Perspectives On Animal Research

Volume 1, Supplement

An Evaluation of Ten Randomly-Chosen Animal Models of Human Disease

Globoid Cell Leukodystrophy (Mice)

(1,2)

Description of the Model:

Globoid cell leukodystrophy (GLD) is a rare, autosomal recessive neurological disease, which is progressive and rapidly fatal by two years of age. Characteristically, there are numerous multinucleated globoid cells in the white matter, which are macrophages containing accumulated galactocerebroside. In 1980, both Duchen et al. and Kobayashi et al. suggested that the neurologic mutant "twitcher" mouse might serve as a model for human GLD, or Krabbe's disease."(1,2) First discovered in 1976, these mice, like human patients with GLD, have an autosomal recessive deficiency of galactosylceramidase activity.(2) This enzyme catalyzes the hydrolysis of galactosylceramide to cerainide and galactose. Twitcher mice present by the 30th day of life with wasting of the trunk and limbs, generalized tremulousness, and progressive muscular weakness.(1) Some twitcher mice have survived for as long as three months, although it is generally very difficult to maintain them beyond 45 days.(1,3) Histologically, there is severe central nervous system (CNS) degeneration, astrocytic gliosis, and an inifitration of globoid cells.(1) The peripheral nerves, including the cranial nerves, undergo similar degeneration. Suzuki and Suzuki concluded:

Rarity and ethical considerations impose severe restrictions on the nature of studies feasible with human patients. The fundamentally identical genetic cause in the twitcher mutant and the human disease provides a strong support for its use as an authentic animal model. It is less suitable than the canine model for experiments that require large amounts of tissue of anatomically defined areas of the brain. On the other hand, its small size, ease of maintenance, and rapid reproduction are advantages for experiments that require a large number of affected and/or heterozygous carrier animals.(5)

Criterion I: Concordance between the Animal Model and the Human Disease

Clinical Presentation:

The twitcher mouse shares several characteristics with GLD patients. For example, both have massive infiltration of CNS white matter by globoid cells containing abnormal tubular inclusions.(1,6,7) Inclusions are also found in histiocytes and Schwann cells of the peripheral nervous system (PNS).(1) In the CNS, oligodendroglia degenerate while astrocytes markedly proliferate.(6) Myelin and axons in both CNS and PNS degenerate.(6)

However, there are important differences in clinical presentation between human and mouse GLD. Twitcher mice have gradual wasting towards flaccid paralysis. Human patients manifest hypertonicity with hyperactive tendon reflexes, spastic quadraparesis, and clonic seizures in the early stages. Later, human patients develop hypotonicity with diminished deep tendon reflexes and flaccid paralysis.(1,6) These findings reflect the predominant central nervous pathology in human patients, in contrast to the predominantly peripheral nervous pathology in the twitcher mice.(3) Indeed, Ichioka et al. noted, "Unlike the pathology of Krabbe disease, CNS involvement is relatively mild in the Twitcher mouse."(9) In the mice, CNS myelination progresses normally until 20 days, when hypomyelination becomes apparent.(3,10) Shortly thereafter, myelin degeneration ensues, but even then "...the only abnormality by electron microscopic examination was thinning of the myelin, mainly for axons of large diameter."(11) In human patients, however, CNS white matter appears diminished in early infancy.(3,12) Furthermore, "...the secondary abnormalities observed in the brains of human patients as a result of tissue devastation were generally not present in the twitcher brain."(12) Thus, human CNS pathology occurs sooner and is more severe than that observed in the twitcher mouse.

Igisu et al. concluded their study of GLD in the twitcher mouse with the observation that, "...the same genetic defect can result in entirely different consequences in different species. Caution must be exercised even when 'authentic animal models' are utilized for studies of human diseases.(4) For example, in 1983 Takahashi et al. reported ultrastructural investigations on the kidney, lung, liver, spleen, lymph nodes, and adrenal glands of the twitcher mouse. They found:

The characteristic cytoplasmic inclusions identical to those found in the glial and globoid cells in the central nervous systems were observed in the epithelial cells of the thin limbs of the loop of Henle in the kidney and also in the macrophages in the lymph nodes. Such inclusions were not detected in the systemic organs in a case of human GLD. This fact may indicate an important species difference between human and murine forms of GLD, although the genetic enzymatic defect is the same.(7)

In accordance with these ultrastructural observations, the kidney of the twitcher mouse contains a 50-fold elevation of hydroxy-fatty acid containing galactosylceramide and a non-hydroxy-fatty acid containing galactosylceramide level five times the norm.(4,8) Furthermore, galactosylceramide in the liver and lung are also greatly increased in the twitcher mouse.(8) Igisu et al. noted:

This finding is in sharp contrast with those in the enzymatically equivalent human disease, globoid cell leukodystrophy ... in which no specific abnormal accumulation of galactosylceramide occurs despite the same genetic block in the catabolic pathway.(4)

Kobayashi and colleagues later discovered two distinct galactosylceramidase enzymes in human beings and mice, and this explained differences in the pathological findings. Their work will be discussed in greater detail below.

