TP508 is a twenty-three-amino-acid peptide that represents a portion of the
receptor-binding domain of human thrombin, the naturally occurring molecule in
the body that is responsible for blood-clotting and the initiation of many of
the cellular events that are responsible for tissue repair. TP508, or drug
products formulated from it, such as Chrysalin (OrthoLogic, Tempe, Arizona),
can be used to mimic part of the thrombin response without stimulating the
events associated with blood-clotting. These drugs can therefore accelerate
the normal cascade of healing events. In this review, we will provide a brief
summary of key cellular studies of TP508, preclinical animal studies, and the
preliminary clinical results concerning diabetic ulcers and
fracture-healing.
Cellular Effects of TP508
TP508 was originally identified as the binding domain of human thrombin
that is responsible for the binding of thrombin to high-affinity receptors on
fibroblasts1. In
these studies, it was shown that TP508 was able to compete with thrombin for
specific binding to a subset of thrombin receptors and to initiate
proliferative signals. Since TP508 has no proteolytic activity, it was thought
that this peptide may be exerting its effects through nonproteolytic
interaction with one of the known protease-activated receptors (PAR-1, PAR-3,
or PAR-4)2 or
through a separate non-protease-activated receptor (NPAR).
To determine if TP508-generated signals were distinct from thrombin
activation of PAR-1 or other PARs activated by thrombin, Sower et al. studied
the effects of TP508 on fibroblast gene expression with use of differential
display polymerase chain reaction (PCR) and compared the effects of TP508 to
both thrombin and the PAR-1-activating
peptide3. These
studies revealed that TP508 was responsible for the up-regulation of a number
of genes, including annexin V, which were not up-regulated by thrombin or the
PAR-1-receptor-activating peptide;
serine-phenylalanine-leucine-leucine-arginine-asparagineproline (SFLLRNP).
These results supported the hypothesis that TP508 acts through a pathway
involving a receptor mechanism distinct from activation of PAR-1 or other
PARs. Furthermore, since annexin V is an inhibitor of protein kinase C (PKC),
the data suggested that TP508 in these cells might attenuate downstream
PAR-1-mediated signals that involve PKC activation.
Further evidence that TP508 was not working through PARs was observed when
the chemotactic effects of TP508 on freshly isolated human polymorphonuclear
(PMN) leukocyte were analyzed. Early studies showed that chemotaxis of PMN by
thrombin was independent of the proteolytic activity of thrombin and that
proteolytically inactivated thrombin stimulated PMN
chemotaxis4. Studies
by Ramakrishnan and Carney and Falk et
al.5,6
showed that TP508 at concentrations as low as 10-8 M stimulated PMN
chemotaxis in modified Boyden chamber (NeuroProbe, Bethesda, Maryland)
trans-well migration assays (Fig.
1), while the fourteen-amino-acid PAR-1 activator had little if
any effect. Interestingly, studies done at about the same time showed that
human PMNs do not express
PAR-17. Therefore,
the effects of TP508 on these cells could not be attributed to non-proteolytic
activation of the PAR-1 receptor.
In addition to the studies that showed that TP508 has chemotactic effects
on PMNs, recent studies by Naldini et
al.8 showed that
treatment of peripheral blood mononuclear cells (PBMC) with TP508 caused a
dose-dependent increase in the production of interleukin-1ß (IL-1 ß)
and interleukin-2 (IL-2) with doses ranging from 0.4 to 43 µM. This effect
was not elicited by a scrambled peptide, indicating the specificity of TP508
and the likelihood that this effect was receptor mediated. The authors had
previously shown that thrombin increased IL-1ß production in PBMCs by
activation of
PAR-19. Therefore,
to determine if TP508 affected the up-regulation of cytokines by activating
PAR-1 or through the activation of a separate NPAR, the authors used a PAR-1
antisense molecule to down-regulate PAR-1. Cells were then treated with SFLLRN
or TP508. As shown in Figure 2,
pretreatment of PBMCs with PAR-1 antisense (AS) inhibited the
SFLLRN-stimulated increase in IL-1ß production but not TP508-stimulated
IL-1ß production. Thus, as in previous studies, these results indicate
that TP508 activates cells by interacting with a specific NPAR distinct from
PAR-1 and other PARs.
Cellular Effects of TP508 Related to Bone and Cartilage Repair
One of the first effects of stimulatory molecules released at the site of
fracture or soft-tissue injury may be to attract inflammatory cells,
endothelial cells, or tissue-specific precursor cells to the site of injury.
