A composite critical-size rabbit mandibular defect for evaluation of craniofacial tissue regeneration

Abstract

Translational biomaterials targeted toward the regeneration of large bone defects in the mandible require a preclinical model that accurately recapitulates the regenerative challenges present in humans. Computational modeling and in vitro assays do not fully replicate the in vivo environment. Consequently, in vivo models can have specific applications such as those of the mandibular angle defect, which is used to investigate bone regeneration in a nonload-bearing area, and the inferior border mandibular defect, which is a model for composite bone and nerve regeneration, with both models avoiding involvement of soft tissue or teeth. In this protocol, we describe a reproducible load-bearing critical-size composite tissue defect comprising loss of soft tissue, bone and tooth in the mandible of a rabbit. We have previously used this procedure to investigate bone regeneration, vascularization and infection prevention in response to new biomaterial formulations for craniofacial tissue engineering applications. This surgical approach can be adapted to investigate models such as that of regeneration in the context of osteoporosis or irradiation. The procedure can be performed by researchers with basic surgical skills such as dissection and suturing. The procedure takes 1.5-2 h, with ?2 h of immediate postoperative care, and animals should be monitored daily for the remainder of the study. For bone tissue engineering applications, tissue collection typically occurs 12 weeks after surgery. In this protocol, we will present the necessary steps to ensure reproducibility; tips to minimize complications during and after surgery; and analytical techniques for assessing soft tissue, bone and vessel regeneration by gross evaluation, microcomputed tomography (microCT) and histology.

Full text

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 1989
INTRODUCTION
Bone defects in the mandibular region due to trauma, pathology
or congenital defects present a substantial challenge to clinicians
attempting to regenerate tissue and restore function and esthet-
ics to patients1. Some of these difficulties include large volumes
of tissue loss, loss of multiple tissue types (e.g., bone, soft tissue,
teeth, nerves) and microbial contamination and/or infection of
regenerating tissues by endogenous oral bacteria. Animal experi-
mental models are a critical component in the development and
preclinical evaluation of new materials aimed at promoting
tissue regeneration, as they provide evidence to federal regulatory
committees that new products or materials can be safe and effica-
cious in an in vivo setting. This evidence is usually required by
regulatory committees in order for new materials or technologies
to be evaluated in human patients and to make the transition to
clinical use, underscoring the importance of animal models in the
translation of basic science to affecting human health.
Although small bone defects can be healed without intervention,
large defects can approach or exceed a critical size, which is classi-
cally defined as the smallest size defect that will not heal over the
lifetime of the animal2. Defects larger than the critical size result
in nonunion, and fibrous tissue grows in the intervening space
instead of regenerated bone3,4. Modern usage of the term ‘critical
size’ usually does not indicate the smallest possible size or extend
healing time to the lifetime of the animal. Instead, ‘critical size’
refers to any size defect that does not heal over a specified time
period. This distinction probably came about because age, sex,
strain and other characteristics of an animal population may influ-
ence healing, making it difficult to accurately determine a smallest
possible defect size, and also because of the practical limitations of
extending healing time to the lifetime of an animal during model
development5. Because of these difficulties, modern researchers
in bone regeneration have focused on developing reproducible
experimental models that result in a nonhealing defect over a time
period relevant to the growth of bone6. Many medical products
and biomaterials have been developed to facilitate the regeneration
of large bone defects, with the eventual goal of clinical translation
for human use7–9. However, despite advances in computational
modeling10 and in vitro techniques11, no ex vivo models exist that
can recapitulate the complexities of the in vivo environment, and
reproducible, validated experimental models in an animal remain
the most stringent option for evaluating the safety and efficacy of
new technologies before clinical translation.
Animal models of bone regeneration
Many factors contribute to the choice of an animal experimental
model for testing bone regeneration strategies2,12. For mandibular
bone regeneration, animal models are generally divided into small-
animal models (mouse13, rat14,15 and rabbit4,16) and large-animal
models (dog17, goat18,19, pig20,21 and sheep22). Small-animal
models are commonly used for ethical, economic and statistical
considerations, whereas large animal models are primarily used if
small-animal models are not suitable for the replication of a clinical
scenario or for proof-of-concept testing before clinical transla-
tion. Recent reviews in the literature discuss the advantages and
limitations of small- and large-animal models, as well as proce-
dural and experimental considerations vital to the implementation
and evaluation of materials in these models23,24. For this model,
the rabbit was chosen because it is an inexpensive and easy-to-
maintain animal that allows for studies with sufficient statistical
power, as determined by power analyses from pilot studies and/or
historical data. In addition, this animal allows for implantation
of biomaterial scaffolds of reasonable size, as the molar/premolar
region of the rabbit mandible is of adequate size (17 mm long,
16 mm high and 6 mm deep) to allow easy surgical access and
A composite critical-size rabbit mandibular defect
for evaluation of craniofacial tissue regeneration
Sarita R Shah1, Simon Young2, Julia L Goldman3, John A Jansen4, Mark E Wong2 & Antonios G Mikos1
1Department of Bioengineering, Rice University, Houston, Texas, USA. 2Department of Oral and Maxillofacial Surgery, University of Texas Health Science Center
at Houston, Houston, Texas, USA. 3Center for Laboratory Animal Medicine and Care, University of Texas Health Science Center at Houston, Houston, Texas, USA.
4Department of Biomaterials, Radboud University Medical Center, Nijmegen, the Netherlands. Correspondence should be addressed to A.G.M. (mikos@rice.edu).
Published online 22 September 2016; doi:10.1038/nprot.2016.122
Translational biomaterials targeted toward the regeneration of large bone defects in the mandible require a preclinical model that
accurately recapitulates the regenerative challenges present in humans. Computational modeling and in vitro assays do not fully
replicate the in vivo environment. Consequently, in vivo models can have specific applications such as those of the mandibular
angle defect, which is used to investigate bone regeneration in a nonload-bearing area, and the inferior border mandibular defect,
which is a model for composite bone and nerve regeneration, with both models avoiding involvement of soft tissue or teeth.
In this protocol, we describe a reproducible load-bearing critical-size composite tissue defect comprising loss of soft tissue, bone
and tooth in the mandible of a rabbit. We have previously used this procedure to investigate bone regeneration, vascularization
and infection prevention in response to new biomaterial formulations for craniofacial tissue engineering applications. This surgical
approach can be adapted to investigate models such as that of regeneration in the context of osteoporosis or irradiation. The
procedure can be performed by researchers with basic surgical skills such as dissection and suturing. The procedure takes 1.5–2 h,
with ~2 h of immediate postoperative care, and animals should be monitored daily for the remainder of the study. For bone tissue
engineering applications, tissue collection typically occurs 12 weeks after surgery. In this protocol, we will present the necessary
steps to ensure reproducibility; tips to minimize complications during and after surgery; and analytical techniques for assessing
soft tissue, bone and vessel regeneration by gross evaluation, microcomputed tomography (microCT) and histology.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
1990 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
implantation. The size of the scaffold should be considered when
choosing an animal model, as the use of larger implants permits
larger doses of drug incorporation for delivery applications and
can reveal important information about mass transport limita-
tions of scaffolds as they approach clinically relevant sizes.
For craniomaxillofacial applications, it is important to match the
intended use of the material to the site of the defect. Calvarial and
long-bone defect models are well described, reproducible, rapid
and economical5. However, cranial defects are nonload-bearing,
and long-bone defects experience different load-bearing patterns
as compared with those imposed by mastication. Therefore, these
models are less appropriate for testing of materials that undergo
the stress of mastication. In addition, the regenerative cell popula-
tions and environment present in the mandible and oral cavity are
unique and cannot be replicated in other areas of the animal. For
mandibular applications, the two most common sites for defects
are the angle and the body of the mandible. The angle of the man-
dible is relatively easy to access surgically, and it produces minimal
impact on the animal’s ability to masticate. However, the angle of
the mandible is a thin piece of bone, is poor in marrow and does
not contain teeth, which may limit the versatility of the model
for bone regeneration applications25,26. The angle defect model
may, however, be suitable for investigation of materials that draw
on areas of limited cellularity and mechanical loading for bone
regeneration. By contrast, defects produced in the body of the
mandible can be leveraged to investigate regeneration of several
tissue types in a load-bearing area with dentition. Depending
on the type of composite tissue defect desired, defects can be
surgically created either in the inferior border of the mandible
or in the body of the mandible with extension into the overlying
dentition4,27. Inferior border defect models allow for the evalua-
tion of combined bone and nerve regeneration28, whereas defects
in the body of the mandible allow for evaluation of bone, tooth
and oral mucosa repair4. Mandibular defects can be easily repro-
duced in a consistent manner, which is of crucial importance in
studies with multiple experimental groups. Reproducibility is also
a critical component in the ethical use of animals, as it minimizes
the total number of animals needed to evaluate a biomaterial or
regenerative strategy.
The mandibular defect in rabbits is a suitable experimental
model for the evaluation of specific hypotheses related to tissue
response, tissue regeneration and drug delivery from biomaterial
scaffolds. Our laboratory has developed a protocol for a mandibu-
lar defect model in the body of the mandible that can be used to
evaluate bone regeneration and neovascularization in response
to new biomaterials or tissue engineering constructs4.
Overview of the procedure
Dissection. Using an extraoral approach, a 10-mm defect is
created in the molar/premolar region of the rabbit mandible.
The extraoral approach involves dissection through the inferior
mandibular region through fascia and muscle, with care taken
to avoid the facial artery, in order to expose a hemi-mandible.
An overview of the procedure outlining the following steps can
be found in Figure 1.
Mandibular defect. After exposure of the hemi-mandible,
a 10-mm trephine bur is used to create a partial-thickness or
full-thickness bone defect4. The partial-thickness defect (Model 1,
Steps 17–19) comprises removal of the buccal mandibular cortex
and underlying tooth roots, leaving the endosteum and periodon-
tal ligament intact as a potential source of regenerative stem cells4.
The partial-thickness defect heals completely after 16 weeks and
is not considered critical size. The full-thickness defect (Model 2,
Steps 17–20) comprising removal of the buccal cortex, tooth roots
and lingual cortex, does not heal after 16 weeks and is considered
Tooth removal
Model 1: Steps 17–19
Partial thickness
defect
Exposure of
hemi-mandible
Removal of buccal
cortex and tooth roots
Removal of
lingual cortex
Implantation of
biomaterial
Application of
fixation plate
Closure in three layers
(muscle, fascia, skin)
Model 2: Steps 17–20
Full-thickness defect
Model 3: Steps 17–22
Full-thickness defect
with intraoral
communication
Pathogen
inoculation
Model 4: Steps 17–23
Full-thickness defect
with intraoral
communication and
pathogen inoculation
Figure 1 | Overview of the rabbit mandibular defect procedure. The different procedural options are highlighted in blue and labeled Models 1–4.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 1991
critical size within that time frame. A round 10-mm trephine
bur is used because it is the maximum size that can be used to
create a defect without interrupting the continuity of the
mandible, and the use of a round trephine bur facilitates the pro-
duction of a reproducible defect4.
Intraoral communication. If communication with the oral
cavity is desired (Model 3, Steps 17–22), a dental bur can be used
to remove a single tooth and expose the defect to the contents
of the mouth29–31. The removal of a tooth results in a compos-
ite defect comprising loss of bone and tooth with an associated
soft-tissue defect, which is useful for stringently evaluating the
resilience of constructs in a nonsterile environment32.
Inoculated defects. The regenerating defect can be further chal-
lenged by inoculation with bacteria (Model 4, Steps 17–23)33,34.
After creation of the defect in the mandible, a tissue engineering
construct or device can be placed into the defect, and then the
wound is closed.
Evaluation. Evaluation of bone regeneration4 and vasculari-
zation35 can be performed by microcomputed tomography
(microCT). Histologic analysis provides information regard-
ing the host response to the implant on a cellular level, and
immunohistochemistry can be performed to evaluate spatial
and temporal expression of growth factors36. Furthermore, this
in vivo model can be leveraged to evaluate systemic effects that
are relevant to the constructs being investigated, and we illus-
trate ancillary testing that may be considered before submitting
protocols for approval.
Applications, advantages and limitations of the protocol. The
protocol outlined here allows the investigator to use a single surgi-
cal approach to evaluate specific anatomic or application-based
variations applicable to a wide variety of tissue engineering con-
structs, such as new biomaterials and drug-releasing implants4,
which are summarized in Figure 1 and Table 1.
