The research, led by Yao Peng and Yong Tan of Nanjing University of Chinese Medicine together with colleagues, set out to map what actually happens inside ovarian tissue after repeated exposure to cyclophosphamide. Working with a rat model of multiple drug exposures, the team combined two powerful approaches: they measured the activity of circadian rhythm genes in the ovary, and they ran proteomic analysis to survey the full complement of proteins whose abundance shifted after treatment. The results converged on a surprising culprit. The ovaries of drug-exposed animals showed abnormal expression of circadian rhythm genes, and proteomics revealed a dramatic drop in one protein in particular: riboflavin kinase, abbreviated RFK, the enzyme that initiates the cellular processing of vitamin B2.

That finding might sound obscure, but riboflavin kinase sits at the head of a metabolic pathway with far-reaching consequences. The enzyme converts riboflavin, the vitamin better known as B2, into flavin mononucleotide, which is then elaborated into flavin adenine dinucleotide, or FAD, one of the most indispensable cofactors in biology. FAD participates in hundreds of redox reactions, shuttling electrons through the metabolic machinery that keeps cells alive. The new study adds a second, less appreciated role to that list: FAD is also a structural partner for the circadian clock, because it competes with the clock proteins cryptochrome 1 and 2 for binding to the ubiquitin E3 ligase, the cellular machine that tags proteins for destruction.

Here is where the chemistry becomes elegant. Cryptochromes are the workhorse repressors of the mammalian circadian oscillator, the proteins that help cells keep a roughly 24-hour rhythm of gene activity. Like all proteins, they are constantly being built and broken down, and the ubiquitin E3 ligase is responsible for their degradation. FAD competes with the cryptochrome proteins for access to that ligase, which means that when FAD levels are adequate, the degradation machinery is partially occupied and the cryptochromes survive longer, stabilizing the clock. When FAD runs short, the ligase turns its full attention to cryptochrome 1 and 2, and the clock proteins are chewed up faster than the cell can replace them.

The rat experiments showed precisely this sequence unfolding after cyclophosphamide exposure. Riboflavin kinase expression fell, riboflavin metabolism faltered, FAD synthesis was inhibited, and the cryptochrome proteins were degraded at an accelerated rate. The circadian rhythm regulation of the ovary, a tissue whose follicular development, hormone secretion and ovulation are all coordinated by daily molecular rhythms, fell into disarray. In other words, the chemotherapy drug was not only attacking the DNA of dividing granulosa cells; it was quietly dismantling the metabolic and temporal infrastructure those cells depend on to function.

To test whether the connection was causal rather than coincidental, the researchers turned to molecular experiments in which they knocked down riboflavin kinase directly. The results reinforced the model. Suppressing the enzyme lowered levels of FAD synthase, the downstream enzyme that completes FAD production, and it worsened the degradation of cryptochrome 1 and 2. Circadian rhythm regulation was disrupted, and the consequences rippled into the function of ovarian granulosa cells, the somatic cells that nurse developing follicles and produce the hormones essential for fertility. With their clock destabilized and their metabolism compromised, the granulosa cells declined, and ovarian aging accelerated.

The most clinically tantalizing result came from the rescue experiment. When the researchers supplemented the system with FAD, the cascade of damage was mitigated. Restoring the flavin cofactor eased the pressure on the cryptochrome proteins, helped re-establish circadian regulation and protected granulosa cell function. In the framework of the study, FAD supplementation acted as a counterweight to the riboflavin metabolism defect that cyclophosphamide had created, suggesting a potential strategy for shielding the ovary during chemo. The authors conclude that cyclophosphamide can impair ovarian function by causing dysregulation of riboflavin metabolism, which in turn drives abnormal circadian rhythms.

The implications extend beyond oncology. Premature ovarian insufficiency affects roughly one percent of women and carries consequences that reach far beyond fertility, including bone loss, cardiovascular risk and psychological distress. If a substantial share of chemotherapy-induced ovarian damage is mediated through a metabolic cofactor rather than through DNA strand breaks alone, then the protective options broaden considerably. Dietary and pharmacological interventions aimed at preserving riboflavin metabolism, or at stabilizing cryptochrome proteins directly, could in principle complement the existing approaches to fertility preservation, which currently rely on egg or embryo freezing and on temporary ovarian suppression with hormonal analogs during treatment.

The study also adds the ovary to a growing list of tissues whose health appears to depend on the integrity of their peripheral circadian clocks. Circadian biologists have shown that nearly every organ runs its own internal timing system, synchronized by the brain’s master clock but capable of autonomous oscillation, and that disrupting these local clocks impairs tissue function. The ovary is an especially rhythm-sensitive organ: follicle maturation, steroidogenesis and ovulation follow daily patterns, and granulosa cells express clock genes whose cycling is thought to coordinate these processes. A chemotherapeutic agent that destabilizes cryptochrome proteins in granulosa cells is therefore attacking the ovary at a level that standard genotoxicity models never captured.

