Recombinant IGF-1 Induces Sex-Specific Changes in Bone Composition and Remodeling in Adult Mice with Pappa2 Deficiency

Deficiency of pregnancy-associated plasma protein-A2 (PAPP-A2), an IGF-1 availability regulator, causes postnatal growth failure and dysregulation of bone size and density. The present study aimed to determine the effects of recombinant murine IGF-1 (rmIGF-1) on bone composition and remodeling in constitutive Pappa2 knock-out (ko/ko) mice. To address this challenge, X-ray diffraction (XRD), attenuated total reflection-fourier transform infra-red (ATR-FTIR) spectroscopy and gene expression analysis of members of the IGF-1 system and bone resorption/formation were performed. Pappa2ko/ko mice (both sexes) had reduced body and bone length. Male Pappa2ko/ko mice had specific alterations in bone composition (mineral-to-matrix ratio, carbonate substitution and mineral crystallinity), but not in bone remodeling. In contrast, decreases in collagen maturity and increases in Igfbp3, osteopontin (resorption) and osteocalcin (formation) characterized the bone of Pappa2ko/ko females. A single rmIGF-1 administration (0.3 mg/kg) induced short-term changes in bone composition in Pappa2ko/ko mice (both sexes). rmIGF-1 treatment in Pappa2ko/ko females also increased collagen maturity, and Igfbp3, Igfbp5, Col1a1 and osteopontin expression. In summary, acute IGF-1 treatment modifies bone composition and local IGF-1 response to bone remodeling in mice with Pappa2 deficiency. These effects depend on sex and provide important insights into potential IGF-1 therapy for growth failure and bone loss and repair.


Introduction
The growth hormone (GH)/insulin-like growth factor 1 (IGF-1) signaling system constitutes a pleotropic axis required for bone development, mineral deposition and skeletal growth [1][2][3][4][5]. IGF-1 exerts its effects on skeletal growth and metabolism by interacting with other bone regulators like thyroid and parathyroid hormones and sex
Overall, the effects of genotype on the femur and tibia length, femur and tibia weight, and relative weights of tibia were found ( Table 1). The overall effects of sex on the femur length and weight, tibia weight and relative weights of femur and tibia were also observed (Table 1), with females having overall lower values than males ( ### p < 0.001). Significant interactions between genotype and sex were found in femur length and weight, tibia weight and femur/body weight ratio of the adult mice (Table 1), with males showing greater effects of Pappa2 deletion ( #/### p < 0.05/0.001) than females (*** p < 0.001).

Pappa2 Deletion Alters Hydroxyapatite Crystallinity in the Femur of Male Mice, But Not Female Mice
Crystallographic indexes of hydroxyapatite-(CaOH) found in the femurs of adult mice were quantified by XRD and Rietveld refinement. The overall effects of genotype on hydroxyapatite crystallinity and crystallite size were found (Table 2). Sex effects on hydroxyapatite crystallinity were also observed, with females having a higher index than males ( ## p < 0.01). A significant interaction between genotype and sex was found in hydroxyapatite crystallinity (Table 2), with a significant increase in Pappa2 ko/ko males compared with Pappa2 wt/wt males ( ## p < 0.01). No difference between WT (wild-type) and KO (knock-out) females was found. Crystallite size was significantly lower in Pappa2 ko/ko mice ( # p < 0.05). No interactions between factors were observed for crystallite size (Table 2) or other parameters such as R-Bragg factor, cell volume, crystal linear absorbance coefficient or crystal density (Table S1).  Figure S1A for representative diffractograms and Table S1 for additional information. ns, not significant.

