Contents lists available at ScienceDirect

Bone

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

Full Length Article

An osteopenic/osteoporotic phenotype delays alveolar bone repair

Chih-Hao Chena,b, Liao Wangb,c, U. Serdar Tulub, Masaki Ariokab,d, Melika Maghazeh Moghimb,e,Benjamin Salmonb,f, Chien-Tzung Chena,g, Waldemar Hoffmannh, Jessica Gilgenbachh,John B. Brunskib, Jill A. Helmsb,⁎

a Craniofacial Research Center, Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital, Chang Gung University School of Medicine, Taoyuan33305, TaiwanbDivision of Plastic and Reconstructive Surgery, Department of Surgery, Stanford University School of Medicine, Stanford, CA 94305, USAc State Key Laboratory of Oral Diseases, West China Hospital of Stomatology, Sichuan University, Chengdu 610041, Chinad Department of Clinical Pharmacology, Faculty of Medical Sciences, Kyushu University, Fukuoka 812-8582, JapaneUniversity College London Medical School, University College London, London WC1E 6BT, UKf Paris Descartes University - Sorbonne Paris Cité, EA 2496 - Orofacial Pathologies, Imaging and Biotherapies Lab and Dental Medicine Department, Bretonneau Hospital,HUPNVS, AP-HP, Paris, Franceg Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital at Keelung, Keelung 20401, TaiwanhNobel Biocare Services AG P.O. Box, CH-8058 Zürich-Flughafen, Switzerland

A R T I C L E I N F O

Keywords:RatsOvariectomyTooth extractionOsteotomyAlveolar boneComputer models

A B S T R A C T

Aging is associated with a function decline in tissue homeostasis and tissue repair. Aging is also associated withan increased incidence in osteopenia and osteoporosis, but whether these low bone mass diseases are a risk factorfor delayed bone healing still remains controversial. Addressing this question is of direct clinical relevance fordental patients, since most implants are performed in older patients who are at risk of developing low bone massconditions. The objective of this study was to assess how an osteopenic/osteoporotic phenotype affected the rateof new alveolar bone formation. Using an ovariectomized (OVX) rat model, the rates of tooth extraction socketand osteotomy healing were compared with age-matched controls. Imaging, along with molecular, cellular, andhistologic analyses, demonstrated that OVX produced an overt osteoporotic phenotype in long bones, but only asubtle phenotype in alveolar bone. Nonetheless, the OVX group demonstrated significantly slower alveolar bonehealing in both the extraction socket, and in the osteotomy produced in a healed extraction site. Most notably,osteotomy site preparation created a dramatically wider zone of dying and dead osteocytes in the OVX group,which was coupled with more extensive bone remodeling and a delay in the differentiation of osteoblasts.Collectively, these analyses demonstrate that the emergence of an osteoporotic phenotype delays new alveolarbone formation.

1. Introduction

Low bone mass diseases including osteopenia/osteoporosis decreasebone quantity, which in turn impacts bone quality [1,2]. An osteo-porotic skeleton is characterized by a decrease in bone volume andbone mineral density, making the skeleton more fragile [3] and there-fore prone to fracture [4]. Osteoporosis is also associated with an in-crease in marrow adipogenesis, commensurate with a loss in osteo-blastogenesis [5,6].

Whether osteopenia/osteoporosis is a risk factor for delayed bonehealing, however, is unclear. For example, rats subjected to an ovar-iectomy (OVX), which reliably produces an osteoporotic phenotype [7],exhibit slower healing of tooth extraction sockets compared to age-

matched non-OVX rats [8,9]. Rodent models of long bone fractures alsoexhibit slower healing (reviewed in [10]).

Clinical data are less clear. In some meta-analyses, bone healingaround dental implants appeared to be unaffected in osteoporotic pa-tients [11,12] But in other analyses, peri-implant bone loss was moresevere in osteoporotic patients ([13]; reviewed in [14]). Most studiesfail to report whether patients are actively taking anti-osteoporoticmedications e.g., bisphosphonates, which may impact bone remodelingrates [15,16]. An additional confounding variable is that most patientsare older, and it is inherently challenging to separate the effects ofosteoporosis from other age-related diseases that can impact the ske-leton. Understanding whether low bone mass diseases impact alveolarbone osteogenesis has direct clinical relevance for dental patients, since

https://doi.org/10.1016/j.bone.2018.04.019Received 28 February 2018; Received in revised form 10 April 2018; Accepted 21 April 2018

⁎ Corresponding author at: Stanford University, 1651 Page Mill Drive, Palo Alto, CA 94304, USA.E-mail address: jhelms@stanford.edu (J.A. Helms).

Bone 112 (2018) 212–219

Available online 25 April 20188756-3282/ © 2018 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/).

T

most implants are performed in individuals over the age of 50 [17,18]and this group is at risk for developing osteopenia and osteoporosis[19].

These data served as a starting point for our study: using the sameOVX rat model we undertook a series of experiments to uncover themechanisms whereby an osteoporotic phenotype impacts bone healing.We used both extraction socket repair and osteotomy repair as injurymodels, and employed a battery of molecular, cellular, and imagingtools to reveal how osteoporosis negatively impacts alveolar bonehealing.

