Hair growth/loss, PPARs and Keratin

[TexugoDoMel]

This post is from Travis, he posted it in 2017~2018.

After viewing many studies showing how prostaglandins inhibit hair growth, I had an idea that keratin itself was likely under the control of the PPARs—the transcriptional receptors for the prostaglandins. After all, you have to take the ghost out of the machine at some point; there has to be a final mechanism for the effects of prostaglandins.

Mice genetically-engineered with no PPARγ receptor on their skin have scarring alopecia. Also, over-expressing PPARγ on the skin protect rats against hair loss. And just as you'd expect, anything that raises prostaglandins on the skin creates similar hair loss: over-expression of cyclooxygenase, over-expression of phospholipase A₂, and over-expression of interferon-γ on the skin all lead to bald mice. In addition to the well-known findings of Louis Garza, you'd be forced to conclude that hair loss is under the control primarily of prostaglandins. You could speculate on an indirect, or even physical mechanism for this; you could say that perhaps that the oils are creating hypoxia, or that prostaglandin D₂ is exerting its well-known ischemic effects—but let's not forget that prostaglandins also transcribe/repress mRNA directly. For this reason, I've decided to check to see if the keratin genes were under the control of the PPAR receptors:

Lemay, Danielle G., and Daniel H. Hwang. "Genome-wide identification of peroxisome proliferator response elements using integrated computational genomics." Journal of lipid research (2006)​

Danielle G. Lemay informs us that the PPAR response element is basically AGGTCA. This is the canonical DNA sequence that the PPAR receptors bind to, and you can expect any DNA sequences with the AGGTCA sequence to be under the control of the prostaglandin·PPAR receptors—and their heterodimers.

The European Nucleotide Database has full sequences for all forty keratin genes. I have searched all forty genes for the PPAR response element. This is a six nucleotide sequence, and as such would only be found every.. . . .

(¼)⁶ = (¹⁄₄₀₉₆)

. . . ..four thousand ninety-six base pairs, on average. Keratin genes are rarely this long, and this sequence is found multiple times in many of them—far exceeding chance. The keratins genes which have been found to include the full six nucleotide sequence, or the canonical PPAR response element, are listed below:

keratin 5
keratin 6
keratin 7
keratin 8
keratin 9
keratin 11
keratin 12
keratin 15
keratin 17
keratin 18
keratin 19
keratin 20
keratin 24
keratin 25
keratin 27
keratin 28
keratin 31
keratin 38
keratin 39
keratin 40

All off the other ones have either a the minus-one-base-pair-truncated five-nucleotide sequence (AGGTC) or the core sequence of AGGT, except for keratins 29, 30, and 33.

Wikipedia will inform you that keratins 31–40 are the hair keratins. With this in mind, you'd have to assume that skin keratins are also under the control of the PPAR receptors. This means: prostaglandins could effect keratin expression in epithelial barriers everywhere—even the lungs.

The hair follicle will continue-on with its cycle after explantation showing that it has what can be considered—although I downright hate the analogy—an internal "clock." This fact alone means that it's not under the control of steroid hormones, molecules which cannot be produced locally. The only steroids which are ever produced locally the ones immediately downstream of cholesterol (i.e. pregnenolone and progesterone). The sex hormones are produced by sex organs, and the adrenocorticoids indicate their origin by their name.

The hair cycle is almost certainly under the control of prostaglandins operating through PPARs, and perhaps modulated or interrupted by cytokines and growth factors which influence prostaglandin production. Peptide hormones such as the transforming growth factor beta (TGF-β) series are released cyclically during the hair cycle, appearing in the same order in which they're numbered. In most observations, TGF-β₂ marks the beginning of catagen. Aldosterone has been shown to upregulate this, powerfully, so should theoretically be able to disrupt the hair cycle. Indeed, shock has been shown to cause shedding. Perhaps this is why mineralcorticoids, such as hydrocortisone, will leave a bald spot when applied to the skin. The transforming growth factors go on to release create prostaglandins through the action of phospholipase A₂, connecting mineralcorticoids with PPAR and keratin.

The recent discoveries that prostaglandin E₂ will protect rats from radiation-induced alopecia, that prostaglandin F₂α stimulates the hair growth of the eyelashes, and that prostaglandin D₂ (and derivative J₂) powerfully suppress hair growth lends further credence to the idea fact that prostaglandins have a fundamental role in the hair cycle; and since they almost certainly control keratin production directly through the PPARs, they must be the essential and indivisible component. There is nothing downstream of prostaglandins, besides the transcription of keratin. What still needs to be understood is how the timing is controlled; If I were to take a guess, I would say that the cystosolic concentration of the enzyme prostaglandin D synthase changes in a way that shadows the hair cycle.

