Contact-Dependent Demyelination by Mycobacterium leprae in the Absence of Immune Cells

See allHide authors and affiliations

Science  03 May 2002:
Vol. 296, Issue 5569, pp. 927-931
DOI: 10.1126/science.1067631


Demyelination results in severe disability in many neurodegenerative diseases and nervous system infections, and it is typically mediated by inflammatory responses. Mycobacterium leprae, the causative organism of leprosy, induced rapid demyelination by a contact-dependent mechanism in the absence of immune cells in an in vitro nerve tissue culture model and inRag1-knockout (Rag1 −/−) mice, which lack mature B and T lymphocytes. Myelinated Schwann cells were resistant to M. leprae invasion but undergo demyelination upon bacterial attachment, whereas nonmyelinated Schwann cells harbor intracellular M. leprae in large numbers. During M. leprae–induced demyelination, Schwann cells proliferate significantly both in vitro and in vivo and generate a more nonmyelinated phenotype, thereby securing the intracellular niche forM. leprae.

Demyelination is one of the central pathologic conditions that ultimately lead to prolonged neurologic disability in many neurodegenerative diseases (1). The process of demyelination involves multiple factors (2). Although inflammatory responses seem to be needed for the complete manifestation of pathologic conditions of demyelination and associated neurological symptoms (1, 3), virtually nothing is known about the mechanisms involved in early events of such neurological injury. One of the classic examples of infectious neurodegenerative diseases of the peripheral nervous system (PNS) is leprosy, which is caused by the obligate intracellular bacterium Mycobacterium leprae (4) and is a leading cause of nontraumatic neuropathies in the world (3). Demyelination is a common pathologic feature in the nerve damage in leprosy (5–8) and is likely to substantially contribute to the neurologic disability in these patients. The nerve damage in leprosy is widely thought to be secondary to the     immune response against M. lepraeitself (3, 4, 8). Because Schwann cells in the PNS provide a privileged site for M. leprae that eludes surveillance by the host immune cells, the early induction of nerve damage may not be associated with immune responses (9). To test this hypothesis and to gain insight into the functional consequences of early M. lepraeinteraction with peripheral nerves, we used a well-characterized myelinating Schwann cell–dorsal root ganglion (DRG) neuron coculture system (10) and mice with null mutation for the recombination-activating gene 1 (Rag1 −/−), which lack mature T and B lymphocytes (11), as in vitro and in vivo models, respectively.

In Schwann cell–DRG neuron cocultures, M. leprae(12) avidly bound to both myelinating and nonmyelinating Schwann cell–axon units after 1 hour of inoculation [Fig. 1B, inset (13)]. We studied the functional consequences of this binding (14) on individual Schwann cell phenotypes. In myelinating Schwann cells,M. leprae induces significant demyelination 24 hours after bacterial attachment in a dose-dependent manner, whereas in nonmyelinating Schwann cells, bacterial invasion in large numbers is the most conspicuous feature. The induction of demyelination byM. leprae was rapid and was observed as early as 24 hours after infection (Fig. 1, B and D) without any signs of apoptosis (15) or cytopathic effects to Schwann cells (16), suggesting that M. leprae–induced demyelination does not result from apoptosis or toxic effects. Demyelination was found over long lengths of the nerve or in short sporadic segments as compared with cultures treated with phosphate-buffered saline (PBS) alone (Fig. 1, A and B). Electron microscopic analysis further showed disrupted myelin sheaths and the separation of lamellae with wide spaces as compared with intact myelin sheaths in controls (Fig. 2B). The most distinctive feature is the formation of myelin ovoids (Fig. 1B), which are characteristics of myelin pathology (17, 18). Because the axons remain intact in most of these demyelinating fibers at early time points (Fig. 2A), this condition resembles classic segmental demyelination (17, 18), which is also common in leprosy patients (3,5–7). Experiments performed in parallel with Mycobacterium smegmatis, a related species that does not cause neuropathy, showed no significant demyelination (<20%) (16).

