SEARCH

SEARCH BY CITATION

Keywords:

  • craniofacial growth;
  • paleoanthropology;
  • human evolution;
  • Neandertals;
  • modern humans;
  • Sus scrofa;
  • rigid plate fixation

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION AND CONCLUSIONS
  6. Acknowledgements
  7. LITERATURE CITED

The influence of the chondrocranium in craniofacial development and its role in the reduction of facial size and projection in the genus Homo is incompletely understood. As one component of the chondrocranium, the nasal septum has been argued to play a significant role in human midfacial growth, particularly with respect to its interaction with the premaxilla during prenatal and early postnatal development. Thus, understanding the precise role of nasal septal growth on the facial skeleton is potentially informative with respect to the evolutionary change in craniofacial form. In this study, we assessed the integrative effects of the nasal septum and premaxilla by experimentally reducing facial length in Sus scrofa via circummaxillary suture fixation. Following from the nasal septal-traction model, we tested the following hypotheses: (1) facial growth restriction produces no change in nasal septum length; and (2) restriction of facial length produces compensatory premaxillary growth due to continued nasal septal growth. With respect to hypothesis 1, we found no significant differences in septum length (using the vomer as a proxy) in our experimental (n = 10), control (n = 9) and surgical sham (n = 9) trial groups. With respect to hypothesis 2, the experimental group exhibited a significant increase in premaxilla length. Our hypotheses were further supported by multivariate geometric morphometric analysis and support an integrative relationship between the nasal septum and premaxilla. Thus, continued assessment of the growth and integration of the nasal septum and premaxilla is potentially informative regarding the complex developmental mechanisms that underlie facial reduction in genus Homo evolution. Anat Rec, 2010. © 2010 Wiley-Liss, Inc.

The genus Homo is characterized by a trend toward reduced craniofacial size and robusticity beginning in the Pleistocene and continuing into the Holocene (Franciscus and Trinkaus,1995; Trinkaus,2003; Holton and Franciscus,2008; Pearson,2008; Maddux and Franciscus,2009). This structural diminution is most pronounced in H. sapiens who exhibit a considerable reduction in overall facial size, projection, and prognathism (Weidenreich,1941; Moss and Young,1960; Franciscus,1995; Vinyard and Smith,2001; Zollikofer, 2001; Lieberman et al.,2002,2004a,b; Ponce de Léon and Trinkaus,2003; Holton and Franciscus,2008). A variety of ultimate causal mechanisms have been proposed to account for this evolutionary dynamic (see Lieberman,2008); however, our ability to test hypotheses regarding morphological change in Homo relies on our understanding of the proximate causal mechanisms that underlie phenotypic variation in fossil forms. To this end, the development and integration of the chondrocranium has been a primary focus in detailing developmental (i.e., proximate) changes upon which evolutionary (i.e., ultimate) mechanisms operate (Moss and Young,1960; Enlow,1990; Ross and Ravosa,1993; Lieberman et al.,2000; McCarthy and Lieberman,2001; Jeffery and Spoor,2002,2004; Bastir and Rosas,2006; Bastir et al.,2006; Lieberman et al.,2008).

As one component of the chondrocranium, the nasal septal cartilage has been argued to have a significant influence on prenatal and early postnatal human midfacial growth. The nasal septal traction model emphasizes the morphogenetic capacity of the nasal septum and considers the midface, including the premaxilla, as responsive to developmentally induced biomechanical forces placed on the facial skeleton (e.g., along the premaxillary suture) during septal expansion (e.g., Scott,1953). Thus, as the nasal septum increases in length, tension is placed on the premaxilla via the septo-premaxillary ligament (Latham,1970; Gange and Johnston,1974; Mooney and Siegel,1986,1991; Siegel et al.,1990). This dynamic results from the nasal septal cartilage acting as a growth plate, which develops through a combination of interstitial cellular division, chondrocyte hypertrophy, and endochondral ossification along its caudal border (Scott,1953; Baume,1961; Catala and Johnston,1980; Copray,1986; Wealthall and Herring,2006).

