Deletion/loss of bone morphogenetic protein 7 changes tooth morphology and function in Mus musculus: implications for dental evolution in mammals

Quantifying regulatory gene effects on dental morphology and function has implications for the underlying mechanisms that generated dental diversity in mammals. We tested the hypothesis that regulatory gene expression changes lead to differences in molars using a neural crest knockout of bone morphogenetic protein 7 (BMP7) in Mus musculus. Three-dimensional geometric morphometric methods were used to quantify the shape of the molar toothrow. BMP7 mutants have extra cusps on the first upper and lower molars, and alterations in cusp orientation and morphology. Furthermore, significant shape differences between control and mutant were found for upper and lower toothrows. Mutant mice also exhibited differences in attrition facets, indicating functional changes that could lead to advantages in chewing new food resources and eventually niche diversification. The size ratio of the molars in the toothrow remained unchanged, implying that BMP7-induced changes in molar size ratio are a result of knocking out epithelial, rather than neural crest, expression of BMP7. Our results indicate that changes in BMP7 expression are sufficient to alter the morphology and function of the toothrow, suggesting that BMP7 or genes affecting its function could have played a role in structuring the dental diversity of extinct and extant mammals.


Mutant Lower Molars
Mutant First Molar (m1) The first molar showed a higher degree of within group variability in mutant mice compared to the second and third molars, especially on the mesial surface of the tooth. Mesial to the labial anteroconid there was a small accessory cusp in all but one of the mutant specimens. This, in addition to a more mesially located labial anteroconid, made the mesial outline surface of the tooth appear squared. The extra cusp on the mesial aspect was either present directly mesial to the labial anteroconid, or the accessory cusp was positioned mesiolingual to the labial anteroconid and mesiobuccal to the lingual anteroconid, these two morphotypes were equally common ( Figure 2). When the extra cusp was mesial to the valley between the anteroconids it was roughly incorporated into the occlusal surface of the anterior chevron (Figure 2), and when the cusp was mesial to the labial anteroconid there was a valley between these two cusps and the lingual anteroconid ( Figure 2). Rather than pointing dorsally, the cusps on the mesial surface of the tooth were slightly posteriorly recurved.
The cusps on the toothrow were individualized and presented less ridging between cusps. This was present in all three chevrons.
The cusps on the distal portion of the tooth had variable wear patterns that were present more commonly in the older age sets, where the wear facets had more time to develop. The entoconid commonly possessed variable morphologies from a mesial or distal view, including a stepped morphology and an M-shaped morphology with the center of the cusp, where the maximum point of the cusp is usually present, dipping ventrally. The hypoconid also had this stepped morphology, and was commonly almost worn off.

Mutant Second Molar (m2)
The shape of the mutant second lower molar was rhomboid, as compared with the square shaped second molar in control specimens. The mesial margin of the m2 did not run parallel to the distal margin of the m1 and instead, the metaconid-protoconid chevron was triangular, with the point of the triangle created by the medial portion of the protoconid, directed mesiobuccally towards the hypoconid on the m1. In general the cusp sizes were variable but the mesial cusps (i.e. protoconid and metaconid) tended to be larger than the distal cusps (i.e. hypoconid and entoconid). In many cases, especially in the older age sets, the hypoconid was nearly worn flat ( Figure 2).

Mutant Third Molar (m3)
The third molars were located lingually on the toothrow in relation to the mesial teeth. Mutant teeth were shorter mesiodistally, and the mesial margin of the tooth was buccolingually wider. Unlike the controls, the two small cusps on the mesial side of the tooth were not even in size. The metaconid was large and contributed to half of the overall tooth size, whereas the protoconid was very small. The distal portion of the tooth came to a point, making the outline shape of the mutant m3 triangular, rather than the rhomboid shape observed in control specimens ( Figure 2).

Mutant First molar (M1)
Mutant first upper molars were broader buccolingually than those of the control mice and their outline shape was squared. Similar to the lower first molar, the upper first molar was most variable in the cusp morphology mesial on the M1. Out of the 25 mutant specimens, fourteen of them had three extra upper mesial cusps, seven had two, and four had one ( Figure 2). Specimens with one extra cusp usually had it placed mesial to the valley between t2 and t3, but one specimen had it present in the notch between t1 and t2 ( Figure 2). In specimens with two extra cusps, the cusps were located either with both mesial to t2 and t3, or one mesial to t2 and t3 and one in the notch between t1 and t2 ( Figure   2). In the specimens that had three extra cusps, one cusp is located mesial to t2, one mesial to the valley between t2 and t3, and the third cusp is in the concave notch of the outline mesial to the space between t1 and t2 ( Figure 2). The addition of these mesial cusps contributed to differences in shape of the mesial outline of the tooth, and they were present across all age sets. The mutant first chevron, t1, t2 and t3, was positioned further distally from the mesial outline edge of the tooth, compared with the location of the chevron in control molars. As with the lower molar morphology, the ridges between the cusps of the tooth were less defined and the cusps were more individualized, meaning that they are not as connected to other cusps via ridges. The anterior lingual cusps of the tooth, t1 and t4, were shifted away from one another, with t1 positioned anteriorly and t4 positioned distally. The posterior aspect of the tooth and the t8 were wider buccolingually.

Mutant Second Molar (M2)
The second molar in the upper row, like the second molar in the lower row, was positioned lingually when compared with the first molar. t4 was shifted distally, which increased the amount of separation between t1 and t4. Rather than the t4, t5 and t6 chevron forming a triangular shape with the tip facing mesially, as seen in controls, t5 was located distally and t6 mesially, making the ridge parallel with the mesial surface of the tooth. The buccal cusps, t6 and t9, were positioned mesially. In some specimens, the maximum of the t1 cusp recurved distally. Additionally, t8 was flattened, broadened, and almost completely worn in some of the specimens from older age sets.

Mutant Third Molar (M3)
The M3 was positioned lingually in the toothrow in comparison with M1 and M2. The tooth was wider buccolingually than it was mesiodistally, compared to the equilateral triangle-shape of the control M3. The t8 extended lingually and mesially, towards t4 on the M2.  Figure 5: Regression of size on shape of the upper cusp landmark set. Black points correspond to control specimens and red points correspond to mutant specimens. Numbers beside specimen points indicate the age set that the specimen belongs to. Warpgrids show the shape differences of the landmark configuration at maximum and minimum centroid size. SI2 Figure 6: Regression of shape on size of the Lower Outline landmark set. Black points correspond to control specimens and red points correspond to mutant specimens. Numbers beside each point correspond to their age set. Warpgrids show the shape differences at minimum and maximum centroid size. SI2 Figure 7: Regression of shape on size of the upper outline landmark set. Black points correspond to control specimens and red points correspond to mutant specimens. Numbers beside each point indicate the age set they belong to. Warpgrids show the shape differences at maximum and minimum centroid size. Table 1. Mus musculus specimens employed in this study, classified according to age class, genotype, and sex.