Figure 1

Attachment of M. leprae and its cell-wall components to Schwann cell–axon units in vitro induces striking demyelination. (A) Control immunofluorescence micrograph showing a Schwann cell–DRG neuron coculture that had myelinated for 3 weeks and was labeled with monoclonal antibody (mAb) against myelin basic protein (MBP), which represents the compact myelin. Note the intact myelin segments, one of which is demarcated by arrows that denote the nodes of Ranvier. (B) M. leprae induce significant demyelination upon attachment. Cocultures were first incubated for 1 hour with M. leprae to attach to Schwann cell–axon units and then further incubated for another 24 hours, fixed, and labeled with MBP-specific antibody. Inset shows the attached M. leprae (green notes) to nerve fibers in similar cultures as visualized by acid-fast labeling (13). Note the demyelinating nerve fibers (arrowheads) with characteristic myelin ovoids (arrows) as compared with intact myelin segments in controls. (C) PGL-1 of the M. lepraecell wall is sufficient to induce significant demyelination. Myelinated cocultures were treated with purified native PGL-1 as in (B). After 24 hours, PGL-1 causes significant demyelination, which is similar to that of whole M. leprae (B). Inset shows the bound PGL-1 (green) to a myelinated nerve fiber after 1 hour of incubation and detected by anti-PGL-1 mAb. Demyelinated fibers indicated by the arrowheads and the myelin ovoids are shown by the arrows. Also shown are myelin debris labeled with MBP-specific antibody as red dots. (D) Quantification of in vitro demyelination induced byM. leprae (ML), the total cell wall (CW), and purified native PGL-1 (PGL-1) as compared with controls (PBS). Fibers were classified as demyelinated if the myelin sheath was crenated or fragmented and the percentage of demyelination was quantified from two to three separate experiments.

Figure 2

Demyelination is an early event inM. leprae infection that leads to axonal damage. (A) Demyelination is an initial consequence of in vitroM. leprae infection. Schwann cell–DRG neuron cocultures that had myelinated for 3 weeks were treated with M. lepraefor 12 hours, fixed, and double-labeled with MBP-specific polyclonal antibody (red) and an mAb against neurofilament mAb (green). A demyelinated nerve fiber with a damaged myelin sheath (top) and its associated intact axon in the merged image (bottom; MBP is in red/orange and neurofilament is in green) are shown by the arrows. The arrowheads denote the Schwann cell nucleus of the demyelinated fiber. Also note the numerous myelin debris (red/orange particles), which are indicated by the small arrows. (B) Axonal degeneration is a late event of M. leprae infection. Electron micrographs showing the ultrastructure of myelinated Schwann cell (SC)–axon (AX) units with surrounding basal lamina (BL) of cocultures that are untreated (left) and treated with M. leprae for 72 hours (right). Note the splitting of the myelin lamellae (arrows), the disorganized myelin sheath (MS), and the displaced degenerating axon in M. leprae–infected cultures as compared with controls.

Viability of M. leprae is not required for its ability to induce demyelination in vitro, because γ-irradiated M. leprae and the M. leprae cell-wall fraction alone (19) were as effective as live bacteria (Fig. 1D). This suggests that the components of the M. leprae cell wall have the capacity to induce demyelination during early infection. Prominent among the biologically active components of the M. lepraecell wall is M. leprae–specific phenolic glycolipid-1 (PGL-1) (19–22), which has recently been implicated for the neural predilection of M. leprae(21). Although PGL-1 represents >2% of the bacterial mass, little is known about its biological functions (20). Here we show that the binding of native PGL-1 (19) to myelinating nerve fibers [Fig. 1C, inset; (21)] results in demyelination in a dose- and time-dependent manner (Fig. 1, C and D). Because the pattern of PGL-1–induced demyelination resembles that of whole M. leprae (Fig. 1, B and C), we propose that the PGL-1 is a crucial molecule in the M. leprae cell wall that is directly involved in the neuropathogenesis of leprosy. We also excluded the possibility that demyelination, induced by the M. lepraecell-wall or PGL-1 preparations, was caused by contamination with lipopolysaccharide or other endotoxins (14).

A time-course analysis showed that demyelination is an initial event that is followed by axonal damage during M. lepraeinfection. Although axons remain intact in early infection (Fig. 2A),M. leprae–treated cultures began to exhibit substantial axonal degeneration in addition to the myelin damage after 72 hours (Fig. 2B). However, no bacteria were found in axons (16). Because our in vitro culture model is completely devoid of immune cells, but only comprised of Schwann cells and neurons, we conclude that such in vitro induction of demyelination and axonal degeneration by M. leprae is mediated by a nonimmune mechanism(s).

