ELSEVIER Bone and Mineral 27 (1994) 183-192

Basic fibroblast growth factor (BFGF)

immunoreactivity as a possible link between head
injury and impaired bone fracture healing

Renate Wildburger *a, Neven Zarkovic b'c, Gerd Egger d,

Walter Petek e, Kamelija Zarkovic f, H.P. Hofer a
aDepartment of Traumatology, Medical School, University of Graz, Auenbruggerplatz 15, A-8036 Graz,
Graz, Austria
bRudjer Boskovic Institute, Zagreb, Croatia
Clnstitute of Biochemistry, University of Graz, Graz, Austria
dInstitute of General and Experimental Pathology, University of Graz, Graz, Austria
9 of Clinical Chemistry, Medical School, University of Graz, Graz, Austria
flnstitute of Neuropathology, Medical School, University of Zagreb, Zagreb, Croatia

Received 15 March 1994; revision received 8 July 1994; accepted 21 September 1994

Abstract

Healing of fractures of long bones or large joints is often accelerated in patients with severe
traumatic brain injury (TBI). However, in these patients an early fracture healing is accompa-
nied by hypertrophic callus formation or heterotopic ossifications which might even result in
an ankylosis of the affected joints. It seems that enhanced osteogenesis in patients suffering
from TBI could be caused by some humeral factors, since the sera of these patients strongly
promote the growth of osteoblast cells in vitro. However, humeral growth promoting factors
which could perhaps induce enhanced osteogenesis are not yet identified. Hence, the aim of
this study was to analyse if basic fibroblast growth factor (bFGF) could be related to the phe-
nomenon of enhanced osteogenesis, since b F G F stimulates the growth of osteoblasts in vitro
and could be found both in the brain and the bone tissue. For that purpose the values of
bFGF immunoreactivity were determined in the sera of patients with TBI and bone fractures
(n = 8) as well as in the sera of patients with either TBI alone (n = 10) or bone fractures alone
(n = 7), during a period of three months after injury. Quantification of the b F G F immune-
reactivity was done using the ELISA based on monoclonal antibodies raised against the

Abbreviations: bFGF, basic Fibroblast Growth Factor; TBI, Traumatic Brain Injury.
* Corresponding author.

0169-6009/94/$07.00 9 1994 Elsevier Science Ireland Ltd. All fights reserved

SSDI 0169-6009(94)00757-8
184 R. Wildburger et al./ Bone Miner. 27 (1994) 183-192

recombinant human bFGF. The bFGF immunoreactivity values obtained were also compared
with the values determined in the sera of normal, healthy persons (n = 9). In the group of pa-
tients with bone fractures alone only a transient increase of bFGF immunoreactivity (three-
fold above the normal values) was observed in the second week after injury. A similar increase
of the values of bFGF immunoreactivity was also determined in the sera of patients with TBI
only, but it lasted longer (from the 1st until the 7th to 8th week after injury). In the case of
patients with TBI and bone fractures a specific pattern of post-traumatic dynamic change of
the values of serum bFGF immunoreactivity was observed. Namely, the increase of bFGF im-
munoreactivity (up to seven-fold above the normal values) was determined even during the
first week after injury. Afterwards, periods of high values of bFGF immunoreactivity
observed during the 2nd, 4th and the 7-10th weeks after injury were interrupted by sudden
decreases even to the normal values (during the 3rd and the 5-6th week after injury). Thus,
post-traumatic increase of the serum bFGF immunoreactivity was induced both by bone frac-
tures and TBI, while the unusual pattern of its changes could be related to the phenomenon
of enhanced osteogenesis in the patients with brain injury.

Keywords: Basic fibroblast growth factor; Enhanced osteogenesis; Heterotop!c ossification;

