By Abby Van Voorhees, MD, June 01, 2012
In this month’s Acta Eruditorum column, Physician Editor Abby S. Van Voorhees, MD, talks with George Cotsarelis, MD, about his recent Science Translational Medicine article, “Prostaglandin D2 Inhibits Hair Growth and Is Elevated in Bald Scalp of Men with Androgenetic Alopecia.”
Dr. Van Voorhees: What have we known about androgenetic alopecia (AGA) up until your most recent findings?
Dr. Cotsarelis: We know a lot about AGA from studies a long time ago, in the ’50s, by Dr. James Hamilton. He was the first one to coin the term androgenetic alopecia. He noticed that men who had been castrated before puberty never went bald, and realized that baldness was androgen-dependent. He did some experiments where he gave men who had been castrated testosterone; the ones who had a family history of baldness started to go bald, so that was how he identified the genetic component. He originally called it androchronogenetic alopecia because you needed androgens, time, and genetics. We’ve known it’s androgen-dependent since then.
Next we learned more from studies by Imperato-McGinley, who studied families in the Dominican Republic who were pseudo-hermaphrodites — they were born looking like girls, with female genitals, and then at puberty they virilized, developing musculature, lower voices, and hair, and would actually develop a penis and could be fertile if they had their undescended testes surgically corrected. They were found to have a deficiency of type-2 5- reductase, the enzyme that converts testosterone into dihydrotestosterone. This was when we understood the importance of dihydrotestosterone in the process. Of course that’s what finasteride blocks; Merck developed that drug based on those studies.[pagebreak]
Dr. Van Voorhees: What is the mechanism that accounts for the miniaturization of hair?
Dr. Cotsarelis: We know that inhibiting the testosterone pathway slows down the miniaturization of the follicle. Jaworsky, Kligman, and Murphy had a paper 20 years ago showing that half the time there is also inflammation around the hair follicle, which led to some thought that maybe inflammatory cells including mast cells were contributing to hair loss. Studies and case reports of transgender operations where men become women and receive high doses of estrogen show that a scalp that was almost completely bald can have, after castration and high estrogen supplementation, a tremendous amount of hair growth.
The overall feeling is that the follicles can be thought of as being in three states. Either they’re terminal, and they’re large, or they’re miniaturized, and they’re small, and the hair they’re creating is microscopic, or they’re in between, called indeterminate. It’s thought that follicles reach a point where they’re producing a hair so small that at that point the chance of reversing that follicle is small. There seems to be a point of no return with respect to androgen removal; even if you castrate someone who’s bald he won’t regrow all his hair. If you give him estrogen, too, he might.[pagebreak]
Dr. Van Voorhees: Do resting stem cells remain in AGA? Do current medications used to treat AGA help us understand this process? What did this research demonstrate about the mechanism of AGA?
Dr. Cotsarelis: About a year ago Luis Garza, MD, did another study in my lab and looked at the stem cells in balding scalps of men undergoing transplants. He got tissue from the donor site and the recipient site and he actually quantitated the number of hair follicle stem cells using flow cytometry, a very accurate cutting-edge technique for quantitating the number of stem cells. And he showed that the percentage of stem cells was very similar in both areas. So the stem cells were still there even in the balding scalp where the follicles were miniaturizing.
However, he did find a population of cells that was markedly diminished in the balding scalp, and those resembled progenitor cells, the immediate progeny of stem cells. At that point we thought there must be either a lack of an activator of the stem cells preventing them from proliferating and doing their job and making hair, or maybe an inhibitor present. Using microarrays of bald and non-bald scalp in the same people, we looked at all 30,000 genes to see whether their expression levels were increased or decreased in the bald scalp. One of the genes that was markedly elevated in the bald scalp was PTGDS, prostaglandin D2 synthase.
We collaborated with Garret FitzGerald, MD, who’s an expert in prostaglandins and has done a lot of work with the COX-2 inhibitor he’s the one who actually showed that the COX-2 inhibitors have dangerous side effects. We gave him tissue from balding and non-bald scalp and he looked at the levels of PGD2 with mass spectrometry and showed that the PGD2 levels were quite high in the bald scalp. That was still just a correlation showing that there were higher levels of the genes and the lipid products in the bald scalp — we didn’t really know functionally whether that meant anything.