To test this in vivo, we used Rag1 −/− mice, which lack mature T and B lymphocytes and thus are incapable of mounting adaptive immune responses against pathogens (11). Although null mutation of the Rag1 gene causes severe deficiency in the immune system, the nervous system of these mice is intact (11). In the PNS, sciatic nerves ofRag1 −/− mice show normal, thick and compact myelin as in wild type [Fig. 3, A and C; (23)]. Intraneural administration of either viable or deadM. leprae or the whole cell-wall fraction to sciatic nerves of Rag1 −/− mice (24) produced significant demyelination (P < 0.0005) in varying degrees as compared with that in the mice injected with PBS (Fig. 3, A to D). Myelin breakdown in various stages was seen in all the fascicles of M. leprae–infected nerves with widely separated myelin lamellae (Fig. 3, B and D). Semithin sections of sciatic nerves and the corresponding electron microscopic analysis revealed that demyelination in infected nerves was time-dependent and quite dramatic after 72 hours (Fig. 3, B and D). Labeling of teased fibers prepared from infected and control sciatic nerves with antibody to myelin-specific P0 protein further confirmed demyelinated fibers and characteristic myelin ovoids in M. leprae–infected nerves (Fig. 3, E and F). As in Rag1 −/− mice,M. leprae injection to sciatic nerves of wild-type C57BL/6 (B6) mice also caused considerable demyelination as early as 24 hours (16). Because inflammatory cells are unlikely to be recruited at such an early stage (24 hours), we suggest that demyelination in wild-type mice in early infection is also caused by a nonimmune mechanism(s). Our finding that demyelination can be induced by dead M. leprae or bacterial cell wall alone both in vitro and in vivo may provide new insights into the underlying causes for continuing neurological injuries in patients who have been cured of leprosy (bacteriological cure) after multidrug therapy (25). It is likely that such nerve damage may be associated with deadM. leprae or residual cell-wall components, such as PGL-1, retained in nerve tissues in treated leprosy patients (26).

Figure 3

In vivo induction of demyelination in early M. leprae infection does not require inflammatory and/or immune response. (A and B) Methylene blue–stained semithin transverse sections of sciatic nerves ofRag1 −/− knockout mice 72 hours after intraneural injection of PBS (A), and viable M. leprae (B). Note the degenerated myelinated fibers in M. leprae–treated nerves. (C and D) Electron microscopy analysis of corresponding sections as in (A) reveals normal compact myelinated fibers with intact axons in the PBS-injected nerves (C), whereas the nerves injected with viable M. leprae (D) show varying degrees of demyelination and axonal damage. The arrows indicate the fibers with extensive demyelination. Also note the collagen deposition and increased endoneural space in infected nerves. (E and F) Demyelination in teased sciatic nerve fibers. Teased fiber preparation from sciatic nerves injected with PBS (E) and M. leprae (F) were labeled with P0-specific antibody. Note the demyelination and formation of myelin ovoids (arrows) as confirmed by antibody to myelin-specific P0 protein. (G) Quantification of in vivo demyelination revealed a significant difference (P < 0.0005) in the number of demyelinated fibers in the sciatic nerves infected with viable M. leprae (ML) and the cell-wall (CW) fraction as compared with those treated with PBS.

In demyelinating diseases, inflammation and/or immune responses are thought to be the cause for the myelin damage, in which macrophages, T lymphocytes, and antibodies play the crucial roles (1,2, 4). However, much less is known about alternative mechanisms for myelin damage. In M. leprae–induced demyelination, we found that there was no evidence of macrophage infiltration, as the number of CD68-positive macrophages was rare in the endoneurium of the sciatic nerves of infectedRag1 −/− mice (16). BecauseRag1 −/− mice are also devoid of functional T and B cells (11), we conclude that the demyelination induced by M. leprae in vivo is not the result of antibody-mediated phenomena or macrophage- or T cell–mediated inflammatory processes. Because these pathological conditions induced by M. lepraein both in vivo and in vitro models resemble neuropathies in human leprosy (3, 7), they may represent an early phase of nerve damage in patients with other neurodegenerative diseases. One possible mechanism for such non-immune-mediated demyelination and/or nerve injury would be the perturbation of homeostasis of the neural microenvironment and signaling network byM. leprae infection.

An important consequence of M. leprae attachment to the Schwann cell–axon units is bacterial invasion. Molecular dissection of M. leprae attachment and early invasion in animal models has been difficult, however, mainly owing to the failure of localization of M. leprae within Schwann cells for as yet unknown reasons (27). Although this was also the case in Rag1 −/− mice described in the present study (16), we could localize significant numbers ofM. leprae in close proximity to myelinated and nonmyelinated Schwann cells within the endoneurium, and some bacteria were found attached to myelinated Schwann cells (Fig. 4C). In contrast, the nerve culture model was found to be suitable for M. leprae invasion studies, because this in vitro model mimics the preferential localization ofM. leprae to nonmyelinated Schwann cells as in patients with lepromatous leprosy (6, 7) (described below). We have therefore focused our efforts on this in vitro model ofM. leprae invasion.

