Uncoupled angiogenesis and osteogenesis in nicotine-compromised bone healing

Abstract

Nicotine is the main chemical component responsible for tobacco addiction. This study aimed to evaluate the influence of nicotine on angiogenesis and osteogenesis and the associated expression of angiogenic and osteogenic mediators during bone healing. Forty-eight adult New Zealand White rabbits were randomly assigned to a nicotine group and a control group. Nicotine pellets (1.5 g, 60-day time release) or placebo pellets were implanted in the neck subcutaneous tissue. The nicotine or placebo exposure time for all the animals was 7 weeks. Unilateral mandibular distraction osteogenesis was performed. Eight animals in each group were euthanized on day 5, day 11 of active distraction, and week 1 of consolidation, respectively. The mandibular samples were subjected to radiographic, histologic, immunohistochemical, and real-time reverse-transcriptase polymerase chain reaction examinations. Nicotine exposure upregulated the expression of hypoxia inducible factor 1α and vascular endothelial growth factor and enhanced angiogenesis but inhibited the expression of bone morphogenetic protein 2 and impaired bone healing. The results indicate that nicotine decouples angiogenesis and osteogenesis in this rabbit model of distraction osteogenesis, and the enhanced angiogenesis cannot compensate for the adverse effects of nicotine on bone healing. © 2010 American Society for Bone and Mineral Research

Introduction

Nicotine is the main chemical component responsible for tobacco addition.1 It is of the highest importance among the potentially toxic substances in tobacco products.1–3 Studies showed that nicotine delays bone healing, but the molecular mechanisms remains unclear.2, 4–8 Recently, we have developed a nicotine-induced rabbit model of mandibular distraction osteogenesis and confirmed the positive correlation between the blood nicotine concentration and compromised bone healing.4, 5 The molecular mechanism of nicotine-compromised bone healing could be explored conveniently with this animal model.

Distraction osteogenesis is a controlled surgical procedure that initiates a regenerative process. It applies mechanical strain to enhance the biologic responses in the injured tissues to create new bone. Distraction osteogenesis shares many features of embryonic growth, fetal growth, and neonatal limb development, as well as fracture repair.9, 10 Compared with bone fracture, in which the molecular signaling lasts only for a few days, the signaling in distraction regeneration is magnified and prolonged as long as the mechanical traction is active. The molecular signaling cascade induced by the mechanical strain plays a key regulatory role in translating traction forces into a biologic response of bone cells.

Angiogenic and osteogenic factors play an important role in bone healing and regeneration. Vascular endothelial growth factor (VEGF) is a potent angiogenic mediator inducing proliferation and migration of endothelial cells. Moreover, it has been shown to promote chemotaxis11 and differentiation of osteoblasts.12, 13 VEGF can interact synergistically with bone morphogenetic protein (BMP) to promote skeletal development and bone healing by enhancing cell recruitment, prolonging cell survival, and increasing angiogenesis.14 BMPs are the most potent osteogenic growth factors inducing the osteogenic differentiation of mesenchymal stem cells.15–17 BMP acts as an important regulator that stimulates production of VEGF in osteoblasts.18–20 Hypoxia is the most potent stimulus for VEGF expression.21, 22 Hypoxia-inducible factor 1α (HIF-1α), a central regulator of hypoxia adaptation in vertebrates, plays a key role in development, physiology, and disease23 and activates downstream hypoxia-responsive genes such as VEGF.24–28 We hypothesized that nicotine exposure affects angiogenesis and osteogenesis by altering the gene expression of angiogenic and osteogenic factors in bone regeneration. In this study, we assessed angiogenesis, osteogenesis, and the expression of HIF-1α, VEGF, and BMP-2 in the nicotine-induced rabbit model of mandibular distraction. Nicotine decouples angiogenesis and osteogenesis in this experimental model. Enhanced angiogenesis cannot compensate for the adverse effects of nicotine on bone healing.

