Investigations of the osteprogenitor cell population and its relevance to the prolongation of the reversal-resorption phase in intracortical remodeling events

Background
Elderly people in Denmark have the highest risk of osteoporotic fractures in the world. Osteoporosis can lead to permanent disability and increased mortality (1, 2). Osteoporotic fractures cost Danish health care more than 11 billion Danish kroner each year, a number that is estimated to increase (1).
The current medical therapies for osteoporosis are most effective in areas rich on trabecular bone such as the spine. Unfortunately, they have only limited effect on the risk of fractures in areas rich on cortical bone (1-3).
It is well known that the femoral neck is one of the skeletal sites most susceptible to osteoporotic fractures (1, 2). It is therefore important to develop a therapy which is also effective in the femoral neck.
In the femoral neck, the cortical bone is especially critical for its overall strength (4-6). Cortical bone is constantly undergoing renewal. The process is conducted by bone resorbing units making intracortical canals that are sub-sequentially refilled by bone forming osteoblasts. Recently, we demonstrated that the initial penetrative resorption carried out by osteoclasts is followed by osteoblastic reversal cells (osteoprogenitors) intermixed with scattered osteoclasts (7). This so-called reversal-resorption phase widens the intracortical canal until bone formation is initiated (Figure 1). Moreover, our studies together with other studies have shown that intracortical remodeling may not only generate new canals, but also remodel existing canals (8-10) (Figure 1).
During aging, cortical bone becomes increasingly thinner and porous (4-6). Recent studies from our research group showed that the reversal-resorption phase of these remodeling events was in general significantly prolonged, leading to a delayed initiation of bone formation (7, 10, 11). A prolonged reversal-resorption phase allows the canals to grow very large. A delayed initiation of bone formation can lead to inadequate refilling of the canals. Interestingly, our very recent studies have shown that accumulation and coalescence of existing canals was the main contributor to age-induced cortical porosity in iliac bone specimens from women (10) and to age-induced cortical thinning and fragility in fibular bone (11).
Aim
Central aim: Investigate the defects responsible for a prolonged reversal-resorption phase causing cortical bone loss and fragility during aging and the effect of the current anabolic drugs parathyroid hormone and romosozumab on this prolongation.
Central hypothesis: An insufficient osteoblast progenitor recruitment and differentiation prolonges the reversal-resorption phase and hereby prevents the initiation of the subsequent bone formation causing bone loss and fragility during aging. Furthermore, this arrest is only partly normalized with parathyroid hormone or romosozumab treatment.
Methods
The project consists of three independent aims. As we mainly apply for funding for aim 2, this will be described in more detail. The studies will be conducted on human cortical bone biopsies collected either nationally or internationally. The national specimen collections and use for this study is approved by the Danish National Committee on Biomedical Research Ethics (project ID# S-20120193, S-20130149, 1506824 and M-20080040) and the Danish National Board of Health (EudraCT# 2008-000606-36, protocol 84421383). ClinicalTrials.gov# NCT00730210.
Aim 1: Investigate whether the accumulation of enlarged coalescing eroded pores contributes to cortical porosity in the femur neck and whether they reflect intracortical remodeling events with a prolonged or arrested reversal-resorption phase.
Hypothesis: Accumulation of enlarged eroded pores is a major contributor to cortical porosity in the femoral neck. Furthermore, these pores reflect remodeling events with an arrested reversal-resorption phase (Figure 2C: type 2c-e events), and not only events with a prolonged reversal-resorption phase (Figure 2C: type 2b events).
Task: The type of remodeling events generating the eroded pores (identified and characterized in a parallel PhD project) will be 3D-traced. The tracing will be conducted on serial sections of femoral neck specimens from 20 young controls and 20 elderly controls (forensic autopsies), and 20 age-matched elderly with osteoporotic fractures. Twenty eroded pores will be 3D-traced in each specimen (400 events per group).
Primary outcome: The events will be 3D-categorized according to the diagram in figure 2C and compared between genders and groups. Importantly, this will allow us to address whether the cumulative eroded pores in elderly reflect remodeling events with an arrested reversal-resorption phase (Figure 2C: type 2c-e events). Moreover, these events will be further divided according to their length (above or below 300 µm; type 2d versus 2c events) and if coalescent (type 2e events).
