James Cray Jr, Department of Plastic Surgery, Children’s Hospital of Pittsburgh, University of Pittsburgh, 3533 Rangos Research Center, 530 45th Street, Pittsburgh, PA 15201, USA. T: +1 412 692 9040; E: James.Cray@chp.edu
Research has shown that Pan and Homo have similar ectocranial suture synostosis patterns and a similar suture ontogeny (relative timing of suture fusion during the species ontogeny). This ontogeny includes patency during and after neurocranial expansion with a delayed bony response associated with adaptation to biomechanical forces generated by mastication. Here we investigate these relationships for Gorilla by examining the association among ectocranial suture morphology, cranial volume (as a proxy for neurocranial expansion) and dental development (as a proxy for the length of time that it has been masticating hard foods and exerting such strains on the cranial vault) in a large sample of Gorilla gorilla skulls. Two-hundred and fifty-five Gorilla gorilla skulls were examined for ectocranial suture closure status, cranial volume and dental eruption. Regression models were calculated for cranial volumes by suture activity, and Kendall’s tau (a non-parametric measure of association) was calculated for dental eruption status by suture activity. Results suggest that, as reported for Pan and Homo, neurocranial expansion precedes suture synostosis activity. Here, Gorilla was shown to have a strong relationship between dental development and suture activity (synostosis). These data are suggestive of suture fusion extending further into ontogeny than brain expansion, similar to Homo and Pan. This finding allows for the possibility that masticatory forces influence ectocranial suture morphology.
The cyclic strains of mastication have a high degree of influence on ectocranial suture morphology and fusion (Herring, 1993, 2008; Byron et al. 2004, 2006). Compensatory bone growth and remodeling occurs at the suture osteogenic fronts in response to both tensile forces and compressive forces (Herring & Teng, 2000; Mao, 2002; Alaqeel et al. 2006; Herring, 2008). Sutures adapt under compressive forces, resulting in more interdigitation, thicker bones and narrower sutures. Widening and bone elongation in sutures are results of tension and cyclic strain forces (Fong et al. 2003; Wu et al. 2007; Herring, 2008). It has been demonstrated that the cranial sutures respond to strain throughout ontogeny by increasing complexity. Wave patterns and interdigitations form, and a relatively constant sutural width is maintained by bony adaption to strain prior to osseous fusion (Yu et al. 2009). Thus, strain magnitudes would have to increase throughout ontogeny to have additive effects on bone morphology.
Despite their close phylogenetic relationship, Pan, Homo and Gorilla suture fusion patterns (suture site fusion progression) have been shown to differ slightly. The suture sites most susceptible to the mechanical influences of the temporal and masseter muscles, the lateral anterior (portions of the coronal, sphenotemporal and sphenofrontal sutures) exhibit similar patterns of fusion for all three species. Gorilla, however, exhibits a strong posterior to anterior gradient for vault sutures, suggesting an independent ectocranial suture pattern (Meindl & Lovejoy, 1985; Cray et al. 2010). It is unknown whether the differences in ectocranial suture morphology and fusion pattern observed for Gorilla are due to underlying genetic differences between these species, or to influences of the bony adaptation to the hypermuscular Gorilla form.
No investigations have been conducted concerning the timing of ectocranial suture activity in Gorilla. Research has shown that biomechanical forces may have a greater influence on ectocranial suture activity because masticatory forces and suture fusion continue further into the life cycle than neural expansion (Meindl & Lovejoy, 1985; Cray et al. 2008, 2010). The current study was designed to test the relationship, in Gorilla, between suture fusion activity and neurocranial expansion and dental development as a proxy for the onset and influence of masticatory forces. In particular, we investigated suture morphology (commencement and termination of osseous activity at the suture site) and its relationships to both cranial volume (as a proxy for brain growth) and dental eruption (as a proxy for the age of the individual and the length of time that it has been masticating hard foods and exerting such strains on the cranial vault). Gorilla represents a hyper-masticatory model and may have earlier masticatory influence on the sutures compared with Pan and Homo due to their growth differences in ontogeny. If suture adaptations are observed to extend later in ontogeny after brain growth is completed, then they likely result from biomechanical forces generated by the masticatory apparatus.
