delayed clinical and pathological signs in twitcher (globoid cell leukodystrophy) mice on a c57bl/6...

Neurobiology of Disease 10, 344–357 (2002)

doi:10.1006/nbdi.2002.0527

Delayed Clinical and Pathological Signs in Twitcher(Globoid Cell Leukodystrophy) Miceon a C57BL/6 � CAST/Ei Background

Sangita Biswas,* Homigol Biesiada,† Todd D. Williams,† andSteven M. LeVine**Department of Molecular and Integrative Physiology, University of Kansas Medical Center,3901 Rainbow Boulevard, Kansas City, Kansas 66160; and †Mass SpectrometryLaboratory, University of Kansas Lawrence, Kansas 66045

Received February 13, 2002; revised May 7, 2002; accepted for publication June 11, 2002

Modifier genes may account for the phenotypic variability observed in the late-onset forms of globoidcell leukodystrophy (GCL) in humans. In order to begin a search for modifier genes, the effect ofgenetic background on the clinical and pathological manifestations of GCL was investigated intwitcher mice. Twitcher mice on a C57BL/6 � CAST/Ei background had an increased life span (61.4 �2.5 vs 37.0 � 0.6 days), a delayed onset of tremor (24 vs 21 days), and a delayed decline in walkingability compared to C57BL/6 twitcher mice. Pathologically, C57BL/6 � CAST/Ei twitcher mice hadfewer lectin-positive globoid cells, less gliosis, and a greater preservation of myelin compared toC57BL/6 twitcher mice under moribund conditions. Similar concentrations of psychosine, the toxicspecies that accumulates in GCL, were measured by tandem mass spectrometry between moribundC57BL/6 twitcher mice (286.5 pmol/mg protein), 40-day C57BL/6 � CAST/Ei twitcher mice (276.5pmol/mg), and moribund C57BL/6 � CAST/Ei twitcher mice (247.0 pmol/mg), suggesting that themilder phenotype in CAST/Ei � C57BL/6 twitcher mice did not correlate with less psychosine. Insummary, the introduction of modifier genes from the wild, inbred CAST/Ei strain had a phenotypiceffect resulting in a significantly slower disease course. © 2002 Elsevier Science (USA)

INTRODUCTION

Globoid cell leukodystrophy (GCL) is an autosomalrecessive disorder caused by mutations in the geneencoding for the lysosomal enzyme galactosylcerami-dase. This enzyme catalyzes the first step in the deg-radation of galactosylceramide, a major glycolipid ofcentral and peripheral myelin, and psychosine, a sec-ondary product of UDP-galactosylceramide galacto-syltransferase, which also synthesizes galactosylcer-amide. A deficiency of galactosylceramidase leads tothe progressive accumulation of psychosine in thebrain and kidney (Miyataki & Suzuki 1972; Svenner-holm et al., 1980; Igisu & Suzuki 1984). Psychosine istoxic to a variety of cells (Igisu, 1989; Tapasi et al.,1998) and it has been postulated to cause early patho-logical changes in oligodendrocytes and Schwann

344

cells (Tanaka & Webster 1993; Cho et al., 1997) thateventually lead to extensive demyelination.

The onset of clinical signs in patients with GCL isvariable. The infantile form has an onset within thefirst 6 months of life, the late infantile form presentsfrom 6 months to 3 years, the juvenile form appearsfrom 3 to 10 years, and the adult variant has symp-toms beginning after 10 years (Kolodny et al., 1991;Wenger et al., 2001). In the infantile form of diseasethere is �5% of the normal levels of galactosylcerami-dase (Luzi et al., 2001), while the residual enzymeactivity is slightly higher in the later onset forms ofdisease (Krivit et al., 1995; De Gasperi et al., 1996, 1999;Jardim et al., 1999). Expression studies have identifiedalleles that result in a lowering of enzymatic activitybelow median normal levels (Furuya et al., 1997). Forexample, a substitution of cysteine by histidine at a.a.

0969-9961/02 $35.00© 2002 Elsevier Science (USA)

All rights reserved.

168 in the human galactosylceramidase gene resultedin a 20% reduction in galactosylceramidase activityand an 80–90% reduction in activity in mice (Luzi etal., 2001). Mice that had this substitution had a diseasethat was milder to that observed in twitcher mice,which have a naturally occurring mutation in theirgalactosylceramidase gene that results in GCL in mice.This substitution resulted in delayed clinical signs, aslower accumulation of psychosine, and a milder pa-thology than that observed in regular twitcher mice.This substitution, however, does not cause GCL inhumans (Luzi et al., 2001). Other studies have shownthat the level of residual activity does not predict theclinical course in the late-onset patient (Fu et al., 1999;Wenger et al., 2000). Furthermore, some individualswithout disease have a pseudo-deficiency in enzy-matic activity that falls in the range observed for pa-tients with late-onset disease (Luzi et al., 2001). Othermutations have been examined to determine whetherthey could explain the phenotypic variability in thelater onset forms of disease, but thus far, there hasbeen no clear correlation between genotype status andthe onset or progression of the late-onset forms ofdisease (Bernardini et al., 1997; Wenger, 1997; De Gas-peri et al., 1999; Fu et al., 1999). This lack of correlationbetween disease progression and enzyme levels ormutation status suggests that environmental and/orother genetic factors influence the course of the late-onset disease.

Examples of environmental factors include the pos-sibility that a blow to the head (Fluharty et al., 1986;Fiumara et al., 1990) or viral infections, like chickenpox or mumps (Thulin et al., 1968; Crome et al., 1973),may trigger the onset of symptoms in humans withGCL, perhaps via the induction of proinflammatorycytokines. This concept is supported by the findingthat lipopolysaccharide (LPS), which is known to in-duce proinflammatory cytokines, accelerates diseaseprogression in twitcher mice (Pedchenko & LeVine,1999; Pedchenko et al., 2000).

An example of a genetic factor is the presence ofmodifier genes that interact with the disease-causing gene or influence the disease process and,thus, impact the disease presentation. A recentstudy indicates that saposin A deficiency can lead toa slowly progressing form of GCL in mice. Thisglycoprotein is a galactosylceramidase activatorprotein and it has been shown to be required foroptimal galactosylceramide degradation by galacto-sylceramidase (Matsuda et al., 2001a,b). Since thediagnosis of GCL is largely based on the assay ofgalactosylceramidase, it is possible that some pa-

tients with a late-onset, chronic leukodystrophy,which appears clinically and pathologically similarto GCL, will have mutations in their saposin A genebut greater than residual levels of galactosylcerami-dase activity as detected by an in vitro assay (Ma-tsuda et al., 2001b). There are other examples ofgenetic influences on the course of GCL in twitchermice. In twitcher mice that were also deficient in theMHC class II locus, clinical and pathological signsof disease were milder than for regular twitchermice (Matsushima et al., 1994), and twitcher micethat were deficient in IL-6 had a more robust pa-thology compared to regular twitcher mice (Ped-chenko & LeVine 1999).

The strain background of the twitcher mouse alsoappears to influence the expression of clinical signs inthis disease. For example, the original description oftwitcher mice by Duchen et al. (1980) described micethat resulted from the mating of CE/J and C57BL/6Jstrains, whereas subsequent studies utilized animalswith the C57BL/6J background. Twitcher mice on themixed background may have had a slower diseasecourse than mice solely on the C57BL/6J background.

