|Home | About | Journals | Submit | Contact Us | Français|
The chronically instrumented pregnant sheep has been used as a model of human fetal development and responses to pathophysiologic stimuli such as endotoxins, bacteria, umbilical cord occlusions, hypoxia and various pharmacological treatments. The life-saving clinical practices of glucocorticoid treatment in fetuses at risk of premature birth and the therapeutic hypothermia have been developed in this model. This is due to the unique amenability of the non-anesthetized fetal sheep to the surgical placement and maintenance of catheters and electrodes, allowing repetitive blood sampling, substance injection, recording of bioelectrical activity, application of electric stimulation and in vivo organ imaging. Here we describe the surgical instrumentation procedure required to achieve a stable chronically instrumented non-anesthetized fetal sheep model including characterization of the post-operative recovery from blood gas, metabolic and inflammation standpoints.
A variety of animal models exist for the study of both normal and compromised pregnancies, including laboratory rodents, non-human primates and domestic ruminants.1,2,3,4,5 The chronically instrumented pregnant sheep has been used extensively for 50 years as a model of human fetal development and responses to pathophysiologic stimuli such as lipopolysaccharide (LPS).6–10 The lesions following LPS exposure mimic exactly what is seen in preterm infants with periventricular leukomalacia, which is due to a similar maturational profile of both species.11, 12
Other pregnancy complications have also been studied in great detail such as the discovery that antenatal glucocorticoids promote lung development 13–15 and understanding the impact of intrauterine growth restriction (IUGR) on the fetus 16,17.
The extensive use of the fetal sheep model is due to the unique amenability of the non-anesthetized fetal sheep to the surgical placement and maintenance of catheters and electrodes, allowing repetitive blood sampling, recording of bioelectrical activity, application of electric stimulation and in vivo brain imaging.18 Telemetry is also possible, although less frequently used yet due to the higher sophistication to set up as well as the initial and maintenance cost.19
Moreover, the fetal sheep model is very versatile as many variations of instrumentation are possible depending on the measures of interest. For example, it is possible to record over days to weeks multivariate signals in real time such as fetal breathing movements, electrical brain activity, cardiovascular responses, electrocardiogram, regional blood flow to a range of organs using flow probes or microspheres, etc. Thanks to this versatility, a wide range of studies have been conducted including the development of the cardiovascular system20,21, hypothalamo-pituitary-adrenal (HPA) axis22, brain development23 and sleep states development in particular24, effects of hypoxia/asphyxia25, therapeutic hypothermia 26, inflammation6–11, combination of both27, glucocorticoids28,29, anti-depressants30, broncho-pulmonary dysplasia (BPD) 31,32, fetal programming33,34,35,36,37,38,39 or development of novel fetal monitoring modalities prior and during labor to name but a few areas of investigation.40,41,42,43
The overall goal of the method presented is to show this versatile model’s basic implementation. It allows establishing a wide variety of acute and chronic experimental protocols studying fetal physiology and pathophysiology on the integrative, organ, cellular and molecular levels.
Animal care followed the guidelines of the Canadian Council on Animal Care and the approval by the Université de Montréal Council on Animal Care (protocol #10-Rech-1560). Detailed information on materials and methods used is provided in the Table 1.
38 pregnant time-dated ewes were instrumented at 128±2 days of gestation (dGA, ~0.88 gestation, term 145 dGA) with arterial, venous and amniotic catheters and electrocardiogram (ECG) electrodes with sterile technique under general anesthesia (both ewe and fetus). In case of twin pregnancy the larger fetus was chosen based on palpating and estimating the intertemporal diameter; alternatively, the fetus to be instrumented may be selected randomly to avoid any potential bias or both fetuses may be instrumented. The total duration of the procedure was 124±27 min. The portion of the fetal upper body to be instrumented remained outside the uterus for 92±19 min. Most of the ewes instrumented were F2 animals. Their dam was F1 (Border Leicester*Romanov) and the sire was a Hampshire ram; they were crossed as follows: Hampshire (50%)-Border Leicester (25%)-Romanov (25%) = HABLRV.
Representative maternal and fetal physiological characteristics during surgery and instrumentation are shown in Figure 1 and were within the physiological norm for the gestational age of the fetus and maternal behavior during anesthesia.
Maternal weights averaged 75±11 kg and 21 out of 38 carried twins (i.e., at a rate of 1.6±0.5). At the time of necropsy (134±3 dGA), among the instrumented fetuses singletons weighed 4,090±800 g while twins’ weight was lower at 3,300±740 g (p=0.003). The weight of uninstrumented twins at 3,300±670 g was similar to the weight of the instrumented twins (p=0.78). 18 of the instrumented fetuses were female and 20 were male.
The dynamics of fetal arterial blood gases, glucose and acid-base status during surgery and post-operative recovery are reported in Table 1. We observed a gradual recovery of fetal acid-base status and a moderate deterioration of oxygenation with little change in glucose and electrolytes from post-operative days 1 to 3. Notably, on post-operative day 3, 42% of fetuses were found to be spontaneously hypoxic with arterial pO2 11 mmHg of and O2Sat of 28%. The normoxic cluster’s fetal pO2 was centered at 22 mmHg and O2Sat at 56%. Twin fetuses were not more hypoxic than singleton fetuses (p=0.26).
Fetal arterial IL-6 ELISA rendered values below the sensitivity threshold of 16 pg/ml throughout the post-operative recovery period. Similarly, the TNF-α levels also remained unchanged and very low at 29 pg/ml with ~30% of the animals also showing values below the sensitivity threshold of 13.9 pg/ml throughout the post-operative recovery period.
The anesthetic and surgical procedures are presented that are required for establishing an animal model for studying fetal physiology and pathophysiology: the chronically instrumented non-anesthetized fetal sheep.