Materials and methods
Two-hundred and fifty-five Gorilla gorilla skulls (all wild-shot), housed at the Cleveland Museum of Natural History, were examined at 10 ectocranial suture sites by the primary author. Seven of the suture sites were classified as cranial vault sutures: midlambdoid, lambda, obelion, anterior sagittal, bregma, midcoronal, pterion; and three of the sutures were classified as lateral–anterior sutures (two from the cranial vault were classified as both): midcoronal, pterion, sphenofrontal inferior sphenotemporal, and superior sphenotemporal (Fig. 1; Meindl & Lovejoy, 1985; Cray et al. 2008). Each suture site was examined to determine suture synostosis activity and dichotomously scored for these two categories: (i) commencement of activity – defined as the earliest onset of bone formation activity within the fibrous joint (scored as a zero if not commenced or one if commenced); or (ii) termination of activity/obliteration – defined as the cessation of that activity or synostosis, i.e. the fibrous joint is replaced by bone (scored as a zero if not completely obliterated or one if obliterated; Fig. 2. The scores were then added to obtain a score for the sum of active suture sites for commencement (0–7 for vault and 0–5 for lateral anterior analyses) and the sum of fused/obliterated suture sites for termination (0–7 for vault and 0–5 for lateral anterior analyses; Meindl & Lovejoy, 1985; Cray et al. 2008). In this study, the primary author observed each skull twice, on separate days. Intraobserver reliability was 98.5%, meaning that over 98% of the suture site observations across study specimens were the same at each of the two observation times.
The direct determination of cranial volume (in cm3) and sex determinations were provided by the Cleveland Museum of Natural History as part of their database of information on skeletal specimens. Cranial volume determination followed the methodology of Simmons (1942), including sand, water and anthropometric techniques.
As a proxy for ontogeny and masticatory muscle influence, dental eruption status was also scored for each skull and assigned to the following categories according to the sequence of permanent dental eruption in accordance with Krogman (1930):
2first permanent molar eruption (defined as erupted into the occlusal plane);
3permanent incisors and premolars eruption (defined as erupted into the occlusal plane);
4second permanent molar eruption (defined as erupted into the occlusal plane);
5permanent canines and last molars eruption (defined as erupted into the occlusal plane);
6wear of age.
To determine if the pattern of differences in cranial volume vary by the sum of suture activity scores (commencement and termination for vault and lateral anterior suture sites) by sex (to account for the great sexual dimorphism in Gorilla), a two-way anova was conducted for each analysis to assess the significance of the interaction term. This analysis elucidated whether there were sex-specific patterns in neurocranial expansion as related to suture morphology.
To assess the statistical relationship between suture activity and cranial volume, a linear regression model, using a single predictor, was used to assess the relationship between cranial volume, and the sums of cranial vault and lateral anterior suture commencement and termination (0–7 for vault, and 0–5 for lateral anterior) were treated as the independent variables, respectively. Cranial volume was treated as the dependent variable or predicted variable. Pearson’s product moment correlation coefficient (r) was used as a measure of linear association (Vittinghoff et al. 2005). This analysis allowed for the determination of whether neurocranial expansion was related to suture activity, bridging and synostosis in Gorilla ontogeny.
To assess the statistical relationship between the sum of active suture sites score (an ordinal variable) and dental status (an ordinal variable), a Kendall’s tau (τ) was calculated (Ferguson, 1976). Sexual dimorphism was assessed for the sum of active and terminated suture sites for vault and lateral anterior sutures, as well as dental status. This analysis allowed for the determination of whether suture activity was related to dental eruption as a proxy for ontogeny, and the length of time that it has been masticating hard foods and exerting such strains on the cranial vault. All analyses were conducted using spss 15 (Chicago, IL, USA).
Demographics were provided by the Cleveland Museum of Natural History as part of their database of information on skeletal specimens. Of the 255 skulls that were analyzed, 69 were classified as subadults and 186 as adults. Of the subadults, 17 were classified as unknown sex, 27 males and 25 females. There were 117 males and 69 females in the adult classification group.
The sample was culled (n = 150) to include only those skulls that had data for cranial volume analysis. A 2 × 8 between-subjects anova was performed on cranial volume as a function of sex (male, female) and vault suture commencement and termination score (0–7), respectively. The assumption of normality was met for all groups, P > 0.05 for each analysis, and the assumption of homogeneity of variance was met for each analysis, P > 0.05. The interaction terms demonstrate there is no significant pattern of difference on cranial volume as a function of sex and vault suture commencement score, F5,137 = 0.921, P = 0.470, or as a function of sex and vault suture termination score, F7,134 = 1.116, P = 0.357.
