Dental peculiarities in the silvery mole-rat: an original model for studying the evolutionary and biological origins of continuous dental generation in mammals

Ontogenic stages in the silvery mole-rat

An embryo of advanced stage of development shows one mineralized cheek tooth per dental row (Fig. 2A, ID119), while the second tooth is developing because a bony crypt can be observed. One day after birth (ID140), the second tooth starts to mineralized. One week later (ID142), the first tooth starts to erupt, and the second one is more developed but not yet covered by an enamel layer. A two-week-old specimen shows both teeth erupting with crown entirely mineralized, but with root still developing (Fig. 2B, ID1bis). After two months, the first teeth are worn and their root is entirely developed, and the third tooth is mineralizing and ready to erupt (Fig. 2C, ID2bis). The two-year-old and four-year-old specimens show the more frequently observed adult dentition, in having four functional teeth occluding with teeth still erupting distally (Figs. 2D–2E, ID13 and ID178). The first specimen shows remnants of teeth in resorption (Fig. 2D, ID13), while the mesial most teeth are highly reduced in the second specimen (Fig. 2E, ID178). The oldest specimen (nine-year-old) presents an abnormal dentition (Figs. 2F–2H, ID180). Upper and lower mesial-most teeth are normally developed, but present malocclusions and are ready to shed, although they are not strongly worn. Moreover, odontomas are formed in the distal part of the dentition and involved both upper dental rows. These odontomas are importantly developed, they spread in incisor alveoli, and include numerous malformed teeth, which subsequently appear in the virtual section (Fig. 2H).

Dental morphology and eruptive sequence in the silvery mole-rat

Description

Cheek teeth of Heliophobius are very high-crowned, but their growth is limited since their single root is closed at the basal end (Fig. 2). As a result, both upper and lower teeth are cone-shaped. Unerupted teeth show enamel free areas (i.e., dentine can be already observed, Fig. 3A). Worn upper teeth shows heart-shaped occlusal surface, while lower teeth show pear-shaped occlusal surface (Fig. 3C).

• DP₄ of Heliophobius is molariform (Fig. 3A). It shows a mesially located metaconid compared to the protoconid, which is less developed. The metalophulid I links these cusps. The metaconid also has a distal spur on the labial border, which reaches the entoconid and closes the mesosinusid. The ectolophid is marked in its mesial part linked to the protoconid, but it is interrupted to faintly marked at the level of the hypoconid. The sinusid is short and transverse. The posterolophid is strong and reaches the distal part of the entoconid. It can be assumed that the hypolophid transversally joins the hypoconid and the entoconid because of the crescent shape of the distal dentine area.

• M₁ is clearly bilophodont, with an ectolophid missing to reduced, but more centrally located (Fig. 3A). As a result, the sinusid and the mesosinusid are equal in size. The mesial lophid consists of the protoconid and the metaconid connected by the metalophulid I. The distal lophid joins the reduced hypoconid to the entoconid. It is difficult to state if the hypolophid is present or fused with the posterolophid.

• M₂ presents the same pattern, but the second lophid is much more reduced, and shows reduced entoconid and crest-like hypoconid (Fig. 3B). Following teeth present the same morphology as well, in having a highly reduced hypoconid (Fig. 3C).

• DP⁴ is molariform. The protocone is strong and linked to the paracone via the anteroloph. The paraloph is short and incipient and does not reach the protocone. Paracone and metacone are contiguous and close the mesosinus. The strong endoloph joins the protocone and the hypocone, which is mesio-distally compressed. The sinus is short and transverse. The hypocone is connected to the posteroloph, which distally joins the metacone. The metaloph is probably absent (Fig. 3A).

• M¹ presents a very close pattern (Fig. 3A). However, the paraloph is absent, and the hypocone is not reduced.

• M² shows a reduced protocone compared to the paracone (Fig. 3B). The anteroloph is weak to interrupted between these cusps. The distal part is much more reduced, with a small hypocone and a highly reduced and crest-like metacone. Following teeth have the same pattern (Fig. 3C).

