Cumulative effects of lifelong systemic excess growth hormone on postcranial skeletal morphology in adult mice

Abstract

Many developmental and metabolic effects of growth hormone (GH) on vertebrate life history traits have been widely studied in biological and biomedical contexts. The scope of alterations in GH/GH receptor (GHR) interactions has generally focused on molecular, cellular, histological, and physiological framing, leaving a gap in understanding of how the cumulative organism‐ and lifespan‐scale manifestations of morphological changes attributable to GH/GHR perturbations may both be influenced by and also influence smaller‐scale study results. The pilot study conducted herein used micro‐computed tomography (μCT) to survey and characterize the axial and appendicular skeletons of adult male and female bGH (transgenic overexpression of bovine GH) mice and to compare them with those of age‐ and sex‐matched wild‐type (WT) controls. Male and female bGH mice in our sample were larger (by linear measurements of skeletal elements) and leaner but not heavier than their WT counterparts. bGH mice exhibit thoracic kyphosis and radiographically detectable incipient sacralization of the last lumbar vertebra, as well as robust and altered muscle attachment sites in both girdle and long bones, re‐orientation of the acetabulum, and dysmorphology of the femur at both hip and knee joints. bGH mice generally lack clear, radiologically determinable differentiation of long bone growth plates and bear larger and differently proportioned sesamoids at the elbow and knee. They are also preferentially subject to the accumulation of inferred heterotopic calcification (IHC) and other radiodense soft tissue (RST) around joints and entheses. Female bGH mice exhibit variable and aberrant morphology in the humerus, innominate, and femur not seen in other groups, and show the most size and shape variation within the four genotype × sex groupings. Our survey illustrates widespread musculoskeletal impacts of excessive GH into adulthood in the model. These data provide an initial whole‐skeleton framework for further efforts characterizing molecular‐, cellular‐, and tissue‐scale alterations ultimately influencing the bGH mouse model skeletal phenotype. This pilot work serves to contextualize future preclinical studies as well as broader investigations into how GH and its metabolic cascade may affect vertebrate morphological and histological development, shape and size disparity, and dimorphism.

Our μCT‐based pilot study reveals size and shape disparity in the adult postcranial skeleton of growth‐hormone model (bGH) mice relative to wild‐type mice. bGH mice have larger and more sexually dimorphic bones, with a systemic radiodensity increase in bony sesamoids and associated, but typically uncalcified, soft tissues. Taken together these observations underscore a need for further histological and transcriptomic characterization of GH overexpression on bone and associated tissues.

1. INTRODUCTION

Growth hormone (GH) plays myriad roles in an organism, including the stimulation of growth and development of bone, cartilage, and muscle tissues by enhancing protein synthesis, collagen production, and fat breakdown via oxidative lipolysis (Kopchick et al., 2022; Olarescu et al., 2015). Its action is most prevalent prior to maturity but continues throughout adulthood. The musculoskeletal system is a major target for both GH's physiological functions when it is within normal range and pathological manifestations when dysregulated. GH exerts profound effects on skeletal growth and remodeling through direct and indirect actions on bone and cartilage (Dixit et al., 2021; Liu et al., 2017; Shim, 2015). GH and its downstream product, insulin‐like growth factor 1 (IGF‐1), have been shown to have both individual and combined effects on postnatal skeletal growth (Lupu et al., 2001; Mazziotti et al., 2022). During development, GH and IGF‐1 together modulate bone growth, skeletal maturation, and bone mass acquisition (Baroncelli et al., 2003; Monson et al., 2002). In adulthood, GH continues to play a role in the maintenance of bone mass. In humans, both over‐ and under‐production of GH during ontogeny have predictable and well‐characterized skeletal phenotypes. For example, untreated patients with acromegaly, a hormonal disorder in which there is an excessive amount of GH production due to a GH secreting pituitary adenoma, can in untreated patients result in an enlarged skeleton (Reid et al., 2015) as well as knee joint pain and degeneration (Claessen et al., 2017) and heterotopic ossification of ligaments associated with the vertebral column (Hoshino et al., 2022). It has also been reported that patients with acromegaly have increased risk of osteoporotic vertebral fractures (Claessen et al., 2013), possibly due to the effect of GH/IGF‐1 on bone remodeling and turnover, and there are marked sexual differences in IGF‐1 serum levels despite similar elevated levels of GH (Colao et al., 2002; Parkinson et al., 2002). In contrast, Laron Syndrome, a rare genetic disorder caused by inactivating mutations in the growth hormone receptor (GHR) gene, results in GH resistance or insensitivity, leading to slow prepubertal growth rates and sexual development, extremely short stature, obesity, and smaller and morphologically distinctive vertebral and appendicular elements with delayed epiphyseal closure (Guevara‐Aguirre et al., 2021; Kornreich et al., 2008; Laron, 2015; Laron et al., 1993).

Established preclinical models with altered GH actions (Young et al., 2022, table 1; Qian et al., 2022) have significantly facilitated our understanding of the specific functions of GH on the human musculoskeletal system while also broadening study into mammals more generally. Mouse lines with excess GH, such as mice with constitutive overexpression of human growth hormone releasing hormone (GHRH), human GH (HGH), and bovine GH (bGH), show increased bone size as well as different bone mechanical characteristics (Wolf et al., 1991; D'Ercole, 1993; Tseng & Goldstein, 1998; Turner et al., 2001; Lim et al., 2015; Dixit et al., 2022; Jensen et al., 2022, table 1). The bGH model produces GH at levels up to 400‐fold that of wild‐type (WT) mice; individuals are generally larger, substantially leaner, and shorter lived than WT controls (Palmer et al., 2009). Studies using the bGH model reported consistently larger limbs as well as radiologically and mechanically determined differences in bone quality and microstructure (Dixit et al., 2022; Turner et al., 2001) and identified sexually dimorphic patterns in femoral midshaft cortical thickness and trabecular structure (Dixit et al., 2022). The bone loss and trabecular changes in adult bGH mice, presumably induced by excess autocrine and paracrine GH production, resemble what is seen in osteoarthritis (OA) and aging (Lim et al., 2015). We and others have reported that bGH mice exhibit altered histological and mRNA evidence of progressive joint degeneration starting as early 3‐months of age (Poudel et al., 2023; Sukul et al., 2025; Zhu et al., 2023) as well as systemic fibrosis in multiple tissue systems by 15 months (Lach et al., 2026). In counterpoint, GHR−/− gene disruption or ‘knockout’ (GHRKO) mice are the longest‐lived laboratory mice known (Pilcher, 2003) and are protected from developing age‐related diseases including diet‐induced diabetes, cancer (Guevara‐Aguirre et al., 2011), and joint degeneration (Liu et al., 2024). These mice have ∼50–60% decreased body size compared with WT mice (List et al., 2019). In the postcranial skeleton, GHRKO mice show decreased trabecular bone volume, as well as reduced cortical bone total cross‐sectional area, bone area, cortical bone thickness, periosteal/endosteal circumference, bone mineral density (BMD), and bone mineral content (BMC) in the femur and tibia (Sjögren et al., 2000).

Further human clinical and preclinical research also has promise to enable understanding of the fundamental effects of the GH/IGF‐1 axis on the broader morphoscape of vertebrate diversity. At present, the evolutionary and comparative biology investigations of GH and GHR modulation remain relatively focused on individual species or else on a limited number of species (Dantzer & Swanson, 2012; Liu et al., 2001; Rasband et al., 2023; Sutter et al., 2007). How GH dynamics work at the macroevolutionary scale remains unclear, although more recent work (e.g., Luo et al., 2024) is beginning to more comprehensively build understanding between molecular‐ and organismal‐scale morphology by elucidating specific growth‐related genes that appear to drive disparity among body size‐diverse clades.

Despite research advancements in characterizing detailed histomorphological changes of the skeletal system at specific anatomical locations in bGH mice, we lack a wholistic understanding of the systemic effects of GH on skeletal morphology and anatomical organization in a broader organismal context that could (1) identify patterns and differences not discernible from previous histological and mechanistic studies, (2) identify additional regions of the skeleton for such studies in the future, and (3) provide insights for broader studies beyond this model. In this pilot study, we sought to survey and characterize the gross anatomical postcranial skeletal phenotypes in adult (15‐month‐old) transgenic mice expressing the bGH gene. We compared postcranial skeletal features of the vertebral column, forelimb, and hind limb of male and female bGH mice with those of age‐ and sex‐matched WT controls, with the aim of identifying differences in size, shape, and condition that might in turn be indicative of sex‐ and genotype‐dependent effects of excess GH on the bGH model system.

2. MATERIALS AND METHODS

2.1. Samples

Transgenic bovine growth hormone (bGH) mice with C57BL/6J genetic background and age‐matched WT control mice were used to characterize the morphological effects of GH overexpression on the postcranial skeleton. Our sample for this pilot study was comprised of specimens frozen after histological and metabolic sample collection for a bGH‐induced fibrosis study (Lach et al., 2026). All mice were housed under standard conditions at 22°C with a 14‐h light/10‐h dark cycle and provided ad libitum access to water and food (ProLab RMH 3000), and all animal care and experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee at Ohio University. The specimens used in this study had undergone varying degrees of dissection by the time of scanning, but all retained an intact vertebral column and at least one forelimb and one hind limb. All were received with the viscera removed, meaning that tissues surrounding and supporting the vertebral column and body wall did not present an in‐life anatomical condition when captured tomographically. Three sampled mice (two bGH males and one bGH female) died prior to sacrifice and experienced soft tissue decomposition, which disallowed mass and body composition measurements as well as intended histological studies. There was a limited sample available to us for each comparison group (genotype × sex), and random sampling was not possible. We attained sample sizes of n = 5 for two of the four groups (WT female and bGH male), but only n = 3 for WT males and n = 4 for bGH females. Female mice in the study had not been used for breeding.

