A cellular (bio) photon field in mitochondrial-chromatin cross-talk: a hypothesis

1 Introduction

Living cells are highly ordered systems, a state that is thermodynamically feasible only through a continuous input of energy. The catabolism of carbohydrates and other nutrients generates ATP, which can then be utilized to drive anabolic reactions responsible for the synthesis of larger and more complex macromolecules. These biosynthetic processes are essential for creating and maintaining intracellular order required for cellular survival, growth, and function. The DNA double helix in chromatin encodes the molecular machinery underlying these anabolic processes, including the synthesis of messenger RNA and proteins. In addition, DNA-guided mechanisms ensure that newly synthesized macromolecules are appropriately transported and organized within the cell. Most spectacular is the supreme (re-)ordering during the process of chromosomal condensation, formation of mitotic spindle and cell division. The high degree of cellular organization requires a substantial fraction of the cell’s ATP budget. Consequently, cellular viability critically depends on a finely tuned coordination between energy production and energy demand.

The classic studies of Gurwitsch were the first that pointed to the significance of mitosis in developing animals and plants that are shaping their form (Gurwitsch, 1910; Gurwitsch, 1911). His theoretical and experimental work focused on the mitotic signal by introducing into biology the notion of a “field” heretofore belonging to the vocabulary of physics (Gurwitsch, 1912). In his Morphogenetic Field Theory the behavior of individual cells is controlled by a field of forces (Gurwitsch, 1922a; Gurwitsch, 1922b). The crucial experiment in 1923 that focused on the verification of a form of radiation demonstrated photon emission as the mitotic signal: emission from a tip of a root into the surrounding space stimulated mitosis in another root (Gurwitsch, 1923; Gurwitsch, 1924). In later studies, Gurwitsch expected “mitogenetic radiation” would be concomitant with metabolic processes. Around the mid-1930’s, the formation and recombination of short-lived free oxygen radicals within reactions became known. Such reactions correspond with high energy, sufficient to generate photon radiation. In line with this, Gurwitsch discovered that atmospheric oxygen is necessary for the mitogenic radiation to have its effect during the enzymatic processes (Gurwitsch, 1937).

In the mid-1950’s the presence of free radicals in biological reactions was discovered, which may account for the presence of gross cellular damage and aging (Harman, 1956; Gerschman, 1981). At that time, after years with biological detectors, this fascinating research continued with the use of a physical detector of the ultra-weak photon emission, the photomultiplier tube. The progress was first made in the USSR where in the 1960’s several Russian groups began to take interest in the ultra-weak photon emission from biological tissues. Most interesting in the present line of thought are the classic studies and reviews suggesting a relationship between mitochondrial respiration and the generation of ultra-weak photon emission. The early studies demonstrated that up to 75% of total photon emission is associated with the recombination of free radicals, such as lipid radicals (Tarusov and Zhuravlev, 1965; Zhuravlev et al., 1968, Zhuravlev et al., 1973; reviewed in Van Wijk et al., 2020). Furthermore, cell homogenization experiments revealed that mitochondria constitute the primary intracellular source of photon emission, with emission closely linked to oxidative phosphorylation processes (Seliger, 1975; Stauff and Ostrowski, 1967; Vladimirov and L’vova, 1964; reviewed by Barenboim et al., 1969). Many more recent reviews in the field of biophoton research and biophotonics have confirmed that virtually all living systems emit extremely low levels of photons in the ultraviolet (UV), visible, and near-infrared spectral ranges (Beloussov et al., 2000, 2007; Fels et al., 2015; Popp and Beloussov, 2003; Shen and Van Wijk, 2005; Van Wijk, 2014; Pospisil et al., 2014; Volodyaev et al., 2023; Moud et al., 2024; Nevoit et al., 2025).

The prevailing view is that biophoton emission originates from electronically excited states generated during biochemical reactions, particularly those involving the electron transport chain and the reactive oxygen species (ROS) produced during lipid peroxidation. With the accumulation of data, the field of biology began to address the significance of photon energy in cellular processes. Does this photon field conception have a direct link to the balance between, on the one hand, the energy requiring chromatin-DNA encoded molecular machinery underlying the anabolic processes and, on the other hand, the energy production delivering the required ATP budget?

Experimental evidence not only supports a causal relationship between the formation ROS and ultra-weak photon emission (UPE), but also the inverse relationship is supported by substantial evidence: ROS production can be stimulated by exposure to low-intensity light (Karu, 1999). This dual causal relationship is illustrated under conditions of ischemia/reperfusion injury. Ischemia results in loss of reduction of oxygen on cytochrome c oxidase of the electron transport chain, resulting in increased neuronal death. Prompt restoration of blood flow, or reperfusion, to the ischemic tissue may reduce neurological damage, but this reperfusion generates additional tissue damage, referred to as reperfusion injury. The injury is due to mitochondrial hyperactivation and ROS production which shows a concomitant increase in UPE intensity. On the other hand, low-intensity laser light therapy (LILT) has been shown to accelerate ATP synthesis, modulate ROS production, and mitigate the deleterious effects of oxidative stress (Tafur et al., 2009). Near-infrared light is known to be primarily absorbed by cytochrome c oxidase due to its two copper centers that are crucial for its enzymatic activity (Karu and Afanasyeva, 1995; Mason et al., 2014; Wharton and Tzagoloff, 1964).

The apparent paradox that UPE appears to arise from ROS while externally applied low-intensity light may also stimulate ROS production has led to the hypothesis that light—both endogenous and exogenous—may serve a regulatory or feedback function, the outcome of which depends on the prevailing oxidative and metabolic state of the cell (Tafur et al., 2009).

