Shock the Ulcer to Life: Do Heat Shock Proteins Trump Growth Factors in Wound Healing?

These interesting data from teams at USC and Xijing Hospital / Fourth Military Medical University based on work just published in JCI.


Scientists say the widely held belief that growth factors represent the best treatment strategy for promoting healing in chronic wounds may be misplaced. They have found that a heat shock protein secreted by both human dermal fibroblasts (HDFs) and human keratinocytes (HKs) as a result of injury is key to removing a natural block to the migration of reparative dermal cells into the wound, effectively speeding cellular infiltration and wound repair and revascularization.

Reporting their findings in the Journal of Clinical Investigation, researchers at the University of Southern California Keck School of Medicine and the Fourth Military Medical University’s Xijing Hospital

describe a series of experiments demonstrating that single topical administration of a functional Hsp90α fragment overrides a natural TGF-β3-mediated barrier that would otherwise prevent HDF and human microvascular endothelial cells (HDMEC) from migrating into wounds until levels of the Hsp90α protein reach a certain threshold. The result is that wounds heal much faster, even under hyperglycemic conditions present in diseases such as diabetes.

The Keck School of Medicine’s Wei Li, M.D.,and colleagues, say the results also explain why the only FDA-approved growth-factor based wound-healing product, becaplermin, effectively speeds healing only in a minority of chronic wounds cases; it essentially can’t override the regulatory effects of TGF-betaβ on cell migration. Drs. Li and colleagues report their findings in a paper titled “A fragment of secreted Hsp90α carries properties that enable it to accelerate effectively both acute and diabetic wound healing in mice.”

Over the last 30 years or so much wound healing research has focused on the use of growth factors to promote tissue repair. Since the mid 1970s, over 30 growth factors have been the subject of extensive in vitro, preclinical, and clinical studies alone or in combinations.

Yet despite the fact that more than a dozen clinical trials with growth factor-based wound-healing candidates have been conducted since the first report of an EGF clinical trial back in 1989, only human recombinant PDGF-BB (Ortho-McNeil’s becaplermin gel called Regranex) has received marketing approval. Moreover, they add, studies conducted since the approval of becalpermin in 1997 indicate that the drug aids in the complete closure of just 15% of chronic wounds at best, while its extensive use is reportedly associated with an increased risk of cancer mortality.

Dr. Li’s investigations into the mechanisms involved in wound-healing have been built on the finding that that fetal bovine serum (FBS), which is widely used in studies of human skin cells and wound healing, isn’t equivalent in composition to human serum, the medium in which wound-healing factors and cells are bathed in vivo. The Keck researchers previously found that while FBS stimulates the migration of human epidermal and dermal cells to the wound site, human serum only promotes human keratinocyte (HK) migration and in fact halts HDF and HDMEC migration.

This discrepancy was due to the presence of TGF-β3 in human serum, which plays a regulatory role in controlling the migration of epidermal and dermal cells. While this acts to ensure fast and proper closure of the wound in vivo, they suggest it also means that “conventional growth factors such as PDGF-BB for dermal fibroblasts and VEGF-A for endothelial cells may not be able to do the job, as they were hoped, in human wounds, because of the co-presence of TGF-β3.”

The team then hypothesized that as a result of injury, other proteins secreted by human skin cells at the wound edge may actually be driving wound closure. Their previous research found that HDFs and HKs produced a secreted form of Hsp90α but only as a result of pathophysiological (e.g., cancer) or environmental stress cues (e.g., hypoxia or gamma-irradiation).

In their latest work, the team compared the the effects of HSp90α with those of beclaplermin in the treatment of acute and diabetic wounds in mice. They first used deletion mutagenesis to generate various fragments of human HSp90α. They identified a 115 amino acid fragment, F5, as the smallest HSp90α fragment that was capable of stimulating HDF migration as effectively as the full-length Hsp90α protein, albeit at higher concentrations.

A number of the identified peptides and becaplermin gel were then tested in athymic hairless mice. Different concentrations of each peptide were initially used to identify the optimal concentrations that most effectively prompted full-thickness excision wound closure after a single application. Peptides in their optimized concentrations were subsequently compared with becaplermin gel in terms of their ability to promote wound closure over time.

Interestingly, the researchers found that full-length Hsp90α, F-5, and a 392-amino acid fragment (F-2), all strongly accelerated wound closure, with F-5 showing the strongest effect. When compared with placebo therapy, F-5 treatment shortened the time of complete wound closure from about 17 days to 10 days, irrespective of how many applications of F-5 were given.

More surprisingly, the researchers note, becaplermin gel therapy showed only limited ability to promote acute wound closure. Three independent experiments confirmed that the Hsp90α peptides performed better than the FDA-approved gel in this respect.

Examination of wounds by staining indicated that F-5 acted by promoting re-epithelialization—the mechanism by which human skin wounds heal—rather than by contraction, a mechanism which is possible in mice because of the nature of the animal’s loose skin and dense hair follicles.

The team then moved on to evaluate whether F-5 could promote the healing of chronic wounds as well as acute wounds. They used the db/db diabetic mouse model and showed that a single F-5 treatment led to wound closure within about 14–18 days, whereas placebo-treated wounds took about 35 days to close. H&E staining on day 14 again showed that F-5 treatment promoted a greater degree of epidermal re-epithelialization than placebo. “These results show that F-5 has an even more prominent effect on diabetic wounds than acute wounds,” the researchers claim.