Following μCT data collection, the scanned specimens were thawed and used for additional studies related to the effects of bGH on musculoskeletal tissues (Sukul et al., 2025; Liu et al., in revision).

2.2. Data acquisition, processing, and measurements

Our characterizations and comparisons were carried out using micro‐computed tomography (μCT). Specimens for μCT scanning were prepared and vacuum‐sealed to standardize anatomical positioning and to minimize movement due to thawing during the scanning protocol. Scans were completed on a GE eXplore CT120 Small‐Animal X‐ray CT Scanner housed at the Institute for Molecular Medicine and Aging/Konneker Research Laboratories, Ohio University using a 50 μm fast‐scan protocol (continuous bed movement, 120 kV energy, 32 mA current, 20 ms exposure, no frame averaging) to minimize thaw‐induced movement. The resultant reconstructions yielded 0.0493 mm isotropic voxel datasets.

We processed, analyzed, and imaged reconstructed volumes of raw data in Avizo (ThermoFisher Scientific). We used the Segmentation Editor tool to perform digital dissections (segmentation) of selected skeletal elements and sesamoids; we did not include the manus, pes, or the more distally positioned caudal vertebrae in our survey. In addition to bone, we segmented numerous non‐bony but nevertheless radiodense materials associated with the skeleton. We utilized Generate Surface, Isosurface and Triangle Editor tools to create the polygon outputs and comparative assessments that comprise the bulk of our description. Details relating to lab protocols for generation of images for figures and supplementary videos have been reported previously (Evans et al., 2014; Krause et al., 2014; Krause et al., 2020 [supp. Info.]). Resultant images used for figures were resized and backgrounds cropped using Adobe Photoshop, and figures were created using Adobe Illustrator (Creative Cloud).

Measurements of mass and body composition were taken the day prior to sacrifice with a Bruker Minispec ND2506 (Bruker Corporation); linear measurements of digitally dissected appendicular elements were taken with the Linear Measurement tool in Avizo (Table 1). We used slice data to qualitatively categorize whether a radiographically distinct growth plate was present at the distal end of the femur, tibia‐fibula, humerus, radius and ulna, and at the proximal end of the tibia and radius, as well as whether a boundary between cartilaginous portions of the proximal femur and humerus and the olecranon process of the ulna could be distinguished (Table 2). We differentiated distinct state by a continuous separation across presumed tissue boundaries (i.e., sharing no trabecular openings), and by the presence of radio‐transparent voxels within those boundaries inferred to represent uncalcified tissue between cartilage and bone. Volumetric measurements of the sesamoids associated with the knee joint (i.e., suprapatella + patella, medial and lateral fabellae) were also calculated using the Material Statistics module in Avizo (Table 3). To do this, we first defined the range of histogram values representing the voxels of a 50 mm³ cylinder of hydroxyapatite contained within a phantom that was included in each scan. This bone‐dense range of histogram values was then applied to the digitally dissected sesamoids.

TABLE 1.

Specimen details and linear measurements (mm) of left side appendicular elements. OUVC 12374, 12377, and 12378 died prior to mass and body composition measurements.

OUVC #	Type	Sex	Mass (g)	Fat%	Lean%	Clav L	Scap L	Hum L	Hum ML	Hum CC	Ulna L	Rad L	Inom L	Fem L	Fem CC	Fem ML	Tibfib L	Calc L
12373	bGH	f	37.0	4.4	75.5	9.5	11.5	14.3	1.2	1.4	14.7	11.8	22.3	18.6	1.8	2.0	20.4	4.0
12374	bGH	f	/	/	/	9.5	12.4	14.8	1.2	1.5	14.8	12.3	22.5	18.7	1.7	2.0	20.3	4.0
12375	bGH	f	41.8	6.5	72.0	10.0	12.4	14.1	1.1	1.4	15.0	12.1	22.1	18.8	1.3	1.7	20.8	3.9
12376	bGH	f	35.7	10.9	68.6	8.7	10.9	12.6	1.1	1.2	14.1	11.3	18.4	15.8	1.7	1.9	18.7	3.9
12377	bGH	m	/	/	/	9.8	12.7	14.7	1.4	1.3	15.5	12.3	21.0	18.4	2.1	2.4	19.9	4.1
12378	bGH	m	/	/	/	10.0	12.5	14.8	1.4	1.8	15.7	12.5	21.1	18.5	2.1	2.6	20.3	3.9
12379	bGH	m	47.7	2.9	74.4	9.8	12.6	14.7	1.4	1.6	15.6	12.5	21.2	18.4	2.3	2.8	20.0	4.0
12380	bGH	m	49.0	2.6	74.4	10.7	12.9	15.5	1.6	1.7	16.0	12.4	21.2	18.5	2.3	2.4	20.1	4.1
12381	bGH	m	44.2	3.6	73.3	10.2	12.9	15.2	1.4	1.6	15.6	12.3	21.4	18.5	2.3	2.6	19.2	4.1
12382	WT	f	42.3	31.9	52.0	8.1	10.9	12.9	1.1	1.1	13.7	10.9	19.3	16.2	1.5	1.8	18.3	3.6
12383	WT	f	37.8	26.8	56.2	7.9	11.0	12.4	1.1	1.2	13.5	11.0	19.3	16.1	1.6	1.9	17.9	3.6
12384	WT	f	24.9	11.9	70.4	7.7	10.6	12.1	1.0	1.0	12.7	10.4	19.0	15.3	1.5	1.8	17.3	3.5
12385	WT	f	38.7	29.8	54.1	8.1	11.0	12.7	1.1	1.5	14.0	11.5	19.8	16.4	1.7	1.9	18.4	3.9
12386	WT	f	34.3	22.8	60.3	7.9	10.8	12.4	1.2	1.1	13.3	11.0	19.0	15.9	1.6	2.0	18.3	3.7
12387	WT	m	43.9	23.1	59.4	8.4	11.1	12.6	1.0	1.4	14.3	11.4	18.8	15.6	1.7	2.1	17.8	3.9
12388	WT	m	46.1	19.8	63.5	8.0	10.4	12.7	1.2	1.1	14.0	11.4	18.7	15.6	1.8	2.0	17.7	3.9
12389	WT	m	37.0	20.7	60.5	8.3	10.9	12.6	1.1	1.3	14.0	11.7	19.0	15.7	1.7	2.2	18.0	3.9

Abbreviations: Calc L, calcaneal length; Clav L, clavicular length; Fem CC, femur midshaft craniocaudal diameter; Fem L, femur length; Fem ML, femur midshaft mediolateral diameter; Hum CC, humerus midshaft craniocaudal diameter; Hum L, humerus length; Hum ML, humerus midshaft mediolateral diameter; Inom L, innominate length; Rad L, radius length; Scap L, scapular length; Tibfib L, tibia‐fibula length; Ulna L, ulna length.

TABLE 2.

Radiographic categorization of whether limb bone growth plates (gp) remain intact and whether a distinct cartilage cap (cc) exists on limb bone ends. '?' indicates inability to determine condition from slice data.

OUVC #	Type	Sex	Femoral head (cc)	Dist femur (gp)	Prox tibia (gp)	Prox humerus (cc)	Dist humerus (gp)	Olecranon (gp)	Dist antebrachium (gp)
12373	bGH	f	N	N	N	N	N	N	N
12374	bGH	f	N	N	N	N	?	?	?
12375	bGH	f	N	N	N	N	N	N	N
12376	bGH	f	N	Y	Y	Y	N	?	?
12377	bGH	m	N	N	?	N	N	N	N
12378	bGH	m	N	N	N	N	N	N	N
12379	bGH	m	N	N	N	N	N	N	N
12380	bGH	m	N	N	N	N	N	N	N
12381	bGH	m	N	N	N	N	N	N	N
12382	WT	f	N	Y	Y	Y	N	N	N
12383	WT	f	N	Y	Y	Y	N	N	N
12384	WT	f	N	Y	Y	Y	N	N	N
12385	WT	f	N	Y	Y	Y	N	N	N
12386	WT	f	N	Y	Y	Y	N	N	N
12387	WT	m	N	Y	Y	N	N	N	Y
12388	WT	m	N	Y	Y	N	N	N	Y
12389	WT	m	N	Y	Y	Y	N	N	Y

TABLE 3.

Volume (mm³) and proportion of bone‐dense tissue in sesamoids of the left knee. OUVC 12375 could not be used because a calibration phantom was not included in the scan; OUVC 12380 was excluded due to an inability to determine the edges of the medial fabella relative to surrounding radiodense soft tissues.