Bone fracture healing; Head injury

1. Introduction

Excessive callus formation in patients with traumatic brain injury (TBI) and frac-
tures of long bones and large joints has often been described and is well documented
[1-4]. However, the mechanisms underlying this phenomenon are still unknown.
Since the sera of patients with TBI strongly stimulate the growth of osteoblastic cells
in vitro, it appears that TBI is accompanied by an increased amount of certain
growth stimulating factors in the serum, which could consequently induce enhanced
osteogenesis in patients with TBI [5]. Among the growth factors which could,
perhaps, induce enhanced osteogenesis in patients with TBI, basic fibroblast growth
factor (bFGF) seems to be the most interesting.
b F G F was purified from bovine pituitary gland tissue by Gospodarowicz in 1975
[61. When isolated from natural sources (different tissues), b F G F usually has an
apparent molecular weight (MW) of 14-16 kDa, while in serum it seems to form
high MW (180-250 kDa) conjugates with the serum components which show the im-
munoreactivity of the free b F G F [7-10]. One of the best known effects of b F G F is
the induction of angiogenesis and stimulation of wound healing, while it also has a
strong influence on proliferation and differentiation of a wide variety of cells, such
as fibroblasts, chondroblasts, osteoblasts, adrenocortical cells, etc. [7,11-13].
Although the physiological role of b F G F in bone metabolism still remains to be
defined, due to its biological effects b F G F could be an important regulator of bone
growth and fracture healing. It was shown that b F G F stimulates proliferation and
differentiation of costal chondrocytes in vitro [14] as well as chondrocyte prolifera-
tion and cartilage healing in vivo [15]. Furthermore, osteoblasts themselves syn-
thesize b F G F [7,13,16], while transient treatment with b F G F in vitro increases
 R. Wildburger et al. / Bone Miner. 27 (1994) 183-192 185

collagen synthesis in fetal rat osteoblasts [17]. The hypothesis that bFGF stored in
the bone matrix has an influence on bone repair was proved [18], at least in part,
by the finding of dramatic hyperostosis in rat femurs after repeated intravenous ad-
ministration of bFGF [19]. On the other hand, a significant transient increase of
bFGF immunoreactivity in the brain areas damaged by a traumatic brain injury or
stroke was confirmed [20-24]. Hence, bFGF could possibly also be released from
the damaged brain tissue. If so, the increased level of bFGF immunoreactivity after
TBI should be also found in sera of the injured patients. On the other hand, it seems
likely that even bone fracture healing should be accompanied by an increased release
of bFGF, especially in the case of enhanced osteogenesis which develops in patients
with TBI. The aim of this study was to test these possibilities and to determine
whether bFGF could be one of the unidentified humoral factors which cause
enhanced osteogenesis in patients with TBI.

2. Materials and methods

2.1. Subjects
Blood specimens were obtained from thirty-four subjects who were assigned to
one of the following four groups.
Group I included seven patients with fractures of long bones or large joints only.
Five patients were male and two female. The range of age was 18-64 years (the mean
age was 31.2 years). They were mostly injured in traffic accidents. Six of the patients
had fractures of the femur, one on both sides. One patient had a dislocation fracture
of the elbow, while another one had fractures of both acetabula. The consolidation
of the femur fractures and the fractures of the pelvis lasted for about 12 weeks on
average. No pathologically increased callus formation and no heterotopic ossifica-
tions were observed in these patients.
Group II included ten patients with TBI only, eight male and two female. The
mean age was 25.2 years. Nine of the patients had a traffic accident, one had a fall.
All had either intracerebral or subdural hematoma or both. Four patients had to be
craniotomied. The average coma lasted 11 days, while one patient died after 37 days
of coma. Five of the patients had transitory neurological deficits such as, for exam-
ple, hemiparesis. All of the surviving patients recovered.
Group III was comprised of eight patients with TBI and fractures of long bones
or large joints. Six of the patients were male, and two female. The mean age was 19.2
years. With the exception of one patient the cause of injury was a traffic accident.
Five patients had intracerebral and/or intracranial hematomas, three had diffuse
edema only, while all were in coma from 3-20 days (average of 9 days). Only one
of the patients had to be craniotomied, while in three patients facial fractures re-
quired surgery. Six of the patients had cranial or facial fractures. All of them had
fractures of the femur, one on both sides. Two patients also had fractures of the
pelvis.
The first callus formation occurred between the second and the fourth week after
injury. It was extended in every patient. One patient had heterotopic ossifications
around the elbow and on the lateral femur condyl after Steinmann nail extension,
186 R. Wildburger et al. / Bone Miner. 27 (1994) 183-192

and one patient had heterotopic ossifications in the small pelvis after a fracture of
the iliac wing.
The patients of all three groups received subcutaneous injections of heparin
(2500-5000 I.U., depending on the body weight) daily during their stay in hospital
(on the average for 4-6 weeks). Antibiotics were mostly given as a prophylaxis,
analgetics were used occasionally, while steroids were not applied.
Group IV included nine normal, healthy subjects, five male and four female (mean
age 25.9 years).