The potential role of the nasal septum in the growth of the facial skeleton is particularly interesting given that components of this structure (i.e., the septal cartilage and perpendicular plate of the ethmoid) are developmentally continuous with the anterior cranial base and yet are spatially located within the upper and midfacial skeleton. Furthermore, in contrast to the rest of the chondrocranium, endochondral growth of the nasal septum continues into adulthood (Van Loosen et al.,1996) following a developmental trajectory similar to the facial skeleton. Thus, although nasal septal growth may be influenced by larger cranial base dynamics, it is potentially more highly integrated with the facial skeleton. Given the recent emphasis placed on elucidating developmental factors that affect the intrinsic growth of the facial skeleton (Lieberman et al.,2008; Holton et al.,2010; Bastir et al.,2010), including during prenatal development (Jeffery and Spoor,2002,2004), determining how the nasal septum may influence facial form is important to provide a more thorough understanding of the complex ontogenetic processes that were presumably altered during genus Homo evolution.

A number of experimental studies have emphasized the importance of the nasal septum in facial growth largely through surgical extirpation of all or part of the nasal septum, which typically results in a deficiency in anteroposterior growth of the maxilla and premaxilla (e.g., Wexler and Sarnat,1961; Ohyama,1969; Riesenfeld,1970; Latham et al.,1975; Wada et al.,1980; Siegel and Sadler,1981; Squier et al.,1985). Experimental work by Sarnat and Wexler (e.g., Wexler and Sarnat,1961; Sarnat and Wexler,1966,1967) as well as Rhys-Evans and Brain (1981), for example, found that resection of the nasal septal cartilage in a rabbit model resulted in a reduction in the length of the snout including the premaxilla, nasal bones, and palate. Surgical procedures that affect the growth of the vomer-premaxillary suture similarly result in midfacial retrusion (Friede,1978; Friede and Morgan,1976). Integration between the nasal septum and premaxilla is particularly interesting given variation in the timing of premaxillary suture fusion between archaic and recent humans, which may contribute to variation in facial prognathism (Maureille and Bar,1999).

Previous experimental studies largely suggest a direct causal relationship between nasal septal and anterior midfacial growth. There are, however, exceptions to these studies (e.g., Strenström and Thilander,1970; Freng,1981; Cuparo et al.,2001), leading some to minimize the importance of the nasal septum as a craniofacial growth center. The negative results reported by these authors, however, are potentially due to the variation among animals in the regions of the cartilage that are most proliferative during growth (e.g., Long et al.,1968; Strenström and Thilander,1970; Searls and Kinser,1972; Copray,1986; Van Loosen et al.,1996; Ma and Lozanoff,1999; Wealthall and Herring,2006). As such, partial resection of the septum may not include these proliferative regions. Furthermore, researchers have contended that facial reduction resulting from septal extirpation may be a function of surgical trauma rather than an interruption of a normal developmental process (Moss et al.,1968; Latham et al.,1975; Squier et al.,1985; Copray,1986; Roberts and Lucas,1994; Enlow and Hans,1996). However, minimally invasive septal extirpation techniques also produce significant reductions in facial length (e.g., Ohyama,1969) suggesting that facial reduction is not a secondary result of the surgical procedure.

The present analysis builds on previous experimental work to further assess the role of the nasal septum on facial growth (i.e., premaxillary length), particularly to better understand the potential affects of this developmental dynamic on genus Homo facial evolution (i.e., reduction in facial size and prognathism). In particular, we have taken a novel experimental approach, using a pig model, to experimentally reduce the length of the facial skeleton via rigid plate fixation of the frontonasomaxillary and zygomaticomaxillary sutures (Holton et al.,2010). This research design allowed us to test specific hypotheses regarding the integration of the nasal septum and premaxilla that follow directly from the nasal septal traction model without surgical alteration to the septum itself. As such, rather than altering the nasal septum to assess the effects on facial growth, we have altered other aspects of the craniofacial skeleton (i.e., experimentally induced synostosis) to examine the effects on the nasal septum and its integration with other components of the facial skeleton. Our hypotheses are as follows:

Hypothesis 1: A reduction in anteroposterior facial length, via rigid plate fixation of the circummaxillary sutures, results in no change in nasal septal length. If the nasal septum contributes to the anterior growth of the facial skeleton, we predict that septal length in pigs with experimentally shortened faces will not be significantly different from pigs with normal facial lengths. That is, we predict the nasal septum to reach its normal length in spite of facial growth restriction.

