The Canine Baculum: the Structure and Mechanical Properties of an Unusual Bone ⇑ A

Journal of Structural Biology 175 (2011) 451–456

Contents lists available at ScienceDirect

Journal of Structural Biology

journal homepage: www.elsevier.com/locate/yjsbi

The canine baculum: The structure and mechanical properties of an unusual bone ⇑ A. Sharir a, D. Israeli a, J. Milgram a, J.D. Currey b, E. Monsonego-Ornan c, R. Shahar a, a Koret School of Veterinary Medicine, The Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Israel b Department of Biology, University of York, UK c School of Biochemistry and Nutrition, The Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Israel article info abstract

Article history: The baculum is an extraskeletal bone located in the penis of a few species in several orders of mammals Received 25 March 2011 such as carnivores, insectivores, rodents, bats and primates. Received in revised form 9 June 2011 This study aims to describe the structure, architecture and mechanical properties of the canine bacu- Accepted 13 June 2011 lum. To this end canine bacula from castrated and uncastrated dogs were collected and examined by light Available online 25 June 2011 microscopy, micro-computed tomography (microCT) scanning, histological staining, and mechanical testing. Their mineral density and mechanical properties were compared with those of a typical skeletal bone Keywords: (the radius) in the same dog. Furthermore, a numerical model of a representative baculum was created Baculum and its mechanical performance analyzed using the ﬁnite element method, in order to try to elucidate its Mechanical properties Bone mineral density function. Examination of light microscopy images of transverse sections shows that the baculum consists of a typical sandwich structure, with two cortical plates separated, and joined, by loose cancellous bone. MicroCT scans reveal that the mineral density is lower in the baculum than in the radius, both in castrated as well as in uncastrated dogs, resulting in much lower stiffness. Castration was found to decrease the mineral density in both the baculum and the radius. The most likely function of the baculum of the dog is to stiffen the penis to assist intromission, and its much lower mineral density compared to that of the radius may be a mechanism designed to decrease the stiffness somewhat, and thus reduce the risk of fracture during copulation. Ó 2011 Elsevier Inc. All rights reserved.

1. Introduction used as a diagnostic taxonomic character; however its structure and material characteristics have not yet been described in detail. The success of the process of copulation in all mammals re- The baculum is a heterotopic (extra-skeletal) bone that is de- quires that the male achieves erection and high flexural stiffness rived from connective tissue. The development of the baculum of his penis. These features are achieved by the synergistic function has been investigated primarily in rats and mice. It has been shown of a high-pressure system (corpus cavernosum) and low pressure that this process consists of two stages. In the first stage mesen- system (corpus spongiosum). However, some mammals, belonging chymal condensation occurs in the penile organ; it is independent to such different orders as carnivores, insectivores, rodents, bats of estrogen, and is observed in both males and females. In the sec- and primates, also have a penile bone – the baculum (os penis) (Ro- ond stage, differentiation into chondrocytes and osteocytes occurs mer, 1970; Williams-Ashman, 1990). This bone is found in the only in males (and incidentally, also in androgen-treated females, glans tissue at the distal end of the penis, dorsal to the urethra. Murakami, 1986). Usually its proximal end abuts the distal end of the corpus cav- The function of the baculum is not well understood and its pres- ernosum. The morphology and size of the bacula in different orders ence in different orders, and only in some species within these or- and species vary greatly, so that its species specific-shape is often ders, is puzzling. Furthermore, the variability in form in the different orders is extreme. Thus in carnivores, for instance, it is present in bears and dogs, but absent in cats. In the dog it serves as a channel for the urethra, in the wolverine it is forked at the Abbreviations: microCT, micro-computed tomography; BMD, bone mineral tip and in the raccoon it is s-shaped and terminates in an enlarged density; GPa, gigapascal. condyle. ⇑ Corresponding author. Address: Laboratory of Bone Biomechanics, Koret School The most obvious biomechanical explanation for the function of of Veterinary Medicine, The Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Israel. Fax: +972 39604079. the baculum is to stiffen the penis, thus assisting intromission. Yet E-mail address: [email protected] (R. Shahar). many species copulate successfully without it. Furthermore, its

