Musculoskeletal study of cebocephalic and cyclopic lamb heads illuminates links between normal and abnormal development, evolution and human pathologies

This paper is part of the emerging field of Evolutionary Developmental Pathology, dedicated to study the links between normal and abnormal development, evolution and human pathologies. We analyzed the head musculoskeletal system of several ‘natural mutant’ newborn lambs displaying various degrees of abnormality, from mild defects to cebocephaly and to cyclopia, and compared them with humans. Interestingly, muscle defects are less marked than osteological ones, and contrarily to the latter they tend to display left-right assymetries. In individuals with cebocephalic and even cyclopic skulls almost all head muscles are normal. The very few exceptions are some extraocular muscles and facial muscles that normally attach to osteological structures that are missing in the abnormal heads: such muscles are instead attached to the ‘nearest topological neighbor’ of the missing osteological structure, a pattern also found in cyclopic humans. These observations support Alberch’s ill-named “logic of monsters” - as a byproduct of strong developmental/topological constraints anatomical patterns tend to repeat themselves, even severe malformations displayed by distantly related taxa. They also support the idea that mammalian facial muscles reverted to an ancestral ‘nearest-neighbor’ muscle-bone type of attachment seen in non-vertebrate animals and in vertebrate limbs, but not in other vertebrate head muscles.


Platysma myoides
Mainly from skin of the neck/pectoral region to skin and muscles of mouth region region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal the superior part of buccinator area, goes diagonally forward and inserts directly in front of the orbita. The other part lies more superficial, originate from the posterior area of the cheek from the sphincter colli profundus, and inserts onto the orbicularis oculi in the area directly under the orbita region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal, although on the left side of the head the dorsal attachment was more anterior to the eye region (where it normally attaches in normal newborn lambs) region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal region of the facial tuber, which is markedly more posterior than normally) to reach the eye region, and not posterodorsally as normally; that is, basically the size and direction of the fibers changes, but the attachments are mainly normal

