unexpectedly, then, there is variation in reported periods of respiratory gas susceptibility in development of the chicken embryo.In fact, even when experiments using similar protocols are repeated by the same experimental laboratory, both qualitative and quantitative differences may emerge.However, despite the absence of clear patterns of change in developmental plasticity, an overarching factor in the interpretation of these data on hypoxia and hypercapnia is that there is no reason a priori to expect that the lungs or heart share identical critical windows for development.Indeed, analysis of the literature on hypoxia effects and avian critical windows suggests an overall pattern in which the greatest vulnerability to phenotypic modification appears in the middle of avian development (Figure 3).While there is the above-mentioned considerable variation around this theme of hypoxia susceptibility, early in avian incubation the developing structures are not sufficiently developed to be heavily perturbed by hypoxia, while towards the end of incubation the embryo is developing the ability to physiologically mitigate hypoxia through adjustments in cardiac output, blood oxygen transport and, late in development, through the onset of pulmonary respiration as internal pipping starts.Thus, the middle period of incubation -a period of rapid cell growth -appears to be the most likely period in which phenotypic plasticity of the cardiorespiratory system arises.In (partial) contrast, induction of phenotypic plasticity (or death) by hypercapnia appears to be greatest in early development, with a growing tolerance to elevated CO 2 during the second half of the incubation period suggested by several longstanding studies (Riddle, 1924;Barott, 1937;Sadler et al., 1954;Taylor and Kreutziger, 1966).
Noteworthy is that the perspective of greatest hypoxia sensitivity in the mid-incubation period (or the reduced hypercapnic sensitivity) depicted in Figure 4 is heavily driven by observations of morphological perturbations to normal development.Our understanding of how cardiorespiratory physiology is altered by hypoxia, hypercapnia, or other stressors is far more fragmentary (see "Fetal programming" in the next section).

Studying heterokairy with avian embryo models
Avian embryos may also prove to be highly useful animal models for studying the malleability or plasticity of the order of appearance of developmental landmarks.Such malleability is incorporated in the concept of heterokairy, the change in developmental sequence of the onset of physiological regulatory mechanisms or supporting morphological structures within an individual or population (Spicer and Burggren, 2003;Spicer and Rundle, 2007;Spicer et al., 2011).This contrasts with heterochrony, which is a change in the development sequence between species over evolutionary time (Spicer and Burggren, 2003;Blom and Lilja, 2005;Spicer et al., 2011;Keyte and Smith, 2014;Mueller et al., 2015b).The relatively few experimental demonstrations of heterokairy have centered on either invertebrates or lower vertebrates.Thus, for example, exposure to predator cues alters the timing of onset of several key developmental events in embryos of the gastropods Radix balthica and Radix auricularia (Rundle et al., 2011).More recently, rearing oxygen level has been used as a stressor on larval air breathing fishes, where hypoxia accelerated the onset of air breathing in larval Betta splendens, but delayed the onset of air breathing in larval Trichopodus trichopterus (Mendez-Sanchez and Burggren, 2014).Additionally, development of rivulid annual killifish, as well as their diapause state, can reflect elements of heterokairy (Varela-Lasheras and Van Dooren, 2014).Collectively, these data on fishes show not only developmental phenotypic plasticity and heterokairy, specifically, but also show that a complex behavior that can be moved back and forth in the normal developmental timetable must be supported by equally complex morphology and physiology (Warkentin, 2011;Mendez-Sanchez and Burggren, 2014).
Despite their promise as experimental models, avian embryos have rarely been examined in the context of heterokairy.In one of the few such studies, hypoxia during incubation was used to investigate the ontogeny of pulmonary surfactant lipids in the lungs of chicken embryos (Blacker et al., 2004).Hypoxic exposure during key critical windows around day 16 of incubation accelerated the production of surfactants in developmental time in chicken embryos, likely through the mechanism of increased corticosterone production.Why the paucity of studies on the plasticity of developmental timing in individuals or populations of bird embryos?Perhaps this results from the perception that avian development -especially development of the chicken -is viewed as extraordinarily invariant and predictable (Bellairs and Osmond, 2014;Hamburger and Hamilton, 1951), even accounting for differences between species ascribed to heterochrony (Blom and Lilja, 2005).On the other hand, studies of heterokairy pivot around inherent variability in development.Yet, the very fact that avian development has been characterized in such detail is advantageous.Thus, subtle changes in developmental timing (i.e., heterochrony), which has been described for intraspecific comparisons of avian embryo development (Blom and Lilja, 2005), may actually be more likely to be teased apart when individuals or populations of avian embryos are perturbed by environmental modifications.