Fig.1. (a) 3-D drawing of the entire canine baculum derived with permission from Millers Anatomy of the dog (Evans, 1993). (b) 3-D reconstructions (based on microCT scan) of the entire baculum and cross-sectional sections at three different regions: base, mid-region, and tip of the bone. (c) Photograph of a cross section of the baculum at mid-length, showing the cortical and cancellous regions. small size compared to the corpus cavernosum in rats, makes this consent) from eight mature castrated dogs which died at the veter- hypothesis unlikely, at least in that species (Kelly, 2000). Several inary teaching hospital of the Koret school of veterinary medicine other roles have been suggested, such as providing added stimula- from causes unrelated to the skeletal system, and from eight ma- tion to the female, helping the male penetrate the hymen, or even a ture intact dogs which were euthanatized due to population con- means of functional specificity of genital morphology, inhibiting trol regulations at local animal shelters. The bacula were cross-species copulation (Meczynski, 1974; Patterson and Thaeler, prepared in the same way as the bacula in the first part. 1982). In any case, no single interpretation of bacular function seems satisfactory and well-supported (Baryshnikov et al., 2003). 2.2. Light microscopy The canine baculum is long and tapers in the proximodistal direction (see Fig. 1). The proximal end (base of the baculum) lo- The bacula evaluated in this study ranged in length from 83 to cated in the body of the penis just caudal to the bulbus glandis, 128 mm. Each baculum was transversely sectioned by handsaw to is quite broad. It is thicker dorso-ventrally than from side to side. obtain three 20 mm-long sections – from the proximal baculum, The distal end (tip) of the bone is small in diameter and it is ex- mid-baculum and distal baculum. One millimeter-thick slices were tended by a slightly curved fibrocartilagenous projection. The then cut from the centers of these sections by low-speed water- proximal two thirds of the bone are indented ventrally by a distinct cooled diamond saw (Buehler isometÓ, USA). These slices were groove in which the penile urethra and corpus spongiosum reside. ground and polished using a Minimet polisher (Buehler MinimetÒ, During erection, the corpus spongiosum receives the greater part of USA) by successively more refined abrasive paper (from 200 to the blood through the artery of the bulb of the penis. 4000 grit), followed by 3 and 1 lm cloth with diamond paste. This manuscript describes the results of a study of the micro- The polished surfaces were studied using reflected light micros- structure of the dog baculum, the mechanical properties of its copy (Olympus BX-51 microscope) and images were captured material, compares them to those of a typical long bone (radius), using a 12.1 megapixel-resolution dedicated camera attached to and evaluates the effects of early castration on its mineral density the microscope (Olympus DP-71). and distribution.

2.3. Histological staining 2. Methods Samples obtained from three bacula from the first (intact) group 2.1. Samples of dogs were fixed overnight in 4% paraformaldehyde (Sigma Chemical, St. Louis, MO, USA), then in phosphate buffered saline In the first part of this study, bacula were obtained from nine at 4 °C for 3 days, then subjected to 3 weeks of decalcification in mature medium-sized intact (non-castrated) male dogs weighing 0.5 M EDTA, pH 7.4. Following dehydration in graded ethanol solu- 25–35 kg, which were euthanatized due to population control reg- tions and histoclearÓ (National Diagnostics, USA), the tissue was ulations at local animal shelters. Bacula were removed by careful embedded in Paraplast (SPI supplies, USA), cut into 5 lm sections dissection, cleaned of soft tissue, wrapped in saline-soaked gauze and mounted on glass slides (Superfrost, Thermo Scientific). The and stored at À20 °C. In the second part of the study, both the embedded sections were deparaffinized in xylene, washed twice baculum and radius bones were harvested (after receiving owner with ethanol and rehydrated through a graded series of ethanol A. Sharir et al. / Journal of Structural Biology 175 (2011) 451–456 453

Fig.2. (a) Reflected light microscopy image of the cortical region of the baculum at mid-length. Note internal remodeled (secondary osteonal) region and external lamellar region. Scale bar 200 lm. (b) Polarized light microscopy image of the lamellar region, showing the orientation of the collagen fibers. (c). Transmission light microscopy of a transverse section of the mid-baculum, stained with Alcian blue stain. Note the inner, secondary osteonal bone, consisting of concentric lamellae (black arrows). An external, layered zone, with a high density of osteocytes within lacunae (black arrowheads). (d) The layered region seen in Fig. 1c, shown at larger magnification, demonstrating the canalicular network between adjacent lacunae (black arrow heads). solutions. The sections were then stained with Alcian blue stain. the intersections of a rotated test grid with the trabecular structure Alcian blue stains acid mucopolysaccharides and glycosaminogly- and calculates the fabric ellipsoid (3D ellipse) (Whitehouse, 1974; cans light blue-green, and cell nuclei black. The stained sections Harrigan and Mann, 1984)), trabecular thickness, and trabecular were visualized using a transmission light microscope (Nikon separation. Eclipse E600) and images captured by a dedicated camera (Olym- pus DP 71). 2.5. Mechanical testing of cortical bone

