Abstract
Interleukin-10 (IL-10) delivered by intrathecal (i.t.) gene vectors is a candidate investigational new drug (IND) for several chronic neurological disorders such as neuropathic pain. We performed a preclinical safety study of IL-10. A syngeneic large animal model was used delivering porcine IL-10 (pIL-10) to the i.t. space in swine by adeno-associated virus serotype 8 (AAV8), a gene vector that was previously found to be nontoxic in the i.t. space. Unexpectedly, animals became ill, developing ataxia, seizures, and an inability to feed and drink, and required euthanasia. Necropsy demonstrated lymphocytic meningitis without evidence of infection in the presence of normal laboratory findings for body fluids and normal histopathology of peripheral organs. Results were replicated in a second animal cohort by a team of independent experimenters. An extensive infectious disease and neuropathology workup consisting of comprehensive testing of tissues and body fluids in a specialized research veterinary pathology environment did not identify a pathogen. These observations raise the concern that i.t. IL-10 therapy may not be benign, that previously used xenogeneic models testing the human homolog of IL-10 may not have been sensitive enough to detect toxicity, and that additional preclinical studies may be needed before clinical testing of IL-10 can be considered.
Keywords: meningitis, interleukin-10, large-animal model, swine, intrathecal, preclinical toxicology
Porcine interleukin-10 (IL-10) was delivered to the intrathecal (i.t.) space in swine by AAV8. Animals developed ataxia and seizures requiring euthanasia. Necropsy demonstrated sterile lymphocytic meningitis. Syngeneic i.t. IL-10 therapy may not be benign. Previous xenogeneic models testing the human homolog may not have been sensitive enough to detect toxicity.
Introduction
Meningitis is an inflammation of the protective linings of the CNS and pia and arachnoid mater, in response to microbial organisms or autoimmune diseases such as systemic lupus erythematous, sarcoidosis, and vasculitis.1, 2 Immune-regulatory mechanisms in meningitis have been extensively investigated in small animal models3, 4 and through correlative studies of body fluids in patients demonstrating a pleiotropic cytokine response.5, 6, 7, 8, 9 To date, no individual cytokine or other endogenous immune regulator has been found to be sufficient for inducing fatal meningitis.
A prominent new agent under development as an investigational new drug (IND) is interleukin-10 (IL-10), a cytokine most often classified as anti-inflammatory, which can be detected in the nervous system of patients with meningitis7, 8 and (used as a therapeutic) is a candidate for treating a variety of CNS disorders such as multiple sclerosis and neuropathic pain.10 In the case of analgesic testing of IL-10, gene vectors are used to deliver the cytokine to the intrathecal (i.t.) space, the cerebrospinal fluid (CSF)-filled compartment lined by pia and arachnoid mater. At least seven laboratories10 (including ours11) demonstrated activity of IL-10 in neuropathic pain without seeing toxicity in rodents.
Here, we report preclinical safety testing of syngeneic IL-10 in a large animal model (porcine IL-10 [pIL-10]) in swine, employing a form of administration that we previously used for the i.t. delivery of other proteins in rats11 and dogs12 (i.t. adeno-associated virus [AAV] vectors). Unexpectedly, we observed clinically severe neurotoxicity when testing pIL-10.
Results
A Syngeneic Large Animal Model
Human IL-10 (hIL-10) is poorly conserved, diverging markedly from the IL-10 homologs of animal species used previously for in vivo testing. 27%, 20%, and 25% of amino acids are different in the mouse, dog, and swine, respectively, compared to the human homolog.13 Xenogeneic IL-10, such as when testing hIL-10 in animals, may not only lack activity14 but, importantly, is cleared from the CSF and blood by neutralizing anti-hIL-10 antibodies against the foreign protein, as we found in dogs starting as early as 1 week after initiation of hIL-10 treatment.12 To stay clear of such confounders, the present study tested IL-10 in a syngeneic model (AAV-encoded pIL-10 in swine).
Signs of Neurotoxicity in First Cohort
pIL-10 was delivered to the i.t. space (Figure S1) using previously established techniques of AAV vector administration11, 12 at a dose of 7 × 1012 genome copies/animal. In the first experimental cohort, three of five swine developed motor weakness and ataxia in all extremities. Animals lost the ability to stand up and to move in their pens. They became somnolent or developed recurrent generalized seizures, appeared moribund, and were unable to feed. Fever was up to 106.7°F (Figure S2A). Progression of neurologic demise was acute, requiring euthanasia on day 7, 9, or 28.
