Diphtheria Toxin Causes Off-Target Neurotoxicity in Mouse Depletion Models

The challenges of modeling myeloid cell depletion in the central nervous system

Microglia and tissue-resident macrophages are increasingly recognized as central mediators in the pathogenesis of neurodegenerative disorders, acting as both protective sentinels and drivers of neuroinflammation [1, 2]. To isolate the specific contributions of these cells to disease progression, researchers frequently employ depletion models to observe how the central nervous system responds in their absence [3, 4]. These models are essential for validating potential therapeutic targets, such as those involving Adeno-associated virus (a viral vector used to deliver genetic material into cells) delivery or small-molecule inhibitors aimed at modifying microglial activity [5]. However, the reliability of these preclinical tools depends entirely on the assumption that the depletion method itself does not introduce confounding pathology. A new study now examines the safety profile of a widely used toxin-based depletion method, revealing unexpected complications that may necessitate a reevaluation of previous findings.

Dose-dependent mortality and motor deficits in wild-type mice

The researchers evaluated the safety and cellular impact of diphtheria toxin by administering it via intracerebroventricular (i.c.v.) injection, a procedure where the toxin is delivered directly into the fluid-filled ventricles of the brain. The study utilized both wild-type mice and CD11b-DTR mice, which are genetically modified to express the diphtheria toxin receptor on specific myeloid cells. Over a 10-day period, the animals received bilateral i.c.v. infusions of either the toxin or a vehicle control to assess tolerability, behavioral outcomes, and cellular responses. A critical finding was that diphtheria toxin induced dose-dependent mortality in wild-type mice, a group that lacks the specific receptor typically required for the toxin to enter and kill cells. This observation is clinically significant because it indicates that the toxin exerts off-target neurotoxicity independent of transgene expression, suggesting that the delivery of the toxin itself may be inherently toxic to central nervous system tissues. This finding challenges the assumption that wild-type animals are immune to the toxin's effects, suggesting that high local concentrations in the cerebrospinal fluid may bypass the need for specific receptor binding.

Beyond survival rates, the researchers conducted a battery of tests to quantify functional impairment resulting from toxin exposure. Both genotypes exhibited significant dose-dependent impairments in rotarod performance, a standardized test used to measure motor coordination and physical endurance by requiring mice to maintain balance on a rotating cylinder. Furthermore, the toxin exposure led to significant dose-dependent impairments in Y-maze spontaneous alternation, a behavioral assay that evaluates cognitive function and spatial working memory based on the natural tendency of mice to explore new arms of a maze. Interestingly, while motor coordination and memory were compromised, open-field mobility was largely preserved among the animals. This preservation of general movement in an open area suggests that the observed deficits in the rotarod and Y-maze were not merely the result of generalized lethargy or systemic illness, but rather reflected specific neurological dysfunction induced by the toxin. For the clinician, this distinction is vital: the toxin appears to selectively disrupt complex circuits required for coordination and memory while sparing basic locomotor output.

Variable myeloid depletion and localized neuroinflammation

The researchers utilized CD11b-DTR mice, a transgenic model engineered to express the diphtheria toxin receptor on myeloid cells (a lineage of immune cells that includes microglia and macrophages), to evaluate the efficacy and safety of targeted cell depletion. The study utilized a comprehensive set of assessment parameters including survival, motor and cognitive behavior, myeloid cell changes, and neuropathology to characterize the effects of the toxin. While the toxin proved toxic to all groups, CD11b-DTR mice exhibited greater susceptibility to diphtheria toxin than wild-type mice, demonstrating reduced survival and the emergence of clinical illness at lower doses than their non-transgenic counterparts. This increased sensitivity suggests that the presence of the receptor on immune cells significantly amplifies the physiological impact of the toxin beyond its inherent off-target effects, potentially through the rapid release of inflammatory cytokines as the targeted cells die.

A critical finding of the study was the lack of uniformity in how the toxin affected different areas of the central nervous system. Region-specific analysis in CD11b-DTR mice showed robust myeloid cell depletion in the midbrain at higher diphtheria toxin doses, yet this effect was not observed throughout the brain. Specifically, hippocampal myeloid cell numbers remained unchanged in CD11b-DTR mice even at higher diphtheria toxin doses, indicating a failure of the depletion model in this vital cognitive region. These findings indicate that diphtheria toxin-mediated myeloid cell responses vary across brain regions, a phenomenon the authors suggest potentially reflects differential toxin exposure following ventricular delivery. For the clinician, this regional variation is significant because it suggests that experimental outcomes in these models may be confounded by incomplete depletion in specific neuroanatomical sites, such as the hippocampus, which is central to Alzheimer's and other dementias.