Characteristics of dental resorption and bone remodelling

Evidence of continuous mesial drift can be observed in bone tissues for specimens of a pygmy-rock wallaby, a manatee and a silvery mole-rat (Figs. 4A–4E). In the wallaby, the interalveolar septum presents a distally compacted aspect resulting from bone resorption, while the mesial part is illustrated by a high number of openings of vascular channels due to bone apposition (or formation, Fig. 4A), clearly observed in the manatee (Fig. 4B). In virtual cross-section of upper cheek teeth in the silvery mole-rat, bone resorption is shown on mesial side of teeth by the serrated aspect of the distal part of interalveolar septa, and by the corresponding alveolar wall showing a surface strongly etched (Figs. 4C–4D). Bone apposition is conversely illustrated on distal side of teeth by the presence of numerous openings of vascular channels in bone septa and alveolar surface (Figs. 4C and 4E). In this area, dental resorption also occurs because the enamel layer is reduced to absent and the outline is irregular due to the indirect compressive force of erupting distal molars also leading to periodontium disruption. Dental resorption is strongly efficient at the mesial-most side of the cheek tooth row, where the layer of cementum tends to be thicker (Fig. 4C).

Normal initial dental development

Before developing supernumerary molars, juveniles of manatees and of the pygmy rock-wallaby show a dental eruptive sequence similar to other mammals in having deciduous premolars before molars (Sanson, 1980; Domning, 1982; Gomes Rodrigues et al., 2011; Fig. 5). One of the major constraints when studying the dentition of the silvery mole-rat is the very high wear, which does not permit accurate morphological descriptions. The investigation of very young specimens helps us to determinate the eruptive sequence of this rodent. As a result, the crown morphology of both first lower and upper erupted cheek teeth strongly looks like DP4 of other African mole-rats. In DP⁴ of Heliophobius, the metaconid is more mesially located as in Cryptomys and Georychus, and the pattern is more complex than molars, which have a marked bilophodont structure, as in Bathyergus. The difference between DP₄ and lower molars is more subtle, but the paraloph is present as in Cryptomys, and the second loph (hypocone-posteroloph-metacone) is not reduced. This means that the initial development of the cheek tooth row is comparable to what it is observed in most Bathyergidae, and in mammals with CDR as well (Fig. 5). These results also corroborate the hypothesis of Landry (1957), who proposed that the fourth premolar is coupled with molars and “that the extra teeth in Heliophobius represent entirely new dental elements added on to the end of the usual mammalian series” (p. 68).

Heliophobius has an eruptive dental sequence which differs from Heterocephalus, due to the occurrence of premolars, as seen in first known bathyergids from the Miocene of Africa (Mein & Pickford, 2008). Heterocephalus and Heliophobius clearly  display derived but opposite dental evolutionary trends, involving dental reduction in the oldest bathyergid lineage (only three or two molars, Fig. 5), and development of additional teeth in the lineage of Heliophobius. Data on extinct relatives of Heliophobius are thus essential to assess the evolutionary origins and the potential adaptive nature of this mechanism. More precisely, knowledge of the climatic, environmental and geological contexts relative to the appearance of CDR is needed. In the meantime, the functional characteristics of CDR in Heliophobius also remain to be precisely described in vivo, inasmuch as its dentition including originally four high-crown cheek teeth seems already highly efficient compared with other mole-rats, having higher bite force at molars, while digging with their incisors (e.g., Fukomys; Van Daele, Herrel & Adriaens, 2009).

The timing of dental eruptive sequence of Heliophobius is also very close to the timing known in Georychus, in which a new born has one teeth, a two-week-older specimen has two erupted teeth, and a four-week specimen has three erupted teeth (Taylor, Jarvis & Crowe, 1985). Unfortunately, we have no data concerning the development of the fourth teeth in Heliophobius, which might be tightly associated with the development of additional molars. It can be only observed that M2 and subsequent molars display a highly reduced distal parts. The additional molars are very similar to the penultimate molar, as observed in wallabies, and as noticed by Domning (1982) in manatees: “Trichechids were turning out unlimited numbers of “second molars” …” (p. 609). More material, as well as further developmental analyses are needed to precisely describe the formation of supernumerary teeth in Heliophobius. More generally, presence of premolars rapidly replaced by additional molars is another evidence of close convergent dental evolution between very different mammals, involving a rodent, a sirenian, and a macropod which had to face abrasive diet and/or terrigenous matter (Domning, 1982; Sanson, Nelson & Fell, 1985; Gomes Rodrigues et al., 2011; Šklíba, Šumbera & Vítámvás, 2011), probably just after weaning.