2. Materials & methods

2.1. Animals, experimental plan, and surgeries

Stanford APLAC approved all procedures, which conform to ARRIVEguidelines. In total, forty-eight 6-week-old (young) and five 12-month-old (aged) female rats (Wistar, Charles River) were used in this study.Three surgeries were carried out: an ovariectomy (OVX); a tooth ex-traction; and an osteotomy. Each surgery is described below. In allcases, anesthesia was reached via intraperitoneal injection of Ketamine(100mg/kg) and Xylazine (10mg/kg); and analgesia via subcutaneousinjection of Buprenorphine-SR (0.5 mg/kg). Rats were randomly as-signed to experimental groups that are outlined in Table 1.

OVX surgery followed previous guidelines where a dorsal midlineincision was made between the mid-back and tail base. The peritonealcavity was accessed through bilateral muscle layer incisions, the ovarywas identified, the connection between the fallopian tube and theuterine horn was suture-ligated, and the ovary was removed. Woundswere closed layer by layer.

Bilateral maxillary first molars (e.g., M1) extractions were per-formed using micro-forceps. After extraction, bleeding was controlledby local compression. Rats recovered in a controlled, heated environ-ment, then were fed a soft food diet for the duration of the experimentand housed in groups of two. Weight changes were< 10%. No adverseevents (e.g., uncontrolled pain, infection, prolonged inflammation)were encountered. In Group 1, M1 extraction was performed concurrentwith the OVX procedure. In Group 2, M1 extraction was delayed8weeks following the OVX. Animals were sacrificed at post-extractiondays (PEDs) 0, 1, 7 and 21 for analysis.

Osteotomy site preparation followed previously established proce-dures [20]. Briefly, osteotomies were produced in the healed M1 ex-traction site using a low-speed handpiece (KaVo Dental UK; 5333 rpm)using 0.3 mm drill bits (Drill Bit City, NY) with saline irrigation. Noadverse events were encountered during the healing process. Tissueswere collected at post-osteotomy days (PODs) 0.5 and 7.

2.2. Tissue preparation and histology

After harvesting, samples were prepared for histology as described[21]. Briefly, samples were fixed in 4% paraformaldehyde (PFA), dec-alcified in ethylene diamine tetra-acetic acid (EDTA), dehydrated usingan ascending graded ethanol series, and embedded into paraffin blocksfor sectioning. Before staining or other histological/cellular activityanalysis, all sections were de-paraffinized in Citrisolv (#1601, DeconLabs Inc. PA), and hydrated via a descending graded ethanol series.After staining, sections were dehydrated in a graded series of ethanoland Citrisolv, and subsequently cover-slipped with Permount (#SP15,Fisher Scientific) mounting media.

For Aniline blue staining, slides were treated with a saturated so-lution of picric acid, followed by a 5% Phosphotungstic acid solutionand staining in 1% Aniline blue. Pentachrome staining was performedas described [22]. In brief, after dehydration, slides were stained with1% Alcian Blue (#A5268, Sigma), Verhoeffs Hematoxylin (#S71299,Fisher Scientific), Sodium Thiosulfate (#14518, Alfa Aesar, MA), Cro-cein-Scarlet-Acid Fuchsin solution (#22914, Chem Impex International,IL; #F8129, Sigma), 5% Phosphotungstic Acid (#P4006, Sigma) andSaffron (#3801, Harlecon), with washing steps between each stainusing ethanol, acetic acid and distilled water. For Picrosirius Redstaining, slides were stained with picrosirius solution (0.5 g Sirius red(#35780, Pfaltz & Bauer, Inc., CT) dissolved in 500mL saturated picricacid solution), and then viewed under polarized light.

2.3. Alkaline phosphatase (ALP) and Tartrate-resistant acid phosphatase(TRAP) activities

To detect ALP activity, tissue sections were treated with ALP-de-tection solution containing BCIP (5-bromo-4-chloro-3-indolyl phos-phate; Roche, #11383221001) and NBT (nitro blue tetrazoliumchloride; Roche, #11383213001) at 37 °C for 30min according to themanufacturer's instructions. TRAP activity was observed using aLeukocyte Acid Phosphatase Staining Kit (catalog #386A-1KT, Sigma-Aldrich, St. Louis, MO). Tissue sections were processed according to themanufacturer's instructions.

2.4. Detection of viable/dead cell nuclei

The number and distribution of viable cells were determined using anuclear stain, DAPI. Slides were mounted with ProLong Gold AntifadeMountant containing DAPI (#P36935, Life Sciences) then viewed underfluorescent light. TUNEL (#11684795910, Roche, Indianapolis, IN) wasperformed as described by the manufacturer.

To quantify the extent of apoptotic osteocytes due to drilling,TUNEL stained tissue sections of drill sites were analyzed from 4 to 6different samples taken from trabecular bone. Each section was pho-tographed using a Leica digital image system. The number of TUNEL-labelled osteocytes, corresponding to apoptotic cells, was determinedand grouped according to their distance from the cut edge. The corre-sponding area for each group was then calculated. The number ofapoptotic cells per unit area was found by dividing number of apoptoticcells to the corresponding area (cell/mm2).

2.5. Immunostaining

Immunostaining was performed using standard procedures [23]. Inbrief, endogenous peroxidase activity was quenched by 3% hydrogen

Table 1Experimental groups.