Materials and Methods

Animal care

The rabbits were kept in a dedicated animal holding facility under veterinary supervision in the Laboratory Animal Unit of Li Ka Shing Faculty of Medicine, University of Hong Kong. The animal experiment was approved by the Committee on the Use of Live Animals for Teaching and Research of the University of Hong Kong.

Nicotine implantation

Forty-eight male adult New Zealand White rabbits (9 months old, 3.4 to 4.0 kg) were randomly assigned to a nicotine group and a control group (n = 24 for each group). Then 1.5-g, 60-day time-release nicotine pellets or placebo pellets (Innovative Research of America, Sarasota, FL) were implanted in the neck subcutaneous tissue of the rabbits. The total nicotine exposure time was 7 weeks, and the animals were exposed to nicotine for at least 4 weeks before mandibular osteotomy (Fig. 1).

Figure 1. Time line of nicotine exposure and distraction osteogenesis. The time of euthanasia: on day 5 (A), on day 11 (B), and on day 18 (C) respectively, after the commencement of active distraction. Nicotine exposure: 7 weeks; latency period: 3 days; active distraction: 11 days.

Osteotomy and distraction procedures

After nicotine implantation, a standard procedure of mandibular osteotomy and distraction used in our previous study5 was performed. Briefly, the animals were given a preoperative dose of antibiotic and analgesic. After anesthesia, the skin was incised along the ventral border on one side of the mandibular body. A straight-body osteotomy was made immediately cranial (anterior) to the first premolar root. A custom-made bone-borne distractor was placed along a plane perpendicular to the osteotomy and fixed by 2-mm-diameter titanium screws. The periosteum, muscle, and skin were repositioned and closed with 3–0 sutures. Each animal remained under close observation by a veterinary technician until it regained consciousness. Postoperative antibiotic and analgesics were administered. The clinical condition, weight, and food consumption of the animals were monitored. After 3 days of latency, the distractor was activated at 0.9 mm per day. Eight animals in each group were euthanized with an overdose of pentobarbital sodium on day 5 (middle of active distraction), day 11 (end of active distraction), and day 18 (week 1 of consolidation), respectively, after the commencement of active distraction. Three of the eight animals were subjected to radiographic, histologic, and immunohistochemical examinations, and the other five were subjected to mRNA expression analysis.

Plain radiography

The mandibular samples were harvested and fixed in 10% neutral phosphate-buffered paraformaldehyde. Each specimen was placed on an occlusal film with the lingual side touching the film. Plain radiography was performed by an Orthoralix 9200 X-ray machine (Gendex, Des Plaines, IL) under a standard conditions of 50 kV and 16 mA.

Micro-computed tomography (µCT)

After plain radiographic examination, the distracted regenerate and 2 to 5 mm of neighboring host bone were harvested. The specimens were subjected to quantitative examination by a µCT20 system (Scano Medical AG, Bassersdorf, Switzerland) using a standard protocol described in our previous study.5 Each harvested specimen was placed in a 17-mm-diameter sample holder with the sagittal plane of the mandibular regenerate vertical to the X-ray tube. The serial scanned images of each specimen were inspected on the computer. On each scanned image, the total area of the distraction regenerate was defined as the region of interest (ROI). The bone volume fraction (the ratio between bone volume and total volume, BV/TV) of the ROI on each section was calculated individually, and a mean value of BV/TV for the total regenerate was obtained.

Histologic examination

After µCT examination, the specimens were decalcified in a solution of 14.5% ethylenediaminetetraacetic acid (EDTA) buffer (pH 7.2) at room temperature. The specimens were dehydrated and embedded in paraffin. Axial sections of 5 µm in thickness were cut with a microtome and stained with hematoxylin and eosin for light microscopy.

