Mold Toxicity Is More Than Exposure: The Mitochondria–Gut–Immune Cascade Behind Chronic Illness

From my research into mold, mycotoxins, mitochondrial physiology, gut ecology, immune signaling and host genetics, and from the patterns I continue to examine while working through complex gastrointestinal, autonomic, neurological, mast-cell and post-infectious cases, one question keeps coming back to me: why can two people encounter what appears to be a similar exposure and experience completely different biological outcomes? One person may encounter mold or contaminated food and recover without an obvious long-term consequence, while another gradually develops bloating, food intolerance, recurrent SIBO, abnormal bowel motility, profound fatigue, exercise intolerance, neurological symptoms, dysautonomia, chemical sensitivity, mast-cell-type reactions or an increasingly narrow ability to tolerate things that were previously completely normal.

The usual explanation is that one person simply “detoxifies mold better,” but I think that description misses most of the biology. Detoxification is not one pathway. It is a sequence of biochemical handoffs. A toxin first has to enter the body and cross a biological barrier. Enzymes then transform it. Sometimes that transformation makes the compound easier to remove, but sometimes it creates a molecule that is substantially more reactive than the original toxin. The host then has to neutralize that reactive intermediate before it reaches DNA, proteins or cellular membranes. Glutathione has to be available and continuously replenished. Mitochondria have to provide enough ATP to maintain cellular repair and membrane transport. Conjugated metabolites have to leave the cell. The liver has to move them toward bile. The intestine has to preserve its epithelial barrier, maintain normal microbial ecology and keep material moving toward elimination. Meanwhile, the immune system has to respond to injury without becoming trapped in a self-amplifying inflammatory state.

This is the foundation of what I call the Host Capacity Model. The exposure matters, but the host is not passive. The final biological effect of a toxin depends partly on what happens after that molecule encounters the metabolism, genetics, mitochondria, immune system and intestinal environment of that particular person.

Aflatoxin B1, abbreviated AFB1, is one of the clearest examples of this. AFB1 is produced primarily by aflatoxigenic strains of Aspergillus flavus and Aspergillus parasiticus and can contaminate agricultural foods such as corn, peanuts, grains and tree nuts. What makes AFB1 especially interesting is that the compound entering the body is not necessarily the most reactive molecule in the pathway. The host’s own metabolism can create something considerably more dangerous.

When AFB1 is ingested, it first enters the gastrointestinal lumen. At this point it is physically inside the digestive tract but remains separated from the internal environment of the body by a single layer of intestinal epithelial cells. That thin layer separates intestinal contents from blood vessels, immune cells, enteric nerves and deeper tissues. The intestine is therefore not merely a tube delivering the toxin to the liver. Enterocytes contain mitochondria, cytochrome P450 enzymes, antioxidant machinery, nuclear receptors, xenobiotic transporters and tightly regulated junctional structures. The interaction between AFB1 and the host can begin directly at the intestinal wall.

Once AFB1 crosses the epithelium, it enters portal circulation and travels toward the liver. This arrangement normally provides an important protective advantage because hepatocytes can metabolize foreign compounds before they reach the wider systemic circulation. AFB1, however, reveals an important paradox in toxicology: metabolism does not always equal detoxification.

Two of the major enzymes involved are CYP1A2 and CYP3A4, with CYP3A5 contributing as well. These belong to the cytochrome P450 family. Phase I metabolism is often casually described as detoxification, but Phase I simply changes the chemical structure of a molecule. Whether the result is safer depends entirely on what is produced.

With AFB1, CYP-mediated oxidation of the 8,9 double bond can generate Aflatoxin B1-8,9-epoxide, or AFBO. The exo form of this epoxide is particularly reactive. This is the first critical transition in the entire story. The mold produces AFB1, but the host can produce AFBO.

AFBO is an electrophile. In simple terms, that means it is electron deficient and strongly attracted to electron-rich structures inside the cell. DNA contains those targets. Proteins contain them. Glutathione contains a highly reactive sulfur group capable of capturing electrophiles as well. Once AFBO appears, the cell enters a molecular race. Does glutathione intercept AFBO first, or does AFBO reach DNA or proteins before the protective systems can capture it?