Group (figure #) Procedure Waitingperiodafter OVX

Characteristics Sample size

Young (Fig. 1) None N/Aa Young ratsNormal bonedensity

N=10

Young+OVX(Fig. 1)

OVX 8weeks Young ratsOsteoporoticbone

N=10

Aged (Fig. 1) None N/A Aged ratsOsteoporoticbone

N=10

Group 1 (Fig. 2) OVX, M1extraction

No waiting Young ratsNormal bonedensity

N=30

Group 2 (Fig. 2) OVX, M1extraction

8weeks Young ratsOsteoporoticbone

N=30

Young (Figs. 3, 4) M1 extraction,osteotomy

N/A Young ratsNormal bonedensity

N=12

Young+OVX(Figs. 3, 4)

OVX, M1extraction,osteotomy

8weeks Young ratsOsteoporoticbone

N=12

a N/A=not applicable.

C.-H. Chen et al. Bone 112 (2018) 212–219

213

peroxide for 5min, and then washed in PBS. Slides were blocked with5% goat serum (S-1000, Vector Labs, CA) for 1 h at room temperature.The appropriate primary antibody was added and incubated overnightat 4 °C, then washed in PBS. The primary antibodies used in this paperare: Osterix (1:1200; ab22552, Abcam), CathepsinK (1:200; ab19027,Abcam), Periostin (1:200; ab14041, Abcam). Samples were incubatedwith appropriate biotinylated secondary antibodies (Vector BA-x) for30min, then washed in PBS. An avidin/biotinylated enzyme complex(Kit ABC Peroxidase Standard Vectastain PK-4000) was added and in-cubated for 30min and a DAB substrate Kit (Kit Vector Peroxidasesubstrate DAB SK-4100) was used to develop the color reaction.

2.6. Micro-computed tomography (μCT)

Scanning and analyses followed published guidelines [24]. Three-dimensional μCT imaging was performed at various times after surgery.In brief, samples were fixed in 4% PFA at 4 °C overnight and washed inPBS. During scanning, the samples were kept in 70% ethanol solution.μCT scanning was performed on Viva CT data-acquisition system(Scanco, Brüttisellen, Switzerland) at 10.5 μm voxel size (70 kV, 115 μAand 300ms integration time). Bone morphometry was performed usingthe acquisition system's analysis software (Scanco). Multiplanar re-construction and volume rendering were carried out using Avizo (FEI,Hillsboro, OR) and ImageJ (NIH, Bethesda, MD) software. Images wereorganized using Adobe Photoshop and Illustrator (Adobe, San Jose,CA). Results were presented in the form of mean ± standard deviation,with N equal to the number of samples analyzed.

2.7. Histomorphometric analyses

Histomorphometric measurements were performed using ImageJsoftware (NIH, Bethesda, MD). To quantify the amount of new boneformation in tooth extraction sockets, Aniline blue-stained histologicsections that spanned the distance from the furcation to the apex wereused. Each section was photographed using a Leica digital image systemat 20× magnification. To find the percentage of new bone, the numberof Aniline Blue+ve pixels within an osteotomy were calculated and di-vided by the number of the total pixels in the same osteotomy area.

2.8. Statistical analyses

Results are presented in the form of mean ± standard deviation. Allstatistical analyses were performed using the GraphPad website.Histomorphometric results were based on t-tests and unpaired t-tests.Significance was attained at p < 0.05 (*), at p < 0.01 (**), atp < 0.001 (***).

3. Results

3.1. Ovariectomy produces an osteoporotic phenotype in both appendicularand alveolar bones

Our first objective was to determine if an OVX surgery created anosteopenic/osteoporotic phenotype that was detectable in the jaw-bones. A cohort of young rats served as a positive control for optimalbone health (Fig. 1A). The same aged young rats that had undergoneOVX 8weeks prior constituted the test group (Fig. 1B), and a cohort ofaged (~12-month-old) rats served as a negative control for age-relatedchanges in bone health (Fig. 1C).

Rats were evaluated when they were 15 weeks old. Volume ren-dering of the distal femur revealed a dense pattern of trabecular bone inthe young group (Fig. 1D). In the young+OVX group, trabecular bonevolume was significantly reduced (Fig. 1E), comparable to the agedgroup (Fig. 1F; quantified in G). Trabecular number and trabecularseparation in the young+OVX group were also equivalent to that seenin the aged group (Fig. 1G).

Clinical data indicate that with age the buccal plate of maxillarybone thins, and alveolar bone height recedes [25,26] therefore weevaluated these characteristics in the experimental groups. The youngrat cohort exhibited the thickest buccal plate (Fig. 1H) and exhibitedthe most coronal alveolar crest height (Fig. 1H′; quantified in K). In theyoung+OVX group the thickness of the buccal bone did not differsignificantly from the young group (Fig. 1I) but the height of the al-veolar crest was significantly receded (Fig. 1I′; quantified in K). In theaged group, the buccal bone was significantly thinner (Fig. 1J), and thealveolar bone showed the maximum resorption (Fig. 1J′; quantified inK). These additional analyses indicated that OVX produced an osteo-penic/osteoporotic phenotype affecting both appendicular and cranio-facial skeletons, although in the latter the bony changes were moresubtle. While the appendicular skeletons of aged and young+OVX ratsshowed clear evidence of osteopenia/osteoporosis, transverse μCTsections through the maxillary alveolar bone did not (Fig. 1L–N).Quantification of a region of interest in the alveolar bones of the threegroups (dotted lines in Fig. 1L–N) showed no significant differences intrabecular thickness and separation (Fig. 1O). Only trabecular numberwas significantly diminished in the young+OVX group compared tothe young group (Fig. 1O).

