The pharmacokinetics and dosing of oral 4‐methylumbelliferone for inhibition of hyaluronan synthesis in mice

Summary

Recently, there has been considerable interest in using 4‐methylumbelliferone (4‐MU) to inhibit hyaluronan (HA) synthesis in mouse models of cancer, autoimmunity and a variety of other inflammatory disorders where HA has been implicated in disease pathogenesis. In order to facilitate future studies in this area, we have examined the dosing, treatment route, treatment duration and metabolism of 4‐MU in both C57BL/6 and BALB/c mice. Mice fed chow containing 5% 4‐MU, a dose calculated to deliver 250 mg/mouse/day, initially lose substantial weight but typically resume normal weight gain after 1 week. It also takes up to a week to see a reduction in serum HA in these animals, indicating that at least a 1‐week loading period on the drug is required for most protocols. At steady state, more than 90% of the drug is present in plasma as the glucuronidated metabolite 4‐methylumbelliferyl glucuronide (4‐MUG), with the sulphated metabolite, 4‐methylumbelliferyl sulphate (4‐MUS) comprising most of the remainder. Chow containing 5% but not 0·65% 4‐MU was effective at preventing disease in the experimental autoimmune encephalomyelitis (EAE) mouse model of multiple sclerosis, as well as in the DORmO mouse model of autoimmune diabetes. While oral 4‐MU was effective at preventing EAE, daily intraperitoneal injections of 4‐MU were not. Factors potentially affecting 4‐MU uptake and plasma concentrations in mice include its taste, short half‐life and low bioavailability. These studies provide a practical resource for implementing oral 4‐MU treatment protocols in mice.

Introduction

Hyaluronan (HA) is an extracellular matrix (ECM) glycosaminoglycan (GAG) with many roles in normal tissue function and development 1, 2, 3. HA is synthesized by three independently regulated hyaluronan synthase (HAS) proteins, which generate predominantly high molecular weight HA (HMW‐HA) of >5×10⁵ Da 4. These enzymes lengthen hyaluronan by adding uridine diphosphate (UDP)‐glucuronic acid and UDP‐N‐acetyl‐glucosamine repeatedly to the anchored UDP‐reducing end of the nascent hyaluronan polysaccharide as it is extruded through the cell membrane into the extracellular space 4. HA levels are elevated greatly in injured tissues and this accumulation of HA is thought to contribute to inflammation in many chronic diseases, including type 2 diabetes 5, 6, multiple sclerosis (MS) 7 and others 8, 9, 10.

Recently, we reported that insulitis in autoimmune type 1 diabetes (T1D) was associated with islet‐specific deposition of HA 11. We have made similar observations in animal models of autoimmune diabetes, including non‐obese diabetic (NOD) mice 12 and D011.10xRIPmOVA (DORmO) mice 13. Together with recently published histological and biochemical analyses by our group and others of islet ECM in non‐diabetic human and murine islets 14, 15, these data implicate HA and islet ECM in the onset of T1D.

Similarly, HA is highly abundant within central nervous system (CNS) tissues under autoimmune attack. High levels of HA are present within demyelinated lesions in MS and in the model of the disease, experimental autoimmune encephalomyelitis (EAE), where it facilitates the extravasation of activated T cells into the CNS 7, 16. HA has also been implicated in other autoimmune diseases, including rheumatoid arthritis 17 and lupus 18.

HA is known to have substantial effects on immune phenotypes. At sites of inflammation HA typically takes the form of catabolized low molecular weight fragments (LMW‐HA), which trigger damage‐associated pattern recognition receptors, such as Toll‐like receptors 2 and 4 (TLR‐2, TLR‐4) 19, 20, 21. This, in turn, drives the activation and maturation of dendritic cells (DC) 22, the release of proinflammatory chemokines and cytokines 23, 24, and proliferation 25 and migration 26 of leukocytes. These signals may be particularly relevant in settings of sterile inflammation 27, 28. HA and its receptor interactions are also known to influence both the number and function of the T cell effector subsets that both drive and prevent autoimmunity 29, 30, 31, 32, 33, 34, 35.