Fetal programming of the avian cardiorespiratory system
What is fetal programming?
Fetal programming is the shaping of the adult animal by events that occur early in its development.First described by Barker in more than 25 years ago (Barker et al., 1989;Barker, 1990), fetal programming has proven to be a robust conceptual framework for understanding developmental phenotypic plasticity, particularly in the context of disease -for an entry in to the voluminous literature, see Alexander et al. (2015), Chiossi et al. (2011), Fernandez-Twinn et al. (2015), Mueller et al. (2015b), Roberts et al. (2015), Sedaghat et al. (2015), Stangenberg et al. (2015), and Thornburg and Louey (2005).Essential, fetal programming describes the phenomena by which early fetal experiences in the uterus actual "program" physiologically and morphologically modified phenotypes subsequent to birth.Importantly, such phenotypes may change their expression during development to adulthood, and may actually not present themselves until later in adult life.A classic example of this is fetal programming caused by fetal intrauterine growth restriction (IUGR), which can be caused by maternal malnourishment, by placental blood flow insufficiencies, or other toxicological factors such as alcohol or tobacco exposure, e.g., Cohen et al. (2016).The resulting very low birth weight (VLBW) infants are undersized at first, but at varying rates show "catch up" growth due to a "thrifty phenotype" that is highly efficient at assimilating nutrients into their body tissues for accelerated tissue growth (Bergmann et al., 2008;Luo et al., 2010).However, the accompanying so-called "metabolic syndrome" results in adult onset problems including diabetes, obesity, and cardiovascular and renal problems, to name but a few (Gonzalez-Bulnes et al., 2016).

Fetal programming -Beyond mammals to birds
Although the term "fetus" implies mammals in popular usage, the concept of fetal programming should, in theory, extend beyond mammals within the vertebrates to the embryos of birds and the fetuses of live bearing reptiles, and the embryos and larvae of fish and amphibians.Yet, there is little evidence of this; not because studies have failed to show the consequences of early developmental experiences on adult phenotype, but rather because few studies outside mammals have been attempted.The following questions thus arise: Is fetal programming of cardio-respiratory biology uniquely mammalian, or is it more broadly a vertebrate phenomenon? and If fetal programming occurs specifically in birds, are there unique aspects to bird development that can aid our understanding of the fetal programming phenomenon in all animals?
Some long-term experiments that go beyond just embryonic plasticity evident through embryonic development (see above) have also been carried out in birds."Catch up growth" occurs in a variety of birds that experience some form of malnutrition that occurs after hatching, with achievement of normal body mass following a recovery period from the original perturbation (Palo et al., 1995;Criscuolo et al., 2008;Honarmand et al., 2010;Hector and Nakagawa, 2012;Burness et al., 2013;Chin et al., 2013).Catch up growth related to embryonic malnutrition (i.e., lowered provisioning of the egg manifested in smaller egg size) or other perturbation has also been demonstrated in late stage chicken embryos (Azzam and Mortola, 2007).Despite such catch-up growth, earlier experiments have shown that albumin removal from the egg produces smaller individuals at hatching (Hill, 1993;Finkler et al., 1998) and smaller eggs (with less yolk and albumin) within a species result in smaller hatchlings (Williams, 1994;Dzialowski and Sotherland, 2004).However, catch up growth evident in juveniles or adults specifically following abnormal embryonic experiences has not been observed in adult birds to our knowledge.Indeed, in one recent study, a lack of catch up growth was evident.Albumen removal from chicken eggs resulted in reduced body weight of the hens 55 weeks after hatching (Willems et al., 2015a), and egg laying performance of adult hens is also altered (Willems et al., 2013).Similar results occurred in a separate study by this research group, but feed intake was increased in juveniles undernourished as embryos, apparently without accompanying body mass increases up to normal levels (Willems et al., 2015b).Interestingly, albumen removal also results in adult hens with modified DNA methylation and long-term alterations in the hepatic transcriptome (Willems et al., 2016).The extraction of a percentage of the egg contents reduces embryonic growth and results in a low hatching weight, but apparently without any vascular consequences (Ruijtenbeek et al., 2003).
Chronic hypoxia in chicken embryos causes cardiac enlargement and beta adrenoceptor desensitization in juvenile birds (Lindgren and Altimiras, 2009).Incubation of chicken eggs at high altitude causes reduced arterial PO 2 and hemoglobin saturation, elevated hematocrit, lower systemic arterial blood pressure, and gender-dependent alterations in the gain of the arterial baroreflex (Herrera et al., 2013).Hypoxia incubation in bobwhite quail results in modifications to both the structure and function of the cardiovascular system (Flores Santin, 2016) (Figure 5).In ovo administration of glutamate inhibitors reduces intestinal size in young chickens (Li et al., 2014).Embryonic hypoxic incubation in chicken embryos leads to adult birds with dilated hearts (Tintu et al., 2009).Embryonic incubation also results in altered vascular structure and function in late chicken embryos, and the authors suggest that these effects may carry into the adult (Rouwet et al., 2002).
To conclude, the limited data on embryonic programming of adult modified phenotype in birds appears to strongly resemble the fetal programming of laboratory mammals as well as those human populations.This suggests that fetal programming is not exclusive to placental vertebrates and likely extends to other animal groups.The use of the avian model, and its highly tractable nature in experiments, introduces new and different insights into the comparative developmental physiology of vertebrates and more specifically to the phenomenon of fetal programming.