Meningitis Identified as a Pathomechanism
Neuropathologic workup of all affected animals revealed a singular presentation of sterile meningitis. Diffuse, lymphocyte-predominant meningeal infiltrates were found in the cerebellum, forebrain, and brainstem (Figures 1A–1D and S3A–S3D). Examination of the peripheral nerves (Figures S4A and S4B), heart, kidneys, and lungs (Figures S5A–S5C) was within normal limits (Figures S4C, S4D, and S5D–S5F). Microbial cultures of CSF and blood were negative. Complete blood counts with differential available for all animals were unremarkable neither indicating signs of immune deficiency nor a systemic inflammatory response and no indication of microangiopathic hemolysis. The complete metabolic profile, including electrolytes, bicarbonate, and tests for renal and liver function, was normal without any indication of organ failure or an anion-gap acidosis. Coagulation studies, prothrombin time, partial thromboplastin time, and fibrin split products were within normal limits, excluding disseminated intravascular coagulation. Thus, laboratory studies provided in Table S1 did not reveal any sign of sepsis or pending septic shock. Therefore, the cause of death was aseptic meningitis without systemic inflammation or organ dysfunction.
Figure 1.
Neuropathology of Moribund Animals
(A–F) Representative H&E-stained images of the cerebellum (A and B), forebrain (C), brainstem (D), and spinal cord (E and F). The meninges were diffusely expanded by lymphocytic infiltrates (A–C and F), vascular congestion (arrowhead; A), and lymphatic distension (plus sign; C). Meningeal infiltration extended along the Virchow-Robin space forming perivascular cuffs (asterisk; D and E) and into adjacent gray matter with mild gliosis (arrows; C). Meningeal infiltrates were predominately lymphocytic with mild numbers of eosinophils in cerebellar meninges (B) and few plasma cells and macrophages. Evidence of an infectious agent was not identified in any image. Scale bars represent 75 μm. Representative images in (A)–(D) are from three moribund animals of the first cohort (n = 25 images). Representative images in (E) and (F) are from two moribund animals of the replication experiment (n = 55 images). For additional images, see also Figure S3.
Assessment of Significance Using the Statistical Framework of a Single-Cohort Clinical Toxicity Study
Initially, the experiment had been designed as a single-cohort study similar to a phase I/II clinical trial to document safety of chronic IL-10 administration to the CNS. Statistical testing of the results (fatal neurotoxicity in three of five animals) was therefore performed against a null hypothesis (H0) specified by the clinical judgment (which rate of fatal complications would be acceptable to patients treated for a non-terminal illness such as chronic pain). A recent clinical trial of a candidate analgesic drug was closed after 6 of 128 enrolled patients developed severe CNS toxicity and 1 died as a result.15 Therefore, we assumed that at most, 1 of 100 patient deaths would be acceptable, and we accordingly set H0 = 0.01. Testing the result of our first cohort against this H0 yielded a highly significant result (p = 10−6).
Replication Cohort and Additional Infectious Disease Workup
To address rigor and reproducibility, a replication experiment was conducted by independent experimentalists not involved with the first cohort using a new preparation of pIL-10/AAV8. Tissue analysis was performed by a separate team of veterinary pathologists from a separate academic research center, including a swine infectious disease specialist (J.E.C.) and a neuropathologist (A.G.A).
The first two animals in the replication experiment both exhibited acute neurotoxicity and moribundity that required euthanasia within 48 hr of onset on day 7 or 24, triggering the statistically defined early stopping rule (at p < 0.01). The clinical presentation consisted of paresis and ataxia, inability to rise, cessation of feeding, lethargy, and fever up to 105.4°F, similar to the findings in the first cohort (Figure S2A). Microscopic examination of the forebrain and cerebellum revealed diffuse meningitis, vascular congestion, and scattered neuronal death, also closely resembling results of the initial cohort. Examination of the spinal cord revealed dense, lymphocyte-predominant meningeal infiltrates, demonstrating that meningitis extended throughout the neuroaxis (Figures 1E, 1F, and S3E). Microbial culture of seven tissue types did not show growth of any microorganism. No parasitic life forms were present in feces. Molecular diagnostics for 7 bacteria and 17 viruses in 10 tissue types were negative. Gas chromatography and mass spectrometry did not detect elevated levels of known chemicals or the presence of an unexpected compound, suggesting no underlying toxidrome to explain the clinical findings (Table S2). Thus, the clinical course and clinicopathologic studies in replication animals closely resembled those of the first cohort, confirming a pathological diagnosis of primary, idiopathic meningitis.