Pathogenesis:

Anatomical and enzymatic differences between mice and people have helped explain many of the observed differences in clinical presentation of GLD. For example, because the mouse brain contains relatively little myelin, Igisu et al. suggested that the milder CNS pathology in the twitcher mouse may reflect the relative amounts of myein in the two species.(12) Alternatively, the earlier CNS demyelination in human patients, which starts in the first week of life, may be due to a relative difference in the onset of myelination. In the human brain, myelination commences at 26 weeks of gestation, while in the mouse brain it begins postnatally at about day 5 to 10.(13) However, these explanations do not account for the more severe pathology in the twitcher mouse PNS. In contrast to the extensive abnormalities in the twitcher mouse, human peripheral nerves show only moderate segmental demyelination.(14) Toyoshima et al. wrote:

Peripheral neuropathy has been well documented in human Krabbe's disease. However, the pathologic process in nerve does not appear to be as severe even in infants as that in the Twitcher, and many axons and myelin sheaths remain intact.(15)

In 1983, Eto et al. suggested that the difference in demyelination between human and mouse GLD was not one of degree but of mechanism. Even though morphological changes are more pronounced in twitcher peripheral nerves than in CNS tissue, myelin lipids are not abnormally reduced in twitcher mouse peripheral nerves. Thus, Eto et al. concluded, "...these biochemical findings in twitcher mice are different from those of human GLD, and different mechanisms must be involved in the demyeinating process."(10) Because paranodal demyelination is more effective in slowing impulse conduction than uniform demyelination, Koles and Rasminsky suggested that paranodal demyelination accounts for nerve conduction block.(16) Toyoshima et al. observed nerve conduction block in the early stages of twitcher mouse PNS dysfunction, and they concluded that paranodal demyelination was responsible.(15) In contrast, Miller et al. reported four patients with Krabbe's disease who had uniform slowing of motor conduction velocity but no evidence of conduction block.(17) Toyoshima et al. summarized, "In human GLD, remyelination may be the predominant lesion with little acute paranodal demyelination. This may account for the mild or even absent peripheral nerve signs and symptoms, which may be masked by the predominant CNS dysfunction."(15)

Criterion II: Citations

Duchen et al.(1) and Kobayashi et al.(2) first described the twitcher mouse model in 1980, and both advocated its use for the study of human GLD. We located 65 articles in the Science Citation Index from 1980 to April, 1988, which cited one or both of these articles. (See Appendix A.) Six of these papers were studies that used human tissues as part of the experimental protocol.

Igisu and Suzuki developed a sensitive and specific test to determine the presence of galactosylsphingosine (psychosine) in the brain, and they applied it successfully with human, canine, and murine tissues.(18) This development was not derived from the animal model, but it did demonstrate that a diagnostic procedure could be useful for the study of twitcher mice as well as patients with GLD.

The research of Kobayashi et al. on cultured human fibroblasts revealed that there are two distinct enzymes that are involved in the hydrolysis of galactosylceramide.(19) They cited Kobayashi et al.(2) when they noted, "...this enzyme is deficient in globoid cell leukodystrophy (GLD) occurring in humans, dogs, and mice."(19)

Alroy et al. published two papers in 1986 that described lectin histochemistry of GLD in mice, dogs, cats, and human beings.(20,21) They concluded that this technique may become useful in distinguishing between human storage diseases because "...receptors can be stained with specific lectins and could serve as markers to characterize and differentiate among the various glycolipid storage diseases."(20) Interestingly, the staining patterns of GLD in different species, while having some similarities, also have differences. In the other paper, Alroy et al. concluded, "The variable lectin staining revealed by these studies suggests the presence of several oligosaccharides in globoid cells whose nature is species or individually determined."(21) They did not comment on whether this technique would be relevant for the use of the animal model.