In order to begin a search for possible modifiergenes, we generated twitcher mice that were on amixed background for C57BL/6 and castaneus (Musmusculus castaneus; CAST/Ei) strains. The CAST/Eistrain was selected for several reasons: (1) CAST/Eimice have low levels of plasma phospholipids, a lowpercentage of HDL lipoprotein particles, low levels ofphospholipids transfer protein activity, and low levelsof plasma HDL cholesterol compared to C57BL/6mice (Albers et al., 1999; Mehrabian et al., 2000). Sincemyelin has a relatively high concentration of phospho-lipids and cholesterol, and myelin is the primary tar-get of pathology in GCL, the differences in phospho-lipids and cholesterol turnover could influence thedisease course; (2) CAST/Ei mice have been used forlinkage analysis studies with common laboratorymouse strains utilizing PCR amplification of simplesequence repeat variants (Taylor et al., 1994). Thismethodology has been predicted to be able to screen94% of the autosomal genome between CAST/Ei andcommon laboratory strains of mice; (3) CAST/Ei miceare considered good breeders.

In this study we have observed substantial effects ofa mixed background of C57BL/6 � CAST/Ei on thedisease course in twitcher mice compared to twitchermice on a C57BL/6 background. These findings couldlead to the identification of additional modifier genesthat influence the course of GCL.

345Twitcher Disease in C57BL/6 � CAST/Ei Mice

© 2002 Elsevier Science (USA)All rights reserved.

MATERIAL AND METHODS

Mice

A female mouse heterozygous for the twitcher mu-tation (C57BL/6J,twi/�) (Jackson Laboratories, BarHarbor, ME) was crossed with a male CAST/Ei mouse(Jackson Laboratories) which is an inbred strains de-rived from a wild population of the subspecies M.musculus castaneus. Offspring in the F1 and F2 gener-ations were genotyped, and heterozygotes for thetwitcher mutation were used for breeding.

The genetic status for the twitcher mutation wasdetermined by diagnostic PCR on genomic DNA ex-tracted from clipped tails from 5- to 6-day-old pupsaccording to the method of Sakai et al. (1996) exceptthat a touchdown PCR protocol was utilized.

Animals were maintained under standard housingconditions in a pathogen-free environment with freeaccess to food and water. Animal handling protocolswere in accordance with institutional guidelines foranimal care and use.

Life Span

In order to determine the survival time, eachtwitcher mouse was allowed to live as long as it couldbe maintained humanely according to the norms oflaboratory animal care ethics without forced feedingor drinking. When a mouse reached a moribund con-dition, characterized by the animal’s incapability ofvoluntary movement and severe wasting, it was eu-thanized by halothane inhalation and the brain wasremoved and immersion fixed in buffered formalin forpathological analyses or frozen and stored at �80°Cfor biochemical analyses.

Clinical Evaluation

In order to follow the course of disease, the onsetday of tremor was noted and walking ability wasscored at every 10th day starting from day 30 afterbirth. Walking ability was scored as follows: 0, normalwalking; 1, wobbly gait but able to maintain uprightposture; 2, weakness in hind limbs characterized bydrooping of hindquarters while walking; 3, strugglingto walk with hind limbs with occasional dragging ofhind limbs; 4, dragging of hind limbs with propulsionentirely by forelimbs.

Neuropathology

Paraffin sections, 8 �m thick, of the hind brain wereprocessed for Griffonia simplicifolia lectin I, myelin ba-sic protein (MBP), and glial fibrillary acidic protein(GFAP) staining as previously described (Pedchenko& LeVine, 1999). G. simplicifolia lectin I has been uti-lized to label globoid cells and infiltrating macro-phages in the CNS of twitcher mice (Taniike & Suzuki1994), but it fails to stain resting microglia. MBP im-munohistochemistry, used to detect demyelination(Matsushima et al., 1994), and GFAP staining, used tomeasure the extent of astrocyte gliosis, were evaluatedin the hindbrain.

Axonal pathology in the cerebellum was studied viaparvalbumin immunohistochemistry using a VectorM.O.M. Immunodetection kit (Vector Laboratories).This procedure was performed as follows: Paraffinsections, 8 �m thick, of the hind brain were deparaf-finized, hydrated through a graded alcohol series;PBS, 5 min; 3% H2O2/PBS, 5 min; PBS, 2 � 5 min;mouse M.O.M. mouse Ig blocking reagent, 1 h; PBS,2 � 5 min; M.O.M. diluent, 5 min; anti-parvalbuminantibody (1:3000 in M.O.M. diluent, Sigma ChemicalCo., St. Louis, MO), 30 min; PBS, 2 � 5 min; M.O.M.biotinylated anti-mouse IgG Reagent, 10 min; PBS 2 �5 min; Vectastain ABC reagent, 10 min; PBS, 5 min;and 0.05% DAB/0.01% H2O2/10 mM Tris buffer, pH7.6, 7–8 min. Sections were then washed in PBS, de-hydrated, cleared, coverslipped, and studied underthe light microscope.

Sciatic Nerve

Sciatic nerve samples were dissected from approxi-mately the same position from the hind legs of controland twitcher mice. The nerves were immersion fixedin buffered formalin and embedded in a plastic resin(see Electron Microscopy, below). Sections, 1 �mthick, were prepared and stained with 1% toluidineblue.

Electron Microscopy

Hind brains were fixed in 2% glutaraldehyde/1%formaldehyde in 0.1 M phosphate buffer for 48 h.Tissues were then rinsed with 0.01 M cacodylatebuffer and postfixed in 1% osmium tetroxide bufferedin 0.01 M cacodylate for 1 h. Tissues were rinsed indistilled water, dehydrated in graded alcohol, incu-bated with propylene oxide for 20 min, and thenplaced in a half propylene oxide, half Embed 812 resin

346 Biswas et al.


(Electronmicroscopy Science, Inc., Fort Washington,PA) at 70°C overnight. Eighty-nanometer-thick sec-tions were cut on a Leica UCT ultramicrotome, pickedup on 300-mesh thin-bar grid, and stained with 7%uranyl acetate and lead citrate for contrast. Pictureswere taken using a JEOL 100CXII transmission elec-tron microscope.

Psychosine Quantification

Extraction of psychosine was performed following amodification of the method of Sullards & Merrill(2001). Briefly, 40–60 mg of tissue from pons/medullawas homogenized in 300 �l of PBS in an all-glass,hand-held homogenizer and transferred to 1.5-mlglass HPLC autosampler vials. The following wereadded to each homogenate in sequential order: 0.5 mlof methanol, 0.25 ml of chloroform, and 0.2 ml oflactosylsphingosine (2 pmol/�l), which is the internalstandard (Whitefield et al., 2001). Samples were soni-cated for 60 s at room temperature and incubatedovernight at 48°C in a water bath. After being cooled,113 �l of 1 M KOH was added and samples weresonicated at room temperature for 60 s followed byincubation at 37°C for 2 h in a water bath. After beingcooled, 6.1 �l glacial acetic acid was added and mixedby vortexing for 20 s. Samples were dried under re-duced pressure in a Savant Speed Vac and reconsti-tuted in 400 �l of 2:1 methanol:chloroform. The tissuedebris was precipitated by centrifugation and the clearsupernatant was transferred to a fresh autosamplervial and stored at �20°C until analysis. All chemicalswere added or transferred with glass Hamilton sy-ringes since polypropylene pipette tips adsorb sphin-golipids. Protein estimations in the pons/medullawere performed following the modified Lowrymethod (Sigma).