Highest CSF pIL-10 Levels Found in Affected Animals
pIL-10 was undetectable in the CSF at baseline and levels rose in all animals following pIL-10 vector delivery (Figure 2). Peak CSF pIL-10 levels correlated with the clinical syndrome: the highest pIL-10 concentrations were seen in five animals that developed fulminant neurotoxicity, while lower levels were detected in two animals that survived. This observation suggests that the highest- or longest-sustained IL-10 levels were causally linked to fatal neurotoxicity.
Figure 2.
IL-10 Concentration in CSF and Survival Period in Swine
CSF obtained by serial lumbar puncture was measured for IL-10 concentration. Each symbol represents the IL-10 concentration measured from one CSF sample per time point. Abbreviated time series (red) were obtained from animals exhibiting fulminant neurotoxicity, requiring premature euthanasia (time of death, black cross). Gray time series represent animals that survived to final study day 56 (survived, black pig). The first cohort (open symbols) comprised 28 CSF samples from five animals. The second cohort (closed symbols) comprised six CSF samples from two animals. Time is defined as weeks after pIL-10/AAV8 administration. CSF IL-10 concentrations obtained prior to vector administration (baseline) were below the assay detection level (5 × 10−3 pg/mL, horizontal dotted line) for all animals.
Characterization of the Inflammatory Response
The majority of meningeal infiltrates were identified as CD3+ T lymphocytes (Figure 3A). T lymphocytes extended along the spinal Virchow-Robin spaces and contributed a large portion of cells contained in perivascular cuffs (Figure 3B). T lymphocytes were also identified within the vicinity of the central canal (Figure 3C). Scattered CD20+ B lymphocytes were identified among T lymphocytes and contributed a small proportion of total cellular infiltrates in meninges (Figure 3D), in perivascular cuffs (Figure 3E), and adjacent to the central canal (Figure 3F). T lymphocytes occurred diffusely throughout the meninges (Figure 3A), while B lymphocytes were focused in close proximity to the i.t. space (Figure 3D).
Figure 3.
Cellular Characteristics of CNS Inflammatory Infiltrates
(A–F) Representative images of immunoperoxidase-stained spinal cords depicting CD3+ T lymphocytes (A–C) and CD20+ B lymphocytes (D–F). The majority of infiltrating cells found within the spinal leptomeninges (A and D), perivascular cuffs of spinal gray matter (B and E), and along the central canal (C and F) were T lymphocytes and a minority were B lymphocytes. (A) T lymphocytes infiltrated the leptomeninges diffusely and extended along the Virchow-Robin space. (D) Infiltrating B lymphocytes localized to the CSF-facing side of the leptomeninges facing the i.t. space but did not extend along the Virchow-Robin space. Scale bars represent 150 μm.
Compared to IL-10 naive controls (Figure S6A), immunoreactivity for Iba-1+ microglia was mildly increased along spinal parenchymal perivascular cuffs (Figure S6B) and was moderately increased for Iba-1+ monocytes in the meninges (Figure S6C). Compared to naive controls (Figure S6D), mild isomorphic astrocytic gliosis seen by GFAP conformed to perivascular inflammation (Figure S6E). As expected, meninges did not show glial fibrillary acidic protein (GFAP) immunoreactivity (Figure S6F).
Markers of apoptosis were assessed for neurotoxicity unrelated to inflammation. The terminal deoxynucleotidyl transferase deoxyuridine triphosphate (dUTP) nick-end labeling (TUNEL) assay (Figures S7A and S7B) and immunohistochemistry to detect Caspase-3 (Figures S7C and S7D) were used. Both markers showed no signs of neuronal apoptosis and only scattered positivity mainly of infiltrating leukocytes consistent with primary inflammation without evidence for independent (i.e., non-inflammatory cell death).