There are actually two types of prostaglandin D synthase: one cytosolic, and one blood-borne one (lipocalin-like prostaglandin D synthase). Besides kamikaze cytokines inducing phospholipase A₂ and cyclooxygenase-2, an upregulation of lipocalin-like prostaglandin D synthase could have similar effect. Louis Garza analyzed the mRNA for this enzyme, but he failed to give the PCR primer sequence in his paper. There appears to be no way of knowing whether he measured the mRNA for ptgds or L-ptgds. This is important, as these two enzymes have radically-different origins yet both produce prostaglandin D₂.

Zuo, Xiangsheng. "Oxidative metabolism of linoleic acid modulates PPAR-beta/delta suppression of PPAR-gamma activity." Oncogene (2006)​

This scientist says that the PPARs compete with eachother:

"PPAR-β/δ overexpression suppresses the activity of PPAR-γ and PPAR-α."​

(He makes no mention of PPARΨ, probably because he's studying human cells and this receptor is found only in unicorns.)

This is interesting. Apparently caffeic acid inhibits the enzyme that turns linoleic acid into 13-S-HODE,* a hydroxylated linoleic acid which can bind PPARs β and δ. This is formed from linoleic acid through the enzyme 15-LOX-1.†

"whereas adding caffeic acid (a 15-LOX-1 inhibitor) in a concentration that specifically inhibits 15-LOX-1 enzymatic activity reduced their 13-S-HODE production."

"We inhibited 15-LOX-1 enzymatic activity with caffeic acid, which significantly reduced PPARγ activity in the 15-LOX-1-transfected cells"​

So at this point, he has shown that this linoleic acid metabolite (HODE) increases PPARγ dependent expression. He did this by injecting luciferase DNA spliced next to a PPAR response element‡ into the cell. So whenever PPARγ is activated, luciferase will be produced. This is a fluorescent protein, and I think the high tyrosine and phenylalanine content must have something to do with it. I can't see how you'd have a fluorescent protein without these amino acids (or tryptophan).

"Among the PPARs, PPARγ selectively induces Keratin 20 expression in colon cancer cells. 15-LOX-1 increased CD36 and Keratin 20 expression in LoVo and HCT-116 cells, and inhibition of 15-LOX-1 enzymatic activity suppressed that increased expression. Ad-15-LOX-1 transfection of LoVo cells increased the expression of CD36 by a factor of 2.5 and Keratin 20 by a factor of 3 relative to that in the control. Inhibition of 15- LOX-1 enzymatic activity blocked those increases in expression levels. Similarly, in HCT-116 cells, 15-LOX-1 increased the expression of both CD36 and Keratin 20, whereas inhibiting 15-LOX-1 enzymatic activity blocked these effects."​

So activating the PPARγ with his hydroxylated linoleic acid (hydroxylated by LOX) caused a threefold increase in keratin 20. This is what you'd expect if you had read my post, since I showed that keratin 20 had a PPAR response sequence (AGGTCA). This scientist confirms directly that keratin 20 is under the control of PPARγ and hydroxylated linoleic acid. He also says that PPARβ and PPARδ inhibit PPARγ (the receptor shown to prevent hair loss in rats), so perhaps prostaglandin D₂ works through one of these. I did read before that prostaglandin D₂ binds to PPARα. Perhaps this receptor inhibits keratin formation in the cell when activated?

Okay, now he proves that the PPAR receptors antagonize eachother. This is huge, and will explain why prostaglandins have different effects. He uses small interfering RNA (siRNA) to hybridize the PPARβ and PPARδ messenger RNA. This is simple, as every nucleic acid has a corresponding sequence that it coils with. These corresponding sequences coil with the mRNA, forming a double-helix and inactivating it—preventing the expression of PPARβ and PPARδ:

This sequence here: AGGTCAGTTACCCGTAACGAGGTCAGTTACCCGTAACGAGGTCAGTTACCCGTAACGAGGTCAGTTACCCGTAACG
Will hybridize with: TCCAGTCAATGGGCATTGCTCCAGTCAATGGGCATTGCTCCAGTCAATGGGCATTGCTCCAGTCAATGGGCATTGC

"PPARβ/δ siRNA downregulated PPARβ/δ mRNA expression by approximately 89% in LoVo cells and 85% in HCT-116 cells as measured by real-time polymerase chain reaction assays. Western blot analyses showed marked downregulation of PPARβ/δ by PPARβ/δ siRNA in both cell lines. The resulting PPARβ/δ downregulation increased PPARγ activity in both LoVo and HCT-116 cells and did so significantly more than mock (with medium only), nonspecific siRNA, or PPARα siRNA transfections did."