The calibration curve for psychosine was con-structed by adding increasing concentrations of psy-chosine standard (0–8 pmol/�l final concentration)and a constant concentration of lactosylsphingosine (1pmol/�l final concentration) to homogenates of pons/medulla from wild C57BL/6 mice and processing itthrough the entire extraction procedure. The ratio ofthe response of psychosine to the response of lactosyl-sphingosine (internal standard) was plotted againstthe concentration of psychosine. The calibration curvewas corrected for endogenous psychosine in the con-trol brain matrix. All standard concentrated stock so-lutions were prepared in 2:1 methanol:choloform.

The crude lipid extracts were separated on a re-verse-phase HPLC column. Psychosine and the inter-

nal standard, lacotsylsphingosine, were separatedfrom salts in the injection front and other lipids on aZorbax (Agilent Technologies, Wilmington, DE) C18

reverse-phase column (2.1 mm id � 15 cm) developedat 0.25 ml/min (Alliance 2690, Waters Corp., Milford,MA). The following ternary gradient was used withthe mobile phase composition of A � 99% H2O/1%MeOH/0.08% formic acid, B � 99% MeOH/1%H2O/0.06% formic acid, and C � 90% isopropanol/10%H2O.

Time A% B% C%

0.0 50 50 0.02.0 50 50 0.05.6 0.0 100 0.015 0.0 100 0.025 10 10 80

The retention times for psychosine and lacotsyl-sphingosine were 8.4 and 8.3 min, respectively. Theisopropanol portion of this busy gradient is requiredto elute and separate more hydrophobic sphingolipidsalso under study.

Psychosine and the internal standard lacotsylsphin-gosine were detected and quantitated by electrosprayionization and tandem mass spectrometry. The colli-sion spectra from MH� of most sphingolipids containabundant ion current from the base portion of themolecule, and multiple reaction monitoring methodsare becoming common for their sensitive, specific de-tection (Mano et al., 1997; Sullards, 2000; Sullards &Merrill 2001; Whitefield et al., 2001). Psychosine andlacotsylsphingosine were detected in HPLC effluentusing multiple reaction monitoring of transition spe-cific for each molecule, for psychosine m/z 462 to 264and lacotsylsphingosine m/z 624 to 264 (Whitefield etal., 2001). Tandem mass spectra were acquired on a“triple” quadrupole instrument (Ultima, MicromassLtd., Manchester, UK). The electrospray source blockwas 100°C and probe desolvation temperature was200°C. Ar collision gas was set to attenuate the beamto 20% (1.7 e�3 mbar on a gauge near the collision cell).The sampling cone voltage was 20 V. For both transi-tions collision energy was 22 eV and 0.1 s acquisitiondwell was used.

Statistical Analyses

Student’s unpaired, two-tailed t test was used forcomparisons of survival, onset day of tremor, andpsychosine concentrations between C57BL/6 and



C57BL/6 � CAST/Ei twitcher mice. This test also wasused to evaluate the life span between agouti andblack C57BL/6 � CAST/Ei twitcher mice. Statisticalsignificance was achieved when P � 0.05.

RESULTS

Life Span

The average life span of twitcher mice on a C57BL/6 � CAST/Ei background was 61.4 � 2.5 days (n �19), which was significantly greater than the averagelife span of 37.0 � 0.6 days for twitcher mice (n � 5) onthe C57BL/6 background (Fig. 1A). The life span ofC57BL/6 twitcher mice in this study was consistentwith that described in previously published studies(Kobayashi et al., 1980; Seller et al., 1986). There was a

FIG. 1. (A) Life span of twitcher mice on C57BL/6 � CAST/Eibackground vs twitcher mice on a C57BL/6 background.Twitcher mice on a mixed background lived significantly (P �0.001, n � 19) longer than mice on a C57BL/6 background (n �5). (B) Comparison of the life span of C57BL/6 � CAST/Eitwitcher mice with agouti coat color vs black coat color. Twitchermice with agouti coat lived significantly (P � 0.05) longer thanmice with black coat color.

FIG. 2. (A) The day of onset of tremor in twitcher mice on C57BL/6 � CAST/Ei background (n � 15) vs C57BL/6 twitcher mice (n �5). Twitcher mice on the mixed background had a significantlydelayed (P � 0.001) onset day of tremor compared to C57BL/6twitcher mice, which was at 21 days. (B) Mean score for walkingability of the twitcher mice on the mixed background. Scoring wasperformed every 10th day according to the rank scale for walkingability (see Material and Methods).

348 Biswas et al.


wide variation in the life span of C57BL/6 � CAST/Eitwitcher mice with a range of 45–77 days, but therewas no difference in the average life span for male vsfemale C57BL/6 � CAST/Ei twitcher mice. Since thetwitcher mice on the C57BL/6 � CAST/Ei back-ground had either an agouti or a black coat color, thelife span data were divided into two groups based onthe coat color. The mean life span of C57BL/6 �CAST/Ei twitcher mice with an agouti coat color was65.5 � 2.9 days (n � 11), which was significantly (P �0.02) longer than that of C57BL/6 � CAST/Ei twitchermice with a black coat color, which lived up to 55.0 �2.2 days (n � 7) (Fig. 1B).

Clinical Observations

Twitcher mice on the C57BL/6 � CAST/Ei back-ground had a delayed age of onset of tremor, with amean of 24 days and a range of 23–26 days comparedto 21 days for C57BL/6 twitcher mice (Fig. 2A). Mostof the twitcher mice on a C57BL/6 � CAST/Ei back-ground had a normal gait up to 30 or 35 days (Fig. 2B),and mice which lived beyond 60 days showed nosigns of an abnormal gait even at day 40. In compar-ison, twitcher mice on a C57BL/6 background wouldtypically display a tremor at 21 days after birth (Fig.2A), develop a wobbly gait around 28–30 days thatwould rapidly progress to paralysis and wasting ofthe hind limbs, and die at �37 days. Notably, thelonger living (72–77 days) twitcher mice on theC57BL/6 � CAST/Ei background showed more se-vere wasting of the hind limbs at moribund conditions(Fig. 3C) compared to C57BL/6 twitcher mice at mor-ibund conditions (Fig. 3B). Several C57BL/6 �CAST/Ei twitcher mice had a reasonably normal ap-

pearance at 40 days (Fig. 3A), which is past the mor-ibund age of C57BL/6 twitcher mice.

Pathology

Griffiona simplicifolia lectin. Control mice did nothave any lectin-positive cells in the cerebellum (Fig.4A). Twitcher mice on a C57BL/6 � CAST/Ei back-ground had fewer G. simplicifolia lectin-positive cells inthe cerebellum at a moribund condition (i.e., 76 days;Fig. 4C) compared to twitcher mice on C57BL/6 back-ground at a moribund condition (i.e., 38 days; Fig. 4B).In the pons and medulla, control mice did not haveany lectin-positive cells, while fewer lectin-positivecells were observed in twitcher mice on the C57BL/6� CAST/Ei background compared to twitcher miceon a C57BL/6 background.