Discussion
IL-10 delivered to the brain side of the blood-brain barrier induced clinically overt, life-threatening neurotoxicity in large animals, which was fatal in the context of animal care conditions (euthanasia was required once animals lost the ability to feed and drink) and, if it occurred in humans, would require intensive care unit (ICU)-level medical care with a high risk of death. The pathomechanism was lymphocytic meningitis throughout the spinal, brainstem, cerebellar, and forebrain regions of the CNS. To our knowledge, the most extensive infectious disease workup available to date in the swine pathology field was executed by a specialized veterinary infectious disease research laboratory and did not identify an infectious cause, suggesting that meningitis occurred as a direct effect of IL-10. Reproducibility was demonstrated in a replication cohort. The AAV vectors used were provided by an established AAV vector core (NIH-dedicated) with a long-standing track record of producing AAV vectors that are nontoxic in the CNS with other therapeutic genes. To our knowledge, our experiments represent the most realistic preclinical testing protocol reported for IL-10 to date, because unlike previous experiments (by us and others), this new study used syngeneic IL-10 and a large animal species. IL-10 is currently under development for the treatment of neuroimmune disorders in humans under the explicit claim that its toxicity profile is benign,10 an assumption that may have resulted from shortcomings of preclinical studies performed to date. Large animal toxicity studies can be critical for drug development, as was exemplified by the tragic outcome of a recent clinical drug trial in which 6 of 128 enrollees suffered severe neurotoxicity and 1 patient died,15, 16 which could have been averted if fatal toxicity seen in large animals (but not in rodents) had been disclosed to clinicians.17 IL-10 levels that were toxic in swine fell into the target range of multiple published studies reporting on the therapeutic benefit of IL-10 in various animal models, as depicted in Figure S8. Therefore, the model used (syngeneic IL-10 in a large animal species), rather than the drug level, appears to have been the critical reason why the present study detected toxicity.
The notion that IL-10 could initiate toxic inflammation may appear counterintuitive because IL-10 is mainly known to be anti-inflammatory in the context of its recent development as a putative therapeutic.10 Yet other studies of the immune-modulatory effect of IL-10 in the nervous system would predict the opposite (a pro-inflammatory effect in the meninges). Koedel et al.18 found in mice that IL-10 induced meningeal pleocytosis with lymphocytic predominance after intracisternal (i.c.) delivery. This result supports the veracity of our findings in all aspects except the degree of severity of the pathology. The fact that mice in the study by Koedel et al. did not develop clinical symptoms may have been because those were difficult to diagnose in rodents or because the human homolog of IL-10 was tested (lacking full activity in mice) and given only as a single injection. In the present study, prolonged delivery of IL-10 to the meninges in a syngeneic model was able to reproduce the neuropathologic findings originally reported by Koedel et al. for IL-10, with only one critical difference, severity. A pair of elegant studies by Ding et al.19, 20 may provide a unifying explanation for why IL-10 may have been found to be anti-inflammatory by some and pro-inflammatory by others. IL-10 can downregulate its own receptor (IL-10R1), dampening or even reversing its anti-inflammatory effect at high concentrations. While the findings by Ding et al. were derived from tissue culture experiments only, our study raises the possibility that the molecular model proposed by Ding et al. may also be applicable in vivo. Regardless of the underlying mechanism, however, our study demonstrates that syngeneic IL-10 in the CSF can be highly neurotoxic.
Extensive post-mortem studies did not find evidence of an infection. Toxicity appeared to be directly related to pIL-10, rather than the vector used to deliver it, because no similar toxicity has ever been reported or observed by us after i.t. administration of AAV encoding EGFP (Figure S9). AAV vectors expressing other transgenes have been tested in dozens of large animal studies exposing the meninges and CNS in dogs, swine, cats, and non-human primates (Table S3) and in humans (Table S4). While this finding does not definitively exclude the possibility of an undetected pathogen or a modifying role of AAV vectors, neither appears likely. The experiments therefore suggest that IL-10 may be the primary driver of the observed meningitis. In summary, this report raises concerns about meninges-directed IL-10 therapy that may require further study before the initiation of clinical trials.