The development of this protocol for our laboratory has focused
on the use of implants for bone regeneration and drug-release
strategies to address infected defects. However, this protocol could
also be used to address other areas of importance in craniofacial
tissue engineering and dental applications, such as tooth regen-
eration, simultaneous alveolar bone and tooth regeneration, bar-
rier membrane regeneration and tissue regeneration combined
with dental implants23,24. This model can also be expanded to
include rabbits that undergo treatment to model compromised
populations. For example, rabbits can be ovariectomized before
surgery in order to evaluate tissue regeneration in an osteoporotic
population37. Bone regeneration and infection prevention in
irradiated bone are also common clinical concerns, and rabbits
may undergo radiation before or after the proposed procedure in
order to evaluate biomaterials or implant performance in bone
compromised by radiation38,39.
Computational and in vitro methods. Computational and/or
in vitro methods of evaluating tissue engineering constructs should
be considered before attempting in vivo work. Computational
modeling in tissue engineering is generally aimed at understand-
ing the interface between cells and biomaterials10. Molecular,
kinetic and chemomechanical adhesion modeling can provide
insight into single cell–biomaterial or cell population–biomaterial
TABLE 1 | Summary of potential models, applications, tissues involved and advantages/disadvantages of each model.
Option
Procedure
steps Applications
Regeneration/
infection Advantages Disadvantages
Model 1:
partial-thickness
defect
17–19 Evaluate strategies to
accelerate or augment
tissue regeneration
Alveolar bone,
tooth
A simple procedure with
minimal risk—tooth crowns
and lingual cortex are intact
Results are applicable only to
a healing defect
Model 2:
full-thickness
defect
17–20 Evaluate strategies
that promote bridging
of alveolar bone in a
critical-size defect
Alveolar bone,
tooth
A simple procedure with low
risk—tooth crowns are intact
Applicable to a limited
number of clinical scenarios
in which there is neither
mucosal damage nor loss of
tooth crown
Model 3:
full-thickness
defect with
intraoral
communication
17–22 Evaluate therapies for
tissue regeneration
in a critical-size
defect with bacterial
contamination from
saliva and teeth
Alveolar bone,
tooth, oral
mucosa
Simulates clean-contaminated
bacterial inoculations from
saliva and tooth plaque (e.g.,
resections due to benign or
malignant neoplasms in the
operating room)
Higher risk of postoperative
complications due to loss of
tooth crown that may briefly
disrupt mastication; risk of
mandibular infection if
no antibiotics are given
Model 4:
full-thickness
defect with
intraoral
communication
and pathogen
inoculation
17–23 Evaluate therapies for
(i) infection prevention
and/or treatment and
(ii) tissue regeneration
in a critical-size
defect with bacterial
contamination from
saliva and teeth
Alveolar bone,
tooth, oral
mucosa,
infection
Simulates traumatic bacterial
inoculations from saliva, tooth
plaque and exogenous bacteria
(e.g., motor vehicle accidents,
gunshot wounds)
Requires additional
protocols and precautions
for use of bacteria; may
lead to increased rabbit
morbidity/mortality

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
1992 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
interaction in order to design biomaterials that maximize a desired
interaction, such as the adhesion of cells to biomaterial scaffolds10.
Furthermore, mass transport and fluid flow through scaffolds can
be modeled computationally in order to provide insight into the
suitability of scaffolds for supporting cell survival and growth40.
However, computational modeling cannot account for the wide
variety of cell types and signaling pathways present in the in vivo
environment—for example, incorporation of ligand–receptor
interactions and cytoskeletal components that are essential in cell
behavior and adhesion. These models are therefore most use-
ful for early design of scaffolds and to understand phenomena
observed in vitro or in vivo.
In vitro experimentation is an essential step in the process of
biomaterial scaffold development. These experiments usually
involve single- or multiple-cell cultures with the material in
order to evaluate toxicity to cells and the ability of cells to survive
and proliferate on or within the material9. More sophisticated
techniques for culturing cells in vitro, such as the use of flow
perfusion bioreactors, can control other variables, particularly
shear stress and mass transport, that affect cells11. Furthermore,
important data regarding the ability of the material to drive cell
behavior such as stem cell differentiation can be evaluated in a
setting in which variables such as mechanical loading and growth
factor/cytokine delivery can be controlled9. The value of in vitro
experimentation lies in the ability to evaluate specific questions
by strictly controlling the environment to which cells are exposed.
Such experiments, similar to computational modeling, help to
refine the design of biomaterials and to better understand or
explain phenomena seen in the complex in vivo environment.
In vivo experimentation is an indispensable part of translating
biomaterials into the clinical environment. Although compu-
tational and in vitro methods can relay important information
about specific interactions between cell types and biomaterials,
they cannot replicate the entire signaling repertoire and cellular
response seen in vivo. These three methods work in concert with
each other to optimize scaffold design, elucidate mechanisms of
response and evaluate host tissue response while minimizing the
number of animals used by promoting testing of only the most
promising scaffold materials.
Experimental design
Critical-size defect. The model uses a validated critical-size
defect, which is an important feature if the intent is to show a
statistical benefit for a new regeneration strategy. The model was
validated by evaluating bony bridging by microCT and histol-
ogy at 8 weeks and 16 weeks postoperatively in full- and partial-
thickness defects without intraoral communication to confirm
lack of bridging in full-thickness defects at 16 weeks4. The rabbit
mandible is adequately sized to receive constructs ~10 mm in
diameter and 6 mm in height (470 mm3).
Controls. Positive and negative controls should be included in
the experimental design. For investigations of bone regeneration,
negative controls are usually empty defects, whereas positive
controls are autograft-filled defects. For investigation of bone
regeneration of infected defects, the negative control is a construct
with no antimicrobial treatment (i.e., infected, no antimicrobial
treatment), and positive controls include antibiotic-loaded
bone cement, as is used clinically. It is highly recommended
that a pilot infected defect study be performed to optimize
conditions for the creation of a true infection, as a healthy
animal may be able to clear an infection regardless of the treat-
ment provided12,41.
Analysis and evaluation methods. Experimental end points
should be determined by the outcome to be measured. Our labora-
tory has typically evaluated bone regeneration and implant-tissue
response by imaging (microCT) and histology. For experiments
in which bone regeneration is the primary outcome, our labora-
tory has generally performed 12-week studies, although 8- and
16-week studies have also been used. However, evaluation of dif-
ferent outcomes, such as development of osteogenic membranes
or quantification of gene expression within healing tissues, will
require controls and optimization to determine appropriate end
points. If appropriately planned, animal samples can be used
for more than one type of analysis, maximizing the amount of
data collected per animal and allowing fewer animals to be used
overall. Refer to Figure 2 for a flowchart that details how to max-
imize data collection when evaluating either bone regeneration/
molecular states or angiogenesis. Essentially, tissue collection
can be optimized by arranging data collection from least to most
destructive. Observational data are collected before dissecting the
mandible, and live tissue specimens (e.g., for culture, quantitative
PCR or western blot protein analysis) are taken before fixation
(histology). If undisturbed tissue is necessary, live specimens
are taken from a consistent portion of the sample, leaving the
remainder untouched for histologic evaluation. Preparation
of samples for histology/immunohistochemistry and electron
microscopy are terminal end points, after which further manipu-
lation of the samples is not advisable.
Sample size. A power analysis should be performed for
each experiment, but our laboratory has typically found that
sample sizes of 6–10 are acceptable to achieve statistical
significance. Sample sizes should be based on data from pilot
studies relevant to the desired outcome (e.g., regenerated
bone volume), and the analysis is usually based on a one-way
ANOVA. Our laboratory uses the commonly accepted biostatis-
tical parameters for significance level
α
= 0.05 (Type I error),
β
= 0.05–0.2 (Type II error), power (1−
β
) = 80–95% and a
medium to large effect size.
Animal care and attrition. Rabbits typically return to full activ-
ity within 3–7 d after surgery. If no tooth is removed, rabbits
recover earlier and have less difficulty eating than if a tooth is
removed. Providing pellets softened with water improves their
ability to eat earlier. In our experience, supplementing softened
food with Critical Care diet and one tablespoon of yogurt for
the first 4–7 d after surgery reduces postoperative diarrhea.
Anticipated attrition for this protocol is ~10–15%, based on
our previous experience. Deaths are typically due to postopera-
tive diarrhea, and they occur within the first 2 weeks, allowing
adequate opportunity for replacement.
The defect described in this protocol is a critical-size defect
produced through the limitation of size. This should be taken
into account when considering the use of this model to address

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 1993
situations in which underlying pathology is the cause of absent
or poor bone growth.
Animal record form. Detailed animal record forms must be
kept for each animal. Examples of these forms can be found in
Supplementary Notes 1–3 (anesthesia record, postoperative
monitoring form and postoperative recovery form). Apart from
satisfying legal requirements for animal welfare reporting, these
records are essential for noting discrepancies in surgical tech-
nique or incidents during the procedure that may later be used to
evaluate outliers in the study. For the most efficient use of these
forms, veterinary personnel who are maintaining anesthesia
should be responsible for recording vitals and major procedural
milestones and times on the forms. These records should be kept
with the animals or within the vivarium at all times. The records
should include the following:
Date
Animal identification information: species, sex and identifica-
tion number
Protocol information: principal investigator and approved
animal protocol number
Personnel information: names of surgeons, assistants,
technicians and anesthetists
Procedure information: type of procedure and details of each
step, including all animal preparation, such as intubation,
and all surgical procedures
•
•
•
•
•
Preoperative vitals: temperature, pulse, respiratory rate and
weight
Substances administered: concentration and dose of all sub-
stances administered to the animal throughout the procedure,
including drugs and fluids; if biomaterial implants and/or
exogenous bacteria are used in the procedure, these must be
recorded on the anesthesia record form as well
Anesthesia monitoring record: type of anesthesia, anesthetic
percentage, oxygen flow rate and vital signs every 15 min
Explanation and time of events: e.g., intubation and extuba-
tion, procedure start and end time, beginning or ending of
ventilation, medication administration, surgical events
Procedure outcome: recovered, died or euthanized
Recovery care: record vital signs (heart rate, respiratory rate,
pulse oximetry and temperature) and observations (e.g.,
monitor the incision site for bleeding) every 15 min until the
rabbit is sternal and maintains stable vitals
Postoperative record: monitor and record general health status,
incision site healing, fecal and urine output, analgesia and food
intake twice daily until postoperative analgesia is discontinued,
and then once daily for the remainder of the study
These forms ensure that the animal is kept in good condition,
from a health and animal welfare perspective, during the proce-
dure. In addition to these forms, a simplified form detailing the
personnel, study codes and major procedural milestones may be
recorded by the nonsterile assistant in order to ensure that the
•
•
•
•
•
•
•
Mandibular dissection
(Step 32B)
Angiogenesis
• SEM
• TEM
• Plain radiography
• MicroCT
• qPCR
• Western blot • TEM
No Yes
Undisturbed sample
Euthanasia
(Step 30)
Bacterial swab
(Step 30)
Bone regeneration and/or
molecular analysis
Type of
analysis
Neck dissection/Microfil infusion
(Step 32A)
Mandibular dissection
(Step 32B)
Gross tissue evaluation
(Step 33)
Fixation in Formical-2000
(Step 36C)
• Plain radiography
• MicroCT
• SEM
• Methacrylate-
embedded histology
Light Microscopy
(Step 36D)
Gross tissue evaluation
(Step 33)
Fixation in 10% (vol/vol)
NBF (Step 35)
Molecular
analysis?
Tissue sampling
(Step 33A,B)
• Paraffin-embedded
histology
Electron microscopy
(Step 33C)
X-ray analysis
(Step 36A,B)
Molecular analysis
(Step 33A,B)
Electron microscopy
(Step 33C)
X-ray analysis
(Step 36A,C)
Light microscopy
(Step 36D)
Figure 2 | Flowchart for sample preparation. Tissue collection can be maximized and total animals used can be minimized if samples are prepared in the
correct order. In general, live cells, such as bacterial culture, must be collected first, and molecules that can degrade, such as RNA and protein, must be
stabilized as soon as possible. Major steps are indicated by rectangles, and decision points are denoted by diamonds. Boxes outlined by thick black lines
indicate destructive end points and gray boxes indicate choice of primary outcome measurement.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
1994 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
experimental design is being followed, especially in experiments
in which there are multiple groups undergoing procedures in a
randomized manner. An example of this form can also be found
in Supplementary Note 4 (surgical record).
Pathogen considerations. The decision to introduce pathogens
into this protocol requires additional considerations, both regula-
tory and procedural. First, the use of pathogens usually requires
an additional biosafety protocol to be submitted. These proto-
cols should be submitted to any institution in which the bacteria
will be grown or used, and a materials transfer agreement should
be in place if bacteria will be transferred between institutions.