3.2. Extraction socket healing is negatively impacted by emergence of anosteoporotic phenotype

To evaluate whether an osteopenic/osteoporotic phenotype affectedalveolar bone healing, we created two experimental groups (Fig. 2A). InGroup 1, OVX and M1 extractions were performed concurrently, andthe rate of extraction socket healing was monitored. In Group 2, theOVX surgery was performed and then, after an eight-week periodduring which the osteoporotic phenotype developed, M1 extractionswere performed (Fig. 2A). We first examined Group 2 rats to determinewhether intact alveolar bone showed any overt changes in bone vo-lume. Using a histomorphometric analysis of inter-radicular bone, nosignificant differences in the area of mineralized matrix/region of in-terest were noted (Fig. 2B, C; quantified in D).

H&E staining on tissues collected within 12 h of tooth extractionrevealed a blood-filled socket in Group 1 (Fig. 2E) and a similar blood-filled socket in Group 2 (Fig. 2F). Remnants of the periodontal ligament(PDL) were visible by Periostin immunostaining in Group 1 (Fig. 2G),but the PDL remnants appeared thinner in Group 2 (Fig. 2H). Within1 day (e.g., post-extraction day 1, PED1), the sockets in Group 1 werefilled with granulation tissue (Fig. 2I), while the sockets in Group 2were still filled with blood (Fig. 2J). On PED7, ALP staining indicatedthat new bone had filled the extraction sockets in samples from Group 1(Fig. 2K; quantified in M), but extraction sockets had significantly lessALP activity in samples from Group 2 (Fig. 2L; quantified in M). Newbone filled the extraction sockets in samples from Group 1 (Fig. 2N;quantified in P), but not so in samples from Group 2 (Fig. 2O; quantifiedin P). By PED21, the area of Aniline Blue+ve bone tissue in Group 1healed sites was equivalent to surrounding, intact bone (Fig. 2Q;quantified in S). At this time point, there was significantly less bonetissue in Group 2 extraction sites (Fig. 2R; quantified in S). Because theamount of new bone in samples from Group 2 was still significantly lessthan in pristine, uninjured sites (Fig. 2S), we concluded that the ex-traction sockets of osteoporotic rats were still in the process of healing.

3.3. The same drilling procedure produces a wider zone of cell death if theskeleton is osteoporotic

Cutting bone kills osteocytes [27]; we next evaluated whether anosteoporotic phenotype impacted the extent of this osteocyte cell death.Osteotomies were produced in healed extractions sites in osteoporoticrats, and in control rats. The drill diameter, drill speed, irrigation, andthe site of osteotomy were identical between the two groups (seeTable 1 for a complete description of the experimental cohorts).

C.-H. Chen et al. Bone 112 (2018) 212–219

214

Immediately after osteotomy, aniline blue staining revealedequivalently-sized, smooth-edged osteotomies in both young andyoung+OVX cohorts (Fig. 3A, B). Collagen organization was visua-lized by picrosirius red staining and when viewed under polarized light,a characteristic basket weave pattern was observed on the cut boneedges in both young and young+OVX cohorts (Fig. 3C, D). Penta-chrome staining verified comparable cut surfaces in both groups(Fig. 3E, F).

The distribution of viable, dead, and dying osteocytes was mappedas a function of distance from the cut edge of the osteotomy.Differential interference contrast (DIC) microscopy was used to

visualize osteocyte lacunae; on the same tissue sections, TUNEL stainingidentified dying (apoptotic) osteocytes, and DAPI staining identifiedviable osteocytes. Osteotomies created in young rats produced a~100 μm wide zone of TUNEL+ve osteocytes interspersed with someviable osteocytes (Fig. 3G). The same osteotomy produced a sig-nificantly wider (~250 μm) zone of TUNEL+ve, DAPI−ve osteocytes inthe young+OVX cohort (Fig. 3H). In a 50 μm-wide zone nearest to theosteotomy edge, the young+OVX cohort had a greater number ofdying cells compared to young rats (Fig. 3I, J). From 150 μm outward,there were hardly any apoptotic osteocytes in the young cohort,whereas the young+OVX cohort still had apoptotic osteocytes

Fig. 1. OVX produces both appendicular and alveolar bone loss resembling the bone loss seen in aged animals. Representative transverse μCT sections of distal femurin animals of (A) young, (B) young+OVX, and (C) aged groups. Three-dimensional visualization of trabecular bone in (D) young, (E) young+OVX and (F) agedrats. (G) Quantification of trabecular thickness (Tb.Th), trabecular number (Tb.N) and trabecular separation (Tb.Sp) in the three groups. Representative coronal μCTsections, with yellow lines demonstrating the width of the buccal bone from young (H), young+OVX (I) and aged (J) rats. Higher magnification, with dotted linesdemonstrating the height of the alveolar bone crest (red) relative to the dento-enamel junction (DEJ, yellow) in (H′) young, (I′) young+OVX and (J′) aged animals.(K) Quantification of width of the buccal bone and height of the alveolar bone crest relative to DEJ across the different groups. Representative transverse μCT sectionsof alveolar bone surrounding mxM2, from (L) young, (M) young+OVX, and (N) aged rats; equivalent regions of interest are indicated by dotted lines. (O)Quantification of Tb.Th, Tb.N and Tb.Sp in alveolar bone around second maxillary molar. Abbreviations: Tb.Th, trabecular thickness; Tb.N, trabecular number;Tb.Sp, trabecular separation; DEJ, dento-enamel junction. Scale bars= 1mm. (For interpretation of the references to color in this figure legend, the reader is referredto the web version of this article.)