"PPARβ/δ downregulation increased Keratin 20 expression in LoVo and HCT-116 colon cancer cells by factors of 3.3 and 2.8, respectively. Furthermore, in KO1 cells, Keratin 20 expression levels were 5.3 times higher than they were in their parental HCT-116 cells. Transient transfection of KO1 cells with PPARβ/δ expression vector significantly reduced Keratin 20 expression by approximately 35%."​

Okay. This is proof that PPARs β and δ antagonize PPARγ and prevent the expression of keratin 20. Perhaps this is why PPARγ is the most popular PPAR to research, it's the only one that works. All of the others seem to inhibit PPARγ in some way. Perhaps the others transcribe other proteins, but they seem to prevent the transcription of keratins. I can only think that prostaglandins E₂ and F₂α are PPARγ ligands, while prostaglandins D₂ and J₂ preferentially activate one of the others.

He then got weird results using a different ligand, results that I can only assume can result from PPARα still being active. It appears as though both PPARα and PPARγ both transcribe keratins, while PPARβ and PPARδ inhibit this.

And he test the effect in vitro, with DNA and the PPARs. He finds that only PPARγ will bind to the response element, and that PPARβ and PPARδ inhibit this:

"We also quantitatively examined the effects of PPARβ/δ on PPARγ binding to the peroxisome proliferator response element. PPARβ/δ alone changed the optical density reading minimally, confirming the assay’s specificity in measuring PPARγ binding to PPRE. When PPARβ/δ was added with PPARγ, PPARβ/δ significantly decreased PPARγ binding to the PPRE."​

I took a look at the graphs, and linoleic acid alone does not activate PPARγ or produce much keratin 20. The hydroxylated linoleic acid works a bit better, but troglitazone worked twice as well as that. Taking a look at the structure, it has a resemblance to oleoropein—a molecule produced by the olive tree which actually stimulates hair growth. Oleoropein then likely works by activating PPARγ and transcribing keratin mRNA.

So you'd think from all of this that the subtypes PPARβ and PPARδ are activated by prostaglandin D₂ as ligand—suppressing transcription of keratin by inhibiting PPARγ and PPARα directly in the nucleus, at a transcritptional level: fighting for the AGGTCA response element.

Warning about PPARα agonists: "Clofibrate may occasionally produce hair loss and hair brittleness."

Tosti, Antonella. "Drug-induced hair loss and hair growth." Drug Safety (1994)​

Clofibrate is a classic PPARα agonist, and it's effect on hair confirms the idea that the PPAR receptors transcribe keratins. It appears as thought PPARα isn't so simple. Here is what Dr. Nelly has to say about clofibrate:

'Effects of clofibrate on in vitro culture of human hair follicles: We investigated the effects of clofibrate, a PPARα ligand, on the survival of freshly dissected human hair follicles grown in vitro. As shown in Fig. 3, we found that the survival of follicles was enhanced by clofibrate dose-dependently up to 10⁻⁸ M, after which a decrease in survival time was observed. Survival of clofibrate-treated follicles was similar to that of
untreated control hair follicles during the first 3 days of culture but, at day 11, none of the 12 untreated control hair follicles were still growing whilst hair growth could still be detected in the presence of 10⁻¹⁰ M and 10⁻⁸ M clofibrate. A statistically significant positive effect on survival was observed at 10⁻⁸ M. However, a higher concentration of clofibrate (10⁻⁶ M) induced cessation of follicle growth compared with the control conditions. Indeed, although the survival percentage of 10⁻⁶ M clofibrate-treated follicles was similar to that of untreated control hair follicles during the first 3 days of culture, at day 4 the survival rate fell to 75% in treated hair follicles vs 92% in the control group and, at day 8, none of the treated hair follicles were still growing, whilst 17% of control untreated hair follicles were still growing. Similar results were obtained in 2 independent experiments, with follicles obtained from 2 different donors.'