Hypothesis 2: A reduction in anteroposterior facial length results in a compensatory increase in premaxillary length. Due to the morphogenetic capacity of the nasal septum, we predict that pigs with experimentally shortened facial lengths will exhibit an increase in the length of the premaxillary component of the palate compared to pigs with normal facial lengths.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION AND CONCLUSIONS
  6. Acknowledgements
  7. LITERATURE CITED

Samples and Experimental Methods

Our study sample consisted of n = 30 domestic Sus scrofa (i.e., 10 female sibship cohorts). Rather than use miniature pigs, which are a more commonly used animal model, we opted to use the larger domestic pig, which exhibits a faster rate of growth and is therefore more appropriate for studies involving growth and development (Smith and Swindle,2006; Swindle and Smith,2008). The sibship cohorts were divided into three trial groups. The experimental group (n = 10) underwent bilateral rigid plate fixation of the circummaxillar sutures (i.e., zygomaticomaxillary and frontonasomaxillary) at 2 months of age (Fig. 1). Oxytetracycline (Liquamycin LA-200, Pfizer Pharmaceuticals, 200 mg/mL) was administered (20 mg/kg) 1 day before surgery, and at 5 and 10 weeks postsurgery. A combination of medetomidine HCL (Dormitor, Orion Pharma, Orion Corporation Espoo, Finland, 1 mg/mL) 0.08 mg/kg/butorphanol tartrate (Torbutrol, Fort Dodge Animal Health, Fort Doge, IA, 0.5 mg/mL) 0.2 mg/kg/glycopyrrolate (Robinul, Baxter Healthcare Corporation, Deerfield, IL, 0.2 mg/mL) 0.01 mg/kg was administered intramuscularly as a premedication. Ketamine HCL (Ketaset, Fort Dodge Animal Health, Fort Doge, IA, 100.0 mg/mL) was used as the induction agent and was delivered intramuscularly at a dose of 10.0 mg/kg. All experimental pigs were intubated intratracheally and placed on isoflurane (Forane, Baxter Healthcare Corporation, Deerfield, IL).

thumbnail image

Figure 1. Rigid miniplates were bilaterally affixed across the frontonasomaxillary and zygomaticomaxillary sutures in the experimental pigs. Note that this figure is for illustrative purposes only and this pig was not included in the analysis.

Download figure to PowerPoint

To restrict growth at the zygomaticomaxillary suture, a 1.5 mm CP titanium microplate (KLS-Martin L. P., Jacksonville, Fla.) was affixed over the suture using 1.5 mm titanium alloy cross driven microscrews. Placement of the plate was ∼10.0 mm ventral and 20.0 mm caudal to the medial canthus of the eye. With respect to the frontonasal and nasomaxillary sutures, a 1.5-mm CP titanium L-shaped microplate was affixed across both sutures. A sham group (n = 10) also underwent surgery at age 2 months receiving only screw implantation. A control group (n = 10) underwent no surgical procedure.

After 4 months of postsurgical growth (i.e., 6 months of age), pigs were euthanized using a combination of 2.5 mL xylazine HCL (Rompun, Bayer HealthCare, Shawnee Mission, KS, 100 mg/mL)/2.5 mL Ketamine HCL (Ketaset, Fort Dodge Animal Health, Fort Doge, IA,100 mg/mL) added to 5 mL of Tiletamine HCL/Zolaepam HCK (Telazol,100 mg/mL, Fort Dodge) administered at 1 mL/100 pounds intramuscularly as a tranquilizer and 30 mL of 26% sodium pentobarbital and 7.8% isopropyl alcohol intravenously as euthanasia solution. A 4-month experimental period (i.e., 2–6 months of age) was chosen as female domestic Sus scrofa exhibit a considerable increase in facial length during this time (∼40%; Holton, et al., unpublished data).

One of the control pigs did not survive to age 6 months and the cranium of a sham pig became disarticulated during processing. These individuals were therefore excluded from the analysis. Diet was kept constant among all three trial groups. To ensure that the surgical procedure did not differentially affect the dietary behavior of the pigs, thus inducing nonexperimentally controlled growth alterations, the trial groups were weighed at ages 2 and 6 months. Weight differences among the trial groups were assessed using the Kruskall–Wallis test. All procedures were approved by the animal care committees of the University of Iowa and Iowa State University.