In this review and position paper, we further explore the relationship between biophoton emission, oxidative metabolic stress, and the biological effects of external low-intensity light, seeking convergence among multiple independent and complementary lines of evidence. Together, these “co-oriented” arguments support the hypothesis that UPE may play a regulatory role in cellular metabolism. The paper is organized as follows. In Section 2, the standard theory of biological ultra-weak photon emission is reviewed and extended by incorporating recent insights into mitochondrial dynamics, including ROS generation and concomitant photon production. This extension considers the multiple regulatory mechanisms governing mitochondrial behavior as cells adapt to changes in the balance between energy production capacity and metabolic demand during growth and function. The prevailing paradigm has shifted from viewing mitochondria as isolated, passive producers of ATP, ROS, and light to recognizing them as interconnected components of a dynamic mitochondrial network. This network exhibits synchronized, pulsatile behavior in its functioning. Importantly, the strong intracellular absorption of mitochondrial light raises fundamental questions regarding the identity of the cellular structures responsible for light absorption [Reviewed earlier by Van Wijk et al. (2020)].

Section 3 focuses on the role of externally applied low-level (laser) light as a therapeutic intervention that feeds into mitochondrial energy metabolism. Particular attention is given to cytochrome c oxidase as a key photoacceptor. To clarify the relationship between endogenously produced biophotons and externally applied laser light, we compare spectral data from cellular photon emission with the action spectra of low-intensity light known to stimulate growth and anabolic processes, including DNA and RNA synthesis.

In Section 4, we build upon the conclusion from Section 2 that endogenously produced UPE is strongly absorbed within cells. Here we examine the light-absorbing properties of cells that give rise to long-term delayed luminescence, which varies with the degree of cellular differentiation. A comparison of normal and tumor cells reveals that highly differentiated cells exhibit the greatest capacity for long-term photon storage after light exposure. This storage capacity appears to be predominantly localized within nuclear chromatin, the macromolecular complex of DNA and associated proteins. Before addressing the functional significance of photon storage in chromatin, we examine recent insights into chromatin structure and dynamics. Contemporary views of chromosome organization highlight the importance of chromatin mobility and structural dynamics, raising the fundamental and unresolved question of what drives chromatin motion at the molecular level.

In Section 5, we explore the hypothesis that light provides a key physical principle enabling chromatin motion. Photon energy has been linked to vibrational modes in protein α-helices as well as to photoexcitation processes in DNA. We argue that advances in exciton research, along with emerging experimental techniques, must be integrated into biophoton and biophotonics research to adequately assess potential regulatory roles of light at both microscopic and macroscopic scales.

Finally, in Section 6, we propose a speculative model in which two alternating, pulsatile spectra—originating from mitochondrial photon emission and delayed chromatin-DNA luminescence—collectively sustain a dynamic equilibrium between mitochondrial energy production and the anabolic energy demands imposed by chromatin organization and gene regulation. We conclude by briefly outlining experimental strategies that could be employed to evaluate and substantiate this model.

2 Biological photon emission

2.1 Basic concepts of biological photon emission

At the most fundamental level, luminescence arises from the radiative relaxation of electronically excited molecular states. In the case of ultra-weak biological photon emission (UPE), these excited states are generated almost exclusively by endogenous chemical reactions. Early investigations into the molecular origin of this biological chemiluminescence demonstrated that free radical reactions constitute the dominant source of emission. Up to 75% of visible photon emission in biological systems originates from reactions involving the recombination of free radicals, including lipid-derived radicals (Tarusov and Zhuravlev, 1965). Free radicals are highly reactive species that initiate chain reactions by abstracting hydrogen atoms from neighboring molecules, thereby converting these molecules into new radicals. In parallel, certain radical-derived intermediates—most notably organic hydroperoxides—can decompose to generate additional radicals, leading to branching chain reactions characterized by exponential increases in reaction rates. These radical chain processes form the physicochemical basis for biological photon emission.

Carefully controlled cell homogenization studies, which preserve functional integrity of subcellular organelles, provided early evidence that mitochondria represent the principal intracellular source of UPE (Vladimirov and L’vova, 1964). Photon emission was shown to be closely associated with oxidative phosphorylation, a conclusion that has been repeatedly confirmed in subsequent decades [see the recent overview by Volodyaev et al. (2023)]. Within the mitochondrial respiratory chain, photon emission is ultimately linked to the stochastic leakage of single electrons to molecular oxygen, resulting in the formation of the superoxide anion. The extent of electron leakage depends on the functional state of the three transmembrane, electron-driven proton-pump complexes, whose relative contributions vary among organs and tissues.

Photon emission arises downstream of superoxide formation during its conversion to hydrogen peroxide, primarily through dismutation reactions catalyzed by superoxide dismutase (SOD) or occurring spontaneously at low pH. Dismutation of hydrogen peroxide yields water and singlet oxygen (¹O₂), the excited state of which emits photons at approximately 780 nm, with additional emissions extending further into the infrared. Excited singlet oxygen can also form dimers that emit broad spectral bands centered around approximately 585 nm, 635 nm, and 703 nm. The emissions near 585 nm and 635 nm are generally assigned to vibrational transitions associated with the simultaneous relaxation of paired ¹O₂ molecules.

A broad emission band in the 500–560 nm range is characteristic of biological lipid peroxidation. Excited singlet oxygen readily reacts with unsaturated lipids, particularly at methylene groups adjacent to double bonds that contain labile hydrogen atoms. Through the dismutation of peroxyl radicals and cleavage of lipid peroxides, excited carbonyl compounds are formed in singlet and triplet states, along with additional excited singlet oxygen dimers. Excited carbonyls exhibit emission maxima in the 520–540 nm range, but may also contribute to shorter-wavelength emissions between 350 and 500 nm. Furthermore, recombination reactions of certain radical species release sufficient energy to generate photons in the ultraviolet range (approximately 230–300 nm). Taken together, the entire emission spectrum from the ultraviolet to the infrared can be traced back to primary superoxide formation and the subsequent peroxidation of mitochondrial membrane lipids.