2.4. MicroCT scanning Cortical beams were prepared from the cortex of the dorsal aspect of the center of bacula and of the mid-diaphyseal cortex of the All bacula from the first (intact) group of dogs were scanned by radii of the 16 dogs of the second group (eight castrated dogs and a high-resolution microCT scanner (GE Healthcare). Due to their eight intact dogs). The beams were 22–26 mm long, with a cross length they had to be scanned twice, such that each baculum section of 1.5 Â 1.5 mm, and their long axis was oriented in the ax- had its proximal half scanned first, then turned over and the distal ial direction of the bones. The beams were cut with the water- half scanned. Scans were performed on wet specimens at 14 lm cooled slow-speed diamond blade saw described above. All beams nominal resolution and the results were analyzed by the appropri- were first scanned by microCT, and no significant imperfections ate software (Microview 2.0). such as large voids or cracks were found in any of them. The effect The baculum consisted of cortical bone and cancellous bone re- of beam damage on the results of mechanical testing was ignored gions (see Fig. 1b and c). Three sub-regions were selected for anal- as it was considered minor. ysis of the cortical region (at 20%, 50% and 80% of the length of the The beams were then tested in four-point bending. The beams bone). In each region five cylindrical regions of interest (ROI), each were placed in a materials testing machine (Instron 3345) within 0.7 mm long and 0.7 mm in diameter, were selected and local mea- a testing chamber which contained two stainless steel base sup- surements of bone mineral density (BMD) were made. For cancel- ports which were 12 mm apart. The moving anvil of the system lous bone regions measurements were made in the proximal and was fitted with a two-prong device which was centered between distal halves of each bone (the 50% sub-region did not contain suf- the bottom supports and the distance between the prongs was ficient cancellous bone for analysis), and the following properties 3 mm. The four-point bending method ensures that the section of were measured: bone volume fraction (BVF), degree of the anisot- the beam between the two upper prongs is loaded by a bending ropy of the trabeculae (based on the mean intercept length (MIL) moment of constant magnitude. All parts of the testing system method which calculates the structural anisotropy by measuring were made of stainless steel. The parts contacting the sample 454 A. Sharir et al. / Journal of Structural Biology 175 (2011) 451–456

(bottom supports and upper prongs) had rounded profiles (1 mm Table 3 diameter) to reduce stress concentration. The specimens were pre- Results of four-point bending tests performed on cortical beams obtained from eight radii from intact dogs, seven radii from castrated dogs and seven bacula of castrated cisely cut, so that uniform beam depth was achieved and the upper dogs are presented. Due to a technical error, beams from bacula of castrated dogs prongs contacted them concurrently. The upper prongs were were not tested. brought to contact with the beams, the chamber was filled with Young’s modulus (GPa) physiologic saline solution at room temperature, and bending tests were conducted under displacement control at a rate of 1 mm/min. Radius intact Radius castrated Baculum intact Baculum castrated Load and displacement were measured at 100 Hz until failure (in 10.8 ± 1.47 10.3 ± 0.53 5.3 ± 1.3 – the bacula failure did not occur by fracture due to the extreme duc- tility of the samples, and testing was ended when the post-yield deformation was so extreme that the resistance to further move- value used frequently for bone material (see for instance Müller ment of the prongs started to rise due to contact of the beam with and Rüegsegger, 1995; Van Rietbergen et al., 1998). The model con- the base). The resulting load–displacement curves were used to sisted of 411,679 tetrahedral elements and 100,794 nodes. Several calculate the slope of the linear portion, which was used to esti- nodes of the base of the baculum were fully fixed (movement not mate the Young’s modulus of the beam material (Draper and Good- allowed in any direction) to restrain the model. The model was ship, 2003). It should be noted that since shear deformation was loaded in two alternative ways: by a vertical bending force or a lat- not taken into consideration, estimated values of the Young’s mod- eral bending force, both applied to the tip of the baculum. FE anal- ulus were slightly underestimated. Due to a technical error, beams ysis of the baculum model yielded 3-D displacements, and all prepared from the bacula of the castrated dogs were not mechan- stress and strain component distributions within the entire bone ically tested. for both ways of loading.

