Marian W. Kies

CHEMICAL STUDIES ON AN ENCEPHALITOGENIC PROTEIN FROM GUINEA PIG BRAIN

Many attempts to identify the material responsible for the encephalitogenic activity of mammalian central nervous system (CNS) have been made since the demonstration by Kabat et al. and Morgan2 that experimental allergic encephalomyelitis (EAE) could be induced in experimental animals by a single injection of nervous tissue combined with Freund’s adjuvants. On the basis of its solubility the activity was ascribed to phospholipids by A l ~ o r d , ~ protein containing fractions by Koprowski,‘ a dialyzable constituent of CNS by Hottle et U I . , ~ a “complex cerebral lipoid” by Lumsden,6 and proteolipids by Waksrnan et al.,’ Olitsky and Tal,a and Goldstein et al.” The almost prophetic work of Hottle et aL5 on the dialyzability of the encephalitogenic factor is particularly noteworthy. It was observed that calcium acetate extraction removed the encephalitogenic constituent of the CNS in rabies virus vaccine thus making the vaccine safer for therapeutic use.l0 Subsequently it was reported by the same group that the calcium acetate-soluble encephalitogen was dialyzable. The calcium acetate extract was thought to contain protein, although lack of a positive biuret test led the authors to conclude that the dialyzable material was not a protein. Nevertheless, Waksman in a review” cited this report as the inspiration for further work at NIH on the protein nature of the encephalitogen. Actually we have never attempted to repeat the experiments on the calcium acetate-soluble encephalitogen and must admit now with chagrin that perhaps the authors were prophets without honor in their own country. Suffice it to say that our own philosophy on the question of dialyzability of the encephalitogen is best summed up by the words of Carr,12 “. . . there is no sharp dividing line that separates one group of substances from another on the basis of their permeability through membranes. There is a complete spectrum of s?lutes ranging in size from that of the smallest molecules to that of substances having particle weights of several million. It is also possible to have a complete spectrum of membrane porosities, ranging from membranes which are scarcely permeable to water to membranes which will allow substances as large as viruses to pass through. . , . “In dialysis with conventional membranes the nondiffusible substance has a particle weight of 40,000 or more. If the non-diffusible substance is smaller than this, then special effort must be made in selecting the proper membrane.” After the report of Hottle et aL5 on the dialyzable encephalitogen there were several papers on proteolipids, and then sporadically publications began to appear on the encephalitogenic activity of protein fractions. The first report on a purified protein which had been subjected to quantitative bioassay was that of Roboz et aI.,l3 who found that a collagen-like protein (CLP) from CNS was encephalitogenic. In comparison to the basic encephalitogenic protein, CLP is not an important factor in the etiology of EAE; its isolation nevertheless represented a significant advance in neurochemistry. It was the first protein of the CNS to be characterized by modern biochemical techniques. That this protein had encephalitogenic activity seemed highly significant at the time, since no other substance had been isolated with equivalent specific activity. We now know that the specific activity of CLP was insignificant compared to the specific activity of basic protein fractions isolated

The Guillain-Barré syndrome and multiple sclerosis. In vitro cellular responses to nervous-tissue antigens.

Abstract Blood lymphocytes from 83 subjects were evaluated for the presence of cellular hypersensitivity to peripheral and central-nervous-tissue antigens by means of the in vitro production of macrophage migration inhibitory factor (MIF). In the group of 25 patients with peripheral neuropathies, only lymphocytes from patients with the Guillain-Barre syndrome produced MIF in response to peripheral-nerve antigen. Lymphocytes from five of 15 patients with multiple sclerosis produced MIF when incubated with central-nerve antigen. An unexpected finding was that MIF was produced in response to central-nervous-tissue antigen by lymphocytes from six of nine patients who had had cerebrovascular accidents. Although these results further demonstrate that cellular hypersensitivity to components of nervous tissue is present in some neurologic diseases, the data from studies in patients with cerebrovascular accidents suggest that cellular hypersensitivity can be present as a result of nervous-tissue damage.