Analytical Methods

Measurements for univariate and bivariate analyses are found in Fig. 2. Given that the sample used in this analysis was taken from a series of macerated crania used in a previous study (Holton et al.,2010), the cartilaginous portion of the nasal septum was not available. We thus used vomer length, which spans nearly the entire length of the nasal cartilage both in adults and during fetal development (Fig. 3), as a proxy for nasal septal cartilage length. Although the nasal septal cartilage and vomer have different embryologic origins, previous research has documented a highly integrated relationship between these structures. For example, as with surgical extirpation of the nasal septal cartilage, extirpation of the vomer and interruption of the vomer-premaxillary suture in long-snouted animals also reduces the anterior growth of the maxilla and premaxilla (Wexler and Sarnat,1961; Latham et al.,1975; Wada et al.,1980). A similar result has also been documented in humans (Friede,1978; Freide and Morgan,1976). Furthermore, in cases of cleft lip and palate both the nasal septal cartilage and vomer exhibit an increase in growth when compared with normal individuals (Kimes et al.,1992). Similarly, in cases of nasal septal deviation in both human and nonhuman mammals, both the nasal septal cartilage and the vomer can be affected (Gray,1978; Takahashi,1987). Combined, the results of these studies suggest integrated growth between the nasal septal cartilage and vomer. As such, in lieu of the nasal septal cartilage, the use of vomer length is justified. To ensure measurement repeatability, intraobserver error was assessed by retaking linear measurements on 25% (n = 7) of the total sample.

thumbnail image

Figure 2. Linear measurements used in univariate and bivariate analyses. All measurements were calculated as the distance between coordinate landmarks. Facial length (ac) was measured as the horizontal distance from the foramen cecum to prosthion. This measurement was taken using CT scans of the pig crania. Vomer length (be) was measured as the distance from the anterior vomeral tip to the posterior aspect of the vomer near the palate. Palate length (cf) was measured as the distance between prosthion and staphylion. Premaxillary length (cd) was measured as the distance between prosthion and the most posterior aspect of the premaxillary suture in the midsagittal plane. Post-premaxillary palate length (df) was measured as the distance between the most posterior aspect of the premaxillary suture in the midsagittal plane and staphylion. We note that while a subset of the pig skulls were hemisected, this was done in a parasagittal plane to preserve midline structures. As such, the vomer was not affected.

Download figure to PowerPoint

thumbnail image

Figure 3. Sagittal section of a neonatal pig cranium. The vomer (dotted outline) approximates the length of the nasal septal cartilage. The vomer similarly runs nearly the entire length of the snout in our study sample pigs (see Fig. 2).

Download figure to PowerPoint

To test Hypothesis 1, we first assessed univariate differences in vomer length (Fig. 2, measurement b–e) among the trial groups using a nonparametric Mann-Whitney U test. Following from the nasal septal traction model, we predict no significant differences in vomer length among the trial groups. In addition, we assessed vomer length relative to facial length (Fig. 2, measurement a–c) in bivariate space using least-squares linear regression analysis with facial length as the independent variable and vomer length as the dependant variable. We predict that the while the slopes of the regression lines should be similar, the regression line of the experimental group should exhibit a greater y-intercept. Thus, following from the nasal septal traction model, vomer length, relative to facial length should be greater in the experimental group. Statistical significance in regression line parameters was assessed using ANCOVA (Sokal and Rohlf,1995).

We tested Hypothesis 2 by assessing variation in absolute premaxilla length (Fig. 2, measurement c–d) and post-premaxillary palate length (Fig. 2, measurement d–f). We similarly assessed relative changes in the length of the premaxilla and post-premaxillary palate by standardizing these measurements against total palate length (Fig. 2, measurement c–f). We predict that restriction of maxillary growth (and thus the post-premaxillary palate) in the experimental pigs will result in an increase in both absolute and relative premaxilla length. Univariate differences in these variables were also assessed using a nonparametric Mann-Whitney U test. In addition, we assessed the bivariate relationship between premaxilla length relative to facial length using least-squares linear regression. We predict, that for a given facial length (independent variable), the experimental pigs will exhibit an increase in the length of the premaxilla (dependent). As such, the experimental group is predicted to exhibit a significantly greater y-intercept.

Hypothesized differences in the relative lengths of the premaxilla and post-premaxillary palate were further tested using geometric morphometrics. A series of unilateral two-dimensional coordinate palatal landmarks (Fig. 4) were digitized from computed tomography scans with the crania oriented along the palatal plane (i.e., along a horizontal line through prosthion and staphylion). Coordinate landmarks were first aligned and scaled for size using generalized Procrustes analysis. Next, multivariate shape differences were quantitatively assessed using principal components (PCs) analysis on the Procrustes scaled shape variables.

thumbnail image

Figure 4. Landmarks used in the geometric morphometric analysis. (1) prosthion; (2) most anterior aspect of the premaxillary suture located just lateral to the incisive canal; (3) most posterior aspect of the premaxillary suture in the midsagittal plane; (4) most posterior aspect of the premaxillary suture along the lateral aspect of the palate; (5) posterior aspect of the maxillary tuberosity; (6) staphylion.