For bacteria that we have used in our studies, we obtained
approval from the institutional biosafety committees at both
Rice University and University of Texas Health Science Center at
Houston. A biosafety protocol should specify the following:
The biological agent
A description of the bacteria, including general background
information, pathogenicity, modes of transmission and whether
it can affect humans
A description of the inoculation procedure and surgical model,
including timing of inoculum preparation, amount to be in-
oculated, expected outcomes and plans for pathogen disposal
at the end of the procedure
Housing requirement for infected animals
Experimental groups
Study personnel and relevant animal and biological agent
training for each person
Transportation and containment of the biological agent
Precautions, including personal protective equipment, engineering
controls and spill procedures
Point of contact for the study
Bacterial pathogens can be obtained either through the American
Tissue Culture Collection (ATCC) or from a clinical collaborator
who can provide patient-derived samples. In this rabbit man-
dibular defect model, our laboratory has used a clinical strain
of Acinetobacter baumannii33 obtained from collaborators and
a strain of Prevotella melaninogenica34 obtained from the ATCC
(ATCC 25845) as the infecting agent. Carefully consider the
biosafety level of the pathogen and whether appropriate safety-
certified areas are available for handling and culture of the
pathogen, both in the laboratory and at the site of the surgeries.
We recommend a thorough search of the literature to identify
whether models with controls using the specific desired pathogen
already exist. The veterinarian should be consulted for advice on
systemic antibiotics that would be used against the chosen bacte-
rial species if needed, and antibiotic susceptibility testing should
be performed using control strains and methods as indicated in
a standard such as ISO 20776. Bacteria must be evaluated for
survival under transport and storage conditions before surgery.
This can be done by creating a growth curve for the bacteria under
the planned transport and storage conditions. If procedures on
multiple animals are planned for the day, evaluate survival and
growth over the entire time period the bacteria will be on ice
before inoculation (i.e., at the time of each inoculation).
For aerobic bacteria, we have typically suspended bacterial
colonies in sterile normal saline and stored them on ice before
•
•
•
•
•
•
•
•
•
implantation. The inocula are stored and transported on ice in
sealed secondary containment. For anaerobic bacteria, colonies
can be suspended to the appropriate OD in sterile normal saline
in an anaerobic chamber and stored on ice. Primary and second-
ary containers are equilibrated within the chamber to achieve the
appropriate gas mixture and sealed in order to prevent prema-
ture exposure of the pathogen to oxygen. A reducing agent such
as Oxyrase for Broth or thioglycolate may be used to scavenge
oxygen, if appropriately tested with the bacteria. In particular,
for anaerobic cultures, we recommend making separate inoculum
tubes for each procedure planned for the day, such that each tube
is only opened once and then discarded. We also recommend
plating the inoculum, aerobic or anaerobic, at the end of the
procedures for the day to ensure viability of the inoculum.
The researcher must also decide how the inoculum will be
placed within the defect. A pipette or a syringe may be used to
inject a liquid culture into the defect. This is a simple method
that can be performed quickly and does not require additional
materials, but a liquid culture is difficult to contain in a single
area. Another possibility is to use a carrier, such as a collagen
sponge, for the bacteria. The bacterial inoculum is pipetted onto
the carrier and the carrier is placed within the defect. This has
the advantage of ensuring that the inoculum stays in the gen-
eral area of implantation, but introduces another material into
the defect that may confound evaluation of tissue response to an
implant. Alternatively, the implant itself may be inoculated with
the bacteria in the laboratory and transported on ice in second-
ary containment to the operating room. This would be particu-
larly useful for applications such as implant-associated biofilm
infections, as materials may need to be immersed in an inocu-
lum for several hours in order to form a biofilm. This option
would still require rigorous presurgical evaluation to ensure
that a reproducible inoculum could be reliably produced on the
material. It is also important to consider whether the inoculum is
placed before or after the implant. The inoculum may be placed
before the implant if the researcher desires to allow the bacteria
to encounter all of the tissues before placement of a drug-eluting
device. Placement of the implant could occur before inoculation
to sequester bacteria within the bone.
Finally, it is important to determine, in consultation with
the veterinarian, a prospective humane end point of the study.
Although the inoculum concentration should ideally be titrated
to produce an infection with minimal morbidity, a substantial
infection that causes pain and distress is possible. In this case,
the researcher must decide before beginning the study whether
to provide systemic antibiotics or to exclude the animal from the
study and euthanize it. This decision should be guided by the
objectives of the study, in consultation with a veterinarian.
Perioperative antibiotics. Administration of perioperative anti-
biotics depends on the design of the experiment. For local anti-
biotic delivery applications in infected mandibular defects, our
laboratory has typically declined to use perioperative antibiotics
in order to minimize the confounding factor of systemic antibiotic
administration on bacterial clearance and local tissue regenera-
tion. However, as clinical scenarios would involve the use of sys-
temic antibiotics in combination with local antibiotics, it is also
reasonable to use perioperative antibiotics to evaluate local drug
release in a more clinically relevant situation. For nonantibiotic

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 1995
delivery applications, the decision to use perioperative antibiotics
should be made in consultation with the veterinarian.
Implant. Our laboratory has typically used this defect model to
evaluate host tissue response to either new biomaterials30,35 or
new formulations of biomaterials29,31,33. Polypropylene fumarate
was synthesized and fabricated into implants within our labora-
tory, whereas polymethyl methacrylate was obtained from a com-
mercial source. Any synthesized material can be used, provided
that the synthesis is reproducible and the polymer has been shown
to be reasonably nontoxic in vitro. Our laboratory has typically
used hard polymer materials, such as polymethyl methacrylate29
and polypropylene fumarate30, for these defects. However, we have
also developed injectable hydrogel systems that would be suitable
candidates for evaluation in the mandibular defect9,42. Implants
can be either prefabricated or fabricated at the time of implanta-
tion, depending on the intended clinical use of the material.
Fixation plate. The fixation plate is probably not necessary in
partial-thickness defects and in full-thickness defects without
intraoral communication. However, our laboratory has always
used the fixation plate as a prophylactic measure against iatro-
genic fracture in all models. In particular, animals with full-
thickness defects with intraoral communication should always
be plated, as only a small strip of bone on the inferior mandible
separates this defect from a segmental defect. Iatrogenic fracture
in the absence of a fixation plate would result in euthanasia of the
animal. A disadvantage of using a fixation plate in the application
of bone regeneration is that the plate can act as an osteoconduc-
tive material and may facilitate bony bridging across the defect.
TABLE 2 | Key decisions necessary before submission of animal protocol.
Decision Options Advantages Disadvantages
Potential clinical
applications
Population
type
Healthy Does not require additional
procedures
Does not take into account
any underlying pathologies
Trauma in otherwise healthy
patients
Compromised
(i.e., ovariectomized,
irradiated)
Introduces challenges
specific to clinical
applications
Requires additional
justification, procedures and
equipment
Ovariectomy: osteoporosis;
irradiation: regeneration
of irradiated bone after
malignant neoplasm resection
Infection Clean-contaminated Does not require additional
biosafety protocol
Unpredictable with regard to
clinical infection development,
which must be considered when
requesting additional animals
for attrition
Resection of benign
pathologies (i.e.,
odontogenic cyst)
Exogenous
bacteria
Stringent model for
evaluating antimicrobial
strategies
Requires additional biosafety
protocol and safe handling
procedures
Traumatic inoculation
(i.e., shrapnel, gunshot)
Perioperative
systemic
antibiotics
Administer systemic
antibiotics
Replicates clinical
prophylaxis against
development of infection in
clean-contaminated
resections
Antibiotics may affect tissue
regeneration
Open mandibular fractures
or defects with tooth
involvement
No systemic
antibiotics
Can evaluate efficacy of
antimicrobial strategies
and materials; no exogenous
drug effects on tissue
regeneration
Clean-contaminated:
undesired infection may occur;
exogenous pathogen: protocol
optimization to titrate
inoculation dose is required
Closed fractures
Fixation
hardware
Fixation plate Minimizes risk of euthanasia
due to iatrogenic fracture;
may assist in retaining
implant within the defect
May confound bridging results
because of osteoconduction;
additional equipment cost
Mandibular fractures and
implants are usually fixed
with hardware
No fixation plate Does not allow conduction of
bone across defect
Potential iatrogenic fracture,
especially in Models 3 and 4,
resulting in mandibular
instability
Note: some committees
will require the use of
fixation hardware in Models 3
and 4 because of the
potential for fracture in the
inferior mandible

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
1996 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
However, an advantage of using the fixation plate is that clinical
applications would require the use of a fixation device, and thus
the model replicates the clinical situation. We highly recommend
the use of a fixation plate, particularly in full-thickness defects
with intraoral communication. The final decision on the use of
fixation should be made in consultation with the veterinarian and
the institutional animal care and use committee.
Population type. Our laboratory has typically used this model
to replicate the specific clinical population of young males who
receive large, traumatic composite tissue defects. Thus, the rab-
bits used in our studies have no underlying pathology, and limits
to regeneration exist because of size. There are many underlying
pathologies that contribute to delayed or poor bone regeneration,
particularly irradiation38,39 and osteoporosis37. Rabbits can be
given radiation treatment to produce a model for the evaluation
of tissue regeneration in an irradiated area, whereas osteoporosis
can be modeled by using ovariectomized female rabbits. Although
the protocols for producing these underlying pathologies are
outside the scope of this protocol, we mention these other popula-
tions here to emphasize that regenerative capacity can vary and
that this mandibular defect protocol may be used to evaluate
regeneration in these different populations.
A summary of the key decisions discussed above is presented
in Table 2.
Required personnel and expertise. Surgical personnel require
basic surgical skills (i.e., dissection and suturing), and we recom-
mend practice on rabbit cadavers before beginning experiments,
if possible. In addition to the primary surgeon, a sterile surgi-
cal assistant is necessary. Rabbit anesthesia requires veterinary
personnel to intubate the rabbits, regulate ventilation and moni-
tor vitals during the surgeries. If possible, licensed veterinary
technicians are preferred. Although not strictly necessary, a sec-
ond nonsterile assistant is helpful, particularly for beginners and
for studies involving the use of bacteria.
MATERIALS
REAGENTS
Skeletally mature rabbits (New Zealand White Rabbits, age: ~6 months, weight:
3.5–4.5 kg, Charles River Laboratory, Oakwood Facility, strain code 571)
! CAUTION Animal protocols must conform to all relevant animal
ethics and welfare regulations and be approved by the appropriate insti-
tutional care and use committees. Protocols should be approved by all
institutions involved in the work. Consult with research administration at
each institution, if multiple institutions are involved, to determine whether
multiple approvals are needed. For the protocol outlined here, all surgical
procedures followed protocols approved by the University of Texas Health
Science Center at Houston Institutional Animal Care and Use Commit-
tee. When necessary, approval was also obtained from the Rice University
Institutional Animal Care and Use Committee and/or the Department
of Defense Animal Care and Use Review Office  CRITICAL Our
laboratory has used male animals for all previous studies; however,
sex should be determined by experimental design and requirements
of the funding agency.
Investigational implant material, sterilized
Bacterial inoculum for infection studies (e.g., clinical isolate from a
hospital33, bacteria obtained from commercial source such as ATCC34)
! CAUTION All institutional regulatory permissions regarding the use
of biohazardous agents must be obtained before use (i.e., institutional
biosafety committee or equivalent) from all institutions where the bacteria
will be grown or used. We received approval from the institutional biosafety
committees at Rice University and the University of Texas Health Science
Center at Houston. Consider biohazard limitations of the laboratory and
the animal facility. Animals inoculated with a pathogen must be housed
in the appropriate biosafety level housing as designated by the relevant
institutional safety committee.
Buprenorphine (Patterson Veterinary, cat. no. 07-891-9756)
Isoflurane (Henry Schein, cat. no. 050033) ! CAUTION Isoflurane is a
halogenated anesthetic that can cause skin, eye and respiratory irritation.
Halogenated anesthetics may be associated with reproductive problems
and developmental defects. Isoflurane has been categorized as a Pregnancy
Class C drug by the United States Food and Drug Administration.
Personnel who are pregnant should avoid exposure to this drug.
A scavenging system should be in place to minimize exposure, and
badges can be worn to monitor cumulative exposure. Isoflurane can
also cause skin, eye and respiratory irritation.