2.6. Radius/ulna mineral density 3. Results

Small cubes (5 Â 5 Â 5 mm) were removed from the same loca- The external morphology of the canine baculum is shown in tions in the radii (mid diaphysis and medial aspect) and bacula Fig. 1a and b. (middle of bone and dorsal aspect) collected in the second part The overall length of the bacula varied between 83 and 128 mm of the study (see above). All cubes were scanned by microCT at (n = 9; mean: 104.4 mm, standard deviation: 14.7 mm), and the high (8 lm) nominal resolution, and their BMD determined. greatest external dimension varied between 8.7 and 13.6 mm (n = 9; mean: 10.9 mm, standard deviation: 1.5 mm). 2.7. Finite element model The macro-architecture of the baculum consists of two thick external cortical layers, and an inner trabecular region (see In an attempt to understand the mechanical function of the bac- Fig. 1c). The thickness of the cortical shell varied between 0.7 ulum, and explain the division between intensely remodeled areas and 2 mm. and the less remodeled zones in the bone, the stress and strain distribution within the baculum when loaded were evaluated using a 3.1. Light microscopy finite element model of the baculum. The entire set of 2-D slices (bitmap files) of the micro-CT scans of one baculum were seg- Light microscopy images of transverse sections of the baculum mented and reconstructed into a 3-D model (cortical bone only, reveal that the cortical shell of this bone has two zones. The exter- as cancellous bone was considered to contribute little to the nal zone consists of lamellar bone structure (see Fig. 2a and b). It mechanical behavior of the entire bone), then meshed (4-node tet- extends to 15–25% of the complete cortical thickness, and is dense rahedral elements) using AmiraÒ 4.1.2 (Visage Imaging Inc., Carls- and poorly supplied by blood vessels. The internal region (75–85% bad, CA, USA). Next, the model was exported to PatranÒ finite of the cortex) was extensively remodeled with a dense population element solver software (Patran 2008r1, MSC Software Corpora- of secondary osteons. tion, Santa Ana, CA, USA). The model consisted of one material (cortical bone) assumed to be homogeneous, linearly elastic and 3.2. Histology isotropic. Young’s modulus was assigned according to values obtained in this study for beams of cortical bone from bacula of intact The stained cross-sections of the baculum show distinctly the dogs (see Table 3): (5.3 GPa), while Poisson’s ratio was set at 0.3, a inner, secondary osteonal bone, with concentric lamellae, and osteocytes residing in their lacunae, and the intricate canalicular Table 1 network connecting lacunae (see Fig. 2c and d). Mean and standard deviation of the average cortical BMD in different regions in nine bacula from the first (intact) group of dogs. 3.3. MicroCT scans Cortical BMD (mg/cc) Base Mid Tip Nine bacula obtained from castrated dogs were scanned, and values of various structural properties were determined for the 696.2 ± 57.6 746.2 ± 38.1 758.0 ± 46.0 cortical region and the cancellous region of these bones. The results obtained by analysis of these scans are presented in Table 1 (for the cortical region) and Table 2 (for the cancellous region). Note that Table 2 bone mineral density values are not provided for cancellous bone Mean and standard deviation of trabecular thickness, trabecular separation and bone volume fraction in eight bacula analyzed. since they are unreliable due to partial volume effects.

Trabecular thickness Trabecular spacing Bone volume fraction 3.4. Mechanical testing (lm) (lm) (%)

45.1 ± 9.5 44.3 ± 13.5 30.8 ± 6.6 Young’s modulus of cortical bone of the bacula and radii, based 44 ± 11.4 41 ± 13.0 on the four-point bending tests are provided in Table 3. Fig. 3 is a A. Sharir et al. / Journal of Structural Biology 175 (2011) 451–456 455

Fig.3. Typical mechanical test results (four-point bending of cortical beams) of specimens of the same size from a radius and baculum of the same dog, showing the lower stiffness and greater toughness of the baculum compared to the radius.