As described above, TP508 was shown to be chemotactic for PMNs; therefore, we
hypothesized that TP508 may also recruit endothelial cells and/or precursor
cells to the site of injury. To test this hypothesis, Li et al. determined the
effects of TP508 and other molecules that have been shown to stimulate
fracture-healing on chemotaxis of human microvascular endothelial cells
(HMVEC) and
osteoblasts10. With
use of a directed migration assay, it was shown that TP508, along with
vascular endothelial growth factor (VEGF), bone morphogenetic protein-2
(BMP-2), and pleiotrophin (PTN), caused significant (p < 0.001) directional
chemotactic migration both of endothelial cells and osteoblasts.
Interestingly, the mean migration speed of osteoblasts was significantly
faster than controls (p = 0.05), as might be expected, when cells were
exposed to a gradient of PTN or BMP-2. In contrast, HMVEC migration was only
seen to be significantly faster than controls when treated with TP508 at 10
µg/mL (p < 0.01) or 100 µg/mL. In both cell types, the rate of
chemotaxis stimulated by TP508 was dose dependent. Thus, these studies confirm
that TP508 is chemotactic for both osteoblasts and HMVEC and suggest a
possible mode of action that involves angiogenesis as well as the recruitment
of progenitor cells.
The angiogenic potential of TP508 was also assessed with use of a
microvessel angiogenic sprouting
assay11. In these
studies, microvessel fragments were cultured and the number of sprouts and
length of each sprout were measured to determine effects of these agents on
early angiogenic events. Results showed that TP508 stimulated sprout outgrowth
to an extent similar to or greater than that achieved with VEGF, but neither
VEGF nor TP508 increased the number of sprouts relative to controls. The
TP508-stimulated outgrowth was inhibited in these studies, as was the control
outgrowth, by addition of tissue inhibitors of matrix metalloproteinase-1
(TIMP-1) or a soluble VEGF-receptor cocktail, suggesting that the TP508
effects are dependent on matrix metalloproteinase-2 MMP-2 and VEGF, although
PCR data showed that TP508 did not up-regulate MMP-2 or VEGF relative to
controls. Interestingly, thrombin and PAR agonists appeared to inhibit or have
no effect on sprout outgrowth. Thus, these studies demonstrated that TP508 is
a potent angiogenic factor and that it exerts this effect through a receptor
mechanism that is distinct from that of thrombin or PAR activation.
Cellular studies of TP508 show direct effects on chondrocyte proliferation
and differentiation. In studies by Schwartz et
al.12, primary
cultures of rat growth-plate chondrocytes were used to assess the effects of
TP508 on proteoglycan synthesis and cell proliferation as measured by both
cell number and [3H]-thymidine incorporation. The novelty of this
model is the ability to study differential effects of regulatory molecules on
resting and growth-zone cells of the chondrocytic lineage as an in vitro model
of endochondral repair. In these studies, TP508 showed a dose-dependent
increase in cell proliferation in cultures of growth-zone cells, but no
increase was observed in resting-zone cells. The increases in proliferation in
the growth-zone cells indicate that TP508 may target early events in the
endochondral process. In contrast, TP508 increased proteoglycan production, as
measured with assays of 35-sulfate incorporation, in resting-zone cells, with
no effect seen in growth-zone cells. Therefore, TP508 may increase
proliferation in less differentiated cells of the chondrocytic lineage while
stimulating matrix production at later stages of chondrocyte differentiation.
Similar to that seen in prior studies, a peptide synthesized with scrambled
sequence was inactive in this model, indicating specificity for TP508
interaction in stimulating these cellular events.
Together with the previous studies that showed that TP508 was chemotactic
for osteoblasts and endothelial cells and that it had distinct angiogenic
effects on vessel outgrowth, these studies suggested that TP508 may be an
effective modulator of tissue repair and regeneration in both dermal and
musculoskeletal tissues.
Dermal, Musculoskeletal, and Myocardial Tissue-Repair Models
A number of preclinical animal models have been used to determine the
potential efficacy of TP508 treatment for a variety of acute and chronic
wound-healing applications. These models included dermal surgical incisions,
full-thickness excisional wounds in normal and ischemic tissue, and
orthopaedic models for fracture repair, critical size defects, and distraction
osteogenesis.
Early dermal studies showed that a single topical application of TP508 to
incisional wounds on rats accelerated healing by approximately 80% as
evidenced by the breaking strength of the
tissue13. In
addition, histological analysis of these incisions at day seven showed more
type-1 collagen, less evidence of inflammation, and an increased number of
capillaries in the TP508-treated incisions relative to controls. Angiograms
confirmed that, in TP508-treated incisions, there were up to 65% more
functional vessels crossing the incision line. Thus, this peptide appeared to
enhance or accelerate both neovascularization and healing of these wounds.
