J. Biomedical Science and Engineering, 2009, 2, 24-29
Published Online February 2009 in SciRes. http://www.scirp.org/journal/jbise JBiSE
In vivo testing of a bone graft containing chitosan,
calcium sulfate and osteoblasts in a paste form in a
critical size defect model in rats
Jerome G. Saltarrelli Jr.1, Debi P. Mukherjee1*
Louisiana State University Health Sciences Center Department of Orthopaedic Surgery, 1501 Kings Highway, Shreveport, Louisiana 71130, * Corresponding
author to Debi P. Mukherjee: ([e-mail])
Received September 26th, 2008; revised November 12th, 2008; accepted November 13th, 2008
ABSTRACT
Bone loss associated with musculoskeletal
trauma or metabolic diseases often require
bone grafting. The supply of allograft and auto-
graft bones is limited. Hence, development of
synthetic bone grafting materials is an active
area of research. Chitosan, extracted from chitin
present in crawfish shells, was tested as a de-
livery vehicle for osteoblasts in a 2-3 mm size
defect model in rats. Twenty-seven male Lewis
rats, divided into three groups with sacrifice
intervals of 3, 6 & 9 months were used. In the
experimental samples, a critical size defect was
filled with chitosan bone graft paste and fixed
with a plate, while in the operated control group,
a critical size defect was repaired only by a plate
(no paste was applied). An unoperated control
group was also included. Bone growth was
evaluated histologically by examining undecal-
cified and decalcified stained sections. The fe-
murs were also examined non-destructively by
micro-computed tomography (µCT). Defects
filled with chitosan bone graft paste demon-
strated superior healing across all time periods
compared to unfilled defects as examined by
histology and micro-computed tomography.
Crawfish chitosan has successfully been used
as a cell delivery system for osteoblasts for use
as a synthetic bone graft material.
Keywords: Chitosan, Synthetic Bone Graft, Cell
Delivery, Histology, Animal Model Running
Head: Chitosan Based Synthetic Bone Graft
Material
1. INTRODUCTION
Bone loss associated with musculoskeletal trauma or
metabolic diseases often require bone grafting. Autograft
or allograft bones are limited in supply. Therefore, de-
velopment of synthetic bone graft materials is an active
area of research.
Chitin is a polysaccharide that exists in fungi, exo-
skeleton of insects and the outer shell of crustaceans
[1,2,3]. It is biocompatible [4], osteoconductive [5], an-
timicrobial [6,7], biodegradable [8], non-toxic [9],
haemostatic [10], fungicidal [11], and the second most
abundant natural polysaccharide on earth [12]. Removal
of the acetyl group from chitin forms chitosan. Chitosan
is more useful due to the presence of amino groups that
impart a positive charge to the molecule. Chitosan has
been investigated in a number of biomedical applications
due to its purity coupled with a positive charge
[13,14,15,16]. This positive charge interacts with cells or
can act as a binding site for other functional groups
thereby expanding the role of the chitosan molecules.
Recently, a new patented process has been developed
to purify chitosan from crawfish shells [17]. The objec-
tive of this current study was to evaluate this crawfish
chitosan as a delivery system for osteoblasts to promote
bone growth in a 2-3mm defect in rats. For this purpose,
prepared crawfish chitosan was compounded with cal-
cium sulfate to form a paste. The bone growths at 3, 6 &
9 months were estimated by histology and microcom-
puted tomography (µCT). It was hypothesized that chi-
tosan provides a delivery system, keeping the cells in the
defect area for a longer period of time allowing defect
repair.
2. MATERIALS AND METHODS
2.1. Chitosan Extraction
Chitosan was extracted from crawfish shells following a
patented process [17]. Briefly, shells were first washed,
and dried in an 80°C oven for 48 hours. Following dry-
ing, the shells were immediately quenched in liquid ni-
trogen then treated by: (1) 3.5% NaOH to remove pro-
teins; (2) 1N HCl to remove minerals; and (3) 50%
NaOH to remove acetyl groups. The extracted chitosan
was purified through a 12,000-14,000 Dalton dialysis
membrane and dried.