So then Luis worked with mouse models to assess the function of PGD2. Luis applied the PGD2 and another breakdown product, PGJ2, topically to the back of a mouse and he showed that hair growth slowed. Next, we collaborated with another group that was good at growing human hair follicles in culture. They showed that PGD2 in the culture media slowed down hair growth. So now we had functional evidence that prostaglandins played a role in hair growth. Luis also looked in the mouse, which has a better-defined hair cycle than humans because all of the hairs tend to be in the same stage whereas humans have mixtures of hair follicles in different stages of growth. In that model we showed that PGD2 started to go up during the end of the growing phase and was highest at catagen, the stage of regression. That, again, was good correlative evidence that PGD2 was inhibiting hair growth and even in a spontaneous hair cycle played a role in controlling hair growth.
There was a mouse that was made in the ’90s by Sue Fisher, who studies skin cancer; it’s a transgenic mouse that makes the COX-2 enzyme in the skin, so you drive COX-2 gene expression with keratin K14 promoters. It’s up higher in the pathway so if you overexpress COX-2 you get increases in PGD2 and also PGE2. E2 and F2 are actually known to promote hair growth — bimatoprost (Latisse) is an F2 analog. We looked at levels of D2 in that mouse, which develops alopecia, and they were sky high. E2 was also higher than the control but D2 levels were much higher. What’s amazing is if you looked at that mouse’s skin histologically the hair follicles were miniaturized and the sebaceous glands were enlarged, just like in androgenetic alopecia.
Finally, in genetic experiments done in collaboration with FitzGerald, we applied D2 to mice that lacked two different receptors that D2 binds to. The DP-1 knockout mouse had suppressed hair growth. The DP-2 (or GPR44) knockout mouse did not have inhibited hair growth.[pagebreak]
Dr. Van Voorhees: Does this research open up possible avenues for treatment approaches that have not yet been considered? Are there currently drugs under development which might be utilized?
Dr. Cotsarelis: This work suggested that D2 was working through the GPR44 receptor to inhibit hair growth. It turns out that there are compounds under development by a number of companies to inhibit this receptor. They are being developed for asthma and allergic rhinitis. PGD2 causes bronchoconstriction — when you inhibit its receptor you relax smooth muscle so it helps with lung disorders. No one is developing a topical formulation, but we think if you did, it would be a potential treatment for alopecia.
Dr. Van Voorhees: Do you expect female pattern hair loss to have the same pathologic mechanism?
Dr. Cotsarelis: The only evidence we have is that when you take hair follicles from women, put them in the culture medium, and add D2, it inhibits their hair growth as well. But we haven’t studied women with female pattern hair loss to see if D2 is elevated there as well. The problem is that women who get hair transplants don’t have the donor scalp removed like men do; the grafts go in between existing hair follicles. The role of PGD2 in female pattern hair loss is an important question to address.
Dr. Van Voorhees: Should this allow for reversal of previously lost hairs or do you expect that it will only play a role in retaining hairs that have not yet miniaturized?
Dr. Cotsarelis: We don’t know. Like anything you’d have to test a large number of people to see how they respond. We don’t know if people who are already completely bald will regrow hair. We do know that the stem cells are present in men who are balding so if this is indeed the inhibitor preventing stem cells from making progenitor cells, there’s a possibility this would help there as well.
Dr. Van Voorhees: Why do you think there’s that threshold where miniaturization becomes non-reversible?
Dr. Cotsarelis: That’s an interesting question. There might be other genes downstream of testosterone playing a role. We already know that if you inhibit testosterone that hair doesn’t revert once it’s completely miniaturized. The fate of the hair follicle is determined very early on during development; there’s patterning, that’s why you have follicles responding to androgens on the top of the scalp and in the beard area in completely different ways. Androgen receptors are set up very early in development; I think the more we understand about that patterning the more we’ll likely be able to figure out what’s going on with hair loss.
Dr. Cotsarelis is chairman and Milton B. Hartzell Professor in the department of dermatology at the University of Pennsylvania. His article was published in Science Translational Medicine, Vol. 4, Issue 126, p. 126ra34.doi:10.1126/scitranslmed.3003122.