For many diseases, the patient population exhibits a wide range of severity, but the cause of this variation is largely unknown. This is the case for many structural craniofacial diseases, including holoprosencephaly (HPE). In HPE patients, the forebrain fails to develop into two hemispheres and the midline of the face does not fully form; however, the disease manifests in a broad spectrum of severity, with phenotypes ranging from mild midfacial narrowing to cyclopia. Although mutations in Sonic hedgehog (SHH) and in other members of the SHH signaling pathway have been linked to HPE, the source of the extreme phenotypic variability characterizing the disease is not understood. Intriguingly, research suggests that development largely buffers against variation in morphological form occurring, but that nonlinearities in development are a rich source of such buffering. Previous research in an exogenously manipulated chick model suggests that there is a non-linear relationship between SHH-signaling and midfacial shape, but it has not been tested whether such a non-linear relationship exists in a genetic model. This dissertation utilizes an allelic series of embryonic mice with discrete, genetic reductions in Shh and in low-density lipoprotein receptor-related protein (Lrp) 2 to test whether nonlinearities in SHH signaling underlie the HPE spectrum. LRP2 is an endocytic receptor required to activate SHH-signaling in the developing forebrain, and Lrp2-/- mice exhibit mild HPE. Geometric morphometric (GM) analysis uncovered no statistically significant difference in craniofacial shape between wild-type, Lrp2+/-, Shh+/-, and Lrp2+/-;Shh+/- embryos. However, the upper jaw is significantly more narrow relative to the lower jaw in Lrp2-/- and Lrp2-/-;Shh+/- embryos in comparison to other genotypes, and there is significantly higher face shape variance in Lrp2-/- and Lrp2-/-;Shh+/- embryo groups than in other genotypes. Surprisingly, SHH pathway activation (as defined by Shh, Ptch1, Gli1 mRNA levels) did not predict face shape or face shape variance. Interestingly, however, GM and gene expression analysis suggest that that growth is delayed in embryos missing functional LRP2 protein, and that this delay may underlie the HPE phenotype—but that a threshold of such delay is required for HPE to manifest.

Craniofacial anomalies are disfiguring, debilitating birth defects that afflict a large portion of the world's population. These malformations can be caused by genetic and/or environmental factors, and one such factor, prenatal hypoxia, is of particular interest in our research on craniofacial dysmorphology. There is experimental and clinical evidence for a correlation between hypoxia and craniofacial malformations, however, the mechanisms underlying such defects are not yet understood. The goal of our research has been to understand the mechanisms whereby hypoxia causes abnormal craniofacial morphology in early embryonic development.

Chick embryos were incubated in either normoxic (21% O2) or hypoxic (7%-19% O2) conditions and collected on days 2-6 for morphological and cellular analyses. Embryos were photographed for morphological analyses, and facial shape variation was quantified via two-dimensional geometric morphometrics. 13-day embryos were cleared and stained for analysis of skeletal and cartilaginous development. Cell proliferation was assessed via BrdU staining, and apoptosis was assessed by whole-mount and section TUNEL and caspase 3 immunoassay. Embryos were examined for oxidative stress using a phospho-AMPK immunoassay.

In morphometric analyses of normal growth, older embryos (HH22-28) showed greater shape variation among embryos than younger embryos (HH16-21). All the embryos fell along a well-defined nonlinear curve of normal facial growth in relation to size, and displayed variation in relation to chronological age and discrete morphological stage. Within stages, older embryos displayed greater variation within stages than younger embryos. In studies of hypoxic embryos, their survival was reduced in a step-wise manner in comparison to normoxic control embryos. Hypoxic embryos showed a wide range of craniofacial anomalies, from mild asymmetry and eye defects to more severe frontonasal and cephalic anomalies. They also displayed delayed skull bone development, with some skeletal defects. Abnormal facial shape variation occurred in relation to centroid size and age among individuals in hypoxic groups versus the normoxic population. Hypoxia disrupted cell proliferation and caused apoptosis of neural crest progenitor cells. Hypoxic embryos also displayed increased metabolic stress response.

Objectives: Craniofacial malformations are among the most common birth defects, affecting ~1 in every 700 live births. However, our understanding of the mechanisms that result in these diseases is limited. Many cases have been associated with genetic predispositions that have allowed us to attempt to investigate the underlying mechanisms for these diseases. Sonic Hedgehog (SHH) signaling pathway has undergone significant investigation due to the involvement of Shh in the development of the midface. In humans, deletion of a single copy of Shh is associated with a spectrum of phenotypes comprising Holoprosencephaly (HPE), ranging from mild hypotelorism and midfacial hypoplasia to cyclopia indicates that diseases like HPE are multifactorial. Although our long-term goal is to discover the underlying mechanisms that cause these diseases, we must first understand the normal progression of development before trying to understand the abnormal. Therefore, the goal of this study was to determine how Shh expression and face shape relate with each other during normal development.

Methods: We collected wild-type chicken embryos at 72 hrs, 96 hrs, and 120 hours of incubation. Chickens infected with RCAS-wnt3a at 72 hrs of incubation were also included in our sample. We used in situ hybridization to identify Shh expression domains and 2D geometric morphometrics to quantify changes in shape in Shh expression domains and face shape. We performed Principal Components Analysis (PCA) as well as multivariable regression analyses of Shh expression shape and face shape on somite number and centroid size. We used Partial Least Squares (PLS) to evaluate covariation in shape between Shh expression domains and facial shape.

Results: Changes in Shh expression shape and face shape are dependent on developmental time. As the embryos progressed in development, there were significant changes in both Shh expression shape and face shape. While the overall size of the embryo grew, both the Shh expression shape and face shape constricted. More specifically, Shh expression shape tapered into a narrow ¬V-shaped band in the ectoderm of the stomodeum while face shape constricted as a result of the nasal pits growing closer together while the mouth became smaller. Our PLS regression identified that the changes in Shh expression and face shape are correlative in which Shh expression shape is associated with face shape at specific timepoints during development.

Conclusion: Changes in Shh expression shape correlate with changes in face shape. This suggests that Shh expression shape may serve as a predictor for face shape during embryonic growth. Although we are unable to determine if Shh is directly responsible for the observed changes in face shape, this potential predictive relationship could be valuable for future studies to identify when and how disease progression initiates.