This is where I think the concept of a kinetic mismatch becomes useful. Imagine Phase I metabolism as a faucet producing AFBO and Phase II conjugation as the drain removing it. If both operate at similar rates, free AFBO remains relatively controlled. If Phase I accelerates while downstream neutralization remains limited, reactive epoxide begins accumulating. The external exposure may be the same in two people, yet the amount of free AFBO present inside their cells can be completely different.

This is where host genetics begins to matter. One variant I follow is CYP1A2 rs762551. In the genomic framework I am studying, A/A represents the more highly inducible configuration. The significance is not that A/A automatically means somebody will develop mold illness. The important question is whether a highly inducible CYP1A2 system can increase AFB1 bioactivation under the relevant environmental conditions. If Phase II glutathione conjugation can keep pace, the system may remain balanced. If AFBO formation accelerates while downstream defense cannot match that speed, the intracellular burden of reactive epoxide rises.

CYP3A4 creates another important branch. One variant I follow is rs2740574, associated with CYP3A4*1B. CYP3A4 is abundant in the liver and is also expressed in intestinal enterocytes, meaning part of the bioactivation can occur within the gut-liver axis rather than only after the toxin reaches hepatocytes. CYP3A4 can move AFB1 toward different metabolic products, including destructive AFBO but also less reactive metabolites such as AFQ1. This is why I do not think of metabolic enzymes as simply “good” or “bad.” Their significance depends on which direction they push chemical flux and whether the next pathway has enough capacity to handle what they produce.

CYP3A5 adds another layer. One SNP I follow is rs776746. The A allele corresponds to the CYP3A51 expression phenotype at this locus, whereas the common G allele creates the splice-defective CYP3A53 allele. This becomes particularly interesting because human aflatoxin studies have linked greater predicted CYP3A5 expression with higher aflatoxin-albumin adduct burden. So even before we discuss antioxidant defense, two people may already be generating different quantities of reactive metabolites from the same original compound.

Once AFBO is formed, the host needs mechanisms to stop it. One of the most important systems involves glutathione S-transferases, or GSTs. GST enzymes bring glutathione together with electrophilic molecules and help form conjugates that are less capable of attacking cellular structures. One variant I follow is GSTP1 rs1695, the Ile105Val polymorphism. The G allele changes the amino acid at position 105 and alters the catalytic properties of the GSTP1 enzyme. What matters to me is not this SNP by itself but whether altered GST activity occurs at the same time as increased Phase I bioactivation.

But even an intact GST enzyme cannot continue protecting the cell if glutathione itself becomes depleted. Glutathione is synthesized from glutamate, cysteine and glycine. One of the major rate-limiting steps is controlled by glutamate-cysteine ligase, whose catalytic component is encoded by GCLC. The variant I follow here is rs17883901, the −129 C>T promoter polymorphism. This brings us to an important distinction between glutathione concentration and glutathione reserve. A cell can have an adequate pool under resting conditions yet still fail when demand suddenly rises. If AFBO production accelerates while oxidative stress simultaneously consumes antioxidant capacity, the important question becomes whether the cell can synthesize new glutathione quickly enough to replace what is being used.

There is another defense system involving EPHX1, microsomal epoxide hydrolase 1. The variants I follow here are rs1051740, producing Tyr113His, and rs2234922, producing His139Arg. EPHX1 can chemically hydrolyze epoxides by adding water across the epoxide ring. In the case of AFBO, this can produce AFB1-8,9-dihydrodiol. I think of this as an additional structural defense. If glutathione-dependent capture fails to intercept every reactive epoxide, epoxide hydrolysis offers another metabolic fate.

However, this does not mean the problem simply disappears. AFB1-8,9-dihydrodiol can undergo further chemistry toward reactive dialdehyde species capable of modifying proteins. EPHX1 therefore changes the form of the toxic burden rather than making the entire pathway harmless. This illustrates something I think is fundamental in toxicology: every metabolic rescue pathway produces another molecule that still has to be processed.