Larger excisional dermal wounds (2 cm in diameter) in
normal14 and
ischemic15 rat skin
also showed accelerated healing when treated with a single topical application
of TP508. In both of these studies, TP508 treatment was associated with early
accumulation of inflammatory cells, early angiogenic responses that resulted
in larger and more functional blood vessels, and an earlier resolution of the
inflammatory phase of wound-healing.
These dermal studies suggested the possibility of a simple topical
treatment and application of TP508 in both acute and chronic wounds.
Additional studies looking at a radiation delayed wound-healing model also
showed enhanced healing with TP508
treatment16. This
study suggested that TP508 may affect the recruitment of circulating
inflammatory cells as well as cells that reside at the site of injury and that
the action of TP508 is distinct from that of platelet-derived growth factor or
transforming growth factor-beta.
The effect of a single injection of TP508 on fracture repair was assessed
in rats with use of an established closed femoral fracture model. In these
studies, a single injection of TP508 increased mechanical strength of the
femora as early as three weeks after
fracture17. These
fractures underwent histo-logical analysis and gene-array analysis of the
fracture callus. Histological results showed that TP508 treatment increased
the number of vessels per unit area three weeks after fracture by
approximately 40% and increased the area occupied by these vessels by
approximately 80%. Thus, TP508 appeared to cause an increase both in the
number of vessels and in the average size and functional development of the
vessels. Analysis of early fracture calluses by gene array showed that TP508
appeared to promote fracture repair through a mechanism that involved an
increased induction of a number of growth factors, of enhanced expression of
inflammatory mediators, and of angiogenesis-related genes.
It is interesting to note that TP508 treatment of larger bone defects also
resulted in regeneration of bone. In a study by Sheller et al., TP508 was
delivered to 0.5-cm radius defects and "critical-sized" 1.5-cm
ulna defects in rabbits in a bio-degradable controlled-release microsphere
carrier18. In both
cases, TP508 treatment caused an increase in mechanical strength. The larger
defects were also evaluated with use of three-dimensional synchrotron
tomography (Fig. 3). This
technique showed that the new bone in the TP508-treated defects had a less
porous surface and more open marrow space, suggesting that these bone repairs
were progressing to a normal state of remodeling. An interesting aspect of
this model is that, unlike many defect treatments in which a test agent is
delivered in a matrix that fills the defect gap, TP508 was delivered in a
volume of microspheres that was less than 5% of the gap volume. Thus, TP508
appears to be able to recruit cells and establish a matrix that provides an
osteoconductive as well as an osteoinductive environment.
The effects of TP508 do not appear to be limited to endochondral bone
formation. Studies by Li et al. with use of a rabbit distraction osteogenesis
model demonstrate that TP508 can enhance the rate of bone
consolidation19.
Interestingly, in this model, as in other models, the authors noted increased
blood-vessel development and less inflammation in the treated samples relative
to controls, suggesting that the inflammatory phase of healing was resolved
and that these regenerates had progressed into the proliferative and
maturation phases of healing.
It is interesting and perhaps important that, both in dermal and
musculoskeletal preclinical studies, a common series of events appears to be
associated with the acceleration of healing and regeneration of tissues.
Collectively, the studies indicate that TP508 causes early changes that allow
for inflammatory cells, endothelial cells, and perhaps progenitor cells to
move more rapidly into the injured tissue, resulting in revascularization and
an early resolution of the inflammatory phase of healing. This type of
response to a single application or injection of the TP508 peptide suggests
that it initiates a cascade of events that accelerates the natural healing and
tissue-regeneration process.
Because TP508 treatment results in increased revascularization in dermal
and orthopaedic tissues, and even in skin with surgically induced
ischemia13, it
seemed possible that this peptide might induce revascularization of ischemic
myocardium. As an initial test of this hypothesis, Coleman et al. used a
rabbit ameroid constrictor model to inhibit blood flow to the left ventricle
of the heart20.
After ischemia was established for three weeks, the ischemic region of the
heart muscle was injected with TP508 in saline solution (five locations) and
analyzed three weeks later using blinded histological assessment and digital
morphometry to determine capillary density. Looking at average densities from
regions of fourteen rabbit hearts, it was found that TP508 treatment resulted
in a 28% increase in capillary density relative to controls (p < 0.02).
These preliminary revascularization data confirm that injecting the peptide
into ischemic tissue results in an angiogenic response leading to
revascularization of the tissue. Since all tissue injury involves a disruption
of blood flow at and near the site of injury, this angiogenic effect further
suggests that TP508 may be clinically effective in stimulating repair of soft
and hard tissues.