Pappa2 Deletion Alters Bone Composition in the Femur of Male Mice, but Not Female Mice
Relative contents of compounds containing phosphates (v 1 v 3 PO 4 3− ), carbonates (v 2 CO 3 2− ) and amides I (vC=O) in the femurs of adult mice were identified by ATR-FTIR spectroscopy, and quantified to calculate relevant parameters related to bone composition, such as mineral-to-organic matrix (M/M) ratio, an index of the relative amount of phosphate per amount of collagen; carbonate substitution (C/p ratio), an index of phosphate-to-carbonate-substituted apatites; mineral crystallinity, a degree of order in a solid; and collagen crosslink ratio, an index of collagen maturity. The overall effects of genotype on M/M ratio, C/P ratio, mineral crystallinity and collagen maturity were found (Table 3), with Pappa2 deletion inducing a higher M/M ratio ( # p < 0.05 in males) and lower C/P ratio ( ## p < 0.01 in males), mineral crystallinity ( # p < 0.05 in males) and collagen maturity ( ##/&&& p < 0.01/0.001 in both sexes) than in respective WT mice. The overall effects of sex on the M/M ratio, C/P ratio and collagen maturity were observed (Table 3), with females having an overall higher M/M ratio ( # p < 0.05) and lower C/P ratio ( # p < 0.05) and collagen maturity ( ### p < 0.001) than males. Interestingly, significant interactions between genotype and sex were found in mineral crystallinity, and reflected significant decreases in Pappa2 ko/ko males compared with controls ( # p < 0.05) and no significant effect of Pappa2 deletion in females (Table 3).  Figure S1B,C for representative spectra and Table S2 for additional information. 2 Mineral-to-organic matrix ratio: Amount of mineral (phosphate) per amount of organic matrix (collagen) per volume analyzed. Ratio of phosphate peak area (v 1 v 3 PO 4 3− : A900-1200 cm −1 ) and amide I peak area (vC = O: A1585-1720 cm −1 ). 3 Carbonate substitution: Relative amount and type of carbonate substitution in the bone mineral apatite lattice. Ratio of carbonate peak area (v 2 CO 3 2− : A850-890 cm −1 ) and phosphate peak area (v 1 v 3 PO 4 3− : A900-1200 cm −1 ). 4 Mineral crystallinity: Transformation of non apatitic domains into apatitic ones. Ratio of absorbance height at 1030 cm −1 and 1020 cm −1 after the application of second derivatives. 5 Collagen maturity: A measure of collagen crosslink ratio of pyridinium (an older, trivalent collagen crosslink) at 1660 cm −1 to dehydrodihydroxylysinonorleucine (a younger, divalent, and freshly synthesized collagen crosslink) at 1690 cm −1 after the application of second derivatives.

rmIGF-1 Induces Sex-Specific Changes in Bone Composition of Pappa2 ko/ko Mice
Short-term effects of rmIGF-1 on bone composition were also evaluated over time (0, 30, 120 and 240 min) in male and female femurs ( Figure 1). The overall effects of genotype, sex and time on M/M ratio, C/P ratio, mineral crystallinity and collagen maturity were detected (F 1.114 > 18.67, p < 0.001). Significant interactions between genotype, sex and time were found in M/M ratio, C/P ratio, mineral crystallinity and collagen maturity (F 3.114 = 3.50, p = 0.02; F 3.114 = 4.04, p = 0.009; F 3.114 = 5.58, p = 0.001; F 3.114 = 30.85, p = 0.001, respectively), suggesting that rmIGF-1 treatment affected bone composition depending on sex and genotype over time.
Ratio of absorbance height at 1030 cm −1 and 1020 cm −1 after the application of second derivatives. 5 Collagen maturity: A measure of collagen crosslink ratio of pyridinium (an older, trivalent collagen crosslink) at 1660 cm −1 to dehydrodihydroxylysinonorleucine (a younger, divalent, and freshly synthesized collagen crosslink) at 1690 cm −1 after the application of second derivatives.