The in vitro demyelinating activity of sera from guinea pigs sensitized with whole CNS and with purified encephalitogen

Experimental allergic encephalomyelitis (EAE) was induced in guinea pigs with either homologous whole spinal cord in Freunds complete adjuvant, or purified homologous myelin basic protein (BP) in complete adjuvant. In order to induce high levels of anti-BP-antibody, another group of guinea pigs was hyperimmunized by multiple injections of BP in incomplete adjuvant, followed by challenge with BP in complete adjuvant. The presence of antibodies to BP was determined by a specific binding assay capable of detecting antibodies in all three γ-globulin classes. Sera from animals with EAE resulting from sensitization with whole CNS tissue contained no detectable levels of anti-BP-antibody, but caused demyelination of cultures of mouse cerebellum. Sera from guinea pigs with EAE after sensitization with BP and sera from hyperimmunized animals with no EAE contained detectable levels of anti-BP-antibodies, but failed to cause demyelination in vitro. This complete dissociation of EAE, antibodies to BP, and demyelinating antibodies indicates that antibodies induced by the encephalitogenic BP are not the serum factors responsible for in vitro demyelinating activity, and that the demyelinating antibodies must be evoked by some antigen in whole spinal cord other than encephalitogenic myelin BP.

Differences between the two myelin basic proteins of the rat central nervous system. A deletion in the smaller protein.

Abstract Myelin of the rat central nervous system contains two highly basic proteins which differ in molecular size, amino acid composition, and encephalitogenic activity. The larger rat protein is very similar to the myelin basic proteins of beef and human in total polypeptide chain length, amino acid composition, encephalitogenic activity, and length of the polypeptide chain between the two methionyl residues. The length of polypeptide chain between the two methionyl residues of the smaller rat protein is considerably less than the corresponding segment of the larger. Both proteins contain 1 mole of tryptophan per mole of protein. The difference in amino acid compositions of the two rat proteins, together with the amino acid compositions of the tryptic peptides present in the larger rat protein but missing in the smaller indicate a deletion in the smaller protein corresponding to bovine and human residues 117–156 or 118–157. The new tryptophan-containing peptide created by the deletion has the composition (Phe, Ser, Trp, Gly 2 ) Arg. This deletion removes a major part of the peptide reported to be encephalitogenic in the guinea pig. Loss of the Gln-Lys portion of this latter peptide explains our observation that the smaller protein is much less encephalitogenic in the guinea pig than the larger.

Myelin basic proteins of the rat central nervous system Purification, encephalitogenic properties, and amino acid compositions

Abstract Rat myelin basic protein isolated from chloroform-methanol-pretreated spinal cords has been resolved into two components by electrophoresis in polyacrylamide gels at acid pH. The two components were separated by repeated fractionation on a Sephadex G-100 column and characterized by their ability to induce experimental allergic encephalomyelitis in guinea pigs and by their amino acid compositions. The larger component, which has the lower electrophoretic mobility, was found to be comparable to central nervous system myelin basic proteins of other mammalian species in encephalitogenic activity and amino acid composition. The smaller component, which comprised approximately three-fourths of the myelin basic protein extracted, was much less encephalitogenic than the larger, and its amino acid composition was somewhat different. Per 100 moles of total residues the smaller component is considerably richer in arginine and poorer in lysine than the larger component; however, both components contain approx. 23 moles of total basic residues and 14 moles of total acidic residues (including amides) per 100 moles of total residues. Per mole of protein the smaller component appears to have significantly fewer lysyl, glutamyl and/or glutaminyl, glycyl, alanyl, leucyl, tyrosyl, and phenylalanyl residues. It is suggested that the two components may be the products of two structurally related nonallelic genes; however, the possibility that the smaller component arise from the larger as the result of the action of a species-specific proteolytic enzyme system cannot be definitely excluded.

CORRELATION OF EXPERIMENTAL ALLERGIC ENCEPHALOMYELITIS WITH DELAYED-TYPE SKIN SENSITIVITY TO SPECIFIC HOMOLOGOUS ENCEPHALITOGEN.