Based on HA's role as an inflammatory mediator, there has been an increasing number of studies targeting HA synthesis using the small molecule inhibitor of HA synthesis 4‐methylumbelliferone (4‐MU) 36. This drug has been shown to inhibit HA production in multiple cell lines and tissue types both in vitro and in vivo 17, 37, 38, 39, 40. We have reported that 4‐MU treatment prevented or reversed autoimmune disease in mouse models of T1D and MS 13, 16.

4‐MU is thought to inhibit HA production in at least two ways. First, it functions as a competitive substrate for UDP‐glucuronyltransferase (UGT), an enzyme involved in HA synthesis 41. When 4‐MU is present, UGT conjugates glucuronic acid to 4‐MU, resulting in 4‐MU‐glucuronide (4‐MUG), rather than forming UDP‐glucuronic acid, one of the two precursors of HA, consequently inhibiting its synthesis. Secondly, 4‐MU reduces expression of HAS mRNA as well as mRNA for UDP glucose pyrophosphorylase and dehydrogenase 42. The pharmacokinetics and metabolism of 4‐MU in humans are well characterized, as detailed in our recent review on the therapeutic potential of 4‐MU 36. This reflects the extensive study of 4‐MU that was performed in the course of its approval as a therapeutic agent in humans. Under the name ‘hymecromone’, it is used in multiple countries in Europe mainly as a treatment for bile disorders, due to its choleretic and biliary anti‐spasmodic activity 43.

Here, we have addressed several practical issues pertaining to the use of 4‐MU in mouse models that have arisen during the course of our recent studies of this molecule. These issues include the dosing, treatment duration, metabolism and plasma drug levels of 4‐MU. In contrast to previous rodent pharmacokinetic studies with 4‐MU given intravenously (i.v.) 44, we present data on the bioavailability of 4‐MU when given as an oral supplement, providing a practical resource to facilitate further studies of 4‐MU in mouse models of autoimmunity and other diseases.

Materials and methods

Mice

All animals were bred and maintained under specific pathogen‐free conditions, with free access to food and water, in the vivarium at the Benaroya Research Institute (Seattle, WA, USA) and the animal facilities at Stanford University Medical School (Stanford, CA, USA). DO11.10 transgenic mice were purchased from The Jackson Laboratory (Bar Harbor, ME, USA) and bred with BALB/c mice expressing RIPmOVA (available at the Benaroya Research Institute) to generate the DORmO double‐transgenic mice. All animal experiments and use procedures were approved by the Institutional Animal Care and Use Committee at the Benaroya Research Institute and/or at Stanford University Medical School.

4‐MU treatment

4‐MU (Alfa Aesar, Haverhill, MA, USA) was pressed at 5% (w/w) into mouse chow by TestDiet^® (St. Louis, MO, USA) and irradiated before shipment, as described previously 37. This dose was calculated previously to deliver approximately 150–250 mg/mouse/day 37. Mice were initiated on the 4‐MU chow between 5 and 6 weeks of age, unless noted otherwise, and were maintained on this diet until they were euthanized, unless noted otherwise.

Pharmacokinetics of 4‐MU

Mice were placed on chow containing 5% (w/w) 4‐MU for 2 weeks, after which groups of three mice were euthanized every 4 h by cardiac puncture under 2,2,2‐tribromoethanol anaesthesia. Blood was collected in heparin‐coated syringes and kept at 4°C until spun down to collect the plasma. Plasma levels of 4‐MU, 4‐MUG and 4‐MUS were determined by high‐performance liquid chromatography (HPLC)‐mass spectrometry.

