Osteoderms as calcium reservoirs: Insights from the lizard Ouroborus cataphractus

Abstract

The functional significance of osteoderms—bony elements embedded in the dermis—remains a topic of much debate. Although many hypotheses have been put forward in the past, the idea that osteoderms can serve as calcium reservoirs has received little experimental attention thus far. In this study, we use micro‐computed tomography to investigate inter‐ and intrasexual variation in osteoderm density in the viviparous lizard Ouroborus cataphractus and conduct histochemical analyses to unravel the potential mechanism involved in mineral resorption from the osteoderms. Our results show that females have denser, more compact osteoderms than males of similar body sizes, regardless of the season during which they were collected and their reproductive state. Furthermore, a histochemical study demonstrates the presence of mononucleated TRAP‐positive cells in the vascular canals of the osteoderms. Based on the findings of this study, we suggest that the mineral storage hypothesis merits further attention as a candidate explanation for osteoderm evolution.

“The hypothesis that osteoderms can serve as calcium reservoirs has received little attention in past studies. Using density‐calibrated micro‐computed tomography, we demonstrate that female Armadillo lizards (Ouroborus cataphractus) have denser, more compact osteoderms than males of similar body sizes. Furthermore, a histochemical study reveales mononucleated TRAP‐positive cells in the vascular canals of the osteoderms.”

1. INTRODUCTION

Osteoderms—mineralized integumentary structures located in the dermis—have evolved convergently in multiple lineages of tetrapod vertebrates (reviewed in Vickaryous & Sire, 2009). Whilst significant progress has been made with respect to their morphological aspects and phylogenetic distribution, their functional significance has remained poorly understood. Osteoderms have been hypothesized to play a role in defence against predators or conspecifics (e.g. Broeckhoven et al., 2015; Broeckhoven, du Plessis, & Hui, 2017; Laver et al., 2020) and to be involved in thermoregulation (Clarac & Quilhac, 2019; Inacio Veenstra & Broeckhoven, 2022) or other physiological processes (Clarac et al., 2020). The mineral storage hypothesis which postulates that osteoderms serve as calcium reservoirs has been discussed in the past but has received little attention so far from the scientific community. Seidel (1979) was one of the first to suggest that the highly vascularized osteoderms of alligators may serve as mineral storage and nearly four decades later, Dacke et al. (2015) provided some evidence that the osteoderm density of females with ripe ovarian follicles was greater compared to those that had recently laid or contained eggs. This finding corroborated the observations of Hutton (1986) and Tucker (1997) that osteoderms of crocodiles undergo considerable remodelling during breeding and may represent an important source of calcium during oogenesis. Similarly, Curry Rogers et al. (2011) proposed that osteoderms may have provided an important reservoir of minerals in titanosaurs. The authors suggest that high levels of fecundity coupled with inhabiting highly seasonal, semi‐arid environments could place significant demands on bone mineral reservoirs (Curry Rogers et al., 2011) and that the presence of hollow osteoderms in titanosaurs is a sign of mineral remobilization (Curry Rogers et al., 2011; Vidal et al., 2017). Nevertheless, the fact that minerals could be stored and drawn from any part of the skeleton, and not necessarily from the osteoderms, argues against the mineral storage hypothesis (Salgado, 2003; Seidel, 1979). Furthermore, Laver et al. (2020) proposed that in geckos with enlarged extracranial endolymphatic sacs, osteoderms require rather than provide calcium. To date, there is little evidence to support or refute the mineral storage hypothesis, particularly that pertaining to potential differences between sexes or reproductive states of females. Furthermore, despite studies showing resorptive lacunae consistent with osteoderm resorption (Curry Rogers et al., 2011; Dacke et al., 2015), the presence of osteoclasts and their contribution to the mobilization of calcium in adult individuals remains to be determined (Dacke et al., 2015, but see Dubansky & Dubansky, 2018).