OUVC #	Type	Sex	Patella	Medial fabella	Lateral fabella	Patellar/Fabellar volume proportion	Avg	Fabellar volume proportion (M/L)	Avg
12373	bGH	f	2.298	0.445	0.569	2.266		0.781
12374	bGH	f	4.284	0.553	0.426	4.375	3.513	1.297	0.888
12376	bGH	f	1.803	0.171	0.291	3.898		0.587
12377	bGH	m	2.795	0.331	0.315	4.324		1.051
12378	bGH	m	3.190	0.673	0.353	3.040		1.907
12379	bGH	m	3.668	0.648	0.507	3.174	2.701	1.277	1.708
12380	bGH	m	3.132	2.649*	0.943	0.872		2.810
12381	bGH	m	3.194	0.913	0.611	2.095		1.494
12382	WT	f	1.001	0.164	0.252	2.419		0.650
12383	WT	f	1.051	0.136	0.256	2.675		0.532
12384	WT	f	0.771	0.137	0.197	2.305	2.505	0.696	0.602
12385	WT	f	1.149	0.145	0.278	2.715		0.519
12386	WT	f	1.217	0.192	0.313	2.412		0.614
12387	WT	m	0.966	0.137	0.293	2.244		0.468
12388	WT	m	0.933	0.159	0.276	2.143	2.43	0.576	0.541
12389	WT	m	1.343	0.169	0.293	2.909		0.578

2.3. Data interpretation

Interpretation of relative radiodensity is qualitative and correlative to (though not indicative of) tissue type, but general protocols for the scanning device employed produce bimodal grayscale histograms dividing a spectrum of bone‐dense voxels from air and most soft tissues. Our digital dissection protocol allowed us to produce consistent representations of skeletal elements for comparisons in this study. In addition to bony skeletal elements, our qualitative survey and digital dissections revealed numerous radiodense structures around the postcranial skeleton of WT and especially bGH mice that provide essential context for the characterization of the bGH postcranial skeletal phenotype. The limitation of using computed tomography as the sole method of reporting on the nature of these structures is manifest. However, the physical specimen processing noted above for downstream studies and further decomposition of soft tissues did not allow for direct dissection of specific regions of radiodensity revealed by our survey by the time our digital dissections were completed, and hence we can only infer the relationship between increased radiodensity and its biological nature for the purposes of this survey. We did not exhaustively document these regions but noted overall patterns. These regions are similar to bone in density, generally discrete, and distinct (i.e., visibly separable via segmentation) from bone surfaces. We employ the terms inferred heterotopic calcification (IHC) and radiodense soft tissue (RST) to differentiate types of densification anatomically coincident with non‐bony, collagen‐dominated skeletal connective tissues. Structures identified as IHC have that determination supported by data from recent histological studies confirming chondrification and ossification within Achilles tendon tissue (Liu et al., in revision: figures 1–3) and hypertrophic chondrocytes within vertebral growth plates (Sukul et al., 2025; figures 1a, S4) in bGH individuals which were scanned for this μCT survey. Another of our recent studies of bGH animals from the same genetic lineage and rearing protocols also found histological evidence of increased calcification within epiphyses, menisci, and ligaments of the knee (Zhu et al., 2023; figure 1). We therefore use these results to infer calcification is the likely cause of radiodensity when seen in anatomically‐equivalent regions in our specimens. Another of our recent papers used the mice scanned for this study in a broad histological soft tissue survey and found fibrosis to be widespread, most relevant to this study in the quadratus musculature of the hind limb (Lach et al., 2026; figures 3e and 4e). Because of widespread thickened and fibrotic tissues found in our sample mice, RST is used here to describe additional and unexpectedly radiodense structures anatomically coincident with tendons, ligaments, fascia, retinacula, and joint capsules in the postcranial skeleton without inference to cause of their densification (i.e., heterotopic ossification, heterotopic calcification, calcified or uncalcified fibrosis, or other causes).

Similarly, whereas the hydroxyapatite phantom included in scans does not provide an accurate characterization of tissue‐level organization of bone in our specimens, it nonetheless provides a consistent means by which to compare bone‐dense radiodensities across individual datasets. The established methodological standards and calibrations for using μCT to quantify the volume of complex bony features such as trabecular volume relate to specific algorithms, higher scan resolutions, and smaller regions of interest than our survey employed. Nonetheless, volumetric comparisons calibrated to bone mineral density provide valuable documentation of differences in larger‐scale, uniform, and anatomically‐equivalent bone‐dense structures, specifically sesamoids (this study) and calcification within Achilles tendon tissue (Liu et al., in revision). It is understood that the scale, resolution, and inconsistent positioning of long bones affect the interpretation of the structure of long bone cartilage‐bone differentiation zones, but the results of this initial survey indicate the need for further investigation into the timing and extent of tissue changes related to long bone growth in this mouse model.

Finally, we recognize that our CT‐derived description of degree and frequency of morphological differences of the genotype × sex groupings is necessarily hampered by the small sample of individuals surveyed. General trends we see as evident and relevant morphology are summarized in Table 4. An attempt to comprehensively elaborate upon the variation and occurrence of differences across the entirety of the postcranial skeleton, even within our small sample, was deemed to be beyond the scope of this pilot study. We provide examples of variation we see as most relevant in our figures, and also in a series of supplementary videos of sequenced and matched 360‐degree rotations with multiple individuals. In instances where individual specimens possess morphology relevant to providing nuance to general patterns, and for exploration of volumetric and polygon data more generally, additional resources are provided (see Data Availability Statement).

TABLE 4.

Summary of qualitative morphological observations as described in Results. 1. Incipient sacralization of the last lumbar vertebra; 2. Proximal caudal hemal arch deformation; 3. Tuberosity on scapular spine; 4. Mm. pectoralis scar not connected by ridge to humeral head; 5. Enclosure of biceps tendon groove on humerus in radiodense canal; 6. Mm. pronator quadratus sulcus across proximal antebrachium; 7. Deformation of caudoventral margins of obturator foramen on innominate; 8. Dorsoventral thickening and compression of the iliac crest on innominate; 9. Caudoventral orientation of acetabulum on innominate; 10. Craniodorsal orientation of femoral head; 11. Proportional narrowing of femoral midshaft; 12. Proximally extended patellar groove on distal femur; 13. Enclosure of foot extensor tendon groove in radiodense canal of tibia‐fibula; 14. Enclosure of groove for peroneal tendons in radiodense canal on calcaneus; 15. Presence of inferred heterotopic calcification (IHC) within Achilles tendon tissue.

OUVC #	Type	Sex	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
12373	bGH	f	y	y	n	y	y	y	y	y	y	y	y	y	y	y	y
12374	bGH	f	y	y	y	y	y	y	y	y	y	y	y	y	y	y	y
12375	bGH	f	n	y	y	y	n	n	y	y	y	y	y	y	y	y	y
12376	bGH	f	n	y	n	n	n	n	y	y	y	y	n	y	y	n	y
12377	bGH	m	y	y	y	n	n	n	n	y	y	y	n	y	y	n	y
12378	bGH	m	y	y	y	n	n	n	n	y	y	y	n	y	n	n	y
12379	bGH	m	y	y	y	n	n	y	n	y	y	y	n	y	y	n	y
12380	bGH	m	y	y	y	n	n	y	n	y	y	y	n	y	y	y	y
12381	bGH	m	y	y	y	n	y	n	n	y	y	y	n	y	y	y	y
12382	WT	f	n	n	n	n	n	n	n	n	n	n	n	n	n	n	y
12383	WT	f	n	n	n	n	n	n	n	n	n	n	n	n	n	n	y
12384	WT	f	n	n	n	n	n	n	n	n	n	n	n	n	y	n	y
12385	WT	f	n	n	n	n	n	n	n	n	n	n	n	n	y	n	y
12386	WT	f	n	n	n	n	n	n	n	n	n	n	n	n	y	n	y
12387	WT	m	n	n	n	n	n	n	n	n	n	n	n	n	n	n	y
12388	WT	m	n	n	n	n	n	n	n	n	n	n	n	n	n	n	y
12389	WT	m	n	n	n	n	n	n	n	n	n	n	n	n	n	n	y

2.4. Statistical analysis

Due to the small and varied sample sizes available for this pilot anatomical survey, we performed limited analysis on collected data. Two‐sided permutation p‐tests were used for pairwise comparisons of linear measurements of sexual dimorphism within genotype (i.e., M WT and F WT, M bGH, and F bGH) and of dimorphism between males and females across genotype (i.e., M WT and M bGH, F WT and F bGH). We also compared knee sesamoid volumes in this manner. Statistical tests were conducted in R Statistical Software (v4.5.2; R Core Team, 2025); box and whiskers plots were made using BoxPlotR (Spitzer et al., 2014).

3. RESULTS

3.1. General postcranial skeletal morphology

The bGH mice differ notably from WT controls in overall size, morphology, mass, and body composition (Figure 1). The thoracic vertebral column exhibits an enhanced kyphosis in living bGH mice that is variably preserved in our dissected bGH sample. Additionally, several bGH mice show radiodensity coincident with incipient sacralization of the sixth lumbar vertebra, and the genotype × sex groupings exhibit differences in caudal vertebral‐hemal arch joints (Figure 2) and other morphology on the ventral aspect of the midline (Figure S1). bGH animals have longer appendicular elements than sex‐matched WT mice, as well as differences in craniocaudal and mediolateral cross‐sectional diameters of those elements at bone midshaft (Table 1). Differences in bone shape, muscle attachment morphology, and linear measures are notable on elements of the pectoral girdle and forelimb (Figure 3; Video S1), the innominate (Figure 4; Video S2), femur (Figure 5; Video S3), and tibia‐fibula and calcaneus (Figure 6).

