Self-Heal: Antiviral and Herpes Research

This is the interesting one. Strip away four centuries of "heal-all" and there is a real, reproducible, independently confirmed laboratory finding at the centre of the modern Prunella vulgaris literature: water extracts of the plant, and a purified sugar-based fraction from those extracts, block herpes simplex virus from infecting cells in culture. Several separate research groups have shown it. It is not a fluke, it is not a single lab, and it is not a marketing claim invented by a supplement company.

It is also not evidence that drinking self-heal tea does anything to a cold sore. This page explains, in detail and in both directions, why both of those statements are true at once — and it does something unusual, because for this particular claim the underlying mechanism has actually been tested in humans. Not with self-heal, but with the same class of molecule, at proper doses, with real endpoints. Those results exist, and they are not encouraging. That makes this one of the very few claims about this plant that sits in the negative column rather than the absent one, and we are not going to bury it at the bottom of the page.

Table of Contents

1. What Was Actually Found
2. Prunellin and the Sulfated-Polysaccharide Story
3. The HSV Cell-Culture Work, Read Carefully
4. The Aciclovir-Resistant Strain Claim
5. The 1990s HIV Work and What Became of It
6. The Mechanism Itself Was Tested in Humans
7. Route and Dose: Why a Cup of Tea Is Not the Experiment
8. What a Real Trial Would Look Like
9. Comparators That Do Have Human Trials
10. Absent, Negative, and Where Each Claim Sits
11. Key Research Papers
12. Connections

What Was Actually Found

The core of the finding, stated as narrowly as the data allow:

1. Aqueous (water) extracts of the dried aerial parts of Prunella vulgaris inhibit herpes simplex virus in cell culture. Both HSV-1, the usual cause of oral cold sores, and HSV-2, the usual cause of genital herpes.
2. The activity survives purification. When researchers fractionate the extract, the antiviral activity tracks with a large, water-soluble, sulfate-bearing polysaccharide rather than with the small phenolic molecules.
3. The effect is on entry, not on replication. The extract works when it is present at the moment the virus meets the cell. Pre-treating cells, or treating cells that are already infected, is much less effective. This is the signature of an attachment or entry blocker.
4. Infected cells make less viral antigen when the polysaccharide fraction is present — a downstream confirmation that fewer virus particles got in.
5. Activity has been reported against virus strains resistant to aciclovir, which is discussed on its own below because it is routinely misread.

Every one of those five statements is about cells in plastic dishes. That is not a dismissal — cell-culture antiviral screening is how essentially every antiviral drug in the pharmacy was first identified. It is a statement about where on the development path this finding sits, which is at the very beginning.

Prunellin and the Sulfated-Polysaccharide Story

The name "prunellin" enters the literature in 1989, in a paper describing an anti-HIV component isolated from water extracts of the plant. That is worth pausing on, because the name has since escaped into supplement marketing as though it referred to a characterised drug. It does not. It refers to a partially characterised polysaccharide fraction from a 1989 isolation.

What kind of molecule is it? A sulfated polysaccharide — a long chain of sugar units carrying negatively charged sulfate groups along its length. Think of it as a large, floppy, strongly negatively charged ribbon, with a molecular weight in the tens or hundreds of thousands of daltons. It is not a small molecule; it is closer in scale to a protein.

That structure explains the antiviral mechanism neatly. Herpes simplex virus begins infection by binding to heparan sulfate proteoglycans on the cell surface — which are themselves sulfated sugar chains. The virus has evolved surface glycoproteins that grip negatively charged sugar. Flood the system with a competing negatively charged sugar and the virus grips that instead, never reaching the cell. This is a real, well-understood mechanism and it is why every sulfated polysaccharide tested tends to show antiviral activity in this assay: carrageenan from seaweed, dextran sulfate, fucoidan, heparin itself, pentosan polysulfate, and prunellin.

That last point is the crucial one and it cuts against the plant. Self-heal is not special here. It is one member of a large class of anionic polymers that all do the same thing in the same assay for the same physical-chemical reason. A finding that generalises across an entire chemical class is a finding about the class. Attributing it to Prunella vulgaris specifically — and then to self-heal tea — is precisely the substitution error that makes herbal pages overstate themselves.

The HSV Cell-Culture Work, Read Carefully

Three methodological cautions apply to this literature, and a reader who knows them will read any antiviral herb claim more accurately.