In this study, we investigate the anatomy and histology of osteoderms in the Armadillo lizard, Ouroborus cataphractus (Figure 1) using micro‐computed tomography (micro‐CT) and TRAP staining. Armadillo lizards are heavily armoured lizards inhabiting the highly seasonal, semi‐arid environments of western South Africa. Armadillo lizards display several unique physiological and behavioural traits associated with energy constraints resulting from having a permanent group‐living lifestyle in these environments (Shuttleworth et al., 2013). During the dry season, which spans from December to April, individuals remain inactive in their rock crevices and feeding activity is almost entirely ceased (Broeckhoven & Mouton, 2015). Nevertheless, pregnancy takes place during the dry season and females give live birth to one exceptionally large offspring in April or May (Flemming & Mouton, 2002; Figure 1a). Such a lifestyle would undoubtedly have placed strong demands on mineral reserves; hence, Armadillo lizards provide a unique model system for further exploring the potential role of osteoderms as calcium reservoirs. We, therefore, aimed to address the following questions: (1) Do females differ from males in their osteoderm anatomy and mineral density, and are potential differences influenced by seasonality? (2) What is the underlying mechanism responsible for mineral resorption in osteoderms?

FIGURE 1.

(a) Adult female Ouroborus cataphractus with newborn. (b) Three‐dimensionally‐rendered micro‐CT image and corresponding slice view of a single osteoderm and the calibration phantom. (c) Visualization of the grey value analysis used to calculate osteoderm mineral density. (d) Visualization of the wall thickness analysis used to quantify differences amongst thicknesses of osteoderm tissues

2. MATERIALS AND METHODS

2.1. Micro‐computed tomography

A single osteoderm was dissected from the dorsal side of the body of preserved specimens of O. cataphractus belonging to the Ellerman Collection at Stellenbosch University. These specimens were collected by Mouton et al. (1999) under Cape Nature Conservation permits (no. 4/96 and 140/96) in a relatively small geographical area (ca 100 km²) for use in morphological, ecological and evolutionary studies. In total, samples were obtained for 29 adult females and 26 adult males that were collected in spring (October) or summer (December and April). These specimens were fixed in 10% formalin for 5 days and were consequently transferred to 70% alcohol. The fixation, preservation and storage protocols were identical for all the specimens. Because the specimens were stored in 70% ethanol, all samples were transferred to 0.9% physiological saline solution for 48 hours. Each osteoderm was placed together with a bone mineral density calibration phantom (QRM GmbH, Möhrendorf, Germany) containing hydroxyapatite in a concentration of 1200 mg/cm³ in an Eppendorf tube containing 0.9% saline solution and was subjected to micro‐CT scanning (Figure 1b). Micro‐CT scanning was performed using a Phoenix Nanotom S system (GE Sensing & Inspection Technologies GmbH, Wunstorf, Germany) at the Stellenbosch University CT facility (du Plessis et al., 2016), using a tube voltage of 60 kV and a current of 240 μA, resulting in a spatial resolution of 9.7 μm. The acquired micro‐CT datasets were subsequently processed and analysed with VGStudio Max 3.2 (Volume Graphics GmbH, Germany). First, a non‐local means filter was employed to remove the noise of raw micro‐CT images. Second, an advanced surface determination was applied, followed by the selection of region of interest (ROI) from the surface. Third, the ROI was split into components to measure the mean grey values of the calibration phantom and the osteoderm, respectively. A cylindrical volume in the saline solution was also selected to calculate background density. The grey value analysis tool was used to obtain a mean value from the distribution of measured counts (Figure 1c). Lastly, we calculated the volume of each osteoderm and used a wall thickness analysis to quantify differences amongst thicknesses of osteoderm tissues (see Broeckhoven, du Plessis, & Hui, 2017; Figure 1d). MS Office Excel 2016 software was used for the calculation of the regression equations of the osteoderm grey levels with their densities. Analyses of covariance (ANCOVA) were conducted using RStudio 2022.02.0 to test for differences in osteoderm mineral density and morphometrics between the sexes and seasons (i.e. month of collection). Previous research (Broeckhoven, El Adak, et al., 2018; Laver et al., 2020) and preliminary analyses demonstrated that osteoderm mineral density, volume and thickness correlated strongly with body size (r = 0.56–0.77; all p < 0.001); hence, the lizard's body size (snout‐vent length) was introduced as a continuous covariate in the analyses.

