The combined cartilage growth – calcification patterns in the wing‐fins of Rajidae (Chondrichthyes): A divergent model from endochondral ossification of tetrapods

Abstract The relationship between cartilage growth – mineralization patterns were studied in adult Rajidae with X‐ray morphology/morphometry, undecalcified resin‐embedded, heat‐deproteinated histology and scanning electron microscopy. Morphometry of the wing‐fins, nine central rays of the youngest and oldest specimens documented a significant decrement of radials mean length between inner, middle and outer zones, but without a regular progression along the ray. This suggests that single radial length growth is regulated in such a way to align inter‐radial joints parallel to the wing metapterygia curvature. Trans‐illumination and heat‐deproteination techniques showed polygonal and cylindrical morphotypes of tesserae, whose aligned pattern ranged from mono‐columnar, bi‐columnar, and multi‐columnar up to the crustal‐like layout. Histology of tessellated cartilage allowed to identify of zones of the incoming mineral deposition characterized by enhanced duplication rate of chondrocytes with the formation of isogenic groups, whose morphology and topography suggested a relationship with the impending formation of the radials calcified column. The morphotype and layout of radial tesserae were related to mechanical demands (stiffening) and the size/mass of the radial cartilage body. The cartilage calcification pattern of the batoids model shares several morphological features with tetrapods' endochondral ossification, that is, (chondrocytes' high duplication rate, alignment in rows, increased volume of chondrocyte lacunae), but without the typical geometry of the metaphyseal growth plates. Research Highlights 1. The wing‐fins system consists of stiff radials, mobile inter‐radial joints and a flat inter‐radial membrane adapted to the mechanical demand of wing wave movement. 2. Growth occurs by forming a mixed calcified‐uncalcified cartilage texture, developing intrinsic tensional stresses documented by morphoanatomical data.


| INTRODUCTION
The cartilage plays a fundamental role in the skeletal development of vertebrates; in mammalians, this is then replaced by bone through endochondral ossification (Hall, 2005). Characteristic transformations (such as "hypertrophy") occur in chondrocytes whose surrounding matrix is undergoing to mineral deposition, as widely investigated in the literature on bio-mineralization (Bonucci, 1971;Pazzaglia et al., 2016Pazzaglia et al., , 2018. In endochondral ossification, cartilage is a temporary tissue before it is replaced by bone after the foetal and the metaphyseal growth phases, remaining later only on the surface of the joints during the whole lifetime. Otherwise, from tetrapods and birds, the cartilage of Chondrichthyes forms the definitive skeleton segments, which do not undergo secondary substitution (remodeling). The endoskeleton and fins cartilage of the latter has drawn deep interest in their phylogenesis (Compagno, 1973(Compagno, , 1977Maisey, 1984) because it represents a distinctive and unique model of cartilage growth associated with mineral deposition diverging from the pattern observed in most living vertebrates and the fish class of Actinopterygii. Batoids developed wing-like structures by fusing the pectoral fins used to swim. The swinging movement of the waves across them during steady swimming gives this movement power (Heine, 1992;Klausewitz, 1964;Rosenberger, 2001). The wing-fins skeleton consists of a series of parallel rays formed by serially aligned radials linked by amphiarthroses with a limited range of movement and restrained laterally by a thick inter-radial membrane Schaefer & Summers, 2005). The radials of the first median line articulate with the pterygia (propterygia, mesopterygia, neopterygia, and metapterygia). Other, shorter rays articulate with the pelvic girdle basipterygia (Compagno, 1999). Overall, the calcified cartilage segments of wings form a functional and mechanical flat system of variable stiffness/ flexibility that can translate the contraction of the powerful dorsal and ventral wing muscles into the undulatory-oscillatory locomotion of these fish species (Rosenberger, 2001).
The Rajidae skeletal morphology offered a cartilage model stiffened by mineral deposition and adapted to the mechanical demand of their motion. Basic units, that is, tesserae, form this composite tissue of calcified and uncalcified matrix.
This study has been carried out in the species Raja clavata with the following purposes: 1) to analyze with X-rays and transillumination morphometry the wing-fins radials growth pattern; 2) to compare the calcified layout of the wing-fins, the pelvic girdle basipterigium fins radials and the tessellated endoskeleton segments; 3) we used a combination of light, scanning electron microscopy (SEM) and heat deproteination to study the mineral deposition progression of cartilage scaffold.