Immunohistochemical staining

The sections were incubated with primary goat antibodies against type IV collagen (Col IV, Southern Biotech, Birmingham, AB), BMP-2 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), VEGF (Santa Cruz Biotechnology, Inc.), and primary mouse antibody against HIF-1α (Abcam, Cambridge, MA) and CD31 (Abcam) overnight at 4°C. The antibodies have been confirmed to recognize the rabbit-specific signals in previous studies.5, 29–32 For negative controls, the primary antibodies were omitted. Goat and mouse ABC staining system kits (Santa Cruz Biotechnology, Inc.) were used to detect the reaction. The sections were counterstained with hematoxylin and observed using a computer-assisted image-analyzing system (Eclipse LV100POL and DS-Ri1, Nikon, Melville, NY) with morphometric software (NIS-Elememts AR 3.0, Nikon).

Col IV is a component of the basal lamina of vessels. Its staining allows the proper identification of blood vessels by immunohistochemical analysis.5, 33, 34 The intensity of staining by CD31 is far less than that by Col IV. It dose not always stain all endothelial cells making up a vessel.34 CD31 may be a suitable marker to combine with Col IV for the estimation of blood vessels.34 To evaluate the neovessel density (NVD), the distraction regenerate on each slide was divided into six areas (three rows and two columns for the sections on day 5 of active distraction) or nine areas (three rows and three columns for the sections on day 11 of active distraction and week 1 of consolidation) at ×1 magnification of the objective lens. The vessels in the center of these areas were counted at ×5 magnification of the objective lens. According to a standard technique described previously, any single brown-stained cell or cluster of endothelial cells that was clearly separated from adjacent vessels, histiocytes, and other connective tissue elements was considered a vessel, and the branching structures were counted as a single vessel unless there was a discontinuity in the structure.5, 33 NVD was calculated by vessel number per observation area.

Real-time reverse-transcriptase polymerase chain reaction (RT-PCR)

The distraction regenerate samples were harvested before the animals were sacrificed. Under anesthesia, the skin and muscle were incised and elevated to expose the distraction regenerate. The regenerate tissue was removed and homogenized using Mikro-11 Dismembrator U (Braun Biotech International, Melsungen, Germany). Total RNA was isolated with an RNeasy Tissue Midi Kit (Qiagen, Hilden, Germany).

cDNA was synthesized using the Superscript first-strand synthesis system (Invitrogen, Carlsbad, CA). The primers were VEGF forward: 5'-TCCAGGAGTACCCTGATGAGA-3'; VEGF reverse: 5'-CCCTGGTGAGGTTTGATCC-3' (157 base pairs; GenBank Accession Number AF022179); BMP-2 forward: 5'-CACTTGGAGGAGAAGCAAGG-3'; BMP-2 reverse: 5'-GCTGTTTGTGTTTCGCTTGA-3' (172 base pairs; GenBank Accession Number AF041421); HIF-1α forward: 5'-TTACAGCAGCCAGATGATCG-3'; HIF-1α reverse: 5'-TGGTCAGCTGTGGTAATCCA-3' (178 base pairs; GenBank Accession Number AY273790); and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) forward: 5'-TCACCAGGGCTGCTTTTAAC-3'; GAPDH reverse: 5'-GCTGAGATGATGACCCTTT-3' (317 base pairs; GenBank Accession Number L23961). Amplification was carried out for 35 cycles (94°C for 1 minute, 56°C for 1 minute, and 72°C for 1 minute) for each in a 50-µL reaction solution containing 1 µL of each cDNA, 0.5µM of each pair of primers, 0.2 mM of each dNTP, 1× PCR buffer, 1.5 mM MgCl₂, and 1 U of taq DNA polymerase (Invitrogen). Standards were constructed by cloning each PCR product into a 3.9-kb pCR 2.1-TOPO with a TOPO TA cloning kit (Invitrogen) and then purified with AIAGEN Plasmid Minikit (Qiagen). Real-time RT-PCR was carried out for each in a 30-µL reaction solution containing 10 µL of each cDNA, 15 µL of 2× Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA), and 0.5 µM of each pair of primers. The standard curve was created by 10-fold dilutions of the standard samples. The quantification of mRNA expression was analyzed using Real Time PCR System software (Applied Biosystems). Absolute quantification was performed by comparing the target threshold cycles directly with the absolute standard curve for each amplification. The copy numbers of BMP2, VEGF, and HIF1α genes were normalized with the copy numbers of GAPDH.