2.2. Sera samples

Sera were prepared from the blood samples (5-6 ml) taken from the patients, as
well as from the healthy subjects, between 08:00 h and 10:00 h, at regular time inter-
vals after traumatic event. After the injury, blood was taken weekly, and after the
release of patients from the hospital at intervals of 2 weeks over a period of 3 months
in total. The first samples were taken in the first 2-4 days after the accident. Blood
samples were centrifuged within 1 h at +4~ for 10 min at 500 times gravity. Super-
natants were collected, and their aliquots were immediately frozen and stored at
-30~

2.3. Determination of bFGF immunoreactivity

The determination of bFGF immunoreactivity was done with a human bFGF im-
munoassay (QuantikineTM HS Assay, R & D Systems, Inc., Minneapolis, MN,
USA). This assay employs the quantitative 'sandwich' enzyme immunoassay tech-
nique (ELISA) using alkaline phosphatase as an amplification system. The assay is
based on murine monoclonal antibodies raised against the recombinant human
bFGF. According to the declared specification of the Quantikine TM HS Assay,
intra-assay precision is > 95% for the sera containing high bFGF immunoreactivity
(> 15 pg/ml) and >87% for the sera containing low bFGF immunoreactivity (< 1
pg/ml). Inter-assay precision is declared to be > 89 and 77% for the sera containing
high and low bFGF immunoreactivity, respectively. The declaration of R & D Sys-
tems regarding the accuracy of the kit were in part confirmed by testing the serial
dilution of the recombinant human bFGF standards enclosed with the kit, as well
as by testing the pool of normal human sera further enriched with the recombinant
human bFGF prior to the analyses of the sera used in the study. Each of the samples
was analysed twice and the means of the results obtained were used for evaluation.

2.4. Statistics
Statistical evaluation of the results obtained was done by the Mann-Whitney test,
and the values of P < 0.05 were considered significant.

3. Results

The mean values of bFGF immunoreactivity in the sera of injured patients deter-
mined during the 3-month period after injury are presented in relationship to the
kind of injury and comparison to the normal values on Fig. 1.
 R. Wildburger et al. / Bone Miner. 27 (1994) 183-192 187

25

20
T
I 0
INF
00

15 II

0f

O)
O. 0#

ii #q

CO
i, I0

.O

E
2
u)
10

NyI
"""'" N- - - "'''"
L "N" " , , /l " N

I
I
III III
i \
. 1 t ~ "~"
1
/ f, 9

Normal values (Mean+SEMI

I

2.35+_0.64 (pglml)

o I I I I I I I I
1 2 3 4 56 78 910 11 14
Weeks After Injury

Fig. l. Dynamic change of b F G F imrnunoreactivity values in the sera of injured patients. The mean values
and S.E.M. obtained from each group are presented (head injury + fractures, n = 8; head injury only,
n = 10; fractures only, n = 7; normal subjects, n = 9). S.E.M. for the first week ranged from 34-41% of
the respective mean. *N, significant difference when compared to values obtained from healthy subjects
(P < 0.05 according to Mann-Whitney test) *F, significant difference when compared to values obtained
from patients with bone fractures only (P < 0.05 according to Mann-Whitney test). - - [] - - fractures
only, - - - O - - - head injury only, 9 , head injury + fractures.
188 R. Wildburger et al./ Bone Miner. 27 (1994) 183-192

