Effects of bisphosphonates on different zones of the epiphyseal growth plate of rats

Bisphosphonates (BIS) are indicated for several clinical disorders (e.g., osteoporosis). However, BIS has been associated with osteonecrosis and alterations in osteoclastogenesis and skeletal development. This study aimed to evaluate the effects of BIS (zoledronic acid ZA and alendronate sodium AS) on zones of the growth plate of rat femur. Animals (Wistar rats, n = 19) were divided into groups: 1) AS Group: animals received alendronate sodium orally (3 mg/kg per day); 2) ZA Group: ZA was administered intraperitoneally (0.2 mg/kg per week); and 3) Control Group (CG): a vehicle was administered. Animals were euthanized 21 days after the treatment, and femurs were collected for histological analysis. The images of all zones (resting, proliferative, hypertrophic, and calcified) were processed by the Qcapture® software providing a 40 and 400-fold increase. ZA decreased epiphyseal growth plate cell zones (ZA Group vs. CG) in most cases. Likewise, AS diminished the proliferative zone (AS Group vs. CG). Furthermore, ZA increased the calcified zone (ZA Group vs. CG). Previous works demonstrated that BIS decrease the epiphyseal disc. This reduction is probably due to the shortening of the cellular zones that undergoes calcification/ossification. The present results suggest that BIS should be carefully indicated because these drugs might accelerate epiphyseal closure.


Introduction
Bisphosphonates (BIS) are analogous to the pyrophosphate (Fernandes et al., 2005). They may be classified as aminobisphosphonates or non-aminobisphosphonates, depending on the presence or absence of nitrogen in the structure (Lin, 1996). Generally, BIS that contain nitrogen are more potent than non-aminobisphosphonates (Russell et al., 2008). Zoledronic acid (ZA) and alendronate sodium (AS) are the most prescribed aminobisphosphonates (Patntirapong & Poolgesorn, 2018).
These drugs are indicated for disorders of bone metabolism due to the ability to bind to the hydroxyapatite and inhibit osteoclast-mediated bone resorption (Dominguez et al., 2011). Bone remodeling occurs through resorption and formation events mediated by osteoclasts or osteoblasts, respectively (Mackie et al., 2011). Primarily, BIS are used to treat osteoporosis in the elderly population (Lin, 1996). However, BIS are also administered at different growth stages to treat pathological conditions, such as osteogenesis imperfecta (Biggin & Munns, 2017), Paget's disease (de Oliveira et al., 2019), and juvenile osteoporosis (Batch et al., 2003).
The clinical benefits of BIS have been associated with alterations in osteoclastogenesis, bone formation, and skeletal growth (Rezende et al., 2017). Nevertheless, the influence of BIS on bone growth is not fully known (Bianchi, 2005). Therefore, animal models might be a valuable tool to investigate the impact of BIS on the epiphyseal plate and, consequently, on human growth. Particularly, the choice of rats as an animal model may be relevant because similar methodologies have been applied to study different bone diseases (Erdogan et al., 2014;Oyhanart et al., 2015;Özenci et al., 2013). Furthermore, the closure of the growth plate does not usually occur in rats. However, skeletal growth slows down around 7 to 8 months of age in these rodents (Quinn, 2005).
The epiphyseal growth plate causes the longitudinal growth of the bones and is composed of specialized cartilage (Ballock & O'Keefe, 2003). This structure is divided into zones (i.e., resting, proliferative, hypertrophic, and calcified) corresponding to the stage of maturation (Wilsman et al., 1996). Previous studies have shown that BIS may modify the thickness of growth plate zones (Junges, 2013). Therefore, considering the numerous factors capable of altering bone structure during the BIS administration, the present study aimed to evaluate the effects of aminobisphosphonates (alendronate sodium and zoledronic acid) on the growth plate zones of the rat femur.

Methodology
This study had a quantitative approach and a basic experimental design (De Barros Silva et al., 2015;Huang et al., 2005).