FIGURE 1.

General differences in morphology, size, mass, and body composition of adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. Top, threshold‐derived polygon visualizations highlighting general size and skeletal differences in example (a) wild‐type female (WT F) and (b) growth‐hormone model (bGH F) transgenic specimens in left lateral view. Right limbs, left ribs, and most caudal vertebrae digitally removed for visual clarity, with forelimbs and hind limbs digitally separated to allow unobstructed view of the axial skeleton. Scale bar = 1 cm. Bottom, box plots of (c) body mass and (d) percentage of body fat composition at sacrifice for wild‐type males (WT M), wild‐type females (WT F), growth‐hormone model males (bGH M), and growth‐hormone model females (bGH F) used in this study. Center lines median; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; data points are plotted as circles. Asterisks indicate permutation p < 0.05. At, Achilles tendon; C2, second cervical vertebra; Ca1, first caudal vertebra; ic, iliac crest of innominate; L1, first lumbar vertebra; L6, sixth (last) lumbar vertebra; pa, patella; pu, pubic portion of innominate; sg, shoulder girdle; T2, second thoracic vertebra; tk, thoracic kyphosis.

FIGURE 2.

Threshold‐derived polygon visualizations of lumbar, sacral, and proximal caudal regions of the vertebral column of representative male (M) and female (F) wild‐type (WT) and growth‐hormone model transgenic (bGH) mice in ventral view. Top, (a) WT male, (b) WT female, (c) bGH male, and (d) bGH female. Scale bar = 1 cm in a–f; g and h are not to scale to allow alignment of morphology for comparison. Below, variation in sixth lumbar transverse processes and hemal arch morphology in sampled (e) WT male, (f) WT female, (g) bGH male, and (h) bGH female mice. Dotted horizontal lines represent vertebral divisions, and dashed lines represent lateral and caudal portions of the innominate digitally dissected to allow unobstructed visualization of vertebrae. Asterisks (*) indicate radiodense tissue accumulation around transverse processes, vertebral centra, and hemal arches. Ca1, first caudal vertebra; Ca4, fourth caudal vertebra; ha, hemal arch (normal morphology); L6, sixth lumbar vertebra; S1, first sacral vertebra; S4, fourth sacral vertebra.

FIGURE 3.

General differences in morphology, size, and linear measures in the shoulder girdle and forelimb of adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. Top, digitally dissected polygon models of (a) WT male, (b) WT female, (c) bGH male, and (d) bGH female scapulae, clavicles, humeri, and antebrachia. Scapula (top left) in lateral view; clavicle (bottom left) in cranial view; humerus (middle) in (left to right) cranial and caudal views; radius and ulna (right) in (left to right) flexor and lateral views. Dashed line in (d) represents enclosure of the intertubercular sulcus within a calcified retinaculum. Asterisks (*) on a–d indicate notably different morphologies noted in text. Scale bar = 1 cm. Bottom, box plots of length of (e) humerus, (f) humerus midshaft craniocaudal cross‐section, (g) humerus midshaft mediolateral cross‐section, (h) scapula, (i) radius, and (j) ulna for the four genotype × sex groupings. Center lines show the median; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; data points are plotted as circles. Asterisks indicate permutation p < 0.05. aa, acromion articulation of clavicle; ap, acromion process; cp, coracoid process; dc, deltoid crest; its, intertubercular sulcus; lt, lesser tubercle; op, olecranon process; pe, pectoralis insertion; ps, pronator sulcus; rr, radial ridge; sa, sternal articulation of the clavicle; ss, scapular spine.

FIGURE 4.

General differences in morphology, size, and length in the innominate of adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. Top, digitally dissected polygon models of (a) WT male, (b) WT female, (c) bGH male, and (d) bGH female innominates in dorsal (top), lateral (middle), medial (bottom), and cranial (bottom right) views. Dashed lines trace the circumference of the acetabulum. Asterisks (*) on a–d indicate notably different morphologies discussed in text. Scale bar = 1 cm. Bottom, box plot of (e) innominate length for the four genotype × sex groupings. Center lines show the median; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; data points are plotted as circles. Asterisks indicate permutation p < 0.05. ac, acetabulum; ic, iliac crest; is, ischium; sn, sciatic notch; ps, pubic symphysis.

FIGURE 5.

General differences in morphology, size, and linear measures in the femur of adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. Top, digitally dissected polygon models of (a) WT M, (b) WT F, (c) bGH M, (d) bGH F femora in cranial (left) and lateral (right) views, and (e) caudal view of female bGH femora (n = 4) to illustrate variability of morphology in the group. Asterisks (*) on a–d indicate notably different morphologies discussed in text. Scale bar = 1 cm. Middle, box plots of lengths of (e) femur, (f) femur midshaft craniocaudal cross‐section, and (g) femur midshaft mediolateral cross‐section diameter for the four genotype × sex groupings. Center lines show the median; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; data points are plotted as circles. Asterisks indicate permutation p < 0.05. Bottom, variation within each of the four genotype × sex groupings (i–l) in midshaft cross‐sectional shape and distal condyles, with lines representing positions of cross‐sectional measurements in (g) and (h). Images not to scale but sized to approximate equivalent craniocaudal length at the distal end of the element. fh, femoral head; lt, lesser trochanter; ms, midshaft; pg, patellar groove; tt, third trochanter.

FIGURE 6.

General differences in morphology, size, and linear measures in hind limb elements of adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. Top, digitally dissected polygon models of (a) WT male, (b) WT female, (c) bGH male, (d) bGH female tibiae‐fibulae, astragali, and calcanei. Tibia‐fibula in cranial (left), lateral (middle), and caudal (right) views; articulated astragalus and calcaneum in lateral (upper right) and caudal (bottom right) views. Scale bars = 1 cm for tibia‐fibula and 5 mm for calcaneum and astragalus. Asterisks (*) on a–d indicate notably different morphologies discussed in text. Bottom, box plot of lengths of (e) tibia‐fibula and (f) calcaneus for the four genotype × sex groupings. Center lines show the median; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; data points are plotted as circles. Asterisks indicate permutation p < 0.05. at, trochlea of astragalus; ct, calcaneal tubercle; fs, fibular shaft; lm, lateral malleolus; mm, medial malleolus; nf, navicular facet of astragalus; pp, peroneal process of the calcaneum; sp, insertion for semimembranosus and popliteus mm.; su, sustentaculum of calcaneum; ta, origin for tibialis anterior; tc, tibial crest.

bGH mice in the sample have relatively less distinct differentiation of cartilage and bone boundaries on long bone ends (Table 2; Figure 7) and larger and differently proportioned sesamoids (Table 3; Figure 8). In addition to differences in calcified tissues, bGH mice show a disproportionate degree of radiodense thickening corresponding anatomically to collagen‐dominated soft tissues (i.e., intervertebral disc, tendon, retinaculum, ligament, meniscus) associated with joints and entheses that is most remarkable around the knee (Figure 9; Videos S4, S5). Sex‐specific differences are also evident. Males attain larger size and robusticity of skeletal elements, with general morphological similarity between the two genotypes. By contrast, female bGH mice display higher variation relative to WT females and, in some cases, clearly pathological skeletal elements in comparison with all other groups. A summary of these qualitative morphological observations is provided (Table 4).

FIGURE 7.

Representative computed tomography slice views of selected appendicular joints of 15‐month‐old male (M) and female (F) wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. (a) Shoulder, (b) wrist, and (c) knee joints. Views and sides vary due to different joint angles and positions at time of scanning. White arrows indicate regions interpreted to have an intact growth plate in one or more elements at the joint. Scale bars = 1 mm. Fe, femur; Hu, humerus; Pa, patella; Ra, radius; Ti, tibia; Ul, ulna.

FIGURE 8.

Morphology, bone‐dense volume, and bone‐dense volume proportions of knee sesamoids of adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. Top, (a) WT F sesamoids at knee joint in cranial (left) and caudal (right) views; (b) bGH F sesamoids at knee joint in cranial (left) and caudal (right) views. Fabellae and patellae in dark gray, and suprapatellae in red. Bottom, box plots of calibrated bone‐dense volumes of the (c) patella, (d) medial fabella, (e) and lateral fabella; box plots of proportions of (f) patellar‐to‐fabellar bone‐dense volume, (g) medial to lateral fabellar bone‐dense volume, and (h) total knee sesamoid bone‐dense volume to body mass. Center lines show the median; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles; data points are plotted as circles. Asterisks indicate permutation p < 0.05. Scale bar = 1 cm. Fe, femur; lf, lateral fabella; mf, medial fabella; Pa + s, patella+suprapatella; Tf, tibia‐fibula.

FIGURE 9.