Pappa2 Deletion Affects the Gene Expression of Bone Remodeling Markers in the Tibia
To further understand the effects of Pappa2 deletion and local IGF-1 system dysregulation on bone composition, we evaluated bone turnover or remodeling (resorption/formation) by analyzing relative mRNA expression of osteopontin, osteocalcin and Col1a1 in the tibia ( Table 5). The overall effects of the genotype on the mRNA levels of osteocalcin (bone resorption) were found, with Pappa2 deficient mice having overall higher expression than WT (Table 5). Significant interactions between genotype and sex were found on the mRNA levels of osteopontin and osteocalcin (bone formation), and reflected significant increases in Pappa2 ko/ko females compared with controls (* / **p < 0.05/0.01).  In males, treatment with rmIGF-1 decreased the mRNA levels of Col1a1 and osteopontin in the tibia of both genotypes over time (30′, 120′ and 240′ versus 0′: * / ** / *** p < 0.05/0.01/0.001; Figure 3A,B). In contrast, rmIGF-1 treatment increased the mRNA levels of osteocalcin in the tibia of Pappa2 ko/ko males, but not Pappa2 wt/wt males, at 30′ and 240′ after In males, treatment with rmIGF-1 decreased the mRNA levels of Col1a1 and osteopontin in the tibia of both genotypes over time (30 , 120 and 240 versus 0 : * / ** / *** p < 0.05/0.01/0.001; Figure 3A,B). In contrast, rmIGF-1 treatment increased the mRNA levels of osteocalcin in the tibia of Pappa2 ko/ko males, but not Pappa2 wt/wt males, at 30 and 240 after administration (30 and 240 versus 0 : * / ** p < 0.05/0.01; Figure 3C). No sustained differences between genotypes (Pappa2 ko/ko versus Pappa2 wt/wt males) were found in the mRNA levels of the three bone remodeling markers.