Animals
The sample (n=19) consisted of one hundred-day-old male Wistar rats weighing 250-300 grams from the Hospital de Clínicas de Porto Alegre (Unit of Animal Experimentation). The rats were randomized by weight and allocated into different groups. Rats were kept in cages of polypropylene (49x34x16cm). The animal facility was maintained under controlled conditions: 12:12 light-dark (on at 7.00 a.m. and off at 7.00 p.m.), controlled temperature room (22±2ºC), and food (Nuvilab, Moinhos Purina, Porto Alegre, RS -Brazil) and water ad libitum. The experimental protocol complied with the ethical and methodological standards of the ARRIVE guidelines (Kilkenny et al., 2010). Attempts were made to minimize animal suffering and decrease pain and discomfort, as well as to use the minimum number of animals required to produce reliable scientific data. The Institutional Animal Care and Use Committee (IACUC) of Hospital de Clínicas de Porto Alegre (HCPA) previously had approved the experimental project (protocol N. 09-366). The experiments followed the Guide for the Care and Use of Laboratory Animals 8th edition 2011 and law 11.794 (Brazil), which establishes procedures for the scientific use of animals. The research leading to these results received funding from Institutional Scientific Initiation Scholarship Program (PIBIC) and Incentive Fund for Research and Events (FIPE/Hospital de Clínicas de Porto Alegre -HCPA) -FIPE.

Treatment protocols
The rats were randomly distributed into groups: control group (CG, n=5), AS group (n=7), and ZA group (n=7). The drug doses were adjusted to the rat's body weight using two protocols that had been already established in the literature (De Barros Silva et al., 2015;Huang et al., 2005). For example, the oral dose of AS in humans is 1mg/kg per week (Porras et al., 1999). However, the bioavailability of AS orally is approximately 1% for humans. On the other hand, AS has low bioavailability in rats, and the elimination rate of this drug is much greater in rats than in humans (Huang et al., 2005).
ZA can be administered intravenously at a dose of 4mg/patient/month in humans (Khosla et al., 2007). Here, animals received intraperitoneal injections of ZA (Ácido Zoledrônico®, Eurofarma S.A, RJ, Brazil) at 0.2mg/Kg per week intraperitoneally for three weeks, which is in accordance with previous studies (De Barros Silva et al., 2015).
After receiving the treatments, animals were euthanized in a carbon dioxide gas (CO2) chamber. The control animals (CG) did not receive any pharmacological intervention.

Histological preparation
Both femurs of each rat were collected and fixed in 10% neutral buffered formalin. After 2 days, bones were decalcified in nitric acid (10%) for 7 days. Afterward, the pieces were dehydrated, and prepared in paraffin blocks. Using microtome, sagittal sections (5μm) were cut and stained using hematoxylin-eosin (H&E) (Fischer et al., 2008).

Evaluation of epiphyseal disc
Three different areas of distal epiphysis were evaluated: the medial, intercondylar (the region between the condyles), and lateral regions. Two slides (right and left sagittal femoral cut) were prepared for each studied area. The average of the two slices was used for the analysis. Images of those regions were acquired using Qcapture® software (magnification of 400x).
The area of the disc (Figure 1a) was determined using the Adobe Photoshop CS3 extended program (Saad et al., 2008). Areas of the epiphyseal disc were classified into four zones: resting, proliferative, hypertrophic, and calcified ( Figure 1b). After identification, each zone was delimited and assessed. Finally, the relative area that each zone occupies in the disc was determined relative to the total thickness of the epiphyseal growth plate (i.e., percentage).

Statistical analysis
The normality of data distribution was assessed by the Kolmogorov-Smirnov test. Each area was expressed by the percentage of the total thickness of the grow plate, according to the following formula: Percentage of a zone = (thickness of the zone*100)/(total thickness of the epiphyseal disc) Average area (pixel 2 ) ± standard deviation of the mean (S.E.M.) expressed the data. To compare results from different groups, a one-way analysis of variance ANOVA, followed by Fisher's L.S.D. was used. SPSS 20.0 packages for windows were used to performed all analysis. P < 0.05 was considered to be statistically significant.

Results
The summary of the findings (optical image of H&E strain) is presented in Figure 2. The present results corroborate the literature showing that chondrocytes are distributed differently depending on their location. Therefore, chondrocytes are irregularly scattered in resting zone and arranged in columns in hypertrophic and proliferative zones. The deepest cartilage layer has a calcified matrix. The results showed alterations in the cartilage zones in animals treated to BIS (see details below). Lateral region: The CG had larger cell zones than both bisphosphonate groups, except for the HZ of AS group. The ZA and AS groups had larger calcified zones than CG.  Comparing the two drugs, ZA showed a more significant decrease in the resting and hypertrophic (AS vs. ZA) zones.