The assay is unusually generous to anionic polymers

In a dish, a virus particle and a cell membrane meet in a simple salt solution. On a real mucosal surface they meet through mucus — a thick layer of heavily glycosylated, itself negatively charged mucins — alongside secreted antibodies, shed epithelial cells, and in a lesion, serum proteins. Every one of those competes for a charge-based interaction. An entry blocker that works beautifully in a dish can be functionally invisible on a real surface, and that gap is the single commonest reason polyanion antivirals have failed to translate.

Selectivity has to be checked, not assumed

Anything that damages or stresses a cell will appear to reduce viral replication in that cell. Good antiviral papers therefore report a selectivity index — the ratio of the concentration that harms cells to the concentration that inhibits the virus. Plant-extract papers report this inconsistently. When you read that an extract "inhibited HSV at 50 micrograms per millilitre", the follow-up question is always: at what concentration did it start killing the cells?

"Extract" is not a defined substance

A hot-water infusion, a cold-water maceration, a lyophilised aqueous extract redissolved at a stated concentration, a 70% ethanol extract, and a chromatographically purified polysaccharide fraction are five different interventions. They contain different molecules in different ratios. The good news for self-heal is that here, unusually, the tradition and the science agree on solvent: the active fraction is water-soluble, and the traditional preparation is a water infusion or decoction. That is a point in the plant's favour and we should say so, because on many herbs the solvent mismatch quietly invalidates the entire comparison. The bad news is everything in the next two sections.

The Aciclovir-Resistant Strain Claim

You will see this claim presented as the headline: self-heal extract works even against herpes strains that resist the standard drug. It is reported in the literature and we have no reason to doubt it. It is also, on inspection, the least surprising thing on the page.

Aciclovir is a prodrug. It has to be phosphorylated by the virus's own thymidine kinase before it becomes active, and it then interferes with the viral DNA polymerase. Almost all clinical aciclovir resistance arises from mutations in that thymidine kinase — the virus loses or alters the enzyme that activates the drug — with a minority arising in the polymerase itself.

Now ask what an entry blocker cares about. Nothing in the entry step involves thymidine kinase or DNA polymerase. An agent that stops the virus reaching the cell would be expected, on first principles, to be completely indifferent to the mutations that cause aciclovir resistance. So "active against aciclovir-resistant strains" is not evidence of unusual potency. It is a restatement of the fact that the mechanism is different, and it would be equally true of carrageenan, dextran sulfate or plain heparin.

This matters clinically, because aciclovir-resistant HSV is overwhelmingly a problem of immunocompromised patients — people with advanced HIV, transplant recipients, those on chemotherapy. That is exactly the population for whom a herbal tea is the worst possible substitute for medical care, and for whom real alternatives exist and are used. Presenting a cell-culture finding as hope for that group would be irresponsible, and we are not going to do it.

The 1990s HIV Work and What Became of It

The prunellin story began as an HIV story, and the history of what happened next is the most useful thing on this page.

In the late 1980s, sulfated polysaccharides looked like a promising anti-HIV strategy for the same charge-based reason described above: HIV's gp120 also engages negatively charged cell-surface molecules, and polyanions blocked infection in culture at low concentrations. Dextran sulfate became the lead candidate of the class, was widely publicised, and reached human study.

Then it ran into the wall. When investigators actually measured what happened after oral administration in people, the answer — reported in Annals of Internal Medicine in 1989 — was that dextran sulfate is poorly absorbed after oral administration. A large, highly charged sugar polymer does not cross the intestinal epithelium in meaningful quantity. The clinical work in AIDS patients from the same period found no useful benefit, and the class was effectively abandoned as an oral antiviral.

This is not an obscure footnote. It is the direct, human, measured answer to the question "if I drink an infusion containing a sulfated polysaccharide, will it get into my blood and block a virus?" The answer, established in people nearly four decades ago for the best-studied member of the class, is essentially no. Heparin, the most familiar sulfated polysaccharide of all, has to be injected for exactly this reason.

So the oral route for prunellin is not merely untested. It is tested by proxy, in humans, and the proxy failed on pharmacokinetic grounds. That is a stronger and less flattering statement than "more research is needed", and it is the honest one.

The Mechanism Itself Was Tested in Humans

The strongest honest move available on any herb page is to find where the proposed mechanism was tested properly, even if the plant never was. For self-heal's antiviral claim, that test exists, at scale, in randomised controlled trials with thousands of participants. It was done on the topical route, which is the route the chemistry actually permits, and the results were disappointing.