| Experimental animals
The study was carried out on five adult specimens of Raja clavata Linnaeus, 1758, family Rajidae, class Chondrichthyes (Serena et al., 2010) captured in the Western Mediterranean and purchased from commercial sources (Mercato Ittico di Milano and Mercato Ittico di Sassari, Italy). Animal welfare laws, guidelines and policies were not applicable.
The fresh fishes were weighed and measured fresh before dissection (Table 1). Wing-fins of specimens 1 and 5 were used for morphometry, while those of specimens 2, 3, and 4 were dissected and processed for histology, heat deproteination and SEM study.

| X-Rays morphometry and trans-illumination imaging
X-rays of each specimen's right and left-wing fins and pelvic girdle basipterygia were taken in the dorsoventral projection with Siemens equipment (Siemens AG, Munich, Germany). The number of rays and radials along each ray was determined in wing-fins X-rays, while for the pelvic basipterygia, only the number of rays was determined. In the right wing-fin of the youngest and oldest specimen (weight 110 and 1520 g, respectively), the rays of the central sector, whose first radials line articulated with the mesopterygium and neopterygium (Compagno, 1999), were divided with two parallel cuts from the basal pterygia to the outer border of the wing. The skin and muscles on the dorsal and ventral surface were gently dissected with a scalpel (leaving only the inter-radials membrane and skin layer of ≈2.5 cm on the wing outermost border) to avoid damaging the most external, very thin radials. High-definition radiograph images of the calcified segments were obtained with a NewTom CT (New Tom, Verona, Italy) for radials length measurements. The same calcified segments (plunged in Petri capsules in a glycerol solution bath) provided the trans-illumination images of radials at magnifications from 1.25 to 40x with a stereo microscope Olympus SXZ 7 and an Olympus BX 51 (Olympus Ltd, Tokyo, Japan). The selected central sector included nine rays, corresponding to the longest rays of the fin. The sketchy mesopterygium T A B L E 1 Sex, weight, disk-width, total-length and ratio disk-width/totallength of the 5 studied specimens of Raja clavata (Chondrichthyes). and neopterygium 1st segment of radial (characterized by high variability of the distance between propterygia border-1st radial joint, that is, line 0) was excluded from statistics. Trans-illumination images of fixed and unembedded specimens (without skin and radial muscles) of radial series were obtained through observation in a large Petri capsule filled with glycerol solution. The central sectors of specimens 1 and 5 were selected for morphometry based on key morphotraits: 1) the longest, straight rays of the wing with the higher number of radials; 2) a regular line of bifurcating radials (line 9) occurring at ≈2/3rd of the whole length, with 8 radial lines interposed between line 0 and 9 (bifurcation line) and approximately more than 8 lines externally. It was not possible to assess with precision the number of the outer calcified segments due to their very-thin diameter (the ventral and dorsal skin layer with dermal denticles was left to avoid damage with dissection of the inner tissue texture); 3) the medial joints between meso/neopterygia-1st radials were not definable. Therefore the latter were excluded from measurements (line 0); 4) the 8 proximal radials lines below bifurcation was divided into two zones 1a, 1b (1 = youngest specimen) and 5b, 5b (5 = oldest specimen), corresponding respectively to radials lines 1-4 and 5-8, while the external zones above bifurcation 1c, 5c corresponded to radials lines 10-13 ( Figure 1d,e). The bifurcating radials of line nine were not measurable. However, the latter's identification accuracy was high through the criterion of a single inter-radial joint medially and two distinct joints laterally.
This experiment analyzed high-definition digitalized images with the program Cells (Soft Imaging System GmbH, Munster, Germany).
The statistical analysis was carried out with the MedCalc program (MedCalc Software Ltd, Ostend, Belgium) using the Student t-test to compare the radial mean length difference between zones 1a -1b -1c, 5a -5b -5c and 1a -5a, 1b -5b, 1c -5c. A probability of p < .05 was considered statistically significant. Two investigators obtained repeated length measurements independently (MR and GZ). The difference in the mean analysis (Bland & Altman, 2010) was applied to these data sets. The difference of each paired measurement (intraobserver, repeated after 30 days, and inter-observers) was plotted against differences of the observers whose only source of variability was the measurement error. The differences between the interobserver and intra-observer data sets had a degree of agreement >95% confidence interval for both.
Trans-illumination allowed highlighting of the "tiles" calcified texture and layout of radials and pterygia.