2.2. Preparation of Osteoblasts
Stromal osteoblast cells were obtained from the marrow
SciRes Copyright © 2009

J. G. Saltarrelli et al. / J. Biomedical Science and Engineering 2 (2009) 24-29 25
SciRes Copyright © 2009 JBiSE
of young adult male (125-149g) Lewis rats. Following
euthanasia by CO2 asphyxiation, femora were aseptically
excised, cleaned of soft tissue, washed in DMEM+ anti-
biotic-antimycotic (concentration of antibiotic- antimy-
cotic was 10 times the normal amount used in cellular
media). The metaphyseal ends were cut off and the mar-
row flushed from the midshaft with 5ml of media
(DMEM+10% FBS+antibiotic-antimyctotic) using a
syringe equipped with a 22-gauge needle and collected
in a sterile test tube. Cell clumps were broken up by re-
peated pipetting of the cell suspension. The cells were
centrifuged at 1200 rpm for 10 minutes. The cell pellet
was resuspended in media (DMEM+10% FBS + antibi-
otic-antimycotic+10% IL-3) and seeded in a flask. On
the following day, the media was removed and the cells
were washed with 10x concentration of antibiotic- anti-
mycotic in PBS; and complete media was introduced
into the flask. Cells were fed every two to three days
until confluent, and then trypsinized.
2.3. Paste Formulation
The ratio of chitosan to calcium sulfate was 1:4, ideally
0.125g of crawfish chitosan to 0.5g calcium sul-
fate. The actual average paste contents were 0.5030g of
calcium sulfate and 0.1234g of crawfish chitosan with an
average osteoblast concentration of 4.06 x 105/ml. The
calcium sulfate (CaSO4) and extracted chitosan, after
sterilization under dry heat, was mixed with 1ml of os-
teoblast cell suspension (106 cells/ml) to form a paste.
This was accomplished immediately prior to implanta-
tion under the sterile field. The same procedure was used
in a previous study repairing a cranial critical size defect
in rats [18].
2.4. Experimental Design and Animal Sur-
gical Procedure
Twenty seven male Lewis rats were divided into 3
groups: (1) operated control (2) experimental and (3)
non-operated control. Animals in the operated control
and experimental groups went through a surgical proce-
dure which created a segmental defect of 2-3mm which
is a modification of the 8mm critical size segmental de-
fect model [19,20]. This modification was necessary
because a four-hole 23mm long plate was used, which
had a solid section of about 4mm between the second
and third hole covering the femoral defect. This modifi-
cation in defect size was necessary for proper femoral
repair. The operated control group animals received only
plate fixation. The experimental group received both
plate fixation and bone graft paste. The unoperated
group was a control. Each group, of three animals per
time period, was studied for 3, 6 and 9 months.
The surgical procedure is as follows: animals were
anesthetized with 1.5% Isoflurane, shaved and cleaned
with 70% alcohol and Betadine. An incision was made
dorsally to the femur and a 4-hole titanium plate was
applied. In both operated groups (experimental and op-
erated control), once the plate was installed a 2-3 mm
size defect was applied to the femur using a burring bit
on a Dremel tool. In the operated control group, the inci-
sion was closed. In the experimental group, the chito-
san-bone graft paste was applied to the defect. Post sur-
gery, animals were housed in individual cages and
monitored for surgical complications (foot drop, lethargy,
loss of use and pain). Upon euthanizing by CO2 as-
phyxiation, operated and contra-lateral control femurs
were removed and stored frozen until testing in a 0.9%
NaCl solution.
2.5. Undecalcified Histological Sample
Preparation and Staining
All undecalcified histological samples were fixed in
10% formalin for at least 2 weeks. Following formalin
fixation, the samples were taken through an increasing
gradient of 2-hydroxyethyl-methacrylate and nanopure
water then a transition of 2-hydroxyethyl-methacrylate
and Technovit 7200 and finally 100% Technovit 7200.
Samples were embedded in resin by submersion in fresh
Technovit 7200 and exposure to an ultraviolet light
(λ=450nm). Ultraviolet light causes polymerization of
Technovit 7200. Sample blocks were attached to slides
using Technovit 7210LVC, cut using an Exakt band saw
and ground into 10-20µm thick slices using an Exakt sur-
face grinder (Exakt Technologies, Oklahoma City, OK).
All undecalcified histological sections were stained
using Goldner’s trichrome method whereas bone stains
green, muscle stains orange, and fibrous tissue stains
red.