Avian epigenetic phenomena and mechanisms
Interest in epigenetic phenomena and mechanisms has burst onto the forefront of experimental biology.Increasing numbers of studies are examining intragenerational effects, including the epigenetics of disease and fetal programming (see above) as well as transgenerational epigenetic inheritance, either or both forms of epigenetic phenotype modification potentially induced by a wide variety of stressors.Interestingly, the vast majority of the focus of epigenetic studies has been on the underlying mechanisms (e.g., DNA methylation, histone modification, microRNA) for intragenerational epigenetic effects (Burggren, 2016).
A comprehensive review of this fascinating field of epigenetics is well beyond the scope of this article, so the reader is referred to reviews that emerge from medicine and physiology (Ho and Burggren, 2010;Zhang and Ho, 2011;Holland and Rakyan, 2013;Burggren, 2014;Burggren and Crews, 2014;Hala et al., 2014;Burggren, 2015), evolutionary biology (Kuzawa and Thayer, 2011;Zhang and Ho, 2011;Duncan et al., 2014;Varriale, 2014;Mendizabal et al., 2014;Burggren, 2016), ecology, and environment (Crews and McLachlan, 2006;Baccarelli and Bollati, 2009;Crews and Gore, 2012;Kim et al., 2012;Collotta et al., 2013;Kilvitis et al., 2014), etc. Curiously, although avian embryos have been used extensively in many types of developmental biology studies, they have been far less exploited in either intraor transgenerational epigenetic studies.Certainly, DNA methylation and histone modification patterns have been investigated as a function of development in avian embryos and hatchlings (Ellis et al., 2012;Li et al., 2015), and in some cases these molecular findings have been related to the respiratory transition from allantoic to pulmonary gas exchange (Gryzinska et al., 2014).Perhaps the biggest contributions to date from studies of avian embryos attempts to answer the question What is the role of "maternal provisioning" in epigenetic inheritance?While the terminology and semantics of "epigenetics" are still quite messy and continue to evolve (Burggren and Crews, 2014;Felsenfeld, 2014;Deans and Maggert, 2015), many embracing a broad definition of epigenetics regard maternal effects and the associated maternal provisioning as a form of transgenerational epigenetics.Bird eggs are, of course, pivotal in the investigation of maternally-derived yolk factors and their effect on the F 1 generation.A variety of hormones, nutrients, antibodies, and other substances that the mother puts into the egg yolk during oogenesis ultimately affect the development and eventual phenotype of the offspring (Hasselquist and Nilsson, 2009;Ho et al., 2011;Reed and Clark, 2011;Adkins-Regan et al., 2013;Ho, 2014).The bird egg and embryo should prove to be an ideal model for more detailed study of maternal effects through provisioning.Not only can the nature of the egg be changed by modifying parental diet (Karadas et al., 2005;Koutsos et al., 2007;Labaque et al., 2013), but also by direct injection into the yolk prior to or during incubation (Tissier et al., 2014;Herrington et al., 2015).
Studies on transgenerational epigenetics have focused far more on the inheritance of molecular phenotypes and far less on system-level physiology (Burggren, 2016).The bird embryo can also help us answer the question Can modified phenotypes of cardiovascular and respiratory processes be epigenetically inherited?Because of the relative precision with which stressors like hypoxia or osmotic stress can be delivered to the avian embryo, it is likely to be much easier to look for causation for subsequent phenotypic plasticity in the F 1 generation than in mammals, for example, where maternal influences can mitigate or at least modify the stressors as externally applied.