Induction of EAE

EAE was induced in female C57BL/6 mice (The Jackson Laboratory) at 8–12 weeks of age by subcutaneous immunization in the flank with an emulsion containing 200 μg of myelin oligodendrocyte glycoprotein peptide 35–55 (MOG_35–55) in saline and an equal volume of complete Freund's adjuvant (CFA) containing 400 ng of mycobacterium tuberculosis H37RA (Difco Laboratories, Detroit, MI, USA). All mice were administered 400 ng of pertussis toxin (List Biological, Campbell, CA, USA) intraperitoneally (i.p.) at 0 and 48 h post‐immunization. Mice were given food and water ad libitum and were weighed and monitored daily for clinical symptoms as follows: 0, no clinical disease; 1, tail weakness; 2, hindlimb weakness; 3, complete hindlimb paralysis; 4, hindlimb paralysis and some forelimb weakness; 5, moribund or dead. Mice were given 4‐MU chow starting 4 days before immunization, or were injected i.p. once daily with 1 mg of 4‐MU in 200 μl 0.08% carboxymethylcellulose (4‐MU was diluted ×100 from a stock solution in dimethyl sulphoxide), as described in the figure legends.

Weight and diabetes monitoring

Beginning at 6 weeks of age, mice were weighed daily both before and after the administration of 4‐MU chow. For blood glucose measurements, mice were bled via the saphenous vein or tail vein for the determination of their blood glucose level using a Contour blood glucose meter and blood glucose monitoring strips (Bayer Healthcare, Tarrytown, NY, USA). When two consecutive blood glucose readings of 250 mg/dl were recorded, animals were considered diabetic. When two consecutive blood glucose readings of 300 mg/dl were recorded, animals were euthanized.

HA quantification

Plasma samples were thawed and assayed for HA levels in triplicate in a single batch using a modified HA‐enzyme‐linked immunosorbent assay (ELISA), as described earlier 45. Briefly, total amount of HA was determined by a modified competitive ELISA‐like assay in which the samples to be assayed were first mixed with biotinylated HA‐binding protein (b‐HABP) and then added to HA‐coated microtitre plates, the final signal being inversely proportional to the level of hyaluronan added to the b‐HABP.

Immune cell activation and cytokine analysis

Mouse splenocytes were isolated from BALB/c mice, as described previously 16. Briefly, spleens were homogenized through a strainer and red blood cells were lysed in the cell suspensions. Splenocytes were then cultured in triplicate at 5 × 10⁶ cells/ml and stimulated with concanavalin A (ConA, 2 μg/ml) in RPMI‐1640 supplemented with L‐glutamine (2 mM), sodium pyruvate (1 mM), non‐essential amino acids (0·1 mM), penicillin (100 U/ml), streptomycin (0·1 mg/ml), 2‐mercaptoethanol (5 × 10⁻⁵ M) and 10% fetal bovine serum. Supernatants were collected after 48 h and cytokine levels were quantified using anti‐mouse OPTEIA ELISA kits from BD Pharmingen [San Jose, CA, USA; interferon (IFN)‐γ and interleukin (IL‐4)] and R&D Systems (Minneapolis, MN, USA; IL‐17).

Statistics

Data are expressed as means ± standard error of the mean (SEM) of n independent measurements. The comparison between two groups was performed with unpaired t‐tests for normally distributed data or a Mann–Whitney U‐test for non‐parametric data. A P‐value less than < 0·05 was considered statistically significant. Multiple group analysis was performed using analysis of variance (anova). Data analysis was performed with GraphPad Prism version 5·0 software.

Results

4‐MU‐mediated suppression of HA serum levels requires prolonged treatment

One practical question in using oral 4‐MU treatment in mice is how long it is necessary to treat animals to suppress HA production. To address this, we incorporated 5% 4‐MU into standard mouse chow, a formulation which we previously established delivers 250 mg/mouse/day 37. We then studied C57BL/6 mice given 5% 4‐MU chow over a 2‐week treatment period and tracked serum HA levels over this time.

We found that a significant decrease in serum HA levels was not seen until day 7 with this regimen (Fig. 1). Previously, we have shown that 4‐MU treatment reduces clinical symptoms significantly in both the DORmO model of T1D [using offspring of ovalbumin (OVA)‐specific T cell receptor transgenic DO11.10 mice crossed with RIPmOVA mice, which express OVA in conjunction with the insulin gene promoter] and the EAE model of MS 13, 16. The time–course in reduction of serum HA levels shown here is consistent with our functional biomarker data in these models, which indicated that a period of 2–7 days on 5% 4‐MU chow was required until full effects on blood glucose or neurological scores were seen 13, 16.