Digitally dissected examples (displayed as unlabeled and in color) of atypical accumulations of radiodense soft tissues in the postcrania of surveyed adult wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. (a) bGH M presacral vertebral column in ventral view with ribs and sternal elements digitally removed; (b) bGH M glenoid, medial view; (c) WT and (d) bGH female innominates in lateral view; (e) bGH M proximal femur, medial view; (f) WT F, (g) bGH F and (h) bGH M knee, medial view; (i) WT M and (j) bGH M ankle, medial view. Scale bar = 2 cm in a, 1 cm in b–j. Ac, acetabulum; ap, acromion process of the scapula; As, astragalus; At, Achilles tendon inferred heterotopic calcification; C2, second cervical vertebra; Ca, calcaneus; Fe, femur; fh, femoral head; Hu, humerus; is, ischium; L1, first lumbar vertebra; L6, sixth lumbar vertebra; mf, medial fabella; mn, meniscus; Pa + s, patella+suprapatella; ps, pubic symphysis; Sc, scapula; T2, second thoracic vertebra; Tf, tibia‐fibula.

3.2. Postcranial axial skeleton

All bGH mice surveyed had the normal precaudal vertebral patterning of WT mice (i.e., seven cervical, 13 thoracic, six lumbar, and four sacral vertebrae). Individual vertebral and bony rib morphology in our assessment did not vary notably between bGH and WT mice, but overall patterns of presacral column curvature indicate kyphosis‐induced shape differences in bGH cervical, thoracic, and lumbar spines that were variably preserved after dissections.

In the lumbosacral portion of the column (Figure 2), the transverse process of the sixth lumbar vertebra in bGH males exhibits incipient sacralization in being laterally extended, generally dorsoventrally broader, and more dorsally oriented than it is in female bGH or WT mice. In contrast, the sixth lumbar vertebra of most females and WT males in our sample exhibits narrower, cranioventrally projecting lumbar transverse processes, although some bGH females do show widening of the processes (Figure 2h). Despite these morphological differences in transverse process shape and size, the vertebral body and zygapophyseal contacts of the last lumbar vertebra with the sacrum appear in our assessment to remain unfused in our sample.

The transverse process of the first sacral vertebra forms all or (in a few cases) most of the fusion with the ilium; the bone on either side of this suture is decidedly more robust in bGH mice. The first and second sacral vertebrae are always completely fused (both body and neural arch), with the second fused most frequently to the third at the body (Figure S1A) and less consistently at the zygapophyses and transverse processes. The fourth sacral vertebra rarely exhibits fusion with the third at either the centrum or the zygapophyses and only occasionally along the transverse process, with no observed patterns of difference relative to sex or genotype.

At the sacral‐caudal transition, WT males lack RST spanning the S4‐Ca1 joint (Figure 2e). By contrast, bGH male mice exhibit irregularly shaped RST along the ventral portion of the S4‐Ca1 joint and at subsequent joints in a manner that appears functionally contiguous across the first few proximal caudal vertebrae (Figure 2g; Figure S1A). Whereas it is likely that some of this material represents the actual hemal arch, there is limited passage for vasculature between the ossifications and the ventral surface of the vertebral bodies/intervertebral disc (Figure S1B). These same mice as well as bGH F mice exhibit RST at the distal margin of caudal vertebral transverse processes (Figure 2c,d). All WT females also have radiodense regions bridging the ventral portion of the sacral‐caudal transition, but unlike in males or bGH females, two specimens also exhibit prominent RST at the ventral junction between S3 and S4 (Figure 2b,f). At the S4‐Ca1 joint, three of the WT females exhibit incipient (or relatively less radiodense) hemal arch morphology like that of the Ca1–Ca2 transition in the WT males (Figure 2f); our survey found hemal arch morphology in the group to be otherwise normal. bGH females share with bGH males the ventral irregularity of RST at the S4‐Ca1 joint, but to a less massive degree and with some discernible, anatomically normal hemal arch partitioning (Figure 2d,h).

All 17 mice also share with wild mice (i.e., naturally occurring individuals of the species) a pattern of having five discrete portions of the sternum (i.e., a fused manubrium and individualized first, second, and third sternebrae, succeeded by fused fourth and fifth sternebrae, and the xiphisternum) (Figure S1C). The sixth and seventh ribs are the last to articulate with the sternum and do so at the fifth sternebral‐xiphisternal synchondrosis, while the distal ends of the eighth through thirteenth ribs are free. Some bGH mice have relatively thick costal cartilages, and in some cases, radiographically distinct portions of floating cartilages within the more caudally positioned ribs (Figure S1D).

3.3. Pectoral girdle and forelimb

The clavicle of bGH mice is proportionally more robust, particularly where it articulates with the scapula and through the mid‐portion of the element (Figure 3c,d). Its acromial articulation is disproportionately enlarged and curved dorsally relative to the WT condition (Figure 3a,b). bGH groups have relatively robust coracoid, acromion, and metacromion processes on the scapula. bGH specimens share thickening of the scapular spine at approximately the widest part of the scapular blade, where it flattens into a distinct tuberosity that WT mice in the sample lack.

Both bGH sexes have an enlarged and irregularly shaped lesser tubercle on the proximal end of the humerus and exhibit a substantially thicker margin at the deltoid crest. Most bGH females bear a raised scar for the insertion of the pectoralis musculature that is largely disconnected from the lesser tubercle (Figure 3d; Video S1); among other specimens, this scar is variable in size but contiguous with the humeral head via a ridge of bone. Two females and one male bGH specimen also display RST anatomically coincident with retinacular tissue around the intertubercular sulcus near the humeral head (Figure 3d; Video S1), forming a radiodense enclosed canal for the long head of the biceps tendon.

The olecranon process of the ulna in some surveyed bGH mice is thicker mediolaterally than that in WT mice (Figure 3c). The ridge on the dorsal aspect of the radius is accentuated, and some specimens exhibit a small, raised scar on the ridge not observed in WT mice (Figure 3c). The radius in some bGH mice also exhibits a distinctly defined sulcus on the lateral aspect, likely a scar from mm pronator quadratus (Figure 3c,d).

Males possess longer scapulae and forelimb elements than females regardless of genotype, and bGH mice have a dimorphic signal not seen in WT mice (Figure 3e,h–j).

3.4. Pelvic girdle and hind limb

The innominate bears sex‐specific, genotype‐specific, and (in the case of bGH females) group‐specific morphologies (Figure 4; Video S2). Male mice exhibit more robust innominates than females within respective genotypes, and bGH male innominates are longer than WT innominates. The sex‐based morphological differences are particularly notable along the dorsal aspect of the iliac crest, at the sacro‐iliac joint, and within the ischium and pubis. The iliac crest and sacral joints of male and female bGH mice are enlarged and dorsoventrally compressed, and the acetabulum projects caudoventrally, not laterally, to receive the femoral head.

Female bGH mice variably express presumably pathological lateral deformation of the caudal portion of the innominate, as well as loss of bone that in WT and bGH male mice forms the posteroventral border of the obturator foramen and pubic symphysis (Video S2; Figure S1e). In a lateral view, bGH females more resemble males of both genotypes in exhibiting a roughly co‐equal caudal extension of both the ischial and pubic portions of the innominate. By contrast, the ischial tuber in female WT mice is shorter caudally (Figure 4b), resulting in a sexually dimorphic difference in the length of the element (Figure 4e). Whereas it may be due in part to our sample condition and preparation, all female specimens had separated pubic symphyses while all but one male had intact ones; in bGH females, however, this displacement is accompanied by deformation of the pubis (Video S2). This discrepancy is not due to pregnancy as female bGH mice used in this study were not used as breeders. The length of bGH female innominates in our survey is also highly variable compared with the other groupings.

Consistent with the morphology of the acetabulum, the femoral head of bGH mice projects dorsally and cranially relative to the shaft, with the third trochanter canted cranially, not laterally (Figure 5; Video S3). The length of the element is similar within each genotype, but bGH male and female femora are longer than those of WT males and females (Figure 5f). The distal third of the femur in bGH mice bears more prominent muscle attachment sites on both medial and lateral surfaces and a dorsally extended patellar groove that rises well proximally relative to the condyles (Figure 5c,d). Male femora are not notably different in shape between genotypes, though in bGH males they are wider at the midshaft cross‐section and appear to have a thinner cortical component (Figure 5g,h,k). They share a robust midshaft that does not winnow much in circumference between the third trochanter and distal condyles. Whereas females share a relatively more gracile morphology overall compared with males irrespective of genotype, intragroup variation in the length, cross‐section properties, and overall midshaft shape of bGH females is striking (Figure 5e,l). They maintain cross‐sectional midshaft measurements similar to WT females despite generally larger element lengths, resulting in a proportionally winnowed midshaft relative to similar dimensions at the third trochanter and near the distal end. This is accompanied by different shapes in cross‐section and varying degrees of bending and torsion (Figure 5e,h; Video S3). Distal femoral morphology of male and female WT mice is invariant relative to bGH mice; bGH females bear grooves and ridges for muscle attachments not seen on male bGH mice.

In contrast to the innominate and femur, morphology of the tibia‐fibula varies little among groups (Figure 6). bGH males appear to have a slightly more robust tibial crest and fibular shaft relative to other groups; the fibular shaft also appears to be somewhat straighter along its length (Figure 6c). bGH females have relatively thinner and more curved fibular shafts than bGH males and may exhibit slightly more variation in the orientation of the proximal condyles, but they lack the notable within‐group morphological differences and deformation observed in either the innominate or femur. The distal fusion of the tibia and fibula has a variable RST enclosure on the cranial surface that presumably surrounds a tendon for extensor musculature of the foot (Figure 6c,d). Total length of the tibia‐fibula is significantly longer in the bGH genotype in male–male and female–female comparisons, but within‐genotype sexual dimorphism is not evident in either case (Figure 6e).