| Heat Deproteination
Dry specimens (1 Â 0.5 cm) of the wing-fins radials and 1 mm thick transverse sections of the pterygia lying on glass slides were subjected to heat deproteination in a muffle furnace at 400 C for 24 h, then mounted with a cover slip and observed at higher magnification (40-200x) with an Olympus BX51 microscope (Olympus Optical Co LTD, Japan). Germany), and 150-200 μm thick sections were prepared with the Exact cutting/grinding system (Exact Advanced Technology GmbH, Norderstedt, Germany) and stained with methylene blue-acid fuchsine. 3). Single, heat-deproteinated "tiles" were mounted on glass slides and observed under the light microscope.

| Scanning electron microscopy
Longitudinally sectioned surfaces of radials were dehydrated in increasing concentration of ethanol solutions, CO 2 critical point dried and secured on stabs with a bi-adhesive tape. They were coated with a thin layer of gold in a vacuum sputter-coater (Emitech) and examined with a Philips XL30 scanning electron microscope using secondary electron imaging (SEI) and backscattered electron imaging at 20 kV and 10 mm of working distance.

| X-rays
The wing-fins of the studied Rajiae were symmetric, showing an approximate triangular outline with two rounded lateral and posterior angles. The anterior wing skeleton ends with the sharpened propterygium tip, as shown by X-rays, but not evident from the fish's outer look because the skin joins the head and rostrum together with the wings.
We have counted 68 to70 rays in the wing fins of the studied specimens. This slight variation is explained by superimposition in X- ( Figure 1c). The pelvic girdle basipterygia articulate with ≈20 shorter rays than the wing fins, with no bifurcation. The first line of radials is characterized by longer and thicker segments than those following outwardly along each ray. Also, their X-ray mineralized texture appeared different from the wing-fin radials ( Figure 2b).

| X-rays morphometry
Statistical analysis of radials mean length measured on X-rays of the right wing-fin central sector (Figure 2c,d) documented a significant decrement among zones a, b, and c and between the corresponding zones 1a, 1b, 1c and 5a, 5b, 5c of the youngest and oldest specimen (Table 2), representing the somatic growth of the species Raja clavata.
However, the radials length decrement along each ray in both the target specimens 1 and 5 was not regular, suggesting that the length growth of single radials could have been set in such a way to keep the curvature of the inter-radials joints in the wing ( Figure 1b). The most F I G U R E 2 Raja clavata (Chondrichthyes) X-rays of specimens 1 & 5. (a) Specimen 5, high-resolution X-rays of the propterygium wing-fin anterior sector (after dissection of skin and muscles) showing the increasing number of radials from the anterior tip towards the central sector and the joint angles. In the outer band (with skin not dissected), the dermal denticles are superimposed on the thin, most external radials. (b) Specimen 5, X-rays of pelvic basipterygium with the 1st line of long and thick radials. The thicker compound radial forms a diarthrodial joint with the pelvic girdle. (c-d) X-rays of the 9 central sectors rays of specimens 1 and 5 with the labeled references for morphometry.
outer radial length of the ray was not assessable with X-rays due to super-imposition of skin and small dermal denticles ( Figure 2a).