Figure 1.

Kinetics of hyaluronan (HA) suppression in blood upon in‐vivo treatment with 4‐methylumbelliferone (4‐MU). C57BL/6 mice were fed chow containing 5% 4‐MU or control chow ad libitum. Serum from both treatment groups was taken at baseline (day 0) and on days 2, 4 and 7, and HA levels were measured in triplicate using an HA‐specific enzyme‐linked immunosorbent assay (ELISA) assay. *P < 0·05, n = 3–5 mice per group.

These data indicate that a pretreatment period is required in vivo if reduced HA serum levels are desired at disease‐onset in mouse models.

4‐MU induces temporary weight loss

To understand how consumption patterns influence oral intake of 4‐MU, we tracked weight and food intake of C57BL/6 and BALB/c mice fed 4‐MU versus control chow. 4‐MU chow evidently has a poor taste, as mice given chow without supplemental flavouring would not eat this at all (data not shown). Chocolate flavouring was required to render the chow palatable to mice. Even then, mice tended to shred the chow and not eat it for the first week. Consequently, they had a significant decrease in weight upon initiation of this treatment. After approximately 1 week they began eating the chow and gained weight at a similar pace to that of mice fed control chow. This was the case for both C57BL/6 (Fig. 2a) as well as BALB/c mice (Fig. 2b). We note that C57BL/6 mice tend to lose more weight on this regimen than BALB/c mice. A similar pattern was seen in two disease models, the MOG‐induced EAE model of MS (which uses C57BL/6 mice; Fig. 2c) and the DORmO model of autoimmune diabetes (which uses mice on a BALB/c background; Fig. 2d). In the DORmO model, the weight of mice fed 4‐MU chow eventually overtook those of mice fed control chow, due probably to 4‐MU‐mediated amelioration of disease severity in these mice 13, 16.

Figure 2.

Body weight is temporarily reduced in mice fed 4‐methylumbelliferone (4‐MU). (a,b). Body weight of C57BL/6 (a) and BALB/c (b) mice fed chow containing 5% 4‐MU or control chow. Mice were 9 weeks of age at the initiation of the experiment and were followed‐up to 16–17 weeks of age. (c). Body weight of C57BL/6 mice after induction of experimental autoimmune encephalomyelitis (EAE). Mice were fed chow containing 5% 4‐MU or control chow starting at 8 weeks of age (black arrow), 4 days before immunization (grey arrow). The dotted line depicts the onset of clinical symptoms in this model. (d). Body weight of D011.10xRIPmOVA (DORmO) mice fed chow containing 5% 4‐MU or control chow. Mice were 8 weeks of age at the initiation of the experiment and were followed‐up to 20 weeks of age, at which point 100% of DORmO mice fed control chow were diabetic, while none of the DORmO mice fed 4‐MU were. *P < 0·05, n = 7–12 animals per group.

These data indicate that a period of time is required to allow mice on 4‐MU to regain weight and normal food intake patterns and suggest that the time lag before serum HA is reduced on this treatment probably reflects limited initial 4‐MU chow intake.

Oral but not i.p. 4‐MU prevents autoimmunity

We demonstrated recently that 4‐MU treatment prevents autoimmune disease in the EAE mouse model of MS 16. We also examined whether the route of delivery impacted the efficacy of 4‐MU in this model. Indeed, we found that oral 4‐MU treatment inhibited disease (Fig. 3a), whereas once‐daily i.p. injections of 50 mg/kg 4‐MU (a dose estimated to be in the range of that reaching the blood by a 5% formulation of 4‐MU chow 37) did not (Fig. 3b).

Figure 3.