Obtaining detailed and quantitative spectra of cellular and tissue UPE has proven technically challenging due to the extremely low emission intensities, and consequently the available data remain limited. However, spectral analyses were conducted using rat liver tissue and liver homogenates subjected to controlled auto-oxidation (Inaba, 1988; Miyazawa et al., 1981, Miyazawa et al., 1983). Such measurements required highly sensitive photon-counting instrumentation combined with carefully designed spectral selection methods. A particularly noteworthy approach employed a 24-band spectral detection system covering the 380–700 nm range, developed by Inaba and colleagues (Inaba et al., 1979, Inaba et al., 1982; Nakano et al., 1975). Instead of dispersive elements such as prisms or gratings, this system utilized successive colored-glass filters with sharp short-wavelength cutoffs, offering a larger acceptance angle without significant wavelength shifts. Spectral windows were defined by subtracting transmission curves of successive filters, followed by calibration of the photocathode spectral sensitivity. The resulting spectral resolution was approximately 30–50 nm in the 275–450 nm range and 20–25 nm between 450 and 670 nm. They studied (a) the spectral distribution of UPE from the liver homogenates from rats administered intragastrical autoxidized linseed oil and (b) the emission spectrum of UPE of liver homogenates after addition of autoxidized linseed oil. In both cases, four dominant bands appeared in the 450–700 nm region. Based on band positions and relative intensities, emissions at approximately 520 nm, 575 nm, and 640 nm were attributed to singlet oxygen transitions corresponding to simultaneous 2[¹Δ_g] → 2[³Σ_g ⁻] transitions with vibrational quantum numbers 2 → 0, 1 → 0, and 0 → 0, respectively. A weaker emission at 480 nm was assigned to the [¹Δ_g] + [¹Σ_g ⁺] → 2[³Σ_g ⁻] transition. An integrated representation of these emission bands has been presented previously (Van Wijk and Schamhart, 1988).

2.2 New insights into the dynamics of mitochondrial ROS production

Beginning in the early 2000s, several discoveries fundamentally altered our understanding of mitochondrial behavior and its relationship to reactive oxygen species (ROS) production, necessitating a reconsideration of UPE dynamics. Three major developments demonstrated that mitochondria are not static energy producers but highly dynamic organelles.

First, the advent of deconvolution and confocal microscopy enabled three-dimensional visualization of mitochondria in living cells, revealing continuous morphological remodeling through coordinated cycles of fission and fusion. These processes dynamically regulate mitochondrial size, shape, and connectivity (Liesa et al., 2009). Wikström and colleagues (2007), Wikström and colleagues (2009) further demonstrated that fission events can generate daughter mitochondria with distinct bioenergetic properties, a phenomenon thought to be linked to mitochondrial quality control (Molina et al., 2009; Mouli et al., 2009; Twig et al., 2008a). Through selective fission and subsequent autophagic removal of dysfunctional fragments, cells prevent the accumulation of inefficient ATP-producing mitochondria under conditions of nutrient excess and oxidative stress (Parone et al., 2008). Conversely, during fasting or low-energy states, enhanced fusion homogenizes the mitochondrial population, improving bioenergetic efficiency and protecting mitochondria from autophagic degradation (Twig et al., 2008b).

Second, dynamic mitochondrial behavior became evident through the use of targeted fluorescent probes sensitive to intramitochondrial processes such as superoxide production, calcium uptake, membrane potential, and pH. Wang et al. (2008) first reported transient, stochastic bursts of superoxide generation within the mitochondrial matrix, lasting approximately 10–15 s. These “mitochondrial superoxide flashes” have since been observed in multiple mammalian cell types and tissues (Li et al., 2012; Wang et al., 2012, Wang et al., 2013). Subsequent development of fluorescent indicators for Ca²⁺, pH, and membrane potential revealed that these parameters fluctuate in a synchronized manner during individual flash events (Wang et al., 2016). Simultaneous imaging of multiple probes demonstrated that a single flash reflects a coordinated, transient reorganization of mitochondrial physiology, strongly suggesting that ROS generation—and by extension photon emission—occurs as a pulsatile on–off process.

Third, studies in skeletal muscle and cardiac tissue revealed that mitochondrial flashes are not independent stochastic events but often occur synchronously among spatially clustered mitochondria (Pouvreau, 2010; Wang et al., 2013; Wei et al., 2011a; Wei et al., 2011b). These findings prompted a paradigm shift from viewing mitochondria as isolated, passive producers of ATP and ROS to understanding them as components of an interconnected, dynamically coordinated network. Within this network, mitochondrial energy production, ROS generation, and associated photon emission are regulated through amplitude and frequency modulation, providing a potential mechanism for matching cellular energy supply to fluctuating metabolic demands.

2.3 Absorption of mitochondrial photons

The hypothesis that synchronously pulsating mitochondrial photon emission might serve a regulatory role within cells encounters an apparent difficulty: the extremely low photon flux detected experimentally from cells and tissues. For example, when a photomultiplier tube is positioned above a 25 cm² liver cell culture containing approximately 6 × 10⁶ cells, the recorded signal rarely exceeds 10² photons s^-1 cm^-2. Accounting for geometric constraints and detector quantum efficiency, this corresponds to roughly one detected photon per second per 2,000 cells, or per 1–2 × 10⁶ mitochondria. On the basis of detected photon counts alone, it would thus be difficult to argue for a meaningful regulatory role of such emissions.