Oxygen, USP grade (Matheson Tri-Gas)
Meloxicam (Patterson Veterinary, cat. no. 07-891-7959)
Critical Care diet (Patterson Veterinary, cat. no. 07-849-4321)
Acepromazine (Patterson Veterinary, cat. no. 07-869-7632)
SomnaSol (Henry Schein, cat. no. 024352)
Bupivicaine (Patterson Veterinary, cat. no. 07-890-4881)
•
•
•
•
•
•
•
•
•
•
•
Ketamine (Patterson Veterinary, cat. no. 07-803-6637)
Sterile normal saline (Henry Schein, cat. no. 031005)
Lactated Ringer’s Solution (Patterson Veterinary, cat. no. 07-883-6643)
Chlorhexidine scrub (VEDCO, cat. no. 50989-048-29) and solution
(VEDCO, cat. no. 50989-351-29)
Formalin (Fisher Scientific, cat. no. SF100) ! CAUTION Formalin can cause
eye, skin and respiratory irritation, and it is a probable carcinogen.
This chemical should be used and stored only in chemical fume hoods.
Formical-2000 (American MasterTech, cat. no. DCF20)
Heparin (Henry Schein, cat. no. 1162402)
Microfil and curing agent (Flow Tech, cat. no. MV-122)
Ethanol, 70% (vol/vol) in water (VWR, cat. no. 89125-180)
Parafilm (Bemis, cat. no. PM992)
Ethylene oxide (Andersen Products, cat. no. AN73) ! CAUTION Ethylene
oxide is a toxic gas, and it should be handled with care. The liquefied
gas is extremely flammable and explosive when exposed to elevated
temperatures. Ethylene oxide causes respiratory tract irritation and skin
and eye irritation/burns.
EQUIPMENT
Electric clippers (Oster)
Scale (Tanita, model no. 1584)
Puralube veterinary ophthalmic ointment (Henry Schein, cat. no. 008897)
Rabbit anesthesia system (Midmark)
Bain rebreathing circuit (Hudson RCI, cat. no. BD2498) and Bain circuit
adapter (Dre Veterinary Equipment, cat. no. 1487)
Vital signs monitor (Surgivet, model no. V9204)
Warming blanket (Gaymar T/Pump, Model TP 700)
Electrocautery system (Bovie Medical Corporation)
Bovie pencil (Bovie Medical Corporation)
Monopolar Colorado microdissection needle (Stryker Leibinger,
cat. no. N103A)
22-gauge butterfly (Patterson Veterinary, cat. no. 07-857-2978)
Sterile disposable scalpel blades, no. 15 (Miltex, cat. no. 4-115)
Scalpel handle (KLS Martin, cat. no. 10-130-03)
Surgical/dental drill (NSK Surgic XT Plus, cat. no. Y141246)
Contra-angle handpiece (NSK Ti-Max, cat. no. SG20L)
Straight handpiece (NSK Ti-Max, cat. no. SG65L)
Spatula (VWR, cat. no. 82027-530)
Adson-Brown forceps (KLS Martin, cat. no. 12-244-12)
Micro Jewelers forceps (KLS Martin, cat. no. 12-418-04)
Small hemostat (KLS Martin, cat. no. 13-310-12)
Needle drivers (KLS Martin, cat. no. 20-632-14)
Periosteal elevator (Hu-Friedy, cat. no. PFITR1/2)
10-mm Trephine bur (Ace Surgical Supply, cat. no. 04-9487-02)
1.1-mm Drill bit (KLS Martin, cat. no. 25-452-05-91)
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 1997
Cross-cut bur, 1 mm (Stryker Leibinger, cat. no. 277-10-210)
1.5-mm Titanium fixation plate, 20 hole (KLS Martin, cat. no. 25-320-00-91)
1.5-mm Titanium fixation screws, 5 mm (KLS Martin, cat. no. 25-878-05-91)
Screwdriver (KLS Martin, cat. no. 25-402-99-07 and 25-489-97-07)
Plate benders (KLS Martin, cat. no. 50-502-10-07)
Plate cutters (KLS Martin, cat. no. 50-502-11-07)
Angiocatheter, 20 gauge (Venisystems Abbocath-T,
cat. no. G717-A01, 4535-20)
Tubing extension set, 20 inches (Hospira, cat. no. 4429-48)
Male–male Luer lock adapter (Cook Medical, cat. no. G15265)
•
•
•
•
•
•
•
•
•
4-0 Vicryl (Patterson Veterinary, cat. no. 07-808-9962)
Gauze, sterilized (Patterson Veterinary, cat. no. 07-838-4518)
Needle, 25 Gauge (Becton Dickinson, cat. no. 305122)
Syringe, 5 ml (Becton Dickinson, cat. no. 309646)
Syringe, 60 ml (Becton Dickinson, cat. no. 309653)
Personal protective equipment, including eye protection, sterile
gloves (Patterson Veterinary, cat. no. 07-834-7223), disposable gowns
(Patterson Veterinary, cat. no. 07-890-7966), surgical caps (Patterson
Veterinary, cat. no. 07-868-9278) and surgical masks (Patterson Veterinary,
cat. no. 07-844-3672)
•
•
•
•
•
•
PROCEDURE
Preoperative preparation ● TIMING 45 min, excluding acclimation and instrument sterilization
1| Acclimate rabbits to soft food at least 7 d before surgery. A minority of rabbits may require 10–14 d to acclimate to
the diet. Sterilize surgical instruments and gauze in an autoclave, and allow them to cool to room temperature (20–25 °C).
Sterilize implants or instruments that cannot be autoclaved by ethylene oxide treatment and allow to off-gas for at least
24 h in a chemical fume hood before implantation.
2| Remove food from rabbits the morning of surgery.
3| On the morning of surgery, administer 0.02–0.03 mg/kg buprenorphine s.c. to each animal in the scapular region for
analgesia using a 25-gauge needle and appropriately sized syringe. Disinfect the operating table with a solution of
quaternary ammonium or 70% (vol/vol) ethanol.
4| For induction, administer 35–55 mg/kg ketamine and 1.25–1.75 mg/kg acepromazine s.c. for sedation using
25-gauge needles and appropriately sized syringes. Weigh the animal and record the weight, shave the right neck,
the lower back and the right side of the face below the eye. Attach a mask with isoflurane at 3.5% to the animal,
and transfer it to the operating table in the dorsal position. Secure the legs to their respective sides of the table with
soft rope. Place a folded towel underneath the animal’s head on the side receiving surgery in order to elevate the area.
Maintain warmth (37 °C) with a hot water blanket and/or forced air warming unit. Verify the depth of anesthesia
by lack of reflex to toe pinch.
? TROUBLESHOOTING
5| Intubate the rabbit and maintain anesthesia with 2–3% isoflurane in 100% oxygen. Maintain ventilation throughout
the procedure.
 CRITICAL STEP Intubation of rabbits should be performed by proficient veterinary technicians. Improper intubation can
lead to inadequate ventilation, airway swelling, asphyxiation and death.
6| Attach the pulse monitor to either hind foot and attach a three-lead ECG monitor to the back. Apply enough Puralube
to each eye to cover the eye with a thin film of the jelly.
7| Monitor heart rate, oxygen saturation, carbon dioxide and respiratory rate every 5 min throughout the procedure.
Monitor temperature and nonresponse to hind-limb toe-pinch reflex every 10–15 min. Please refer to Table 3 for guidelines
on normal values for these vitals.
? TROUBLESHOOTING
8| Place a 22-gauge intravenous catheter in the ear vein
and provide Lactated Ringer’s Solution at a drip rate of
10 ml/kg/h throughout the procedure.
9| Prepare the incision site with alcohol and chlorhexidine.
Place a sterile drape over the animal.
10| Put on personal protective equipment, to include cap,
shoe covers, mask, eye protection, sterile gown and sterile
gloves. If performing multiple surgeries in a single session,
change the sterile gown and gloves between animals.
TABLE 3 | Guide for monitoring vital signs under the written
anesthetic protocol.
Vital sign Expected range
Heart rate 210–300 Beats per min
Oxygen saturation 95–100%; >90% is acceptable
Carbon dioxide 35–45 mmHg
Ventilation respiratory rate 12–18 Breaths per min
Temperature 98–100 °F under anesthesia;
101–103°F awake

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
1998 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
Operative procedure ● TIMING 90–120 min
11| With the animal in the dorsal position, cut a window in the sterile drape to expose the inferior mandible. Use a
scalpel with a no. 15 blade to make a midline incision from the mentum posteriorly to the midpoint between the left and
right angles of the mandible through the skin, as shown in Figure 3a, stopping when the fascia is exposed.
12| While the surgical assistant provides retraction, use a Bovie cautery with a monopolar Colorado microdissection needle
to continue dissecting through the fascia until the inferior border of the mandible is visible. Small vessels may be cauterized
with the Bovie. Note the use of electrocautery in the animal record form.
 CRITICAL STEP Be aware of the large facial artery running along the inferior border of the mandible. A representative
image showing the facial artery during dissection can be seen in Figure 3b.
? TROUBLESHOOTING
13| Locate the inferior border of the mandible, and use a scalpel to make an incision through the visible periosteum
between the attachment of two muscles on the inferior border, as shown in Figure 3c.
14| Using a periosteal elevator, expose the buccal cortical plate in the area of the premolars through the incision in
the periosteum by carefully lifting away muscle and periosteum. Start anteriorly and move posteriorly.
15| Manually retract soft tissue away from the bone, and observe the superior aspect of the mandible. As shown in
Figure 3d, the sloped curve toward the incisor anteriorly from the body of the mandible should be visible; this is a
useful landmark for creation of the defect.
16| Approximate the margins of the defect using the trephine bur. First, estimate the location of the second tooth
(each tooth is ~2–3 mm wide). Center the trephine bur in the anterior–posterior axis on the second molar from the
incisor. Leave 2 mm between the bur and the inferior aspect of the mandible, if possible. Check to ensure that the bur
does not cut through the superior aspect of the mandible by centering the bur in the intended area and retracting the
soft tissue to visualize the superior aspect mandible.
 CRITICAL STEP The surgical assistant should retract the soft tissue anteriorly at the level of the incisor and posteriorly at
the buccinator muscle.
 CRITICAL STEP Leaving adequate bone (~ 2 mm) between the inferior mandible and the defect is essential to preventing
iatrogenic fracture of the mandible.
? TROUBLESHOOTING
17| Use a 10-mm trephine bur on the contra-angle 1:20 handpiece at 2,000 r.p.m. under constant irrigation with saline to
begin making the mandibular defect (Figure 4). Lightly score the buccal cortical plate by applying the trephine bur firmly to
the desired area for ~1 s, and then verify the margins of the defect, as shown in Figure 4a. Continue drilling under irrigation
through the buccal cortex until the bur hits the teeth. The surgeon will feel the trephine break through the cortical plate
and feel a change in the resistance of the drill, as the teeth are harder than the mandibular bone. Check that all sides of
the plate are completely free by using a small periosteal elevator to gently lift the plate away from the mandible, as shown
a
m
la
ra
b c d
Figure 3 | Dissection of the mandible. Note that the superior and inferior aspects of the mandible are located at the bottom and top of the image,
respectively, as the rabbit is in the dorsal position. (a) A window is cut into the drape to reveal the inferior mandible. An incision (dashed line) is made from
the mentum (m) to the midpoint between left and right mandibular angles (la and ra, respectively). (b) During dissection through the fascia and muscle,
it is important to avoid inadvertently damaging the facial artery (black arrow). (c) The periosteum is divided on the inferior mandible along the white line
indicated by the blue arrow. (d) The buccal (lateral) cortex of the mandible is exposed, with the white arrowhead indicating the diastema for reference.
All surgical procedures followed protocols approved by the University of Texas Health Science Center at Houston Institutional Animal Care and Use Committee.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 1999
in Figure 4b. If substantial resistance is encountered, drill through the area that is not free. Lift the buccal cortex off the
mandible. This bone and all other biological materials removed during this procedure should be disposed of as biohazardous
waste. If desired, this bone can also be morcellized for use as an autograft.
 CRITICAL STEP Score the mandible without penetrating too deeply, in case adjustment needs to be made to the defect.
Always lift the bone away from the defect from the anterior or posterior sides of the defect, as pressure on the superior or
inferior aspect of the defect can cause iatrogenic fracture.
? TROUBLESHOOTING
18| Irrigate with 10 ml of normal saline to remove bone fragments. One full tooth root surrounded by two half tooth
roots should be visible, as shown in Figure 4c. Continue drilling through the roots with the trephine bur under continuous
irrigation. Completely cut through the teeth before lifting them out with periosteal elevators. Free the teeth from the
lingual cortex with periosteal elevators, and lift the teeth out gently. Only lift teeth from the anterior or posterior sides of
the defect. Irrigate with 10 ml of normal saline to remove tooth remnants.