MBP. MBP staining of the cerebellar white matterrevealed an abundance of myelin in control animals(Fig. 4D). In twitcher mice on a C57BL/6 background,there was a loss of MBP staining (Fig. 4E) indicative ofextensive demyelination in this region as establishedin earlier studies (Matsushima et al., 1994; Pedchenko& LeVine 1999). In twitcher mice on the C57BL/6 �CAST/Ei background, there was evidence of greatermyelin preservation than for C57BL/6 twitcher mice,although there were still focal areas of demyelination(Fig. 4F).

Parvalbumin. In control mice, parvalbumin immu-noreactivity was observed in the cerebellar Purkinjeneurons and neurons in the molecular layer (Fig. 5A).Generally, there was an absence of staining in thegranular layer. In twitcher mice on the C57BL/6 �CAST/Ei background, parvalbumin immunostainingwas seen in focal swellings of various sizes in pro-

FIG. 3. (A) Appearance of a 40-day-old twitcher mouse on C57BL/6 � CAST/Ei background compared to (B) a moribund twitcher mouseon a C57BL/6 background at 38 days. (C) The longer living C57BL/6 � CAST/Ei twitcher mice had more severe wasting of hind legs thanC57BL/6 twitcher mice at moribund conditions. The mouse shown is at 76 days.



cesses in the granular layer (Fig. 5B) in addition to thestaining observed in control mice. These swellings,indicative of axonal pathology, were also present in

the cerebellar white matter (Fig. 5C). Axonal swellingswere generally not seen in the cerebellum of moribundtwitcher mice on a C57BL/6 background.

FIG. 4. Lectin staining (A–C) and MBP immunostaining (D–F) in the cerebellum of control C57BL/6 � CAST/Ei mice, twitcher mice on aC57BL/6 background, and twitcher mice on C57BL/6 � CAST/Ei background. (A) Control mice (76 days old) did not show any lectin-positivecells whereas (B) twitcher mice on a C57BL/6 background had numerous lectin-positive cells at moribund age (38 days) as reported earlier.(C) C57BL/6 � CAST/Ei twitcher mice had relatively less infiltration of lectin-positive cells even at moribund conditions (e.g., 76 days)compared to C57BL/6 twitcher mice. (D) Control mice had abundant myelin staining while (E) C57BL/6 twitcher mice showed extensive lossof myelin staining at a moribund age. (F) C57BL/6 � CAST/Ei twitcher mice had relatively greater MBP staining than C57BL/6 twitcher miceat moribund age, indicative of greater myelin preservation in the cerebellum. Bar � 100 �m.

FIG. 5. Parvalbumin immunostaining in the cerebellum of control and twitcher mice on C57BL6 � CAST/Ei background. (A) In control mice,parvalbumin immunoreactivity was seen in neurons in the molecular layer and in Purkinje neurons and their processes in the molecular layer.There was typically an absence of staining in the granular layer. (B) In C57BL6 � CAST/Ei twitcher mice axonal spheroids (focal swelling ofvarious sizes) (arrows) were seen in the axonal plexus of the granular layer and in (C) the cerebellar white matter. Bar � 40 �m.

350 Biswas et al.


GFAP. Control mice had a few lightly stainedGFAP-positive processes and/or astrocytes in the cer-ebellum and pons (Figs. 6A and 6D). An increase instaining above control levels was observed in the cer-ebellar granular layer in both C57BL/6 and C57BL/6 � CAST/Ei twitcher mice (Figs. 6B and 6C), but itwas more extensive in the latter. However, in the ponsof twitcher C57BL/6 � CAST/Ei mice the density ofGFAP-immunoreactive astrocytes was less than thatobserved in C57BL/6 twitcher mice (Figs. 6E and 6F).Furthermore, the large degree of swelling observed inthe somas of astrocytes from C57BL/6 twitcher mice(Fig. 6E) appeared to be qualitatively less in astrocytesfrom C57BL/6 � CAST/Ei twitcher mice (Fig. 6F).

Sciatic nerve. Since long-surviving C57BL/6 �CAST/Ei twitcher mice displayed severe wasting oftheir hindquarters (Fig. 3C), their sciatic nerves wereexamined following dissection. The nerves fromC57BL/6 � CAST/Ei twitcher mice had a greaterdegree of swelling than nerves from C57BL/6 twitcher

mice, and the latter nerves were swollen in compari-son to nerves from normal animals, which has beendescribed earlier (Powell et al., 1983). In order to com-pare the sciatic nerves between these groups, 1-�m-thick cross sections of nerves were prepared and in-spected via light microscopy. In nerves from long-surviving C57BL/6 � CAST/Ei twitcher mice, therewas evidence of demyelination and myelin debris, butthinly myelinated axons, which retained an ovalshape, were still abundant (Fig. 7C). There was a lossof discrete fascicules and an increased cellularity oflipid-filled cells appeared to account for an increasedoverall density within the nerve (Fig. 7C) compared tonormal and C57BL/6 twitcher nerves. In nerves fromC57BL/6 twitcher mice in a moribund state, there wasevidence of demyelination, myelin debris, and someC-shaped and several thinly myelinated axons (Fig.7B). Individual fascicules were still apparent in thesenerves, albeit the fascicules were somewhat disorga-nized. In control mice, there was a high density of

FIG. 6. GFAP staining in the cerebellum (A–C) and pons (D–F) of control C57BL/6 � CAST/Ei mice, twitcher C57BL/6 mice, or twitcherC57BL/6 � CAST/Ei mice. (A) Occasionally, light staining of a few GFAP processes was present in the cerebellum of control C57BL/6 �CAST/Ei mice. (B) In C57BL/6 twitcher mice, mild staining of astrocyte processes was seen occasionally in the granular layer of the cerebellum.(C) In C57BL/6 � CAST/Ei twitcher mice, astrocyte staining in the granular layer was greater than that observed in C57BL/6 twitcher mice.(D) A few lightly stained astrocytes were seen in the pons of control C57BL/6 � CAST/Ei mice. (E) In pons, gliosis was relatively severe withswollen astrocytes cell bodies in C57BL/6 twitcher mice. (F) In C57BL/6 � CAST/Ei twitcher mice, gliosis was present in the pons but theastrocyte somas were typically not as large as those that were readily apparent in C57BL/6 twitcher mice. Bar � 20 �m (A–C), 40 �m (D–F).



myelinated axons that were organized into fascicules.Each axon had a thick myelin sheath surrounding it(Fig. 7A).

Ultrastructure. In the pons of twitcher mice with aC57BL/6 � CAST/Ei background, there were charac-teristic globoid cells containing filamentous inclu-sions, a granular cytoplasm, and myelin fragments(Fig. 8A) that were similar to globoid cells describedfor twitcher mice on the C57BL/6 background (Koba-yashi et al., 1980; Takahashi & Suzuki 1982). There wasa significant preservation of myelin sheaths in thecervical spinal cord region of twitcher mice on theC57BL/6 � CAST/Ei background (Fig. 8B). This pres-ervation of myelin was similar to the findings ob-served in the cerebellum using MBP immunohisto-chemical staining (Fig. 4F).

