Technical Comments

Response to Comment on "Separate Evolutionary Origins of Teeth from Evidence in Fossil Jawed Vertebrates"

See allHide authors and affiliations

Science  13 Jun 2003:
Vol. 300, Issue 5626, pp. 1661
DOI: 10.1126/science.1084686

We welcome the comments of Burrow (1), but find the majority of her assertions inaccurate and ambiguous. Pivotal to our argument for a convergent evolution of teeth in placoderms is that basal groups lack the ordered and patterned dentition we described in the Arthrodira. Our choice of basal groups was not an innovative phylogeny, but was the most recent (2, 3). None of the basal clades noted in (1) have dental plates satisfying the structural criteria we carefully defined as rows of teeth restricted in number, with new teeth added at specific loci in a consistent and uniform pattern and “out of the bite” (4).

Regular rows of denticles that increase in height along the row and are not randomly distributed (1) precisely characterize basal arthrodires such as the Phlyctaeniida (3, 5), but not those placoderms basal to arthrodires. Burrow insists that the more basal Antiarchi, Rhenanida, and Acanthothoraci groups have organized “toothlike” structures. However, the structures identified on the dental plates of Antiarchi are only “irregular tooth-like cusps” (6). In the Rhenanida (Jagorina), the dentition comprises polygonal units with no ordered arrangement, and previous reconstructions of the dentition are simply incorrect (7). The dentition of Gemuendina, drawn schematically with neat rows of denticles in (8), needs a critical reinterpretation. The plates shown in figure 1 (A to C) in (1), described as “probably from an acanthothoracid,” are isolated plates of uncertain taxonomy, and present no solid evidence placing them as gnathal plates— especially not in the Placodermi. Teeth never overlap in this way, nor do they exhibit multiple cusps. Rather, most features conform to those of the Acanthodii (9). One articulated placoderm specimen of the acanthothoracid Romundina sp. (Fig. 1A) has both upper anterior gnathal plates in situ, demonstrating denticles in a unique and distinctive organization, unlike both those of basal arthrodires and of the adjacent dermal armor.

Fig. 1.

Comparison of dermal tubercles with oral denticles. (A) Romundina sp. ventral view of head shield showing premedian plate with dermal tubercles (der. tub.) and paired anterior supragnathals (ASG) with oral denticles in packed concentric array (photo courtesy of D. Goujet). (B) Dicksonosteus arcticus [figures 3 and 4 in (5)] reconstruction of the head in ventral view with posterior supragnathal (PSG) in position, and occlusal view showing radial rows of small to larger denticles, anterior supragnathal (ASG) not present. Note smaller denticles on the palate (p. den., parasphenoid denticles).

Relevant to our argument for a prepattern for the dentition, refined in more advanced groups as radial rows of teeth (3), is a template of oral denticles, as featured in the phylctaeniid Dicksonosteus arcticus, a basal arthrodire (5). Here, there is a radial order to the multiple rows of denticles on the supragnathal plate (Fig.1 B) with gradation of denticle size—an ideal prototype for toothed plates.

Enlarged tubercles do occur in specific dermal locations. These dentine modules can also vary with functional requirement, such as the spinelets associated with the pectoral fin. However, they are not arranged in a prepattern for tooth sets such as those described for the oral and pharyngeal denticles (10).

Contrary to Burrow's interpretation, we did not assert that “regular tubular dentine” of the teeth arises “de novo” only in the teeth of advanced brachythoracids, nor did we assert that it is absent in dermal spinelets. The significance of our observation is that regular dentine grows from cells located in a pulp cavity, without trapped cells (semidentine), which is contrary to previous accounts (11). This, together with tooth structural characteristics, reinforces our main conclusions (3).

The variety of dentine types in the dermal armor indicates a great degree of plasticity of this tissue, and has led to enormous terminological confusion. It is quite inappropriate and misleading to compare the outer zone of part of the dentine in dermal tubercles, as Burrow does “of regular sub-parallel dentine tubules,” with the entire width of the dentine in the teeth as we described (3).

Finally, we propose that denticles and teeth on the inside of the gnathostome oropharynx and the putative pharyngeal denticles (10) and teeth of placoderms (3) have different origins from those of the dermis. Specifically, they did not acquire their pattern information (as do dermal tubercles) from the ectoderm, but from the endoderm. To accept (1) the former derivation of teeth ignores new research showing that the putative viscero-skeletal cells of chick and zebrafish (12, 13) are instructed by the endoderm “as to size, shape and position of all the ventral head skeleton” of which these denticles are a part. In this way, it is suggested that pharyngeal denticles and teeth were regulated differently from those of the dermal skeleton (3, 10).


Stay Connected to Science

Navigate This Article