Download figure to PowerPoint

The geometric morphometric analysis was conducted using Morphologika (O'Higgins and Jones,1998). All other statistical analyses were conducted using NCSS (Hintze,2001).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION AND CONCLUSIONS
  6. Acknowledgements
  7. LITERATURE CITED

There were no significant differences between the control and sham groups in any of the variables. These groups were therefore combined into a single control/sham trial group. Furthermore, it is of note that there were no significant differences in weight between trial groups at the time of the surgical procedure (P = 0.899) and at euthanasia (P = 0.911). This indicates that the morphological differences between the trial groups are a function of sutural restriction and that the surgical procedure did not have any detrimental effects to the general health of the pigs nor did it likely induce any nonexperimentally modified growth patterns. We further note that intraobserver error averaged less than 1.0% for our linear measurements.

Hypothesis 1

As a result of rigid plate fixation of the circummaxillary sutures, facial length was significantly reduced (P < 0.001) in the experimental group, which exhibited a mean facial length value of 132.1 mm compared with a mean value of 144.0 mm for the control/sham group (Table 1, Fig. 5a). In contrast, vomer length was identical in the trial groups with a mean value of 135.5 mm in the experimental group and 135.6 mm in the control/sham group (Table 1, Fig. 5b). As is evident from the scatter plot in Fig. 6, there was a significant positive relationship between vomer length and facial length in both the experimental (r = 0.70; P = 0.025) and the control/sham (r = 0.86; P < 0.001) groups. However, although there was no significant difference in the slopes of the regression lines between the two groups (F = 0.06; P = 0.809), the experimental group had a significantly greater y-intercept (F = 43.90; P < 0.001). Thus, for a given facial length, the experimental pigs were characterized by a relatively longer vomer when compared to the control/sham pigs.

thumbnail image

Figure 5. (a) Distribution of facial length in the control/sham and experimental groups. The experimental group, relative to the control/sham group, exhibited a significant reduction in the length of the facial skeleton (P < 0.001). (b) Distribution of vomer length in the control/sham and experimental groups. There was complete overlap of the distributions for vomer length, and thus there was no significant reduction in the experimental group (P = 0.992).

Download figure to PowerPoint

thumbnail image

Figure 6. Scatter plot of vomer length on facial length. Vomer length and facial length were significantly positively correlated within both the experimental (r = 0.70; P = 0.025) and the control/sham (r = 0.86; P < 0.001) groups. There was no significant difference in the slopes of the regression lines (F = 0.06; P = 0.809); however, the y-intercepts were significantly different (F = 43.90; P < 0.001). Relative to facial length, the experimental pigs (black circles) are characterized by an increase in vomer length when compared to the control/sham pigs (gray circles).

Download figure to PowerPoint

Table 1. Descriptive Statistics and Mann-Whitney U Test Results for All Measurements
MeasurementsExperimentalControl/ ShamP
Mean (s.d.)Mean (s.d.)
  1. Significant differences are in bold.

Facial length (mm)132.10 (5.20)144.00 (8.50)<0.001
Vomer length (mm)135.52 (7.23)135.55 (8.07)0.992
Palate length (mm)158.95 (8.20)158.37 (7.76)0.360
Premaxilla length (mm)41.63 (2.72)38.25 (2.46)0.002
Relative premaxilla length (%)26.52 (1.01)24.15 (0.85)<0.001
Post-premaxilla palate length (mm)116.03 (6.51)119.92 (6.06)0.115
Relative post-premaxilla palate length (%)73.92 (1.36)75.73 (1.04)0.001

Hypothesis 2

Circummaxillary sutural restriction had no effect on palate length (P = 0.360) with the control/sham group exhibiting a mean of 158.4 mm compared with a mean of 159.0 mm in the experimental group (Table 1, Fig. 7a). Nevertheless, there were differences in the lengths of the individual skeletal components that comprise the palate. Premaxillary length was significantly increased (P = 0.002) with a mean of 41.6 mm in the experimental group compared with 38.3 mm in the control/sham group (Table 1, Fig. 7b). When standardized against total palate length, premaxillary length, in the experimental group, comprised, on average, 26.52% of palate length, whereas the premaxilla in the control/sham group comprised, on average, 24.15% of palate length. The differences in relative premaxillary length were highly significantly different (P < 0.001) and, as evidenced by the box plot in Fig. 7c, there is little overlap between the distributions, and no overlap between the interquartile ranges.