Power assessment: The power between young versus elderly is 1.0 (α=0.01) if the prevalence of type 2c-e remodeling events increases from 20±10% to 40±10%. Part of the tracings will subsequently be converted into 3D-reconstructions using Amira.
Aim 2: Investigate if the prolonged or arrested reversal-resorption phase in elderly is the result of an insufficient recruitment of osteoprogenitor cells.
Hypothesis: The prolonged and arrested reversal-resorption phase in elderly is the result of an insufficient recruitment and differentiation of osteoprogenitors, observable as an altered density, proliferation, and expression profile of the osteoprogenitors. This is the result of cellular senescence or/and changes in the local expression of pro-osteoblastic factors. Cellular senescence is when normal cells cease to divide and change to a secretory phenotype affecting the function of neighboring cells (13).
Task 2A: Using multiplex immunostaining and in situ hybridization, we will investigate the osteoprogenitor expression of selected osteoblastic and catabolic markers as well as their proliferation index and cell density as done before (7, 14), but now within the reversal-resorption phase of the remodeling events identified in aim 1 (7).
Primary outcome: The density of osteoprogenitor cells and percentage of these cells expressing the described markers in given types of events. This will allow us to address whether remodeling events with a prolonged or arrested reversal-resorption phase have an altered density of osteoprogenitors compared to remodeling events with a shorter reversal-resorption phase, and whether their osteoprogenitor had an altered proliferation or expression profile.
Power assessment: The individual events (n=300) or osteoprogenitor cells are considered as the n but corrected for their clustering (specimen and group), using clustered logistic regression analysis to address the significance of the differences. The power is estimated to 1.0 (α=0.01) if the osteoprogenitor density (cells/mm) declines from the reported 20±10 in type 1 events (n=100) to 10±10 in type 2c-e events (n=100), and even better when addressing the likelihood that osteoprogenitor cells in given type of events are positive for one of the specified markers.
Task 2B: Using the same approach, specimens and 3D-categorized events as in task 2a, we will investigate whether events with a prolonged or arrested reversal-resorption phase have more cells positive for markers of cellular senescence (15) compared to remodeling events with a shorter reversal-resorption phase. Moreover, investigate whether the expression of pro-osteoblastic factors or their receptors are different between categories of events.
Primary outcome: The prevalence of cells expressing the described factors, receptors and markers in given types 2b and 2c-e events versus the other types of events.
Power assessment: Statistics is performed as in task 2a and the power is estimated to 1.0 (α=0.001) if the prevalence of senescent cells 4±2% (n=1000) in type 2c-e events versus 1±1% (n=1000) in other types of events.
Aim 3: Investigate whether the reversal-resorption phase is shortened by parathyroid hormone in patients treated with hypoparathyroidism and by romosozumab in osteoporosis patients compared to placebo treatment.
Hypothesis: Parathyroid hormone and romosozumab can to some extent shorten the reversal-resorption phase and transition to the formation phase, hereby partly attenuate the cortical bone loss and fragility.
Task 3A: Using our classification of intracortical pores and their remodeling characteristics; we will investigate how parathyroid hormone affects cortical bone remodeling in transiliac biopsies. The investigation will be conducted on transIliac biopsies from hypoparathyroidism patients (n=37) treated randomly with: placebo for 30 months, parathyroid hormone for 30 months, or parathyroid hormone for 6 months followed by placebo for 24 months.
Primary outcome: The prevalence of eroded and formative pores categories as previous in iliac bone specimens (10), as well as their contribution to the cortical porosity, which will be compared to patients treated with parathyroid hormone, placebo, or both. Here we will be able to investigate whether coalescing eroded pores upon existing intracortical canals are affected by the treatment.
Task 3B: Using our classification of intracortical pores and their remodeling characteristics; we will investigate whether romosozumab affects the cortical bone remodeling in transiliac biopsies. The investigation will be conducted on transiliac bone biopsies from osteoporotic patients treated randomly with placebo (n=18/33) or romosozumab (n=16/40) for 2/12 months available from the FRAME study available at INSERM, Lyon.
Primary outcome: The prevalence of eroded and formative pores categories as previous in iliac bone specimens (10), as well as their contribution to the cortical porosity, which will be compared to patients treated with romosozumab or placebo. Here we will be able to investigate whether coalescing eroded pores upon existing intracortical canals are affected by the treatment.