Avian embryos in toxicology and pollution research
It has long been appreciated that the development of the bird embryo, especially its cardiovascular and respiratory system, can be affected by anthropogenic toxicants as well as abiotic factors such as hypoxia and temperature (Mueller et al., 2015a).Do developing bird embryos respond to toxicants in unique ways that reflect their development within the bird egg, or do their responses more generally reflect the responses of other vertebrates?The answer of relevance to other vertebrates is indirectly answered by the existence of an extensive literature using the bird embryo for toxicology testing, e.g., Kishore et al. (2008), Kormos et al. (2009), Kue et al. (2015), Sharma et al. (2001), andVarnagy et al. (2001).However, beyond screening of cosmetics and putative drugs for human use, the bird embryo is increasingly being used to understand the effects of environmental pollutants on animals, including men, as we will now explore.

Inorganic toxicants as developmental disruptors
Anthropogenic environmental toxicants can be divided into inorganic substances (e.g., heavy metals, metalloids, non-metals, and radioactive isotopes.Some metals (e.g., arsenic, cadmium, cobalt, copper, iron, mercury, and zinc) may be extremely toxic at low levels, while in trace amounts some comprise essential nutrients required for biochemical and physiological functions.Some inorganic substances including metals not only have general effects in bird embryos such as reduced hatchability (Oliveira et al., 2015), but also specifically affect cardiovascular system through altered angiogenesis (Bertossi et al., 2004;Gheorghescu et al., 2015;Mroczek-Sosnowska et al., 2015), disruption to hematology (Newkirk et al., 2014), or in vitro changes in cardiomyoctes (Canga et al., 1993;Natarajan et al., 2006).Inorganic pesticides, organometals, and halogenated polycyclic aromatic hydrocarbons (HPAH) are some of the chemicals identified as endocrine disruptors (ED) in part because of their high lipophilicity, their ability to irreversibly bind to macromolecules (i.e., DNA), and their ability to reversibly react at specific sites of the receptors and enzymes (Yu et al., 2011).The exploration of ED in early development has extensively used the avian embryonic model -for an entry into the literature see Malir et al. (2013) and Roig et al. (2014).As one example, however, compounds that interfere with gonadal estrogen (responsible for differentiation of female phenotype in birds) such as the ERα agonists propyl-pyrazole-triol (PPT) and diarylproprionitrile (DPN) and the ERα antagonist methyl piperidino pyrazole (MPP) disrupted the normal development of the female reproductive tract in chicken embryos, resulting in embryonic retention of the right Müllerian duct (which normally regresses) and the malformation of both ducts (Mattsson et al., 2011).