C.-H. Chen et al. Bone 112 (2018) 212–219

215

~300 μm from the cut edge (Fig. 3I). From these analyses, it was clearthat the zone of apoptotic osteocytes produced by osteotomy was sig-nificantly larger in the osteopenic/osteoporotic group.

3.4. Finite element model predicts larger apoptotic area in osteoporoticbones

In an attempt to understand why site preparation in an osteoporoticanimal resulted in significantly greater cell death, we evaluated themechanical impact of drilling on young and osteoporotic bones. Themajor variance between the finite element (FE) model of young andyoung+OVX bone was the BV/TV, which was taken the literature

[28]. In modeling the young cohort, a BV/TV of 50% was used whereasmodeling the young+OVX cohort used a BV/TV of 6% (Fig. 4A). Tosimulate the effect of drilling [29] the bony trabeculae lining the cutsurface of the osteotomy were subjected to an outwardly-directed radialforce of 0.8 N (indicated in blue, Fig. 4B). Results from modeling theyoung cohort indicated minimal compressive strains on each trabeculae(i.e., 0.54%, Fig. 4C). Because the trabeculae were thinner and fewer,compressive strains on each trabeculae in the young+OVX cohortwere much greater (i.e., 8%, Fig. 4C) and several times larger than theyield strain of bone. These FE results are consistent with a modelwhereby an osteoporotic skeleton sustains greater damage than does ahealthy skeleton in response to the same mechanical force [30,31].

Fig. 2. The rate of extraction socket healing depends on the emergence of the OVX phenotype. (A) Schematic of the experimental model where 6-week-old ratsunderwent OVX and bilateral first molar (M1) extractions. In Group 1, tooth extraction occurred immediately after OVX, followed by sacrifice at post-extraction days(PEDs) 0, 1, 7 and 21. In Group 2 OVX was followed by an 8-week recovery period during which the OVX phenotype developed; afterwards tooth extraction wasperformed, followed by sacrifice at the same time points. Representative transverse tissue section stained by Aniline blue from intact alveolar bone in Group 2 at (B)post-OVX week 8, and (C) post-OVX week 11. (D) Quantification of Aniline Blue+ve pixels/total pixels in the intact bone from Group 2. Representative sagittal H&E-stained tissue sections from rats in (E) Group 1 and (F) Group 2 showing the extraction socket on PED0. Adjacent tissue sections immunostained with Periostin toidentify the residual PDL in the extraction sockets of rats in (G) Group 1 and (H) Group 2. H&E-stained sagittal tissue sections on PED1 indicate granulation tissuepresence in the rats of (I) Group 1 and (J) Group 2. Representative transverse histologic sections of M1 healing extraction sockets on PED7, evaluated for alkalinephosphatase (ALP) activity in (K) Group 1 and (L) Group 2 rats; dotted lines demarcate extraction site boundaries. (M) Quantification of ALP+ve pixels within theregion of interest for this time point. Transverse tissue sections through the extraction socket stained using Aniline blue on PED7 from rats in (N) Group 1 and (O)Group 2. (P) Quantification of Aniline Blue+ve pixels/total pixels in the healing extraction site on PED7. Aniline blue-stained tissue sections on PED21 from rats in (Q)Group 1 and (R) Group 2. (S) Quantification of Aniline Blue+ve pixels/total pixels in the intact alveolar bone and the healing extraction site on PED21 in both groups.Abbreviations: ab, alveolar bone; es, extraction socket; p, periodontal ligament. Scale bars= 100 μm. (For interpretation of the references to color in this figurelegend, the reader is referred to the web version of this article.)

C.-H. Chen et al. Bone 112 (2018) 212–219

216

3.5. Ovariectomy negatively affects bone formation in an osteotomy

The more extensive osteocyte death observed in the osteopenic/osteoporotic cohort correlated with slower, new bone formation. Forexample, on POD7 there appeared to be fewer osteogenic precursors(Fig. 4D, E), and less osteoclast activity (Fig. 4F, G) in the young+OVXcohort compared to young controls. Qualitatively, the extent of ALP andTRAP staining was also less in the young+OVX cohort (Fig. 4H, I),which suggested that bone remodeling was less robust. Quantificationof the amount of reparative bone in the osteotomy revealed sig-nificantly less new mineralized matrix in the osteoporotic group(Fig. 4J–L). Collectively, these data demonstrated that an osteopenic/osteoporotic phenotype showed greater damage than a young skeletonin response to the same mechanical force, which resulted in a delay inalveolar bone healing, in both extraction sites and in osteotomies pro-duced in healed extraction sites.

4. Discussion

4.1. Aging, bone health, and healing potential

As we age, molecular and cellular damage accrues and coupled withthe loss of tissue-specific progenitor cells, a functional decline in tissuehomeostasis ensues [32,33]. For example, bone repair is slower in anaged population [34] but osteopenia/osteoporosis also increases withage, which makes it difficult to discriminate between the effects ofaging on bone healing [4] versus the effects of low bone mass disease onbone healing. In an experimental setting, one can remove the aging

variable from investigation. Here, we achieved this by producing anosteoporotic phenotype in young rats, then compared their rate of bonerepair against a group of age-matched young rats. This allowed us toeffectively test whether low bone mass conditions directly impacted therate of alveolar bone healing without the confounding variable ofchronologic age.