2.2. Histochemistry

Osteoderms were extracted from non‐preserved, defrosted dorsal skins of five O. cataphractus individuals (4 females and 1 male) collected by Broeckhoven et al. (2015) under a Northern Cape Province permit (no. FAUNA 1541/2015). The samples were fixed in Histofix® (Roth, Karlsruhe, Germany) for 72 hours, decalcified in Osteosoft® (Merck, Darmstadt, Germany) for 10 days, dehydrated and embedded into paraffin. 5 μm sections were collected sequentially on uncoated microscope slides (i.e. 4 sections per slide—12 slides per sample). Sections were then deparaffinized, rinsed with double distilled water and immersed in a TRAP incubation medium (0.1 M sodium acetate and 50 mM L[+] tartaric acid in ddH2O, pH 4.7–5.0) for 30 min at 37°C. After washing, the sections were counterstained for 30 s with 0.02% Fast Green, rinsed in distilled water, dehydrated rapidly in a graded series of ethanol and cleared with xylene. Mouse bones were included as positive control tissue.

3. RESULTS

Analyses of covariance revealed that the osteoderm mineral density was significantly higher in females than in males of similar body size (Table 1; Figure 2). Osteoderms had similar sizes in both sexes, but those of females were characterized by smaller vascular cavities (Table 1; Figure 2). None of the osteoderm morphometrics differed between the seasons and the interaction effect between sex and season was also not statistically significant (Table 1).

TABLE 1.

Results from ANCOVA analyses examining the differences in osteoderm mineral density, volume and wall thickness between the sexes and the months during which the specimens were collected. Body size was included as a covariate in the analyses

	Mineral density		Volume		Wall thickness
	F‐value	p‐value	F‐value	p‐value	F‐value	p‐value
Sex	F 1,50 = 35.47	<0.001	F 1,50 = 2.94	0.09	F 1,50 = 6.42	0.01
Month	F 2,50 = 0.82	0.45	F 2,50 = 0.15	0.86	F 2,50 = 0.66	0.52
Body size	F 1,50 = 64.69	<0.001	F 1,50 = 86.04	<0.001	F 1,50 = 51.21	<0.001
Sex * Month	F 2,48 = 0.53	0.59	F 2,48 = 0.25	0.78	F 2,48 = 1.00	0.38

Statistically significant differences are indicated in bold.

FIGURE 2.

Relationship between osteoderm mineral density (a), wall thickness (b), volume (c) and body size in male and female Ouroborus cataphractus. The outer bands represent the simultaneous 95% confidence intervals, whilst the inner bands represent the pointwise 95% confidence intervals

In the control mouse tissue, the histochemical study demonstrated positive staining of multinucleated osteoclasts in the ossification zone of the epiphyseal plates. In O. cataphractus, mononucleated TRAP‐positive cells were detected in the vascular cavities of two osteoderm samples belonging to female individuals (Figure 3). These TRAP‐positive cells, however, were only present in very low numbers.

FIGURE 3.

(a) Histological section of dorsal osteoderm in the transversal plane with details of Sharpey's fibres, vascular canal showing large adipocytes and bone microarchitecture. Stained with Alcian blue hematoxylin‐Orange G (data taken from Broeckhoven et al., 2015). (b) TRAP‐positive cells, identified as mononucleated osteoclasts, in the vascular canals of the osteoderms