? TROUBLESHOOTING
19| If a fixation plate is to be used, measure and cut the fixation plate to span the defect with at least two holes on either
side, as shown in Figure 5. Bend the plate to match the contours of the mandible, drill a pilot hole and apply a single screw
on the posterior side of the defect. Swing the plate out of the way, as shown in Figure 5a. If a full-thickness defect, with
or without intraoral communication (Models 2–4, see Table 1), is desired, proceed to Step 20. If a partial-thickness healing
defect (Model 1, see Table 1) is desired, proceed to Step 24.
20| Drill through the lingual aspect of the plate under irrigation, as shown in Figure 4d, stopping every few seconds to
assess the evenness of the cut. During the entirety of the drilling process, ensure that the bur does not catch on the exposed
facial artery, as seen in Figure 4e. Check the depth of the cut frequently, using a periosteal elevator to assess the looseness
of the lingual cortical plate. When the plate is almost free, a periosteal elevator can be used to gently separate the lingual
plate from the surrounding bone. Gently tilt the circular piece of bone with forceps, and carefully dissect and remove the
lingual periosteum and soft tissue from the plate with the flat edge of a periosteal elevator. Irrigate the defect well with 15 ml
of saline, and apply pressure with gauze until the bleeding stops. If a full-thickness defect with an intraoral communication
(Models 3 and 4) is desired, proceed to Step 21. If a full-thickness defect without an intraoral communication (Model 2) is
desired, proceed to Step 24.
 CRITICAL STEP Stop frequently to assess the lingual plate, gently loosening it with the periosteal elevator. Do not drill all
the way through into the soft tissue on the lingual aspect, as the lingual nerve may be damaged, leading to autophagia.
? TROUBLESHOOTING
21| Locate the insertion point into the superior mandible of the tooth whose full root was removed, usually the central
tooth. Use the straight handpiece and a 1-mm cross-cut bur at 40,000 r.p.m. under irrigation to drill straight downward
toward the superior aspect of the mandible, using the margins of the tooth as a guide, as shown in Figure 6a.
 CRITICAL STEP Take care not to snag the soft tissue in the bur.
22| Use a pair of jeweler forceps to remove the tooth, creating an intraoral communication, as seen in Figure 6b. Irrigate
well with 15 ml of saline.
***
a b c d e
Figure 4 | Creation of the mandibular defect. (a) A 10-mm trephine bur is used to score the mandible, being careful to drill evenly through the cortex. The white
arrow indicates an area of shallower cutting as compared with the area indicated by the black arrow. (b) The buccal cortex is lifted away from the teeth with a
periosteal elevator. (c) The exposed tooth roots of one full tooth and two half teeth (*) are drilled through. (d) After the teeth are removed, the lingual (medial)
cortex is revealed and drilled through. (e) The facial artery (black arrow) is exposed throughout this process, and care should be taken not to damage it with the
bur. All surgical procedures followed protocols approved by the University of Texas Health Science Center at Houston Institutional Animal Care and Use Committee.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
2000 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
23| Use a sterile syringe to draw up the appropriate volume, and inject it into the area of the defect. If appropriate, the
inoculum can be loaded onto a carrier, such as a collagen sponge, in order to minimize leakage of the inoculum into other areas.
 CRITICAL STEP The inoculum should be kept apart from the sterile instruments, preferably on another sterile table, in
order to prevent contamination of the instruments.
24| Place the implant within the defect. If infection without an intraoral communication is desired (Model 4), the inoculum
can be placed either before or after implantation of material within the defect. If a fixation plate is being used, rotate the
plate into the desired orientation. Drill pilot holes and apply the remaining screws, one on the anterior side of the defect
and then one more on each side to maintain fixation, for a total of four screws. The plate can be fixed along the inferior
border, as seen in Figure 5b, or across the defect if needed to hold an implant in place.
If no fixation plate is used, proceed to Step 25.
25| Close the muscle and fascia over the defect using running stitches of 4–0 Vicryl suture, taking care to leave as little
dead space as possible. Dead space results when tissues are not properly closed, leaving a cavity into which blood or serum
can accumulate and infection can take hold. Irrigate the closed muscle with 20 ml of saline, and wipe it vigorously with
gauze. Close the skin with subcuticular stitches of 4–0 Vicryl.
 CRITICAL STEP If incision site infection is a concern, a monofilament suture such as 4–0 Monocryl (degradable) or
4–0 Prolene (non-degradable) should be used to close the skin. Monofilament suture should be used in defects inoculated
with exogenous pathogens. We recommend using monofilament suture in defects with intraoral communication, although
this is not strictly necessary if the wound is stringently irrigated before closing. Sterile gloves should be changed before
beginning to suture.
? TROUBLESHOOTING
Postoperative care ● TIMING 2 h
26| Clean the incision with warm saline or room-temperature 3% (vol/vol) hydrogen peroxide to remove blood.
27| Administer 0.5 mg/kg of 0.25% bupivicaine intradermally along the incision site with a 25-gauge needle and
appropriately sized syringe. Trained veterinary personnel should wean the rabbit from the ventilator and administer
0.5 mg/kg of Meloxicam s.c. 30–45 min before expected recovery (~1.5–2 h after weaning).
28| Maintain temperature support and physiologic monitoring. Check the heart rate, respiratory rate and pulse oximetry
every 5 min until extubation. Extubation should be performed only by trained veterinary personnel. Monitor the incision site
for bleeding, as well as bleeding within the mouth. After extubation, monitor every 15 min until the animal is sternal.
29| When animals can maintain sternal recumbency, return them to housing. Commence a postoperative analgesia regimen
by administering 0.02–0.03 mg/kg buprenorphine s.c. every 12 h for 48 h, and after 48 h administer buprenorphine as
a b
Inferior Inferior
Incisor
Incisor
SuperiorSuperior
Angle
Angle
Figure 5 | Placement of fixation hardware. Note that the superior and
inferior aspects of the mandible are located at the bottom and top of the
image, respectively, as the rabbit is in the dorsal position. The incisor and
the angle are located on the left and right of the image, respectively, as
labeled in the image. (a) The fixation plate is screwed into the mandible on
one side and then rotated out of the defect area. (b) After implantation of a
polypropylene fumarate scaffold, the fixation plate is screwed into place
with two screws on each side of the defect, either at the inferior mandible,
as shown here, or spanning the defect. All surgical procedures followed
protocols approved by the University of Texas Health Science Center at
Houston Institutional Animal Care and Use Committee.
a b
Figure 6 | Creation of the intraoral communication. (a) After the removal of
the lingual cortex, the three crowns in the superior mandible are observed
(black box) and the central tooth crown is removed. (b) The final 10-mm
bicortical defect with intraoral communication. All surgical procedures
followed protocols approved by the University of Texas Health Science Center
at Houston Institutional Animal Care and Use Committee.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 2001
needed, determined by observation of distress (e.g., vocalization, abnormal posture, decreased appetite, decreased activity,
poor grooming); administer 0.5 mg/kg Meloxicam s.c. once daily for 4 d, and then as needed. Rabbits should be fed regular
rabbit chow soaked in water and mixed with applesauce.
? TROUBLESHOOTING
Euthanasia and implant/bone collection ● TIMING 1–1.5 h
30| At 12 weeks or at another desired time point, sedate the rabbit with 35–55 mg/kg ketamine and 1.25–1.75 mg/kg
acepromazine s.c. Using sterile swabs, take bacterial cultures from the oral cavity and/or defect area, if required.
31| Euthanize the animal with 1 ml (390 mg) of pentobarbital solution (Somnasol) given i.v. through the ear vein.
Verify the absence of cardiovascular function.
32| Perform a neck dissection (option A) if Microfil perfusion will be used to assess vasculature. It is possible to also
perform a mandibular dissection (option B) sequentially if the implant/bone will be used for gross tissue evaluation,
histology or microCT evaluation of bone regeneration.
(A) Neck dissection
 CRITICAL Modifications to this procedure may be necessary if additional analysis, such as histology, will be performed.
Please see the publication by Sarhaddi et al.43 for more information on how to perform Microfil perfusion and histology on
the same samples.
(i) Shave the lower jaw and neck of the rabbit after sedation (Step 30) and before administration of the euthanizing
agent (Step 31).
(ii) Immediately after euthanasia, use a no. 15 blade to make an incision through the epidermis of the neck, starting
posterior to the original midline incision and extending to the sternum, working quickly to avoid blood clotting within
the vasculature.
(iii) Use the blade to dissect through the fascia and paratracheal musculature until the trachea is reached.
(iv) Using blunt dissection through the sternohyoid strap muscles, locate the right and left neurovascular bundles
containing the external jugular vein, carotid artery and vagus nerve, as shown in Figure 7a.
(v) Pass a 2–0 silk suture around each carotid artery without tying, in anticipation of securing a catheter.
(vi) Cannulate each carotid artery with a 20-gauge angiocatheter, and thread the catheter into place. Tie the 2–0 silk
suture around the cannula to secure the catheter, as shown in Figure 7b,c.
(vii) Locate the external jugular veins and make incisions in them bilaterally, as shown in Figure 7d.
(viii) Assemble a 3-way stopcock, male–male Luer-lock adapter and two sets of extension tubing, as shown in Figure 7e,
without the cannula attached. Flush the lines with heparinized saline (100 U/ml in a 60-ml syringe).
(ix) Connect the two sets of extension tubing to the two cannulas, and push the heparinized saline at a rate of 1 ml/min.
A constant stream of blood should exit the external jugular veins and start to turn pale when ~30 ml of volume has
perfused through the vasculature. Flush 60 ml total of heparinized saline through the vasculature.
(x) Flush the vasculature with 190 ml of normal saline to remove as much blood from the vessels as possible.
(xi) Prepare 50 ml of Microfil MV-122 yellow silicone with 5% curing agent and inject it into the vasculature at a rate of
1 ml/s, as shown in Figure 7f. Do not dilute the Microfil with the diluent. The gingiva, iris and ears of the rabbit
should begin to appear yellow after ~20 ml has perfused, and Microfil should exit the external jugular veins when
~30 ml has perfused. Perfuse the entire 50 ml.
(xii) Carefully wrap the rabbit in a bag and allow the Microfil to cure inside the vasculature. Either place the rabbit in the
refrigerator (4 °C) overnight or wait for 90 min at room temperature.
 PAUSE POINT Rabbits can be left at 4 °C for up to 24 h. Continue to retrieve the implant as described under op-
tion B ‘Mandibular dissection.’ If the rabbit is refrigerated, keep it at room temperature for 1 h before collecting the
implant.
(B) Mandibular dissection
(i) Place the rabbit in the dorsal position on a clean table. Shave the neck and the face below the eye. Locate by
palpation the inferior border of the mandible and the angle of the mandible.
(ii) Make an incision from the angle of the mouth toward the ramus of the mandible, taking care to stay above
the mandible, ideally at the level of the top teeth. Cut through the skin and observe the fascia, clamping medium
or large visible vessels before continuing to cut through fascia and muscle until reaching the inside of the
mouth. On the posterior side, make a horizontal cut through the large buccinator muscle to reach the ramus
of the mandible.
(iii) Make a midline incision from between the front incisors, extending posteriorly to at least the midpoint between
the mandibles. Cut straight down through into the mouth.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
2002 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
(iv) Lift the skin by the incisor and remove the skin by
passing the scalpel between the skin and fascia.
Continue to remove the skin until the entire hemi-
mandible is exposed. Starting at the most posterior
portion of the midline incision, cut around the
angle of the mandible to loosen as many muscle
attachments as possible, being careful not to disturb
the implant site.
(v) Use shears to cut in between the incisors, resulting
in two separated hemi-mandibles. Lift the right
mandible slightly and cut across the ramus of the
mandible with shears, completely severing the
bony connection of the lower mandible to the rest
of the face.
(vi) Remove the remainder of the soft-tissue connections
using surgical scissors. Cut along the midline, freeing
the tongue from its connection to the soft tissue on
the mandible. Gently lift the mandible up and sever
any remaining connections, being careful not to cut
into the defect area.
Tissue sampling for molecular analysis and/or microscopy
and preparation for radiography and/or histology
● TIMING ~72 h–2 weeks
 CRITICAL It is important to determine how the
animal tissue will be evaluated before collection in order to
maximize the use of tissues and to minimize the number of
animals used. Several evaluations can be performed on the
same sample if the sample is processed in the correct order.
Refer to Figure 2 for a visual representation of how samples
should be prepared to maximize data collection.