Through the 2000s, sulfated and otherwise anionic polymers were the leading candidates for vaginal microbicides to prevent HIV transmission — a gel applied topically, at the site of exposure, at drug concentrations, exactly the scenario an entry blocker should win. The major trials:

• Carraguard, a carrageenan-based gel, went to a Phase 3 trial in South African women, published in The Lancet in 2008. It did not reduce HIV acquisition.
• Cellulose sulfate (Ushercell) trials were stopped early in 2007 when an interim analysis raised the possibility of more infections in the gel arm; the negative-efficacy result was published in the New England Journal of Medicine in 2008. The excess-risk signal was not confirmed in every subsequent analysis, and we will not overstate it — but the efficacy result was clear, and it was a failure.
• PRO 2000, a naphthalene sulfonate polymer working by the same polyanion mechanism, showed a non-significant hint of benefit in one trial and no benefit in the larger one that followed.

Read that in both directions, because both directions matter.

Against the claim: this is the nearest thing to a direct test of the prunellin hypothesis that exists. The mechanism was given every advantage — correct route, correct site, correct timing, purified compound, drug-level concentrations, thousands of participants, proper randomisation — and it did not prevent viral infection in humans. A self-heal infusion offers a weaker version of the same mechanism at a tiny fraction of the concentration. The inference is not favourable.

Not a scare about the tea: nothing here says self-heal is dangerous. These were concentrated gels, applied repeatedly to vaginal mucosa, at doses orders of magnitude above anything in a cup of herb tea. The cellulose sulfate safety signal belongs to that product in that setting and there is no basis whatever for transferring it to a plant that people have been eating as a wild green for centuries.

And one honest complication: the class is not uniformly negative. Iota-carrageenan nasal sprays have been trialled for the common cold with some positive results, including a placebo-controlled trial reported in Respiratory Research in 2010, and carrageenan gels have been studied against HPV transmission. The pattern that emerges is coherent: a sulfated polysaccharide can plausibly work where it is applied directly and continuously to the exposed surface at high concentration, and fails where transmission conditions are harsher or the dosing is intermittent. That pattern is not kind to a swallowed tea, but it does mean the mechanism is not dead — and it points at where a genuinely interesting self-heal study might one day be done.

Route and Dose: Why a Cup of Tea Is Not the Experiment

Pull the three threads together and the route problem is decisive.

1. The active fraction is not orally bioavailable. Large anionic polysaccharides do not cross the gut wall in meaningful amounts. Measured in humans, for this exact class.
2. Even if it were absorbed, the mechanism does not work systemically. An entry blocker has to be physically present between virus and cell at the moment of contact. For a reactivating cold sore, the virus travels down a sensory nerve and infects epithelium from the basal side outward — where no swallowed extract, and arguably no topically applied one, is waiting.
3. The concentration is not in the right range. Cell-culture work uses defined concentrations of purified fraction in a small volume of medium. A cup of infusion made from a few grams of dried herb delivers an unknown quantity of an uncharacterised mixture into 40-odd litres of body water. Even ignoring absorption entirely, the arithmetic does not begin to close.

The one route that survives this analysis is direct topical or mucosal contact — a cooled infusion held in the mouth, used as a gargle, or applied to a skin surface. There the extract is undiluted, unabsorbed and in contact with the tissue, which is exactly the condition the mechanism requires and exactly what the tradition actually did. That is a genuinely interesting convergence, and it is the honest place to point a curious reader. It is also still untested: nobody has run the trial.

Notice what has happened. Careful analysis of the modern antiviral chemistry ends up supporting the traditional topical and gargle use over the modern oral-supplement framing. Capsules and teas sold on the strength of the herpes research are being sold on a mechanism that cannot operate by that route. The old-fashioned mouthwash is the one with the coherent rationale.

What a Real Trial Would Look Like

The commonest defence of an untested herb is that traditional remedies cannot be trialled. For recurrent herpes labialis, that is simply false. The methodology is mature, the endpoints are validated, the trials are run routinely by pharmaceutical and consumer-health companies, and they are neither large nor expensive by modern standards. Here is exactly what one would look like.

Design

Randomised, double-blind, placebo-controlled, patient-initiated. Recruit adults with a history of at least three to four cold sores a year. Give each participant study product plus a diary and instructions to begin treatment at the first prodromal tingle. Because typical effect sizes for antivirals in this condition are on the order of half a day to a day and a half of healing time, several hundred participants are needed for adequate power — that is the main cost driver, and the reason under-powered herbal studies keep producing uninterpretable results.