A 2 × 6 between-subjects anova was performed on cranial volume as a function of sex (male, female) and lateral anterior suture commencement and termination score (0–5), respectively. The assumption of normality was met for all groups, P > 0.05 for each analysis, and the assumption of homogeneity of variance was met for each analysis, P > 0.05. No significant pattern of difference on cranial volume as a function of sex and lateral anterior commencement score, F5,138 = 0.465, P = 0.802, or as a function of sex and vault termination score, F5,138 = 0.866, P = 0.506 was found, as demonstrated by the interaction terms.
The resulting regression model for cranial volume predicted by the sum of vault suture commencement score was not significant (r2 = 0.019, P = 0.078), cranial volume = 463.582 + 4.931 × vault suture commencement score. Likewise, the regression model for cranial volume predicted by the sum of vault suture termination score was not significant (r2 = 0.007, P = 0.275), cranial volume = 485.631 + 1.868 × vault suture termination score. A regression model for cranial volume predicted by the sum of lateral anterior suture commencement score was not significant (r2 = 0.017, P = 0.100), cranial volume = 474.188 + 5.004 × lateral anterior suture commencement score. The regression model for cranial volume predicted by the sum of lateral anterior suture termination score was again not significant (r2 = 0.002, P = 0.621), cranial volume = 491.589 + 1.121 × lateral anterior suture termination score. Figure 3 represents the resulting regression models with constants included.
All 255 specimens were included in the dentition analysis. Sex was not significantly correlated with dental status (τ = −0.152, P = 0.12), nor was it correlated with sum of vault suture commencement score (τ = −0.079, P = 0.210), sum of lateral anterior suture commencement score (τ = −0.063, P = 0.304), sum of vault suture termination score (τ = −0.093, P = 0.114) or lateral anterior suture termination score (τ = −0.045, P = 0.455). Figures 4 and 5 exhibit the vault and lateral anterior suture activity as a function of dental status. Mean vault and lateral anterior commencement appear as early as deciduous dentition to permanent incisor and canine eruption, but it is at third molar eruption where it appears most suture sites have commenced activity. Mean vault termination first appears to occur after the incisor and canines emerge, but exhibit variability. A mean value > 1 is not observed for vault termination until third molar eruption. Mean lateral anterior termination does not begin until third molar eruption and proceeds thereafter.
Dental status was significantly correlated with sum of vault suture commencement score (r = 0.721, P < 0.001), sum of lateral anterior suture commencement score (r = 0.774, P < 0.001), sum of vault suture termination score (r = 0.839, P < 0.001) and sum of lateral anterior suture termination score (r = 0.601, P < 0.001). All analyses exhibited highly significant positive correlations, suggesting a strong association between advanced dental status and osseous suture bridging and synostosis.
Gorilla does not have a strong association between suture activity and brain volume, similar to what has been reported for Pan (Cray et al. 2010) and Homo (Meindl & Lovejoy, 1985; Cohen & MacLean, 2000). Our regression models for cranial volume and suture activity represent a positive association. However, the resulting models were not significant and r2-values were very low, suggesting a poor predictive model. These data suggest that neurocranial expansion is completed before sutures begin to adapt to biomechanical forces (suture serration and bridging or fusion in non-pathological cases). Thus, the expanding brain continues to be the predominant force that shapes the neurocranium early in ontogeny (Weidenreich, 1941; Moss, 1969; Moss & Salentijn, 1969; Enlow & Hans, 1996; Mooney et al. 2002; Richtsmeier et al. 2006), even in this truncated ontogeny where Gorilla display relatively early suture adaptation activity. Suture patency appears to persist at least through the evolutionarily selected neurocranial expansion for Pan, Homo and Gorilla.
In contrast, the data presented for suture scores and dental status exhibit statistically significant positive relationships. Gorilla displays early suture activity unlike that reported for both Pan and Homo, where suture activity has a strong relationship with third molar eruption (Cray et al. 2010). This relationship is interpreted as the result of the allometrically larger masticatory apparatus for Gorilla (Hylander, 1979; Shea, 1983, 1985; Daegling, 1989, 2007; Taylor, 2002; Daegling & Hotzman, 2003; Taylor et al. 2008). However, even with earlier suture activity, it appears that the ectocranial suture sites do not fuse until after third molar eruption, similar to that reported for Pan and Homo (Meindl & Lovejoy, 1985; Cray et al. 2010). It is observed that mean vault termination first appears to occur after the incisor and canines emerge, but exhibit variability. In fact, the mean values for stages 3 and 4 are both < 1, suggesting a propensity of the individuals observed had not begun termination, and those cases that had were only observed for on average one suture site. Thus, statistically, it appears the onset of termination is not observed for most specimens until eruption of the third molar. This relationship is observed for lateral anterior termination as well, which does not begin until third molar eruption and proceeds thereafter. The extended delay observed between dental eruption and cessation of suture activity is not as pronounced in the Gorilla as in Pan and Homo (Cray et al. 2010). However, there was no suture fusion observed before the third molar eruption in Gorilla. This earlier display of suture adaptation appears to be due to the increased masticatory musculature and apparatus of the Gorilla. Thus, mastication appears to effect ectocranial suture activity of Gorilla much more than the expanding neurocranium.