If AFBO escapes glutathione conjugation and epoxide hydrolysis, DNA becomes one of its major targets. AFBO can react with the N7 position of guanine and create an AFB1-N7-guanine adduct. An adduct means that an aflatoxin-derived chemical structure has become physically attached to the DNA molecule. DNA is not merely information written as letters. It is a three-dimensional molecular structure. Polymerases have to move across it. Repair proteins constantly inspect it. Transcription machinery has to read it. When a bulky foreign molecule becomes attached to guanine, that geometry changes.

Replication machinery may stall or misread the damaged template. Repair systems have to recognize and remove the lesion. If the damage survives until replication, mutations become more likely. One of the best-known signatures associated with substantial aflatoxin exposure is a guanine-to-thymine transversion affecting TP53, particularly codon 249 in hepatocellular carcinoma.

Think about the depth of that pathway. An Aspergillus species produces a secondary metabolite outside the human body. The molecule enters the intestine, crosses the epithelium, reaches the liver through portal blood, is transformed by human CYP enzymes into AFBO, enters the nucleus and chemically attaches itself to guanine. What began as an environmental molecule has now become a structural problem inside the human genome.

At this stage another genetic layer becomes important: DNA repair. Once damage occurs, the question is no longer only how efficiently the toxin was detoxified but how efficiently the host can repair what detoxification failed to prevent. One marker I follow is XRCC3 rs861539, the Thr241Met variant. XRCC3 participates in homologous recombination and genomic repair. Another is XRCC4 rs28383151, involved in non-homologous end joining and repair of double-strand DNA breaks. Additional DNA-repair variants I follow include XRCC1 rs25487, PRKDC/XRCC7 rs7003908, ERCC2/XPD rs13181 and XPC rs2228001.

These genes add another dimension to the Host Capacity Model. Two people could produce similar amounts of AFBO, yet one may repair DNA damage more efficiently. Susceptibility therefore exists both before and after injury. One set of pathways determines whether the toxic lesion occurs; another set determines whether the lesion persists.

At the same time, AFB1-related oxidative stress can affect mitochondria. This is where the biology begins expanding beyond toxicology into cellular energetics. Mitochondria are often described as the powerhouses of the cell, but I think it is more useful to describe them as the machinery that determines cellular work capacity. ATP powers the sodium-potassium ATPase, calcium pumps, protein processing, DNA repair, membrane transport, bile secretion, intestinal epithelial maintenance and neuronal repolarization. When ATP falls, many systems that appear separate begin losing function together.

Inside the inner mitochondrial membrane sits the electron transport chain. NADH supplies electrons primarily through Complex I. Succinate supplies electrons through Complex II. Those electrons move through ubiquinone, Complex III, cytochrome c and Complex IV, where oxygen functions as the final electron acceptor. Complexes I, III and IV use the energy of electron transfer to pump protons across the inner mitochondrial membrane. This establishes the mitochondrial membrane potential, or ΔΨm. ATP synthase then allows those protons to flow back toward the mitochondrial matrix and captures the energy to generate ATP.

When oxidative toxic stress begins disturbing electron transport, electrons can escape prematurely and react with molecular oxygen, producing superoxide. Superoxide can participate in the formation of hydrogen peroxide and other reactive oxygen species. Mitochondrial membrane lipids undergo oxidative damage. Respiratory proteins become injured. Mitochondrial DNA becomes vulnerable. The respiratory chain becomes less efficient, which creates even more electron leak.

A feedback loop develops. Mitochondrial injury produces more reactive oxygen species, and those reactive oxygen species create more mitochondrial injury. As the mitochondrial membrane potential declines, ATP production falls.

This becomes critically important because the host requires ATP to handle the toxic burden. The liver needs ATP to operate transport proteins. DNA repair requires energy. The intestinal barrier requires ATP to maintain itself. Neurons require enormous amounts of ATP to restore ionic gradients after firing. Smooth muscle requires energy for coordinated contraction and relaxation. The toxin can therefore damage the same cellular machinery the host needs in order to recover from the toxin.