Evidence of continuous dental replacement

One of the most interesting point concerns evidence of CDR in Heliophobius. Landry (1957) proposed that extra teeth can be present in this rodent, while Gomes Rodrigues et al. (2011) went further and proposed a CDR by analogy with what happens in manatees and the pygmy rock-wallaby. But, here, the study of the oldest specimen provides another evidence of continuous generation of teeth. The presence of bilateral odontomas in the distal part of the dentition in a rodent specimen supposed to have its last molars erupted for a while (e.g., between one and two years in Georychus; Taylor, Jarvis & Crowe, 1985) is intriguing, if we are not aware of the possibility of continuous dental development in supposed diphyodont animals. But, in the case of mammals with CDR, it can be related to developmental disruption probably linked to ageing. This was observed in a few specimens of old mice having bilateral odontomas at the level of the proximal end of incisors, in relation to their continuous growth (Dayan et al., 1994). This is obviously an evidence of the maintenance of active dental stem cells at the back of jaw. As incisors were also involved in this disorder and because their developmental origin is back and very close to last erupting molars (i.e., in outgrowth of pterygoid processes), a possible molecular interaction between incisor and molar tissues, which regulates their continuous activity, cannot be rejected and remains to be investigated. In any event, such results validate the hypothesis of continuous dental generation in the silvery mole-rat. Only when its breeding will be possible, precise and detailed investigation of mechanisms of dental development would permit to test the assumption of an extension of the dental lamina allowing CDR, conversely to its regression in diphyodont mammals, in which part of developmental mechanisms have already been described at the histological level (Buchtová et al., 2012).

Evidence of intense dental drift

Three stages of dental mesial drift (or physiological drift) were defined in rodents (Fig. 5). The first one, observed in Georychus and Bathyergus, corresponds to a slight dental displacement and does not involve any dental loss (Gomes Rodrigues et al., 2011; Gomes Rodrigues, 2015; Fig. 5). The moderate mesial drift is present in all gundis (Ctenodactylidae), which only lose anterior teeth under the pressure of erupting posterior teeth (Gomes Rodrigues et al., 2012; Figs. 4 and 5). The strong or continuous mesial drift is only known in the silvery mole-rat because of the CDR leading to the constant loss of teeth at the front of the jaw. This dental dynamic involves bone remodelling and resorption of dental tissues whose intensity increases according to the degree of dental drift. This is evidenced when comparing alveolar bone characteristics between gundis (e.g., Felovia, Figs. 4F–4H; Gomes Rodrigues et al., 2012) and the silvery mole-rat. This latter shows a higher number of openings of vascular channel when bone is forming, and a more intensively etched surface when bone is in resorption, as similarly does the pygmy rock wallaby and the manatee. Moreover, alveolar septa are thinner in Heliophobius compared to other mole-rats in relation to constant bone remodelling (Gomes Rodrigues, 2015). Mesial teeth are highly resorbed and constantly covered by a thick layer of cementum, which might correspond to cementum repair protecting teeth (Jäger et al., 2008), necessary to maintain anchoring of the teeth in the dental alveolus via the periodontal ligament (Saffar, Lasfargues & Cherruau, 1997; Fleischmannová et al., 2010). Cemental modifications have also been reported in manatees (Miller, Sanson & Odell, 1980). The constant formation of teeth at the back of the jaw drives intense compressive forces which induce strong physical constraints on the whole dentition and on surrounding tissues (periodontium, alveolar bone). Nonetheless, dental disorders and misalignments are rare, because of the presence of a diastema in front of cheek teeth in these mammals.