Discussion
In the present study, we showed that acute treatment with recombinant IGF-1 modulates bone composition and remodeling in Pappa2-deficient mice, in accordance with human clinical studies [25][26][27][28][29][30][31][32][33][34]. The effects of Pappa2 deficiency itself, as well as the short-term response to rmIGF-1 treatment on bone composition parameters, and the expression of relevant components of local IGF-1 system (Igfbp3, Igfbp5, Igfals) and bone remodeling (Col1a1, osteopontin, osteocalcin) were shown to be sex-dependent. Bone composition was analyzed using physical-chemical approaches to detect changes in bone crystallinity and relevant parameters (mineral-to-collagen matrix ratio and carbonate substitution) that reflect the relative content of bone compounds containing phosphates (v 1 v 3 PO 4 3− ), carbonates (v 2 CO 3 2− ) and amides I (vC=O), including mineral crystallinity (a degree of order in a solid) and collagen maturity (collagen crosslink ratio) in bone. The main results of the present study are as follows: (1) Pappa2 KO in mice of both sexes reduces body length, and bone length and weight, as expected from previous studies [30][31][32][33]; (2) Pappa2 KO in mice alters crystallographic parameters (hydroxyapatite crystallinity and crystallite size) in the femur of males, but not in that of females; (3) Pappa2 KO in mice impairs the M/M ratio and carbonate substitution, as well as mineral crystallinity in the femur of males, but not in females; (4) Pappa2 KO in mice of both sexes reduces collagen maturity; and (5) Pappa2 KO in female mice specifically increases the mRNA expression of Igfbp3, osteopontin (a marker of bone resorption) and osteocalcin (a marker of bone formation) in the tibia, but not in that of males. These results suggest that Pappa2 deficiency alters bone length, weight and composition, probably through changes in bone remodeling, in a sex-dependent manner.
A single administration of rmIGF-1 to adult mice induced short-term effects on bone composition and remodeling in a sex and genotype-dependent manner. The main results are as follows: (1) There is an increase in the M/M ratio and decreases in carbonate substitution and mineral crystallinity in Pappa2 KO mice of both sexes (Pappa2 WT mice showed a short transitory effect of rmIGF-1 on these parameters); (2) Administration of rmIGF-1 resulted in a specific increase in collagen maturity in Pappa2 KO females (no effect of rmIGF-1 in Pappa2 WT females); and (3) Specific increases in the mRNA expression of Igfbp3, Igfbp5, Col1a1 and osteopontin in Pappa2 KO females (no effects of rmIGF-1 in Pappa2 WT females) were also observed. Except in the higher expressions of Igfbp3 in Pappa2 KO males, no further differences between genotypes in the local IGF-1 system or bone remodeling markers were found in males.
These results suggest that rmIGF-1 treatment at a dose of 0.3 mg/kg may facilitate changes in bone composition, as addressed by parameters that reflect relative amounts of bone compounds containing phosphates, carbonates and amides I, which in turn, are controlled by local IGF-1 signaling and molecular mechanisms of bone mass homeostasis, such as a balance between bone resorption and formation. Some of these responses to exogenous IGF-1 were genotype-dependent. For instance, specific increases in Igfbp5 and/or Igfbp3 in the bone of Pappa2 KO mice likely reflect the requirement for PAPP-A2 in this response, as the exogenous IGF-1 can become locally bound in ternary complexes. Clinically, these results further contribute to our understanding of the therapeutic efficacy of recombinant IGF-1 in patients who display a novel syndrome characterized by short stature, skeletal abnormalities and increased formation of ternary complexes due to mutations in PAPP-A2 resulting in low IGF-1 bioactivity [25,[27][28][29].
Many of the results observed here were sex-dependent and sex-specific regulation of bone properties and skeletal growth by GH/IGF-1 signaling system [5,7,[35][36][37], including IGFALS, IGFBP-2, IGFBP-5, IGFBP-4, PAPP-A and PAPP-A2, have been reported [20,33,[38][39][40][41]. Sex differences were also observed in basal bone physiology of Pappa2 deficient mice. Sex effects reflect an overall lower expression of Igfbp5 and Igfals, likely resulting from restricted IGF-1 production, could participate in the shorter bone length (12% less) and lower weight (34% less) compared to male bone. Sex effects are also associated with an overall difference in hydroxyapatite crystallinity, as well as M/M ratio, carbonate substitution and collagen maturity parameters, which were calculated from the relative content of bone compounds containing phosphates (v 1 v 3 PO 4 3− ), carbonates (v 2 CO 3 2− ) and amides I (vC=O). It is clear that sex steroid hormones are implicated in the differences between males and females in expression