Materials and Methods
Animals
All animal experiments were approved by the Mayo Clinic Institutional Animal Care and Use Committee (IACUC) and were carried out in accordance with IACUC policy and the Guide for the Care and Use of Laboratory Animals.21 A total of seven domestic swine of Landrace origin (n = 5 female and n = 2 male) were administered pIL-10/AAV8 in two cohorts. The first cohort consisted of five animals (namely, Y872, Y876, Y877, Y965, and Y971). The second cohort consisted of two animals (namely, Y342 and Y343). At the time of vector administration, animals were 4 months old and weighed between 23.0 and 30.5 kg. Animals were closely observed for overall appearance, appetite, activity, changes in behavior, body temperature, weight, heart rate, and respiratory rate. The study duration was 56 days after vector administration, at which point the primary endpoint was met and animals were humanely euthanized. Animals that became moribund satisfied the humane endpoint according to institutional policy and were humanely euthanized prior to completing 56 days of study. Institutional and IACUC rules mandate euthanasia of animals that are destined to die from dehydration and subsequent systemic organ failure. Therefore, neurotoxicity is described as fatal in this article.
Definition and Collection of Vital Signs and Laboratory Parameters
Significant variation in the normal range of vital signs and laboratory parameters exists for swine due to many factors, including sex, breed, growth rate, diet, age, feeding method, housing and husbandry, season, and sampling stress.22, 23 To account for this variation, we used breed-specific parameters22 to define reference values and ensured that the same veterinary personnel to which animals were acclimatized also collected measurements and blood samples.
EGFP/AAV8 Swine Experiments
To determine the safety of AAV administered via the i.t. route in swine, a separate cohort five swine was used. These animals were housed and treated in an identical fashion to those receiving pIL-10/AAV8 except that AAV8 encoding the EGFP transgene was used with this cohort.
Experimental Design
The first cohort (n = 5) was designed for studying the safety of IL-10 administration to the CSF. Animals in this cohort were housed at the Mayo Clinic, where vector administration and CSF collection (J.P., L.F.H.), veterinary assessment and blood collection (L.K.N., J.A.S.), and euthanasia and tissue harvest (L.K.N., J.P.) were completed. Laboratory analysis and pathologic examination of moribund animals was completed at a commercial veterinary laboratory (J.L.B.).
A replication experiment (n = 2) was designed to test reproducibility of the first cohort’s results. The experimental protocol was conducted exactly as for the first cohort. A new batch of pIL-10/AAV8 vectors was tested and experiments were performed by different experimentalists. Vector was administered by an interventional, diagnostic neuroradiologist with greater than 30 years of experience (T.P.M.). Moribund animals were transported alive for final laboratory analysis, euthanasia, tissue harvest, and pathologic examination by a diagnostic veterinary pathologist specializing in swine infectious disease (J.E.C.) and a comparative veterinary neuropathologist (A.G.A.).
Vector Quality and Design
Two vector batches were used (lots V4349MI-R and CS0615-3CS; Penn Vector Core). Per request by the authors, both batches were assessed for purity and endotoxin concentration (Table S5). Self-complementary AAV encoding pIL-10 coding DNA, under control of the cytomegalovirus (CMV) promoter/enhancer, was used. Viral particles were packaged in AAV8 capsid, suspended in PBS, and administered at a dose of 7.0 × 1012 genome copies/animal.
Vector Administration
All procedures were performed under general anesthesia and with a strict aseptic technique. Anesthesia was induced by intramuscular injection of tiletamine/zolazepam, xylazine, and glycopyrrolate. Animals were then intubated and anesthesia was maintained by 1.5%–2.5% isoflurane in 50% O2 titrated to effect. The lateral recess of the lumbar i.t. space was then accessed under computed tomography (CT) fluoroscopic guidance as described previously.24 Briefly, the lumbosacral spine was visualized on a pre-procedure CT scan using a Beekley biopsy grid (Beekley Medical) to aid in needle placement and trajectory. A 25-gauge Quincke point spinal needle (Kimberly-Clark) was advanced through the interlaminar space and into the subarachnoid space lateral to the spinal cord. The i.t. position of the needle tip was verified by imaging after injection of 200 μL Omnipaque 240 contrast media (GE Healthcare) diluted 1:3 in PBS to a final concentration of 80 mg/mL. Vector was administered using a programmable syringe pump (model 10050; Chemyx) at a flow rate of 1 mL/min to a final dose of 7.0 × 1012 genome copies/animal. Following infusion, the needle was left in place for 3 min to prevent reflux during needle withdrawal. A post-injection CT scan was obtained to document delivery of the vector, indicated by persistence of contrast within the i.t. space, absence of contrast in the epidural space, and dilution of i.t. contrast by the vector suspension. All injections were completed successfully on the first attempt and each animal received a single injection.