The PPARα has unusual ligands, and you might suspect that oleuropein is a PPARα agonist based on it's structure (although it does also inhibit COX-2.)*

View attachment 7331

They all have that ester group, and also substituted phenyl groups. The ester group is most unusual for PPAR ligand, and I haven't seen an others that have it.

It makes you wonder what the natural PPARα ligand is? Could it be . . . a cholesterol ester?

View attachment 7332

And perhaps retinol esters, as PPARα is highly-expressed in the liver. I did just find this:

"In macrophages, PPARα activation also reduces the cholesteryl ester content due to reduced cholesterol esterification rates and Acyl-CoA:cholesterol acyltransferase-1 (ACAT₁) activity." ―G. Chinetti†​

It really makes you think this could be so: The PPARα receptor could be sensing cholesterol esters and lowering their production in a negative-feedback manner. This could explain why: clofibrate is used clinically in an attempt to lower cholesterol, activation of the PPARα receptor has been shown experimentally to lower cholesterol esterification rates (a.k.a. acylation) through downregulation of enzyme responsible, and also why all of the PPARα ligands appear to have ester groups attached to a carbon ring structure.

I'm going to file-away the PPARα receptor as the cholesterol ester sensor. The Acyl-CoA:cholesterol acyltransferase-1 gene doesn't have the canonical PPAR response element, but J D Tugwood has narrowed two down for PPARα—specifically:

"We have examined the 5' flanking region of the rat ACO gene for sequences that mediate the transcriptional effect of peroxisome proliferators and have identified an element located 570 bp upstream of the ACO gene that confers responsiveness to the hypolipidaemic peroxisome proliferator Wy-14,643. This peroxisome proliferator response element (PPRE) contains a direct repeat of the sequence motifs TGACCT and TGTCCT and binds PPAR." ―J D Tugwood§​

Now, Wy-14,643 is a PPARα ligand (as shown here). This means that TGACCT and TGTCCT are PPARα response elements. This also shows why PPARα transcribes the enzyme Acyl-CoA:cholesterol acyltransferase-1, as it has a TGTCCT response element.

I decided to check the hair keratins, the ones not transcribed by PPARγ. All of the hair ketatins (№'s 30–40) not transcribed by PPARγ have the PPARα response element TGACCT, with the lone exception of keratin 33. That is to say: keratins № 32, № 34, № 35, № 36, and № 37 all have the PPARα response elements. All of the other "hair keratins" have PPARγ response elements with the lone exception of keratin № 33. And like the PPARγ response element the PPARα response element is six nucleotides long, making the odds of its appearance one in four thousand ninety six. All of the keratins just described were much shorter in length, between 1,000 and 1,500 base pairs. Moreover, this response element often appeared multiple times in their sequence. I don't think it's unreasonable to assume that these keratins are under the control of PPARα.

This could be why clofibrate promoted hair growth slightly, but could not maintain this growth at high concentrations. Perhaps too much PPARα cannot sustain hair growth without PPARγ keratins in the mix?.. . .(unbalanced keratin ratio?). . ..Also, the brittleness described by Antonella Tosti could perhaps have been the result of clofribrate-induced keratin imbalance—working through PPARα.

Should anyone like to confirm these observations, the sequences I had used were kindly provided by the European Nucleotide Archive. The nomenclature for the keratin genes is quite simple: it is krt××, with the '×' indicating a Roman numeral—an integer. I can say that these response elements far exceed chance, but any statistical analysis would certainly be welcome.

Current State of the Theory:

The hair keratins are numbered between thirty and forty. Some of them have PPARα response elements, and the others have PPARγ response elements:

PPARα Keratins: № 32, № 34, № 35, № 36, and № 37
PPARγ Keratins: № 31, № 38, № 39, and № 40

It has also been shown that PPARβ and PPARδ inhibit the general transcription of PPARα and PPARγ (Zuo, 2006). For this reason, it should perhaps be assumed that prostaglandin D₂ binds to either PPARβ or PPARδ.

PPARα Agonists: cholesterol ester, oleuropein, pirinixic acid, and all fibrates
PPARγ Agonists: prostaglandin E₂, biochanin A, resveratrol, and all thiazolidinediones

Prostaglandins—working through the PPAR receptors—control the hair cycle with the help of the transforming growth factor beta (TGF-β) series. They do this primarily through regulating steroid esters, fatty acids, enzymes, and keratin proteins themselves.