Primary and secondary endpoints, all standard and all validated

• Time to lesion healing — loss of crust, assessed by a blinded examiner. The conventional primary endpoint.
• Time to cessation of pain and of all symptoms — participant-reported on a fixed scale.
• Proportion of episodes aborted at the prodrome — the outcome patients actually care most about.
• Maximum lesion area or vesicle count — measured, not estimated.
• Viral shedding — quantitative PCR on swabs, which turns a subjective assessment into a number.
• For a suppression trial: annualised recurrence rate, time to first recurrence, and proportion recurrence-free at six and twelve months.
• For genital HSV-2: subclinical shedding rate measured by daily self-collected swabs with PCR — a well-established technique.

Comparators

Placebo at minimum; ideally a three-arm design against topical or oral aciclovir, so that a positive result can be sized against standard care rather than merely declared.

None of this has been done with Prunella vulgaris. We searched for registered and published trials and could not find one; readers can verify the current position themselves at ClinicalTrials.gov and on the PubMed searches below. The tools exist. They have been pointed at lemon balm, at lysine, at honey, at docosanol, at zinc oxide, and at a dozen pharmaceutical candidates. They have never been pointed at this plant.

Comparators That Do Have Human Trials

Naming these is not a way of dismissing self-heal. It is the direct rebuttal to the "nobody trials traditional remedies" defence, since every item below is a traditional or over-the-counter remedy that got tested anyway.

• Lemon balm (Melissa officinalis) — the closest possible comparator, a mint-family herb with a similar rosmarinic acid profile and a similar traditional reputation. A standardised balm-mint cream was tested in randomised, placebo-controlled trials for recurrent herpes labialis, with results reported in Phytomedicine in the 1990s. Whatever one makes of the size of the effect, the point stands: a mint-family herb for cold sores was trialled properly, so self-heal could have been.
• Lysine — multiple small randomised trials since the 1980s, with genuinely mixed and partly contradictory results. This is what a claim looks like when it has been tested and the answer is uncertain, as opposed to untested.
• Topical docosanol 10% — an over-the-counter cream with a multicentre randomised placebo-controlled trial behind it, reported in the Journal of the American Academy of Dermatology in 2001. The measured benefit was modest, roughly half a day of healing time. Useful calibration: this is the size of effect that a well-run trial detects in this condition, and it is far smaller than the impression herbal marketing usually creates.
• Aciclovir, valaciclovir and famciclovir — the reference standard, with a very large trial literature for episodic treatment and for suppression of recurrences.

Set self-heal beside that list and the position is clear. It is not that self-heal lost a fair fight. It is that it has never entered one.

Absent, Negative, and Where Each Claim Sits

Sorting the antiviral claims into the two columns that actually matter:

Established (laboratory)

1. Water extracts and purified polysaccharide fractions of Prunella vulgaris inhibit HSV-1 and HSV-2 in cell culture, reproducibly and across independent groups.
2. The mechanism is consistent with blockade of viral attachment or entry.
3. The plant contains sulfated polysaccharides capable of that action.

Negative or strongly unfavourable (human data, borrowed but directly relevant)

1. Oral bioavailability of the active class. Measured in humans for dextran sulfate and found to be poor. This undercuts any oral self-heal antiviral claim on pharmacokinetic grounds.
2. Topical polyanion antivirals for preventing viral infection. Large randomised HIV microbicide trials of carrageenan and cellulose sulfate gels failed. Correct route, correct site, drug doses, thousands of participants.
3. Partial counterweight: iota-carrageenan nasal sprays have shown benefit for the common cold, so continuous direct application to an exposed surface may still be viable. This is the one door the human data leaves open.

Absent (never tested)

1. Any human trial of self-heal for cold sores, genital herpes, shingles, or any other viral illness — oral or topical.
2. Any human pharmacokinetic study of any self-heal constituent.
3. Any dose-finding work, so nobody can say what a therapeutic dose would even be.
4. Any standardisation of the polysaccharide content of commercial self-heal material, so two products need not contain comparable amounts of the compound the claim rests on.
5. Any animal model of HSV lesion healing using the traditional preparation. We looked for a controlled animal efficacy study of this kind and could not verify one; the animal literature on this plant is thinner than the cell-culture literature, not richer.

The bottom line. The herpes research on self-heal is real science and it is genuinely interesting. It supports one narrow, honest conclusion: Prunella vulgaris contains a sulfated polysaccharide that behaves like other sulfated polysaccharides in an entry-inhibition assay. It does not support taking self-heal for a herpes infection, and the human data on the mechanism argues actively against the oral route. If you have recurrent cold sores or genital herpes, the treatments with trials behind them work, are cheap, and are safe. Enjoy the tea for what it is.

Key Research Papers

All links are live PubMed searches by title or topic rather than fixed identifiers. Where we could not verify a paper's exact details we describe the finding and link the topic instead.