Psychosine Concentrations

The concentration of psychosine in the pons/me-dulla was dramatically increased in C57BL/6 twitcherand C57BL/6 � CAST/Ei twitcher mice relative totheir respective control nontwitcher mice (Fig. 9). Thelevels of psychosine were comparable between mori-bund C57BL/6 twitcher mice and 40-day C57BL/6 �CAST/Ei twitcher mice (Fig. 9), which were at approx-imately the same age. Furthermore, moribund C57BL/6 � CAST/Ei twitcher mice at 60–63 days also had

concentrations of psychosine similar to levels in theabove-mentioned twitcher mice (Fig. 9).

DISCUSSION

Twitcher mice on a C57BL/6 � CAST/Ei back-ground had a 66% increase in life span and a slowerprogression of clinical signs compared to C57BL/6twitcher mice. Twitcher mice on the mixed back-ground also had less infiltration of lectin-positive mac-rophages in the cerebellum, pons, and medulla andless gliosis in the pons at the terminal stage of thedisease compared to moribund C57BL/6 twitchermice. Also, there was evidence of greater myelin pres-ervation although areas of demyelination were stillpresent. These results suggest that modifier genes inthe CAST/Ei background altered the clinical andpathological manifestations of the twitcher disease.

At a moribund stage, C57BL/6 � CAST/Ei twitchermice were more alert, in general, despite completehind limb paralysis compared to moribund C57BL/6twitcher mice. The sciatic nerve was found to be moreswollen with an increased cellularity in two of thelonger living twitcher mice, compared to C57BL/6twitcher mice, indicating that the peripheral pathol-ogy was perhaps more severe than the CNS pathology

FIG. 7. Cross section of a sciatic nerve from a C57BL/6 � CAST/Ei control mouse, a C57BL/6 twitcher, and a C57BL/6 � CAST/Ei twitchermouse. (A) The nerve bundles in C57BL/6 � CAST/Ei control mice had a high density of myelinated nerve fibers organized into fascicles.Generally, a thick myelin sheath surrounded each axon. (B) In a C57BL/6 twitcher mouse, at a moribund condition, there was loss of myelinaround axons and myelin debris was present. The spaces between axons were enlarged and the cellularity was increased compared to thatobserved in control mice. (C) In C57BL/6 � CAST/Ei twitcher mice, at a moribund stage, there was demyelination and presence of myelindebris. Many thinly myelinated axons retained their oval shape. There was an increased cellularity compared to control and C57BL/6 twitchermice and many of these cells contained extensive vacuoles and/or lipid droplets. Bar � 20 �m. Filled arrowheads point to C-shaped myelinatedaxons and open arrowheads identify densely stained material which could be myelin debris.

352 Biswas et al.


FIG. 8. Electron micrographs from the pons of C57BL/6 � CAST/Ei twitcher mice. (A) Globoid cells with filamentous inclusions were present similarto previous descriptions of globoid cells in C57BL/6 twitcher mice. (B) There were regions with significant preservation of myelin. Bar � 10 �m.



and possibly a critical factor in determining the lifespan in these mice.

Astrocyte gliosis has been observed in some CNSgray matter regions of C57BL/6 twitcher mice (Koba-yashi et al., 1986; LeVine et al., 1994), but not typicallyin the cerebellum. In the C57BL/6 � CAST/Eitwitcher mice there was an increase in GFAP stainingin the granular layer of the cerebellum. Since astro-cytic reactions have been associated with neuronalpathology in the cerebellum in other disease states, weexamined the integrity of axons by parvalbumin im-munohistochemical staining. We detected somepathological alterations, i.e., focal swellings of varioussizes in axons in the granular layer and the cerebellarwhite matter of C57BL/6 � CAST/Ei twitcher mice.These alterations could be indicative of Purkinje neu-ron degeneration since the axonal swellings were sim-ilar to those seen in a mouse model of genetic cerebel-lar degeneration (Higashi et al., 1993), in patients withspinocerebellar degeneration (Ishikawa et al., 1995), orin spinocerebellar ataxia (Yang et al., 2000), which allinvolve significant loss of Purkinje neurons. AlthoughC57BL/6 twitcher mice had a small increase in GFAPstaining in the granular layer, axonal pathology as

revealed by parvalbumin staining was not seen in thecerebellum. Therefore, axonal spheroids observed inC57BL/6 � CAST/Ei twitcher mice could be associ-ated with a late phase of the disease in long-survivingmice. Interestingly, extensive loss of granule cells andmoderate loss of Purkinje cells have been found in apatient with juvenile GCL disease (Hirato et al., 1994).Thus, the cerebellar changes together with the slowerclinical and pathological changes suggest thatC57BL/6 � CAST/Ei twitcher mice may have a dis-ease that is similar to juvenile GCL while C57BL/6twitcher mice have a disease that is similar to theinfant form of GCL.

Recent studies have shown that the various genesencoding for products involved in the sphingolipidmetabolic pathway, e.g., saposin A, saposin C, andacid �-galactosidase, may act as modifiers for the phe-notypic expression of the twitcher disease. For exam-ple, accumulation of the toxic metabolite psychosinehas been shown to result not only from deficiency ofgalactosylceramidase but also from a selective defi-ciency of saposin A and/or C (Harzer et al., 2001) anda mutation in the saposin A domain of the prosaposingene resulted in a late-onset, chronic form of globoidcell leukodystrophy (Matsuda et al., 2001b). Similarly,acid �-galactosidase gene dosage exerts a paradoxicalinfluence on the twitcher phenotype. Twitcher micewith a complete deficiency of acid �-galactosidasehave a milder phenotype with a longer life span andsubstantially reduced CNS pathology, whereas a sin-gle dose of the gene produces a more severe disease,with a shorter life span and increased psychosine lev-els compared to a twitcher mouse with normal acid�-galactosidase activity (Tohyama et al., 2000). If sa-posin A, saposin C, or acid �-galactosidase were re-sponsible for disease attenuation in C57BL/6 �CAST/Ei twitcher mice, then the level of psychosineshould have been reduced in these mice compared tothe level in C57BL/6 twitcher mice, since the concen-tration of psychosine is expected to correlate with thedegree of pathology (Igisu & Suzuki, 1984; Shinoda etal., 1987) and the former mice had substantially lesspathology. However, it appears that alleles of thesethree proteins are not responsible for the slower dis-ease in C57BL/6 � CAST/Ei twitcher mice since thelevels of psychosine are equivalent between moribundC57BL/6 twitcher mice and C57BL/6 � CAST/Eitwitcher mice at an equivalent age. These results sug-gest that the disease-modifying effects in C57BL/6 �CAST/Ei twitcher mice may be due to the ability ofvarious cells, e.g., oligodendrocytes, and/or myelin toresist the toxic effects of psychosine or that myelin

FIG. 9. Psychosine concentrations in the pons/medulla of mori-bund C57BL/6 twitcher mice (filled bars, n � 4) and age-matchedC57BL/6 controls (open bars, n � 2), 40-day C57BL/6 � CAST/Eitwitcher mice (n � 4) and age-matched C57BL/6 � CAST/Ei con-trols (n � 3), and moribund C57BL/6 � CAST/Ei twitcher mice(n � 3) and age-matched C57BL/6 � CAST/Ei controls (n � 3).Although psychosine levels were always elevated in twitcher micecompared to levels in control animals, there was no appreciabledifference in psychosine levels between C57BL/6 twitcher mice andC57BL/6 � CAST/Ei twitcher mice.