levels of components of the GH/IGF-1 axis and skeletal structure [7,42]. In humans, sex steroids contribute to the differences in pubertal growth that are mainly caused by greater periosteal expansion, smaller marrow diameter and greater longitudinal growth velocity and bone mass in males, compared to females [43][44][45], as well as to the earlier pubertal rise in GH and IGF-1 in girls than in boys and protection from age-related bone loss [6,42]. Consequently, bone strength shows a greater gain during postnatal growth and less decline during aging in males than in females [7,46]. In this context, we propose that sex-specific control of IGF-1 bioavailability through regulation of IGF ternary complexes could modify bone composition and at least partially explain the lower mechanical strength in the female bone [47]. Future studies should elucidate whether PAPP-A2 deficiency compromises the effects of estrogens and androgens in bone remodeling during growth and aging.
Constitutive and osteoblast-specific deletion of Pappa2 in mice has been described to affect postnatal skeletal growth, including bone mineral density, in a sex and agedependent manner [31,32]. Interestingly, the bone of male Pappa2 ko/ko mice has been described to be more similar to that of WT females than WT males [31]. Compared to WT males, our results indicated that the femur of Pappa2 ko/ko males was shorter (up to 22% less), weighed less (up to 34% less), had a smaller crystallite size and decreases in carbonate substitution (relative amount of carbonates), mineral crystallinity and collagen maturity, and exhibited increases in sample crystallinity and M/M ratio (relative amount of phosphates). However, these alterations in bone composition in the femur of Pappa2 ko/ko males were not accompanied by significant changes in the local IGF system or bone remodeling. In contrast, the significant reduction in bone length and weight (15-18% less) in Pappa2 ko/ko female mice was associated with lower collagen maturity and a higher expression of Igfbp3. In this regard, bone turnover mechanisms could involve circulating factors such as IGFBP-3 or IGFBP-5 [33]. However, unlike Igfbp3, we found no genotype effect in the bone expression of Igfbp5 and Igfals, which is not consistent with previously described changes in its circulating levels [33,34]. These authors reported that Pappa2 ko/ko mice of both sexes had higher serum levels of IGFBP-5 than their WT counterparts at 19 weeks, although the difference was not significant at 30 weeks [33]. Additionally, in a report where male and female mice were not studied separately [34], circulating levels of IGFBP-5 and IGFBP-3 were higher and lower, respectively, in the serum of Pappa2 ko/ko mice at six weeks of age. These results are in partial contrast with those in male Pappa2 ko/ko mice showing increased Igfbp3 expression in bone, as shown here, and those in patients with PAPP-A2 deficiency showing increased circulating levels of IGFBP-3 [28]. This apparent contradiction could reflect differing roles and regulation of IGFBPs in endocrine versus autocrine/paracrine signaling [20,[48][49][50][51][52]. Indeed, no effects of Pappa2 deletion on the expression of these factors in the liver or kidney were described [34], suggesting tissue specific regulation of some members of the local IGF system. Moreover, locally generated IGF-1 can regulate bone growth in response to GH actions in liver-specific GH-receptor KO mice [53].
Molecular mechanisms balancing bone resorption by osteoclasts and bone formation by osteoblasts reflect the rate of bone turnover and remodeling [9]. In the present study, bone expression of osteopontin and osteocalcin (bone markers implicated in bone resorption and bone formation respectively) were increased in Pappa2 ko/ko females. However, no change was found in Col1a1 expression, a major component of type I collagen that strengthens bone. These results suggest that higher bone turnover may also associate with lower collagen maturity in the bone of Pappa2 ko/ko females. Although circulating levels of bone remodeling markers were not measured here, and these results contrast with the lower circulating levels of other markers of bone resorption (TRACP 5b) and bone formation (PINP), previously described in the plasma of female mice with Pappa2 deletion [33]. Again, this could reflect differences between local and circulating mechanisms controlling bone mass homeostasis. Bone turnover markers can also be modified by factors, such as feeding (lower resorption), bone loss and fractures (increased levels of all markers), and sources other than bone, such as platelets, liver, lungs and heart, can contribute significantly to circulating levels [54]. Higher levels of all bone turnover markers are highly correlated with the increased rate of bone loss in women [54][55][56], as observed here in Pappa2 ko/ko females. Overall, our data suggest that an elevated rate of bone turnover, including lower collagen crosslink ratio, is a key determinant of bone immaturity in Pappa2 deficiency.
