The Cuprizone Mouse Model
Jennifer MacArthur and Theodora Papanikolaou, Ph.D.
When administered orally in mice, cuprizone, a copper chelator, causes rapid demyelination and gliosis, or rapid proliferation of glia subtypes. The cuprizone mouse model captures several aspects of MS pathology, bypassing the autoimmune component in other preclinical models.
Cuprizone treatment is the most frequently used among the toxicant-induced MS models, which include lysolecithin and ethidium bromide, and is used to study mechanisms of oligodendrocyte turnover, astrogliosis, and microgliosis (Blakemore and Franklin, 2008). Unlike other toxicant-induced models, which are introduced in the brain by stereotaxic microinjections and result in focal demyelination, oral administration of cuprizone produces a global insult. Different rodent strains react in idiosyncratic ways to cuprizone, which is best characterized in the C57BL/6 strain. For example, unlike C57BL/6 mice, BALB/cJ mice present with a delayed and incomplete demyelination in certain brain regions following cuprizone treatment (Skripuletz et al., 2008), while rats are resistant to demyelination and therefore do not represent a suitable model to study remyelination events (Love, 1988).
Because a number of studies have reported that female mice are more resistant to toxic demyelination induced by cuprizone, typically, male C57BL/6 mice are fed with cuprizone, which leads mainly to selective oligodendrocyte death and reproducible demyelination in the brain (Hibbits et al., 2009; Steelman et al., 2012).
In this acute paradigm, robust remyelination follows a predictable time course. The animal’s gender, age, and exposure time to cuprizone are important determinants of the reproducibility and sequence of de- and remyelination events. Prolonged cuprizone exposure induces chronic demyelination and impairs the capacity of the brain to repair. Similarly, age decreases oligodendrocyte turnover and progenitor recruitment to the demyelination site (Doucette et al., 2010). Delayed expression of growth factors, recruitment of histone deacetylases, and further differences in epigenetic and transcriptional control in older animals account for the age-related decline in remyelination efficiency. Gender is another rate-limiting factor to consider when using the cuprizone model. Taylor et al. (2009) documented resistance of female mice in specific strains to remyelination due to possible hormonal and genetic differences. Consistent with this, combined treatment of estradiol and progesterone protect against cuprizone-induced demyelination (Acs et al., 2009).
In the acute cuprizone paradigm, male C57BL/6 mice at 6 to 9 weeks of age are fed a diet of chow mixed with 0.2% cuprizone over the course of 6 weeks. By the third week of cuprizone feeding, consistent demyelination can be observed in the corpus callosum, the largest white matter tract in the mouse brain. Demyelination reaches a maximum at 5 or 6 weeks (Hibbits et al., 2009; Skripuletz et al., 2008; Steelman et al., 2012). When mice are returned to a diet of normal chow, spontaneous robust remyelination occurs in the corpus callosum, which reaches completion by 3 to 5 weeks after the toxicant is withdrawn.
Consistent reports from different laboratories emphasize the rostrocaudal demyelination pattern within the corpus callosum giving a patchy appearance (Binder et al., 2008; Stidworthy et al., 2003; Wu et al., 2008). The midline of the superficial portion of the genu in the rostral corpus callosum appeared substantially demyelinated, while in the dorsal part of the corpus callosum, oligodendrocytes were completely depleted, as indicated by the 90% glutathaionine S-transferase–positive neurons in the splenium of cuprizone-treated animals (Steelman et al., 2012). Other brain regions including the dorsal hippocampal commissure and to a lesser extend the cortex are also vulnerable to cuprizone (Koutsoudaki et al., 2009).