4.2. OVX-induced osteoporosis affects both long bones and alveolar bones

OVX results in a significant loss in trabecular number and a sig-nificant increase in trabecular separation, most notably in long bones[7]. These changes are similar to age-related alterations in the skeleton[4]. OVX only slightly altered trabecular number in alveolar bone of thesame animal (Fig. 1). The evaluation of BV/TV took place ~8weeksafter surgery, as per established protocols (see [35–38]). This time-frame might not be long enough for an osteoporotic phenotype to fullydevelop in alveolar bones; alternatively, it could indicate that alveolarbone is somehow “resistant” to osteoporotic changes [39–41].

We evaluated alveolar bone using additional parameters and founda significant reduction in alveolar bone height (Fig. 1), similar to whathas been reported in aged patients [42]. We also found that OVX ratshad significantly slower extraction socket healing (Fig. 2), which wouldargue against a resistance hypothesis. The surgery itself was eliminatedas a contributing factor to slower healing because both Groups 1 and 2underwent OVX. The only remaining variable was that in Group 2, theosteoporotic phenotype was allowed to develop for 8 weeks beforetooth extraction was performed (Fig. 2). Consequently, we can concludethat lack of estrogen due to OVX is associated with slower bone repair.

Fig. 3. Despite an equivalent surgical procedure, significantly more osteocyte apoptosis occurs in OVX rats versus age-matched controls. Experimental model inwhich M1 extraction was followed by osteotomy preparation and sacrifice at post-osteotomy day 0.5 (POD0.5). Young+OVX group underwent an 8-week recoveryperiod to allow for phenotype expression prior to osteotomy preparation and sacrifice. Aniline blue staining of transverse tissue sections from representativeosteotomies prepared in (A) young and (B) young+OVX rats. Representative transverse tissue sections stained using Picrosirius Red and visualized under polarizedlight from (C) young and (D) young+OVX rats. High magnification of pentachrome-stained transverse tissue sections of (E) young and (F) young+OVX rats.Representative tissue sections stained using TUNEL and DAPI in (G) young and (H) young+OVX rats; dotted lines show the edge of the osteotomy. Arrows indicatethe approximate extent of the radial zones of cell death. Quantification of (I) TUNEL+ve apoptotic osteocyte and (J) DAPI+ve viable osteocyte distribution as afunction of distance from cut edge of osteotomy. Abbreviation: os, osteotomy. Scale bars= 100 μm. (For interpretation of the references to color in this figure legend,the reader is referred to the web version of this article.)

C.-H. Chen et al. Bone 112 (2018) 212–219

217

4.3. Potential implications of these data for patients with sub-optimal bonehealth

Our experimental data and computational modeling predict that, allother variables being equal, creating an osteotomy in osteoporotic pa-tients will result in more osteocyte death than in patients with normalbone density. The greater the amount of dead or dying bone, the greateris the extent of bone resorption [21,27]. Therefore, our results heresuggest that patients with uncontrolled osteoporosis may show agreater degree of bone remodeling compared to an equivalent patientpopulation with normal bone density. If true, these data have directclinical relevance: most dental implant patients are over age 50[18,43], and this age group is typically osteopenic or osteoporotic [19].While these inferences for patients remain speculative, it is nonethelessclear that methods to reduce osteocyte apoptosis caused by osteotomysite preparation will be of considerable value to any patient, especiallythose with sub-optimal bone health.

Acknowledgements

This work is supported by a grant from the NIH (R01 DE024000-12to J.A.H.) and funds from Nobel Biocare (grant #2015-1400). J.B.B.and J.A.H. are paid consultants for Nobel. We thank Maziar Aghvamifor his helpful discussions and Pedro Cuevas for his assistance in his-tological experiments. All other authors declare they have no conflictsof interest.

References

[1] E. Seeman, P.D. Delmas, Bone quality—the material and structural basis of bonestrength and fragility, N. Engl. J. Med. 354 (2006) 2250–2261.

[2] L.G. Raisz, Pathogenesis of osteoporosis: concepts, conflicts, and prospects, J. Clin.Invest. 115 (2005) 3318–3325.

[3] L.G. Koniaris, K.D. Lillemoe, C.J. Yeo, R.A. Abrams, J. Colemann, A. Nakeeb, H. Pitt,J.L. Cameron, Is there a role for surgical resection in the treatment of early-stagepancreatic lymphoma? J. Am. Coll. Surg. 190 (2000) 319–330.

[4] A.L. Boskey, R. Coleman, Aging and bone, J. Dent. Res. 89 (2010) 1333–1348.[5] J. Li, X. Liu, B. Zuo, L. Zhang, The role of bone marrow microenvironment in

governing the balance between osteoblastogenesis and adipogenesis, Aging Dis. 7(2016) 514–525.