Kobayashi et al. identified psychosme in somatic organs of a patient with GLD and in twitcher mice.(22) They concluded, "These findings indicate that GLD is a generalized galactosylsphingosine storage disease."(22) Once again, human tissue demonstrated the pathology in human patients. Similar lesions in mice confirmed that the murine model resembles the human disease, but the authors did not report a clinical insight derived from the animal model.

Konola et al. studied in vitro hybrids of twicher mouse skin cells and human fibroblasts. They suggested, "These results indicate that twitcher-human somatic cell hybrids will express galactocerebrosidase activity and thus may be useful for determining the human chromosome or chromosomes associated with this expression."(22a) Their approach, which uses a transformed cell line derived from the twitcher mouse, seems to hold some promise. Although it might have been possible to develop this cell line with skin tissue from humans with GLD, it seems reasonable to use the established cell line for further studies.

Criterion III: Historical Impact

Krabbe first characterized globoid cell leukodystrophy in 1916. In 1948, Hallervorden suggested that the globoid cells may contain cerebroside. The chemical (23,24) and histochemical (25) studies of human patients that followed from 1954 to 1963 demonstrated that cerebroside was in fact present in the globoid cells. In 1965-1966, it was shown that a globoid cell condition could be experimentally induced in rats by intracerebral implantation of galactosylceramide. This finding "...further supported the close relationship between the globoid cells and cerebroside"(6) that had already been established in human studies. In 1970, Suzuki identified a deficiency of galactosyltransferase as the underlying genetic defect of the disease.(26) Several months after publication of this paper, Suzuki demonstrated a similar enzymatic defect in a canine model of GLD.(27) Subsequent studies using human tissues revealed enzymatic dysfunction in brain, liver, spleen, kidney, peripheral leukocytes, serum, and in cultured flbroblasts.(28-30)

Several investigators addressed a surprising finding in GLD patients. Despite the deficiency of galactosylceramidase, there was no CNS accumulation of galactosylceramide. In 1985, Kobayashi et al,, using human cultured fibroblasts, revealed the existence of two genetically distinct galactosylceramidase enzymes. Although both galactosylceramidase I and II can hydrolyze galactosylceramide, galactosylceramidase I is deficient in GLD, and only galactosylceramidase I can hydrolyze psychosine. Thus, patients with GLD can catabolize galactosylceramide, but not its metabolic product, psychosine. It was shown previously, using human tissue, that the psychosine that accumulates in the brain is cytotoxic.(31) Thus, Kobayashi explained why galactosylceramide accumulation was absent in human patients despite galactosylceramidase deficiency. Furthermore, Kobayashi et al. used this finding to explain why galactosylceramide accumulates in mice but not in human patients. Both human and murine GLD result from a deficiency of galactosylceramidase I, but the human galactosylceramidase II has much more activity against galactosylceramide than the analogous murine enzyrne.(32)

Developments in the diagnosis of human patients with GLD have come from clinical studies. Early recognition of the condition can be difficult. In general, diagnosis is suspected by clinical presentation and confirmed by biopsy. Because both the clinical presentation and the pathology of GLD differ between people and mice with the disease, the most relevant information about diagnosis has come from clinical studies. An important diagnostic advance was the development of an in utero diagnostic test. Suzuki et al. demonstrated a technique for identifying OLD from human amniotic fluid in 1971.(33)

In 1983, Suzuki and Suzuki noted, "There is no specific treatment for patients with globoid cell leukodystrophy other than supportive care, nor is it expected that any effective specific treatment can be developed in the near future."(6) However, Hoogerbrugge et al. found that, following bone marrow transplantation in twitcher mice:

Histologically, a gradual disappearance of globoid cells ... and the appearance of foamy macrophages capable of metabolizing the storage product were seen in the CNS. By immunohistochernical labeling it was demonstrated that these foamy macrophages were of donor origin.(34)

While this report is encouraging, Suzuki et al., who also studied effects of bone marrow transplantation on CNS histology, concluded, "...the presence of inclusions in oligodendrocytes in the twitcher with BMT, even after 100 days of age indicates that the basic enzymatic defect involving oligodendrocytes could not be completely corrected with BMT."(35) Thus, it is unclear at this time whether bone marrow transplantation will be a useful treatment for human GLD. Among many of the difficulties one might expect in human patients is graft-versus-host disease, which is not as problematic in immunologically homogeneous mice.