A paradigm having different ligands for different PPAR receptors which selectively allow, or inhibit, hair growth can resolve such contradictory findings such as:

Garza, Luis A. "Prostaglandin D₂ inhibits hair growth and is elevated in bald scalp of men with androgenetic alopecia." Science translational medicine (2012)
Larson, Allison R. "A prostaglandin D‐synthase‐positive mast cell gradient characterizes scalp patterning." Journal of cutaneous pathology (2014)
Johnstone, Murray A. "Prostaglandin-induced hair growth." Survey of ophthalmology (2002)
Michelet, Jean-Francois. "Activation of cytoprotective prostaglandin synthase-1 by minoxidil as a possible explanation for its hair growth-stimulating effect." Journal of investigative dermatology (1997)
​
The PPARβ, or PPARδ, receptor seems to be the main target for prostaglandin D₂. I have yet to find direct proof of this, largely because these receptors are not thorougly studied: all eyes are on PPARγ. But I did just now find a study which indicates as much—that PPARs β and δ are, in fact, the pathological receptor subtypes:

Romanowska, Malgorzat. "Activation of PPARβ/δ causes a psoriasis-like skin disease in vivo." PloS one (2010)​

They're mainly concerned about psoriasis here, but they confirm the PPARβ/δ with the PPARα/γ antagonism.

"As shown in figure 1a, both data sets confirm highly significant upregulation of PPARβ/δ in psoriatic skin whereas both PPARα and PPARγ are downregulated, consistent with the notion that PPARβ/δ acts antagonistically to PPARγin psoriasis, as previously proposed."​

Transgenic mice were created by injecting DNA for the PPARδ receptor next to a CYP1A1 promoter. This had given them extra PPARδ receptors. The experimenters also put the agonist drug GW501516 in the food, to activate these receptors.

"Psoriasis-like skin disease in PPARb/d transgenic mice: As early as seven days after initiation of PPARβ/δ - activation by GW501516, scaling, inflammation, and skin thickening was notable in all mice. Skin roughening and concomitant hair loss was maximal in regions subjected to mechanical friction, such as abdomen, the paws, or the chin (fig. S2)."​

I have seen 'fig. S2' and can confirm the lack of hair. These mice also upregulated the killer cytokine interleukin-1β, which I'd rank in the same category as interferon-γ and tumor necrosis factor alpha. They confirmed that the psoriasis lied downstream of the cytokines, showing that prostaglandins are ultimately responsible.

"PPARβ/δ expression is enforced downstream of TNFα in this model."​

Like all the A-list-terrible cytokines (such as interleukin-1β and interferon-γ) tumor necrosis factor alpha induces transcription of phospholipase A₂ once inside the cell. This means that after it gets done signalling, prostaglandins will continue to be produced for quite some time. This is why the three aforementioned cytokines are the worst cytokines of the 30-or-so in existence.

There should be no inherent teleological reason why prostaglandin D₂—working through PPARβ/δ—should cause hair loss. That is to say: the body is not shedding hair for defense, that would be absurd. What I think is going on is this: These PPAR receptors control the differentiation of the epithelial cells—all of them. The relative concentration of prostaglandins obviously determines their phenotype, to some extent, on the skin—a fact plainly obvious by observing genetic manipulations as seen above. The endothelial cells in the lungs and arterial lumen, for instance, would have a different set of keratins and other proteins being expressed. This is necessary to have different epithelial cell types at all, and they must have the ability to create different proteins. The precise balance—or ratios—of prostaglandins can help turn any epithelial cell into either an arterial cell, lung cell, or some other chimera without any sort of usual definition. An epithelial stem cell can become anything depending on where its placed. For a skin cell to remain a skin cell—and to be recognized as such—it must retain the natural balance of prostaglandins found in the skin, or else the difference in prostaglandin ratio will variably and pathologically activate the PPARs resulting in protein ratios unfit for it's location. This is scarring, psoriasis, and hair loss.

[mnemosyne]

Hugely overlooked when it comes to hair loss is the structural aspect. The blood and nutrients need to actually be delivered and directed away from the place again. Notice how the top of your head is literally the last place your blood is going to reach if problems arise.

[basednigga2006]

@TexugoDoMel bump

[Vonshlovens]

@mnemosyne so would blood flow (through heart rate?) be a big priority in your opinion? Do you think inversion therapy (headstands, etc) could be helpful?

[MasterPeatler]

@Vonshlovens I microneedle.