Objective
The purpose of this study was to biomechanically characterize and compare human, porcine, equine, and ovine fetal membranes.
Study Design
Noncontact metrology was used for topographic analyses. Uniaxial tensile testing was performed to resolve specific biomechanical values. Puncture force and radial stresses were determined with biaxial puncture testing. Microstructure and surface tortuosity were analyzed histologically.
Results
Equine and human membranes sustained larger magnitude loading, but ovine and porcine membranes exhibited stronger material properties. Biaxial puncture validated uniaxial results; human and equine groups accommodated the largest loads but lowest stresses. Equine membranes were mostly vascularized; tortuosity was highest in porcine membranes. Species’ gestation length was correlated positively with membrane thickness.
Conclusion
The anatomy of placentation and length of species gestation show distinct relationships to membrane biomechanics. Unlike other species, human fetal membranes do not compensate for structural weakness with a thicker membrane. This finding may explain the high incidence of preterm premature rupture of membranes in humans.
It is understood that the most primitive vertebrate life forms were marine in nature and that, as terrestrial life emerged, additional evolutionary innovations such as viviparity became necessary to increase species survivability. Many variations have risen among different taxonomic groups that use viviparity to allow for successful reproduction (ie, superfetation and superfecundation).
In the viviparous mammal, the fetal membranes (FMs) play a crucial role in enabling the physiologic progression of pregnancy. From a biomechanical viewpoint, they are ultimately responsible for the containment and sustenance of pregnancy to term gestation. This fundamental function of the FM is related directly to its mechanical and structural properties. The evolution of upright posture in humans has presented a unique challenge to the FM that is not present in other species. This biomechanical dilemma of containment has resulted in a significant incidence of complications during pregnancy, especially those that pertain to our ability to carry offspring to term gestation.
Human FMs have been characterized mechanically and structurally because of the high incidence of preterm premature rupture of membranes (PPROM) during human gestation and the increasing interest in amniotic membranes as a biomaterial in the field of tissue engineering. To our knowledge, no direct comparison of the mechanical and structural properties of FMs has been conducted across varying mammalian species. By providing insight to the cross-species variable properties of FMs, a better understanding of the role, function, and physiology of the membranes as they relate to our species and evolution can be attained.
Such comparative biomechanical studies can be quite beneficial to the understanding of membrane-related complications that are common in human gestation, particularly PPROM. PPROM complicates 3-4.5% of all pregnancies in the United States and is the leading cause of preterm birth in women; 30-40% of all preterm births occur as result of PPROM. PPROM, which likely embodies a multifactoral cause, occurs through multiple pathologic pathways throughout the reproductive structures that result in biochemical fluctuations that are capable of changing the structural stability within the FM. This possibility combined with the effects of biomechanical events (such as repeated membrane stretching, which can strain-harden the FMs and make them less elastic) can make the FMs vulnerable to rupture.
Many studies regarding human reproduction have been carried out using animal models. Although animal models remain relatively valid and accurate when other organ systems (ie, cardiovascular, respiratory, musculoskeletal) are being studied, the immense variation in reproductive function across species can result in decreased validity of animal models that are used for studying reproduction. It has been suggested that more variation exists in the placental organ than any other mammalian organ. As such, a multispecies comparison of FMs as it relates to structure and mechanical function can be advantageous in the justification of animal models that frequently are used in reproductive studies, especially those studies that focus on membrane-related diseases and complications.
In this study, we used well-established biomechanical techniques to analyze the mechanical properties of FMs from 4 different species: Homo sapiens, Sus scrofa domesticus (domestic pig), Equus ferus caballus (horse), and Ovis aries (domestic sheep). We supplemented our mechanical results with histologic analyses of the membranes to analyze the microstructure of the FMs. Our goal was to identify unique membrane properties that are pertinent to each species’ form of gestation and properties that are conserved across species.
Materials and Methods
Tissue retrieval
The reported study was approved by the Mississippi State University Institutional Review Board (#08-275) and Institutional Animal Care and Use Committee. Term human placentas were collected with informed consent from 9 vaginal deliveries of singleton pregnancies that showed no evidence of infection. Thirteen porcine placentas were obtained from term gestation sows after vaginal delivery at a commercial breeding facility and from slaughtered sows at a local abattoir. A total of 8 equine placentas were obtained from mares (American Quarter Horse) after vaginal delivery at the Morgan Freeman Equine Reproduction Research Unit at the Mississippi State University College of Veterinary Medicine. Vaginally delivered ovine placentas were obtained from 9 Kataden-Dorpor ewes from a local farm. The Table summarizes specific maternal and fetal metrics of patients and animals within the study.
| Species | n | Avg. fetal weight, kg | Avg. placental weight, g | Avg. placental weight to fetal weight ratio | Avg. gestational age, wk | Gravidity | Parity | Body mass index (human only) |
|---|---|---|---|---|---|---|---|---|
| Human | 9 | 3.547 (±0.082) | 562.67 (±61.5) | 0.143 (±0.014) | 39.1 (±0.2) | 2.65 (±0.44) | 1.35 (±0.33) | 33.1 (±1.8) |
| Porcine | 13 | 1.275 (±0.110) | — | — | 16.5 a | — | — | — |
| Equine | 8 | 38.2 (±4.4) | 5080 (±320) | 0.142 (±0.019) | 47.8 a | 1.5 (±0.29) | — | — |
| Ovine | 9 | 4.385 (±0.772) | — | — | 21.7 a | 1.67 (±0.33) | — | — |
a The average gestational age data for porcine, equine, and ovine groups were obtained from Senger.