Power assessment: Here the power between groups is 0.99 (α=0.05) if the prevalence of eroded pores is reduced from the observed 20±5% (n=14) to 10±5% (n=39).
Statistics
Spearman’s rank correlation will be used to compare the pore categories with other parameters. Significant differences of parameters from different pore categories will be identified using a Friedman test followed by Dunn’s posttest. The percentage of osteoprogenitors positive for different markers will be examined with Student’s t-test. Student’s t-test will also be used to compare the type, stage, and positions of pores between drug and placebo groups. D’Agostino and Pearsson Omnibus normality test will be used for testing whether the data are normally distributed.

N/A

Danske ældre har verdens højeste risiko for frakturer forårsaget af knogleskørhed. Knoglebrud kan give varige mén og øge dødeligheden. Desuden koster behandling af knoglebrud årligt sundhedsvæsenet over 11 milliarder kroner.
Der findes to typer af knoglevæv: den svampelignende inderste knogle (trabekulær) og den hårde yderste knogle (kortikal). De lægemidler, der bruges mod knogleskørhed i dag, virker bedst i områder med meget trabekulær knogle, f.eks. rygsøjlen. Desværre er de mindre effektive i områder med meget kortikal knogle, f.eks. arme og ben.
Lårbenshalsen er en af de knogler, der påvirkes mest af knogleskørhed, og knoglebrud sker ofte her. Studier af lårbenshalsen viser, at den kortikale del af knoglen spiller en stor rolle for knoglens samlede styrke. Kortikal knogle er opbygget af såkaldte Haverske kanaler, hvori gammel eller beskadiget knogle nedbrydes af knoglenedbrydningsceller, hvorved kanalen udvides. Kanalen fyldes derefter ud med ny frisk knogle vha. knogledannende celler. Disse to faser er knyttet tæt sammen af en såkaldt reversal-resorption fase. Vores nyeste studier af hoftebiopsier viser, at reversal-resorption fasen med alderen forlænges hos kvinder. Det medfører, at nogle af kanalerne bliver meget store, smelter sammen med andre og nogle bliver aldrig fyldt tilstrækkeligt op igen. Det medfører, at hoftens kortikale knoglevæv bliver mere udtyndet.
For at kunne optimere fremtidige behandlingsmuligheder er det nødvendigt at vide præcist, hvorfor den er forlænget. Vores hypotese er at reversal-resorption fasen er forlænget på grund af mangel på knogledannende celler. Hvis de knogledannende celler ikke rekrutteres til områderne, hvor knogleomdannelsen foregår eller hvis de ikke udvikles i den rigtige retning, kan det medføre en forsinket knogledannelse. Formålet med dette projekt er at finde ud af de mere præcise cellulære årsager til denne forsinkede knogledannelse.
Projektet består af tre studier.
Studie 1 undersøger, om store sammensmeltede kanaler er årsagen til øget porøsitet og skrøbelighed i lårbenshalsen ligesom i hoften. Knoglebiopsier fra ældre med knoglebrud grundet knogleskørhed sammenlignes med raske i forskellige aldre vha. 3D-tracing. 3D-tracing er en teknik, som giver et strukturelt billede af kanalerne og ved at kombinere 3D-tracing med histologiske snit opnås helt ny og meget detaljeret information om kanalernes forandring.
Studie 2 tester hypotesen om at knogledannende celler ikke rekrutteres eller udvikles tilstrækkeligt. Ved hjælp af immunfarvning og in situ hybridisering er det muligt at se, hvor mange knoglenedbrydende og knogledannende celler der er og om de udvikles optimalt. Målingerne udføres på det væv, som undersøges i studie 1. Herved kan vi afgøre, om porer med en forlænget reversal-resorption fase har få knogledannende celler eller om de ikke udvikles optimalt.
Studie 3 skal undersøge effekten af parathyroidhormon og romosozumab (anti-sclerostin antistof). Biopsier fra patienter behandlet med enten parathyroidhormon/romosozumab eller placebo sammenlignes for at se, om lægemidlerne genskaber balancen mellem knoglenedbrydning og -dannelse. Biopsierne fra romosozumab-behandlede patienter har vi fået adgang til gennem vores samarbejde med INSERM, Lyon. Vores analyser vil give svar på om de anvendte præparater har haft en positiv indvirkning på knoglefornyelsen.