1. Tabba HD, Chang RS, Smith KM. "Isolation, purification, and partial characterization of prunellin, an anti-HIV component from aqueous extracts of Prunella vulgaris." Antiviral Research, 1989. The origin of prunellin. Note the words "partial characterization" in the title of the founding paper — that is where the compound's definition still largely stands. — PubMed search
2. Xu HX, Lee SHS, Lee SF, White RL, Blay J. "Isolation and characterization of an anti-HSV polysaccharide from Prunella vulgaris." Antiviral Research, 1999. The purification that tied the antiviral activity to the polysaccharide fraction, including the report of activity against an aciclovir-resistant strain. — PubMed search
3. Chiu LCM, Zhu W, Ooi VEC. "A polysaccharide fraction from medicinal herb Prunella vulgaris downregulates the expression of herpes simplex virus antigen in Vero cells." Journal of Ethnopharmacology, 2004. Independent confirmation with a downstream readout. Vero cells are a monkey kidney line — standard for HSV work, and not human epithelium. — PubMed search
4. Psotová J, Kolář M, Soušek J, et al. "Biological activities of Prunella vulgaris extract." Phytotherapy Research, 2003. Antioxidant and anti-HSV activity in the same extract, from a group working on the plant's skin-cell effects. — PubMed search
5. Lorentsen KJ, Hendrix CW, Collins JM, et al. "Dextran sulfate is poorly absorbed after oral administration." Annals of Internal Medicine, 1989. The single most important citation on this page. Human pharmacokinetics for the class, measured directly, with an unambiguous answer. — PubMed search
6. Oral dextran sulfate in AIDS and AIDS-related complex. The clinical arm of the same story, reported in Annals of Internal Medicine in 1989. The class did not deliver. — PubMed search
7. Skoler-Karpoff S, et al. "Efficacy of Carraguard for prevention of HIV infection in women in South Africa: a randomised, double-blind, placebo-controlled trial." The Lancet, 2008. The carrageenan microbicide Phase 3. The mechanism, given every advantage, in thousands of women. — PubMed search
8. Van Damme L, et al. "Lack of effectiveness of cellulose sulfate gel for the prevention of vaginal HIV transmission." New England Journal of Medicine, 2008. The trial that was stopped early. Read the paper rather than the headlines for what the interim safety signal did and did not establish. — PubMed search
9. PRO 2000 vaginal gel trials. The other major polyanion candidate, with a suggestive smaller result and a null larger one. — PubMed search
10. Eccles R, et al. Iota-carrageenan nasal spray versus placebo in early treatment of the common cold in adults. Respiratory Research, 2010. The counterweight: the same mechanism, applied continuously to the exposed surface, with a positive result. — PubMed search
11. Koytchev R, Alken RG, Dundarov S. "Balm mint extract (Lo-701) for topical treatment of recurring herpes labialis." Phytomedicine, 1999. Proof that a mint-family herb for cold sores can be, and was, properly trialled. — PubMed search
12. Sacks SL, et al. Clinical efficacy of docosanol 10% cream for herpes simplex labialis: a multicentre randomised placebo-controlled trial. Journal of the American Academy of Dermatology, 2001. Calibration for how large a real effect in this condition actually is. — PubMed search
13. Aciclovir resistance mechanisms in herpes simplex virus. The thymidine kinase and DNA polymerase literature behind the resistant-strain discussion above. — PubMed search
14. Sulfated polysaccharides as antiviral agents, and their translation problem. The class review literature, which is where the pattern described on this page is set out properly. — PubMed search
15. Live search for any human trial of Prunella vulgaris in viral infection. Please run it. If this page ever goes out of date, this is the link that will show it. — PubMed search

Connections

• All Herbs
• Self-Heal (Prunella vulgaris) — the main topic page for the plant
• The Name and What It Promises — why expectations arrive before the evidence
• Sore Throat and Topical Use — the route the chemistry actually supports
• Rosmarinic Acid, Safety and Unknowns — what is genuinely not known
• Lemon Balm: Antiviral and HSV — the mint-family comparator that was actually trialled
• Lemon Balm — another rosmarinic-acid-rich Lamiaceae herb
• Lysine: Herpes Simplex and Cold Sores — a claim that has been tested, with mixed results
• Lysine — the amino acid behind the best-known cold-sore supplement
• Herpes Simplex — the infection itself, and the treatments with evidence behind them
• Herpes Zoster — shingles, a different herpesvirus and a different urgency
• Elderberry — another traditional antiviral herb, with its own evidence problems
• Immune Boosting — what the phrase does and does not mean

Back to Table of Contents