The architecture of the skin microbiome
Human skin is not a uniform environment. The 1.5 m² surface is divided into ecologically distinct niches defined by temperature, pH, moisture content, and lipid availability, and each niche selects for different microbial communities. Sebaceous follicles — hair follicles associated with sebaceous glands — are rich in lipids and colonised predominantly by Cutibacterium acnes (previously Propionibacterium acnes), an obligate anaerobe that metabolises triglycerides to free fatty acids that maintain the acid mantle of the skin and inhibit many potential pathogens. Moist areas including the axillae, inguinal folds, and interdigital spaces favourStaphylococcus epidermidis, Corynebacterium species, and other Gram-positive cocci that can tolerate the osmotic stress of secreted eccrine sweat. Dry areas like the forearm and back are more diverse, with representation fromMalassezia fungi, Micrococcus species, and lower-abundance communities that shift considerably between individuals and seasons.
This spatial organisation is not static. The microbiome changes across the lifespan: neonatal skin is colonised at birth (primarily with maternal vaginal or skin flora depending on delivery mode), the composition shifts through childhood and puberty (sebum production increases, enriching Cutibacterium), and continues to evolve in adulthood. Diet, personal care products, topical medications, environmental exposures, and genetic factors all shape individual skin microbiome composition. The key point for understanding atopic dermatitis is that a healthy skin microbiome — dominated by commensal staphylococci and cutibacteria — actively maintains barrier function and local immune homeostasis. When that community is disrupted, the consequences can be severe.
Atopic dermatitis and the Staphylococcus aureus bloom
Atopic dermatitis (AD) is a chronic inflammatory skin disease affecting approximately 10–20% of children and 5–10% of adults in high-income countries, with rising prevalence consistent with the hygiene hypothesis. Its hallmarks are intense pruritus, disrupted skin barrier, and cycles of remission and flare. One of the most reproducible microbiological findings in AD is the bloom of Staphylococcus aureus: in healthy skin, S. aureus constitutes less than 5% of the bacterial community in most studies; in lesional AD skin, it can constitute up to 90% of the microbiome during flares. This is not simply a correlation — the density of S. aureuscolonisation correlates with AD disease severity, and targeted antibiotic treatment of S. aureus improves AD outcomes even when not accompanied by other interventions.
The mechanistic story of how S. aureus drives AD pathology involves multiple virulence factors acting on distinct aspects of skin barrier and immune function. Delta-toxin (encoded by the hld gene, part of the RNAIII regulatory system) is a surfactant-like amphipathic peptide that directly activates mast cells to release pre-formed IL-4 and IL-13 — key Th2 cytokines that skew the immune response away from the Th1/Th17 profile needed for effective antimicrobial defence and toward the Th2 allergic profile characteristic of AD. This creates a positive feedback loop: Th2 cytokines suppress antimicrobial peptide production (discussed below), allowing more S. aureus proliferation, which generates more delta-toxin, which drives more Th2 skewing.
V8 protease (glutamyl endopeptidase, encoded by sspA) cleaves desmoglein-1 (Dsg1), a desmosomal cadherin essential for keratinocyte cohesion in the superficial epidermis. Dsg1 cleavage by V8 protease disrupts cornified envelope integrity in a manner mechanistically similar to, though less severe than, the exfoliative toxins of bullous impetigo. In AD patients with the compromised barrier associated with filaggrin mutations, even partial Dsg1 cleavage dramatically worsens transepidermal water loss and allergen penetration. Alpha-toxin (alpha-hemolysin) is pore-forming and directly cytotoxic to keratinocytes; in AD skin, which has reduced sphingomyelin content compared to healthy skin, keratinocytes are more susceptible to alpha-toxin cytotoxicity.
Filaggrin mutations and the compromised barrier
The genetics of AD converge substantially on filaggrin (FLG). FLG encodes a large protein (profilaggrin) that is processed during terminal keratinocyte differentiation into multiple filaggrin repeat units. Filaggrin has two main functions: it aggregates keratin filaments in the stratum granulosum, providing mechanical structure to the cornified envelope; and it is ultimately degraded to the component amino acids — pyrrolidone carboxylic acid, trans-urocanic acid, and others — that constitute the natural moisturising factor (NMF) of the stratum corneum, maintaining water-holding capacity and acidic pH.