In group I, only a transient increase of bFGF immunoreactivity compared to nor-

mals (on the average three-fold above the normal values) was observed in the second
week after injury (P < 0.05). Afterwards, the values of the growth factor decreased
to the normal until the l l-14th week when a second increase was realized.
A similar increase of the values of bFGF immunoreactivity was also determined
in group II, but it appeared earlier and lasted longer (from the 1st until the 7-8th
week after injury). Although it seems that TBI was followed by a more pronounced
increase of the serum bFGF immunoreactivity than bone fracture, the values obtain-
ed were significantly higher in the former group only in the 1st and the 3rd post-
traumatic week (P < 0.05) when compared to normals and patients with fractures
only.
As in the case of patients with bone fractures only, in the sera of patients with TBI
only, after reaching the maximal increase the values of bFGF immunoreactivity also
decreased to normal. However, the values determined in the sera of patients with
TBI were significantly higher than normal values during the 2 months (from the 1st
until the 7-8th week after injury, for the first 3 weeks P < 0.002, afterwards P <
0.02 or 0.05).
In the sera of patients of group III a specific pattern of post-traumatic dynamic
change of the values of serum bFGF immunoreactivity was observed. Namely, the
increase of the growth factor concentration (on the average up to seven-fold above
the normal values) was determined even during the first week after injury (P <
0.002). Afterwards, periods of high bFGF immunoreactivity observed during the
2nd, 4th and the 7th-8th week after injury were interrupted by sudden decreases
even to normal values (during the 3rd and the 5th-6th week after injury). Further-
more, after the peak observed during the 7th-8th week after injury the values of
bFGF immunoreactivity in the sera of patients with combined injury gradually
decreased and reached the normal values at the end of the period that followed
(9-10th week after injury).
During the intervals of maximal increase the values of bFGF in group III were
always significantly higher than in group I (P < 0.05), except in the second week
(P > 0.1), when bFGF showed the maximal increase also in the sera of patients with
fractures only. As opposed to this, significant differences were never observed be-
tween group II and group III (P > 0.05).
Furthermore, since there were individual variations of the values of bFGF im-
munoreactivity observed for the patients within the groups, a comparison of the
maximal values determined in the sera of each of the patients and healthy persons
was carried out (Table 1). While the lowest individual values of the bFGF maximum
determined for the patients of group I or II were within the range of values determin-
ed for the healthy persons, in the case of group II even the lowest one was more than
three-fold higher than the highest normal value. Furthermore, for all three groups
of injured patients, maximal values of serum bFGF immunoreactivity were
significantly higher than the normal values (P < 0.002). They were also higher, in
the sera of patients of group III, than the values determined in the sera of patients
of group I (P = 0.002), as well as when compared with the values obtained for the
sera of group II (P < 0.05). However, there was no difference of the maximal bFGF
 R. Wildburger et al. / Bone Miner. 27 (1994) 183-192 189

Table 1
Comparison of the values a of bFGF immunoreactivity determined in the sera of healthy subjects and
injured patients

Normal subjects TBI only Fracture only TBI + fracture

(n = 9) (n = 10) (n = 7) (n = 8)

Individual values of 0.7 5.0 5.6 17.8

serum bFGF 1.1 8.3 9.2 20.2
(pg/n~) 1.2 12.2 9.3 26.5
1.8 13.3 12.7 37.3
2.2 16.5 16.0 37.9
2.5 17.3 18.2 38.0
2.6 19.9 23.2 39.0
3.6 30.5 41.8
5.5 35.0
39.0

Mean -~- S.E.M. 2.35 • 0.64 19.70 • 3.82 13.46 • 2.48 32.29 • 3.56
PN < 0.002* PN < 0.002 PN < 0.002
PF > 0.1 PF = 0.002
PTBI < 0.05
aThe maximal values obtained for each patient and the mean values • S.E.M. for each of the groups are
presented. The post-traumatic period of 14 weeks was monitored, while the maximal values of bFGF im-
munoreactivity for each of the patients was measured during the first month after injury.
*Levels of significance (according to the Mann-Whitney test) if the values are compared with those ob-
tained from normal persons (PN), patients with bone fractures only (PF) and patients with head injury
only (PTsl).

immunoreactivity determined in the sera of patients with TBI only and bone frac-
tures only (P > 0.1). Thus, it appears that combined TBI and bone fractures induced
a specific pattern of dynamic change of bFGF serum concentration during the post-
traumatic period of 3 months, and the maximal values reached were higher than for
the patients with bone fractures or TBI alone.

4. Discussion

There are numerous data showing the involvement of bFGF in the bone metabo-
lism and pathophysiology of the brain damage [7,13,16,20-25]. The results obtained
show that the phenomenon of enhanced osteogenesis in patients with TBI is accom-
panied by a particular pattern of post-traumatic dynamic change of the serum values
of bFGF immunoreactivity. However, it is not certain if bFGF could be released
from the damaged tissue (brain) into the blood and modify the growth of the cells
at the distant 'target' organ (bone). Furthermore, determination of the bFGF values
in the serum (as well as its purification) is difficult, since the growth factor is
presented in very low concentrations and in different high MW (180-250 kDa) forms
which have the immunoreactivity of the free lower MW (14-16 kDa) bFGF [8-11].
Although the biochemical nature of high MW forms of bFGF is not entirely
190 R. Wildburger et al./Bone Miner. 27 (1994) 183-192