Statistical analysis

The mRNA expression values between the two groups were compared by two-sample t test (Version 11.0 of Statistical Package of Social Sciences software, SPSS, Inc., Chicago, IL). A statistical result of less than .05 was considered significant.

The t test assumes that the data are sampled from populations that follow Gaussian distributions and have equal standard deviations. To compare the values of mRNA expression, five animals are necessary in each subgroup to pass the assumption tests. The sample size for the radiographic and immunohistochemical analysis was estimated based on the results from our previous studies,5 and three animals in each subgroup were considered adequate for the present experiment.

Results

Clinical examination

All rabbits completed the experimental process uneventfully. The animals showed mild weight loss after the operation and started to regain the weight within 2 weeks. None of the animals experienced any postoperative complications, and the distractors remained stable until the time of sacrifice.

Plain radiography

The distraction regenerate was detected by the low radiodensity between the host bone segments. On day 5 of active distraction, the distraction gap in both the control and nicotine groups was radiolucent without obvious signs of new bone formation (Fig. 2A, B). On day 11 of active distraction, radiopaque streaks extending from the bony edges with nonunion in the center were noted. The radiodensity of the regenerate in the nicotine group was lower than that in the control group (Fig. 2C, D). On week 1 of consolidation, partial union was noted in the center of the distraction regenerate in all animals belonging to the control group (Fig. 2E). Nonunion in the center was noted in 2 of the 3 animals in the nicotine group (Fig. 2F).

Figure 2. Lateral radiographic view of the rabbit hemimandibles (distraction regenerates shown by arrows). (A) Control group on day 5. (B) Nicotine group on day 5. (C) Control group on day 11. (D) Nicotine group on day 11. (E) Control group on day 18. (F) Nicotine group on day 18.

µCT

The bone formation in the distraction regenerate was quantified by µCT analysis. A gradual increase in bone volume fraction from active distraction to consolidation was noted in both the control and nicotine groups. When the two groups were compared, the difference in bone volume fraction was not significant on day 5 of active distraction. However, the bone volume fraction in the nicotine group was significantly less than that in the control group on day 11 of active distraction and week 1 of consolidation (Table 1).

Table 1. BV/TV (mean ± SD, %) in the rabbit mandibular distraction regenerates (n = 3).

Group	Day 5	Day 11	Day 18
* p < .05 is considered statistically significant.
Control	1.36 ± 0.443	11.96 ± 1.383	19.95 ± 1.318
Nicotine	0.93 ± 0.181	6.34 ± 1.875	17.24 ± 0.910
P value	0.1923	0.0140*	0.0425*

Histology

On day 5 of active distraction, the distracted gap was bridged by fibrous tissue, and hemorrhage was seen in the central area of the distraction regenerate in both the control and nicotine groups (Fig. 3A, B).

Figure 3. Histologic sections of the distraction regenerate in rabbit mandibles (H&E stain). (A) Control group on day 5. (B) Nicotine group on day 5. (C) Control group on day 11. (D) Nicotine group on day 11. (E) Control group ont day 18. (F) Nicotine group on day 18. H = hemorrhage; F = fibrous tissue; T = trabeculae.

On day 11 of active distraction, the distracted gap in the control group was mainly filled with thin longitudinal bony trabeculae aligned in the direction of the distraction vector from both sides of the bony margins. Fibrous tissues were observed in the central area of the distraction regenerate (Fig. 3C). In the nicotine group, the distraction zone was bridged mostly by fibrous tissue, and hemorrhage was seen in the central area of the distraction regenerate (Fig. 3D).