However, these measurements do not account for intracellular photon production and absorption. Chance was among the first to estimate cellular oxygen consumption based on organismal metabolism, calculating that an average rat cell processes approximately 10¹² oxygen molecules per day (Chance et al., 1979). If roughly 2% of consumed oxygen undergoes single-electron reduction to form superoxide, this yields approximately 2 × 10¹⁰ superoxide molecules per cell per day. Similar calculations by later authors across different organisms and tissues produced values within one order of magnitude of this estimate, reflecting tissue-specific metabolic activity (Ames et al., 1993; Bokkon et al., 2010; Doraiswamy and Finefrock, 2004; Hoidal, 2001; Perry et al., 2002). According to Slawinski (1988), at least 1% of these reactions contribute to photon production. Based on an average production of approximately 2.3 × 10⁵ superoxide molecules per cell per second this implies the generation of roughly 2,300 photons per cell per second, corresponding to a few photons per mitochondrion per second.

The calculated values are compatible with a signaling function within a mitochondrial network characterized by pulsations lasting on the order of seconds. The striking discrepancy between calculated photon production (∼10³ photons s^-1 cell^-1) and detected emission (∼10^–3 photons s^-1 cell^-1) implies that intracellular absorption of mitochondrially produced photons must be extraordinarily high—on the order of 10⁶ or greater. This conclusion shifts attention from photon emission toward photon absorption. Within the cellular architecture, the nucleus emerges as a prime candidate for photon absorption. The nucleus typically measures 6–10 µm in diameter, occupies roughly 10% of cell volume, and is surrounded by thousands of mitochondria distributed throughout the perinuclear cytoplasm. Given this spatial configuration, synchronously emitted mitochondrial photon pulses are likely to impinge upon the nucleus with high probability.

Consequently, any analysis of the fate and potential regulatory significance of internally generated photons must consider the nuclear compartment and, in particular, nuclear chromatin as a key light-absorbing structure. This issue is addressed further in Section 4. In the following section, we first turn to the biological effects of externally applied low-intensity laser light on mitochondrial function.

3 Low-intensity light therapy (LILT)

3.1 Photobiomodulation of mitochondrial respiration

Low-intensity light therapy (LILT), also referred to as photobiomodulation, has been shown to stimulate cellular metabolism by enhancing ATP production. Cells in an actively proliferating state—which is typically associated with a more reduced redox balance and elevated metabolic activity—often exhibit enhanced responsiveness to LILT. Indeed, rapidly dividing cell types such as HeLa cells, fibroblasts, and epithelial cells show pronounced photobiological responses. Such cells have been systematically investigated for the wavelength dependence of LILT—commonly referred to as the action spectrum. Action spectra plot relative biological effectiveness as a function of wavelength and thus provide insight into the identity of intracellular photoacceptors. These studies expanded rapidly during the 1980s with the advent of semiconductor medical lasers (Karu et al., 1982, Karu et al., 1984a; Karu et al., 1984b). In the context of energy-demanding cellular processes, particular attention has been given to proliferation and to stimulation of DNA and RNA synthesis.

Using semiconductor lasers, Karu and colleagues investigated approximately 25 discrete spectral bands between 330 and 860 nm, with spacing of 20–35 nm. Remarkably, the action spectra for cell proliferation, DNA synthesis, and RNA synthesis were found to be nearly identical, suggesting a common underlying mechanism related to mitochondrial energy production (Karu et al., 1982, Karu et al., 1984a; Karu et al., 1984b; Karu et al., 1985, Karu et al., 1996). In a generalized action spectrum for cultured mammalian cells, four major activity regions were identified between 600 and 860 nm: (i) a red band with a peak between 613.5 and 623.5 nm, (ii) a far-red band peaking at 667.5–683.7 nm, and (iii) two near-infrared bands centered at approximately 750.7–772.3 nm and 812.5–846.0 nm.

In addition, an active band was observed in the violet–blue region, with a maximum between 400 and 450 nm.

Interpretation of these peaks relied on comparison with absorption spectra of known chromophores. The red (620–680 nm) bands correspond well to charge-transfer transitions in metal–ligand systems, whereas the near-infrared bands (760–830 nm) are characteristic of d–d electronic transitions in metal centers (Hughes, 1987; Wilkinson et al., 1987; Siegel, 1971; Wharton and Tzagoloff, 1964). The blue–violet band (400–420 nm) matches π–π* transitions of porphyrin rings (Gouterman, 1978). Collectively, this spectral fingerprint strongly implicated cytochrome c oxidase as the primary photoacceptor mediating LILT effects (Karu, 1988; 1999; Karu and Afanasyeva, 1995; Mason et al., 2014). This conclusion has been supported by biochemical, physiological, and spectroscopic studies demonstrating light-induced modulation of cytochrome c oxidase activity, mitochondrial membrane potential, and ATP synthesis (Eels et al., 2003, Eels et al., 2004; Karu, 2008; Karu et al., 2008; Pastore et al., 2000; Wong-Riley et al., 2001). Thus, LILT acts primarily by enhancing mitochondrial respiratory efficiency under conditions of suboptimal oxidative phosphorylation.

3.2 Low-intensity light therapy and the oxidative stress enigma

LILT has been shown, paradoxically, to mitigate oxidative stress—despite the fact that its primary photochemical effects are generally oxygen-dependent and are accompanied by transient increases in ROS levels (Eels et al., 2004; Karu et al., 1995, Karu et al., 2004; Pal et al., 2007). Importantly, cellular sensitivity to LILT strongly depends on the intracellular redox state (Karu et al., 2005; Maiya et al., 2005). In actively proliferating cells, proliferation is commonly accompanied by increased mitochondrial ROS production (Karu et al., 1994; Keszeli et al., 1994; Tafur and Mills, 2008; Zhang et al., 2003). From a conventional oxidative stress perspective, the application of LILT under such conditions might be expected to exacerbate oxidative damage rather than alleviate it.