Clinical Studies
TP508 (Chrysalin) has been studied in two therapeutic areas: diabetic
ulcer-healing and fracture repair. The following is summary information on the
phase 1/2 clinical studies in diabetic ulcer and fracture-healing.
Diabetic Foot Ulcers
Chronic diabetic ulcers of the lower extremities represent a major
health-care problem today, with over 850,000 diagnoses made in the United
States each
year21,22.
On the basis of the pre-clinical studies summarized above, it was hypothesized
that Chrysalin may be efficacious in diabetic ulcer-healing. Phase 1/2 trials
were conducted to evaluate the safety and efficacy of Chrysalin in the
treatment of diabetic ulcers. The study was a four-site, prospective,
randomized double-blind placebo-controlled clinical trial. Chrysalin was
topically applied twice weekly in saline solution in combination with standard
of care for up to twenty weeks. The study was a three-arm, sixty-subject trial
including all lower-extremity (below the knee) ulcers that had been present
for more than eight weeks and that were classified as meeting diabetic ulcer
grade (Wagner grade) I, II, or III without bone
involvement23. This
study was conducted as a part of the United States Food and Drug
Administration study of investigational new drug (IND) #56,811.
In this study, subjects were randomized to one of three treatment groups: 1
µg Chrysalin, 10 µg Chrysalin, or placebo. The wound bed was prepared by
sharp débridement, with treatment administered topically in a volume of
0.1 mL of saline solution, followed by application of Cutinova Foam
(Beiersdorf, Hamburg, Germany). The investigators analyzed adverse events,
chemical and hematological parameters (including hemoglobin A1c values) as
well as local wound reactions that were graded for erythema, edema, pain, and
overall condition. The primary efficacy end point was the incidence of ulcers
that progressed to complete ulcer closure during the twenty-week study.
Secondary end points included the time to 100% closure of the wounds and the
linear rate of wound
closure24. The
per-protocol population included patients with ulcers located distal to the
kneecap on the leg, ankle, and foot. There were fifteen subjects who received
placebo, eleven subjects who were treated with 1 µg Chrysalin, and fourteen
subjects who were treated with 10 µg Chrysalin. A further analysis of the
subpopulation with only foot ulcers was also performed (thirteen placebo;
twelve Chrysalin 1µg; and ten Chrysalin 10 µg).
From a safety perspective, laboratory values showed no significant changes
from baseline or for any of the treatment groups at any of the time points.
Chrysalin treatment did not result in any adverse events that were probably or
definitely drug related. No significant differences were found between the
groups with regard to the prevalence of infection or other adverse
effects.
In the per-protocol population, the prevalence of complete ulcer closure
was five of fifteen in the saline-solution placebo control group compared with
five of eleven in the 1-µg Chrysalin treatment group and eight of fourteen
in the 10-µg Chrysalin treatment group. In the subpopulation with foot
ulcers, complete healing of the ulcer was achieved in four of thirteen of the
placebo control patients compared with nine of twelve in the 1-µg Chrysalin
group and seven of ten in the 10-µg Chrysalin group. The median time to
wound closure in the placebo group was not reached, as only four of thirteen
foot ulcers healed in the twenty-week study period. The median time to closure
was seventy-two days in the 10-µg Chrysalin group (p = 0.033). The
wound-healing rate, calculated from digital images, increased approximately
80% in the limbs that were treated with 10-µg Chrysalin compared with the
limbs that were treated with placebo (p = 0.044).
Figure 4 shows an example of a
Chrysalin-treated ulcer that healed at eighty-seven days after initiation of
treatment.
In conclusion, the safety and efficacy rates seen in this study suggest
that topical application of Chrysalin may have considerable therapeutic value
for patients with diabetic foot ulcers. Current efforts are focused on
manufacturing a gel formulation of Chrysalin and initiating a design for
further human clinical trials.
Distal Radial Fracture Repair
There are approximately seven million fractures per year in the United
States, with distal radial fractures representing one-sixth of all fractures
seen in emergency
rooms25. Therefore,
there is a medical need for a product that is safe and that could accelerate
the rate of bone-healing, leading to earlier cortical union, functional bone
strength, and, ultimately, reduction in the required immobilization time. The
acceleration of the healing process would be beneficial to patients with wrist
fractures because patients who experience a shorter time of joint
immobilization should regain more complete hand function
earlier26. On the
basis of the preclinical studies summarized above, it was hypothesized that
Chrysalin might accelerate fracture-healing at the distal portion of the
radius.