There is no notable morphological disparity in the astragalus. Instead of a groove on the plantar aspect of the peroneal process of the calcaneus for mm. peroneus digiti quarti, digiti quinti, and longus, some bGH specimens (Figure 6c,d; Table 4) exhibit RST that appears to completely enclose the peroneal tendons within a canal similar to other RST in retinacula‐supported tendons noted on the tibia and humerus. The general pattern of sex comparisons across genotypes seen in the length of the tibia‐fibula is also seen in the length of the calcaneus (Figure 6f).

3.5. Long bone epiphyseal plate patterns

A radiographic assessment of long bone growth plates reveals a variable degree of patency generally consistent within genotype × sex groupings (Table 2; Figure 7). We found all sampled mice lacked distinct epiphyseal differentiation in elements comprising the hip and ankle joints (i.e., innominate, proximal femur, and distal tibia‐fibula), at the distal humerus, and at the proximal radius. All bGH males lacked detectable separation of cartilage and bone on limb ends and had epiphyseal continuity in all examined limb bones, except one mouse with a partially intact growth plate at the proximal tibia (Figure 7c, bottom left). One of three bGH female specimens maintained intact growth plates at the knee, a cartilage cap on the humerus, inconclusively closed plates in the distal antebrachium, and possibly a portion of cartilage on the olecranon. WT males and females uniformly possess radiographically distinct growth plates at both the distal femur and proximal tibia but show different patterns of forelimb fusion. While all WT females appear to maintain a distinct cartilaginous portion of the humerus, two of the three males do not. In contrast, WT females lack distinct growth plates in the distal radius and ulna whereas males appear to have retained them.

3.6. Sesamoid, inferred heterotopic calcification, and skeletally‐associated radiodense soft tissue

Our anatomical survey of bony and cartilaginous structures known to develop within tendon tissue identified all specimens with consistent sesamoid organization (number and position) at the sternum, elbow, and knee joints. Paired omosternal elements of similar size and shape sit at each side of the sternoclavicular joint. The sesamoid situated between the proximal radius and lateral epicondyle of the humerus tends to be larger and more variable in shape in bGH specimens. At the knee joint of male and female WT mice, the bone‐dense volume of the patellar and fabellar sesamoids, as well as patellar‐to‐fabellar bone volume proportion (i.e., intra‐tendinous bone‐dense growth in the extensor versus flexor compartments of the joint) and fabellar proportion (i.e., medial to lateral sesamoids) are all remarkably consistent (Table 3; Figure 8d–f). In contrast, bGH mice have inconsistently larger and differently proportioned sesamoids relative to WT mice, more variability in both volume and proportion from specimen to specimen, and notably higher amounts of bone‐dense volume when body mass is taken into account (Figure 8h).

In addition to sesamoids, all mice examined in this survey exhibited other radiologically detectable structures anatomically coincident with non‐bony, collagen‐dominated skeletal connective tissues around joints and entheses (Figure 9). However, these regions of IHC and RST are more pervasive, larger, and numerous in bGH mice than in WT counterparts and therefore represent anatomy essential to characterization of the overall phenotype of the adult bGH mouse postcranial skeleton.

Spinal IHC and RST are most frequent in the bGH thoracic region but also occur in cervical, lumbar, and proximal caudal regions. The degree of development varies between individuals, but IHC/RST is prevalent along the ventral surface of bGH presacral vertebrae, anatomically coincident with intervertebral discs and within the ventral longitudinal ligament (Figure 9a). In the forelimb, bGH mice develop RST around the omosternal region, with larger additional nonskeletal RST regions within or around the shoulder joint (Figure 9b).

Near the hip joint, RST is associated with the capsule and muscle attachment sites (Figure 9c–e). While some female WT mice exhibit RST dorsal and caudal to the acetabulum, all bGH mice express between two and three large, overlapping volumes of RST in this region, which is anatomically coincident with fleshy tendinous attachments of the adductor compartment of the hip (mm gemellus, caudofemoralis, et obturator [Charles et al., 2016; Figure S1]). These are accompanied by numerous (2–5) structures associated with the proximal femur that also likely represent musculotendinous or fibrocartilaginous entheses that cross the hip joint (Figure 9e).

IHC and RST around and within the knee joint is by far the most variable between genotypes (Figure 9f–h; Video S4). While all four quadrants of the meniscus indicate thickening in all 17 specimens, WT males bear no additional IHC or RST within or near to the joint capsule itself. All WT females except one bear small, generally spherical IHC on the medial and lateral aspects of the knee joint, likely within the medial and lateral ligaments (Figure 9f). The condition in bGH females (Figure 9g; Video S5) and males (Figure 9h) indicates extreme pathology and likely severely altered mobility. Many exhibit extensive IHC throughout the joint and display dramatic enlargement of RST within the patellar tendon (Figure 8b), a condition coincident with the enlarged patellar groove of the femur (Figure 5c,d). Moreover, in several bGH males a separation between the patella and suprapatella cannot be distinguished. All bGH mice bear largely discrete regions of variably sized and shaped (3–25+ in number) IHC and RST around the margins of the joint capsule that probably represent fibrocartilaginous, ligamentous, and tendinous thickening (Video S4). Sex‐based differences appear to exist. Some bGH males have larger, plate‐like structures around parts of the joint not seen in female bGH mice and have more growths of variable, but generally smaller, sizes expressed along cranial, caudal, medial, and lateral aspects of the knee. Females in the sample generally form fewer, larger growths. Two females have RST along the lateral aspect of the femur that projects toward the tibia‐fibula, representing what may be the only examples of bony (osteophytic) outgrowth from a bone into the joint. On the medial aspect of the knee, two females have a long, rod‐like RST that spans the entire joint from the distal femur to the proximal tibia‐fibula (Figure 9g; Videos S4 and S5), a condition that presumably would have severely limited or even prevented movement at the joint. All 17 mice express some level of IHC within the Achilles tendon (Figure 9i,j) but bGH mice (males and females alike) exhibit larger volumes of the intratendinous IHC. The position of these volumes generally indicates IHC of tendon within the two heads of the gastrocnemius muscle. Three of the five bGH males also possess IHC within the soleus tendon.

4. DISCUSSION

4.1. Postcranial skeletal disparities in the context of excess GH

Our radiological findings indicate that vertebral segmentation is normal within our sample of bGH mice, and thus that the presacral kyphotic curvature seen in intact bGH animals (Junnila et al., 2013) is not somitogenic or pleiotropic in nature but rather likely the result of GH‐driven growth within normally partitioned segments that is reinforced over lifespan by intervertebral IHC/RST. However, this supposition based on our preliminary survey must await further studies that utilize intact specimens from additional time points scanned at higher resolutions. An important (though expected) corroborative finding for the bGH model axial skeleton is the patterning of rib and sternal segments are consistent with the WT condition. It is less clear if the observed incipient sacralization of the last lumbar vertebra or differences detailed for the proximal caudal region relate to the processes causing lifespan‐accumulated IHC/RST, and/or to the bony morphological differences seen in the ilium, femur, and hemal arches, or to other factors earlier in life. Sacralization of the last lumbar vertebra is within normal variation in WT mice as well as in some mouse models (Munro & Barnett, 1969). A recent large‐scale mouse genomic survey that was focused on vertebral anatomy identified over 200 genes involved in morphology and development; three with broad metabolic and developmental effects (Dnase1l2, Duoxa2, Fbn2) were associated with vertebral fusion and transverse process and hemal arch morphology in the sacral region (Ibarra‐Soria et al., 2026). Although there are no data to suggest a direct link between the bGH transgene and lumbar, sacral, or proximal caudal condition, our study does underscore the lumbosacral region as a future target for histological and mechanistic studies to clarify the bGH axial phenotype, beginning with comparisons of morphology at earlier timepoints to better understand effects of age and habitus. Experimental studies tracking changes in rodent caudal vertebrae helped to define understanding of the effects of posture and positioning on bone resorption and deposition (Ellender et al., 1989; Feik & Carach, 1988) and variation in hemal arch morphology in mammals as it relates to function remains understudied (Zavodszky & Russo, 2020). Since the configuration of the acetabular joint and (in females) lateral displacement of the pubis and caudal portion of the ischium in bGH mice suggest substantial differences in muscle action in the hip compartment, it is reasonable to expect that flexor musculature at the base of the tail would be similarly affected.