Nevertheless, extensive experimental and clinical studies demonstrate that LILT improves tissue survival and functional recovery in conditions such as cardiac and cerebral ischemia, both of which are characterized by marked elevations in ROS following hypoxic injury and mitochondrial respiratory failure (Oron et al., 2001a, Oron et al., 2001b; Oron et al., 2006). These therapeutic effects exhibit a remarkable dependence on wavelength, intensity, energy dose, and pulse frequency (Agaiby et al., 2000; Nussbaum et al., 2002a, Nussbaum et al., 2002b; Young et al., 1989).

It was initially believed that near-infrared light consistently activated cytochrome c oxidase activity, which is true for certain wavelengths (Wong-Riley et al., 2005; Eels et al., 2004; Morse et al., 2021). However, other authors have identified and characterized specific wavelengths of near infra-red light (750 and 950 nm) that inhibit cytochrome c oxidase enzymatic activity (Sanderson et al., 2018; Wider et al., 2023; Strubakos et al., 2020). These inhibitory wavelengths provide neuroprotection in cultured human neuroblastoma cells (Pham et al., 2025).

These findings pose not only the fundamental paradox: how can a LILT intervention that transiently increases ROS nonetheless protect tissue from oxidative damage, but also show both activating and inhibiting responses? A productive approach to resolving this enigma is to distinguish between different qualities and contexts of light. In this line of thinking, we have first asked whether endogenously generated photon emission associated with ROS formation is equivalent, in spectral terms, to externally applied low-intensity laser light that stimulates mitochondrial respiration and attenuates cellular stress.

3.3 Spectral comparison with endogenous photon emission from ROS

The action spectrum of LILT, which selectively stimulates cytochrome c oxidase and ATP synthesis, can now be compared with the spectrum of ultra-weak endogenous photon emission arising from mitochondrial ROS formation. As discussed in Section 2.1, spontaneous emission spectra of biological systems have been measured using highly sensitive multi-band photon-counting approaches, notably the 24-band spectral system developed by Inaba and colleagues covering the 380–700 nm range (Inaba, 1988; Inaba et al., 1979, Inaba et al., 1982; Miyazawa et al., 1981, Miyazawa et al., 1983; Nakano et al., 1975).

These studies consistently reported emission bands at approximately 480, 520, 575, and 640 nm. When these endogenous emission peaks are compared with the biologically active LILT wavelengths in the same spectral range—namely, the red peak around 620 nm, the far-red interval around 670–680 nm, and the violet–blue peak between 400 and 450 nm—it becomes evident that, although both spectra are multi-peaked, their maxima are clearly shifted with respect to one another.

This spectral mismatch suggests that the photons produced directly during mitochondrial ROS formation are unlikely to efficiently activate cytochrome c oxidase in neighboring mitochondria playing a role in the synchronization of pulsations. Endogenous UPE does not appear to function simply as an internal analog of externally applied LILT. This observation reinforces the notion that the biological role of endogenous photon emission differs fundamentally from that of external photobiomodulation.

However, an alternative interpretation may reconcile these findings. It is conceivable that LILT wavelengths correspond not to primary mitochondrial emission, but rather to red-shifted, delayed luminescence originating from secondary intracellular chromophores—most notably nuclear chromatin—following excitation by mitochondrial photon pulses. In this framework, mitochondrial light would act as a primary trigger, while chromatin-mediated delayed emission would generate spectral components capable of modulating mitochondrial respiration.

To evaluate this possibility, it is necessary to examine recent insights into chromatin structure, dynamics, and light–matter interactions. These topics form the focus of the next section.

4 Absorption of mitochondrial light: a role for nuclear chromatin?

4.1 Cellular architecture favors absorption of mitochondrial light by chromatin

As discussed in Section 2.3, basic considerations of the spatial arrangement of mitochondria surrounding the nucleus, it is highly probable that at least part of the mitochondrial photon flux reaches the nuclear compartment—either within the same cell or, in densely packed tissues, neighboring cells.

The present section examines the interaction of photons with chromatin at the molecular level, focusing on fluorescence, delayed luminescence, and photon storage. A central observation emerging from decades of biophysical research is that photon absorption and emission in cells are dominated by proteins and nucleic acids. In proteins, light absorption is primarily mediated by aromatic amino acid residues—tryptophan, tyrosine, and phenylalanine. Importantly, the photophysical properties of these residues are profoundly altered when incorporated into polypeptide chains: fluorescence spectra, quantum yields, and excited-state lifetimes depend strongly on protein conformation and molecular packing.

A comparable situation exists in nucleic acids. In DNA and RNA, excitation is dominated by the purine and pyrimidine bases, whose photophysical behavior is likewise highly sensitive to molecular context. Base stacking, hydrogen bonding, and interactions with associated proteins modulate excitation lifetimes, energy transfer pathways, and emission properties. Thus, both proteins and nucleic acids function not merely as isolated chromophores but as cooperative molecular assemblies whose structural organization determines their interaction with light.

These considerations suggest that chromatin—the highly ordered complex of DNA and associated proteins—constitutes a particularly plausible intracellular sink for photon absorption. The ability of chromatin to absorb, transiently store, and re-emit photons may therefore reflect fundamental aspects of its structural state and functional activity.

4.2 Long-term delayed luminescence and cellular differentiation

The hypothesis that nuclear DNA, or more precisely chromatin, possesses a capacity for photon storage was originally articulated by Popp and co-workers (Nagl and Popp, 1983; Popp et al., 1984; Rattemeyer et al., 1981) in the context of cellular differentiation and oncogenesis. Differentiation was proposed to involve progressive structural organization of chromatin, accompanied by an increased capacity for energy—and photon—storage. Tumorigenesis, by contrast, was hypothesized to involve a loss of this organizational capacity, manifested as reduced photon storage and increased photon emission.