Supplementary Information 2, including summarized methods, as well as results, of morphometric analysis
For the morphometric analysis of all the left and right sides of the skull of each abnormal specimen dissected by us, we first removed all the soft tissues of those specimens, and then used landmarks that are consistently used for morphometric analysis of mammalian skulls (junction of the front tooth and maxilar bone; connection between the zygomatic and temporal bones in the zygomatic arch; facil tuber; around the eye; see e.g. Bookstein 1991), in order to run our analyses, which included Proctustes ANOVA. Skull halves were photographed on a flat horizontal surface with a camera Nikon D5000 fixed in a strict vertical position. Shape of skulls was analyzed using landmark-based geometric morphometric methods (Bookstein, 1991;Rohlf, 1990). Images were digitized using tpsUtil (Rohlf, 2004) and tpsDig2 (Rohlf, 2010). Series of landmarks along the mandible and eye hole were placed equidistant along the curvature. We applied the generalized Procrustes analysis (GPA) (Dryden and Mardia, 2002;Slice, 2005) in MorphoJ 1.05f (Klingenberg, 2008(Klingenberg, , 2011 to align the landmarks. Right and left halves were analyzed separately. Shape variation was studied by performing (1) Principal Component Analysis (PCA) of the data with the allometric component included (total shape variation) and (2) PCA after the regression of the centroid size on shape (pure shape) and visualizing the shape with a scatter plot and morphological differences with thin-plate spline (TPS) deformation grids (Bookstein, 1991;James Rohlf and Marcus, 1993;Slice, 2005;Thompson, 1917). To visualize the association between size and shape, we plotted shape scores against WCS. The amount of shape variation was given as a percentage of the total variation around the sample mean. The percentage numbers were computed to shows the relative importance of allometry for shape variation in each part of the skull. A permutation test with 10,000 runs (Good, 1994;Pitman, 1937) was applied to test independence between size and shape changes. A discriminant function analysis (DFA) and canonical variate analysis (CVA) were used to distinguish between groups. Procrustes ANOVA was used to estimate the fluctuating asymmetry (FA, asymmetric variation within one individual) and directional asymmetry (DA, one side is systemically different from the other one) in both centroid size and shape.
Because the lower jaw is movable, it is difficult to always keep it exactly within the same position in the skulls: because this factor may affect the overall result we analyzed the jaw separately. The results for the lower jaw are shown in Fig. 6 of the main paper, and are also shown in the figure just below. Other parts of the skull were also have to split into regions. For example, when we analyze the snout, we cannot use true cyclops (stage 4) because they basically don't have a ossified snout. On the other hand, we can analyze the brain case in all specimens. Thus, we obtained different groupings depending on what regions were included in the various morphometric analyses. The main results are summarized below: Lower jaw: Fig. PCA Lower jaw. Variation of shape of mandible in animals with different degree of defects. PCA of shape scatter plot (PC1 and PC2) and associated shape changes of non-allometric shape component of eye curvature and snout. The TPS deformation grids illustrate shape changes indicating the relative shifts of landmarks along the PC1 axes with PC scale factor +/-0.1 and along the PC2 axes with PC scale factor +/-0.05. Shape of lower jaw is significantly different between degrees 2, 3 and 4. The difference between groups is clear along PC1, which is mainly a change in curvature of lower surface of mandible and position of coronoid and condylar processes relative to body of the mandible.
• Contrarily to what we found concerning soft tissue pattern (i.e. usual asymmetry between left and right sides), there is no directional asymmetry (DA, i.e. in which one side would be systemically different from the other one) in the skull in size (Centroid Size, p= 0.7863) or shape (p= 0.7141) revealed by Proctustes ANOVA (see Table Proctustes ANOVA below): • However, there is a strong fluctuating asymmetry (FA, asymmetric variation within one individual) in both size (Centroid Size, p<0.005) and shape (p<0.0001) ( and right sides. The left and right sides develop as more or less separate copies of each other within the same genome and in nearly the same environment. There are a variety of random processes at the molecular and cellular levels that can affect development and therefore may produce small differences between body parts even if genetic and environmental differences are absent. The differences between left and right sides are an opportunity to measure this variation and FA can therefore be used as a measure of the 'imprecision' of developmental processes, or developmental instability. • All analysis ware performed with true landmarks (LM), no semilandmark (semiLM) as the curvature of the lower jaw is very different between individuals. Relaxation of semiLM leads to their rearrangement and does not depict the trend. Repeated measurements show that the error of LM analysis is relatively low (Table Proctustes ANOVA) and suggest that the results obtained are reliable. • R and L sides were not averages as each of them includes important information information.
• When the allometric component were included, the Principal Component 1 (PC1) explains 84.5% of the shape variation. Other PCs explain <5% (PC2 4.2%). Transformation grids located in corners of the graph illustrate shape changes along PC1 and PC2. For the PC1, it's mainly changing of the jaw curvature. In more defective specimens, the lower jaw is wider and bent upward; in less defective specimens, the jaw is straighter. Another shape alteration is the widening of the condylar process in cyclop specimens (stage 4). • The allometric component is very strong and can explain 49.1% of the variation (p<0.0001). In general, most defective specimens are smaller than others. However, stages 3 and 4 significantly overlap in size, as can see in this figure showing Regression: A distribution of different shapes of mandible (top panel), brain case (middle panel), and face (bottom panel) across specimens with different skull sizes. There is a clear distribution of mandible shape along size of skull and degree of defect. Specimens with stage 4 have the largest and the smallest brain cases. Animals with the stage 2 tend to have larger frontal parts of the skull than animals with stage 3, but they still substantially overlap. • Specimens 199 and 993 (the most defective, stage 4, specimens) are often outliers in the morphometric analyses. Specimen 993 has the jaw most bent upward. Specimen 199 has almost no brain, but the analysis shows that not only the brain case is defective but also the facial bones and the lower jaw, in particular. Fig. PCA Brain case. Variation of shape of brain case in animals with different degree of defects. PCA of shape scatter plot (PC1 and PC2) and associated shape change of non-allometric shape component of skulls. The TPS deformation grids illustrate shape changes indicating relative shifts of landmarks along axes with the PC scale factor +/-0.1. Specimens 993 and 199 are outliers. Other specimens do not show essential overlap in the shape of the brain case. PCA Face Stages 2 and 3: Fig. PCA Face (Stages 2 and 3). Variation of shape of frontal part of skull in animals with different degree of defects. PCA of shape scatter plot (PC1 and PC2) and associated shape changes of non-allometric shape component of eye and snout. The TPS deformation grids illustrate shape changes indicating the relative shifts of landmarks along the axes with PC scale factor +/-0.15. Shape of frontal part of skull is significantly different between stages 2 and 3. The difference between groups is present along PC1, which mainly concerns the length of the snout.  4). Variation of shape of frontal part of skull in animals with stage 4. PCA of shape scatter plot (PC1 and PC2) and associated shape changes of non-allometric shape component of the eye curvature and snout. The TPS deformation grids illustrate shape changes indicating relative shifts of landmarks along axes with PC scale factor +/-0.1. Specimens 3 and 11 have similar shapes of the face; specimen 993 is rather unique as it is distant from other specimens along both PC1 and PC2.