In addition to observations in the axial skeleton, we find substantial size (by linear measure) and morphological differentiation throughout the appendicular skeleton in bGH mice. Specifically, we find that: (1) overall size is consistently larger in bGH compared with WT mice; (2) overall morphology is consistently more similar in male WT‐to‐bGH comparisons than between females of the two genotypes; (3) the combination of these disparities appears to drive skeleton‐wide sexual dimorphism in the bGH model; (4) the dimorphism may result from different patterns of bGH growth (as seen especially in the femur) rather than from solely differential amounts of growth, and (5) these disparities exhibit high variance in measures and variability in morphology among bGH female individuals not seen in any of the other genotype × sex groups. The initial survey of limb bone ends of animals in our sample also hints that different patterns of epiphyseal closure that are evident even after 15 months of age could have previously overlooked sex‐ and element‐specific trajectories in WT mice, and that these could also ultimately factor into excess‐GH‐mediated length and shape outcomes. Variation in GH secretion and response are fundamentally linked to the influences of sex hormones (Chagin et al., 2004; Leung et al., 2004; Börjesson et al., 2010) as are patterns of pituitary GH pulsatility/GHRH response (Jessup et al., 2003; Johnson, 1988; Sanchez‐Cardenas et al., 2010). The extreme autocrine/paracrine level of release in bGH mice (Palmer et al., 2009) confounds larger extrapolation of GH effects on morphology in any biologically plausible manner but nonetheless points toward profound bone tissue effects of increased levels of GH on mediating the disparate size and shape of the male and female postcranial skeleton. Further, we have added nuance to recent work demonstrating bGH sexual dimorphism in intestinal gross anatomy (Jensen et al., 2022) and structural bone properties (Dixit et al., 2022) by demonstrating that bGH female skeletal variability may encompass a wider spectrum of morphologies. We suggest that this could be interpreted as a relative but inconsistent ability of females to mediate or attenuate the excess GH. For example, one female bGH specimen (OUVC 12376) was below median weight for all (WT and bGH) females while retaining a body fat percentage intermediate between the two genotypes. Its linear skeletal measurements are consistent with WT females (Table 1), but it exhibited femoral morphology distinctly bGH in character (i.e., a tall patellar groove, a distinctly enlarged distal end, and cranial orientation of the head; Figure 5e). In the innominate, this same specimen also exhibited less severe (though present) dorsoventral compression of the iliac crest seen in all bGH specimens, a caudally oriented acetabulum seen only in (and in all) bGH animals, and a laterally deflected pubic symphysis seen only in bGH females. Our sample size is small but this outlier, especially in the context of different expressions of bone growth and pathology observed in bGH females, nonetheless supports the supposition that female skeletal metabolism resulting in this morphology responds in a markedly different manner to the excess GH trajectory than that of bGH males.

4.2. bGH entheses in the context of GH effects on bone and skeletal muscle

Our secondary finding of widespread IHC and RST in scanned bGH mice corroborates the myriad histological and molecular evidence of hypertrophic chondrification, heterotopic ossification, and fibrosis in the model, but importantly identifies organism‐wide patterns at gross anatomical scale. Of particular relevance to the interpretation of our reported morphological differences are regions anatomically coincident with the bone‐muscle interface. Across the sample, male and female bGH limb bones share more scars, sulci, and thickening at muscle attachment sites, as well as RST of (likely ligamentous) retinacula associated with long tendons (e.g., mm. biceps brachii, mm. peroneus). The long, pennate tendinous connection of mm. pectoralis can be presumed to have caused the prominent (and in females distinctly separated) scar on the deltoid crest of the humerus, and fleshy entheses in the hip adductor musculature are coincident with large volumes of RST as well as the cranial orientation of the third trochanter of the femur. These patterns have manifest systemic metabolic (e.g., inflammation and enthesitis) and probably differential (i.e., “biological”) aging‐related explanations, but other lifespan‐accumulated influences on entheses are also likely responsible for differences from WT mice.

Though we lack locomotor information or cellular/molecular sampling specific to these sites of enhanced surface bone growth in our sample, whole‐bone morphology in bGH mice near the end of their expected lifespan may warrant studies that also consider organismal‐scale biomechanical interpretations of skeletal shape. On the bone side of these connective tissue interfaces, cumulative mechanical stress from longer limbs and correspondingly altered muscle actions, as well as the fundamentally different body composition of bGH mice (Berryman et al., 2010; Palmer et al., 2009), might drive enhanced periosteal osteogenesis, resulting in more prominent morphology consistent with our radiographic observations. The periosteum is a key mediator for bone growth, repair, and remodeling, particularly in regions subjected to high mechanical stress (Ellender et al., 1989; Feik et al., 1987; Ito et al., 2014). In other bone compartments bGH animals exhibit different and generally weaker cortical and trabecular hind limb and vertebral mechanical properties in comparison to controls (Dixit et al., 2022; Lim et al., 2015; Liu et al., 2017). Chronic stress or strain on this relatively weaker and differently deposited bone could also lead to chronic periosteal response, promoting subsequent and different tissue deposition and mineralization.

On the muscle side of these connective boundaries, RST at the shoulder and especially the hip joints also show evidence of soft tissue thickening coincident with or within the extracellular matrix (ECM) of fleshy tendinous attachments. We recently reported fibrosis in bGH quadriceps muscle tissue of mice used for this study (Lach et al., 2026), in a location consistent with RST prevalence around the osteologically dysmorphic bGH acetabulum. Autocrine‐derived GH plays a major role in the growth and development of muscle size and fiber type (Dudley & Portanova, 1987; Harvey, 2010; Segard et al., 2003; Schuenke et al., 2008; Young et al., 2022), with its action preferentially on collagen synthesis to the exclusion of myofibrillar development regardless of exercise as a variable (Doessing et al., 2010). The mass increase of distinct muscles in bGH mice has been shown to vary relative to body size, with an oxidative/Type I fiber dominant muscle (Mm. soleus) scaling equally and a glycolytic/Type II fiber dominant muscle (Mm. gastrocnemius) scaling negatively (Consitt et al., 2017). Such differences in muscle size and fiber type differentiation could contribute to the disparate density of thickened connective tissues we observe radiologically at these interfaces. Finally, the entheses themselves, whether fibrous or fibrocartilaginous, are tissues responsive to the GH‐IGF‐1 axis and are implicated in human adult idiopathic hyperostosis (e.g., DISH, Littlejohn, 2024). Investigation of the development and change of bGH mouse periosteum, skeletal muscle, and connective tissues more generally over the lifespan will require transcriptional and histological sampling and gross anatomical validation at areas of interest identified by this pilot study, illuminating a path toward the ultimate influences on whole‐bone morphology in this and other enhanced GH systems.

4.3. Sesamoids and inferred heterotopic calcification of connective tissues within and around joints

In this study, we observe an increase in the volume of bone‐dense composition of surveyed sesamoids in the knee of bGH mice. Further, we find that this enlarged size coincides with changes in relative proportions among the sesamoids when compared with WT mice as well as variability in the total bone‐dense volume. The stark differences correspond to a systemic response from a mature musculoskeletal system that is substantially different from WT both biomechanically and metabolically. Variation in the structural properties of sesamoids within tendons and their binding materials relates to the biomechanical demands of the joints around which they develop and facilitate movement (Alexander & Dimery, 1985; Benjamin & Ralphs, 1998; Sarin & Carter, 2000). Because the relative composition and types of cartilage, bone, collagen, and ECM are subject to change over lifespan due to aging and through differential mechanoreception pathways, properties can also be altered by inactivity, defect, or injury (Benjamin et al., 1993; Benjamin et al., 1995; Vailas et al., 1985). Sesamoids typically develop within tendon via Sox9 ⁺/Scx ⁺ chondroprogenitors (Eyal et al., 2019) but thereafter are modulated via a broad range of genetic and epigenetic factors (Abdala et al., 2019) to ultimately become either fibrocartilaginous or bony. Some, such as the patella, are broadly present across mammals, while others vary widely in presence, size, and tissue composition among species (Samuels et al., 2017) and even within populations in a given species (i.e., the human fabella [Berthaume & Bull, 2019; Berthaume et al., 2019; Goldberg & Nathan, 1987; Zeng et al., 2012]), suggesting a clear life history/functional signal in addition to lineage‐specific or phylogenetic influences.

In this context of evolutionary, developmental, and idiopathic plasticity, excessive GH effects at both the molecular and organismal scale in the bGH model likely alter not only the sesamoid morphology we document in this study but other tissue interfaces related to joints as well. Indeed, our study of Achilles tendon tissue in bGH mice (Liu et al., in revision) identified IHC radiographically (using the same method as this study) in 3‐month‐old mice, in contrast to WT mice that only developed minimal bone‐dense volumes in the same region by 15 months of age. Safranin O staining confirmed the presence of chondral nodules within the tendon, and immunohistochemical staining confirmed a protein expression shift from tenogenic to chondrogenic and osteogenic phenotypes. At the structural level, this manifested with significantly different collagen fiber diameters and distributions relative to WT mice. Taken together, these observations indicate that excessive GH affects both tendon integrity and the associated sesamoid structures that facilitate joint movement. This may represent a shared biological response aimed at stabilizing aged and/or mechanically weak regions within the tendon system under conditions of altered muscle activity and loading, as fibroblast‐lineage progenitor cell populations undergo chondrogenic and osteogenic transformations in this nonhomeostatic environment (Plikus et al., 2021).