Lateral region
In Figures 2 and 5 The resting zone was reduced in animals that used ZA compared to AS group (AS vs. ZA, F(2-20)=29.60, p <0.05).
Source: Authors. Figure 5 shows that ZA was able to reduce the proliferative, resting, and hypertrophic zones, but increase the calcified zone compared to controls (ZA vs. CG). Likewise, AS induced a reduction in the resting and proliferative zones (CG vs. AS).
However, the calcified zone was thicker in the AS animals compared to control group (CG vs. AS). The resting zone was decreased in the ZA animals compared to those treated with AS.

Discussion
Overall, the present data showed that BIS are able to alter the morphology of the epiphyseal growth plate in rats. For instance, AS was able to reduce the proliferative zone, while ZA decreased most of the cellular zones but increased the calcified zone. The shrinking the cell zone size did not compensate the increase of the calcified zone. This unbalance between reduction/increase between cellular and calcified zones may explain the decrease of the total growth plate thickness during BIS treatment (Junges, 2013).
The long bones growth occurs by endochondral ossification, where the embryonic cartilaginous skeleton is replaced by mature bone tissue (Mackie et al., 2008;Russell et al., 2008). Some drugs may disrupt (i.e., delay, abolish, or decrease) bone formation. For example, BIS, which are antiresorptive drugs, inhibit bone resorption by their effects on clastic cells.
These bone cells are involved in the reabsorption of calcified cartilage septa and in the invasion of osteogenic cells. During the endochondral ossification process, the osteogenic cells secrete bone matrix over the calcified cartilage (Li et al., 2006).
Differences between humans and rats might be considered when age is critical. Epiphyseal disc Closure varies according to each bone and between individuals in humans. However, all bone epiphyses finish growing around 20 years of age. In rats, epiphyseal disc closure does not usually occur. Moreover, skeletal growth reduces around 7-8 months in Sprague-Dawley rats (Quinn, 2005). Here, this study used 120-day-old rats that did not reach musculoskeletal maturity. Therefore, the data were collected during growth and can be associated to human beings during the skeletal growth phase.
The epiphyseal growth plate is composed of specialized cartilage and is related to the growth of long bones (Ballock & O'Keefe, 2003). Based on the ossification stage, growth plate is divided into several zones: calcified, hypertrophic, proliferative, and resting. (Wilsman et al., 1996). Chondrocytes secrete and maintain a highly specialized matrix. This cellular environment contributes to ossification induced by a sequence of events that cause changes in chondrocyte morphology. (Mackie et al., 2011).
The present data demonstrated that the administration of ZA (short-term period) in rats was able to decrease the resting zone in all condyle regions: lateral, intercondylar, and medial (Figures 3, 4, and 5). Likewise, AS decreased the resting zone of the lateral condyle ( Figure 5). Furthermore, the reduction of the intercondylar and lateral regions was found to be more significant in animals that received ZA than those treated with AS (Figures 4 and 5, respectively). The resting zone is a contoured region (narrow and irregular) consisting of chondrocytes organized in cell lines (single or paired).
These chondrocytes are in a relatively latent state where there is a high relationship between extracellular matrix and the cell volum (Ballock & O'Keefe, 2003;Ernst B. Hunziker, 1994). The resting zone plays a key role supplying chondrocytes to the proliferative zone (Pfeil & DeCamp, 2009). In all studied condyle areas, ZA showed a more significant effect on the resting zone when compared to AS. The difference in drug effects may be due to the higher potency of ZA (Russell et al., 2008). Over time, the chondrocytes mature, enlarge to a hypertrophic state, and achieve a proliferative phenotype. Therefore, these cells acquire a flattened morphology, organize themselves longitudinally, and form the proliferative zone (E. B. Hunziker et al., 1987).
The use of ZA decreased the proliferation zone in the lateral, intercondylar, and medial regions. The AS had a similar effect in all three areas (Figures 3, 4, and 5). Previous studies had reported that chondrocytes go through a series of rapid divisions, along the long bones (proximodistal axis) in the proliferation zone (Pfeil & DeCamp, 2009). Furthermore, there is the formation of intracellular matrix and synthesis of collagen and proteoglycan (E. B. Hunziker et al., 1987). The decrease of the proliferative zone by BIS can involve the reduction of the number of cells originating from the resting zone (Pfeil & DeCamp, 2009). Moreover, the diminution of vascular endothelial growth factor (VEGF) expression would regulate the turnover of chondrocytes contributing to reducing this cell zone (Gerber et al., 1999). Furthermore, the low levels of growth factors may also decrease angiogenesis in long bones (Oyhanart et al., 2015).