To experimentally test this hypothesis, Popp developed a modified photon-counting approach in which biological samples were first exposed to an external light source and subsequently recorded in complete darkness. The delayed emission of stored photons was recorded over time, with emission kinetics—rate, duration, and total photon yield—serving as quantitative measures of photon storage capacity.

Initial studies focused on normal adult rat liver cells and two hepatoma cell lines, classified as well-differentiated and poorly differentiated in terms of ultrastructure and metabolism according to the “molecular correlation” concept of cancer (Lo et al., 1968; Shonk et al., 1965; Weber, 1968; Weber and Lea, 1967). Electron microscopy revealed progressive simplification of cellular architecture across the differentiation spectrum, including changes in mitochondrial volume and internal organization (Volman, 1978; Volman and Van Wijk, 1980). Metabolic characterization demonstrated parallel alterations in glucose metabolism, with reduced gluconeogenic capacity and altered hormonal responsiveness accompanying loss of differentiation (Schamhart et al., 1979; Van Rijn et al., 1974; Van Wijk, 1983; Van Wijk et al., 1972, Wijk et al., 1973; Wicks et al., 1973a, Wicks et al., 1973b).

Photon storage measurements corroborated these structural and metabolic distinctions. Poorly differentiated hepatoma cells exhibited rapid and intense photon emission, indicative of minimal photon storage capacity. Normal adult liver cells, by contrast, showed slow, prolonged emission consistent with high photon storage. Well-differentiated hepatoma cells displayed intermediate behavior (Schamhart and Van Wijk, 1987; Van Wijk and Schamhart, 1988; Van Wijk and Van Aken, 1991). These findings were further strengthened by experiments in which cells were cultured at increasing densities. Normal liver cells exhibited enhanced photon retention at higher densities, consistent with reabsorption and storage of photons emitted by neighboring cells. Poorly differentiated hepatoma cells, however, continued to emit large quantities of photons even under dense culture conditions, indicating an inability to absorb light efficiently. Again, well-differentiated hepatoma cells showed intermediate characteristics.

Subsequent studies extended these observations to other biological systems, including comparisons between human amnion cells and WISH cells (Scholtz et al., 1988) and between normal and malignant melanoma cells (Niggli, 1992). A comprehensive analysis examined photon storage across 17 different mammalian normal and tumor cell lines from multiple species and tissues (Van Wijk et al., 1993). Photon storage capacity varied more than 25-fold across cell types. Notably, fibroblasts and connective tissue cells exhibited the lowest photon storage capacity and highest emission, independent of malignancy status. This observation may reflect both their lower degree of differentiation and their relatively isolated functional role within connective tissue matrices. The normal and tumor cells of other tissues show the behavior as mentioned before: a high photon storage in normal cells and a lower storage capacity in tumor cells. Collectively, these studies establish a consistent relationship between long-term delayed luminescence and cellular differentiation state.

4.3 Long-term delayed luminescence originates from chromatin, not naked DNA

Although cellular capacity for long-term photon storage is well established, experimental evidence clearly indicates that this capacity does not reside in DNA alone. Cell fractionation experiments provided decisive insight into the subcellular localization of delayed luminescence (Van Wijk and Van Aken, 1991). Careful separation of cellular components followed by light exposure and subsequent photon recording revealed that the cytoplasmic fraction—including mitochondria and ribosomes—exhibited little or no long-term photon storage. In striking contrast, intact isolated nuclei retained an even greater capacity for delayed luminescence than whole cells. When nuclei were gently fractionated further, the chromatin fraction containing DNA tightly associated with nuclear proteins (notably histones) displayed robust photon storage. However, once DNA was separated from its associated proteins, neither the isolated DNA nor the protein fraction alone retained significant long-term luminescence.

These findings demonstrate that photon storage is an emergent property of chromatin as an integrated macromolecular structure, rather than of DNA or proteins in isolation. The physical organization of chromatin is therefore crucial for its optical properties. In the next paragraph we examine recent insights into chromatin complexity and dynamics as a transition to Section 5, where we consider the potential role of light in regulating chromatin motion and genome function.

4.4 Chromatin structure and dynamics: what brings the genome into motion?

The extraordinary complexity of chromatin organization becomes evident when considering that interphase DNA remains compacted approximately 1,000-fold, yet must remain dynamically accessible. Advances in chromosome conformation capture techniques have revealed a hierarchical organization that depends on observation scale (Lieberman-Aiden et al., 2009; Rao et al., 2014). At the megabase level, the genome is partitioned into transcriptionally active “A” compartments and more condensed, inactive “B” compartments. Individual chromosomal regions can transition between these states over time. At finer scales, DNA is organized into nucleosomes—octameric complexes of histone proteins around which 146 base pairs of DNA are wrapped (Rhodes, 1997). Histone–DNA interactions involve extensive hydrogen bonding and repeated α-helical motifs. Nucleosomes further assemble into higher-order chromatin fibers forming a dynamic, fluid mosaic. At intermediate scales, the genome organizes into topologically associating domains (TADs), constraining local interactions while preserving global flexibility (Lavelle, 2009; Kouzine et al., 2008).

Superimposed on this architecture is continuous motion driven by ATP-dependent chromatin remodeling complexes, loop formation and dissolution, and transcriptional activity itself (Aalfs and Kingston, 2000; Cairns, 1998; Kingston and Narlikar, 1999; Rhodes, 1997). During transcription, RNA polymerase locally opens the DNA double helix, generating transient “transcription bubbles” that travel along genes and subsequently close. Across thousands of genes, such processes occur simultaneously, raising the fundamental question of how tangling and topological stress are avoided. Enzymes such as topoisomerases resolve these constraints by introducing transient single- or double-strand breaks, enabling controlled unwinding and strand passage (Wang, 2009).