Table 4 Mean and standard deviation of the BMD of cubes obtained from cortex of the Fig.5. Finite element results: Transverse cross sections of the baculum, showing baculum and radius in eight intact dogs and eight castrated dogs. maximal principal stresses due to a vertical force and lateral force. Note arbitrary non-linear scale, used to demonstrate peaks, without obscuring distribution of Intact Castrated lower stresses. Radius Baculum Radius Baculum

963.3 ± 40.0 889.7 ± 64.5 895.1 ± 61.4 808.9 ± 22.7 reported in Table 4 for intact dogs are higher than those reported in Table 1. The difference is caused because in Table 1 results are plot of typical results from a four-point bending experiment on a based on an average BMD in a large section of cortex, which in- beam obtained from a baculum and a same-sized beam from the cludes some voids. In Table 4, the results were based on scanning cortex of a radius of the same intact dog. of bacula from eight different dogs, and analyzed in small cubes It should be noted that the although the beams consisted removed from the mid diaphysis. mostly of remodeled bone, they probably contained some primary lamellar bone as well, since the thickness of the cortical plate did 3.6. Finite element analysis not allow the formation of beams of sufficient thickness consisting entirely of a single bone type. Therefore the calculated Young’s The exact loading regime affecting the baculum is unknown. modulus represents an effective value based on a mixture of both The most likely scenario would be some form of cantilever beam bone types, however this mixing is likely to have a minor effect bending, with the baculum anchored at its base, and loaded at its only. tip. The two alternative extremes of this scenario are a vertical bending force and a lateral bending force. The maximal principal stresses within a transverse slice in the middle of the baculum 3.5. Bone mineral density are shown in Fig. 5 for a vertical force and a lateral force, respectively. The mineral density of 5 Â 5 Â 5 mm cubes cut from the radii and bacula of eight intact and eight castrated dogs were determined by microCT scanning, and their BMD values are given in 4. Discussion Table 4 and Fig. 4. It should be noted that the values of BMD The baculum is a unique bone, whose function is still unclear. In particular, the reason for its presence in some species and lack in others is unknown. An explanation for its disappearance in humans was offered by Dawkins in ‘The selfish gene’ (2006, p. 309),who suggested that the disappearance of the os penis may be the result of selection pressure from females, who use the ability of the male to reach a full and stiff erection as a sign of health and the ability to cope with stress. The presence of a bone in the penis would inter- fere with this ‘diagnostic’ measure. Our study demonstrated quite clearly that the baculum pos- sesses the structural features typical of the bones of large mammals. These include cortical and cancellous regions, secondary osteons within the cortical shell, fat and blood-forming cells in the medullary region, etc. However its spatial arrangement, mineral distribution and mechanical properties are, as far as we know, unique. We show that the mineral density of the baculum is consis- Fig.4. Mean BMD values in the radius and baculum of castrated (eight samples tently much lower than that of the radius in the same individual. each, left) and intact (eight samples each, right) dogs, belonging to the second group The radius is a weight-supporting bone and is subjected to muscle of dogs (error bars are standard deviations, ordinate starts at 750 mg/cc). Note forces, while the baculum is not. The relative decrease in BMD is higher mineral density values in the bones obtained from intact dogs than castrated considerable (mean of 8%), and must require tight control, as bone dogs for each bone type, and also higher mineral density values in radii than in bacula, both between intact dogs as well as between castrated dogs. These mineralization in most bones is usually maintained within narrow differences are all statistically significant. bounds (Currey et al., 2009). A prominent example of a bone that 456 A. Sharir et al. / Journal of Structural Biology 175 (2011) 451–456 exhibits much lower mineral density than that of most other bones of remodeling in the rat bacula are not correlated with the regions is the antler of deer. The reported low mineral content of antler that would be under the greatest amount of stress....’’ (Kelly, (59% ash) compared to that of long bones (67% in the bovine femur) 2000). An alternative explanation may be that contrary to com- is thought to be associated with its function – the need of deer to monly held belief, remodeling may not be strain- or damage-in- increase the toughness of their antlers which are used primarily for duced at all, and thus the remodeled areas are determined by the inter-male fighting (Currey et al., 2009). The functional demands age of the bone or an as-of-yet undetermined stimulus (Currey, made during intromission are such that an extremely stiff baculum 2002, p. 377). may hinder successful coitus, and be exposed to very high stresses and possibly fracture. The reason for the low mineral density of the 5. Conclusion baculum could therefore be similar to that making the antler less stiff than other skeletal bones – it may be that the penile bone We have described the detailed structure of the canine bacu- should be more flexible and tough than other skeletal elements. In- lum, its mineral density distribution and its mechanical properties. deed, mechanical test results reported here demonstrate that the The bone material has a lower mineral density, and is less stiff than stiffness of the cortical bone of the baculum is much lower than skeletal bone. Furthermore, its mineral density, as well as that of that of the radius in the same dog, and values reported for most the radius, is reduced by castration. bones (5.5 GPa vs. 10–18 GPa, see for instance Currey et al., 2009). It should be noted that as the process of mineralization is References highly regulated, some mechanism must prevent further mineralization of the baculum. Baryshnikov, G.F., Binida-Edmonds, O.R.P., Abramov, A., 2003. Morphological Also of interest is the finding that the sexual status (castrated/ variability and evolution of the baculum (os penis) in mustelidae (carnivora). J. Mammal. 84, 673–690. intact), significantly affects the BMD of the bacula and of the ra- Currey, J.D., 2002. Bones: Structure and Mechanics. Princeton University Press, dius. It has been stated that in man rapid bone loss occurs after Oxford. castration (Tuck and Francis, 2009). The suggested mechanism is Currey, J.D., Landete-Castillejos, T., Estevez, J., Ceacero, F., Olguin, A., et al., 2009. The mechanical properties of red deer antler bone when used in fighting. J. Exp. Biol. net loss of bone material caused by estrogen deficiency since the 212, 3985–3993. effects of testosterone on the male skeleton are mediated in part Dawkins, R., 2006. The Selfish Gene, 30th Anniversary Edition. Oxford University by aromatization to estradiol. This mechanism may be responsible Press, Oxford. Draper, E.R.C., Goodship, E.A., 2003. A novel technique for four-point bending of also for the decreased BMD of the radii and bacula seen in this small bone samples with semi-automatic analysis. J. Biomech. 36, 1497–1502. study. Evans, H.E., 1993. Miller’s Anatomy of the Dog, third ed. W.B. Saunders, The observation that within the cortex the internal region Philadelphia. Harrigan, T.P., Mann, R.W., 1984. Characterization of microstructural anisotropy in (which abuts the cancellous region) was intensely remodeled orthotropic materials using a 2nd rank tensor. J. Mater. Sci. 19, 761–767. while the outer region was not was somewhat unexpected. If one Kelly, D.A., 2000. Anatomy of the baculum–corpus cavernosum interface in the accepts that the remodeling process occurs as a response to dam- Norway rat (Rattus norvegicus), and implications for force transfer during age caused by high strains (or to the high strains themselves), this copulation. J. Morph. 244, 69–77. Meczynski, S., 1974. Morphohistological structure of the female genital organs in finding suggests that it is the inner region that is more intensely sousliks. Acta Therio. 19, 91–106. loaded. However according to beam theory, if the main load ap- Müller, R., Rüegsegger, P., 1995. Three-dimensional finite element modelling of plied to the baculum resembles cantilever bending the resulting non-invasively assessed trabecular bone structures. Med. Eng. Phys. 17, 126– 133. bending moments cause maximal strains at the outer surfaces of Murakami, R., 1986. Development of the os penis in genital tubercles cultured the beam, where the distance from the neutral axis is maximal. beneath the renal capsule of adult rats. J. Anat. 149, 11–20. (Cantilever bending causes also shear stresses, which are maximal Patterson, B.D., Thaeler, C.S., 1982. The mammalian baculum: hypotheses on the nature of baculum variability. J. Mammal. 63, 1–15. along the neutral axis, but in a typical beam these are negligible in Romer, A.S., 1970. The Vertebrate Body. W.B. Saunders, Philadelphia. comparison to the bending moments.) However finite element Tuck, S.P., Francis, R.M., 2009. Testosterone, bone and osteoporosis. Front. Horm. analysis shows that the maximal principal stress distribution is Res. 37, 123–132. van Rietbergen, B., Odgaard, A., Kabel, J., Huiskes, R., 1998. Relationships between irregular, such that in some regions higher values exist in the inner bone morphology and bone elastic properties can be accurately quantified using region of the cortical plate (see Fig. 5b). This unexpected pattern is high resolution computer reconstructions. J. Orthop. Res. 16, 23–28. likely the result of the complex and highly irregular shape of the Whitehouse, W.J.J., 1974. The quantitative morphology of anisotropic trabecular bone. J. Micros. 101, 153–168. baculum. It should be noted that similar findings (remodeling in Williams-Ashman, H.G., 1990. Enigmatic features of penile development and the internal regions of the bone) were also found in the baculum functions. Perspect. Biol. Med. 33, 335–374. of rats (Kelly, 2000). The authors commented that ‘‘....the zones