The lack of depletion in the hippocampus did not signify a lack of biological impact. Instead of being eliminated, the hippocampal myeloid cells showed marked morphological signs of activation in CD11b-DTR mice, suggesting a shift toward a pro-inflammatory state. This localized activation, occurring in the absence of cell death, indicates that the toxin may trigger inflammatory signaling that complicates the interpretation of behavioral data. Because the toxin induces both targeted depletion in some regions and unintended activation in others, the researchers emphasize that behavioral phenotypes in these models must be interpreted with caution, as they may reflect localized neuroinflammation rather than the intended loss of myeloid cell function.

Focal CNS pathology and clinical implications for research

The researchers identified specific neuropathological changes in the central nervous system that occurred independently of the intended myeloid depletion. In a subset of clinically affected animals treated with diphtheria toxin, histological examination revealed focal brain abnormalities including ventriculitis (inflammation of the ventricular system) and meningoencephalitis (inflammation of the brain and its membranes). Furthermore, the study documented spongiotic changes (microscopic vacuoles in the gray matter) within the brain tissue of these subjects. These findings suggest that the administration of the toxin itself, rather than the loss of specific immune cells, can induce significant structural damage and inflammatory responses within the central nervous system. Such spongiotic changes are particularly concerning as they mimic the vacuolation seen in prion diseases and certain metabolic encephalopathies, potentially masking or mimicking the very disease states researchers aim to study.

Notably, the toxic effects of the intracerebroventricular administration appeared to be sequestered within the neurological compartment. The researchers reported that peripheral organs in diphtheria toxin-treated animals were largely unremarkable, showing no evidence of systemic toxicity. Additionally, hematological changes in diphtheria toxin-treated animals were infrequent, suggesting that the observed mortality and behavioral deficits were not driven by systemic inflammatory responses or bone marrow suppression. This localization emphasizes that the observed pathology is a direct consequence of the toxin's presence within the brain environment rather than a secondary effect of systemic illness. This isolation is important for clinicians to note, as it confirms that the neurological deficits observed in these models are primary CNS events.

For the practicing physician, these findings necessitate a critical appraisal of preclinical studies utilizing diphtheria toxin receptor-based depletion models. Because the toxin can induce focal CNS pathology and behavioral dysfunction in the absence of the target receptor, outcomes attributed to microglial or macrophage loss may instead be artifacts of the toxin itself. When evaluating research that employs these models to study neurodegenerative or neuroinflammatory conditions, clinicians should look for the inclusion of appropriate toxin-treated wild-type controls and careful dose optimization to ensure that the reported therapeutic or pathological insights are truly representative of the underlying disease biology rather than a byproduct of the experimental methodology.

References

1. Malvaso A, Gatti A, Negro G, Calatozzolo C, Medici V, Poloni TE. Microglial Senescence and Activation in Healthy Aging and Alzheimer’s Disease: Systematic Review and Neuropathological Scoring. Cells. 2023. doi:10.3390/cells12242824

2. Colonna M, Butovsky O. Microglia Function in the Central Nervous System During Health and Neurodegeneration. Annual Review of Immunology. 2017. doi:10.1146/annurev-immunol-051116-052358

3. Varol C, Mildner A, Jung S. Macrophages: Development and Tissue Specialization. Annual Review of Immunology. 2015. doi:10.1146/annurev-immunol-032414-112220

4. Spangenberg EE, Severson P, Hohsfield LA, et al. Sustained microglial depletion with CSF1R inhibitor impairs parenchymal plaque development in an Alzheimer’s disease model. Nature Communications. 2019. doi:10.1038/s41467-019-11674-z

5. Zhou L, Wang Y, Xu Y, Zhang Y, Zhu C. A comprehensive review of AAV-mediated strategies targeting microglia for therapeutic intervention of neurodegenerative diseases. Journal of Neuroinflammation. 2024. doi:10.1186/s12974-024-03232-2