33| Immediately after collecting the sample, assess gross
characteristics, such as soft-tissue healing or dehiscence
under a fume hood or in a biological safety cabinet, as seen
in Figure 8. Note observations in the animal record form.
Immediately retrieve samples for molecular analysis such as
quantitative PCR (qPCR, option A) and/or protein analysis
(western blot, option B) or electron microscopy (option C).
If histologic evaluation is also desired, be careful to leave
a portion of the sample undisturbed so as not to distort the
normal tissue architecture.
 CRITICAL STEP To minimize degradation, samples for
molecular analysis should be taken as soon as possible and
put into protective storage immediately. It is recommended
to practice collecting samples on cadavers. With experience,
the mandibular collection procedure can be performed in
less than 10 min after euthanasia.
(A) qPCR for gene expression
(i) Store tissues for qPCR in an RNA-stabilizing reagent
such as RNAlater (Qiagen). Extract mRNA, create a cDNA library and perform qPCR according to the manufacturer’s
instructions.
 PAUSE POINT RNA can be preserved in RNAlater for 7 d at 18–25 °C, 4 weeks at 2–8 °C or several months at
–20 or –80 °C.
(B) Protein analysis
(i) For protein analysis, shock-freeze the tissue samples by immersion in liquid nitrogen, and store them before lysis at
–80 °C. Alternatively, freeze the tissues at –20 or –80 °C in a protein stabilizing reagent such as Allprotect Tissue
n
v
a
t m
n v a
fa
c
c
e
s
c
a b
c d
e f
Figure 7 | Microfil infusion. (a) Fascia (f) and musculature (m) are retracted
to reveal the trachea (t) flanked bilaterally by the left and right neurovascular
bundles (black box), consisting of the vagus nerve (n), jugular vein (v)
and carotid artery (a). Silk suture is looped around the bundle to facilitate
separation of these structures and later to secure a cannula in place.
(b) The artery (a) is lifted gently in order to provide tension on the tissue for
insertion of the angiocatheter. (c) A 20-gauge angiocatheter (c) is
inserted into the carotid artery (black arrow) and secured in place with the
silk suture. (d) A cut is made along the black dotted line in the jugular vein
(v) in order to allow outflow of the heparin and Microfil infusion. The catheter
is not inserted in this image in order to clearly show the artery, vein and
nerve. Note the thick-walled artery (a) medial to the thin-walled vein (v) to
help differentiate between the two structures. The nerve (n) does not have a
lumen. (e) A 60-ml syringe is connected to extension tubing (e) and attached
to the cannula (c) for infusion of heparin and Microfil. Note the assembly
of the syringe to the stopcock (s), followed by attachment to two sets of
extension tubing (e), where fluid then flows out the tubing through the
cannula (c). The outflow is indicated with white arrows. (f) Microfil flowing
through the infusion apparatus in the direction of the white arrows. The two
sets of tubing should be attached to the two cannulas located in the left and
right carotid arteries. Panel c reproduced with permission from ref. 35; the
publisher of this copyrighted material is Mary Ann Liebert, Inc. All surgical
procedures followed protocols approved by the University of Texas Health
Science Center at Houston Institutional Animal Care and Use Committee.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 2003
Reagent (QIAGEN), which will stabilize protein, RNA and DNA. Perform protein extraction with lysis buffer per the
manufacturer’s instructions.
 PAUSE POINT Protein in Allprotect can be preserved for 7 d at 15–25 °C, 12 months at 2–8 °C or archived for
long-term storage at –20 °C or –80 °C.
(C) Electron microscopy analysis
(i) Remove tissue or scaffolding for evaluation by scanning electron microscopy or transmission electron microscopy.
These tissues should be removed carefully, then fixed, dehydrated and processed for microscopy44,45.
34| Cut off the incisor to expose the marrow cavity, and remove as much soft tissue as possible without compromising the
desired tissue. Mark either the anterior or posterior bone with pathology ink to maintain orientation.
35| Place the sample into 10% (vol/vol) neutral-buffered formalin (NBF) at 20 °C for 48–72 h.
! CAUTION Formalin is an eye, skin and respiratory system irritant and probable carcinogen. Use appropriate
personal protective equipment (gloves, eye protection and laboratory coat) and work inside a chemical fume hood in
a well-ventilated area.
 CRITICAL STEP If immunohistochemistry is planned, remove the samples from formalin as soon as they are fixed, as
prolonged formalin fixation can result in decreased epitope recognition46. In general, 3 d is sufficient for fixation if the
marrow cavity is exposed.
 CRITICAL STEP Several analyses can be performed on the same sample if tissues are prepared appropriately. Histologic
analysis is a destructive technique and should be performed after radiography.
 PAUSE POINT Samples can be stored in 70% (vol/vol) ethanol in water for up to 3 months.
Analysis methods
36| Follow option A to perform plain film radiographic evaluation of bone; option B to perform microCT analysis of bone
regeneration; option C to perform microCT analysis of angiogenesis/vasculature; or option D to perform histological analysis.
 CRITICAL STEP Plain film radiography is more suitable for qualitative evaluations. MicroCT can be performed after
radiography if desired, and it is more suitable for quantitative evaluations.
(A) Plain film radiography evaluation of bone ● TIMING ~5 min
(i) Wrap the sample in a towel soaked with 70% (vol/vol) ethanol in water, and lay the sample on a radiography cassette
in the desired orientation. Samples should be oriented in the mount to maximize rotational symmetry. For mandibular
samples, we have obtained the best images by vertically mounting the samples; that is, the proximal end of the
mandible (toward the angle) is embedded in clay on the mount, and the incisor points toward the roof.
(ii) Expose at 70 kV and 3 mA for 30 s in a Faxitron X-ray cabinet. Voltage, amperage and exposure time may have to be
optimized to achieve the appropriate exposure.
(B) MicroCT analysis of bone regeneration ● TIMING 2–10 h
 CRITICAL Scan times range from 2 h with no fixation hardware to 10 h with fixation hardware. If possible, fixation
hardware should be removed from the sample before microCT in order to minimize scattering and scanning time. Carefully
examine all samples before removing hardware, as bone growth over the screws/plate can preclude removal. If hardware
cannot be removed from all samples, do not remove hardware from any samples in order to maintain consistency between scans.
 CRITICAL MicroCT evaluation requires that samples be kept under constant conditions while scanning. Wrap samples in
gauze soaked with 70% (vol/vol) ethanol in water and then wrap with Parafilm to create a moisture barrier. The samples
should be wet without dripping.
(i) To quantify regenerated bone volume, use nondecalcified samples (from Step 31), and follow the protocol by
Kallai et al.47 for general guidance on the analysis of bone regeneration in round critical-size bone defects.
(ii) For samples that contain hardware, use the Cu+Al filter48.
(iii) Score each sample according to the scoring guide detailed in Supplementary Table 1.
 CRITICAL STEP Consider the radio-opacity of the implanted biomaterial before performing microCT. Modifications
to the implant or the image acquisition protocol may be required to differentiate bone from implanted biomaterial.
(C) MicroCT analysis of angiogenesis/vascularization ● TIMING 2–10 h
(i) Place the sample in 100 ml of 10% (vol/vol) NBF at 20 °C for 48–72 h.
(ii) Wash the sample under running deionized water for at least 3 min to remove formalin.
(iii) Immerse the sample in Formical-2000 for ~2 weeks, with daily solution replacement. Check for complete decalcification
through planar radiography or microCT, and immerse longer if necessary.
 PAUSE POINT Samples can be stored in 70% (vol/vol) ethanol in water for up to 3 months.
(iv) To quantify the percentage vascularization within a volume of interest, use decalcified samples and proceed with the
methods described by Young et al.35.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
2004 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
(D) Histological analysis ● TIMING ~15 min per slide
(i) Proceed with either plastic-embedded histology for undecalcified samples (Step 35) or paraffin-embedded histology
for decalcified samples (Step 36C). Refer to An and Martin49 for a more complete discussion of considerations for
preparation of bone samples for histology.
(ii) Score histologic sections according to the established criteria shown in Supplementary Table 2 (ref. 50).
? TROUBLESHOOTING
Troubleshooting advice can be found in Table 4.
TABLE 4 | Troubleshooting table.
Step Problem Possible reason Solution
4 The animal is not
ventilating properly
Pressure on the trachea from
the surgeon or assistant
Ensure that there is no pressure being placed on the
trachea from hands or instruments; evaluate and
reposition the head if needed
7 The heart rate of the
rabbit increases
The level of anesthesia is not
deep enough, so the rabbit
experiences pain
Increase the percentage of isoflurane in oxygen
Respiratory rate is
too low
The level of anesthesia is too high,
causing respiratory depression
Reduce the percentage of isoflurane in oxygen.
If necessary, place the animal on a ventilator.
Carefully monitor color in the nose and ears
12 The facial artery is block-
ing further dissection
The incision through the fascia is
not large enough
The facial artery is too large to cauterize, so extending
the incision line caudally frees up the fascia,
allowing the surgeon to move the facial artery
out of the way
Hemorrhaging during
dissection
Facial artery or other vessel was
cut during dissection
Small vessels can be cauterized using the Bovie.
If the facial artery is cut, it should be clamped
with a hemostat and repaired with a figure-eight
stitch can be attempted. Otherwise, ligate the
vessel with suture
16 There is not enough
mandible visible to fit
the trephine bur in the
appropriate area
Further dissection is needed, or
smaller retractors should be used
The buccinator muscle will be encountered as the
dissection proceeds toward the angle of the mandible.
A small amount of blunt dissection with curved hemostats
through the muscle can be performed in order to increase
the surgical field. Angled periosteal elevators or bent
spatulas can be used to retract tissue and conserve space
in the surgical field
17 Trephine bur snags on
muscle or soft tissue
during defect creation
Inadequate exposure or retraction
of muscle and soft tissue
The surgical field can be expanded anteriorly by further
dissecting the periosteum toward the incisor and
posteriorly by blunt dissection of the buccinator muscle.
A surgical assistant is necessary to retract tissue while
the surgeon creates the defect
The buccal cortical plate is
difficult to remove
The mandible is an irregular
surface, making an even cut
difficult
Keep the trephine bur perpendicular to the plane
of the mandible. Assess the depth of the cut regularly,
and precess the bur perpendicular to areas that
are thicker
17–25
18
The rabbit’s temperature
drops during surgery
An entire tooth is removed
during extraction, leaving
an oral communication
The animal is wet from irrigation
run-off or the warming apparatus
is inadequate
Insufficient amount of tooth
structure was cut before extraction
Either use an additional assistant to suction during
irrigation or place gauze to absorb irrigation run-off.
Change the gauze frequently. Check to ensure that the
warming unit is working
Irrigate and verify that the entire tooth has been
cut through before extraction. Do not pry the tooth
aggressively
(continued)

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 2005
● TIMING
Steps 1–10, preoperative procedure: 45 min, excluding acclimation and instrument sterilization
Steps 11–25, operative procedure: 90–120 min
Steps 26–29, postoperative care: 2 h
Steps 30–32, euthanasia and implant/bone collection procedure: 1–1.5 h
Steps 33–35, tissue sampling for molecular analysis and/or microscopy and preparation for radiography and/or histology:
72 h–2 weeks
Step 36A, plain film radiography evaluation of bone: ~5 min
Step 36B, microCT analysis of bone regeneration: 2–10 h
Step 36C, microCT analysis of angiogenesis/vascularization: 2–10 h
Step 36D, histologic analysis: ~15 min per slide
ANTICIPATED RESULTS
The proposed protocol provides a reproducible load-bearing critical-size defect in the rabbit mandible and methods to
quantify bone regeneration and angiogenesis within these defects and to evaluate microscopic tissue response to biomate-
rial implants. Gross observation, microCT and histology are the primary techniques used for analysis, although other ancillary
evaluations can be performed to further probe in vivo responses to drug-releasing implants or infected models. All of the
techniques presented here have previously been performed in our laboratory4,29–31,33–35.
Gross soft-tissue evaluation
Implantation of biomaterials into the mandible can affect the healing of oral mucosa over the defect, and soft-tissue
healing has been evaluated by gross observation in several studies involving mandibular defects with intraoral
communication29–31,33. We have previously evaluated soft-tissue healing in a study involving the controlled release of
the antibiotic colistin from porous implants placed within an infected mandibular defect with intraoral communication33.
Briefly, colistin was delivered via either burst release over 7 d from gelatin-loaded constructs or extended release over 8 weeks
TABLE 4 | Troubleshooting table (continued).
Step Problem Possible reason Solution
18 or 20 The inferior mandible
fractures
Not enough mandible was left
intact during defect creation.