Common FLG null mutations (R501X and 2282del4 in European populations; different variants in Asian populations) reduce filaggrin production, impairing both the structural and NMF functions. The result is a barrier with elevated pH (less acid mantle), reduced ceramide content, higher transepidermal water loss, and enlarged intercellular spaces through which allergens and S. aureus virulence factors penetrate more readily. FLG heterozygous mutations increase AD risk approximately threefold; homozygous null mutations confer very high penetrance for severe AD. Importantly, FLG null mutations are found in only about 30% of AD patients, indicating that the barrier defect can arise through other mechanisms — including microbiome disruption and Th2 cytokine-driven suppression of ceramide synthesis.
Staphylococcus epidermidis: the commensal protector
The dramatic S. aureus bloom in AD skin is accompanied by a reciprocal decline in commensal staphylococci, particularly S. epidermidis. This matters because S. epidermidis is not merely a passive bystander: it actively competes with S. aureus and provides several layers of protection against pathogen colonisation.
FAME (fatty acid-modifying enzyme) systems in S. epidermidis esterify unsaturated fatty acids (UFAs) liberated from sebum, producing fatty acid methyl esters and fatty acid ethyl esters. Some of these products have direct antimicrobial activity against S. aureus that is greater than the parent UFAs, and additionally activate TLR2 on keratinocytes to stimulate controlled, commensal-type innate immune activation without inducing the inflammatory cascade that pathogen encounter triggers. Serine-rich repeat proteins (SRRPs) on S. epidermidissurfaces compete with S. aureus for epithelial adhesion sites, reducing the density of S. aureus that can physically attach to the skin surface.
Perhaps most importantly, S. epidermidis produces Esp, a serine protease that specifically degrades the S. aureus biofilm matrix and inhibitsS. aureus nasal colonisation. In an elegant set of experiments, Iwase and colleagues demonstrated in 2010 that nasal colonisation with Esp-producingS. epidermidis strains inversely correlated with S. aureusnasal carriage in human subjects, and that purified Esp protein disrupted establishedS. aureus biofilms in vitro. The same mechanisms appear to operate on skin surface colonisation. The practical implication is that interventions supportingS. epidermidis abundance should, by competition and direct Esp-mediated biofilm disruption, reduce S. aureus colonisation density.
Antimicrobial peptides: the endogenous defence landscape
The skin produces a suite of endogenous antimicrobial peptides (AMPs) that contribute to the first line of defence against pathogen colonisation. Human beta-defensins 2 and 3 (HBD-2, HBD-3) are produced by keratinocytes under innate immune stimulation and are broad-spectrum: they disrupt bacterial membranes through electrostatic interaction with anionic phospholipids and are active against S. aureus, Gram-negative pathogens, and some fungi. HBD-2 is strongly induced by IL-1β, TNF-α, and bacterial LPS; HBD-3 is additionally induced by IL-17A and IFN-γ.
Cathelicidin LL-37 is the mature antimicrobial peptide derived from proteolytic cleavage of the human cathelicidin hCAP18. LL-37 is produced by keratinocytes, neutrophils, and mast cells. It is a key AMP against S. aureus: it disrupts the bacterial membrane and additionally neutralises LPS and reduces biofilm formation. Critically, LL-37 production in skin is suppressed by Th2 cytokines IL-4 and IL-13 — the same cytokines that are elevated in AD. This creates a direct mechanistic link between the Th2-skewed immune environment of AD and the failure of endogenous AMP-mediated defence against S. aureus. Dermcidin, a constitutively produced AMP in eccrine sweat that is effective across a broad pH range, rounds out the key components of the endogenous antimicrobial defence.
Zinc is a cofactor for multiple components of this AMP system. Zinc deficiency reduces LL-37 production and impairs defensin induction, and conversely, adequate zinc status supports robust AMP production in response to skin colonisation events.