After a number of divisions, the chondrocytes stop to divide and become hypertrophic (Brighton, 1984). Therefore, cell division ends, and the chondrocytes start to differentiate terminally (Pfeil & DeCamp, 2009). Here, ZA caused a reduction in the hypertrophic zone in all studied regions of the epiphyseal disc (Figures 3, 4, and 5). However, the AS had a more discreet effect. It has been seen that the decrease of the hypertrophy zone by BIS is dose-dependent (Zhu et al., 2014).
Therefore, the dosage of AS used in the present work may not have been sufficient to induce more remarkable tissue changes.
Furthermore, the hypertrophic chondrocytes are involved in cartilage calcification that is used as model for bone formation (Ballock & O'Keefe, 2003). The lower thickness of the hypertrophic zone caused by BIS treatment may be due to the less number of hypertrophic chondrocytes per column (Oyhanart et al., 2015).
In a previous study, our group has shown lower thickness of the epiphyseal disc after a BIS treatment (Junges, 2013).
Corroborating previous work, the present data indicate that BIS may cause a reduction in the hypertrophic, proliferative and resting zones (Oyhanart et al., 2015). It has been seen that AS (0.2 mg/kg/week) for growing animals caused a decrease in the thickness of the epiphyseal cartilage. This effect may be explained by two different reasons: a) a reduction of different cartilage areas or, b) changes in the morphology or number of osteoclasts (Oyhanart et al., 2015).
In this study, the distal femur growth plate showed apparently different responses to ZA and AS. Alternatively, other factors (e.g., dosages, potency, and posology) may induce distinct responses in bone physiology. For example, when 2.5 mg/kg was administered once a day, for 21 consecutive days, BIS cause expansion of cartilage zones (Rezende et al., 2017).
Moreover, a short-term ZA treatment (0.1mg/kg 3 times per week for up to 8 weeks) caused interruption of the development of the proximal tibia growth plate of rats. The authors reported non-physiological alterations in chondrocytes, cell alignment disruption, and decreased physical height (Erdogan et al., 2014).
The variability of treatment response using different BIS can be associated with the drug potency and its affinity to bone hydroxyapatite (Lawson et al., 2005). The highest mineral-binding capacity is related to more significant inhibition of bone turnover, which can influence the drug potency and clinical relevance of different BIS (Lawson et al., 2005). Another factor that contributes to the bisphosphonates potency is the inhibition of the enzyme farnesyl pyrophosphate synthase (FPPS) found in osteoclasts (Harrington et al., 2004). These drug characteristics may justify the more significant effects of ZA on cell zones, which have higher FPPS activity and affinity for hydroxyapatite (Lawson et al., 2005). The ZA potency may justify the enlargement of the calcified cartilage zone in all studied femoral areas (Figures 3, 4, and 5). Therefore, the effect of AS was more restricted than ZA in reducing the calcified zone (Lawson et al., 2005). Even though, in the lateral region of the femoral condyle, the thickness of the calcified zone was narrower in animals that received AS compared to the control group ( Figure   5).
The use of BIS has been associated with bone alterations such as sclerotic lines in the metaphysis (Silva et al., 2010) and spontaneous (Tan et al., 2019) or atypical (Leclerc et al., 2019) fractures. The present findings indicated that BIS might alter the epiphyseal growth plate. While cell areas were reduced, there was an increase in the calcified zone in animals that received BIS. Therefore, the reduction in the epiphyseal growth plate found in earlier studies is probably due to the decrease of the resting, proliferative, hypertrophic zones.

Final Considerations
The use of BIS in clinical treatment should be carefully indicated in patients prior to epiphyseal closure because there are alterations in the process of endochondral ossification and, consequently, a possible reduction in long bone growth. Future studies should address a possible relationship between the histological findings reported in the present study and the possible disruption in growth. For example, differences between sexes or ages are factors that can potentially influence the responses of the femoral growth plate to the use of bisphosphonates. Furthermore, different concentrations of these drugs may have varied results from those found here.