As with any animal study, the road from rat ovaries to human clinics is long, and the findings, published open access with support from the National Natural Science Foundation of China, will need validation in human ovarian tissue and eventually in patients. The dosing, timing and safety of FAD supplementation alongside chemotherapy remain to be established, and it is far too early for patients undergoing cyclophosphamide treatment to draw practical conclusions. But the study reframes a familiar hazard in a new light. A drug long understood as a DNA-damaging agent now appears to sabotage the ovary through a second route, by starving it of a vitamin-derived cofactor that keeps its cellular clocks running, and that second route, unlike DNA damage itself, may be one medicine can actually repair.

Subject of Research: Mechanisms of cyclophosphamide-induced ovarian damage via riboflavin metabolism and circadian rhythm disruption

Article Title: Cyclophosphamide as a reproductive hazard: disrupted riboflavin metabolism and circadian rhythms drive ovarian decline

Article References: Peng, Y., Ma, Z., Hu, D., Wu, S., Shi, J., Chen, J., & Tan, Y. (2026). Cyclophosphamide as a reproductive hazard: disrupted riboflavin metabolism and circadian rhythms drive ovarian decline. Journal of Ovarian Research. https://doi.org/10.1186/s13048-026-02293-6

Image Credits: AI Generated

DOI: 10.1186/s13048-026-02293-6

Keywords: cyclophosphamide, ovarian aging, premature ovarian insufficiency, riboflavin kinase, FAD, cryptochrome, circadian rhythm, granulosa cells, chemotherapy, fertility, vitamin B2, ovary

Cathodoluminescence, or CL, is a spectroscopic technique in which electrons rather than photons excite a material, prompting it to emit light. The method has long been a workhorse for characterizing inorganic semiconductors, phosphors, and wide-bandgap materials, where defect-related emission and excitation dynamics are well understood. Organic molecules, however, have remained largely terra incognita. Their structural complexity, strong nonradiative relaxation pathways, and susceptibility to irradiation-induced chemical modification have made them difficult subjects. Yet the payoff for studying them is considerable: high-energy electrons generate a broad cascade of secondary electrons that produce excitation and ionization events fundamentally different from those triggered by a laser or a lamp, potentially granting access to electronic states that optical methods simply cannot reach.

The research team, led by Wataru Inami and Yoshimasa Kawata with colleagues including Machi Takahashi, Kentaro Nakajima, Yu Masuda, and Kei Hosomi, chose their molecular targets with care. They compared three classes of biologically relevant compounds: the aromatic amino acids tryptophan, tyrosine, and phenylalanine in both their D- and L-forms; the coenzymes NADH and FAD, central players in cellular metabolism; and the carbohydrates glucose, glycogen, and sodium glucose-6-phosphate. All samples were measured as powders mounted on carbon tape inside a scanning electron microscope, with a parabolic collection mirror gathering the emitted light and feeding it through an optical fiber to a spectrometer equipped with a photomultiplier tube. Measurements were performed in vacuum at an acceleration voltage of 10 kilovolts and a beam current of 1 nanoampere, with the beam raster-scanned over the sample to record spatially integrated spectra.

The results were unambiguous in their hierarchy. Aromatic amino acids outshone coenzymes, which in turn outshone carbohydrates. When intensities were normalized to sodium glucose-6-phosphate, the weakest emitter, L-tryptophan registered a relative signal of roughly 305, while L-tyrosine and D-phenylalanine came in at around 10 to 11, NADH at about 3, FAD near 1, and glucose and glycogen at under 2. In absolute terms, L-tryptophan emitted approximately 30 times more light than L-tyrosine and 27 times more than D-phenylalanine. The likely culprit behind this dramatic difference is the indole ring, the fused double-ring structure unique to tryptophan’s side chain, whose conjugated system of delocalized pi-electrons provides an efficient channel for both electron excitation and radiative relaxation.

But here is where the story takes a turn that should make microscopists sit up: the dominant emission from tryptophan appeared not in the near-ultraviolet, where its famous intrinsic fluorescence lives, but as a broad, asymmetric band centered around 570 nanometers in the visible spectrum. The authors caution that this visible band cannot be straightforwardly assigned to the conventional low-lying excited states of the indole chromophore that govern ordinary photoluminescence. Instead, they suggest it may arise from solid-state or irradiation-modified emissive states, possibly involving molecular packing effects or multiple overlapping transitions. Notably, both D- and L-tryptophan showed comparable near-UV emission around 340 nanometers, yet the visible band was markedly stronger in the L-isomer, evidence that the visible component is not simply a scaled copy of the molecule’s familiar fluorescence.