Serial CSF Collection
CSF was collected under general anesthesia and a strict aseptic technique via lumbar puncture. Animals were positioned prone and in the reverse Trendelenburg position. The iliac crest was palpated bilaterally and the most distal lumbar intervertebral space (L4-5 or L5-6) was identified. A 25-gauge Quincke point spinal needle (Kimberly-Clark) was advanced in a ventral and slightly rostral direction into the dorsal subarachnoid space. When spontaneous CSF outflow was observed, up to 2 mL was collected and the needle was withdrawn.
Measurement of pIL-10 Concentration
The Quantikine pIL-10 immunoassay (catalog no. P1000; R&D Systems) was used to quantify pIL-10 levels in CSF. The samples were analyzed exactly according to the manufacturer’s instructions. CSF from animals that had not received the vector served as the negative control and was used to determine the detection limit of the assay (5 × 10−3 pg/mL), defined as the upper 95% confidence interval of the mean. hIL-10 levels up to 600 pg/mL (the upper limit of the pIL-10 standard curve range) were tested and no cross-reactivity was observed.
Euthanasia and Tissue Harvest
All animals were deeply anesthetized by intramuscular injection of tiletamine/zolazepam and xylazine and euthanized by intravenous injection of pentobarbital. For the first cohort (L.K.N. and J.P.), two large-bore, intra-aortic catheters (Saint Jude Medical) were inserted into the ascending aorta and thoracic aorta at the level of the diaphragm and were exsanguinated and perfused with ice-cold PBS followed by 4% paraformaldehyde. Tissue samples were collected, fixed by immersion in 4% paraformaldehyde, and underwent pathologic workup by J.L.B. For the second cohort, tissue samples were collected and fixed by immersion in 10% buffered formaldehyde and underwent pathologic workup by J.E.C. and A.G.A.
Immunohistochemistry
Formalin-fixed spinal cords were embedded in paraffin for detection of antigens using rabbit polyclonal primary antibodies and goat anti-rabbit secondary antibodies conjugated to a chromogenic immunodetection system (catalog no. K4003; Dako/Agilent). Sections were incubated at room temperature (RT) for 45 min in primary antibodies against CD3 (1:20, catalog no. CP 215 A; Biocare), CD20 (1:400, catalog no. RB-9013-P0; Thermo Scientific), Iba-1 (1:400, catalog no. CP 290 A; Biocare), and Caspase-3 (1:100, catalog no. 9661; Cell Signaling Technology). For CD3 and Iba-1 detection, heat-induced antigen retrieval was performed prior to incubation in primary antibodies in premade buffer (catalog no. DV2004 MX; Biocare). To detect GFAP, sections were incubated at RT for 30 min in primary antibodies (1:700, catalog no. Z0334; Dako/Agilent). The TUNEL assay was performed according to the manufacturer’s instructions (catalog no. S7101; Millipore).
Statistical Analysis
Statistical testing was performed to test whether the result (rate of toxicity in each cohort) was significantly different from the null hypothesis H0 = 0.01. The null hypothesis was justified by assuming that fatal toxicity is acceptable in at most 1 of 100 test subjects. This judgment was derived from the observation that a recent clinical trial of a candidate analgesic drug closed after 6 of 128 enrollees developed severe CNS toxicity and 1 died as a result.15 For the first cohort, the binomial test was performed in R software using the following command: binom.test(x = 3, n = 5, p = 0.01, alternative = ‘greater’).
The replication experiment was performed with an early stopping rule and the number of animals used in the first cohort as the maximum allowed size (n = 5). Pocock-type boundaries were calculated for four interim tests keeping H0 = 0.01 (as above) and setting the experiment-wide type 1 error rate at < 1%. Computation was performed as described.25 The design allowed for stopping when two animals suffered fatal toxicity. There was no randomization of animals and no blinding in regard to group assignment, since experiments were designed as single-cohort safety studies. All data supporting these findings are available in the article and in the Supplemental Information.