Oral but not intraperitoneal (i.p.) 4‐methylumbelliferone (4‐MU) delivery prevents autoimmunity. (a,b). Experimental autoimmune encephalomyelitis (EAE) scores of mice following treatment with 4‐MU delivered (a) in chow at 5% or (b) via daily i.p. injections of 50 mg/kg. *P < 0·05, **P < 0·01, ***P < 0·001, Mann–Whitney comparing treated mice with untreated mice (n = 5–10 animals per group).

These data indicate that processes involved in uptake of 4‐MU through oral administration contribute to its therapeutic efficacy.

4‐MU is rapidly metabolized in vivo

We next sought to quantify 4‐MU and its metabolites in the plasma of mice receiving a stable oral dose of 4‐MU. Metabolism of 4‐MU occurs via conjugation to either a glucuronic acid to form 4‐MUG or a sulphate, forming 4‐MU sulphate (4‐MUS) (Fig. 4a). We allowed mice to acclimate to 5% 4‐MU chow for 2 weeks prior to analysing plasma levels. We then sampled plasma from these animals at frequent intervals over a 24‐h time‐period. Upon sampling, we found that plasma drug concentrations of 4‐MU were quite low in mice on 5% 4‐MU chow, averaging 0·89 ± 0·51 µg/ml (5·1 ± 2·9 µM) over 24 h (Fig. 4b).

Figure 4.

4‐methylumbelliferone (4‐MU) is metabolized rapidly in vivo. (a) Molecular structures for 4‐MU and its primary metabolites, 4‐methylumbelliferyl glucuronide (4‐MU) and 4‐methylumbelliferyl sulphate (4‐MUS). (b) Concentrations of 4‐MU and its metabolites in plasma isolated over a 24‐h cycle from mice that had been fed chow containing 5% 4‐MU for 2 weeks. (c) Concentrations of 4‐MU and its metabolites in plasma of animals that had been fed chow containing 5% or 0·65% 4‐MU for 2 weeks. *P < 0·05 comparing 4‐MUG to 4‐MU. #P < 0·05 comparing 4‐MUS to 4‐MU, n = 3 animals per group.

In contrast, the concentrations of 4‐MU metabolites were much higher. In man, 4‐MUG is the predominant metabolite 46. Consistent with this, we observed that in mice fed 5% 4‐MU chow 4‐MUG was the most abundant metabolite found in plasma, yielding an average concentration of 531 ± 66 µg/ml (1519 ± 189 µM), while 4‐MUS was present in much lower concentrations (30·5 ± 2·1 µg/ml; 119 ± 8 µM) (Fig. 4b). However, concentrations of both metabolites were significantly higher than the parent 4‐MU (P < 0·001) (Fig. 4b).

Plasma 4‐MUG drug concentrations varied in mice depending on the time of day. They tended to be higher during the night and lower during the day (Fig. 4b). Given that mice are nocturnal, the fluctuations in concentrations are consistent with the expected higher exposures during periods of feeding (i.e. night) followed by much lower concentrations during periods of fasting (day), due presumably to rapid metabolism. Plasma concentrations of 4‐MU, by comparison, were uniformly low, irrespective of the time of day (Fig. 4b), another indication that it is metabolized rapidly during oral uptake.

4‐MU dosage impacts its efficacy in treating autoimmunity

We then asked whether lower doses of 4‐MU were sufficient to prevent autoimmunity in our mouse models of autoimmunity. Similar to our previously published results 16, oral treatment with chow containing 5% 4‐MU, started after onset of symptoms, decreased the severity of MOG_35–55‐induced EAE in C57BL/6 mice significantly, indicating that 4‐MU is able to affect already established disease processes (Fig. 5a). However, this effect was not observed in mice treated with chow containing 0·65% 4‐MU (Fig. 5b). Similarly, whereas the higher dose of 4‐MU was effective in preventing onset of diabetes in DORmO mice, the lower 0·65% formulation of chow was eventually unable to prevent disease in this model (Fig. 5c,d).

Figure 5.