clarified, it is known that at least some of them are conjugates of the growth factor
with the serum proteins which act as cleavage enzymes [91. Furthermore, one of the
characteristic features of b F G F is its affinity to bind to heparin [71, while the
possibility of binding of this growth factor to the other serum proteins also seems
probable. Preliminary data (not presented) obtained by membrane ultrafiltration of
the sera samples used in the study (pooled according to the groups) indicate that the
MW of the b F G F immunoreactivity in the sera was higher than 30 kDa. Hence, it
seems that the b F G F immunoreactivity was represented by the higher MW forms
of the growth factor but not by the free bFGF. Heparin affinity chromatography
further indicated high (more than 90%) affinity of b F G F immunoreactivity to bind
to heparin. It seems that the type of traumatic event did not have an influence on
the mentioned features of the serum b F G F immunoreactivity. However, further
evaluation of the biochemical characteristics of the serum b F G F immunoreactivity
is necessary and should be accompanied by the analysis of its biological activities,
at least in vitro. On the other hand, since there was not a major difference of the
conservative treatment (including heparin treatment during the initial post-
traumatic phase) between the different groups of injured patients, it does not seem
that the medicaments used could play a major role regarding the amount of b F G F
immunoreactivity in the sera of the patients.
The results obtained showed that the values of b F G F immunoreactivity were
significantly increased above normal values in the sera of injured patients, especially
during the initial post-traumatic period.
Furthermore, b F G F immunoreactivity was increased in the sera of patients of
group III in comparison to the values obtained for the patients of group I, but not
if compared with those obtained for the patients of group II, for any particular time
period followed. Thus, it seems that the increase of b F G F immunoreactivity in the
sera of patients with combined injury was mainly a consequence of brain damage,
at least during the initial post-traumatic time. However, both the values of the max-
imal increase and the particular pattern of post-traumatic change of the b F G F im-
munoreactivity observed for patients with TBI and bone fractures appeared to be
related to the phenomenon of enhanced osteogenesis. The initial peak of the growth
factor concentration in the sera of these patients probably is caused by the release
of b F G F from the damaged brain and bone tissues (including resorption of hemato-
ma). Further, increased values of b F G F immunoreactivity afterwards seem to be
related to the hypertrophic callus formation. The second increase of b F G F im-
munoreactivity was observed during the fourth post-traumatic week and that was
the time when for all of the patients the symptoms of enhanced osteogenesis were
manifested (heterotopic ossifications or distinct callus formation). Later on, the
third increase of the growth factor concentration could be a consequence of abun-
dant callus formation, particularly during the 7-8th week when the fractures con-
solidated (a slight increase was also noticed in the patients with bone fractures only,
but later, i.e. during the 11-14th week after injury). To prove that and further clarify
the humoral mechanisms causing enhanced osteogenesis in patients with severe brain
injury, serum b F G F immunoreactivity of the bone and of the brain origin should
be distinguished, and the involvement of the organs (particularly of the endocrine
system) should be monitored.
 R. Wildburger et al./ Bone Miner. 27 (1994) 183-192 191

Acknowledgements
This study was supported by the AO/ASIF Foundation, Switzerland, project
W25/93. The authors wish to thank Ing. A. Meinitzer, Laboratory of Clinical
Chemistry, Medical School, University of Graz for perfect technical assistance.