thumbnail image

Figure 7. (a) Distribution of palate length in the control/sham and experimental groups. There was no significant difference between the trial groups (P = 0.360). (b) Distribution of premaxilla length in the control/sham and experimental groups, which was significantly greater in the experimental group (P = 0.002). (c) Distributions in relative premaxillary length in the control/sham and experimental groups. Relative premaxillary length was greater in the experimental pigs (P < 0.001) indicating that premaxilla length makes up a relatively greater component of total palate length.

Download figure to PowerPoint

The difference in premaxilla length between the two groups is further evident in the scatter plot in Fig. 8. Premaxilla length and facial length were positively correlated within both the experimental (r = 0.58; P = 0.087) and the control/sham (r = 0.67; P = 0.002) groups, however, relative to the control/sham group, the experimental group occupies a different morphospace. Although there was no significant difference between the slopes of the regression lines (F = 0.16; P = 0.687), the y-intercept of the experimental group was statistically significantly greater (F = 16.00; P < 0.001). Thus, for a given facial length, the experimental pigs are characterized by a larger premaxilla. In some cases (40% of the experimental pigs), the increase in premaxillary length resulted in an anterior displacement of the upper incisors relative to the lower incisors (i.e., increased overjet; Fig. 9).

thumbnail image

Figure 8. Scatter plot of premaxilla length on facial length. Premaxilla length and facial length were positively correlated within both the experimental (r = 0.58; P = 0.087) and the control/sham (r = 0.67; P = 0.002) groups. There was no significant difference in the slopes of the regression lines (F = 0.16; P = 0.687); however, the y-intercepts were significantly different (F = 16.00; P < 0.001). Thus, for a given facial length, the experimental pigs (black circles) are characterized by a relative and absolute increase in premaxillary length when compared to the control/sham pigs (gray circles).

Download figure to PowerPoint

thumbnail image

Figure 9. Comparison of control/sham pig (a) and experimental pig (b). Note that the more procumbent mandibular incisors in the control/sham pig extend anterior to the more vertically oriented maxillary incisors. In contrast to the normal condition, increased growth of the premaxilla resulted in an anterior displacement of the upper incisors (i.e., increased overjet) in a subset (40%) of the experimental pigs.

Download figure to PowerPoint

In contrast to premaxillary length, post-premaxillary palate length (Fig. 10a) in the experimental group was reduced (mean = 116.03 mm) relative to the control/sham group (mean = 119.92 mm). This reduction, however, was not significantly different (P = 0.115). When standardized against total palate length, the experimental group was characterized by a statistically significant reduction in relative post-premaxillary palate length (P = 0.001; Fig. 10b).

thumbnail image

Figure 10. (a) Distribution of post-premaxillary length for the control/sham and experimental groups. There was no significant difference in post-premaxillary length (P = 0.115). (b) Distribution of relative post-premaxillary length (standardized against palate length) for the control/sham and experimental groups. Relative post-premaxillary length was reduced in the experimental pigs (P = 0.001).

Download figure to PowerPoint

The results of the geometric morphometric analysis further illustrate the morphological differences between the experimental and control/sham groups. Individual PC scores along PC1 (30.2%) for the experimental and control/sham groups were significantly different (P < 0.001) with virtually no overlap between the experimental and control/sham pigs (Fig. 11a). Morphological variation along PC1 is largely a function of differences in the relative sizes of the premaxilla and post-premaxillary palate (Fig. 11b). The premaxilla in the experimental group is relatively larger as evidenced by the more posterior positioning of the premaxillary suture (landmarks 2–4 in Fig. 2). As such, the post-premaxillary palate is relatively reduced in the experimental group. In contrast, the control/sham group is characterized by a relatively smaller premaxilla and relatively larger post-premaxillary palate.

thumbnail image

Figure 11. (a) Scatter plot of individual PC scores of the Procrustes shape residuals for the first two principal components. PC1 (30.2%) separates the experimental pigs (gray circles) from the control/sham pigs (black circles) such that there is virtually no overlap between the groups. Differences in individual PC scores along PC1 are highly statistically significant (P < 0.001). (b) Visualization of morphological shape variation along PC1 (30.2%). The experimental configuration (positive PC scores) is illustrated on the right while the control/sham configuration (negative PC scores) is illustrated on the left. The experimental configuration exhibits a more posteriorly positioned premaxillary suture. As such, it is characterized by a relatively longer premaxilla and relatively shorter post-premaxillary palate. In contrast, the control/sham group is characterized by a relatively shorter premaxilla and a relatively longer post-premaxillary palate. Note that the wireframe representation of the control/sham pigs in (b) is mirrored.