At week 1 of consolidation, the distraction regenerate in the control group was composed of primary trabeculae and loose fibrovascular stroma. Small fibrous discontinuities were seen in the central area (Fig. 3E). In the nicotine group, new bone was formed at the edges of the distracted gap, and the central area of the distraction regenerate was occupied with fibrous tissues (Fig. 3F).

Neovessel density

Col IV expression was observed in the cytoplasm of the vascular endothelium. In the distraction regenerate, the signals in the areas adjacent to the host bone were more intense than those in the central area. Figure 4 presents the Col IV–staining sections in the distal middle areas of distraction regenerate. On day 5 of active distraction, the signals occurred in the capillary-like cell clusters (Fig. 4A, B). On day 11 of active distraction, cannular vessels were labeled in the control group (Fig. 4C). In the nicotine group, most of the labeled cell clusters became capillary loops (Fig. 4D). At week 1 of consolidation, the control group showed that primary trabeculae were obvious, and cannular vessels distributed among the trabeculae (Fig. 4E). In the nicotine group, tiny dense vessels were noted among the slender immature trabeculae (Fig. 4F). Compared with Col IV, the expression of CD31 in the endothelia of vessels was weaker or even absent. The signals of CD31 also were observed in osteoclasts (Fig. 5). The low intensity of CD31 staining in vessels may be related to the experimental model, as well as to the time point of observation in the present study.

Figure 4. Col IV expression in the rabbit mandibular distraction regenerates. Blood vessels are visualized by 3,3′-Diaminobenzidine (DAB) (brown coloration). (A) Control group on day 5. (B) Nicotine group on day 5. (C) Control group ont day 11. (D) Nicotine group on day 11. (E) Control group on day 18. (F) Nicotine group on day 18.

Figure 5. CD31 expression in the rabbit mandibular distraction regenerates on day 18. Panel B is a section in panel A at higher magnification. The expression is weaker or even absent in the endothelia of vessels. The signals are also observed in osteoclasts (arrows).

To quantify the new vessels, the density of the blood vessels stained by Col IV was evaluated and represented by NVD. The nicotine group showed a significantly higher NVD than the control group during active distraction and at week 1 of consolidation (Table 2).

Table 2. Neovessel Density (mean ± SD, vessles/mm²) in the Rabbit Mandibular Distraction Regenerates (n = 3)

Group	Day 5	Day 11	Day 18
* p < .05 is considered statistically significant.
Control	11 ± 2.5	9 ± 2.0	17 ± 2.1
Nicotine	22 ± 2.1	19 ± 2.1	25 ± 3.1
p Value	.0100*	.0102*	.0269*

Expression of BMP-2, VEGF, and HIF-1α

Positive signals of BMP-2 were detectable in hemorrhage, fibroblasts, osteoblasts, and fibrous matrix. VEGF was widely expressed in hemorrhage, fibroblasts, osteoblasts, osteocytes, and fibrous matrix and bone matrix of trabeculae. Intense HIF-1α expression was noted in hemorrhage and osteoblasts lining the newly formed trabeculae, and very weak signals also were detected in fibroblasts and some immature osteocytes. The nicotine group showed much weaker BMP-2 signals in osteoblasts, whereas HIF-1α signals in osteoblsts were more intense. The stronger expression of VEGF in fibroblasts, osteoblasts, and osteocytes was detected in the nicotine group (Fig. 6).