Perhaps because of the rarity of GLD in people, we were unable to locate any recent clinical review articles on GLD. There was no reference to the twitcher mouse model in the 1987 edition of Nelson's Textbook of Pediatrics.(36) None of the 11 references in a general review entitled "Lysosomal storage diseases" related to this animal model.(37)

Conclusions:

It appears that the important advances in the understanding of Krabbe's disease have come from investigations using human patients and human tissues, despite the low incidence of GLD in man. Development of diagnostic tests used human tissues, and many of these tests were also applicable to nonhuman GLD. Treatment for human patients remains supportive. Although bone marrow transplantation has been studied in the mouse model and may prove valuable, there remain substantial obstacles to its use in man.

References:

1. Duchen LW, Eicher EM, Jacobs JM, Scaravilli F, Teixeira F: Hereditary leukodystrophy in the mouse: The new mutant twitcher. Brain 1980;103:695-710.

2. Kobayashi T, Yamanake T, Jacobs J, Teixeira F, Suzuki K: The twitcher mouse: An enzymatically authentic model of human globoid cell leukodystrophy (Krabbe’s disease). Brain Res 1980;202:479-483.

3. Suzuki K, Suzuki K: Globoid cell leukodystrophy. Am J Pathol 1983;111:394-397.

4. Igisu H, Takahashi H, Suzuki K, Suzuki K: Abnormal accumulation of galactocerebroside in the kidney of twitcher mouse. Biochem Biophys Res Commun 1983;110:940-944.

5. Suzuki K, Suzuki K: Globoid cell leukodystrophy, Model No. 176. In: Capen CC, Hackel TC, Jones TC, Migaki G (Eds): Handbook: Animal Models of Human Disease, Fasc 12. Washington DC, Registry of Comparative Pathology, Armed Forces Institute of Pathology,1983.

6. Suzuki K, Suzuki Y: Galactosylceramide lipidosis: gioboid cell leukodystrophy (Krabbe's disease). In: Stanbury JB, Wyngaarden JB, Fredrison DS: The Metabolic Basis of Inherited Disease, 5th Ed. New York, McGraw-Hill, 1983 pp 857-880.

7. Takahashi H, Igisu H, Suzuki K, Suzuki K. Murine globoid cell leukodystrophy (the
twitcher mouse): The presence of characteristic inclusions in kidney and lymph node. Am J Pathol 1983;112:147-154.

8. Igisu H, Suzuki K: Glycolipids of the spinal cord, sciatic nerve, and systemic organs of the twitcher mouse. J Neuropathol Exp Neurol 1984;43:22-36.

9. Ichioka T, Kishimoto Y, Brennan S, Santos GW, Yeager AM: hematopoietic cell transplantation in murine globoid cell leukodystrophy (the twitcher mouse): Effects on levels of galactoceramide, psychosine, and galactocerebrosides. Proc NAS 1987;84:4259-4263.

10. Nagara H, Kobayashi T, Suzuki K, Suzuki K: The twitcher mouse: Normal pattern of early myelination in the spinal cord. Brain Res 1982;244:288-294.

11. Eto Y, Umezawa F, Kasai E, Ida I, Maekawa KM: Biochemical studies in mouse Krabbe's disease (twitcher). J Inherit Metabol Dis 1983;6:125-126.

12. Igisu H, Shimomura K, Kishimoto Y, Suzuki K: Lipids of developing brain of twitcher mouse: An authentic murine model of Krabbe's disease. Brain 1983;106:405-417.

13. Matthieu JM: Murine leukodystrophies as tools to study myelinogenesis in normal and pathological conditions. Neuropediat 1984;15(suppl):37-52.

14. Suzuki K: Biochemical pathogenesis of genetic leukodystrophies: comparison of metachromatic leukodystrophy and globoid cell leukodystrophy. Neuropediat 1984;15(suppl): 32-36.

15. Toyoshima E, Yeager Am, Brennan S, Santos GW, Moser HW, Mayer RE: Nerve conduction studies in the twitcher mouse (murine globoid cell leukodystrophy). J Neuro Sci 1986;74:307-318.

16. Koles ZJ, Ranunsky M: A computer simulation of conduction in demyelinated nerve fibers. J Physiol 1972;227:351-364.

17. Miller M, Gutmann L, Lewis RA, Sumner AJ: Acquired versus familial demyelinative neuropathies in children. Muscle Nerve 1985;8:205-210.