In addition to modifications in cartilage, bone, and tendon boundaries relative to WT counterparts, the locations of other IHC/RST volumes revealed in the μCT data implicate GH/IGF‐1 axis impacts to other soft tissues around appendicular joints including ligament, meniscus, and retinaculum, as well as those within and ventral to intervertebral joints. The limitations of our study (i.e., radiographic and at one time point) do not allow us to fully understand as‐yet the nature or progression of these soft tissue structural changes, but our recent site‐specific investigations of Achilles tendon (Liu et al., in revision) and intervertebral disc (Sukul et al., 2025) histology and mRNA work confirm fibrotic remodeling and heterotopic calcification in bGH mice manifest at least in part through altered transcriptional programs. Fibrosis in adult bGH mice is systemic (Lach et al., 2026) and known to occur in cardiac, liver, and kidney tissues (Kopchick et al., 2022), but significantly, we find no clear or prevalent relative radiodensity anatomically coincident with thickened tissue within these organs. This adds further (though indirect) evidence that a generalized process of fibrosis with subsequent tissue remodeling that specifically incorporates a radiodense component (i.e., heterotopic calcification in the form of chondrification and ossification) is a system‐specific manifestation of excessive growth hormone on musculoskeletal connective tissues.

Heterotopic calcification can be triggered by trauma, inflammation, or aberrant signaling. bGH mice are known to have chronic inflammatory responses (Ding et al., 2013; Jara et al., 2014), and recent histological and molecular investigations by our group on bGH mice demonstrate heterotopic calcification in knee joint tissues (Zhu et al., 2023) that is coincident with our observed regions of radiodensity (Video S5). The presence (and in the case of the knee, abundance) of this radiodense tissue around joints also implies that mechanoreceptive stimulus–response may play a role in driving this pattern in other tissue types within and around the joint capsule. Indeed, ligaments associated with the vertebral column are known to calcify and ossify in patients with acromegaly (Hoshino et al., 2022; Yoshizawa et al., 2023), lending further support to our observations that excessive GH production and its downstream effects on growth and metabolism may be a potent driver of calcification broadly throughout non‐bone connective tissues of the postcranial skeleton. As recently elucidated in a review by Littlejohn (2024) of diffuse idiopathic skeletal hyperostosis (a condition resulting in ossification around entheses in which growth factors including GH are implicated), mesenchymal stem cells in the perientheseal periosteum that can undergo chondrogenic and/or osteogenic transformations are proposed as the primary driver of calcification in human spinal ligaments and the annulus fibrosis, as well as in tendons and aponeuroses. These calcifications are also radiographically detectable. If this supposition holds true in animals other than humans, our results suggest that the bGH model might have preclinical utility in understanding the role of GH, IGF‐1, and growth factors in this and other diseases in which calcification impacts soft tissue.

4.4. GH effects beyond the model system (evolution, development, sexual dimorphism)

There remains a need to understand fundamental aspects of the skeletal biology and pathophysiology of GH, GHR, and especially IGF‐1 within naturally occurring nonhuman vertebrate populations (Lodjak & Verhulst, 2020), both in the context of genome evolution (e.g., gene duplication and deletion dynamics) and in the specific gene expression and modulation that may underlie bone shape, structure, and maintenance. Satisfying this need could, in turn, inform questions of body size disparity and diversification dynamics at larger biological and evolutionary scales. There are many examples of strong sexual body size dimorphism within individual species of vertebrates, as well as investigations into how specific expression/modulation dynamics of GH/GHR and the GH/IGF‐1 axis influence body size disparity within a single species or across species (e.g., Dantzer & Swanson, 2012; Liu et al., 2001; Luo et al., 2024; Rasband et al., 2023; Sutter et al., 2007). The results presented here in a mouse model system provide proof of concept for further consideration of how alteration of GH/GHR (and by extension, IGF‐1) may provide a mechanistic basis for species and clade‐level disparity in body size across multiple vertebrate lineages. Future work should target genomic‐ and expression‐specific variation that potentially underlies both within‐species (i.e., normal size variation, sexual size dimorphism) and among‐species disparities in skeleton morphology.

5. CONCLUSION

We demonstrate widespread patterns of gross anatomical differences in the morphology of the adult axial and appendicular skeleton of the bGH model and its associated connective tissues. Some caution regarding extrapolation to more general trends in the genotype is warranted given the small sample size and single time point, especially as neither of our male nor female bGH samples achieve statistically‐differentiable masses as reported for samples in previous studies. Nevertheless, we find whole‐skeleton size (by linear measure) and morphological differentiation in the bGH model, including within‐genotype sexual dimorphism and variability in documented morphology among bGH female individuals not seen in other genotype × sex groupings. Enhanced and relatively robust muscle scarring on appendicular elements, as well as the prevalence of radiodense soft tissues that are most likely explained as heterotopic calcification within connective tissues generally, suggest that the bGH bony phenotype is driven by organism‐scale biomechanical differences as well as intrinsic metabolic pathways. Study at earlier timepoints will be required to further evaluate how factors that relate to general processes of aging play into the observed differences. In a clinical context, these findings support the use of the bGH model to better understand idiopathic development of calcification in human non‐bony tissues, as well as further emphasizing the need to identify early life stage, presymptomatic features of GH overproduction to inform potential intervention strategies. Systemic GH‐induced alterations to adult mouse bone size and shape, different patterns of limb bone growth plate condition, the differential enlargement and proportions of sesamoids, and the densification of associated soft tissues also provide a keyhole into how GH and its metabolic cascade may affect vertebrate morphological and histological development, disparity, and dimorphism more broadly.

AUTHOR CONTRIBUTIONS

Joseph Groenke: Conceptualization; data development and curation; investigation; figures; measurements; writing and editing. Huanhuan Liu: Funding acquisition; writing, review, and editing. Grace Lach: Original histological experimental supervisor; data collection; review and editing. Abhishek Wajpe: Statistical analysis; review and editing. Darlene A. Berryman: Provided samples; review and editing. John J. Kopchick: provided samples; endocrinology input; review and editing. Shouan Zhu: Funding acquisition; writing, review, and editing. Patrick O'Connor: conceptualization; funding acquisition; investigation; project administration; figures; formal analysis; writing, review and editing.

FUNDING INFORMATION

Scanning was funded by the Research & Scholarly Awards Committee (RSAC) Award (RP2405) to HL, JRG, and PMO from Ohio University and National Institutes of Health grant R01AR081804 (SZ, HL, JK). JK is supported by the state of Ohio Eminent Scholars' Program that includes a gift by Milton and Lawrence Goll.

Supporting information

FIGURE S1 Comparative anatomy of midline (axial skeleton and pelvis) of male (M) and female (F) wild‐type (WT) and growth‐hormone model transgenic (bGH) mice. (A) Longitudinal (mediolateral) slice view of generally unfused condition of S3–S4 vertebral centra in a WT M (above), and a partially fused S3–S4 joint in a bGH male (below), with red asterisks indicating IHC or RST present at the S4‐Ca1 and Ca1–Ca2 joints. (B) Axial (craniocaudal) slice view of normal hemal arch canal in a WT M (left), with IHC/RST occluding the canal in a bGH M (right). Dashed circle indicates passage for caudal vein. (C) Morphology of the sternal elements of WT (top) and bGH (bottom) female mice in ventral view, with left costal cartilages digitally removed (cranial to right in image). (D) Differences in rib cage morphology and radiodensity between WT and bGH female mice in right lateral view, with black arrows indicating regions of IHC/RST. (E) Typical pelvic morphology (left) contrasted with aberrant (right) pubic symphyseal morphology in bGH female mouse in posteroventral view. S4 and proximal caudal vertebrae digitally removed to provide visual clarity. All scale bars = 1 cm except (B) and (E) = 1 mm. Ca, caudal vertebra; H, hemal arch (chevron); L6, sixth lumbar vertebra; M, manubrium; S, sacral vertebra; St, sternal element; Xs, Xiphisternum. Numbers subsequent to Ca, S, and St pertain to segment levels with respect to caudal and sacral vertebrae and sternal elements.

TABLE S1. Morphosource DOIs for CT volumes and figured polygon outputs.

VIDEO S1. Examples of humeri and antebrachia from each of the four genotype × sex conditions through one full and synced rotation.

VIDEO S2. Examples of innominates from each of the four genotype × sex conditions through one full and synced rotation.

VIDEO S3. Examples of femora from each of the four genotype × sex conditions through one full and synced rotation.

VIDEO S4. Examples of knee joints from the four genotype × sex conditions through one full and synced rotation showing distal femur (at top) and proximal tibia‐fibula (bottom) in light gray, sesamoids in dark gray, radiodense portions of the meniscus in red, and other radiodense soft tissue as multicolored to display sizes, shapes, and positions of volumes around the joint.

VIDEO S5. Examples of bone and radiodense soft tissues associated with the knee joint in a bGH female. Left, proximal‐to‐distal progression of CT slice data through the femur, knee joint, and tibia‐fibula (cranial to bottom of video). Right, synced transition of the same knee in cranial view, with a volume rendering of the knee below the transition and a segmented polygon model above the transition showing bone in gray, sesamoids in dark gray, meniscus in red, and other radiodense soft tissues in multiple colors to display sizes, shapes, and positions of the volumes around the joint.

DATA AVAILABILITY STATEMENT

Volumetric datasets of all 17 specimens used for the study, as well as relevant polygon meshes representing digital dissections of morphology used in figures, are available for interactive browsing and download on the MorphoSource website (www.morphosource.org) organized under a project entitled “Growth Hormone Skeletal Phenotyping Project (https://www.morphosource.org/projects/000790866?locale=en).” Table S1 provides digital object identifiers (DOIs) for these data. Other polygon models derived from the volumetric data as well as segmentation (label) files used to create them can be made available upon request, as can details of linear measurement positions (including screen captures) and segmentation and long bone cross‐sectional measurement workflows.