The changes in bone mineral composition, resulting from ionic substitution in biological hydroxyapatite, can induce alterations in hexagonal architecture, collagen structure and crystallinity affecting bone resistance or fragility [57][58][59][60][61][62][63][64][65]. Phosphate and carbonate substitutions in the apatite structure are responsible for changes in the degree of crystallinity, weakening the bonds and increasing mineral solubility [62,63]. Whereas, amide content in collagen-containing structures confers flexibility [64]. Carbonated hydroxyapatite contributes to a critical structure that affects mechanical strength of bone [65][66][67]. Here, the higher crystallinity and smaller crystallite size of the resulting hydroxyapatite are accompanied by a higher mineral-to-matrix ratio (relative amount of phosphates) and lower carbonate substitution and mineral crystallinity (related to crystallite size and perfection) in the femur of both WT females and Pappa2 ko/ko males, compared to WT males. This is consistent with previous reports showing lower bone mineral density in Pappa2 transgenic mice in a sex-dependent manner [33]. Clinically, the fragility and fracture risk, related to untreated osteoporotic bone are associated with increases in cancellous crystallinity and mineral-to-matrix ratio [54,68]. Elevated levels of carbonate substitution were also found in fracture and osteoporosis cases [68]. Together, our data suggest that Pappa2 deletion results in the alteration of key parameters of bone composition that could denote a level of bone fragility.
In patients with PAPP-A2 deficiency recombinant human IGF-1 (rhIGF-1) was employed [28,29], but recombinant human PAPP-A2 could be a promising therapy [69]. Benefits of rhIGF-1 treatment include improved growth and increased bone mineral density and trabecular structure [28,29]. The administration of rhIGF-1 improves bone formation in aged mice [35], an effect that was markedly potentiated when combined with IGFBP-3 in ovariectomized rats [70], suggesting possible indications of long-term treatment, even after adult height is reached. In the present study, acute rmIGF-1 treatment induced a sex-and genotype-specific effect on bone composition, mostly reflected in short-term substitutions of the ionic content of phosphates, carbonates and amides I. In Pappa2 ko/ko mice (males and females), rmIGF-1 increases the M/M ratio and decreases the carbonate substitution and mineral crystallinity over time, likely increasing bone strength and stiffness, while ductility decreased [61,62,68,71]. These rmIGF-1 induced changes in bone composition also include an increase in collagen maturity of the bone of Pappa2 ko/ko females specifically, as this effect was not observed in Pappa2 wt/wt females or Pappa2 ko/ko males. Moreover, rmIGF-1-induced changes in the bone composition of Pappa2 ko/ko females was accompanied by higher expression of Igfbp3 and Igfbp5, suggesting an acute response to increased IGF-1 bioavailability and signaling in a context of Pappa2 deficiency. Interestingly, higher expression of osteopontin and Col1a1 that was also specifically found in Pappa2 ko/ko females after rmIGF-1 treatment, suggests an up-regulation of bone matrix resorption/formation. In this case, the concurrence of both higher collagen maturity and increased Col1a1 expression in the bone of rmIGF-1-treated Pappa2 ko/ko females may trigger the correct collagenous matrix formation that can underlie bone strength [9]. The short-term modulation of bone mineral composition, the local IGF-1 system and bone remodeling in a sex and genotype-dependent manner may provide relevant insights into the therapeutic efficacy of recombinant IGF-1. Indeed, the increased expression of Igfbp3 and Igfbp5 in females with Pappa2 deletion, in response to rmIGF-1 treatment, may compromise long-term IGF-1 bioavailability. It could also be associated with an up-regulation of bone remodeling as assessed by increased expression of osteopotin and Col1a1.
In summary, our results support sex-specific regulation of bone composition and remodeling by PAPP-A2. Moreover, Pappa2 deletion altered the response to rmIGF-1 treatment, as indicated by parameters of bone mineral content, the local IGF-1 system and bone resorption/formation, in a sex specific manner. Collectively, our results support the therapeutic impact of IGF-1 to improve bone strength and density, but suggest that its clinical efficacy may differ depending on the sex of the patient with postnatal growth deficiency. However, the scarce experience treating patients with PAPP-A2 deficiency seems to indicate that both sexes respond adequately to this treatment [28].