The development and functionalization of the tooth in time and space occurred in close relation with the development of the surrounding bone and periodontal tissues. Moreover, the homeostasis of the alveolar bone is characterized by a high turn-over during the life of a tooth and by specific interactions with the periodontal ligament (Fleischmannová et al., 2010), which are modified during dental drift. The impact of dental drift has only been studied in the context of orthodontic researches in rat and mouse (e.g., Ren, Maltha & Kuijpers-Jagtman, 2004; Yoshimatsu et al., 2006), in which the migration is activated by an experimental system producing tensile forces (Wise & King, 2008). However, rat and mouse were rather described as having a slight distal drift (Johnson & Martinez, 1998; Fig. 5), without any detail on the cause of migration. The silvery mole-rat is pertinent for investigating the effect of the permanent migration of teeth on surrounding tissues involving notably homeostasis of alveolar bone and the periodontium, and also cementum repairs. The accurate description of the dental dynamics and its impact on both teeth and bone at the histological level in the silvery mole-rat and in rodents showing different degrees of drift will permit to comparatively assess the effects of natural migration and the associated mechanical constraints sustained by the whole masticatory apparatus. This could provide new insights on the functional properties of these various masticatory apparatus in relation to the ecology of these rodents.

Evolutionary insights on the horizontal CDR

Interestingly, if the silvery mole-rat, manatees, and the pygmy rock-wallaby, represent unique cases of CDR in mammals, such an extended horizontal replacement has also been reported in non-mammalian synapsids (i.e., Traversodontidae^†: Hopson, 1971; Liu, Bento Soares & Reichel, 2008; Liu & Sues, 2010; Tritylodontidae^†: Kühne, 1956; Cui & Sun, 1987). These Mesozoic forms were assumed as herbivores and ingested fibrous plants (Kemp, 2005), probably also abrasive. This is more obvious in multi-cuspate tritylodontids (Fig. 5), especially derived taxa, which have higher dental wear in the mesial part of the dentition, and show evidence of dental drift and also dental resorption similar to what happen in manatees (Kühne, 1956; Cui & Sun, 1987; Jasinovski & Chinsamy, 2012).

The central point is that these synapsids acquired an extended or continuous horizontal replacement of more complex and larger teeth combined with the progressive loss of anterior post-canines. In addition, this sequential replacement was superimposed to the initial alternate and vertical replacement known in most non-mammalian synapsids (e.g., Thrinaxodon^†, Fig. 5), which were not limited to two dental generations contrary to mammals (Hopson, 1971; Osborn & Crompton, 1971). Hopson (1971) partly explained the reduction of the number of dental generation in mammals, by the presence of more complex teeth anchored in a socket via periodontal ligaments (i.e., thecodonty) allowing to cope with mechanical stresses resulting from strong dental pressure during occlusion, which in contrast makes them more difficult to replace. In that context, eruption of teeth at the back, drifting then along the jaw, appears as a more suitable means to renew the dentition, inasmuch that such displacement can be favoured by transseptal fibers linking adjacent teeth, as noticed by Miller, Sanson & Odell (1980) in manatees, and also kangaroos, which have dental drift. However, because of the compressive force of erupting teeth, the perfect anchoring of teeth is a necessary prerequisite to avoid dental disorders. Some traversodontids, tritylodontids, manatees, and the pygmy rock-wallaby have elongated roots, which prevent such problems. The silvery mole-rat is the only one chisel-tooth digger among bathyergids to have very high-crowned teeth, which are also cone-shaped as in some traversodontids (e.g., Boreogomphodon^†, Liu & Sues, 2010). That characteristics has probably facilitated the mechanical superimposition of CDR, whatever its function, as well as mesial drift also observed in other high-crowned mole-rats (e.g., Bathyergus).

The biological origins of CDR in mammals remains however to discover. Even if mammals having CDR show striking morphological convergences with non-mammalian synapsid groups, it seems unlikely that the same developmental pathways permitting new generation of teeth distally were reactivated after such a very long period of time (Gomes Rodrigues et al., 2011). Nonetheless, specific biological parameters probably shared between non-mammalian-synapsids and these mammals might have favoured the acquisition of CDR. For instance, it should be noticed that these mammals have low basal metabolic rates (Gallivan & Best, 1980; Janis & Fortelius, 1988; Mc Nab, 2005; Zelová et al., 2007), as show “reptiles” (i.e., Sauropsida), which generally have continuous generation of teeth, even if not horizontal. Direct or indirect relations between CDR and such biological parameters cannot be demonstrated for the moment. In any event, these troubling “reptile-like” similarities clearly need deeper considerations, and the silvery mole-rat could appear as an appropriate candidate for further researches. This will contribute to a better understanding of the origins of the mammalian CDR, but also to the gathering of new data on the biological mechanisms leading to the reduction of the dentition at the origin of mammals.