354 Biswas et al.


repair or oligodendrocyte regeneration is more vigor-ous in C57BL/6 � CAST/Ei twitcher mice comparedto that in C57BL/6 twitcher mice. This possibilitywould explain the reduced loss of myelin and mildergliosis in C57BL/6 � CAST/Ei twitcher mice but stillaccount for the relatively high level of psychosine. Insupport of this idea, high levels of estrogen or preg-nancy were found to significantly delay the diseasecourse in saposin-A-deficient mice but pregnancy didnot lower psychosine levels (Matsuda et al., 2001a).Since estrogen (Toung et al., 1998) and pregnancy mayoffer protection to cells in the brain, it is plausible thatprotective mechanisms, induced in C57BL/6 �CAST/Ei twitcher mice, slow the disease course with-out altering the concentration of psychosine. Anotherprotein implicated in repair mechanisms is apoli-poprotein E, which depending on the allele can resultin a greater susceptibility to Alzheimer’s disease(Strittmatter et al., 1993) or a faster rate of disabilityprogression in multiple sclerosis (Chapman et al.,2001). Cholesterol is also important for maintainingthe integrity of the myelin sheath. It is largely presentin the membranes of glial cells, myelin, and neurons(Dietschy & Turley, 2001), and there is a large amountof cholesterol turnover in oligodendrocytes and neu-rons during myelination, repair, and remodeling.CAST/Ei mice have low plasma levels of phospholip-ids and HDL cholesterol compared to C57BL/6 mice(Albers et al., 1999; Mehrabian et al., 2000). Althoughcholesterol within the brain is derived almost entirelythrough in situ synthesis by brain cells (Jurevics &Morell, 1995), it is possible that it is involved in thedisease-modifying effects observed in C57BL/6 �CAST/Ei twitcher mice.

Previous studies have shown that the course of thetwitcher disease could be slowed by restricting thesynthesis of substrates for galactosylceramidase, i.e.,galactosylceramide and psychosine, via inhibition of3-ketodihydrosphingosine synthase with l-cyclo-serine (LeVine et al., 2000). Similarly, administration ofN-butyldeoxynojirimycin, an inhibitor of the first stepof glycosphingolipid synthesis, alleviated clinicalsigns in an animal model of Sandhoff’s disease (Jeya-kumar et al., 1999) and in humans with Gaucher dis-ease (Cox et al., 2000). Thus, a slower disease course inC57BL/6 � CAST/Ei twitcher mice could have beendue to differences in the metabolic turnover of sphin-golipids which are involved in the formation and/orturnover of myelin. However, the finding that psycho-sine accumulation is equivalent between C57BL/6 andC57BL/6 � CAST/Ei twitcher mice makes it unlikely

that synthetic or turnover rate is responsible for thedifferent disease courses.

CAST/Ei mice are characterized by unique alleles inthe Ly-1 (CD5�) and Ly-2 (CD8�) loci which code forspecific cell surface antigens on the CD5� B cells andCD8� cytotoxic T cells, respectively (Kurihara et al.,1985). In twitcher mice, an initial immune activation inthe CNS with MHC class II antigen expression andinfiltration of CD4� T lymphocytes from the circula-tion occurs with the initiation of demyelination (Ohnoet al., 1993), and the absence of MHC class II moleculesreduced CNS demyelination, macrophage infiltration,and severity of twitching in twitcher mice (Matsu-shima et al., 1994). Although CD5� and CD8� cellswere not detected in the CNS of twitcher mice, theirinvolvement in the pathogenesis cannot be ruled out.Notably, CD5� B cells belong to a subpopulation withpotential autoreactive properties and the CD5 antigenhas been shown to act as an activation marker forsecretion of anti-myelin antibodies in patients withpolyneuropathy (Ekerfelt et al., 1995). The presence ofdifferent alleles for these CD antigens may result in analtered immune response in the C57BL/6 � CAST/Eitwitcher mice which could account for a milder dis-ease course. In addition to possible contributions fromthe immune system, reactive astrocyte gliosis mayserve a protective role in GCL and modulations of thegliotic response could affect the disease course.

In conclusion, C57BL/6 � CAST/Ei twitcher micehad a slower pathological and clinical course thanC57BL/6 twitcher mice, despite having high levels ofpsychosine at 40 days. Thus, modifiers proteins thatprotect against psychosine toxicity may account forthe milder disease course. The identification of modi-fier proteins will help to uncover factors that can delaydisease progression, and this information could beused to design potential therapies for the variousforms of GLD. In addition, twitcher mice on theCAST/Ei � C57BL/6 background may be a usefulmodel of the late-onset form of human GCL.

ACKNOWLEDGMENTS

The authors thank Julie Collins and Marsha Danley for MBP andGFAP immunohistochemical staining, Barbara Fegley and KarenGrantham for help with electron micrographs, Davin Watne, MRRCImaging Center, for photograph layouts, and Brian Smith, LAR-KUMC, for help with handling the CAST/Ei mice. This work wassupported by NINDS (NS36544), NICHD Mental Retardation Re-search Center Grant HD 02528, and the KU Research DevelopmentFund (TDW).



REFERENCES

Albers, J. J., Pitman, W., Wolfbauer, G., Cheung, M. C., Kennedy, H.,Tu, A. Y., Marcovina, S. M., & Paigen, B. (1999) Relationshipbetween phospholipid transfer protein activity and HDL leveland size among inbred mouse strains. J. Lipid Res. 40, 295–301.

Bernardini, G. L., Herrera, D. G., Carson, D., DeGasperi, R., GamaSosa, M. A., Kolodny, E. H., & Trifiletti, R. (1997) Adult-onsetKrabbe’s disease in siblings with novel mutations in the galacto-cerebrosidase gene. Ann. Neurol. 41, 111–114.

Chapman, J., Vinokurov, S., Achiron, A., Karussis, D. M., Mitosek-Szewczyk, K., Birnbaum, M., Michaelson, D. M., & Korczyn, A. D.(2001) APOE genotype is a major predictor of long-term progres-sion of disability in MS. Neurology 56, 312–316.

Cho, K. H., Kim, M. W., & Kim, S. U. (1997) Tissue culture model ofKrabbe’s disease: Psychosine cytotoxicity in rat oligodendrocyteculture. Dev. Neurosci. 19, 321–327.

Cox, T., Lachmann, R., Hollak, C., Aerts, J., van Weely, S., Hrebicek,M., Platt, F., Butters, T., Dwek, R., Moyses, C., Gow, I., Elstein, D.,& Zimran, A. (2000) Novel oral treatment of Gaucher’s diseasewith N-butyldeoxynojirimycin (OGT 918) to decrease substratebiosynthesis. Lancet 355, 1481–1485.

Crome, L., Hanefeld, F., Patrick, D., & Wilson, J. (1973) Late onsetgloboid cell leucodystrophy. Brain 96, 841–848.

De Gasperi, R., Gama Sosa, M. A., Sartorato, E., Battistini, S., Ragha-van, S., & Kolodny, E. H. (1999) Molecular basis of late-lifegloboid cell leukodystrophy. Hum. Mutat. 14, 256–262.

De Gasperi, R., Gama Sosa, M. A., Sartorato, E. L., Battistini, S.,MacFarlane, H., Gusella, J. F., Krivit, W., & Kolodny, E. H. (1996)Molecular heterogeneity of late-onset forms of globoid-cell leu-kodystrophy. Am. J. Hum. Genet. 59, 1233–1242.