Figure 6. The expression of BMP-2 (A, B), VEGF (C, D), and HIF-1α (E, F) in the rabbit mandibular distraction regenerates at week 1 of consolidation. (A, C, E) Control group. (B, D, F) Nicotine group. Bars = 10 µm. Compared with the control group, the nicotine group shows that BMP-2 expression in osteoblasts is much weaker, whereas VEGF signals in fibroblasts, osteoblasts, and osteocytes and HIF-1α signals in osteoblsts are more intense.

mRNA expression of BMP2, VEGF, and HIF1α was detected in the distraction regenerates. Quantified by real-time RT-PCR, their expression levels increased gradually from active distraction to week 1 of consolidation. When the mRNA levels between the two groups were compared, BMP2 expression decreased, whereas expression of VEGF and HIF1α increased in the nicotine group. Significant differences in the expression of BMP2 (day 5 of distraction: p = .0003; week 1 of consolidation: p < .0001) and VEGF (day 5 of distraction: p = .0007; week 1 of consolidation: p = .0002) were detected on day 5 of active distraction and at week 1 of consolidation. HIF1α expression between the two groups showed a significant difference at week 1 of consolidation (week 1 of consolidation: p = .0015). The gene expression for BMP2 on day 11 of active distraction (p = .7293), VEGF on day 11 of active distraction (p = .0536), and HIF1α on days 5 and 11 of active distraction (day 5: p = .0585; day 11: p = .0869) showed no statistically significant difference. (Fig. 7)

Figure 7. The quantification of mRNA expression of BMP2 (A), VEGF (B), and HIF1α (C) in the rabbit mandibular distraction regenerates. *p < .01; **p < .001; ***p < .0001.

Discussion

Many studies show that angiogenesis and osteogenesis are tightly coupled during bone formation.24, 35–37 Angiogenesis plays a pivotal role in skeletal development and bone repair.24, 35–37 Enhanced angiogeneis led to the increased bone coverage and mineral density in bone defect reconstructions,25, 38, 39 whereas the administration of antiangiogenic agents inhibited bone healing.24, 38, 40–44 Interestingly, this study found an uncoupling of neovessel formation and bone formation in the nicotine-induced distraction osteogenesis model. Nicotine-stimulated angiogenesis should be able to facilitate bone formation. However, an impairment of bone healing was noted in this study.

These results revealed a significantly enhanced expression of HIF-1α and VEGF associated with consistently increased neovessel density in the nicotine group. In our previous study using the same experimental model, we found that nicotine exposure reduced blood perfusion, resulting in ischemia and lower oxygen level.5 Tissue hypoxia is the major stimulus for initiating the angiogenic cascade.26–28 Nicotine has been found to stimulate the accumulation of HIF-1α,45 which is a central regulator of hypoxia adaptation and activates downstream hypoxia-responsive genes such as VEGF.23–28 However, the increased vessel formation did not lead to an increased blood supply. Besides carrying oxygen and nutrients to bone tissue, blood flow play an active role in bone formation and remodeling by mediating the interactions among osteoblasts, osteocytes, osteoclasts, and vascular cells at a variety of levels.46 The uncoupled vessel density and blood perfusion implied a complex mechanism of nicotine in controlling angiogenic activity and blood perfusion. The reduced blood flow indicates that nicotine may produce vasoconstriction during bone regeneration. Nicotine was reported to induce vascular endothelial dysfunction.47–49 It has a direct effect on small blood vessels in producing vasoconstriction and systemic venoconstriction,50–54 but this effect in bone healing has not been reported by others. In our bone-healing model, the direct effects of nicotine on blood vessels may be responsible for the reduced blood flow. Our results suggest that hypoxia and ischemia owing to nicotine exposure could stimulate HIF-1α expression, leading to an increased expression of VEGF. This, in turn, stimulates angiogenesis. However, the enhanced vessel formation is incapable of compensating for the adverse effect of the reduced blood flow possibly caused by nicotine-induced vasoconstriction.