| Microscopic trans-illumination morphology
The trans-illumination imaging allowed documenting the inner, mineralized texture of wing-fin radials after dissection of skin and muscles.
The typical pattern, either within or outside the bifurcation line 9, was

| Heat deproteination
Heating at 400 C burns the cartilage's organic phase, revealing the mineral component alone but leaving the segment's calcified structure intact (Pazzaglia et al., 2016). The heated samples are brittle, and the radials "tiles" units are easily separated. The residual, organic phase combustion products leave black sediments in fissures and chondrocyte lacunae of the calcified mass. Separated "tiles" showed three morphotypes: 1). Cylindrical; 2). Bifurcated (observed below the proximal and distal disk of inter-radials joints or forming intercalary connections between multi-columnar radials); 3). Polygonal (alike the

| Histology
The central column, calcified cartilage of the wing-fin radials, has a slender profile and a pattern of aligned, cylindrical tiles (Figure 6c). At the extremities, a calcified disk forms the base of the inter-radial joints, which is sustained by short branches from the central calcified T A B L E 2 Raja clavata (Chondrichthyes). Comparison of radials mean length of specimens 1 and 5 between right wing-fin zones a, b, and c in the central sector 9 rays (radial line 0 and 9 excluded from measurements).

Radial line 0
Zone

| Scanning electron microscopy
Longitudinally fractured radials (at the level of inter-radial amphiarthroses) were examined with SEM in both the secondary and backscattered modes, documenting three calcified cartilage blocks intersected by the fracture plane but leaving undetected the others