Author Contributions
J.P., T.P.M., and A.S.B. designed the study. J.P. and L.F.H. performed animal experiments for the first cohort. T.P.M., L.K.N., and J.A.S. performed animal experiments for the second cohort. J.P., L.K.N., L.F.H., J.A.S., and A.S.B. examined the animals. J.P. and L.F.H. performed the laboratory analysis of IL-10 levels. J.L.B. performed pathologic workup for the first cohort. J.E.C. and A.G.A. performed the pathologic workup for the second cohort. J.E.C. designed and oversaw the infectious disease workup for the second cohort. M.D.U., J.P., J.E.C., A.G.A., J.L.B., T.P.M., and A.S.B. analyzed the data. M.D.U. and A.S.B. wrote the manuscript. A.S.B. is responsible for the overall integrity of the study.
Conflicts of Interest
The authors declare that no conflicts of interest exist.
Acknowledgments
This study was supported by the NIH National Institute of Neurological Disorders and Stroke (R01NS063022 to A.S.B.) and by the Richard M. Schulze Family Foundation (to A.S.B.).
Footnotes
Supplemental Information includes nine figures and five tables and can be found with this article online at http://dx.doi.org/10.1016/j.ymthe.2017.07.016.
Supplemental Information
References
- 1.Hasbun R. The acute aseptic meningitis syndrome. Curr. Infect. Dis. Rep. 2000;2:345–351. doi: 10.1007/s11908-000-0014-z. [DOI] [PubMed] [Google Scholar]
- 2.Hajj-Ali R.A., Singhal A.B., Benseler S., Molloy E., Calabrese L.H. Primary angiitis of the CNS. Lancet Neurol. 2011;10:561–572. doi: 10.1016/S1474-4422(11)70081-3. [DOI] [PubMed] [Google Scholar]
- 3.Frei K., Nadal D., Pfister H.W., Fontana A. Listeria meningitis: identification of a cerebrospinal fluid inhibitor of macrophage listericidal function as interleukin 10. J. Exp. Med. 1993;178:1255–1261. doi: 10.1084/jem.178.4.1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Chiavolini D., Pozzi G., Ricci S. Animal models of Streptococcus pneumoniae disease. Clin. Microbiol. Rev. 2008;21:666–685. doi: 10.1128/CMR.00012-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Westendorp R.G.J., Langermans J.A.M., Huizinga T.W.J., Elouali A.H., Verweij C.L., Boomsma D.I., Vandenbroucke J.P. Genetic influence on cytokine production and fatal meningococcal disease. Lancet. 1997;349:170–173. doi: 10.1016/s0140-6736(96)06413-6. [DOI] [PubMed] [Google Scholar]
- 6.Sprong T., van der Ven-Jongekrijg J., Neeleman C., van der Meer J.W.M., van Deuren M. Influence of innate cytokine production capacity on clinical manifestation and severity of pediatric meningococcal disease. Crit. Care Med. 2009;37:2812–2818. doi: 10.1097/CCM.0b013e3181ab851c. [DOI] [PubMed] [Google Scholar]
- 7.Sato M., Hosoya M., Honzumi K., Watanabe M., Ninomiya N., Shigeta S., Suzuki H. Cytokine and cellular inflammatory sequence in enteroviral meningitis. Pediatrics. 2003;112:1103–1107. doi: 10.1542/peds.112.5.1103. [DOI] [PubMed] [Google Scholar]
- 8.Perdomo-Celis F., Torres M.A., Ostos H., Gutierrez-Achury J., Molano V., Durán L.F., González G., Narváez C.F. Patterns of local and systemic cytokines in bacterial meningitis and its relation with severity and long-term sequelae. Biomark. Insights. 2015;10:125–131. doi: 10.4137/BMI.S35005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Scriven J.E., Graham L.M., Schutz C., Scriba T.J., Wilkinson K.A., Wilkinson R.J., Boulware D.R., Urban B.C., Lalloo D.G., Meintjes G. A glucuronoxylomannan-associated immune signature, characterized by monocyte deactivation and an increased interleukin 10 level, is a predictor of death in cryptococcal meningitis. J. Infect. Dis. 2016;213:1725–1734. doi: 10.1093/infdis/jiw007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kwilasz A.J., Grace P.M., Serbedzija P., Maier S.F., Watkins L.R. The therapeutic potential of interleukin-10 in neuroimmune diseases. Neuropharmacology. 