In parallel and despite the presence of the toxicant, oligodendrocyte precursors mature and begin partial remyelination of the corpus callosum at the height of demyelination (Matsushima and Morell, 2001). Increase in G-ratios—the ratio of axon circumference to myelin circumference—from 0.802 to 0.926 and 0.964 at week 4 and 6 of the cuprizone treatment represents additional evidence that remyelination has taken place (Lindner et al., 2008). Wendy Macklin’s laboratory has recently developed a technique to better separate the demyelination and remyelination phases of the cuprizone model. Investigators inject cuprizone-fed mice with rapamycin, a drug that blocks mTOR and spontaneous remyelination, allowing better quantification of oligodendrocyte turnover. (mTOR is a serine/threonine kinase involved in several processes including cell survival, protein synthesis, transcription, and oxidative stress.) With cuprizone alone, her laboratory documents 30% to 40% myelination at 6 weeks, with remyelination complete by 3 to 5 weeks off the toxicant. With cuprizone and rapamycin together, the group detects 0% myelination at 6 weeks and complete remyelination 7 weeks later. “You have a clean negative and a clean positive,” Dr. Macklin said in an interview with Multiple Sclerosis Discovery Forum (MSDF).
Unraveling the molecular mechanisms of this potentially predetermined remyelination pattern and identifying oligodendrocyte progenitor populations with repair capacity may be the basis for future disease-modifying therapies in MS. A vital but still poorly understood component of this endogenous replacement mechanism is the prominent astrogliosis and microgliosis at the demyelination sites. Microglia are potentially recruited from other brain regions. However, conflicting reports make their function unclear in the context of oligodendrocyte turnover after demyelination. Microglia account for proinflammatory production of cytokines, which are neuroprotective (Kotter et al., 2001; Simard and Rivest, 2007; Turrin and Rivest, 2006), while other research groups have described their neurotoxic effects (Kim and de Vellis, 2005; Walker and Lue, 2005).
Chronic demyelination can be induced if C57BL/6 mice are maintained on a diet with cuprizone for 12 weeks (Hibbits et al., 2009; Matsushima and Morell, 2001; Skripuletz et al., 2010; Skripuletz et al., 2008). Systemic symptoms appear at 14 weeks, and C57BL/6 mice do not survive past 16 weeks. However, certain other mouse strains do live up to 6-7 months. Oligodendrocytes and oligodendrocyte progenitors in this case undergo programmed cell death and eventually become depleted. Interestingly, transplantation of O4 + (a marker for oligodendrocyte precursors) induces repair of chronically demyelinated axons, indicating that chronic demyelination results due to depletion in the oligodendrocyte pool rather than loss of their capacity to repair (Mason et al., 2004).
Key pathological features
Cuprizone (bis-cyclohexanone oxaldihydrazone)—a copper chelator—targets mature oligodendrocytes, the cells that make up the myelin sheath of axons in the CNS (Suzuki and Kikkawa, 1969). Copper is an essential element for the function of several metalloproteases, and decreased levels are observed in a number of neurodegenerative disorders. A consistent pattern of demyelination has been observed mainly within the corpus callosum. Oligodendrocyte depletion was also reported in the cerebellar peduncles, cortex, and dorsal hippocampal commissures. In parallel, reactive astrocytes and microglia populate the areas of demyelination and secrete a battery of proinflammatory cytokines including TNF-α, interleukin-1β, and interferon-γ, which regulate the demyelination events (Arnett et al., 2001; Blakemore, 1972; Hiremath et al., 1998; Mason et al., 2001; Voss et al., 2012). Decreases in transcriptional and translational levels of phospholipases A2 (Palumbo et al., 2011) and myelin-associated glycoprotein (MAG), myelin basic protein, ceramide, and galactosyltransferase precede the onset of demyelination (Jurevics et al., 2001; Norkute et al., 2009). Increased concentrations of cuprizone lead to mitochondrial dysfunction (megamitochondria) by enlargement or fusion and reduction in cytochrome and monoamine oxidase (Suzuki and Kikkawa, 1969).
As oligodendrocyte precursors mature, robust myelin repair can be observed. The pattern of demyelination and spontaneous remyelination in the cuprizone model occurs over a highly predictable time course and within anatomically distinct areas. This reproducibility allows for the focused and quantitative study of lesion generation in addition to molecular processes and cell-to-cell interactions during the repair phase.