2.6. Non-destructive Evaluation by Micro-
computed Tomography (µCT)
µCT was preformed (courtesy of Animal Resources,
Louisiana State University Health Sciences Center,
Shreveport, Louisiana) on a MicroCAT II; model MCII-
UAAI following standard operating procedure. Each
femur, along with the contra-lateral side, was run for 512
scans along the defect site. This data was compiled into
2-D and 3-D images and qualitative information on bone
growth was collected.
3. RESULTS
3.1. Undecalcified Histological Sectioning
Three, six and nine months data for operated control and
experimental samples are shown in Figure 1, A through
I and an example of unoperated mature bone is shown in
Figure 1-J. Goldner’s Trichome stain was used which
stains bone green and muscle and fibrous tissues red. All
photographs shown were taken at 4x magnification un-
der a light microscope.
3.2. Histological Quantitative Data
From bone pixelation, bone percentages from each group
can be compared (Figure 2). In each time interval, more
bone was present in the experimental group compared to

26 J. G. Saltarrelli et al. / J. Biomedical Science and Engineering 2 (2009) 24-29
SciRes Copyright © 2009 JBiSE
Figure 1. Undecalcified Histology. (Figure 1-A & 1-B shows the presence of a defect after 3 months without
chitosan/ Plaster of Paris bone graft paste (operated control group) (1-A=5x; 1-B=20x). Figure 1-C shows
experimental samples after 3 months, the defect was repaired with beginning of intermedullary canal. After
6 months, significant fibrous growth remained in the operated control group (Figure 1-D = 20x). In the 6
month experimental group, rapid repair has occurred (vs. operated control) with the presence of large voids
(Figure 1-E = 20x). Figure 1-F, G & H (20x) show the 9 month operated control group with infiltration of fi-
brous tissue which inhibits bone growth. Figure 1-I (5x) is the experimental group after 9 months with near
complete bone growth. Figure 1-J is an unoperated control (20x).)
the operated control group. During the three month in-
terval, 40.23% bone growth was calculated in the ex-
perimental group while only 29.43% was calculated in
the operated control group. At the six month time inter-
val, 35.58% bone growth was calculated in the experi-
mental group and 30.09% bone growth was recorded in
the operated control group. As expected, nine month
time interval demonstrated the highest percentage of
bone growth in both groups. The experimental group
experienced 42.82% bone growth while the operated
control group experienced 40.16%. In all the sacrifice
intervals the degree of bone growth was slightly higher
for the experimental group than that of the operated con-
trol group. After three months, the experimental samples
demonstrated 26.85% more bone in the defect site than
the operated control samples. After six months, the ex-
perimental samples demonstrated 15.42% more bone
and after nine months, 6.2% more bone was evident in
the defect site. In depth statistical analysis of the data
was not possible due to limited number of animals.
3.3. Non-destructive Testing by Microcom-
puted Tomography (µCT)
µCT was preformed on a MicroCAT II; model
MCII-UAAI following standard operating procedure.
Each femur, along with the contra-lateral side, was run

J. G. Saltarrelli et al. / J. Biomedical Science and Engineering 2 (2009) 24-29 27
SciRes Copyright © 2009 JBiSE
for 512 scans along defect site. This data was compiled
into 2-D and 3-D images. We were unable to extract the
quantitative information like bone volume or connec-
tivity density of trabecular bone from the µCT images
since we did not possess the appropriate software.
Therefore we qualitatively looked at the 3-D rendered
µCT images for bone repair. The 3, 6 and 9 month data are
shown in Figures 3A-3F while Figure 3-G displays unop-
erated control data. MicroCT data showed gradual bone
growth even at the 3 month period as seen with the com-
puterized volumetric rendering data (3-B). At six and
Figure 2. Bone Growth Fractions. (Bone growth fractions were calculated by pixel quantification within the
defect site. In all time periods, more bone is present in the experimental samples versus the operated control
samples, as represented here. Error bars represent variation between specimens.)