After delivery each placenta was rinsed in Ringer’s lactate solution to remove blood and debris. All placentas were arranged with the maternal-side facing out to best expose each species’ placental anatomy ( Figure 1 ). In an effort to increase homogeneity and avoid zones of extreme altered morphology within the sample population of the human group, the FMs were excised by first the removal of a portion of tissue (approximately 3-cm wide) that surrounded the site of rupture and a portion of tissue (approximately 4-cm wide) around the placental disk. The study groups consisted of human chorioamnion (CA), human amnion (AM), equine allantoamnion (AA), porcine AA, and ovine AA.
Thickness and topographic analysis
Sections of FM from each species were further trimmed to 2 × 2 cm and placed in Ringer’s lactate. Four to 8 sections were used per donor. Samples were placed onto a smooth and level stainless steel platform and analyzed with a noncontact surface profiler (Talysurf CLI 2000 Gauge System; Taylor-Hobson Ltd, Leicester, UK; Figure 2 , A). To calculate thickness, linear profiles over approximately 1 cm of each sample were obtained at 100 μm/sec ( Figure 2 , B and C). For topographic analysis, high-resolution scans were made with the scan profile function of the TalySurf. The accurate measurement of membrane thickness avoids errors that likely are caused by other thickness-measuring methods (eg, Mitutoyo thickness gauge might underestimate the membrane thickness because of the applied contact load; digital caliper might bring in human errors). As we know, a small error in thickness measurement might cause a large error in stress and tensile modulus calculation.
Mechanical testing
The Mach-1 Micromechanical System (Biomomentum Inc, Laval, Québec, Canada) was used for analysis of membrane biomechanics. All testing was performed with samples submerged in buffered saline solution at 37°C with a circulating bio-bath system.
Uniaxial tensile testing
Membrane specimens were dissected into dog bone–shaped samples ( Figure 3 , A) and mounted onto the Mach-1 with a clamp-to-clamp distance of 40 mm. As with mechanical testing of most soft tissues, an initial step of cyclic loading and unloading was used to establish a preconditioned tissue state. This step aids in overcoming the effects of tissue handling and ex vivo analyses.
Preliminary experiments were conducted to determine the appropriate loads that were required for the preconditioning protocols for each species. Samples from each species were prepared and pulled to failure at 50 μm/sec. The value for load near the end of the toe-in region was used for preconditioning samples of each species before uniaxial testing. The preconditioning load values that were used were 15 g, 10 g, 20 g, and 10 g for human, porcine, equine, and ovine groups, respectively.
Each sample was preloaded to 1 g and preconditioned for 10 cycles of load-unload at a rate of 500 μm/sec with the Mach I in load-control mode. Immediately after this step, each sample was pulled to failure at 50 μm/sec. Data from uniaxial testing were used to resolve ultimate tensile load, ultimate tensile stress, peak tensile modulus (tissue stiffness), extensibility (percent tissue stretch before significant mechanical response), ductility (strain at failure), and toughness (strain energy that is required to fail tissue) ( Figure 3 , B). Ultimate tensile load and ultimate tensile stress were determined as the load in grams at failure and stress in kilopascals at failure, respectively. All stress values were calculated from the undeformed cross-sectional area.
Biaxial puncture testing
Biaxial failure strength of tissues was determined by the biaxial puncture test with the Mach-1. Samples were mounted between 2 cylindrical discs with a circular opening of 20-mm diameter and sandpaper on the gripping surfaces ( Figure 3 , C). A 3.2-mm spherical probe that was mounted on the Mach-1 head was used to puncture samples at a speed of 50 μm/sec. Thickness data and peak tensile modulus values from uniaxial testing were incorporated with the raw biaxial puncture data to determine biaxial failure strength (maximum radial stress at failure), as previously described by Oyen et al.
Histological analyses
Histological analyses were performed on samples that were fixed with 10% neutral buffered formalin and embedded in paraffin. Samples were sectioned at 5 μm and analyzed with hematoxylin and eosin staining. High-resolution images were taken with the Leica DM 2500 microscope (Leica Microsystems Inc, Bannockburn, IL) in transmitted light mode and analyzed both objectively and quantitatively. Quantitative analysis was performed to determine surface tortuosity of the basement membrane ( Figure 3 , D) of the FMs with ImageJ software (National Institutes of Health, Bethesda, MD).
Statistical analyses
Statistical analyses were performed with SigmaStat software (version 3.0; SPSS Inc, Chicago, IL). One-way analysis of variance was applied for statistical analysis; Fisher’s least significant difference test was used for post hoc multiple comparisons. The differences were considered statistically significant when the probability value was < .05. All results are expressed as means ± SD. Furthermore, linear regression analysis was performed across all species to determine the presence of any species-specific physiological correlations to results of biaxial puncture testing. Results from biaxial puncture were used for this analysis because of more statistically significant differences among study groups. Additionally, gestational period of various mammalian species was tested against birthweight to determine whether any correlation exists between the length of a species’ gestation and its average birthweight. Data for the gestational period to birthweight test ( Appendix 1 ) were obtained from the San Diego Zoo AnimalBytes mammal database ( http://www.sandiegozoo.org/animalbytes/a-mammal.html ).