In contrast, it has been shown that compressive forces dissipate with suture remodeling in rodents where sutures remain patent (Sun et al. 2004; Shibazaki et al. 2007). In addition, Henderson et al. (2004) modeled early sutural strain in humans with results that suggested that tissue level strains were likely too small to effect osteoblast biology. These contrasting data are suggestive of a lack of effect for biomechanical forces on the suture. The data presented here support the model that adaptation in suture morphology is temporally linked to the development of the masticatory apparatus and later chewing. The gorilla has the most robustly developed masticatory muscles of the hominoids. We observe few changes to the morphology of the calvarial sutures prior to the completion of neurocranial expansion similar to the pattern observed for both Homo and Pan (Cray et al. 2010). In fact, these data suggest that suture fusion in the gorilla is only temporally related to eruption of the third molar (dental maturity). It is important to note that the data presented here were abstracted from dry skull morphological investigation. The direct causation of adaptation in suture morphology and eventual fusion likely involves biomechanical adaptation, tension- and compression-related changes to the extracellular matrix and cells of the perisuture area, as well as mechanotransduction of cells eliciting the expression of growth factor gradients and cellular expression related to osteoblast and osteoclast activity.
What may be puzzling about the effects of biomechanics on the calvarial sutures, is the relatively higher strains affecting the facial sutures (Adab et al. 2002; Mao, 2002; Rafferty et al. 2003; Al-Mubarak et al. 2005). The facial sutures in most hominoids do not fuse except in rare cases very late in ontogeny (Todd & Lyon, 1925a,b,c; Krogman, 1930). It has been suggested that facial suture patency may be adaptive to protect the thin bones of the face during dynamic loading, thus acting as a type of shock absorber (Jaslow, 1990; Rafferty & Herring, 1999; Rafferty et al. 2003). It is likely similar growth factors are at work in the morphogenesis, growth and maintenance of both cranial and facial sutures, transforming growth factors (TGFβs), bone morphogenetic proteins (BMPs), fibroblast growth factors (FGFs), twist transcription factor (TWIST), Msh Homeobox 2 (MSX2), Runt-related transcription factor (RUNX2). It may be, however, that these growth factors and others, matrix metalloproteinase (MMP1 and 2) and insulin-like growth factor (IGF1 and 2), for example, have differential expression or different metabolisms between cranial and facial sutures (Rafferty & Herring, 1999; Adab et al. 2002; Rafferty et al. 2003; Al-Mubarak et al. 2005). There is no consensus concerning facial suture patency, and further research is required.
Finally, it is likely that Gorilla, Pan and Homo exhibit heritable age-related changes to the sutures due to heritability of perisutural growth factor concentration gradients, including the TGFβ, FGFs and the Ets-2 CAM-kinase pathways (Cohen & MacLean, 2000; Opperman & Ogle, 2002; Poisson et al. 2004; Rawlins & Opperman, 2008; Yu et al. 2009). It is also likely that the earlier suture remodeling observed in the Gorilla may be related to the increased size of the masticatory muscles and apparatus observed for this species.
In conclusion, like Pan and Homo the expanding neurocranium is unrelated to the onset or cessation of suture fusion activity in Gorilla. However, the Gorilla model does exhibit earlier suture response to the age of the individual and the length of time that it has been masticating hard foods and exerting such strains on the cranial vault sutures. This causes earlier, biomechanically mediated adaptation to the suture during Gorilla ontogeny. Only after neurocranial expansion is complete does this occur, suggesting the expanding brain still is the important driving force for early suture patency. Future research should continue to explore the relationship of ectocranial suture activity among and between hominoids, and primates in general, and the causal mechanism of suture synostosis.
The authors would like to thank Lymann Jellema of the Cleveland Museum of Natural History for access to the study specimens. We would also like to thank Emily Lensie Durham for critical reading of this work.