References

[I] Perkins R. Callus formation and the rate of healing of femoral fractures in patients with severe head
injuries. J Bone Joint Surg 1987;69-B(4):521-524.
[2] Smith R. Head injury, fracture healing and callus. J Bone Joint Surg 1987;69-B(4):518-520.
[3] Spencer RF. The effect of head injury on fracture healing. J Bone Joint Surg 1987;69-B(4):525-528.
[4] Wildburger R, Schwr F, Passler JM, Hofer HP. Heterotope Ossifikationen bei einem Pa-
tienten mit schwerem Sch/idd-Hirn Trauma. Zentr bl Chir 1992;117:515-519.
[5] Bidner SM, Rubins IM, Desjardins JV, Zukor DJ, Goltzman D. Evidence for a humoral mechanism
for enhanced osteogenesis after head injury. J Bone Joint Surg 1990;72-A(8):1144-1149.
[6] Gospodarowicz D. Purification of a fibroblast growth factor from bovine pituitary. J Biol Chem
1975;250:25! 5-2520.
[7] Gospodarowicz D, Ferrara N, Schweigerer L, Neufeld G. Structural characterization and biologi-
cal functions of fibroblast growth factor. Endocrinol Rev 1987;8:95-114.
[8] Baird A, Esch F, Morm6de P, Ueno N, Ling N, B6hlent P, Ying SY, Wehrenberg WB, Guillemin
R. Molecular characterization of fibroblast growth factor: distribution and biological activities in
various tissues. Recent Prog Hor Res 1986;42:143-205.
[9] Gauthier T, Maftouh M, Picard C. Rapid enzymatic degradation of [125I](Tyr 10) FGF (1-10) by
serum in vitro and Involvementin the determination of circulating FGF by RIA. Biochem Biophys
Res Commun 1987;145(2):775-781.
[10] Morm6de P, Baird A, Pigeon P. Immunoreactive fibroblast growth factor (FGF) in rat tissues: mo-
lecular weight forms and the effects of hypophysectomy. Biochem Biophys Res Commun
1985;128(3):1108-1113
[11] Gospodarowicz D, Cheng J, Lui GM, Baird A, B6hlent P. Isolation of brain fibroblast growth fac-
tor by heparin-Sepharose affinity chromatography: identity with pituitary fibroblast growth factor.
Proc Nati Acad Sci USA 1984;81:6963-6967.
[12] Rifkin DB, Moscatelli D. Recent Developments in the Cell Biology of Basic fibroblast growth Fac-
tor. J Cell Biol 1989;109:1-6.
[13] Rodan SB, Rodan GA. Fibroblast growth factor and platdet derived growth factor. In: Maxine
Gowen, ed. Cytokines and Bone Metabolism. Boca Raton, Florida; CRC Press, 1992;115-145.
[ 14] Kato Y, Gospodarowicz D. Sulfated proteoglycan synthesis by rabbit costal chondrocytes growing
in the presence and absence of fibroblast growth factor. J Cell Biol 1985;100:477.
[15] Cuevas P, Burgos J, Baird A. Basic fibroblast growth factor (FGF) promotes cartilage repair in
vivo. Biochem Biophys Res Commun 1988;156(2):611-618.
[16] Hauschka PV, Mavrakos AE, Iafrati MD, Doleman SE, Klagsbrun M. Growth factors in bone ma-
trix. Isolation of multiple types by affinity chromatography on Heparin-Sepharose. J Biol Chem
1986;261:12665-12674.
[17] Canalis E, Centrella M, McCarthy T. Effects of basic fibroblast growth factor on bone formation
in vitro. J Clin Invest 1988;81:1572-1577.
[18] Canalis E. Bone-related growth factors. Triangle 1988;27(1/2):11-19.
[19] Mazue G, Bertolero F, Garofano L, Brughera M, Carminati P. Experience with the preclinical
assessment of basic fibroblast growth factor (bFGF). Toxicol Lctt 1992;329(38):64-65.
[20] Finklestein SP, Caday CG, Kano M, Berlove DJ, Hsu CY, Moskowitz M, Klagsbrun M. Growth
factor expression after stroke. Stroke 1990;21(11 Suppl.)III:122-124.
[21] Frautschy SA, Walicke PA, Baird A. Localization of basic fibroblast growth factor and its mRNA
after CNS injury. Brain Res 1991;12, 553(2):291-299.
192 R. Wildburger et al. / Bone Miner. 27 (1994) 183-192

[22] Logan A, Frautschy SA, Gonzalez AM, Baird A. A time course for the focal elevation of synthesis
of basic fibroblast growth factor and one of its high-affinity receptors (fig) following a localized
cortical brain injury. J Neurosci 1992;12(10):3828-3837.
[23] Riva MA, Gale K, Mocchetti I. Basic fibroblast growth factor mRNA increases in specific brain
regions following convulsive seizures. Brain Res Mol Brain Res 1992;15(3-4):311-318.
[24] Takami K, Iwane M, Kiyota Y, Miyamoto M, Tsukuda R, Shiosaka S. Increase of basic fibroblast
growth factor immunoreactivity and its mRNA level in rat brain following transient forebrain
ischemia. Exp Brain Res 1992;90(1):1-10.
[25] Globus RK, Plouet J, Gospodarowic ZD. Cultured bovine bone cells synthesize basic fibroblast
growth factor and store it in their extracellular matrix. Endocrinology 1989;124(3):1539-1547.