Immunostaining assays suggested that the activity of F-5 wasn’t related to the abnormal increase of wound angiogenesis or wound contraction. Compared with placebo, F-5 therapy caused neither excessive recruitment of endothelial cells to the wound nor increases in myofibroblasts. In fact antibody staining showed visibly fewer myofibroblasts in F-5-treated wounds than in placebo-treated wounds.

Because PDGF-BB has been found to have little effect on acute wound healing in mice but does speed wound healing in db/db animals, the researchers compared the effects of a single treatment with either placebo, F-5, or becaplermin on full-thicknesss wound healing in this mouse model. Becaplermin therapy did speed wound healing when compared with placebo but was still significantly less effective than F-5, which again led to complete wound closure between 14 and 18 days.

The question emerging from these studies was, therefore, what made F-5 more effective than PDGF-BB in healing both acute and diabetic wounds? To try and answer this Drs. Li et al. first analyzed the effects of F-5 on migration of HKs, HDFs, and HDMECs. In the absence of any stimulus, all three types of skin cells exhibited limited levels of motility. However, while F-5 promoted the migration of all three types of cells, PDGF-BB promoted the migration of HDFs but not HKs or HDMECs.

In the hunt for the basis of this difference, the researchers first focused on the presence or absence of the receptors for PDGF-BB (i.e., PDGFRα and PDGFRβ) and secreted Hsp90α (i.e., LDL receptor-related protein-1 [LRP-1]). What they found, was that while PDGF-BB-responsive HDFs expressed both PDGFRα and PDGFRβ, the PDGF-BB non-responsive HK and HDMEC cells expressed neither of these receptors. “If we extrapolate these in vitro findings to equivalent wound-healing events in mice, it suggests that PDGF-BB cannot have a direct role in recruitments of HKs for wound re-epithelialization and HDMECs for wound neovascularization,” the researchers remark.

The second part of their search for differential factors impacting on the effectiveness of F-5 and PDGF-BB centered on the previous finding that TGF-β3 in human serum selectively blocks growth factor-induced HDF and HDMEC migration due to the higher levels of TβRII expression in these dermal cells than in epidermal cells. Further evaluation confirmed that while the presence of TGF-β3 inhibited PDGF-BB-induced migration of HDF, it had no effect on the ability of F-5 to stimulate migration of all three cell types.

In a third set of studies designed to investigate what made F-5 more effective in promoting diabetic wound healing specifically, the researchers tested whether hyperglycemia blocks hypoxia-induced HDF motility and whether F-5 is able to bypass this and rescue HDF migration. Hyperglycemia is reportedly able to destabilize HIF-1α protein, which is a key regulatory of HSP90α in HKs and HDFs.

They found that hypoxia strongly promoted HDF migration in conditions of normal glucose concentrations, while hypoxia plus F-5 resulted in a slightly higher stimulation of cell migration. However, hypoxia failed to simulate HDF migration under conditions of hyperglycemic and also blocked PDGF-BB-stimulated migration.

Conversely, the addition of F-5 prompted cell migration even under hyperglycemic conditions. “This finding provides the third explanation for why F-5 showed a stronger effect on accelerating diabetic wounds healing,” the investigators note.

Finally, the team evaluated whether secretion of Hsp90α is actually a critical factor in wound healing. They used a lentiviral shRNA delivery system known as FG-12 to permanently downregulate LRP-1, to investigate the importance of the Hsp90α receptor in HKs, HDFs, and HDMECs. A lack of LRP-1 did indeed result in all three cell types failing to migrate in response to F-5 stimulation, but migration could be reinstated by the addition of a mini-LRP-1 receptor.

They then administered LRP-1-associated protein (RAP), which directly blocks LRP-1 signaling, to F-5-treated nude mice carrying five-day old wounds. While topically applied F-5 strongly promoted wound healing in nude mice by day five, the addition of purified RAP delayed the wound-healing process, confirming that LRP-1 signaling plays a critical role in normal wound healing.

The researchers have used their combined results to draw up a new hypothesis about the factors that drive epidermal and dermal cell migration in wound closure. According to this hypothesis, the process starts with the migration of HKs laterally across the wound and the secretion of Hsp90α, within the first few hours of skin injury. At this early stage HDFs and HDMECs at the wound edge are prevented from moving into the wound bed due to the presence of TGF-β3.

It is only once secreted Hsp90α tops a threshold concentration that dermal cells are triggered to migrate into the wound bed from the surrounding wound edge, despite the presence of HTF-β3. The migrating HKs then completely close the wound, and the newly recruited HDFs can start to remodel the wounded tissue, while the HDMECs rebuild new blood vessels.

“We propose here that injury-induced secretion of Hsp90α, instead of the conventional growth factors, is the initial driving force of wound closure,” the authors conclude. “The discovery of F-5 challenges the long-standing paradigm of wound-healing factors and reveals a potentially more effective and safer agent for healing acute and diabetic wounds.”

David G. Armstrong

Dedicated to amputation prevention, wound healing, diabetic foot, biotechnology and the intersection between medical devices and consumer electronics.

Submit a comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s