BMPs are the most important osteogenic growth factors.15–17 BMP-2 can reliably induce both ectopic and orthotopic bone formation at the site of administration.55–61 The expression of endogenous BMPs is regarded as one of the indices to evaluate the biologic environment in distraction regenerate.5, 29 The effect of nicotine on BMPs has not been fully studied. Our previous immuhistochemical study demonstrated that nicotine inhibited BMP expression in osteoblasts. In this study, the inhibitory effect of nicotine exposure on BMP2 mRNA expression was detected in the whole block of distraction regenerates, which further confirmed that nicotine depressed osteogenic activity in bone regeneration.

Taking together, two reasons may be responsible for the impaired bone healing in the present experimental model. First, nicotine decreases blood perfusion by its direct effects on blood vessels in producing vasoconstriction and systemic venoconstriction, even though it increases angiogenesis. Second, nicotine directly inhibits the osteogenic activity (Fig. 8).

Figure 8. Schematic showing the effect of nicotine on bone regeneration. Nicotine inhibits BMP expression and associated osteogenesis. At the same time, it causes vasoconstriction, which leads to hypoxia and ischemia. The induced HIF-1α stimulates VEGF expression and associated angiogenesis. However, this stimulatory effect cannot compensate for the adverse effect of nicotine on bone healing.

It is known that VEGF and BMP act synergistically during bone healing.19, 62, 63 The synergistic interaction between VEGF and BMP depends on the ratios of the two factors.14, 19, 63 Excessive VEGF may lead to impairment in bone formation, possibly by promoting mesenchymal stem cell differentiation toward an endothelial lineage,64 consequently reducing the availability of mesenchymal stem cells (MSCs) for osteogenic differentiation.65 Alternatively, excessive VEGF may increase recruitment of osteoclasts into the bone-regeneration sites and lead to an excessive bone resorption.65 The disruption of the optimal ratio between VEGF and BMP caused by nicotine also might contribute to the compromised bone healing. In addition, the inflammatory response to bone fracture or distraction plays an important role in initiating the repair cascade. It activates downstream factors such as cytokines and growth factors that recruit osteoprogenitor and mesenchymal cells to the injury site.64, 66, 67 Nicotine is an anti-inflammatory agent.68–71 The suppression of inflammation by nicotine may have an adverse effect on bone healing. However, the conclusion cannot be drawn on these speculations before finding hard evidence.

Distraction osteogenesis relies on the application of controlled mechanical force to promote bone induction and formation between two osteotomy fronts. It has become a widely accepted surgical approach in the treatment of congenial and acquired bone deformities.9, 72, 73 In this study, the significant differences in mRNA expression levels between the nicotine and control groups were noted on day 5 of distraction and at week 1 after distraction, except for day 11 of active distraction. During distraction osteogenesis, the mechanical strain triggers and sustains molecular signaling. The expression of BMP-2, VEGF, and HIF-1α can be induced gradually during the active distraction.24, 74–78 The effect of mechanical strain on the molecular signaling accumulates gradually and eventually may cover the effect of nicotine. Thus distraction osteogenesis could be the preferred choice among the various bone-reconstruction methods available to treat patients who have compromised healing ability, such as smokers and those taking nicotine medication.

In summary, nicotine exposure decouples angiogenesis and osteogenesis in this experimental model of mandibular distraction osteogenesis. Nicotine enhances blood vessel density and stimulates the associated HIF-1α and VEGF expressions but impairs bone formation and inhibits the associated BMP expression. The uncoupling of angiogenesis and osteogenesis may be explained by the complex effects of nicotine on blood vessels and osteogenic activity during bone healing.

Disclosures

The first two authors contributed equally to this research. All the authors state that have no conflicts of interest.

Acknowledgements

This study was supported by the Small Project Funding Programme (Reference code HKU200507176099) from the University of Hong Kong. We appreciate the valuable advice given by Professor J Glowacki from the Harvard School of Dental Medicine. We also appreciate the technical assistance provided by the Laboratory Animal Unit of the Li Ka Shing Faculty of Medicine and the Centralized Research Laboratories of the Faculty of Dentistry.