Download figure to PowerPoint

DISCUSSION AND CONCLUSIONS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION AND CONCLUSIONS
  6. Acknowledgements
  7. LITERATURE CITED

The results of our analysis indicate that the developmental relationship between the nasal septum and premaxilla, as one component of midfacial growth, contributes to variation in adult facial length (e.g., Wexler and Sarnat,1961; Sarnat and Wexler,1966; Mooney et al.,1989; Wealthall and Herring,2006). Our first hypothesis that facial growth restriction has no effect on the length of the nasal septum was indirectly supported by this analysis. Although the experimental group exhibited a significant reduction in the length of the facial skeleton the length of the vomer, used as proxy for nasal septal cartilage length, was unaffected as would be predicted by the nasal septal traction model. Thus, the nasal septum continued to grow in spite of a reduction in facial length suggesting that nasal septal length is not a secondary response to the growth of the midfacial skeleton (Moss et al.,1968; Moss and Salentijn,1969).

Our second hypothesis that an experimental reduction in facial length results in compensatory growth in the premaxilla was also supported by this analysis. The experimental group exhibited both an absolute and relative increase in premaxillary length. Thus, although facial length was reduced, the continued expansion of the nasal septum presumably produced compensatory premaxillary growth, likely via the septo-premaxillary ligament (e.g., Latham,1970; Mooney and Siegel,1986,1991; Siegel et al.,1990). The relative increase in premaxillary length was further reflected in the multivariate geometric morphometric analysis. The premaxilla was posteriorly elongated at the premaxillary suture in the experimental group indicating a relatively greater contribution of the premaxilla to palatal length. This is further evidenced by the anterior displacement of the upper incisors relative to the lower incisors in some of the experimental pigs (Fig. 9). However, a reduction in mandibular length in the experimental pigs, resulting from a more vertically oriented mandibular ramus (Holton et al.,2010), may contribute to this as well.

When scaled to palate length, the post-premaxillary palate was significantly reduced in the experimental group. In contrast, absolute post-premaxillary palate length, while reduced in the experimental group, was not significantly different. This was due, in part, to the effects of one particularly small individual found at the low end of the control/sham range of variation. If this individual is excluded, the difference in post-premaxillary palate length achieves statistical significance (P = 0.046). It should also be noted that post-premaxillary palate length includes contributions from both the maxilla and palatine bones. As such, there may be experimentally induced variation in palatine length that affects this result as well. Unfortunately, the transverse palatine suture was partially obliterated in some individuals and thus, it was not possible to assess the contribution of this skeletal element independent of the maxilla.

Following what is likely a general mammalian developmental dynamic, the integration between the nasal septum and premaxilla potentially bear on unresolved issues regarding human craniofacial evolution. Although the chondrocranium plays a key integrative role in craniofacial development and evolution in genus Homo (Moss and Young,1960; Enlow,1990; Ross and Ravosa,1993; Lieberman et al.,2000; McCarthy and Lieberman,2001; Jeffery and Spoor,2002,2004; Bastir and Rosas,2006; Bastir et al.,2006,2010; Lieberman et al.,2008), the influence of the nasal septum, as one component of the chondrocranium, has not been as widely considered. There is, nevertheless, fossil evidence to suggest that an integrated nasal septal/premaxillary complex may account for variation in facial size between archaic and recent modern humans as Neandertal and recent human subadults exhibit taxonomic variation in the timing of premaxillary suture fusion. Maureille and Bar (1999) documented that the premaxillary suture in recent humans tends to fuse at an earlier age than in Neandertal subadults. In particular, ∼75% of the Neandertals used in their study retained sutural patency along the palatal surface and the nasal floor. This was in contrast to their modern human sample in which only 20% retained premaxillary suture patency on palatal and nasal floor surfaces. Thus, the potential for early postnatal anterior growth of the premaxilla is reduced in modern humans and may therefore partially account for a reduction in both facial prognathism and overall facial size. A similar relationship has also been documented between premaxillary suture fusion and variation in facial form among modern human populations. Mooney and Seigel (1986) found evidence for prolonged premaxillary suture patency in a sample of African derived subadults compared with a sample of European derived subadults, which vary in subnasal alveolar prognathism. Moreover, Kieser et al. (1999) documented a high frequency of premaxillary suture patency in a sample of Maori crania and suggested that this may contribute to the large facial dimensions that characterize this population.