The chromatin dance seems to be essential for transcription activity. Recently, genome-wide binding studies using techniques such as ChIP-seq and CUT&RUN have revealed the enormous complexity of transcriptional regulation, identifying vast networks of gene targets controlled by specific transcription factors (Lee et al., 2002; Milo et al., 2002). Moreover, single-cell imaging combined with fluorescent reporters has uncovered dynamic pulsing behavior in genetic circuits (Levine et al., 2012, Levine et al., 2013). Transcription factors repeatedly enter and exit the nucleus in discrete pulses, even under constant external conditions.

Taken together, both mitochondrial metabolism and chromatin organization operate via highly orchestrated, pulsatile dynamics. What once appeared as stochastic intracellular motion is now recognized as tightly regulated spatiotemporal choreography. Gene expression depends not only on DNA sequence but also on higher-order chromatin architecture and coordinated movement (Dekker and Misteli, 2015). This convergence leads to a central unanswered question: “what physical principle brings chromatin into motion?” In the following section, we explore the possibility that light—endogenously, i.e., mitochondrial-produced and chromatin-absorbed—contributes to this dynamic regulation.

5 Chromatin dynamics driven by light

5.1 Excitonic processes in proteins

During recent decades, the classical view of proteins as static structural scaffolds has been replaced by a dynamic paradigm in which protein motion is recognized as integral to biological function. With respect to light–matter interactions, photon-induced motions have been linked to small-amplitude vibrational dynamics, particularly within α-helical protein structures (Renger and Marcus, 2002). These vibrations enable the formation of collective excited states known as excitons, which represent delocalized electronic or vibrational excitations. The exciton concept was first introduced to describe electronic excitation in crystalline lattices, where energy could propagate through a medium without net charge transfer (Frenkel, 1931).

In biological macromolecules, excitons analogously describe excited electron–hole pairs that migrate through ordered molecular structures held together by hydrogen bonding networks. In proteins, α-helices provide a favorable environment for exciton propagation due to their regular geometry and strong hydrogen-bond stabilization (Yomosa, 1963). The recombination probability of the electron–hole pair is reduced under these conditions, allowing excitons to persist for relatively long lifetimes—ranging up to milliseconds. Upon relaxation to the ground state, the excitation energy may be released as photon emission, typically red-shifted relative to the absorbed wavelength. The spatial displacement of excitonic energy along the polypeptide chain depends on the degree of excitation delocalization, which itself is tightly coupled to protein conformation and dynamics. These findings indicate that proteins are not merely passive absorbers of light but dynamic structures capable of transiently storing and redistributing photon energy.

5.2 Excitonic and charge-transfer processes in DNA

Photoexcitation dynamics in DNA have been intensively studied by advances in ultrafast spectroscopy (Takaya et al., 2008), and a substantial body of theoretical and experimental literature has addressed charge and exciton transfer processes in nucleic acids (e.g., Crespo-Hernández et al., 2004; Improta et al., 2016; Middleton et al., 2009; Schuster, 2004). As in proteins, exciton and charge propagation in DNA depend not only on extrinsic factors such as solvent and ionic environment, but also on intrinsic structural fluctuations within the double helix (Gutiérrez et al., 2009, Gutiérrez et al., 2010). Theoretical modeling of these processes rapidly becomes intractable beyond a few base pairs, prompting the development of tight-binding and coarse-grained models that incorporate statistical ensembles of DNA sequences (Lambropoulos and Simserides, 2019).

Recent work has provided quantitative support for sequence-dependent excitonic behavior. Herb et al. (2024) demonstrated a direct relationship between exciton lifetime and spatial separation of electron–hole pairs in seven-base-pair DNA sequences, identifying a limited subset of sequences that support exceptionally long-lived excited states and pronounced charge separation. Complementary circular dichroism studies have revealed spatial delocalization of excitons in short single-stranded DNA, with coupling strengths dependent on base composition (Pichiotti et al., 2025). Systematic studies have established that guanine-rich sequences preferentially localize radical cations, while adenine-rich tracts facilitate efficient hole migration. In contrast, thymine and cytosine act as barrier bases that strongly attenuate charge transport. In summary, hole transfer in DNA is a complex process in which many chemico-physical factors play a role.

5.3 Broader implications of excitonic phenomena

Excitons represent only the first tier of light-induced interactions. Coupling between excitons and photons gives rise to exciton–polaritons, hybrid states that extend the lifetime and spatial coherence of excitation. Further complexity arises when excitons interact with one another to form biexcitons or, at sufficiently high densities, collective electron–hole liquids—states experimentally observed in semiconductors and theoretically proposed for nucleic acids under extreme conditions. This is particularly interesting in chromatin with α-helix proteins in complexes with DNA (see 4.4).

Although the formation of electron–hole liquids in DNA remains speculative, such concepts may eventually help explain preferential pathways of energy and charge migration within chromatin. The biological consequences could be substantial. Charge delocalization within chromatin domains would generate transient or semi-stable electric dipoles, capable of altering local conformation, electrostatic potential, and dielectric properties. In turn, these changes could modulate the affinity of regulatory proteins for DNA, influencing gene regulation and transcription dynamics.

6 Discussion and conclusion

In this review, two historically distinct focal points in the study of universal ultra-weak biophoton emission (UPE) are integrated into a unified conceptual model. The first research tradition emerged from mitochondrial biochemistry and aimed to elucidate the mechanisms by which photons arise from energetically excited biomolecules generated during oxygen reduction. The second tradition focused on the photon field associated with chromatin organization and emphasized a potential regulatory role of intracellular light. Over several decades, these approaches evolved largely independently, forming two schools of thought with contrasting interpretations: one viewing biophotons primarily as metabolic byproducts of free radical reactions, and the other attributing to them a functional role in cellular regulation.