2015;96:55–69. doi: 10.1016/j.neuropharm.2014.10.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Storek B., Reinhardt M., Wang C., Janssen W.G., Harder N.M., Banck M.S., Morrison J.H., Beutler A.S. Sensory neuron targeting by self-complementary AAV8 via lumbar puncture for chronic pain. Proc. Natl. Acad. Sci. USA. 2008;105:1055–1060. doi: 10.1073/pnas.0708003105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Pleticha J., Malkmus S.A., Heilmann L.F., Veesart S.L., Rezek R., Xu Q., Yaksh T.L., Beutler A.S. High cerebrospinal fluid levels of interleukin-10 attained by AAV in dogs. Gene Ther. 2015;22:202–208. doi: 10.1038/gt.2014.96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Altschul S.F., Madden T.L., Schäffer A.A., Zhang J., Zhang Z., Miller W., Lipman D.J. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 1997;25:3389–3402. doi: 10.1093/nar/25.17.3389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Moore K.W., de Waal Malefyt R., Coffman R.L., O’Garra A. Interleukin-10 and the interleukin-10 receptor. Annu. Rev. Immunol. 2001;19:683–765. doi: 10.1146/annurev.immunol.19.1.683. [DOI] [PubMed] [Google Scholar]
- 15.Butler D., Callaway E. Scientists in the dark after French clinical trial proves fatal. Nature. 2016;529:263–264. doi: 10.1038/nature.2016.19189. [DOI] [PubMed] [Google Scholar]
- 16.Kerbrat A., Ferré J.-C., Fillatre P., Ronzière T., Vannier S., Carsin-Nicol B., Lavoué S., Vérin M., Gauvrit J.Y., Le Tulzo Y., Edan G. Acute neurologic disorder from an inhibitor of fatty acid amide hydrolase. N. Engl. J. Med. 2016;375:1717–1725. doi: 10.1056/NEJMoa1604221. [DOI] [PubMed] [Google Scholar]
- 17.Jouan A., Mascret D. Le Figaro; 2016. Biotrial: de nouvelles révélations troublantes dans l’essai clinique mortel de Rennes. [Google Scholar]
- 18.Koedel U., Bernatowicz A., Frei K., Fontana A., Pfister H.W. Systemically (but not intrathecally) administered IL-10 attenuates pathophysiologic alterations in experimental pneumococcal meningitis. J. Immunol. 1996;157:5185–5191. [PubMed] [Google Scholar]
- 19.Ding Y., Qin L., Zamarin D., Kotenko S.V., Pestka S., Moore K.W., Bromberg J.S. Differential IL-10R1 expression plays a critical role in IL-10-mediated immune regulation. J. Immunol. 2001;167:6884–6892. doi: 10.4049/jimmunol.167.12.6884. [DOI] [PubMed] [Google Scholar]
- 20.Ding Y., Chen D., Tarcsafalvi A., Su R., Qin L., Bromberg J.S. Suppressor of cytokine signaling 1 inhibits IL-10-mediated immune responses. J. Immunol. 2003;170:1383–1391. doi: 10.4049/jimmunol.170.3.1383. [DOI] [PubMed] [Google Scholar]
- 21.National Research Council . Eighth Edition. National Academies Press; 2011. Guide for the Care and Use of Laboratory Animals. [Google Scholar]
- 22.Bollen P.J.A., Hansen A.K., Olsen Alstrup A.K. Second Edition. CRC Press/Taylor & Francis; 2010. The Laboratory Swine. [Google Scholar]
- 23.Weiss D.J., Wardrop K.J. Blackwell Publishing; 2010. Schalm’s Veterinary Hematology. [Google Scholar]
- 24.Pleticha J., Maus T.P., Jeng-Singh C., Marsh M.P., Al-Saiegh F., Christner J.A., Lee K.H., Beutler A.S. Pig lumbar spine anatomy and imaging-guided lateral lumbar puncture: a new large animal model for intrathecal drug delivery. J. Neurosci. Methods. 2013;216:10–15. doi: 10.1016/j.jneumeth.2013.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ivanova A., Qaqish B.F., Schell M.J. Continuous toxicity monitoring in phase II trials in oncology. Biometrics. 2005;61:540–545. doi: 10.1111/j.1541-0420.2005.00311.x. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.