4‐methylumbelliferone (4‐MU) dosage impacts its efficacy in treating autoimmunity. (a,b). Impact of 4‐MU dose on the severity of experimental autoimmune encephalomyelitis (EAE). Treatment with either (a) 5% or (b) 0·65% 4‐MU in the chow was initiated after the onset of symptoms. *P < 0·05, **P < 0·01, ***P < 0·001, Mann–Whitney U‐test comparing treated mice with untreated mice (n = 10). (c,d). Impact of 4‐MU dose on glycaemic control, indicated by blood glucose levels, in D011.10xRIPmOVA (DORmO) mice fed either (c) 5% or (d) 0·65% 4‐MU chow initiated at 6 weeks of age; n = 5 mice per group. Grey shaded boxes indicate the duration of 4‐MU treatment.

These data are explained most readily by the low bioavailability (< 2%) of 4‐MU 46 and suggest that the 0·65% 4‐MU chow formulation was not able to yield adequate concentrations of 4‐MU in the mice to be effective. Consistent with this interpretation, minimal concentrations of 4‐MU and its metabolites were detectable in the plasma of mice fed the 0·65% 4‐MU chow formulation (Fig. 4c).

4‐MU dosage impacts T helper (Th) subset polarization

In our previous report we also showed that 4‐MU significantly decreased the fraction of IFN‐γ‐producing Th1 T cells and increased the fraction of IL‐4‐producing Th2 T cells in mice fed 5% 4‐MU chow 16. We therefore next examined how 4‐MU dosage affects the impact of 4‐MU on T cell subset polarization, as assessed by cytokine production. For this, we isolated splenocytes from mice fed chow containing 5% or 0·65% 4‐MU or control chow, and assessed these for cytokine production after in‐vitro stimulation. We found that only the 5% 4‐MU chow reduced the production of IFN‐γ significantly (Fig. 6a) and increased the production of IL‐4 (Fig. 6c), whereas the 0·65% 4‐MU chow did not result in a significant change in cytokine production compared to control (Fig. 6a,b,c). We also observed a moderate increase in IL‐17 production by unstimulated cells after 5% 4‐MU treatment, as we have observed previously, but not after 0·65% 4‐MU treatment. These data are consistent with the effect of chow formulation strength on autoimmune disease in the EAE and DORmO models (Fig. 5).

Figure 6.

4‐methylumbelliferone (4‐MU) dosage impacts T helper (Th) subset polarization. (a–c). Cytokine production profiles for (a) interferon (IFN)‐γ, a Th1 cytokine, (b) interleukin (IL)−17, a Th17 cytokine and (c) IL‐4, a Th2 cytokine, of splenocytes isolated from mice fed chow containing 5% or 0·65% 4‐MU or control chow. Cells were cultured for 48 h with (grey bars) or without (black bars) activation using concanavalin A (ConA) and cytokine levels were measured in the supernatants using enzyme‐linked immunosorbent assay (ELISA). *P < 0·05, n = 5 animals per group.

Discussion

We have examined a number of variables that influence the practical implementation of oral treatment protocols to suppress HA synthesis in mouse models of autoimmunity using 4‐MU.

We report that a significant reduction in serum HA in C57BL/6 mice was observed after 7 days of treatment with oral 4‐MU at a dose of 5% of the total chow volume. While this duration may vary between different tissues, from a practical perspective it suggests that a pretreatment period is necessary to suppress systemic HA synthesis. In contrast, using pancreatic islets cultured in vitro with 4‐MU, significant suppression of HA synthesis was observed after 24 h 13. This difference may reflect the pharmacokinetics of 4‐MU in vivo after oral uptake 46. However, mice also take time to adjust to the unfavourable taste of oral 4‐MU, and we speculate that they may not consume sufficient chow initially to achieve suppression of HA synthesis. Regardless, it has become our practice to treat for a minimum of 1 week prior to performing functional analyses of mice on 4‐MU.

Notably, even long‐term oral 4‐MU treatment does not remove all HA in pancreatic islets 13, EAE CNS 16 or other tissues 37, suggesting that some HA is slow to turn over and/or that 4‐MU‐mediated inhibition of HA synthesis is not 100% efficient. We speculate that the residual HA seen in histological stains of mice treated with 4‐MU may be bound substantially to hyaladherins 47, 48 or otherwise less susceptible to catabolism.