18. Igisu H, Suzuki K: Analysis of galactosylsphingosine (psychosine) in the brain. J Lipid Res 1984;25:1000-1006.

19. Kotayashi T, Shinnoh N, Goto I, Kuroiwa Y: Hydrolysis of galactosylceramide is catalyzed by two genetically distinct acid B-galactosidases. J Biol Chem 1985;260:14982-14987.

20. Alroy J, Ucci AA, Goyal V, Woods W: Lectin histochemistry of glycolipid storage
diseases on frozen and paraffin-embedded tissue sections. J Histochem Cytochem 1986;34:501-505.

21. Alroy J, Ucci AA, Goyal V, Aurilio A.: Histochemical similarities between human and animal globoid cells in Krabbe's disease: A lectin study. Acta Neuropathol 1986;71:26-31.

22. Kobayashi T, Shinoda H, Goto I, Yamanaka T, Suzuki Y: Globoid cell leukodystrophy is a generalized galactosylsphingosine (psychosine) storage disease. Biochem Biophys Res Commun 1987;144:41-46.

22a. Konola JT, Lyerla TA, Skiba MC, Raghavan S: Establishment of a galactocerebrosidase-deficient twitcher mouse cell line that expresses galactocerebrosidase activity in hybrids with control human fibroblasts. In Vitro Cell Dev Biol 1988;24:575-580.

23. Austin JH: Studies in globoid (Krabbe) leukodystrophy: Controlled thin layer chromatographic studies of globoid body fractions in seven patients. J Neurochem 1963;10:921-930.

24. Blackwood W, Cumings JN: A histochemical and chemical study of three cases of diffuse cerebral sclerosis. J Neurol Neurosurg Psychiat 1954;17:33-49.

25. Diezel PB: Histochemical investigations of degenerative diffuse sclerosis (leukodystrophy and diffuse sclerosis of the Krabbe type). In: Cumings JN, Lowenthal A (Eds): Cerebral Lipidoses: A Symposium. Springfield, Charles C Thomas, 1957.

26. Suzuki K, Suzuki Y: Globoid cell leukodystrophy (Krabbe's disease): Deficiency galactocerebroside B-galactosidase. Proc NAS 1970;66:302-309.

27. Suzuki Y, Austin J, Armstron D, Suzuki K, Schienker J, Fletcher T: Studies in globoid leukodystrophy: Enzymatic and lipid findings in the canine form. Exp Neurol 1970;29:65-75.

28. Austin J, Suzuki K, Armstrong J, et al.: Studies in globoid (Krabbe) leukodystrophy (GLD) V. Controlled enzyme studies in ten human cases. Arch Neurol 1970;23:502-512.

29. Suzuki K, Suzuki Y, Eto Y: Deficiency of galactocerebroside B-galactosidase in Krabbe's globoid cell leukodystrophy, in, Bersohn J, Grossman HJ (Eds): Lipid Storage Diseases: Enzymatic Defect and Clinical Implications. New York, Academic Press, 1971.

30. Suzuki Y, Suzuki K: Krabbe's globoid cell leukodystrophy: Deficiency of galactocerebrosidase in serum, leukocytes, and fibroblasts. Science 1971;171:73-75.

31. Svennerhohn L: Krabbe disease: a galactosyisphingosine (psychosine) lipidosis. J Lipid Res 1980;21:53-64.

32. Kobayashi T, Shinnoh N, Kuroiwa Y: Metabolism of galactosylceramide in the twitcher mouse, an animal model of human globoid cell leukodystrophy. Biochim Biophys Acta 1986;879:215-220.

33. Suzuki K, Schneider EL, Epstein CJ: In utero diagnosis of globoid cell leukodystrophy (Krabbe's disease). Biochem Biophys Res Commun 1971;43:1363-1367.

34. Hoogerbrugge PM, Suzuki K, Suzuki K, et al.: Donor-derived cells in the central nervous system of twitcher mice after bone marrow transplantation. Science 1988;239:1035-1038.

35. Suzuki K, Hoogerbrugge PM, Poorthuis BJHM, Bekkum DWV, Suzuki K: The twitcher mouse: Central nervous system pathology after bone marrow transplantation. Lab Invest 1988;58:302-309.

36. Behrman RE, Vaughn VC: Nelson's Textbook of Pediatrics, 13th Ed. Philadelphia, WB Saunders, 1987.

37. Glew RH, Basu A, Prence EM: Lysosomal storage diseases. Lab Invest 1985;53:250-269.