Figure 3. Microcomputed Tomography 3-D Images. (Figure 3-A is an operated control specimen after 3 months. An in-
complete intermedullary canal as well as the presence of a surface defect can be visualized. Figure 3-B is an experimental
sample after 3 months. Increase in the intermedullary canal and a decrease in the surface defect (versus the operated
control sample from the same time point) can be seen. Figure 3-C is the operated control sample after 6 months. An ap-
parent surface defect is present compared to the experimental sample (from the same time point) in figure 3-D. Figure 3-E
and 3-F are the operated control and experimental samples, respectively, after 9 months. A complete defect, with boney
ingrowth, can be seen in the operated control sample (Figure 3-E). Lack of significant surface defect can be visualized in
the experimental sample (Figure 3-F) along with complete defect repair. Figure 3-G is an unoperated control.)

28 J. G. Saltarrelli et al. / J. Biomedical Science and Engineering 2 (2009) 24-29
SciRes Copyright © 2009 JBiSE
nine months (Figures 3-D & 3-F respectively) the de-
fect-free bone is seen in the experimental group. However,
the repair was incomplete in operated controls 3-A, 3-C and
3-E. This confirmed bone growth was stimulated in the
experimental group.
4. DISCUSSION
4.1. Undecalcified Histological Sectioning
and Staining
During each time interval (three, six and nine months)
repair occurred faster in experimental groups compared
to than in the operated control groups. The repaired area
in the experimental group lacked fibrous tissue and, his-
tologically, appeared to consist primarily of cortical bone.
In comparison, operated control samples demonstrated
repair with a relative high concentration of voids and
fibrous tissues. Defects were present, in the operated
control femurs, after even nine months of repair time.
The color variation in the sections was due to speci-
men thickness, whereby more tissue was stained. Using
the Exakt system, exact specimen thickness is difficult to
control. General colors were similar with slight shade
variation. Exact duplication in slide thickness was at-
tempted but not always achieved. Thus histological
analysis demonstrated the presence of bone.
At the three month time interval, bone infiltrated into
the defect site although infiltration was not complete, the
defect could still be visualized. In operated control sam-
ples, incomplete healing began at the defect site. Bone
infiltration was sporadic, filled with large voids, and the
defect was still present (Figures 1-A & 1-B). In experi-
mental samples, the defect was completely filled with
mature bone and the intermedullary canal was beginning
to reform (Figure 1-C). In the operated control histo-
logical sections, repaired bone could be visualized, al-
though the rate of repair was far inferior to repair seen in
the experimental group.
Six month undecalcified histological operated control
sections displayed significant fibrous growth at the de-
fect site. Tissue juxtaposed to the defect site was highly
infiltrated with fibrous tissue (Figure 1-D). Analysis of
later sections revealed the area of fibrous tissue de-
creased in the defect site. Six month experimental histo-
logical sections displayed superior growth compared to
the operated control samples. Repaired bone tissue dis-
played continuous growth although large voids, white
spaces, were noticed (Figure 1-E). Six month experi-
mental histological sections showed improved repair
compared to operated control samples. While voids were
present in experimental samples, operated control sam-
ples had complete defects present.
Nine month undecalcified operated control sections
displayed a high relative percentage of fibrous tissue in
the defect site. Unlike the six month group, fibrous tis-
sue was juxtaposed to cortical bone (Figure 1-F). In the
nine month undecalcified experimental sections, defects
were only slightly present with no signs of fibrous tissue
or voids (Figure 1-I). All defects, in experimental fe-
murs, were repaired with, histologically, cortical bone.
After surgery, rats gained weight due to lack of signifi-
cant movement and therefore had less space for move-
ment. This lack of significant movement reduced the
mechanostimulation on all loaded skeletal structures,
including the defected bone. Since repair processes
strive under loaded conditions and these lethargic ani-
mals presented reduced loaded conditions, bone growth
decreased from the three month group to the nine month
group. During all time intervals, animals were confined
to Institutional Animal Care and Use Committee (IA-
CUC) approved cages 10.5" W x 19" L x 8" H (Allen-
town Caging Equipment, Inc., Allentown, N.J.), physical
activity was negligible and feed intake was constant
(free choice). Reduced physical activity led to decreased
mechanostimulation which led to weakened repaired bone
[21,22]. Nevertheless the nine month experimental group
demonstrated a relative higher percentage of bone com-
pared to operated control femurs from the same time pe-
riod.