We found initially that C57BL/6 mice typically lost significant weight on 4‐MU chow, despite the inclusion of chocolate flavouring. After the first week, however, mice fed 4‐MU start to gain weight again at the same rate as mice fed control chow. None the less, they often do not catch up to mice fed control until later in the disease process, when studying disease models. This may be relevant to treatment protocols where body weight is a significant variable. For example, this diminished food intake may account for some of the early improvement in glycaemic control and leptin‐related regulation of neuroinflammation in the DORmO and EAE mouse models in our recent studies 13, 16, 49. However, it is unlikely to account for the long‐term effect seen in these animals, given that mice regained normal food intake within weeks and were maintained on 4‐MU chow for up to a year in these studies 13.

Of note, 4‐MU has fluorescent properties and might therefore be light sensitive. It might therefore better be stored in a darkened container to preserve efficacy (data not shown). Along related lines, it has been suggested that chow formulations older than 4 months are less effective (data not shown), implying that 4‐MU chow formulation might have a limited shelf life.

With regard to in‐vivo metabolism of 4‐MU, we find that 4‐MUG is the predominant metabolite present in plasma in mice on a steady 5% chow regimen; indeed, it was present at concentrations 300‐fold higher than those seen for the parent molecule, 4‐MU. Glucuronidation is a common mechanism to solubilize hydrophobic noxious agents, allowing them to be cleared from the body. Consistent with this, in humans it has been shown that 4‐MU is metabolized extensively and that less than 1% of a given dose is excreted unchanged in the urine 46. Glucuronidation into 4‐MUG accounts for more than 90% of this metabolism and 93% of a single i.v. dose of 4‐MU is eliminated as the 4‐MUG metabolite in the urine 46, 50.

Our data suggest that giving daily i.p. injections of 4‐MU, despite bypassing first‐pass metabolism, is not an effective treatment approach for preventing EAE. This is consistent with the high clearance of hymecromone: using i.v. infusion of 4‐MU it was shown previously that the clearance of 4‐MU is rapid, having a half‐life of ∼30 min 46. In our experiments, each day's dose was made up freshly to avoid issues of formulation stability. In light of this, our interpretation of our i.p. data is that the half‐life of 4‐MU is probably too short for daily injections to be effective. We considered more frequent dosing, but deemed this approach impractical given that weeks to months of treatment are necessary to demonstrate efficacy in some of our autoimmunity models.

The effectiveness of 4‐MU is concentration‐dependent. A 5% 4‐MU chow formulation prevented autoimmunity efficiently in both the EAE and DORmO models and polarized T cell subsets away from pathogenic Th1 cytokine production and towards Th2. However, a 0·65% 4‐MU chow formulation was ineffective. This is consistent with the known low bioavailability of 4‐MU; in one pharmacokinetic study of hymecromone in human volunteers, the systemic bioavailability of hymecromone after oral dosing was less than 2% 46. In rodents, previous studies have demonstrated the extraction of hymecromone by the gastrointestinal system (prehepatic) to be ∼40% and extraction by the liver as high as 97% 50. As a result of this high extraction, the fraction of an administered oral dose of hymecromone that reaches the systemic circulation (post‐hepatic) as unchanged drug is very low. In light of this low bioavailability, using higher doses is warranted in in vivo studies.

Together, these studies provide a practical resource for implementing 4‐MU treatment protocols in mice. Our hope is that this will contribute to further investigation of the role of HA in autoimmunity and other disease conditions and the development of potential therapeutic strategies targeting HA synthesis for these diseases.

Disclosure

The authors declare no disclosures.

Author contributions

H. F. K., N. N., T. N. W., L. S. and P. L. B. designed the experiments. H. F. K., N. N., S. M. R., V. G. K., P. M., J. A. G. and H. D. I. performed the experiments. S. K., T. N. W. and L. S. provided reagents. J. B., A. F. and P. L. B. wrote the manuscript.