Fracture could also be due to
excessive prying to remove
cortical plates or teeth
Use the fixation plates to approximate the pieces of
the mandible, taking care to use two screws on each
side of the defect
20 The lingual cortical plate
is difficult to remove
The mandible is an irregular
surface, making an even cut
difficult
Carefully inspect the cut area in order to determine
the area that is unevenly cut. Precess the drill to cut
perpendicular to that area
The drill cut all the way
through the lingual
cortex and into the
surrounding soft tissue
An insufficient number of stops
were made to assess the looseness
of the lingual cortex
Stop the drill every few seconds to assess the depth
of the cut. If soft tissue is affected, monitor the animal
for autophagia, which will manifest as bleeding from
the mouth
25 The incision does not heal
appropriately and appears
red, warm,
swollen and/or purulent
The skin incision is infected,
probably because of insufficient
closure of the incision
Eliminate dead space when closing muscle and fascia.
Place subcuticular stitches close together to completely
close the incision
29 The rabbit is exhibiting a
poor appetite
The rabbit is not used to soft
food or is experiencing pain
The rabbit must be acclimated to a soft diet before
initiation of the surgeries. Provide additional analgesia
if other signs point to pain or distress
The rabbit does not
recover as expected, as
observed by abnormal
posture, decreased activity,
decreased urine or fecal
output, or poor grooming
The rabbit is experiencing distress
as a result of the procedure,
possibly because of iatrogenic
fracture of the jaw, inflammation,
swelling or implant failure
Daily weights should be recorded to monitor progress.
Food can be supplemented with treats and/or Critical
Care diet as needed to maintain weight. The animal may
need increased or longer pain relief. If iatrogenic fracture
is suspected, a radiograph can be performed

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
2006 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
from constructs containing colistin-loaded poly(DL-lactic-co-glycolic acid) (PLGA) microparticles. The objective of this study
was to evaluate the effects of antibiotic dose and release kinetics on tissue healing and infection clearance. As seen in
Figure 8a–c, soft tissue can be characterized as completely healed (Fig. 8a) or nonhealed (Fig. 8b) over the area of the oral
communication. In some cases, a separate dehiscence can be seen where oral mucosa on the lingual mandible has broken
down over the implant (Fig. 8c). The distinction between nonhealing mucosa and dehiscence is important to consider, as
tissue over the defect in the area of the oral communication has to re-epithelialize in order to heal, whereas the soft tissue on
the lingual and buccal aspects of the mandible was not removed during the procedure. In this study, a high dose of antibiotic
delivered over a short time (gelatin group) inhibited the healing of soft tissue across the intraoral communication as com-
pared with a high dose of antibiotic delivered over a longer period of time (PLGA high group) (Fig. 8d). However, a high-dose
burst release had no effect on dehiscence (Fig. 8e)33. In this study, we demonstrated how to evaluate gross observation of
soft-tissue healing over implanted constructs and how to distinguish soft tissue that has dehisced from soft tissue that has
not healed. Gross evaluation may be used to evaluate drug delivery applications with the appropriate experimental groups.
MicroCT
Although planar radiography is fast and inexpensive, and may allow for qualitative evaluation of bone regeneration and/or
decalcification, it is not suitable for quantitative evaluation of bone regeneration in mandibular defects because of the
bicortical nature of mandibular bone. Therefore, microCT evaluation of bone regeneration and vascularization produces the
best quantitative results. Previous work in our laboratory has demonstrated the visualization of vessel growth within rabbit
10
*
8
6
4
Number of animals
2
0
Nonhealed Healed
Dehiscence No dehiscence
10
8
6
4
Number of animals
2
0
Group
Uninfected
Gelatin
PLGA low
PLGA high
Group
Uninfected
Gelatin
PLGA low
PLGA high
a
b
c
d
e
Figure 8 | Gross evaluation of soft-tissue healing at 12 weeks after surgery.
(a) Healed oral mucosa over the intraoral communications. (b) Nonhealed oral
mucosa, revealing the implanted porous polymethylmethacrylate construct.
(c) A lingual dehiscence, separate from the area of the intraoral communication.
(d) Statistical analysis of healed versus nonhealed samples reveals that
immediate antibiotic release from gelatin (gelatin group) results in less
soft-tissue healing than a high dose of controlled antibiotic release (PLGA high
group) (P < 0.05). Statistics were performed using the Fisher–Freeman–Halton
Test with post hoc analysis by Fisher’s exact test. (e) Statistical analysis of
occurrence of lingual dehiscence shows that there is no difference between
groups. Statistics were performed using the Krukal–Wallis test with post hoc
analysis by the Mann–Whitney U-test. Adapted from Acta Biomaterialia, 9,
Spicer et al., Evaluation of antibiotic releasing porous polymethylmethacrylate
space maintainers in an infected composite tissue defect model, 8832–8839,
copyright 2013 with permission from Acta Materialia, Inc. All surgical
procedures followed protocols approved by the University of Texas Health
Science Center at Houston Institutional Animal Care and Use Committee and
the Rice University Institutional Animal Care and Use Committee.
a b
c d
Figure 9 | Representative microCT images of vasculature surrounding
an implant in a full-thickness defect without intraoral communication.
Note that the images are coronal, with the buccal and lingual aspects
shown on the right and left, respectively, of each image. (a) Empty
control defect at 2 weeks. (b) Scaffold-implanted defect at 2 weeks.
(c) Empty control defect at 4 weeks. (d) Scaffold-implanted defect
at 4 weeks. Scale bars, 2 mm. Reproduced with permission from
ref. 35; the publisher of this copyrighted material is Mary Ann Liebert,
Inc. All surgical procedures followed protocols approved by the
University of Texas Health Science Center at Houston Institutional
Animal Care and Use Committee.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 2007
mandibular defects that are either empty controls
(Fig. 9a,c) or filled with porous polypropylene fumarate
(PPF) scaffolds (Fig. 9b,d)35. The images in Figure 9 are
maximum-intensity projections, which represent a valuable
tool used to qualitatively evaluate vessel network morphol-
ogy before 3D modeling. Quantification of the percentage
of vascularization (PV) and average vessel thickness (AVT)
was performed by microCT to investigate the effect of the
implant on neovascularization in empty control defects and
PPF-implanted defects35. Empty defects had a mean PV of
5.29 ± 0.80% at 2 weeks and 7.18 ± 1.04% at 4 weeks, in
contrast to PPF-implanted defects, which had a mean PV
of 1.58 ± 0.24% at 2 weeks and 2.38 ± 0.75% at 4 weeks
(n = 6 per group)35. Quantification of AVT was also reported,
revealing that empty controls had an AVT of 180.8 ± 7.2 µm
at 2 weeks that decreased significantly to 151.8 ± 10.3 µm
at 4 weeks (P < 0.05; n = 6). PPF-implanted defects showed
a similar trend, although the results were nonsignificant (P > 0.05) in this group, and the decreasing AVT value is thought
to be due to an immature vessel network at 2 weeks that undergoes remodeling by 4 weeks35. Vessel size distribution was
determined to range from 45 to 226 µm in all groups. These types of analysis allow researchers to visualize and quantify
vessel growth within healing defects in order to assess the effects of treatment on neovascularization.
Quantification of bone regeneration can be performed by scoring bony bridging across the defect (Supplementary Table 1)
and by calculation of regenerated bone volume4,34. As seen in Figure 10, the presence of a radio-opaque biomaterial implant
renders quantification of regenerated bone volume difficult, and thus a more sophisticated image acquisition technique, such
as dual-energy computed tomography, may be required to accurately quantify regenerated bone volume51. However, these
images can still be used to quantify bridging across the defect according to Supplementary Table 1.
Histology
Histologic specimens can be blocked through either paraffin or methacrylate embedding, if specimens are appropriately
prepared. In a study of empty partial-thickness and full-thickness mandibular defects, decalcified samples were prepared
by paraffin embedding and then stained with H&E to show microscopic tissue features at the anterior, central and posterior
portions of the defect3. Hematoxylin stains cell nuclei blue and eosin stains cell cytoplasm and fibrous tissue pink.
Our laboratory more commonly uses methacrylate embedding without decalcification because of the use of biomaterial
implants that cannot be sectioned with a microtome. The methacrylate-embedded sections are stained with methylene
blue and basic fuchsin, which stains nuclei purple; collagen and connective tissue blue; and bone red. Figure 11a,b shows
representative histologic images from a study in which porous antibiotic-releasing polymethylmethacrylate constructs were
Score: 3 Score: 2Score: 4
a b c
Figure 10 | Representative microCT scoring images (as described
in Supplementary Table 1) of bone regeneration over a
polymethylmethacrylate implant. The dotted circles indicate the original
defect area where a polymethylmethacrylate construct was implanted.
(a) Bone has fully bridged across the longest axis of the defect, score 4.
(b) Bone has bridged around the borders of the defect but has not bridged
across the longest defect, score 3. (c) Bone has grown in distinct islands
that are not connected to each other, score 2. All surgical procedures
followed protocols approved by the University of Texas Health Science
Center at Houston Institutional Animal Care and Use Committee and the
Department of Defense Animal Care and Use Review Office.
10
Number of animals
8
6
4
2
0
10
Number of animals
8
6
4
2
0
01234 0
Histologic score
*
1 2 3 4 0 12 34 0 1 23 4
01234 0
Histologic score
1234 01 234 0 1 23 4
Uninfected
Gelatin
PLGA low
PLGA high
a
b
c
d
Figure 11 | Representative histologic images and scoring histograms
(as described in Supplementary Table 2). (a,b) Histologic images are
stained with methylene blue/basic fuchsin and sectioned coronally through
the intraoral communication (top) with the buccal and lingual aspects of
the mandible located on the right and left sides of the left-hand images,
respectively. The left-hand images are low-magnification images that display
the entire defect. The insets on the right are high-magnification images of
the fibrous capsule. Scale bars, 2 mm (left-hand images) and 50 µm (insets).
(a) The high-magnification inset shows a mature fibrous capsule (white
arrow) separating the implant from regenerated bone. (b) The high-
magnification inset shows inflammatory cells (*) with a disorganized fibrous
capsule (white arrow) in contact with the implant. (c) Histogram of scores
at the tissue–implant interface denotes significant differences (P < 0.5).
(d) Histogram of scores within the pores of the implant. Histologic scores
were statistically analyzed using the Kruskal–Wallis test with post hoc
analysis using the Steel–Dwass test. Panels c and d adapted from Shah, S. R.
et al. Polymer-based local antibiotic delivery for prevention of polymicrobial
infection in contaminated mandibular implants. ACS Biomater. Sci. Eng., 2,
558–566 (2016). Copyright 2016 American Chemical Society. All surgical
procedures followed protocols approved by the University of Texas Health
Science Center at Houston Institutional Animal Care and Use Committee and
the Department of Defense Animal Care and Use Review Office.

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
2008 | VOL.11 NO.10 | 2016 | NATURE PROTOCOLS
implanted in a contaminated mandibular defect with intraoral communication. The inset of Figure 11a is a high-magnitude
image of the interface between the biomaterial implant and host tissue showing regenerated bone separated from the
implant by a mature fibrous capsule. These sections are also useful for evaluating the in-growth of tissue into the porous
implant, as fibrous tissue can be seen growing around into the pores of the implant in Figure 11a. The inset of Figure 11b
is a high-magnification image of the tissue–implant interface in the area of the intraoral communication, showing
inflammatory cells and a less organized fibrous capsule at the tissue–implant interface. Typical data obtained from
scoring tissue at the interface and within the pores as described in Supplementary Table 2 can be seen in Figure 11c,
demonstrating that a high dose of controlled-release antibiotic into an infected defect results in superior tissue healing
and bone regeneration at the tissue–implant interface33. Figure 11d shows that the tissue within the pores of constructs
is not different between groups and consists primarily of either acellular fluid or fibrous tissue33.
Ancillary evaluations
Because the rabbit mandibular defect model can be modified in several ways to evaluate variations on tissue
engineering scaffolds, ancillary tests may be required to perform a complete evaluation of the implant beyond bone or
vessel regeneration. These tests should be considered before submitting the protocol for institutional approval. In two
studies, our laboratory used this model for evaluation of infected mandibular defects by inoculation with the bacterial
pathogens Acinetobacter baumannii and Prevotella and treating the defect locally with antibiotic-loaded constructs.
Bacterial cultures were taken from rabbit mouths and directly from the implants to determine clearance of the bacteria
with which the animals were inoculated33,34. Blood samples drawn from the central ear artery have been used to
determine systemic antibiotic concentration and to evaluate for end-organ damage (nephrotoxicity) by blood urea
nitrogen/creatinine levels33,34. Finally, if end-organ damage is suspected, the organ of interest can also be collected
and evaluated by histology.