The introduction of low-intensity light therapy (LILT) in the final decade of the 20th century provided a crucial conceptual bridge between these perspectives. It became clear that externally applied light can stimulate mitochondrial activity, and more recently even enhance endogenous biophoton emission (Kam et al., 2025). In this review, we have sought to integrate these three bodies of knowledge—mitochondrial photon production, chromatin-associated photon fields, and LILT—by incorporating recent advances in mitochondrial dynamics, chromatin organization, and photobiology.

Central to this integration is the view of mitochondria not as isolated energy generators, but as active and interconnected participants in a network regulating the cell energy metabolism. Production of ATP, reactive oxygen species, and UPE in a network are synchronized and pulsatile; thus, fluctuations are due to the number of mitochondria involved and the frequance of individual mitochondria involved in pulsation. In other words as a combination of amplitude modularity (AM) and frequency modularity (FM) modes.

Also central is the view that energy production is tuned to required gene activity, via a feedback loop between chromatin photon storage and its delayed photon re-emission that signals back to the mitochondrial network activity. This view emphasizes a definite meaning of photonic signaling as was earlier proposed (Cilento, 1988; Slawinski, 1988). That view considered a photon either as the “substrate” of a photoreaction, or the “product” of a chemiluminescence reaction and proposed (i) that these types of reactions may occur simultaneously, and (ii) that photons are then able to couple these reactions in cells. This principle is now extended in more detail in the presented model.

Many aspects of the proposed role of UPE have remained hypothetical and evidently need experimental verification. Aspects such as: (i) the modulation of chromatin organization by mitochondrial emission; (ii) the extent of photonic chromatin activity, i.e., its delayed luminescence, that may influence mitochondrial network activity,; (iii) the relation between DL and LILT, i.e.,the potential of LILT under conditions of in insufficient signaling. The following is an illustration of the basic principles required for a primary level research approach, including (i) physiological conditions for testing, (ii) protocols for testing, (iii) the accuracy of equipment, and what is already known in these essential research strategies.

For testing whether mitochondrial photon emission can modulate chromatin organization it is important to use physiological and stress situations. Along the line of thinking in AM- and FM-modularity modes, evidence is accumulating suggesting that the FM mode of ROS regulation is uniquely suitable to regulate signals under physiological conditions when respiration is stimulated by physiological respiration substrates, such as oxygen, glucose, fatty acids (Gong et al., 2014, Gong et al., 2015; Ying et al., 2016). Under extreme conditions or strong stress, the FM mode activity becomes uncontrolled and significantly augments ROS production (Zorov et al., 2000; Aon et al., 2003; Wang et al., 2008; Ying et al., 2016). In testing stress conditions, stressors such as sodium troclosene and rotenone might be utilized, each with different mechanisms of action. Sodium troclosene is a strong oxidising agent that slowly releases chlorine and hence changes the pH of the medium to acidic (Heling et al., 2001). Rotenone is an inhibitor of complex I in the mitochondrial electron transport chain and has been shown to increase the production of ROS, lower the production of ATP and decrease the mitochondrial membrane potential (Swarnkar et al., 2010). Interestingly, these conditions have been recently explored in combination with red and near-infrared light treatment, showing that it did not influence UPE emissions when healthy, but did increase it when stressed (Kam et al., 2025).

In any future research project the primary aim is to establish an accurate spectrum of spontaneous photon emission under the physiological and stress conditions. As discussed in paragraph 2, the most successful way is to utilize the super-high sensitivity system of successive colored glass filters with sharp short-wavelengths cutoffs that result in spectral resolution of 20–30 nm in the 400–750 nm range (Inaba, 1988). It is fundamental to identify the characteristic spectral emission bands because these wavelengths will then be used to get insight how the mitochondria-generated wavelengths spectrum may serve as a source of radiation for nuclear chromatin to test their capacity to be stored for longer durations. It will be necessary to use these wavelengths artificially both in the excitation of cells and isolated chromatin.

Experimental evidence has clearly demonstrated that chromatin is capable of storing incident light energy and releasing it over characteristic timescales. Pulsatility, typically synchronously, plays a critical role in this model. The 5–10 s period of mitochondrial pulsations occurring, may be used experimentally to deposit light energy into chromatin. The delayed emission–generally not exceeding a few seconds–will be recorded for each individual spectral band used for excitation, utilizing a detection procedure which will be again wavelength-specific. This requires repetitive excitations and demands a rigorous refinement of the classic long-term delayed luminescence equipment (paragraph 2). In recent decades, important technical expansions have been made, combining advanced photomultiplier tubes with targeted illumination of objects using specific wavelengths–across the spectrum from UV to IR–and reading the emission in microseconds after excitation (Musumeci et al., 2005; Scordino et al., 2014).

This technique offers in principle, the full complexcity of the DL spectrum. Differences in these complex spectra between whole cells and isolated chromatin may then give additional additional also information how the photonic signal is influenced by the crowded and optically intracellular heterogeneous cytoplasmic environment. But the same complexity of the spectrum of delayed luminescence of chromatin can be also used to compare with the spectrum of beneficial effects of LILT in stimulating cytochrome c oxidase and enhancing mitochondrial energy production.

With this short outline of a research program, we now find ourselves at a threshold where future progress in UPE research hinges on a deeper understanding of how specific wavelength bands are absorbed by intracellular structures. Progress in this field will critically depend on methodological refinement and the employment of narrowly defined excitation wavelengths. Through such wavelength-resolved and time-resolved approaches can UPE research meaningfully contribute to understanding disturbed mitochondrial–chromatin cross-talk, particularly in conjunction with low-intensity light therapy. This perspective also provides the conceptual foundation for a new model of cellular regulation, in which photonic processes complement and integrate with metabolic and genetic control mechanisms.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Maurizio Benfatto, Universities and Research, Italy

Reviewed by: Lev Mourokh, Queens College (CUNY), United States

Pavel Pospíšil, Palacký University, Czechia