Fig. 4. Effects of compression strain by osteotomy and alveolar bone healing of osteotomy site on an osteoporotic skeleton. (A) Representative computer-generatedskeletons to model osteotomy into trabeculae of young and osteoporotic bones for finite element (FE) analysis. (B) FE model shows a radial outward force as a resultof drilling into trabeculae of young and osteoporotic bones. This force is distributed along the surfaces of bony trabeculae (shown in blue) near osteotomy. (C) FEanalysis shows compressive strain in trabeculae of young and osteoporotic bones during drilling. Representative transverse tissue sections from POD7, im-munostained with Osterix, as a marker of osteoblast differentiation in the osteotomy sites of (D) young and (E) young+OVX rats; dotted lines show the edge of theosteotomy. Adjacent tissue sections immunostained with CathepsinK, as an indicator of osteoclast activity in the osteotomy sites (F) young and (G) young+OVX rats;dotted lines show the edge of the osteotomy. Co-staining of representative transverse tissue sections of healing osteotomy site evaluated for TRAP and ALP activity in(H) young and (I) young+OVX rats. Aniline blue-stained tissue sections evaluated for (J) young and (K) young+OVX rats. (L) Quantification of Aniline Blue+ve

pixels/total pixels in the osteotomy site. Abbreviation: os, osteotomy. Scale bars= 100 μm. (For interpretation of the references to color in this figure legend, thereader is referred to the web version of this article.)

C.-H. Chen et al. Bone 112 (2018) 212–219

218

[6] E.J. Limonard, A.G. Veldhuis-Vlug, L. van Dussen, J.H. Runge, M.W. Tanck,E. Endert, A.C. Heijboer, E. Fliers, C.E. Hollak, E.M. Akkerman, P.H. Bisschop,Short-term effect of estrogen on human bone marrow fat, J. Bone Miner. Res. 30(2015) 2058–2066.

[7] D.N. Kalu, The ovariectomized rat model of postmenopausal bone loss, Bone Miner.15 (1991) 175–191.

[8] M.C. Pereira, K.G. Zecchin, E.B. Campagnoli, J. Jorge, Ovariectomy delays alveolarwound healing after molar extractions in rats, J. Oral Maxillofac. Surg. 65 (2007)2248–2253.

[9] G. Ramalho-Ferreira, L.P. Faverani, G.A.C. Momesso, E.R. Luvizuto, I. de OliveiraPuttini, R. Okamoto, Effect of antiresorptive drugs in the alveolar bone healing. Ahistometric and immunohistochemical study in ovariectomized rats, Clin OralInvestig 21 (2017) 1485–1494.

[10] W.H. Cheung, T. Miclau, S.K. Chow, F.F. Yang, V. Alt, Fracture healing in osteo-porotic bone, Injury 47 (Suppl. 2) (2016) S21–6.

[11] H. Chen, N. Liu, X. Xu, X. Qu, E. Lu, Smoking, radiotherapy, diabetes and osteo-porosis as risk factors for dental implant failure: a meta-analysis, PLoS One 8 (2013)e71955.

[12] F. de Medeiros, G.A.H. Kudo, B.G. Leme, P.P. Saraiva, F.R. Verri, H.M. Honorio,E.P. Pellizzer, J.F. Santiago Junior, Dental implants in patients with osteoporosis: asystematic review with meta-analysis, Int. J. Oral Maxillofac. Surg. 47 (4) (Apr2018) 480–491.

[13] C.M. Holahan, S. Koka, K.A. Kennel, A.L. Weaver, D.A. Assad, F.J. Regennitter,D. Kademani, Effect of osteoporotic status on the survival of titanium dental im-plants, Int. J. Oral Maxillofac. Implants 23 (2008) 905–910.

[14] G. Giro, L. Chambrone, A. Goldstein, J.A. Rodrigues, E. Zenobio, M. Feres,L.C. Figueiredo, A. Cassoni, J.A. Shibli, Impact of osteoporosis in dental implants: asystematic review, World J. Orthop. 6 (2015) 311–315.

[15] S.L. Kates, C.L. Ackert-Bicknell, How do bisphosphonates affect fracture healing?Injury 47 (Suppl. 1) (2016) S65–8.

[16] M. Zandi, A. Dehghan, P. Amini, L. Rezaeian, S. Doulati, Evaluation of mandibularfracture healing in rats under zoledronate therapy: a histologic study, Injury 48(2017) 2683–2687.

[17] C.B. Starr, M.A. Maksoud, Implant treatment in an urban general dentistry re-sidency program: a 7-year retrospective study, J. Oral Implantol. 32 (2006)142–147.

[18] F. Mangano, C. Mortellaro, N. Mangano, C. Mangano, Is low serum vitamin D as-sociated with early dental implant failure? A retrospective evaluation on 1625implants placed in 822 patients, Mediat. Inflamm. 2016 (2016) 5319718.

[19] S.R. Cummings, L.J. Melton, Epidemiology and outcomes of osteoporotic fractures,Lancet 359 (2002) 1761–1767.

[20] C.H. Chen, X. Pei, U.S. Tulu, M. Aghvami, C.T. Chen, D. Gaudilliere, M. Arioka,M. Maghazeh Moghim, O. Bahat, M. Kolinski, T.R. Crosby, A. Felderhoff,J.B. Brunski, J.A. Helms, A comparative assessment of implant site viability inhumans and rats, J. Dent. Res. 97 (4) (Apr 2018) 451–459 22034517742631.

[21] L. Wang, Y. Wu, K.C. Perez, S. Hyman, J.B. Brunski, U.S. Tulu, C. Bao, B. Salmon,J.A. Helms, Effects of condensation on peri-implant bone density and remodeling, J.Dent. Res. 96 (2017) 413–420.

[22] H.Z. Movat, Demonstration of all connective tissue elements in a single section;pentachrome stains, AMA Arch. Pathol. 60 (1955) 289–295.

[23] S. Minear, P. Leucht, J. Jiang, B. Liu, A. Zeng, C. Fuerer, R. Nusse, J.A. Helms, Wntproteins promote bone regeneration, Sci. Transl. Med. 2 (2010) 29ra30.