Another mitochondrial pathway I continue to study is the PGC-1α, NRF1 and TFAM axis. PGC-1α coordinates mitochondrial biogenesis and oxidative metabolism. NRF1 controls many nuclear genes required for mitochondrial function. TFAM is critical for maintaining, organizing, replicating and transcribing mitochondrial DNA. If toxic stress damages existing mitochondria while this renewal network is also suppressed, the host is dealing with two problems simultaneously: mitochondria are being lost and the ability to replace them is weakening. This may help explain why ending an exposure does not necessarily produce immediate recovery. The exposure can stop before the cell has rebuilt the machinery that was damaged.

This mitochondrial problem becomes particularly interesting in the gut. Healthy colonocytes have an intimate metabolic relationship with anaerobic bacteria. Many beneficial colonic organisms ferment substrates and generate the short-chain fatty acid butyrate. Colonocytes absorb that butyrate and oxidize it inside mitochondria. This produces ATP, but it also performs another important function: it consumes oxygen.

Colonocyte oxygen consumption helps maintain physiological hypoxia near the epithelial surface and contributes to the highly anaerobic conditions of the colonic lumen. In other words, the microbiome feeds the host, and host metabolism creates the environment that favors that microbiome. Anaerobic organisms produce butyrate. Colonocytes oxidize it. Oxygen is consumed. Low oxygen favors strict anaerobes. The system reinforces itself.

Now imagine mitochondrial oxidative metabolism becoming impaired. Colonocytes consume less oxygen. The epithelial oxygen gradient begins to shift. At the same time, inflammation can increase inducible nitric oxide synthase, or iNOS, changing local nitrogen chemistry and creating respiratory opportunities such as nitrate.

The microbial environment has now changed.

Facultative anaerobes such as Escherichia coli, Klebsiella and other Proteobacteria possess metabolic flexibility that strict anaerobes lack. They can ferment when oxygen is unavailable but can switch toward oxygen or nitrate respiration when those electron acceptors become available. This gives them a substantial energetic advantage in an inflamed, increasingly oxygenated mucosal environment.

This is one reason I do not interpret dysbiosis simply as the presence of “bad bacteria.” I want to understand why the ecosystem selected those organisms. If epithelial mitochondrial metabolism changes, local oxygen utilization changes. If inflammation increases nitrate availability, facultative organisms gain another respiratory advantage. If Gram-negative organisms expand, more lipopolysaccharide, or LPS, becomes available.

LPS activates TLR4, one of the major receptors used by the innate immune system to recognize Gram-negative bacterial material. TLR4 signaling can activate MyD88, TRIF and NF-κB pathways and increase inflammatory mediators such as TNF-α, IL-1β, IL-6 and iNOS. The resulting inflammation can further alter epithelial metabolism and mitochondrial function. The original toxin has therefore created conditions capable of maintaining an ecological and inflammatory feedback loop even after the initial exposure has changed.

This is why, when I look at recurrent SIBO or persistent dysbiosis, I do not only ask which organism should be removed. I also ask what biological environment is continuously selecting for it.

The intestinal barrier adds another major part of the story. One pathway I focus on involves FXR and Myosin Light-Chain Kinase, or MLCK. FXR is often discussed primarily as a bile-acid receptor, but it also participates in intestinal homeostasis. Experimental AFB1-related intestinal injury has been associated with reduced FXR signaling, increased MLCK activity and disruption of junctional proteins such as ZO-1, occludin and claudins.

The intestinal barrier is not static glue. Tight-junction proteins are mechanically connected to the actomyosin cytoskeleton. MLCK phosphorylates myosin light chains and increases actomyosin contraction. When that contraction becomes excessive, the cells begin placing mechanical force on their own junctional structures.

In simple terms, the epithelial cells can begin pulling their own barrier apart.

Once permeability increases, the problem expands beyond AFB1. LPS, microbial metabolites, food-derived antigens and other luminal molecules can gain greater access to immune cells and deeper tissue. The original toxicological problem has now created an immunological problem.

More LPS activates more TLR4. NF-κB signaling increases. TNF-α and IL-1β rise. iNOS increases. At the same time, damaged mitochondria may be generating larger amounts of superoxide. Nitric oxide produced through iNOS can react with superoxide to generate peroxynitrite. Peroxynitrite can modify proteins, damage membrane lipids and impair mitochondrial enzymes.