Dietschy, J. M., & Turley, S. D. (2001) Cholesterol metabolism in thebrain. Curr. Opin. Lipidol. 12, 105–112.

Duchen, L. W., Eicher, E. M., Jacobs, J. M., Scaravilli, F., & Teixeira,F. (1980) Hereditary leucodystrophy in the mouse: The new mu-tant twitcher. Brain 103, 695–710.

Ekerfelt, C., Ernerudh, J., Solders, G., & Vrethem, M. (1995) CD5expression on B cells may be an activation marker for secretion ofanti-myelin antibodies in patients with polyneuropathy associ-ated with monoclonal gammopathy. Clin. Exp. Immunol. 101, 346–350.

Fiumara, A., Pavone, L., Siciliano, L., Tine, A., Parano, E., & Innico,G. (1990) Late-onset globoid cell leukodystrophy. Report on sevennew patients. Childs Nerv. Syst. 6, 194–197.

Fluharty, A. L., Neidengard, L., Holtzman, D., & Kihara, H. (1986)Late-onset Krabbe disease initially diagnosed as cerebroside sul-fatase activator deficiency. Metab. Brain Dis. 1, 187–195.

Fu, L., Inui, K., Nishigaki, T., Tatsumi, N., Tsukamoto, H., Kokubu,C., Muramatsu, T., & Okada, S. (1999) Molecular heterogeneity ofKrabbe disease. J. Inherit. Metab. Dis. 22, 155–162.

Furuya, H., Kukita, Y., Nagano, S., Sakai, Y., Yamashita, Y.,Fukuyama, H., Inatomi, Y., Saito, Y., Koike, R., Tsuji, S., Fuku-maki, Y., Hayashi, K., & Kobayashi, T. (1997) Adult onset globoidcell leukodystrophy (Krabbe disease): Analysis of galactosylcer-amidase cDNA from four Japanese patients. Hum. Genet. 100,450–456.

Harzer, K., Hiraiwa, M., & Paton, B. C. (2001) Saposins (sap) A andC activate the degradation of galactosylsphingosine. FEBS Lett.508, 107–110.

Higashi, Y., Murayama, S., Pentchev, P. G., & Suzuki, K. (1993)Cerebellar degeneration in the Niemann–Pick type C mouse. ActaNeuropathol. 85, 175–184.

Hirato, J., Nakazato, Y., Sasaki, A., Hikima, A., Shimizu, S., &Yamanouchi, H. (1994) Krabbe’s disease with giant lamellar bod-ies in Purkinje cells. Acta Neuropathol. 88, 78–84.

Igisu, H. (1989) [Psychosine: A “toxin” produced in the brain—itsmechanism of action]. J. Uoeh. 11, 487–493.

Igisu, H., & Suzuki, K. (1984) Progressive accumulation of toxicmetabolite in a genetic leukodystrophy. Science 224, 753–755.

Ishikawa, K., Mizusawa, H., Fujita, T., Ohkoshi, N., Doi, M., Ko-matsuzaki, Y., Iwamoto, H., Ogata, T., & Shoji, S. (1995) Cal-bindin-D 28k immunoreactivity in the cerebellum of spinocere-bellar degeneration. J. Neurol. Sci. 129, 179–185.

Jardim, L. B., Giugliani, R., Pires, R. F., Haussen, S., Burin, M. G.,Rafi, M. A., & Wenger, D. A. (1999) Protracted course of Krabbedisease in an adult patient bearing a novel mutation. Arch. Neurol.56, 1014–1017.

Jeyakumar, M., Butters, T. D., Cortina-Borja, M., Hunnam, V., Proia,R. L., Perry, V. H., Dwek, R. A., & Platt, F. M. (1999) Delayedsymptom onset and increased life expectancy in Sandhoff diseasemice treated with N-butyldeoxynojirimycin. Proc. Natl. Acad. Sci.USA 96, 6388–6393.

Jurevics, H., & Morell, P. (1995) Cholesterol for synthesis of myelinis made locally, not imported into brain. J. Neurochem. 64, 895–901.

Kobayashi, S., Chiu, F. C., Katayama, M., Sacchi, R. S., & Suzuki, K.(1986) Expression of glial fibrillary acidic protein in the CNS andPNS of murine globoid cell leukodystrophy, the twitcher. Am. J.Pathol. 125, 227–243.

Kobayashi, T., Yamanaka, T., Jacobs, J. M., Teixeira, F., & Suzuki, K.(1980) The twitcher mouse: An enzymatically authentic model ofhuman globoid cell leukodystrophy (Krabbe disease). Brain Res.202, 479–483.

Kolodny, E. H., Raghavan, S., & Krivit, W. (1991) Late-onset Krabbedisease (globoid cell leukodystrophy): Clinical and biochemicalfeatures of 15 cases. Dev. Neurosci. 13, 232–239.

Krivit, W., Lockman, L. A., Watkins, P. A., Hirsch, J., & Shapiro,E. G. (1995) The future for treatment by bone marrow transplan-tation for adrenoleukodystrophy, metachromatic leukodystro-phy, globoid cell leukodystrophy and Hurler syndrome. J. Inherit.Metab. Dis. 18, 398–412.

Kurihara, Y., Miyashita, N., Moriwaki, K., Petras, M. L., Bonhomme,F., Wang, S. C., & Kohno, S. (1985) Serological survey of T-lymphocyte differentiation antigens in wild mice. Immunogenetics22, 211–218.

LeVine, S. M., Pedchenko, T. V., Bronshteyn, I. G., & Pinson, D. M.(2000) l-Cycloserine slows the clinical and pathological course inmice with globoid cell leukodystrophy (twitcher mice). J. Neurosci.Res. 60, 231–236.

LeVine, S. M., Wetzel, D. L., & Eilert, A. J. (1994) Neuropathology oftwitcher mice: Examination by histochemistry, immunohisto-chemistry, lectin histochemistry and Fourier transform infraredmicrospectroscopy. Int. J. Dev. Neurosci. 12, 275–288.

Luzi, P., Rafi, M. A., Zaka, M., Curtis, M., Vanier, M. T., & Wenger,D. A. (2001) Generation of a mouse with low galactocerebrosidaseactivity by gene targeting: A new model of globoid cell leukodys-trophy (Krabbe disease). Mol. Genet. Metab. 73, 211–223.

Mano, N., Oda, Y., Yamada, K., Asakawa, N., & Katayama, K. (1997)Simultaneous quantitative determination method for sphingo-lipid metabolites by liquid chromatography/ionspray ionizationtandem mass spectrometry. Anal. Biochem. 244, 291–300.

Matsuda, J., Vanier, M. T., Saito, Y., Suzuki, K., & Suzuki, K. (2001a)Dramatic phenotypic improvement during pregnancy in a geneticleukodystrophy: Estrogen appears to be a critical factor. Hum.Mol. Genet. 10, 2709–2715.

356 Biswas et al.


Matsuda, J., Vanier, M. T., Saito, Y., Tohyama, J., & Suzuki, K.(2001b) A mutation in the saposin A domain of the sphingolipidactivator protein (prosaposin) gene results in a late-onset, chronicform of globoid cell leukodystrophy in the mouse. Hum. Mol.Genet. 10, 1191–1199.