4.2. Non-Destructive Testing by Microcom-
puted Tomography (µCT)
MicroCT (µCT) is a series of x-ray images compiled
together into a three dimensional image. Individual
slices were available for viewing as well as three dimen-
sional images. Microcomputed tomography images of
the bone graft site displayed progressive repair in ex-
perimental samples compared to operated control femurs,
which displayed a complete defect throughout all time
periods. Operated control femurs, across all time inter-
vals, displayed similar characteristics. All operated con-
trol samples had incomplete intermedullary canals, de-
fect presence (appropriate for a critical sized defect) and
dense tissue growth over bone ends at defect site. Like-
wise, experimental samples displayed similar character-
istics across all time intervals. All experimental samples
had a slight surface defect present (although size de-
creased as time progressed) and intermedullary canal
was continuous throughout entire femoral length. Thus
µCT provided a non-destructive technique to view in-
ternal composition without compromising sample integ-
rity. Non- destructive microcomputed tomography test-
ing allowed femurs to undergo additional testing, thus
maximizing the quantity of data per sample.
In the operated control group, repair was present
along the ventral femoral surface. Repair was incomplete
(discontinuous intermedullary canal with defect on dor-
sal femoral surface) and resultant area of repair visually
appeared as dense cortical bone in a single slice image.
In Figure 3-A, the surface defect can be visualized along
with an incomplete intermedullary canal. In the experi-
mental group, significant bone repair was observed. In
Figure 3-B, the defect was almost completely repaired
with the exception of a small surface disruption and the
intermedullary canal has begun to reform.

J. G. Saltarrelli et al. / J. Biomedical Science and Engineering 2 (2009) 24-29 29
SciRes Copyright © 2009 JBiSE
In the six month operated control group, the defect
was still visible and bone was somewhat deformed
(Figure 3-C). In the experimental six month group, re-
pair was significantly improved compared to the oper-
ated control group during the same time interval. The
intermedullary canal was continuous throughout the fe-
mur. A slight remnant of the defect in cortical bone on
ventral surface remained. Figure 3-D shows the dorsal
plane of the defect completely healed and indistinguish-
able from surrounding bone.
In the nine month operated control group, cortical
bone defined edges of the defect (Figure 3-E) and repair
of the intermedullary canal was discontinuous. In the
nine month experimental group, repair was essentially
complete with only small amounts of defect present; the
three dimensional rendering showed the absence of a
visually apparent defect (Figure 3-F). Figure 3-G
showed the unoperated control femur.
5. CONCLUSIONS
The specific conclusions were:
1. The animal evaluations of the composite paste for 3,
6 & 9 months examined by undecalcified histology and
microCT (µCT) demonstrated a high degree of bone
ingrowth for the experimental group, compared to sam-
ples in the operated control group (defect repaired with-
out paste).
2. The use of µCT as a non-destructive tool to follow
bone ingrowth was found to be very valuable. Although
we used this technique mainly for qualitative data, the
data could be digitized to make more quantitative pre-
dictions.
3. Purified crawfish chitosan compounded with cal-
cium sulfate in paste form can be successfully used to
deliver osteoblasts that will allow bone growth to occur
in defects.
ACKNOWLEDGMENTS
The authors appreciate the help of Errin Robinson and Alan L. Ogden
for surgical assistance, Dollie Smith for growing the osteoblasts and
Shelia Rogers for guidance regarding the sectioning and staining of
slides. The plates were supplied by Synthes (Paoli, PA).
REFERENCES
[1] Q. Li, E. T. Dunn, E. W. Grandmasion, M. F. A. Goosen. (1992)
Applications and Properties of Chitosan. Applications of Chitin
and Chitosan, ed. M. F. A. Goosen. Technomic Publishing
Company, Inc.: Lancaster. 3-29.
[2] H. No, S. Meyers, (1995) Preparation and characterization of
chitin and chitosan-A review. Journal of Aquatic Food Product
Technology. 4(2): 27-52.
[3] H. K. No, S. P. Meyers, (1997) Preparation of chitin and chito-
san, in Chitin Handbook, R. A. A. Muzzarelli, Peter, M. G., Edi-
tor. 475-489.
[4] S. B. Rao, C. P. Sharma, (1997) Use of chitosan as a biomaterial:
Studies on its safety and haemostatic potential. Journal of Bio-
materials Research. 34: 21-28.