Materials and Methods
All procedures were conducted in strict adherence to the principles of laboratory animal care (National Research Council, Neuroscience CoGftUoAi, Research B, 2003) following the European Community Council Directive (86/609/EEC) and were approved by the Ethical Committee of the University of Málaga (Ref. [69]-2016H). Special care was taken to minimize the suffering and number of animals necessary to perform the procedures.

Animals
Adult male and female mice (C57BL/6 background) with constitutive Pappa2 gene deletion (Pappa2 ko/ko ) and littermate controls (Pappa2 wt/wt ) were generated as previously described [31]. Mice were housed on a reverse 12-h light/dark cycle (lights off at 8:00 a.m.) in a humidity and temperature-controlled (22 ± 1 • C) vivarium. Standard rodent food and tap water were available ad libitum. Body weight and body length (including the tail) were monitored. Mice were genotyped by PCR using ear-clip tissue, as previously described [34].

Drugs
Recombinant murine IGF-1 (rmIGF-1; cat. no. 250-19; PeproTech, Inc., Rocky Hill, NJ, USA) was prepared by dissolving in 0.9% saline. The rmIGF-1 solution was injected subcutaneously once at a dose of 0.3 mg/kg in a volume of 2 mL/kg. Pappa2 wt/wt and Pappa2 ko/ko mice (males and females) were sacrificed 30, 120 and 240 min after rmIGF-1 administration. The route, dose and timing were selected based on previous studies of bone formation in aged mice [72].

Sample Collection
Adult mice (8 months of age) were sacrificed by decapitation after the administration of Equitesin ® (3 mg/kg). Both femurs and tibias were extracted, cleaned of adjacent soft tissue, and immediately frozen in liquid nitrogen. Bones were weighed, and their length measured using a caliper. Femur epiphysis and metaphysis were removed and diaphysis were isolated. Frozen bone collections were stored at −80 • C until XRD, ATR-FTIR and RT-qPCR analyses.

Sample Preparation
Both femur diaphysis per mouse were pooled and pulverized in liquid nitrogen using a 6770 Freezer Mill (SPEX CertiPrepFreezerMill, Stanmore, London, UK). The cryogenic milling was carried out under mild conditions (cycles: 2; run time: 2 min; rate: 9 cps) to avoid altering the crystallinity of the materials or the spectral levels of the compounds under study. The resulting powder (50-100 µg particle size) was collected (~250 mg) and kept in a −80 • freezer until XRD and ATR-FTIR analyses.

X-ray Powder Diffraction
Each femur diaphysis sample (~100 mg) was analyzed using an Empyrean Malvern Panalytical automated X-ray diffractometer (Malvern Panalytical, Malvern, United Kingdom) and Rietveld refinement [73][74][75][76]. The patterns of sample crystallinity and crystallite size were collected with a step size of 0.017 • (2θ) and 300 sec/step using Cu-Kα (λ = 1.540598 Å) radiation from a tube operated at an accelerating voltage of 45 kV and a current of 35 mA. The (002) peak was baselined from 4 • to 80 • (2θ) for 30 min and fitted with a Lorentzian curve to determine the peak broadening based on its full width at half maximum ( Figure S1A). Identification of amorphous phase and pure crystalline material was performed with reference to an external standard and the database supplied by the International Centre for Diffraction Data (Powder Diffraction File no. 84-1998), Inorganic Crystal Structure Database and Crystallography Open Database (COD no. 9010050; RRID:SCR_005874). Sample crystallinity (the degree of order in a solid) is defined as the quotient of enthalpy difference between pure amorphous phase and the sample enthalpy over the difference of pure amorphous and pure crystalline material (external standard). Percentage of crystallinity is calculated by: (total area of crystalline peaks) · 100/(total area of crystalline and amorphous peaks). The Scherrer equation (D v = K · λ/β002 · cosθ) and Williamson-Hall method were used to calculate crystallite size (LVol-IB, nm); where D v is the volume weighted crystallite size, K is the Scherrer constant with a value of 1, λ is the x-ray wavelength used, and β002 is the integral breadth of the (002) reflection or length of the apatite crystals along the c-axis. The R-Bragg factor, cell volume, crystal linear absorbance coefficient (1/cm) and crystal density (g/cm 3 ) were also checked. Three patterns were performed and a mean pattern was obtained for each sample.