How spirulina connects: anti-Staphylococcal activity
Phycocyanin has direct antimicrobial activity against S. aureusin vitro. The mechanism appears to be dual: photodynamic antimicrobial activity, in which phycocyanobilin generates singlet oxygen and reactive oxygen species upon light absorption, disrupting bacterial membranes; and direct, non-photodynamic membrane disruption at higher concentrations. The minimum inhibitory concentrations reported in in vitro studies vary considerably (generally in the 100–500 μg/mL range for direct antimicrobial activity), which are higher than concentrations achievable in serum through oral supplementation but potentially relevant for topical application.
Several research groups have investigated topical spirulina or phycocyanin preparations for skin applications. These are early-stage investigations and the human clinical evidence is very limited, but the rationale for direct antimicrobial and anti-inflammatory effects at the skin surface is mechanistically coherent.
The GLA pathway and Th2 modulation
Spirulina is one of the richest dietary sources of gamma-linolenic acid (GLA), an omega-6 fatty acid that bypasses the rate-limiting delta-6-desaturase step in the arachidonic acid pathway to yield dihomo-gamma-linolenic acid (DGLA). DGLA is the precursor to the 1-series prostaglandins (PGE1) and 15-HETrE, which have anti-inflammatory profiles contrasting with the pro-inflammatory 2-series prostaglandins derived from arachidonic acid. Importantly, DGLA competes with arachidonic acid for COX enzyme activity, and PGE1 produced from DGLA stimulates cAMP in T cells — which, in the specific context of Th2 differentiation, can reduce IL-4 and IL-13 production.
The clinical relevance of this pathway for AD is supported by several lines of evidence. GLA deficiency has been proposed as a contributing factor in AD (the delta-6-desaturase hypothesis), and some clinical trials of evening primrose oil (another GLA source) in AD showed modest benefit. The GLA content of spirulina at typical doses (1–3 g/day providing roughly 30–90 mg GLA) is comparable to that in evening primrose oil capsule studies. Spirulina’s GLA-derived pathway modulation of the Th2 cytokine environment — reducing the IL-4/IL-13 milieu that suppresses LL-37 production and favours S. aureus colonisation — represents a systemic mechanism that could complement any direct antimicrobial effects.
Zinc and antimicrobial peptide support
Spirulina contains meaningful quantities of zinc — approximately 0.3–0.5 mg per gram of dried powder, yielding 1–2.5 mg of zinc from a typical 3–5 g daily dose. Zinc is a cofactor for numerous immune functions, but its relevance here is specifically the support of LL-37 production. Zinc-deficient states are associated with reduced cathelicidin expression in skin and neutrophils; zinc supplementation in deficient individuals restores AMP production. For people with low baseline zinc status — which includes a meaningful proportion of AD patients, particularly those on restricted diets — spirulina’s zinc contribution may support LL-37 production and thereby partially counteract the Th2-mediated AMP suppression characteristic of AD.
The caveat is that 1–2.5 mg zinc from spirulina is a modest contribution relative to the recommended dietary intake of 8–11 mg/day for adults, and spirulina’s zinc bioavailability has not been precisely characterised. Phytic acid, which can chelate zinc, is present in spirulina, but at lower levels than in seed-based plant foods, and spirulina’s zinc likely has reasonable bioavailability. This is a supporting rather than a primary mechanism.
Honest limitations
The mechanistic case for spirulina’s relevance to AD and skin microbiome health is coherent: anti-Staphylococcal activity through phycocyanin (primarily topically relevant), Th2 modulation through GLA-derived lipid mediators, and zinc support for AMP production. However, the specific human clinical evidence in AD is essentially absent from the published literature. No well-controlled clinical trial has examined spirulina supplementation in atopic dermatitis patients with microbiome composition and AD severity as co-primary endpoints.
The comparison point here is dupilumab — the IL-4Rα antagonist that blocks both IL-4 and IL-13 signalling and achieves 50–75% improvement in EASI scores in moderate-to-severe AD. That level of cytokine pathway targeting far exceeds what any dietary intervention can achieve. Spirulina is most plausibly positioned as a general anti-inflammatory dietary contribution that supports the broader immune environment, rather than as a targeted therapeutic for active AD.