Four critical steps within the protocol should be emphasized. First, passing the catheters and electrodes through the maternal flank: it is important that this is done at once to avoid any internal organ injuries. Second, securing the uterotomy operating site prior to exteriorizing the fetus: this is crucial to prevent or minimize loss of amniotic fluid and subsequent suturing of the amniotic membrane prior to uterine closure. Third, arterial catheterization: fetal sheep arteries are about 1–2 mm in diameter and hence not difficult to catheterize for an experienced surgeon; a team of two surgeons performs best and quickest in this task which helps save time to minimize the overall length of the procedure. Fourth, careful securing and organization of all catheters and electrodes in the amniotic cavity prior to returning the fetus into the amnion and closing the uterus: this helps to avoid accidental pulling of the catheters or tearing of the ECG electrodes due to maternal or fetal movements after surgery.
Instead of the here presented approach to catheterize brachial vessels, carotid or femoral vessels can also be used. The choice depends on the overall instrumentation approach that in turn will be dictated by the study design. We recommend to minimize the time the fetus spends outside the uterus and to minimize the extent to which the fetus needs to remain outside the uterus to be instrumented. These considerations led to the choice of the vessels instrumented in the presented “minimal approach”. We recommend catheterizing arteries on both sides to allow for intra-arterial blood pressure monitoring and arterial blood sampling with no mutual interference throughout the experiment. An added advantage is the fail-safe redundancy this approach introduces: in case one artery does get blocked during the experimental period, sampling and pressure monitoring are possible from the same vessel with the drawback of interrupting the monitoring when blood sampling is done.
There are three limitations, which preclude a wider adaptation of this animal model. These limitations can be addressed by some modifications suggested below. First, it is the requirement for biosafety level 2 confinements in some jurisdictions. This is due to a risk for Coxiella burnetii infection from pregnant sheep in humans with a weakened immune system. 49,50 Vaccination is available for the animals and exposed humans to decrease this risk 51,52 and PCR tests can be done to ensure that no positive animals are delivered to the research facility from the farm. A solution can be to combine animal vaccination with multiple PCR tests on the farm from the vaginal swabs performed prior to breeding begin, prior to delivery and then again from the amniotic fluid during the surgery. With such precautions, in some jurisdictions the use of sheep in research is not limited as it is in others. Second, the cost per animal is in the lower four-digit range, comparable to some murine knockout strains. With this in mind however, the information gain from each fetal sheep experiment compares only to non-human primates as it pertains to the extensive amount of data that can be collected, the questions that can be asked and the potential of translation to the human due to the timing of the development of organs in sheep. Third, there is the issue of breeding and animal availability during certain times in the year only. Even with hormonal treatment, results of sheep reproduction such as pregnancy rate and vitality of lambs (fetuses) are satisfactory only a few months around natural breeding seasons.53 Hence, experimental scheduling requires careful planning with the years split into a fall and spring seasons. A solution can be to establish September-to-November and an April-to-June experimental ‘sheep seasons’. This issue is also a virtue, as it allows for time to analyze the many data collected during each experimental season.
There is a number of factors contributing to the significance with respect to existing methods. The morphometric, cardiovascular and blood gas data presented were within the range for the species 54,55 and resemble those of human species 56, a major advantage of this animal model. One exception is the higher rate of multiple pregnancies compared to human twinning. 57 This however is also a virtue of the model, as studying the effects of twinning on fetal development is an important biomedical task.55,58,59 Very low levels of post-operative inflammation as measured by IL-6 and TNF-α ELISAs combined with recovery of acid-base status indicate that fetal surgical instrumentation is well tolerated and the post-operative recovery period of 72 hr is adequate to ensure a stable baseline condition of the fetal sheep prior to commencing an experiment. High percentage of spontaneous moderate chronic hypoxia in near-term fetal sheep renders them an interesting model for studying the chronic effects of the human antenatal hypoxia and inflammation on fetal and perinatal development, such as e.g., IUGR, and perinatal insults, such as inflammation and acute asphyxia. 60,61,62 Several IUGR sheep models are used, some relying on spontaneous hypoxia, some inducing it by placental embolization, for example. 16,63–66 On the other hand, severe hypoxia prior to start of an experiment may also be an exclusion criterion in cases where e.g., cardiovascular or central nervous systems are to be studied, as here the responses of chronically hypoxic fetuses are known to differ from those who are normoxic.60 Another important application is the study of prenatal maternal stress impact on fetal and postnatal development. 5,67 Finally, as can be seen in the numerous cited publications with this model, the fetal instrumentation can be made throughout a wide range of gestational ages anywhere from ~70 to 135 dGA corresponding to mid-gestation–near-term studies of fetal development. With the advancing gestational age, instrumentation of ever increasing complexity are possible, but considerations of the duration of surgical instrumentation need to be weighed against the need to obtain a number of multivariate recordings from the same fetus.
A number of very promising future applications of the technique presented is derived from the ever-growing number of sheep-specific molecular biology reagents and recent sheep genome sequencing. These recent developments have further promoted this animal model to be a very promising and powerful approach to understanding healthy and pathological human fetal development on various scales of organization, from (epi)genome to integrative physiology. 68,69–74
authors gratefully acknowledge funding support from the Molly Towell Perinatal Research Foundation, Canadian Institutes of Health Research (CIHR), and Fonds de Recherche du Québec–Santé (FRQS) (to MGF) and CIHR-Quebec Training Network in Perinatal Research (QTNPR) (to LDD).
The authors wish to thank Esther Simard, Marco Bosa, Carl Bernard and Carmen Movila for technical assistance.
The video component of this article can be found at http://www.jove.com/video/52581/
No disclosures have been made.