[5] T. Jiang, C. Pilane, C. (2005) Laurencin, Fabrication of Novel
Porous Chitosan Matrices as Scaffolds for Bone Tissue Engineer-
ing. Materials Research Society Symposium. 845: 187-192.
[6] Y. Chen, Y. Chung, L. Wang, K. Chen, S. Li, (2002) Antibacte-
rial Properties of Chitosan in Waterborne Pathogens. Journal of
Environmental Science and Health. A37 (7): 1379-1390.
[7] Z. Shi, K. Neoh, E. Kang, W. Wang, (2006) Antibacterial and
mechanical properties of bone cement impregnated with chito-
san nanoparticles. Biomaterials. 27: 2440-2449.
[8] T. Don, C. Chuang, W. Chiu, (2002) Studies on the Degradation
Behavior of Chitosan-g-Poly (acrylic acid) Copolymers. Tam-
kang Journal of Science and Engineering. 5(4): 235-240.
[9] P. VandeVord, H. M., S. DeSilva, L. Mayton, B. Wu, P. Wooley,
(2002) Evaluation of the biocompatibility of a chitosan scaffold
in mice. Journal of Biomedical Materials Research. 59: 585-590.
[10] Y. Okamoto, R. Yano, K. Miyatake, I. Tomohiro, Y. Shigemasa,
S. Minami, (2003) Effects of chitin and chitosan on blood co-
agulation. Carbohydrate Polymers. 53: 337-342.
[11] M. Jumaa, F. H. Furkert, B. W. Müller, (2002) A new lipid
emulsion formulation with high antimicrobial efficacy using
chitosan. European Journal of Pharmaceutics and Biopharma-
ceutics. 53: 115-123.
[12] P. A. Felse, T. Panda, (1999) Studies on applications of chitin
and its derivatives. Bioprocess and Biosystems Engineering, 20
(6): 505-512.
[13] T. Chandy, C. P. Sharma, (1990) Chitosan as a Biomaterial.
Biomaterials, Artificial Cells and Artificial Organs. 18(1): 1-24.
[14] P. K. Dutta, J. Dutta, M. C. Chattopadhyaya, V. S. Tripathi,
(2004) Chitin and Chitosan: Novel Biomaterials Waiting for Fu-
ture Development. Journal of Polymer Materials. 21(3):
321-333.
[15] S. Mallapragada, B. Narasimhan, (2006) Handbook of Biode-
gradable Polymeric Materials and their Applications. Vol. 1 & 2.
[16] A. K. Singla, M. Chawla, (2001) Chitosan: some pharmaceutical
and biological aspects-an update. Journal of Pharmacy and
Pharmacology. 53(8): 1047-1067.
[17] D. Mukherjee, (2001) Method for producing chitin or chitosan,
US patent 6310, 188B1, Oct. 30.
[18] T. A. Einhorn, J. M. Lane, A. H. Burstein, C. B. Kopman and V.
J. Vigorita, (1984) The healing of segmental defects induced by
demineralized bone matrix, J. of Bone and Joint Surgery, 66-A:
274-279.
[19] D. W. Wolff, V. M. Goldberg and S. Stevenson (1994) Histo-
morphometric analysis of the repair of a segmental diaphyseal
defect with ceramic and titanium fiber metal implants: effect of
bone marrow. J. of Orthopaedic Research. 12(3): 439-446.
[20] D. P. Mukherjee, A. S. Tunkle, R. A. Roberts, A. Clavenna, S.
Rogers, D. Smith, (2003) An Animal Evaluation of a Paste of
Chitosan Glutamate and Hydroxyapatite as a Synthetic Bone
Graft Material. Journal of Biomedical Material Research, Part B.
67B: 603-609.
[21] T. L. N. Jarvinen, P. Kannus, H. Sievanen, P. Jolma, A. Hei-
nonen, M. Jarvinen, (1998) Randomized controlled study of ef-
fects of sudden impact loading on rat femur. Journal of Bone
and Mineral Research. 13(9): 1475-1482.
[22] Y. Umemura, T. Ishiko, T. Yamauchi, M. Kurono, S. Mashiko,
(1997) Five jumps per day increase bone mass and breaking
force in rats. Journal of Bone and Mineral Research. 12(9):
1480-1485.