A combined phase-1/2 multicenter, prospective, double-blind, randomized
placebo-controlled study was performed to study the safety and efficacy of two
different doses of Chrysalin compared with placebo in the healing of distal
radial fractures. The study was a three-arm, ninety-subject trial of displaced
and unstable distal radial fractures that were treated with percutaneous
pinning or external fixation. This study was conducted as a part of United
States Food and Drug Administration study of IND #59066.
On the day of surgery, subjects were randomly assigned to placebo, 10
µg/mL Chrysalin, or 100 µg/mL Chrysalin treatment groups. Fracture
treatment was performed with standard orthopaedic care at the discretion of
the clinical investigator, and all fractures were reduced and stabilized by
external fixation, or with percutaneous Kirschner wires and immobilized with a
cast, as described by Cassidy et
al.27.
Intra-fracture percutaneous injection of Chrysalin was performed under
fluoroscopic guidance. Subjects were evaluated radiographically weekly for
eight weeks and then at three and six months. Pain was assessed with use of a
visual analog score (VAS), and the patients were assessed with regard to range
of motion and grip strength and were given the Disabilities of the Arm,
Shoulder and Hand (DASH) questionnaire.
Blinded radiographic assessments of fracture-healing were performed by the
clinical investigators, an orthopaedic surgeon (hand specialist), and a
musculoskeletal radiologist. Radiographic criteria for healing included an
assessment of cortical bridging and trabecular bridging and an overall
radiographic assessment of fracture-healing (i.e., healed or not healed).
Safety measurements, including the prevalence of adverse events, serology
results, urinalysis, and postoperative complications, did not differ among the
treatment groups. The subjects' rating of pain by VAS showed no significant
differences among treatment groups. The Kaplan-Meier analysis of the
radiographic evaluation revealed that the time to healing was 23.5 days in
subjects treated with 10 µg Chrysalin compared with thirty-one days in
subjects treated with placebo (log-rank test, p = 0.0067;
Fig. 5). Throughout the healing
curve, the group treated with 10 µg Chrysalin showed earlier
time-to-healing. The Cox proportional regression model, adjusted for
covariates, revealed that patients treated with 10 µg Chrysalin had a
hazard ratio of 1.93 relative to the placebo, p = 0.021. As an example, the
time when 50% of the subjects are healed occurs 10.6 days faster for subjects
treated with the 10-µg dose than for those treated with placebo and the
time when 70% are healed occurs 13.1 days faster. The investigators'
assessment of trabecular bridging and overall radiographic healing was not
significantly different among the three treatment groups.
The blinded radiographic analysis performed by the radiologist determined
that trabecular bridging was significantly accelerated in favor of the
10-µg Chrysalin dose, resulting in a hazard ratio of 1.80 relative to
placebo (Cox proportional-hazard regression model adjusted for covariates; p =
0.04). The independent hand surgeon did not find significant differences among
treatment groups for any of the radiographic measures. The median times to
removal of immobilization were not significantly different among treatment
groups (forty-one, forty, and forty-one days for the placebo group, 10 µg
Chrysalin group, and 100-µg Chrysalin group, respectively). There were no
significant differences among treatment groups with regard to range of motion,
grip strength, and VAS or DASH scores. Stratification of extra-articular and
intra-articular fractures showed a strong trend to earlier removal of
immobilization for extra-articular fractures, indicative of a clinical benefit
to subjects.
In conclusion, the results of this phase-1 and phase-2 clinical trial
demonstrated the safety and preliminary efficacy of Chrysalin at the 10-µg
dose in the accelerating healing of distal radial fractures, resulting in
plans to further evaluate the Chrysalin peptide in human fracture repair.
Currently, there are ongoing phase-2b and 3 clinical trials evaluating the
safety and efficacy of Chrysalin in the acceleration of fracture-healing at
the distal portion of the radius.
The discovery that a synthetic peptide portion of native thrombin could
stimulate cell chemotaxis (migration), proliferation, and other events leading
to the revascularization and repair of dermal and musculoskeletal tissues
clearly has potential implications for therapeutic application of TP508 for
tissue repair. In addition, the results reviewed here provide a better
understanding of the apparent role of thrombin and thrombin fragments in
initiating and regulating key steps in wound repair and regeneration. It is
noteworthy that in both preclinical and clinical studies, TP508 appears to be
effective with either a single local application or, for diabetic foot ulcers,
only two applications per week. This supports the concept that this peptide
drug initiates a cascade of molecular events that result in acceleration of
the tissue-repair process. ?