Time-resolved measurements added another layer of intrigue. Under continuous electron irradiation, the 570-nanometer band of L-tryptophan steadily declined over 30 minutes, suggesting that the visible-emitting states are progressively deactivated or degraded rather than continuously generated. L-tyrosine, by contrast, held remarkably stable, its luminescence barely changing over 35 minutes of bombardment. Then came the surprise: L-phenylalanine did the opposite of fading. Its CL intensity increased with irradiation time, most dramatically between 10 and 20 minutes, while its dominant emission peak shifted from about 500 nanometers to 430 nanometers, a blue shift of roughly 0.40 electron-volts toward higher energy. Something about the electron beam appears to generate or activate additional higher-energy emissive states in phenylalanine, a molecule whose intrinsic fluorescence is normally confined to the near-ultraviolet and far too weak to explain these broad visible bands.

What could account for such divergent behavior among three amino acids that share the same backbone? The authors point to a competition between destructive and constructive processes. Electron irradiation can induce radiolysis and other chemical modifications that degrade existing emissive states, as apparently happens in tryptophan. But it can also create new radiative centers, as the phenylalanine results imply. Previous work has shown that self-assembled phenylalanine exhibits aggregation-induced visible emission, hinting that intermolecular organization and molecular packing, not just molecular composition, shape the CL response. The differing 430-to-500-nanometer intensity ratios between D- and L-phenylalanine reinforce this picture: the visible emission likely involves multiple electron-beam-induced emissive states whose populations depend on solid-state structure and irradiation history. The researchers are careful to note that their measurements alone cannot distinguish among molecular rearrangement, changes in packing, phase transformation, or specific radiolysis products, and that structural characterization before and after irradiation, for example by Raman spectroscopy, would be needed to settle the question.

The coenzymes and carbohydrates, though fainter, told a consistent story. NADH produced a broad band peaking near 510 nanometers but at only about one-sixtieth the intensity of L-tryptophan, while FAD yielded barely detectable signal across the visible range. Glucose and glycogen, despite glycogen’s being a complex branched polymer of glucose units, emitted nearly identical broad spectra centered around 460 nanometers, a neat confirmation that their shared glucose-based composition dictates their spectral character. Sodium glucose-6-phosphate, a central metabolic intermediate, glowed only weakly, with a low signal-to-noise ratio. Across all three classes, the degree of aromatic conjugation emerged as the decisive variable: the more extended and delocalized the pi-electron system, the brighter the electron-beam-induced glow.

The practical implications extend well beyond basic molecular spectroscopy. Electron-beam-induced autofluorescence has already been demonstrated in living cells, and cathodoluminescence has been observed in functional proteins such as enhanced green fluorescent protein and even in biogenic materials like human kidney stones. As label-free electron-beam microscopy techniques mature, understanding which biomolecules light up, at what wavelengths, and how their emission evolves under bombardment becomes essential for interpreting images correctly. A tryptophan-rich region of a protein will not simply mimic its optical fluorescence signature; it will glow green-yellow at 570 nanometers and fade with dose, while a phenylalanine-rich region may brighten and blue-shift as the beam lingers. Irradiation history, in other words, is written into the spectrum itself.

The study also demonstrates that CL measurements can serve as a sensitive readout of irradiation-induced changes in soft organic materials, a capability with potential relevance for radiation chemistry, electron-beam lithography of organic films, and even assessing radiation damage in biological specimens. The authors emphasize that systematic measurements as a function of accelerating voltage and quantified electron dose will be needed to disentangle direct excitation from irradiation-induced modification. For now, the message is clear and a little humbling: the molecules of life respond to electron beams in ways that neither textbooks on fluorescence nor intuition from optical spectroscopy would predict, and each molecule carries its own signature of damage, resilience, and transformation under the electron’s gaze.

Subject of Research: Cathodoluminescence properties and electron-beam irradiation responses of biologically relevant organic molecules

Article Title: Electron-beam-induced responses of biologically relevant organic molecules

Article References: Inami, W., Takahashi, M., Nakajima, K., Masuda, Y., Hosomi, K., & Kawata, Y. (2026). Electron-beam-induced responses of biologically relevant organic molecules. Results in Chemistry, 31, Article 103940. https://doi.org/10.1016/j.rechem.2026.103940

Image Credits: AI Generated

DOI: 10.1016/j.rechem.2026.103940

Keywords: cathodoluminescence, tryptophan, amino acids, electron beam, biomolecules, spectroscopy, radiolysis, NADH, FAD, carbohydrates, fluorescence, scanning electron microscopy