[24] M.L. Bouxsein, S.K. Boyd, B.A. Christiansen, R.E. Guldberg, K.J. Jepsen, R. Muller,Guidelines for assessment of bone microstructure in rodents using micro-computedtomography, J. Bone Miner. Res. 25 (2010) 1468–1486.

[25] R. Fuentes, T. Flores, P. Navarro, C. Salamanca, V. Beltran, E. Borie, Assessment of

buccal bone thickness of aesthetic maxillary region: a cone-beam computed tomo-graphy study, J. Periodontal. Implant. Sci. 45 (2015) 162–168.

[26] J.B. Payne, R.A. Reinhardt, P.V. Nummikoski, K.D. Patil, Longitudinal alveolar boneloss in postmenopausal osteoporotic/osteopenic women, Osteoporos. Int. 10 (1999)34–40.

[27] J.Y. Cha, M.D. Pereira, A.A. Smith, K.S. Houschyar, X. Yin, S. Mouraret,J.B. Brunski, J.A. Helms, Multiscale analyses of the bone-implant interface, J. Dent.Res. 94 (2015) 482–490.

[28] A. Monje, R. Gonzalez-Garcia, F. Monje, H.L. Chan, P. Galindo-Moreno, F. Suarez,H.L. Wang, Microarchitectural pattern of pristine maxillary bone, Int. J. OralMaxillofac. Implants 30 (2015) 125–132.

[29] L. Qi, X. Wang, M.Q. Meng, 3D finite element modeling and analysis of dynamicforce in bone drilling for orthopedic surgery, Int. J. Numer. Method Biomed. Eng.30 (2014) 845–856.

[30] G. Osterhoff, E.F. Morgan, S.J. Shefelbine, L. Karim, L.M. McNamara, P. Augat,Bone mechanical properties and changes with osteoporosis, Injury 47 (Suppl. 2)(2016) S11–20.

[31] B. Van Rietbergen, R. Huiskes, F. Eckstein, P. Ruegsegger, Trabecular bone tissuestrains in the healthy and osteoporotic human femur, J. Bone Miner. Res. 18 (2003)1781–1788.

[32] C. Lopez-Otin, M.A. Blasco, L. Partridge, M. Serrano, G. Kroemer, The hallmarks ofaging, Cell 153 (2013) 1194–1217.

[33] T. Finkel, N.J. Holbrook, Oxidants, oxidative stress and the biology of ageing,Nature 408 (2000) 239–247.

[34] L. Batoon, S.M. Millard, L.J. Raggatt, A.R. Pettit, Osteomacs and bone regeneration,Curr. Osteoporos. Rep. 15 (2017) 385–395.

[35] M.L. Oliveira, E.F. Pedrosa, A.D. Cruz, F. Haiter-Neto, F.J. Paula, P.C. Watanabe,Relationship between bone mineral density and trabecular bone pattern in post-menopausal osteoporotic Brazilian women, Clin Oral Investig 17 (2013)1847–1853.

[36] J. Lopez-Lopez, A. Estrugo-Devesa, E. Jane-Salas, R. Ayuso-Montero, C. Gomez-Vaquero, Early diagnosis of osteoporosis by means of orthopantomograms and oralX-rays: a systematic review, Med. Oral Patol. Oral Cir. Bucal. 16 (2011) e905–13.

[37] G. Jonasson, Bone mass and trabecular pattern in the mandible as an indicator ofskeletal osteopenia: a 10-year follow-up study, Oral Surg. Oral Med. Oral Pathol.Oral Radiol. Endod. 108 (2009) 284–291.

[38] G. Jonasson, G. Bankvall, S. Kiliaridis, Estimation of skeletal bone mineral densityby means of the trabecular pattern of the alveolar bone, its interdental thickness,and the bone mass of the mandible, Oral Surg. Oral Med. Oral Pathol. Oral Radiol.Endod. 92 (2001) 346–352.

[39] C.M. Esteves, R.M. Moraes, F.C. Gomes, M.S. Marcondes, G.M. Lima, A.L. Anbinder,Ovariectomy-associated changes in interradicular septum and in tibia metaphysis indifferent observation periods in rats, Pathol. Res. Pract. 211 (2015) 125–129.

[40] Z. Du, R. Steck, N. Doan, M.A. Woodruff, S. Ivanovski, Y. Xiao, Estrogen deficiency-associated bone loss in the maxilla: a methodology to quantify the changes in themaxillary intra-radicular alveolar bone in an ovariectomized rat osteoporosismodel, Tissue Eng. Part C Methods 21 (2015) 458–466.

[41] X.L. Liu, C.L. Li, W.W. Lu, W.X. Cai, L.W. Zheng, Skeletal site-specific response toovariectomy in a rat model: change in bone density and microarchitecture, Clin.Oral Implants Res. 26 (2015) 392–398.

[42] J. Wactawski-Wende, E. Hausmann, K. Hovey, M. Trevisan, S. Grossi, R.J. Genco,The association between osteoporosis and alveolar crestal height in postmenopausalwomen, J. Periodontol. 76 (2005) 2116–2124.

[43] C. Bural, H. Bilhan, A. Cilingir, O. Geckili, Assessment of demographic and clinicaldata related to dental implants in a group of Turkish patients treated at a universityclinic, J. Adv. Prosthodont. 5 (2013) 351–358.

C.-H. Chen et al. Bone 112 (2018) 212–219

219

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具