Organic toxicants as developmental disruptors
Organic toxicants, including complex pesticides, organometallic compounds, aliphatic compounds (alkanes and alkenes), and aromatic compounds (e.g., polycyclic aromatic hydrocarbons, PAH) represent a significant portion of contaminant mixtures found in ground-and surface waters, coastal areas, soil, and the global atmosphere.One of the main sources of aliphatic and aromatic compounds are petroleum hydrocarbons (alkanes, alkenes, and aromatic hydrocarbons).Such compounds (e.g., DDT, PCBs, TCDD, PAH) resist degradative processes and thus persistence in the environment for extremely long periods of time.Sources of PAH include the incomplete combustion of organic material due to both anthropogenic and natural sources -high temperature combustion of fuels, emissions from crude and refined oil, forest and prairie fires, volcanic eruptions, and direct biosynthesis by microbes and plants (Näf et al., 1992).Some of these PAH are readily metabolized by organisms.For example, benzoperylene (a PAH) has a half-life of 14 months on soil, but it is quickly absorbed and metabolized in animal tissues, including those of the embryos of chickens and eider ducks (Albers, 1978;Näf et al., 1992;Albers, 2006;Fournier et al., 2010).
Polycyclic aromatic hydrocarbons are both mutagenic and carcinogenic -see Yu et al. (2011).They enter tissues through inhalation, ingestion, and dermal contact, and their rate of absorption increases when they are present in mixtures.The liver is generally the most important site where toxicants undergo biotransformation (e.g., Westman et al. (2013)), but generally most if not all organs are affected to some extent by PAH, with their toxicity dependent upon dose, concentration, time, route of exposure, further distribution, sensitivity, and ability of the organ to metabolize these compounds (Canga et al., 1993).
Bird embryos, once again, have been heavily exploited to probe the effects on the cardiovascular system of PAH, either individually or generally as a component of petroleum products.For example, vascularization decreases in the 72 h chicken embryo in responses to diesel exhaust particles (DEP), and between 30-80% of exposed embryos were poorly developed (Simsek et al., 2012).
Accelerated attention is now being devoted to the effects of PAH on animals, driven in part by the devastating Deep Horizon oil spill in the Gulf of Mexico (Burggren et al., 2015a).In this context, the bird embryo is proving to be useful not only as a tractable animal model for studying PAH effects, but more specifically to see what the effects might be on nesting birds and their eggs exposed either to oil transferred to the eggs or as a result of direct exposure to PAH in oil vapors in the air.Particularly useful in this regard is the ability to grow bird embryos in either the opened egg (Romanoff, 1931) or in in vitro culture (Dunn, 1991;Ho et al., 2011;Jourdeuil et al., 2015).Such unparalleled access to the early developmental stages has enabled study of organogenesis, angiogenesis, hematology, and the physiological function of the cardiovascular system in bird embryos just a few days old (Burggren et al., 2004;Datar and Bhonde, 2005;Khorrami et al., 2008;Branum et al., 2013).Figure 6 illustrates the process of growing embryos in shell-less culture as well as physiological information that can be derived from them.Experiments are currently underway to determine the specific cardiotoxic effects of PAH released from the Deep Water Horizon oil spill on early cardiac function in birds.In this respect, bird embryos will be illustrative in terms of the critical windows of sensitivity of the cardiovascular system to PAH in developing vertebrates.However, not only do nesting birds inadvertently foul their eggs with oil they have come in contact with, but the incubating embryos are exposed to PAH vapors emanating from the landed oil on the shoreline (Burggren et al., 2015a).

Temperature as a developmental disruptor
Temperature is sometimes excluded in lists of pollutants per se, but this abiotic factor results from local anthropogenic effects (e.g., power plants, urban environments) but also as a chronic pollutant with broad reach; e.g., climate change in the form of global warming extending over decades.The potential future effects of global warming on developing birds is beyond the scope of this article, but here we introduce some of the known effects of changes in incubation temperature on bird egg development.
Bird embryos chronically incubated under hypothermia (35.5 ºC) not unexpectedly show significant retardation of embryonic development (Tazawa, 1973;Boehm et al., 1987;Lourens et al., 2007;Reyna and Burggren, 2012).Additionally, acute hypothermia increases the P O2 and pH and decreases the P CO2 of both arterialized blood and mixed venous blood and creates cardiac hypotrophy in chicken embryos (Boehm et al., 1987;Tazawa, 1973;Tazawa and Mochizuki, 1978).On the other hand, hyperthermia during incubation decreases hatching success in the chicken and the bobwhite quail (Colinus virginianus) (Stoleson and Beissinger, 1999;Reyna, 2010;Reyna and Burggren, 2012) and alters hemodynamics beyond the expected rate function increases with elevated temperature (Lee and Lee, 2010).Prenatal incubation temperature affects many aspects of post-natal performance.For example, the early postnatal development of neuronal hypothalamic thermal sensitivity in birds, with birds incubated during the last week at 38.5 o C exhibiting elevated neuronal hypothalamic cold sensitivity at the 10th day post-hatching, likely due to an increased proportion of cold-sensitive neurons and a reduced proportion of warm-sensitive neurons (Tzschentke and Basta, 2002).Periodic cooling studies had shown a decrease in embryo mass and yolk reserves in the zebra finch (Taeniopygia guttata), and it also delayed development and increased metabolic costs (Olson et al., 2006).Interestingly, temperature disruption also had disproportionately larger effects on soft tissues compared to long bones and skull in zebra finch (Olson et al., 2008).
Even pre-incubation storage temperatures at the early blastocyst stage of bird embryos can affect viability (Reyna and Burggren, 2012;Branum et al., 2016).While many species are continuously incubated by a parent (e.g., chicken), other species may leave the eggs exposed to the environment either upon first laying or later during incubation while parents are off foraging (e.g., ratites, quail, etc.).Thus, temperature changes (especially habitat warming) may have detrimental effects on embryos in these birds.
The ease with which the temperature of the bird embryo is manipulated has also led to their use as models in many different lines of experimental inquiry.For example, bird embryos have been instrumental in our understanding of the evolution of endothermy (Walter and Seebacher, 2009;Tzschentke and Rumpf, 2011;Sirsat et al., 2016).Hypothermic bird embryos have also led to an understanding of how cardiac function is altered during hypothermic surgery in humans (Sarre et al., 2006;Shao et al., 2007).Finally, temperature manipulation has provided insights into strain-based physiological differences in the poultry industry (Collin et al., 2007).