EAE is readily produced in many mammals and birds by a single injection of central nervous system (CNS) tissue in Freund’s “complete” adjuvants (water-in-oil emulsion with killed mycobacteria). The disease is generally believed to be due to delayed-type hypersensitivity, since it can be passively transferred by sensitized cells,’ but not by serum. Although delayed-type skin reactions to homologous and heterologous neural antigens have been reported in only heterologous reactions have been noted in guinea pigs.““ This discrepancy has been all the more puzzling since the guinea pig is generally regarded as the animal best suited for the demonstration of delayed-type hypersensitivity. Interpretations have, therefore, been cautious: positive reactions may merely have been temporal coincidences (“false witnesses”) of otherwise unrelated phenomena, and negative skin reactions may not have represented the true hypersensitivity reaction which can occur only in the CNS. Correlations between EAE and circulating antibody have been equally ambiguo~s ,~ , 5 , but may be clarified by other papers presented at this rn~nograph .~~’~ One possible explanation for the lack of correlation between EAE and either circulating antibody or delayed-type hypersensitivity may lie in the relative impurity of the encephalitogenic materials which have generally been used as test antigens. The cruder the antigenic mixture, the less correlation can be achieved between EAE and its specific encephalitogen; this is especially true when heterologous antigens are used, since foreign nonencephalitogenic antigens may be stronger than the organ-specific encephalitogenic antigen(s). The availability of a water-soluble, homologous, encephalitogenic protein which previously failed to produce any skin reactivityz9 but which recently has been shown to produce delayed-type skin reactions in guinea pigs3o provided the basis for this reinvestigation of a subject which has provided more controversy than theory suggests should be the case.

Evidence for multiple human T cell recognition sites on myelin basic protein

Myelin basic protein (BP)-specific T cell clones were used to study human T cell recognition sites on the BP molecule. Proliferation assays performed with a panel of xenogeneic BPs of known amino acid sequence and with large peptide fragments of human and guinea pig BPs demonstrated ten different patterns of reactivity. The data provide evidence for at least four different human T cell epitopes within the C-terminal half of the BP molecule, three within the N-terminal half, and three located within the central portion of the molecule. The results indicate that attempts to inhibit anti-BP responses in vivo in an antigen-specific manner will require the suppression of multiple T cell populations.

THE OCCURRENCE OF TWO MYELIN BASIC PROTEINS IN THE CENTRAL NERVOUS SYSTEM OF RODENTS IN THE SUBORDERS MYOMORPHA AND SCIUROMORPHA

Extracts containing myelin basic proteins have been prepared from CNS tissue of representatives of the three suborders of Rodentia—Myomorpha, Hystricomorpha and Sciuro‐morpha. Analyses of the extracts by electrophoresis at low pH showed that one type (L) of myelin basic protein is present in the CNS of all of the rodents examined (rat, mouse, hamster, guinea pig, chinchilla, prairie dog, woodchuck and squirrel). This protein is comparable in molecular size and charge to the CNS myelin basic proteins found in several other mammalian orders. In the CNS of the myomorphs (rat, mouse, hamster) and sciuro‐morphs (prairie dog, woodchuck, squirrel) there is an additional type (S) of myelin basic protein of higher cathodic mobility and smaller molecular size. This additional protein is absent from the CNS of the hystricomorphs (guinea pig, chinchilla). These findings indicate that the presence of two myelin basic proteins originally reported in the CNS of the inbred rat is not an anomaly of inbreeding. These data further suggest that the presence of a single L‐type CNS myelin basic protein might be a general characteristic of hystricomorphs, while the presence of both L‐ and S‐type CNS myelin basic proteins might be a general characteristic of the myomorphs and sciuromorphs.

Myelin Basic Protein: Location of Multiple Independent Antigenic Regions

Immnunization of guinea pigs with homologous myelin basic protein induces antibodies that differ in their ability to bind specific peptide fragments of the protein. Antiserums with differing specificities made it possible to demonstrate at least three mutually exclusive antigenic sites in the protein molecule. One of these sites is located between residues 44 and 89, another between 90 and 116, and the third between 117 and 170.

Peptide specificities of myelin basic protein-reactive human T-cell clones.

Forty myelin basic protein (BP) -reactive T-cell clones were isolated from a patient with multiple sclerosis and used to identify human T-cell recognition sites on the BP molecule. At least three sites have been identified: one in the N-terminal half of the molecule (residues 1–97), one in the C-terminal (residues 98–170), and one which spans residues 97–98. The clones exhibited a marked preference for the C-terminal half of the molecule. No cross-reactivity with measles virus was detected. These clones will be useful for both the further delineation of the human T-cell recognition sites on BP and the generation of anticlonotypic monoclonal antibodies.