Matsushima, G. K., Taniike, M., Glimcher, L. H., Grusby, M. J.,Frelinger, J. A., Suzuki, K., & Ting, J. P. (1994) Absence of MHCclass II molecules reduces CNS demyelination, microglial/mac-rophage infiltration, and twitching in murine globoid cell leuko-dystrophy. Cell 78, 645–656.

Mehrabian, M., Castellani, L. W., Wen, P. Z., Wong, J., Rithaporn, T.,Hama, S. Y., Hough, G. P., Johnson, D., Albers, J. J., Mottino,G. A., Frank, J. S., Navab, M., Fogelman, A. M., & Lusis, A. J.(2000) Genetic control of HDL levels and composition in an in-terspecific mouse cross (CAST/Ei � C57BL/6J). J. Lipid Res. 41,1936–1946.

Miyataki, T., & Suzuki, K. (1972) Globoid cell leukodystrophy:Additional deficiency of psychosine galactosidase. Biochem. Bio-phys. Res. Commun. 48, 538–543.

Ohno, M., Komiyama, A., Martin, P. M., & Suzuki, K. (1993) MHCclass II antigen expression and T-cell infiltration in the demyeli-nating CNS and PNS of the twitcher mouse. Brain Res. 625,186–196.

Pedchenko, T. V., Bronshteyn, I. G., & LeVine, S. M. (2000) TNF-receptor 1 deficiency fails to alter the clinical and pathologicalcourse in mice with globoid cell leukodystrophy (twitcher mice)but affords protection following LPS challenge. J. Neuroimmunol.110, 186–194.

Pedchenko, T. V., & LeVine, S. M. (1999) IL-6 deficiency causesenhanced pathology in Twitcher (globoid cell leukodystrophy)mice. Exp. Neurol. 158, 459–468.

Powell, H. C., Knobler, R. L., & Myers, R. R. (1983) Peripheralneuropathy in the twitcher mutant. A new experimental model ofendoneurial edema. Lab. Invest. 49, 19–25.

Sakai, N., Inui, K., Tatsumi, N., Fukushima, H., Nishigaki, T., Tani-ike, M., Nishimoto, J., Tsukamoto, H., Yanagihara, I., Ozono, K., &Okada, S. (1996) Molecular cloning and expression of cDNA formurine galactocerebrosidase and mutation analysis of thetwitcher mouse, a model of Krabbe’s disease. J. Neurochem. 66,1118–1124.

Seller, M. J., Perkins, K. J., & Fensom, A. H. (1986) Galactosylcere-brosidase activity in tissues of twitcher mice with and withoutbone marrow transplantation. J. Inherit. Metab. Dis. 9, 234–238.

Shinoda, H., Kobayashi, T., Katayama, M., Goto, I., & Nagara, H.(1987) Accumulation of galactosylshingosine (psychosine) in thetwitcher mouse: Determination by HPLC. J. Neurochem. 49, 92–99.

Strittmatter, W. J., Saunders, A. M., Schmechel, D., Pericak-Vance,M., Enghild, J., Salvesen, G. S., & Roses, A. D. (1993) Apolipopro-tein E: High-avidity binding to beta-amyloid and increased fre-quency of type 4 allele in late-onset familial Alzheimer disease.Proc. Natl. Acad. Sci. USA 90, 1977–1981.

Sullards, M. C. (2000) Analysis of sphingomyelin, glucosylceramide,ceramide, sphingosine and sphingosine l-phosphate by tandemmass spectrometry. In Methods in Enzymology, Vol. 312 (A. H.Merrill & K. A. Hunnun, Eds.) Academic Press, New York.

Sullards, M. C., & Merrill, A. H., Jr. (2001) Analysis of sphingosine1-phosphate, ceramides, and other bioactive sphingolipids byhigh-performance liquid chromatography-tandem mass spec-trometry. Protocol from Science’s Stke methods. SciSTKE2001(67):PL1.

Svennerholm, L., Vanier, M. T., & Mansson, J. E. (1980) Krabbedisease: A galactosylsphingosine (psychosine) lipidosis. J. LipidRes. 21, 53–64.

Takahashi, H., & Suzuki, K. (1982) Globoid cell leukodystrophy:Specialized contact of globoid cell with astrocyte in the brain oftwitcher mouse. Acta Neuropathol. 58, 237–242.

Tanaka, K., & Webster, H. D. (1993) Effects of psychosine (galacto-sylsphingosine) on the survival and the fine structure of culturedSchwann cells. J. Neuropathol. Exp. Neurol. 52, 490–498.

Taniike, M., & Suzuki, K. (1994) Spacio-temporal progression ofdemyelination in twitcher mouse: With clinico-pathological cor-relation. Acta Neuropathol. 88, 228–236.

Tapasi, S., Padma, P., & Setty, O. H. (1998) Effect of psychosine onmitochondrial function. Indian J. Biochem. Biophys. 35, 161–165.

Taylor, B. A., Navin, A., & Phillips, S. J. (1994) PCR-amplification ofsimple sequence repeat variants from pooled DNA samples forrapidly mapping new mutations of the mouse. Genomics 21, 626–632.

Thulin, B., McTaggart, D., & Neubuerger, K. T. (1968) Demyelinat-ing leukodystrophy with total cortical cerebellar atrophy. Arch.Neurol. 18, 113–122.

Tohyama, J., Vanier, M. T., Suzuki, K., Ezoe, T., & Matsuda, J. (2000)Paradoxical influence of acid beta-galactosidase gene dosage onphenotype of the twitcher mouse (genetic galactosylceramidasedeficiency). Hum. Mol. Genet. 9, 1699–1707.

Toung, T. J. K., Traystman, R. J., & Hurn, P. D. (1998) Estrogen-mediated neuroprotections after experimental stroke in male rats.Stroke 29, 1666–1670.

Wenger, D. A. (1997) Krabbe disease (globoid cell leukodystrophy).In The Molecular and Genetic Basis of Neurological Disease (R. N.Rosenberg, S. B. Prusiner, S. DiMauro, & R. L. Barchi, Eds.), pp.421–431. Butterworth–Heinemann, Boston.

Wenger, D. A., Rafi, M. A., Luzi, P., Datto, J., & Costantino-Cecca-rini, E. (2000) Krabbe disease: Genetic aspects and progress to-ward therapy. Mol. Genet. Metab. 70, 1–9.

Wenger, D. A., Suzuki, K., Suzuki, Y., & Suzuki, K. (2001) Galacto-sylceramide lipidosis. Globoid cell leukodystrophy (Krabbe dis-ease). In The Metabolic and Molecular Basis of Inherited Disease (C. R.Scriver, A. L. Beaudet, W. S. Sly, & D. Valle, Eds.), pp. 3699–3694.McGraw–Hill, New York.

Whitefield, P. D., Sharp, P. C., Taylor, R., & Meikle, P (2001) Quan-tification of galatosylsphingosine in the twitcher mouse usingelectrospray ionization-tandem mass spectrometry, J. Lipid Res.42, 2092–2095.

Yang, Q., Hashizume, Y., Yoshida, M., Wang, Y., Goto, Y., Mitsuma,N., Ishikawa, K., & Mizusawa, H. (2000) Morphological Purkinjecell changes in spinocerebellar ataxia type 6. Acta Neuropathol.(Berl). 100, 371–376.



delayed clinical and pathological signs in twitcher (globoid cell leukodystrophy) mice on a c57bl/6...

Documents