In this protocol, we describe a reproducible, orthotopic, critical-size mandibular defect suitable for preclinical testing
of biomaterial implants for bone regeneration. Analytical techniques include gross soft-tissue, radiographic and histologic
evaluation to assess functional, mineral/vascular and cellular responses to biomaterials.
Note: Any Supplementary Information and Source Data files are available in the
online version of the paper.
ACKNOWLEDGMENTS A.G.M. acknowledges support toward the development of
biomaterials for tissue engineering applications by the National Institutes of
Health (grant R01 AR068073). A.G.M. and M.E.W. acknowledge support toward
the development of materials and techniques through the Armed Forces Institute
of Regenerative Medicine (award no. W81XWH-14-2-0004). S.R.S. acknowledges
support from a Ruth L. Kirschstein Fellowship from the National Institutes of Health
(F30 AR067606). We acknowledge A. Tatara, J. Lam and S. Lu for their assistance
with photography, P. Spicer for sharing his experience with the model and S. Frazier
for her assistance in providing details of analgesia and intraoperative monitoring.
AUTHOR CONTRIBUTIONS S.R.S. wrote the protocol and developed
adaptations to the protocol; S.Y., M.E.W. and A.G.M. developed the protocol;
J.L.G. developed the veterinary care plan for the protocol; J.A.J. assisted
with the development of analytical protocols.
COMPETING FINANCIAL INTERESTS The authors declare no competing financial
interests.
Reprints and permissions information is available online at http://www.nature.
com/reprints/index.html.
1. Smith, B.T., Shum, J., Wong, M., Mikos, A.G. & Young, S. Bone tissue
engineering challenges in oral & maxillofacial surgery. Adv. Exp. Med. Biol.
881, 57–78 (2015).
2. Schmitz, J.P. & Hollinger, J.O. The critical size defect as an experimental-
model for craniomandibulofacial nonunions. Clin. Orthop. Relat. Res. 205,
299–308 (1986).
3. Takagi, K. & Urist, M.R. The reaction of the dura to bone morphogenetic
protein (Bmp) in repair of skull defects. Ann. Surg. 196, 100–109 (1982).
4. Young, S. et al. Development and characterization of a rabbit
alveolar bone nonhealing defect model. J. Biomed. Mater. Res. A 86A,
182–194 (2008).
5. Spicer, P.P. et al. Evaluation of bone regeneration using the rat critical
size calvarial defect. Nat. Protoc. 7, 1918–1929 (2012).
6. Cooper, G.M. et al. Testing the critical size in calvarial bone defects:
revisiting the concept of a critical-size defect. Plast. Reconstr. Surg.
125, 1685–1692 (2010).
7. Kasper, F.K., Tanahashi, K., Fisher, J.P. & Mikos, A.G. Synthesis of
poly(propylene fumarate). Nat. Protoc. 4, 518–525 (2009).
8. Kinard, L.A., Kasper, F.K. & Mikos, A.G. Synthesis of oligo(poly(ethylene
glycol) fumarate). Nat. Protoc. 7, 1219–1227 (2012).
9. Watson, B.M. et al. Biodegradable, phosphate-containing, dual-gelling
macromers for cellular delivery in bone tissue engineering. Biomaterials
67, 286–296 (2015).
10. Sanz-Herrera, J.A. & Reina-Romo, E. Cell-biomaterial mechanical
interaction in the framework of tissue engineering: insights,
computational modeling and perspectives. Int. J. Mol. Sci. 12,
8217–8244 (2011).
11. Bancroft, G.N., Sikavitsas, V.I. & Mikos, A.G. Design of a flow perfusion
bioreactor system for bone tissue-engineering applications. Tissue Eng. 9,
549–554 (2003).
12. Tatara, A.M., Shah, S.R., Livingston, C.E. & Mikos, A.G. Infected animal
models for tissue engineering. Methods 84, 17–24 (2015).
13. Mankani, M.H., Kuznetsov, S.A., Wolfe, R.M., Marshall, G.W. & Robey, P.G.
In vivo bone formation by human bone marrow stromal cells:
reconstruction of the mouse calvarium and mandible. Stem Cells
24, 2140–2149 (2006).
14. Saadeh, P.B. et al. Repair of a critical size defect in the rat mandible
using allogenic type I collagen. J. Craniofac. Surg. 12, 573–579 (2001).
15. van Leeuwen, A.C. et al. Guided bone regeneration in rat mandibular
defects using resorbable poly(trimethylene carbonate) barrier membranes.
Acta Biomater. 8, 1422–1429 (2012).
16. Stubinger, S. & Dard, M. The rabbit as experimental model for research in
implant dentistry and related tissue regeneration. J. Invest. Surg. 26,
266–282 (2013).
17. Elsalanty, M.E. et al. Reconstruction of canine mandibular bone defects
using a bone transport reconstruction plate. Ann. Plast. Surg. 63,
441–448 (2009).
18. Nienhuijs, M.E.L. et al. Healing of bone defects in the goat mandible,
using COLLOSS (R) E and beta-tricalciumphosphate. J. Biomed. Mater. Res.
B 92B, 517–524 (2010).
19. Zhou, L.B. et al. Reconstruction of curved mandibular angle defects
using a new internal transport distraction device: an experiment in goats.
Br. J. Oral Maxillofac. Surg. 46, 445–448 (2008).
20. Ma, J.L., Pan, J.L., Tan, B.S. & Cui, F.Z. Determination of critical size
defect of minipig mandible. J. Tissue Eng. Regen. Med. 3, 615–622
(2009).

© 2016Nature America, Inc. All rights reserved.
PROTOCOL
NATURE PROTOCOLS | VOL.11 NO.10 | 2016 | 2009
21. Sun, Z.Y., Kennedy, K.S., Tee, B.C., Damron, J.B. & Allen, M.J. Establishing
a critical-size mandibular defect model in growing pigs: characterization
of spontaneous healing. J. Oral Maxillofac. Surg. 72, 1852–1868 (2014).
22. Tatara, A.M. et al. Autologously generated tissue-engineered bone flaps
for reconstruction of large mandibular defects in an ovine model. Tissue
Eng. Part A 21, 1520–1528 (2015).
23. Dard, M. Animal models for experimental surgical research in implant
dentistry. in Implant Dentistry Research Guide: Basic, Translational and
Experimental Clinical Research. (ed. Ballo, A.) 167–190 (Nova Science
Publishers, New York, 2012).
24. Dard, M. Methods and interpretation of performance studies for dental
implants. in Biocompatibility and Performance of Medical Devices
(ed. Jean-Pierre Boutrand) 308–344 (Woodhead Publishing,
Cambridge, UK, 2012).
25. Guo, J. et al. Restoration of critical-size defects in the rabbit mandible
using porous nanohydroxyapatite-polyamide scaffolds. Tissue Eng. Part A
18, 1239–1252 (2012).
26. Campillo, V.E., Langonnet, S., Pierrefeu, A. & Chaux-Bodard, A.G.
Anatomic and histological study of the rabbit mandible as an experimental
model for wound healing and surgical therapies. Lab. Anim. 48, 273–277
(2014).
27. Miloro, M., Haralson, D.J. & Desa, V. Bone healing in a rabbit mandibular
defect using platelet-rich plasma. J. Oral Maxillofac. Surg. 68, 1225–1230
(2010).
28. Rottgers, S.A. et al. Bone morphogenetic protein 2-mediated mandible
reconstruction successfully heals bony defects but inhibits concurrent
inferior alveolar nerve grafting: a rabbit experimental model.
J. Craniofac. Surg. 25, 2241–2245 (2014).
29. Kretlow, J.D. et al. Evaluation of soft tissue coverage over porous
polymethylmethacrylate space maintainers within nonhealing alveolar bone
defects. Tissue Eng. Part C Methods 16, 1427–1438 (2010).
30. Nguyen, C., Young, S., Kretlow, J.D., Mikos, A.G. & Wong, M. Surface
characteristics of biomaterials used for space maintenance in a mandibular
defect: a pilot animal study. J. Oral Maxillofac. Surg. 69, 11–18 (2011).
31. Spicer, P.P. et al. In situ formation of porous space maintainers in a
composite tissue defect. J. Biomed. Mater. Res. A 100A, 827–833 (2012).
32. Petersen, K. et al. Prevention of infections associated with combat-related
eye, maxillofacial, and neck injuries. J. Trauma 71, S264–S269 (2011).
33. Spicer, P.P. et al. Evaluation of antibiotic releasing porous
polymethylmethacrylate space maintainers in an infected composite tissue
defect model. Acta Biomater. 9, 8832–8839 (2013).
34. Shah, S.R. et al. Polymer-based local antibiotic delivery for prevention
of polymicrobial infection in contaminated mandibular implants.
ACS Biomater. Sci. Eng. 2, 558–566 (2016).
35. Young, S. et al. Microcomputed tomography characterization of
neovascularization in bone tissue engineering applications.
Tissue Eng. Part B Rev. 14, 295–306 (2008).
36. Lalani, Z., Wong, M., Brey, E.M., Mikos, A.G. & Duke, P.J. Spatial and
temporal localization of transforming growth factor-beta1, bone
morphogenetic protein-2, and platelet-derived growth factor-A in healing
tooth extraction sockets in a rabbit model. J. Oral Maxillofac. Surg. 61,
1061–1072 (2003).
37. Wen, B. et al. The osseointegration behavior of titanium-zirconium
implants in ovariectomized rabbits. Clin. Oral Implants Res. 25, 819–825
(2014).
38. Schon, R., Ohno, K., Kudo, M. & Michi, K. Peri-implant tissue reaction in
bone irradiated the fifth day after implantation in rabbits: histologic and
histomorphometric measurements. Int. J. Oral Maxillofac. Implants 11,
228–238 (1996).
39. Jegoux, F. et al. Alveolar ridge augmentation in irradiated rabbit
mandibles. J. Biomed. Mater. Res. A 93, 1519–1526 (2010).
40. German, C.L. & Madihally, S.V. Applications of computational modelling
and simulation of porous medium in tissue engineering. Computation 4,
article 7, http://dx.doi.org/10.3390/computation4010007 (2016).
41. Collinet-Adler, S., Castro, C.A., Ledonio, C.G., Bechtold, J.E. & Tsukayama,
D.T. Acinetobacter baumannii is not associated with osteomyelitis in a rat
model: a pilot study. Clin. Orthop. Relat. Res. 469, 274–282 (2011).
42. Vo, T.N., Ekenseair, A.K., Kasper, F.K. & Mikos, A.G. Synthesis,
physicochemical characterization, and cytocompatibility of bioresorbable,
dual-gelling injectable hydrogels. Biomacromolecules 15, 132–142 (2014).
43. Sarhaddi, D. et al. Validation of histologic bone analysis following microfil
vessel perfusion. J. Histotechnol. 35, 180–183 (2012).
44. Fischer, E.R., Hansen, B.T., Nair, V., Hoyt, F.H. & Dorward, D.W. Scanning
electron microscopy. Curr. Protoc. Microbiol. Chapter 2 Unit 2B 2, (2012).
45. Graham, L. & Orenstein, J.M. Processing tissue and cells for transmission
electron microscopy in diagnostic pathology and research. Nat. Protoc. 2,
2439–2450 (2007).
46. Webster, J.D., Miller, M.A., Dusold, D. & Ramos-Vara, J. Effects of
prolonged formalin fixation on diagnostic immunohistochemistry in
domestic animals. J. Histochem. Cytochem. 57, 753–761 (2009).
47. Kallai, I. et al. Microcomputed tomography-based structural analysis of
various bone tissue regeneration models. Nat. Protoc. 6, 105–110 (2011).
48. Meganck, J.A., Kozloff, K.M., Thornton, M.M., Broski, S.M. & Goldstein,
S.A. Beam hardening artifacts in micro-computed tomography scanning
can be reduced by X-ray beam filtration and the resulting images can be
used to accurately measure BMD. Bone 45, 1104–1116 (2009).
49. An, Y.H. & Martin, K.L. Handbook of Histology Methods for Bone and
Cartilage (Humana Press, New York, 2003).
50. Patel, Z.S. et al. Dual delivery of an angiogenic and an osteogenic
growth factor for bone regeneration in a critical size defect model.
Bone 43, 931–940 (2008).
51. Taschereau, R., Silverman, R.W. & Chatziioannou, A.F. Dual-energy
attenuation coefficient decomposition with differential filtration and
application to a microCT scanner. Phys. Med. Biol. 55, 1141–1155 (2010).