Earlier fusion of the premaxillary suture in H. sapiens, relative to archaic Homo, suggests a developmental shift in the dynamics that regulate sutural fusion/patency. Unlike neurocranial sutures, which are regulated via tissue interaction with the dura matter (Opperman et al.,1995; Opperman,2002), the nasal septal cartilage itself may play an important role in the morphogenesis and regulation of facial sutures (Adab et al.,2002,2003). Thus, in addition to passively placing the premaxillary suture under biomechanical strains, thereby promoting osteoblastic and fibroblastic activity (e.g., Rafferty and Herring,1999; Kopher and Mao,2002; Mao et al.,2003; Mao,2006; Katsaros et al.,2006), the nasal septum may also play a more active role in premaxillary suture development.

Given the influence of the nasal septum on premaxillary growth, the potential role of the nasal septum on variation in facial size and projection in genus Homo may be limited to prenatal and early postnatal development prior to the obliteration of the premaxillary suture. This is evidenced, in part, by the increased frequency of septal deviation in humans relative to other mammals in which the premaxillary sutures remains patent through life (Gray,1978; Takahashi,1987; see also Rönning and Kantomaa,1985). Nevertheless, the postnatal influence of the nasal septum potentially extends beyond the facial skeleton proper affecting the spatial relationship between the facial skeleton and neurocranium. Experimentally induced facial suture synostosis coupled with continued growth of the nasal septum results in changes in cranial base flexion (Rönning and Kantomaa,1985), anterior cranial base length (Ruan et al.,2008) and the orientation of the facial skeleton relative to the cranial base and neurocranium (Rönning and Kantomaa,1985; Mooney et al.,1992; Holton et al.,2010). Thus, relative differences in nasal septal and facial growth potentially explain part of the variation in cranial base angulation due to facial size differences between modern and premodern hominins (i.e., facial packing; e.g., Lieberman et al.,2008). The contribution of the nasal septum to this dynamic, however, requires further study.

The precise influence of an integrated nasal septal/premaxillary complex on the development of facial form is incompletely understood. The results of this study, however, contribute to a broader understanding of the growth of this complex as one dynamic that potentially influenced facial reduction in genus Homo. Given the early ontogenetic fusion of the premaxillary suture in humans, the contribution of nasal septal and premaxillary growth may be restricted to prenatal and early postnatal growth. Nevertheless, species-specific (e.g., Neandertal versus modern human) and population-specific variation in facial form begins to manifest early in ontogeny (Minugh-Purvis,1988; Mooney and Siegel,1991; Maureille and Bar,1999; Williams,2000,2006; Ponce de Léon and Zollikofer, 2001, 2006; Viǒarsdottir et al.,2002; Krovitz,2003). Moreover, although the nasal septum likely influences facial prognathism via premaxillary sutural growth, its influence on other aspects of facial growth (e.g., growth at other facial sutures, and anteroinferior displacement of the palate) is unknown. Although the nasal septal cartilage in humans attains its adult size early in development, the perpendicular plate of the ethmoid continues to grow through endrochondral ossification of the septal cartilage into adulthood (Van Loosen et al.,1996). This pattern is similar to that seen in mice (Wealthall and Herring,2006) and is thus likely similar to that in other hominins. Thus, as one aspect of craniofacial development, a continued assessment and more precise understanding of the growth and integration of the nasal septum and premaxilla is likely to help elucidate the complex developmental mechanisms that underlie facial reduction in genus Homo evolution.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION AND CONCLUSIONS
  6. Acknowledgements
  7. LITERATURE CITED

The authors thank Dr. Dean Riedesel and Matt Keller for their assistance with the surgical procedures and Ken Krizan for postsurgical specimen preparation. Dr. Suzanne Stock provided helpful suggestions on the manuscript.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION AND CONCLUSIONS
  6. Acknowledgements
  7. LITERATURE CITED