Figure 4

Distinct functional consequences of nonmyelinated and myelinated Schwann cells in response to M. leprae infection in vitro and in vivo; bacterial invasion, demyelination, and Schwann cell proliferation. (A and B) Nonmyelinated Schwann cells (NMSC) are preferentially invaded by M. leprae, whereas myelinated Schwann cells (MSC) are resistant to bacterial invasion but undergo demyelination (see Fig. 1). (A) Quantification of nonmyelinated and myelinated Schwann cells infected with M. leprae in vitro. (B) Intracellular M. leprae(arrows) were found exclusively in nonmyelinated Schwann cells in high numbers. Axons in nonmyelinated Schwann cells are normal (arrowheads) in spite of the numerous bacteria. (C) Representative electron micrograph of sciatic nerve ofRag1 −/− mice 72 hours after infection showing extracellular M. leprae (arrows) either attached to or close to myelinated Schwann cells, one of which is shown with extensive demyelination (double arrowheads). (D to F) In vitro and in vivo proliferation of Schwann cells in M. leprae–infected Schwann cell–neuron cocultures and sciatic nerves of Rag1 −/− mice that are undergoing demyelination. Quantification of 5-bromo-2′-deoxyuridine (BrdU)-positive cells in cocultures (D) and in teased fiber preparations from sciatic nerves of Rag1 −/−mice (F) reveals that M. leprae (+ML) induce a statistically significant increase in Schwann cell proliferation (P< 0.05 and P < 0.005 in vitro and in vivo, respectively). (E) Representative confocal micrograph showing demyelinated fibers associated with Schwann cell proliferation in vitro as detected by double-labeling with BrdU-specific antibody (green) and MBP-specific mAb (red). The arrows denote the demyelinated fibers (red) with proliferating Schwann cell nuclei (green).

Analyzing the events of M. leprae invasion in myelinated versus nonmyelinated phenotypes in in vitro cultures (28) showed striking differences. The most consistent finding in myelinated Schwann cells is their resistance to M. leprae invasion even after 72 hours (Fig. 4A), although at that time point M. leprae induces significant demyelination and axonal degeneration in this phenotype (Fig. 2B). This clearly indicates that M. leprae invasion is not necessary to cause demyelination, but the bacterial attachment alone is sufficient to induce this nerve pathology via a contact-dependent mechanism(s). This also seems to be the case inRag1 −/− mice infected with M. leprae, because some bacteria were found attached to or in the vicinity of the myelinated Schwann cells that are undergoing demyelination in vivo (Fig. 4C).

In contrast to myelinated Schwann cells, nonmyelinated Schwann cells are highly susceptible to M. leprae invasion in vitro, with intracellular bacteria found in the majority of nonmyelinated cells (>75%) in varying degrees (that is, 1 to 30 bacteria per cell;Fig. 4, A and B). This is also the case in human leprosy, where intracellular M. leprae are frequently seen in nonmyelinated Schwann cells (70 to 80%) in the early stage, but bacilli are rarely seen in myelinated Schwann cells (∼4%) even in advanced lepromatous patients with high bacterial load (6, 7). This susceptibility of nonmyelinated Schwann cells to M. leprae invasion may provide this bacterium with a clear advantage for its intracellular survival within the PNS. M. leprae–induced demyelination and axonal damage in vitro and in vivo resemble peripheral nerve injury such as in Wallerian degeneration, where Schwann cells rapidly proliferate to promote the regeneration of injured nerves (29). Strikingly, Schwann cells in M. leprae–infected cocultures and sciatic nerves of Rag1 −/− mice proliferate (30) significantly (P < 0.05 and P < 0.005, respectively) after 72 hours (Fig. 4, D to F), at which point myelinated fibers both in vitro and in vivo have undergone substantial demyelination and axonal degeneration (Figs. 2B and 3). On the basis of these findings, we propose that M. leprae–induced nerve injury contributes to Schwann cell proliferation, as in other nerve injury responses (29), and, thereby, increases the number of nonmyelinating Schwann cells in the early phase of the infectious process. Because M. leprae is an obligate intracellular bacterium, it must invade its preferred nonmyelinating phenotypes in order to survive within the PNS. However, as the infection progresses and bacteria undergo unrestrained multiplication, the availability of nonmyelinating Schwann cells becomes a limiting factor. To avoid such a situation, leprosy bacilli may induce demyelination and axonal damage as an effective strategy to increase the number of nonmyelinating Schwann cells so that a sufficient intracellular niche is available for bacterial survival. Because myelinated Schwann cells do not serve as an intracellular niche for M. leprae, we propose that M. lepraepropagates a nonmyelinating phenotype by inducing demyelination and nerve injury in myelinated Schwann cells in the early phase of infection, a novel bacterial survival strategy in the nervous system.

  • * To whom correspondence should be addressed. E-mail: rambuka{at}


View Abstract

Stay Connected to Science

Navigate This Article