| DISCUSSION
The regular geometry of the Rajidae wing-fin radials with a high number of elements provides a suitable model for a morphometric/ morphological study of the combined cartilage -mineralization pattern of growth in the class Chondrichthyes. Endochondral ossification in tetrapods provides the stiff substratum for bone apposition, while the mineralized structure of Chondrichthyes is formed inside the cartilage anlagen, remaining the definitive mechanical axis of the segment not undergoing further remodeling. Therefore, these fishes' fully cartilaginous skeletons must develop alternative strategies (different from teleosts and tetrapods) to accomplish their mechanical function throughout the life cycle. The comparative analysis of radials length in the target specimens 1 and 5 allowed to examine the wing-find growth in relation to age, the youngest (110 g) and the oldest fish (1250 g), respectively, documenting a significant decrement of radials mean length among the inner, middle and outer zones of the wing-fin central sector. The distribution of inter-radial joints showed a regular layout of curved rows parallel to the propterygium and the metapterygium, which suggests that the radials' length growth in the whole wing was set in such a way to maintain the ordered alignment of the joints.
In the anterior and posterior wing-fins sectors, the joint alignment is also achieved by reducing the radial number in the ray. This confirms a controlled radial length growth adapted to the mechanical demand. Schaefer and Summers (2005) have reported in skates and rays "wings" a complex and phylogenetically diversified layout, including the calcification pattern of the wing-fin segments. These authors further distinguished a "catenated" or "crustal" mineralized structure of radials, and it is interesting to underscore that the wing images of  (Cohn et al., 2002;Gillis et al., 2009;Hall, 2005Hall, , 2007. This mechanism explains how the number of rays in the wing becomes defined during embryonic development through hatching (Marconi et al., 2019), combining cartilage growth-calcification throughout the fish life (Dean et al., 2009). This suggested that radial segmentation (Stern, 1990) occurred in the early developmental phase before calcification and that in the following growth period, chondrocytes proliferation and matrix calcification concurred to modulate stiffness/ elasticity of the whole wing to the oscillatory, undulatory or mixed swim mode in each batoid species (Blevins & Lauder, 2012;Di Santo et al., 2017;Heine, 1992;Klausewitz, 1964;Rosenberger, 2001).
To the best of our knowledge, the microscopic trans-illumination combined with the heat deproteination technique has not been so far applied to morphological and morphometric studies of cartilage calcification in the Rajidae wing-fins. Our results have provided the following observations: 1) microstructure of the wing cylindrical tiles like that of the endoskeletal tesserae (Maisey et al., 2020;Seidel et al., 2016Seidel et al., , 2017, the difference lying only in shape; 2) assemblage in columns of mono-, bi-, and multi-columnar morphotypes (in the latter with a gradual transition to a peripheral crustal pattern); 3) decrement of wing-fin radials length from the basal line to the wing edge; 4) bifurcated tilel type connecting bi-columnar or multi-columnar radials.
The combination of all these morphoanatomical traits displays a wing fins mechanical model with a diversified distribution of stiff and mobile segments which can be summarized as follows: 1) radials, whose stiffness is provided by the inner, calcified axes; 2) hinges with variable mobility such as the diarthrodial joints between pterygia -girdles; b) amphiarthroses between pterygia -1st line radials and inter- this occurs between the calcified inter-columnar septa, which orient the duplicating chondrocytes along the long bone axis, an action which is strengthened externally by the perichondral bone bark (Dodds, 1930;Pazzaglia et al., 2017a;2017b;2018;Pratt, 1959;Shapiro et al., 1977). This structural layout is combined with increased chondrocyte volume (hypertrophy) and mineral deposition in the inter-columnar cartilage. The latter forms the substratum for osteoblasts apposition, followed by remodeling of the mixed cartilage/ osteoid calcified tissue (Pazzaglia et al., 2019;. With the due differences, in Rajidae wing-fin radials, longitudinal rows of chondrocytes have been observed in the cylindrical tiles, and the same pattern (but with a radial layout) is also present in the polygonal tesserae of the elasmobranchs crustal calcification (Maisey et al., 2020;Seidel et al., 2016Seidel et al., , 2017. The main difference between the elasmobranch's calcifying cartilage is that the alignment of chondrocytes is less regular, and their size increase is not so much developed as that of the metaphyseal growth plate. Chondrocyte hypertrophy (or better cytoplasmic swelling) has been hypothesized to represent a mechanism of Ca and PO 4 ions concentration through the passage of water from the intercellular matrix into the cell (determined by osmotic equilibrium), resulting in calcium-phosphate deposits in the matrix (Maroudas, 1976;Maroudas & Schneiderman, 1987;Pazzaglia et al., 2018Pazzaglia et al., , 2020. In the Rajidae calcifying cartilage the chondrocyte volume increases; they appear more densely packed and are surrounded by less intercellular matrix than the not-calcifying zones . Therefore, no definitive conclusions can be drawn at present by comparison of these two calcification patterns, which certainly deserve more in-depth comparative studies. The ultrastructural and developmental features of the tessellated endoskeleton of elasmobranchs have been extensively studied using synchrotron high-resolution 3-D techniques revealing in the course of development a complex lamellar, high-mineral density layout with structures radiating outward, like spokes on a wheel, from the center of each tesserae to the contact with the neighbors (Seidel et al., 2016(Seidel et al., , 2021. Our study carried out with heat deproteination at 400 C confirms the already reported morphology of the tesserae mineral phase and further documents in the cylinder-shaped radial tiles an ions in the intercellular matrix as suggested for the growth plate cartilage (Pazzaglia et al., 2020). However, the different patterns of interstitial fluids diffusion in the two models could also have a role in the nucleation and CaPO 4 salts deposition into the collagen matrix network. The X-ray imaging and the microscopic observations of heatdeproteinated wing-fins specimens provided so far unreported data on morphology and calcification in the radials tiles showing an evident correlation with the mechanics of the wing-fins.
To characterize and discriminate the taxa of the Mediterranean Rajidae, identification keys are currently used based on external and internal morphoanatomical traits (Serena et al., 2010), but doubtful cases can be occasionally discussed. Additional diagnostic morphotraits provided by X-ray imaging and heat deproteination of the calcified skeleton could be useful to complete the classical morphological data and to solve doubtful taxonomic cases. However, how the mineral deposition in the radial columns and the crustal-like pattern in the endoskeleton segments of the pelvic basipterygia and of the longer and thicker radials remains unresolved. Assuming that the chondrocyte metabolic activity regulates the mineral deposition in the peri-cellular matrix, it should not be considered casual the changes in the chondrocyte's morphology and the organization of the pericellular matrix in the impending calcification zone.