ATR-FTIR Spectroscopy
The infra-red (IR) analysis of each femur diaphysis sample (~100 mg) was carried out in a Bruker Vertex 70 Fourier Transform (FT)-IR spectrophotometer (Bruker Corporation, Billerica, MA, USA). We worked with attenuated total reflectance (ATR) using a Golden Gate System of Individual Reflection [77][78][79]. The material of our internal reflection element was ZnSe (20,000-500 cm −1 ). For the acquisition of spectra, a standard spectral resolution of 4 cm −1 in the spectral range of 500-4000 cm −1 was used, as well as 64 accumulations per sample. The background spectrum in all cases was the air. For the analysis of the raw spectra, the v 1 v 3 PO 4 3− bands were baselined from 1200 to 900 cm −1 , the v 2 CO 3 2− band from 890 to 850 cm −1 , and the amide I band from 1730 to 1585 cm −1 . Spectral analysis was performed in triplicate and a mean spectrum was obtained for each sample ( Figure S1B). After curve-fitting of every individual (not smoothing) spectrum, position, height and area under the curves (baseline correction) were measured.
The following parameters that reflect the bone tissue compositional properties were calculated [80][81][82]: (1) Mineral-to-organic matrix (M/M) ratio, an index of tissue mineral content that characterizes the relative amount of phosphate per amount of collagen present, and is calculated by the ratio of the integrated areas of the respective raw peaks of v 1 v 3 PO 4 3− (900-1200 cm −1 ) and amide I (1585-1730 cm −1 ); (2) carbonate substitution (C/P ratio), an index of phosphate-to-carbonate-substituted apatites that characterizes the extent to which carbonate substitutes into mineral lattice, and is calculated by the ratio of the integrated areas of the respective raw peaks of v 2 CO 3 2− (850-890 cm −1 ) and v 1 v 3 PO 4 3− (900-1200 cm −1 ); (3) mineral crystallinity or maturity (1030/1020 cm −1 intensity ratio), a degree of order in a solid that is related to crystal size and perfection; and (4) collagen maturity (1660/1690 cm −1 intensity ratio), an index related to the ratio of mature, non-reducible collagen crosslinks to immature, reducible collagen crosslinks. We applied the second derivatives of the raw data from ATR-FTIR spectra to determine specific peaks at~1030,~1020,~1660 and~1690 cm −1 , and improve the accuracy of quantification of mineral maturity and collagen crosslink ratio ( Figure S1C).

Data Analysis
Data are presented as means ± S.E.M. and the "n" in figure legends indicates the number of animals per group. Data were normally distributed. For statistical analysis, we used GraphPad Prism 6.0 (GraphPad Software, San Diego, CA, USA) and IBM SPSS software 23.0 (SPSS Inc., Chicago, IL, USA) in order to apply two and three-way ANOVA (genotype, sex and time as factors) followed by Tukey-corrected tests or simple effect analyses where appropriate. A p < 0.05 indicates statistical significance.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/ijms22084048/s1, Figure S1: (A) Representative diffractogram of bone hydroxyapatite-(CaOH) in the femur of Pappa2 wt/wt and Pappa2 ko/ko mice (males and females) by X-ray powder diffraction (XRD). (B) Representative ATR-FTIR spectra of the raw data. The typical ATR-FTIR spectral bands (peaks) that define amide, carbonate and phosphate groups at specific wavenumbers were obtained. (C) Second derivatives of the ATR-FTIR spectra define the amide I band positions in the 1585-1730 cm −1 region (left) and the v 1 v 3 PO 4 3− band positions in the 900-1200 cm −1 region (right). The peaks at 1660 and 1690 cm −1 (red lines in the left) correspond to the pyridinoline and dehydrohydroxylysinonorleucine collagen crosslinks, respectively. The respective peaks at~960, 1020, 1030 and 1115 cm −1 (red lines in the right) correspond to v 1 PO 4 3− , PO 4 3− in stoichiometric apatites, HPO 4 2− and/or CO 3 2− in nonstoichiometric poorly crystalline apatites and other poorly crystalline apatites, respectively; Table S1: Quantitative analysis of crystallographic properties of bone hydroxyapatite-(CaOH) in the femur of Pappa2 wt/wt and Pappa2 ko/ko mice (males and females) by XRD; Table S2: Quantitative analysis of spectral levels (actual absorbance) of specific ionic contents in the femur of Pappa2 wt/wt and Pappa2 ko/ko mice (males and females). Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available on reasonable request from the corresponding author.