Hypoxia as a developmental disruptor
Finally, hypoxia as a natural environmental stressor for birds and especially their embryos can result from natural conditions such as inundation of the nest by water, poorly ventilated burrows, and high altitude habitats (Mueller et al., 2015a).The potentially disruptive effects of hypoxia on cardiovascular and respiratory development have already been discussed in earlier sections of this essay and will not be elaborated upon here.Noteworthy, however, is that for many animals, the developmentally disruptive effects of hypoxia are multiplied by the simultaneous presence of other factors such as environmental temperature and toxicants.Indeed, the suite of hypoxia, elevated temperature, and toxicants are often present concurrently in polluted environments; consider, for example, the so-called "dead zone" of the Gulf of Mexico (Stow et al., 2005).

Conclusions
Developmental biology, comparative physiology, evolutionary biology, and numerous other areas of contemporary biological research continue to benefit from lessons learned using the venerable bird egg and its embryo within.Indeed, as we identify similarities between all vertebrates from molecular to organismal levels, the use of the bird embryo continues to expand, including animal modeling for areas as diverse as nonclinical safety studies of pharmaceuticals and vaccines (Saw et al., 2008;Bjornstad et al., 2015) to toxicology (Kopf and Walker, 2009;Smith et al., 2012) to prototyping surgical and measurement apparatus (Filas et al., 2011;Ford and Mertz, 2013;Lee and Ha, 2013).Moreover, the structure of the bird embryo within its egg has allowed focus on specific organs, such as the chorioallantoic membrane (Branum et al., 2013;Nowak-Sliwinska et al., 2014;Yuan et al., 2014), or on large specific and circumscribed energy and nutrient pools, such as albumin and yolk (Williams, 1994;Dzialowski et al., 2009;Nangsuay et al., 2015).Finally, a compelling aspect of the bird embryo is its size, particularly when considering the embryos of the ratites (ostriches, emus, rheas), which allows surgical and other interventions not possible in many other common embryonic models (Burggren et al., 2000;Burggren et al., 2004;Dzialowski and Greyner, 2008;Khorrami et al., 2008;Shell et al., 2016), including in vitro cultivation (Ho et al., 2011;Brand et al., 2014).
Like the ancient Egyptians, Aristotle, and William Harvey, we continue to peer into the bird egg, transfixed by what we see within.In doing so, we gain insight into developmental processes, structures, and outcomes in all vertebrates.

Figure 1 -
Figure 1 -Representation of papers in the PubMed data base that reference embryos of five major examples of vertebrate animals, determined May 2016.

Figure 2 -
Figure 2 -A schematic illustration of different scenarios of maturation of a physiological regulatory system and its constituent components during normal embryonic development.A) All components appear and mature at a similar rate, and full regulation occurs at the end of embryonic development.B) and C) the normal onset of one of the regulatory components is delayed relative to other components, resulting in a delayed onset and maturation of the physiological regulatory system.

Figure 4 -
Figure 4 -Relationship between susceptibility to hypoxia and hypercapnia induced phenotypic modification, morphological and physiological maturity, and selfrepair capabilities during the avian incubation period.Note that the timing of susceptibility varies for hypercanpia and hypoxia.

Figure 5 -
Figure 5 -Left ventricular anterior wall thickness in 12-month-old adult quail (Colinus virginianus) chronically incubated in either A) normoxic or B) hypoxic conditions.Note the hypertrophy of the adults exposed to hypoxia in ovo.Note the thicker walls of the hypoxia-exposed birds (Flores Santin, 2016).

Figure 6 -
Figure 6 -Shell-less in vitro culture of chicken embryos.A) An individual embryo on yolk surrounded by albumin in a culture dish (left), mass production of embryos for experimentation (center), and using recording electrodes to determine physiological parameters such as heart rate (right).B) The cardiac impedance signal produced by heart contraction and relaxation and the derived instantaneous heart rate in a 72-h chicken embryo at 37.5 °C in shell-less culture.