4. DISCUSSION

During pregnancy or egg production, females must provide sufficient calcium to ensure adequate bone development of their offspring and/or for the formation of eggshells, and several strategies have evolved to this end. From an evolutionary point of view, this would be more important to terrestrial vertebrates than those inhabiting calcium‐rich oceans where there is no need for an internal calcium reservoir and having a solid bone mass is not a constraint (Bouillon & Suda, 2014). In mammals, mobilization of calcium from the maternal skeleton may occur during pregnancy and lactation (e.g. Kalkwarf & Specker, 2002). In contrast, theropods evolved medullary and hollow bone that might develop in reproducing females and provide a mineral reservoir required for shell calcification (Dacke et al., 1993; Prondvai, 2017). Compared to cortical bone, medullary bone can be metabolised at least 10–15 times faster due to the large surface area, extensive vascularization and a higher number of osteoclasts (Dacke et al., 1993). Whilst the influence of reproduction on bone density has been extensively investigated in mammals and birds, very little is known about strategies used by reptiles. In musk turtles (Sternothaerus odoratus) and Nile monitors (Varanus niloticus), the production of eggs relates to seasonal reductions in skeletal bone density (Edgren, 1960; de Buffrenil & Francillon‐Vieillot, 2001). Several other lizards are characterized by the presence of extracranial endolymphatic sacs, which are larger in females than in males, and are depleted during egg development (Lamb et al., 2017; Laver et al., 2020). Based on previous observations (e.g. Dacke et al., 2015) and our own study, we propose that osteoderms provide an additional source of calcium that can be resorbed during reproduction. The results of our study show that the mineral density of osteoderms is higher in females than in males taking into account differences in body size. This higher density of osteoderms in females can be explained by two non‐exclusive causes: firstly, a minimum mineral density level might be necessary for female O. cataphractus to reproduce. Actis et al. (2017), for example, showed that female pygmy armadillos (Zaedyus pichiy) that reproduced had higher bone mineral densities than those that did not successfully breed. Secondly, resorption of minerals from the osteoderms would reduce the mechanical strength of osteoderms, leaving female individuals more vulnerable to predation. Broeckhoven et al. (2015) showed that the force required to puncture the osteoderms O. cataphractus is slightly higher than the actual bite force of their main predators, mongooses. By having denser osteoderms, minerals can be resorbed without threatening the integrity of the osteoderms. In contrast to our expectations, we did not find a difference in osteoderm mineral density between spring and dry seasons in females. There are several explanations for this, which require further investigation. The most plausible explanation is that minerals might only be resorbed from the osteoderms when the calcium content of the diet is below a critical point, for example, during periods of an extreme or prolonged drought that are often experienced in semiarid environments, as proposed by Curry Rogers et al. (2011). Precipitation data from the time period in which the individuals used in this study were collected (1996–1997) confirm that an exceptionally high amount of precipitation was received during that particular dry season (Figure S1), hence there might not have been a need for mineral resorption. Alternatively, the allocation of calcium to the embryo might only occur during the final stages of pregnancy. Unfortunately, the destructive nature of our current method and low sample sizes per season did not allow us to assess individual changes in osteoderm density throughout the pregnancy period. Future studies should investigate this in more detail using, for example, in vivo micro‐CT scanning (Broeckhoven, du Plessis, le Roux, et al., 2017).

Despite differences in osteoderm mineral density within and between sexes, the mechanism for osteoderm resorption remains poorly understood. At least two mechanisms have been put forward: (1) bone resorption performed by osteoclasts or (2) lowering of extracellular pH to aid the dissolution of bone. Dacke et al. (2015) observed secondary osteons, indicative of bone remodelling, but did not find any osteoclasts. Similarly, Curry Rogers et al. (2011) demonstrate resorptive lacunae in the inner bone margin of the osteoderm cavity, suggestive of osteoclastic resorption. The results of our histochemical study show TRAP‐positive cells, which given their position in the vascular canals and the protocol used, are likely to be mononucleated osteoclasts. It remains to be determined, however, whether these mononucleated osteoclasts are the active resorbing cells, as seen in teleosts or whether they are precursors to multinucleated cells (Witten & Huysseune, 2009).

The evidence presented here puts forward the mineral storage hypothesis as a plausible explanation for osteoderm expression, at least in reptiles (but see Actis et al., 2017 for armadillos). Broeckhoven, de Kock, and Mouton (2017); Broeckhoven, de Kock, and Hui (2018) recently demonstrated that in some species of cordyline lizards, males have more or larger osteoderms than females. Osteoderms in males might primarily play a role during male–male contests, whereas osteoderm expression in females might be driven by reproductive needs, both processes governing the evolution of the same structure in a particular taxonomic group. Future studies should investigate whether these trends can be generalized to other osteoderm‐bearing taxa or whether these phenomena are highly context‐specific (e.g. linked to environments or reproductive life history).

AUTHOR CONTRIBUTION

Chris Broeckhoven conceived the idea, analysed and interpreted the data and wrote the manuscript with Anton du Plessis. Anton du Plessis carried out the micro‐CT scanning and analysis.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

Figure S1

Broeckhoven, C. & du Plessis, A. (2022) Osteoderms as calcium reservoirs: Insights from the lizard Ouroborus cataphractus . Journal of Anatomy, 241, 635–640. Available from: 10.1111/joa.13683

DATA AVAILABILITY STATEMENT

The data underlying the results of this study are available on request from the corresponding author.