Key clinical features
One difficulty that researchers face in using this model is that no overt signs—such as paralysis or other obvious nervous system impairments—result from the demyelination. This observation is consistent with what is seen in the clinic; many times demyelinating lesions on a patient’s MRI do not correlate with obvious physical symptoms (see “More Than Meets the Eye”).
According to Brian Popko, Ph.D., a neuroscientist at the University of Chicago, the outward effects that occur in cuprizone-fed mice—motor activity deficits and altered social interactions—are subtle and require specific tests in order to observe them. For example, in a sensorimotor coordination assay (developed by Regina Armstrong’s laboratory at the Uniformed Services University of the Health Sciences), cuprizone-treated mice run more slowly on a hamster wheel with irregularly spaced rungs than do control mice (Hibbits et al., 2009).
In their 2011 review, Martin Stangel and colleagues find “the absence of major behavioral deficits in the face of nearly complete demyelination in the CNS … striking” (Skripuletz et al., 2011). They suggest that cuprizone-treated mice do not show major neurological symptoms because the direct damage to axons is minimal.
Cuprizone is a simple and reliable model for inducing and examining demyelination and remyelination. Other toxicant-induced models require complex stereotaxic surgeries. Furthermore, the insult is reproducible in a highly temporal and spatial manner within the corpus callosum and is straightforward to detect and score through a battery of biomarkers and microscopic analysis.
Other models, such as experimental autoimmune encephalomyelitis, induce demyelinating lesions by triggering T-cell activation and infiltration into the brain, with subsequent immune attack. The cytokines and other factors released by these inflammatory cells introduce additional variables, which complicate interpretation of the oligodendrocyte cell death.
The cuprizone model circumvents the inflammatory component and makes it possible to tease out the factors that enhance or inhibit myelin repair. Also, unlike viral or autoimmune models, cuprizone treatment can be ended; this system therefore enables researchers to study remyelination in the absence of continued demyelination. As a model system it also serves as an excellent platform to conduct high-throughput screens and identify novel compounds with promising effects for oligodendrocyte survival and axonal myelination.
Because the cuprizone model is well characterized in C57BL/6 mice, the strain most commonly used to create targeted gene knockouts, it is possible to study the roles of specific genes and cell-to-cell interactions involved in the remyelination process.
While there is an MRI technique that can show lesions in these mice, no dependable noninvasive method exists to distinguish the rate of remyelination in cuprizone-fed mice from that in untreated animals. Researchers need to sacrifice the animals and dissect out the corpus callosum to document the effects of the toxicant and the various remyelination factors under study. This feature makes it difficult to do longitudinal studies, because the same animal can’t be followed over time. Several mice are needed to power those studies adequately. Fluctuations in hormonal levels among female mice are additional variables that limit the use of cuprizone to male mice. Progesterone and estrogen delay the demyelination process and speed up remyelination events. Such an outcome can significantly interfere with the results and conclusions related to the repair capacity of the brain.
“It is a decent model, but cuprizone can be a pain,” Wendy Macklin, Ph.D., a cell and developmental biologist at the University of Colorado School of Medicine in Denver, told MSDF. She noted that cuprizone’s toxicity can vary from batch to batch. “You need to use so many animals, and it takes a month to know if it works. It’s very labor-intensive. Even though demyelination is clear and localized, rapid underlying remyelination occurs by about 4 weeks after treatment initiation. The error bars are large.”
Disease processes that can be studied
Cuprizone is best suited for studying factors that affect oligodendrocyte turnover and myelin repair. The simplicity of the model, technically and physiologically, makes it ideal for testing hypotheses about what might hinder remyelination in MS patients. Cuprizone-associated oligodendrocyte apoptosis and MAG mRNA loss during early demyelination greatly mimic hallmarks of the pathophysiology of primary progressive MS and to a lesser extent progressive relapsing MS as no other experimental autoimmune model does (Lucchinetti et al., 2000).