Now inflammation worsens mitochondrial function, and impaired mitochondria intensify inflammatory redox chemistry.

The nervous system can become part of the same process. The intestinal wall contains large numbers of sensory neurons, including unmyelinated C fibers. These nerves can release neuropeptides such as substance P and CGRP. Mast cells are positioned close to sensory nerves and can respond not only through IgE-mediated pathways but also through receptors such as MRGPRX2.

An inflamed epithelial environment can sensitize sensory fibers. Those fibers release neuropeptides. Mast cells respond with histamine, tryptase, prostaglandins and other mediators. Those mediators further alter nerve excitability, vascular behavior and epithelial permeability. What began as a toxin exposure can therefore evolve into a much broader neuroimmune reactivity state.

Motility can also become involved. Gastrointestinal smooth-muscle contraction depends partly on the balance between Myosin Light-Chain Kinase, which promotes phosphorylation and contraction, and Myosin Light-Chain Phosphatase, which promotes dephosphorylation and relaxation. Acetylcholine, nitric oxide, VIP and other enteric signals continuously regulate this system. One component of my working model is that persistent inflammatory and nitrergic signaling can disrupt this balance in susceptible people and contribute to impaired migrating motor complex activity, intestinal stagnation and recurrent microbial overgrowth.

The final stage of the pathway is physical export. Chemical neutralization does not mean the compound has left the body. Once a reactive molecule has been conjugated, the conjugated metabolite still has to cross a cellular membrane.

This is where Phase III transport becomes important. One transporter I follow is ABCC2, also known as MRP2, with rs717620 as a candidate modifier. MRP2 sits on the canalicular membrane of hepatocytes and uses ATP to transport many conjugated molecules toward bile. Another transporter is ABCB1, also known as MDR1 or P-glycoprotein, with rs1045642 as a candidate efflux variant. ABCB1 is expressed at intestinal and other biological barriers and helps control intracellular xenobiotic exposure.

The mitochondrial connection appears again here because these transport systems require energy. A cell may successfully conjugate a compound but still struggle to export it if ATP production has fallen. This is why detoxification cannot be understood as a single chemical reaction. The molecule has to be transformed, intercepted, conjugated, transported and finally removed from the host. A bottleneck at one stage changes the pressure on every stage behind it.

When I put these pathways together, I do not see mold susceptibility as one gene or one abnormal laboratory value. I see a biological architecture. One person may carry a highly inducible CYP1A2 rs762551 A/A phenotype, a CYP3A4 configuration that increases Phase I throughput and a CYP3A5 expression phenotype related to rs776746. The same person may also carry GSTP1 rs1695, reduced glutathione reserve related to GCLC rs17883901, altered epoxide handling through EPHX1 rs1051740 or rs2234922, DNA-repair variants such as XRCC3 rs861539, XRCC4 rs28383151, XRCC1 rs25487, PRKDC/XRCC7 rs7003908, ERCC2/XPD rs13181 or XPC rs2228001, and Phase III transport variants involving ABCC2 rs717620 or ABCB1 rs1045642.

What interests me is not one SNP in isolation. It is how these pathways stack. A person who produces AFBO quickly, neutralizes it slowly, replenishes glutathione less efficiently, repairs DNA damage less effectively, loses mitochondrial ATP and has constrained cellular export is operating under a very different biochemical environment from someone exposed to the same external molecule with greater reserve at each step.

That is the reason I keep returning to host capacity.

Aflatoxin B1 gives us a way to follow the biology almost molecule by molecule: from the intestine, through CYP-mediated bioactivation, into electrophilic stress, glutathione demand, epoxide handling, DNA adduction, DNA repair, mitochondrial dysfunction, altered epithelial metabolism, microbial ecological shifts, LPS signaling, barrier disruption, neuroimmune amplification and finally ATP-dependent export.

The central question is therefore not only whether the exposure occurred.

It is what the host was able to do with that exposure once it arrived.

That, from my research, is where the deeper biology of susceptibility begins.