Disease processes that cannot be studied
Cuprizone is a poor model for understanding the immune system’s involvement in lesion formation or other potential causes of MS. The etiology of demyelinating disease in humans is unknown, but there’s ample evidence that environmental and genetic factors are involved in the disease pathogenesis. However, cuprizone causes demyelination by directly killing mature oligodendrocytes in the mouse brain, without inflammation around the blood vessels or involvement of the immune system. Demyelination does not underlie a perivenous distribution, and the borders are not clearly defined. Thus, the toxicant’s effects are unlikely to mimic the complex pathology of MS in humans and specifically relapsing-remitting MS and secondary progressive MS.
Utility in uncovering relevant biology
This model can be used to identify important pathways or molecules involved in myelin repair, including transient increase in Olig 2 levels before levels of myelin-specific proteins are decreased and proinflammatory cytokines are increased during the demyelination-remyelination phase (Morell et al., 1998). Cuprizone-treated mice or neuronal cultures could also be used in preclinical studies to test the efficacy of drugs designed to enhance lesion repair. Fumaric acids did not appear to spare oligodendrocyte cell death in cuprizone-treated mice (Moharregh-Khiabani et al., 2010), while glatiramer, one of the most common approved treatments, was tested in this preclinical model. The beneficial effects were attributed to the secretion of interleukin-10 and -4 (Rosato Siri et al., 2013). Moreover, targeting knockdown of FGF2 and FGFR1 accelerated the remyelination process in cuprizone-treated animals (Mierzwa et al., 2013).
In an interview with MSDF, Popko pointed out that cuprizone results in a reproducible lesion. “You have to know exactly where to look in the brain,” he said. The literature indicates the specific coordinates to focus on within the corpus callosum.
Although cuprizone experiments traditionally use male mice, more recent studies have demonstrated no gender differences between male and female C57BL/6 mice. Nevertheless, most studies still use only male mice. To assess the extent of myelination, researchers need to sacrifice the animals, perfuse them to fix the brain tissue, and dissect out the brain. Electron microscopy allows precise measurement of the myelin sheath thickness and the number of myelinated axons. Immunohistochemistry and biochemistry with labeled antibodies specific for axons, myelin, or oligodendrocytes give a more qualitative picture of the demyelinating lesions and the remyelination process.
Cuprizone (bis-cyclohexanone oxaldihydrazone) can be purchased from Sigma-Aldrich (cat# C9012). To add cuprizone to standard mouse chow, the chow must be ground into a powder and mixed with the cuprizone (under a fume hood; cuprizone is highly toxic when inhaled or swallowed as well as upon contact with eyes and skin!). Vendors such as Harlan (cat# TD.01453) will prepare the 0.2% cuprizone diet by special order if the researcher provides the chemical. Cuprizone is heat-sensitive, so it is recommended that it be stored at between 2-8°C to retain effectiveness. Cuprizone chow must also be stored at 4°C until use.
The copper-chelating properties of cuprizone were first discovered in the 1950s (Nilsson, 1950), and the compound was first used to study the effects of copper in biological specimens like serum and plasma (Peterson and Bollier, 1955). Toxic effects of cuprizone associated with spongiform encephalopathy, brain edema and liver mitochondrial dysfunction were later reported by Carlton (Carlton, 1967). The cuprizone mouse model was initially developed in the 1970s by William Blakemore at Cambridge University, Samuel Ludwin at Queen’s University in Ontario, and Kinuko Suzuki at the University of Pennsylvania to study demyelination and myelin repair (Blakemore, 1972, 1973). However, the high doses of cuprizone used caused liver toxicity or death. In the 1990s, Glenn Matsushima at the University of North Carolina, Chapel Hill, School of Medicine systematically determined the dose of cuprizone (0.2%) that would produce demyelination and subsequent remyelination without affecting the liver (Hiremath et al., 1998). Matsushima’s group also established the age, duration of treatment, and the main location of the insult focused on the corpus callosum, where the extent of demyelination could be scored easily and consistently. Since then the cuprizone model has been used routinely to understand molecular mechanisms of oligodendrocyte turnover and identity agents that prevent or reverse their degeneration.