Caveat to protein replacement therapy for genetic disease. Immunological implications of accurate molecular diagnosis

The impact of the immune system on the safety and efficiency of enzyme replacement therapy in lysosomal storage disorders

In the light of clinical experience in infantile onset Pompe patients, the immunological impact on the tolerability and long-term efficacy of enzyme replacement therapy (ERT) for lysosomal storage disorders has come under renewed scrutiny. This article details the currently proposed immunological mechanisms involved in the development of anti-drug antibodies and the current therapies used in their treatment. Given the current understanding of the adaptive immune response, it focuses particularly on T cell dependent mechanisms and the paradigm of using lymphocytic negative selection as a predictor of antibody formation. This concept originally postulated in the 1970s, stipulated that the genotypically determined lack of production or production of a variant protein determines an individual's lymphocytic repertoire. This in turn is the key factor in determining the potential severity of an individual's immunological response to ERT. It also highlights the need for immunological assay standardization particularly those looking at describing the degree of functional impact, robust biochemical or clinical endpoints and detailed patient subgroup identification if the true evaluations of impact are to be realised.

Immunological evaluation of repeated administration of erythrocyte-entrapped protein to C3H/HeJ mice

The immunogenicity of heterologous protein entrapped in autologous erythrocytes was evaluated in a mouse system. Multiple intravenous administrations of an erythrocyte-entrapped bovine beta-glucuronidase preparation resulted in little, if any, immunological response to the entrapped protein or erythrocyte carrier. Erythrocyte entrapment also protected protein from circulating antibody elicited by prior subcutaneous sensitization with unentrapped enzyme in adjuvant. Thus erythrocyte entrapment may protect potentially immunogenic agents from immunological surveillance.

A serological investigation into the acidic alpha-D-mannosidase in normal Angus cattle and in a calf with mannosidosis

Antiserum was raised in a rabbit against bovine kidney acidic alpha-mannosidase that had been purified 570-fold by affinity chromatography on concanavalin A--Sepharose and Sepharose 4B-xi-aminohexanoylmannosylamine. The antiserum precipitated the acidic but not the neutral alpha-mannosidase in normal calf tissues. Human acidic alpha-mannosidase cross-reacted partially with the antiserum, emphasizing the close structural resemblance between the enzyme in the two species. The residual acidic alpha-mannosidase in the tissues of a calf with mannosidosis was also precipitated by the antiserum, the same volume of antiserum being required to precipitate a unit of alpha-mannosidase activity from the normal and pathological tissues. The concentration of cross-reacting material detected by antibody-consumption experiments in the organs of the calf with mannosidosis appeared to be proportional to the concentration of the residual acidic alpha-mannosidase. It is suggested that the residual acidic alpha-mannosidase in mannosidosis accounts for the cross-reacting material detected and that it is unlikely that enzymically inactive but cross-reacting material is present. The residual acidic alpha-mannosidase could be a decreased concentration of the normal gene product or an altered enzyme with a decreased specific enzymic activity and a correspondingly decreased antigenicity.

Human beta-glucuronidase deficiency mucopolysaccharidosis: identification of cross-reactive antigen in cultured fibroblasts of deficient patients by enzyme immunoassay

An enzyme immunoassay for human beta-glucuronidase was developed to determine the presence or absence of antigenically cross-reactive material (CRM) in patients with beta-glucuronidase deficiency mucopolysaccharidosis. This assay provided a sensitive means of measuring the primary interaction between the enzyme molecule and antibody but required neither pure antigen nor monospecific antiserum, an important consideration, since neither of these was available. Goat antiserum to partially purified human placenta beta-glucuronidase did not recognize differences in normal enzyme from human placenta, liver, fibroblasts, or blood platelets. CRM was identified in fibroblast extracts from all four of the unrelated beta-glucuronidase-deficient patients studied, but titration patterns indicated genetic heterogeneity among these four mutant proteins. Fibroblast enzymes from two obligate heterozygotes were distinguishable immunologically from normal enzyme. The enzyme immunoassay was also used to compare human enzyme with liver enzyme from other mammalian species. CRM was present in liver extracts of all species tested, but the liver enzymes, except for the rabbit, were weakly cross-reactive. We conclude that despite certain limitations, the enzyme immunoassay for human beta-glucuronidase is useful and that all four beta-glucuronidase-deficient patients studied possess CRM.

Increased nerve-growth-factor beta-chain cross-reacting material in familial dysautonomia

To determine whether dysautonomia arises from alteration in nerve-growth factor (NGF), we measured serum levels of NGF subunits in normal and dysautonomic subjects using a biologic assay based on neurite outgrowth from chick ganglions, a binding assay based on displacement of radiolabeled betaNGF from rabbit-ganglion microsomes, and radioimmunoassays of chi, gamma and betaNGF subunits via antiserum to mouse NGF polypeptides. A threefold increase (P less than 0.001) in serum antigen levels of the biologically active subunit (betaNGF) was found for dysautonomic as compared with normal subjects. By all other assays, the groups were alike. The marked discrepancy in betaNGF levels between antigenic and functional (biologic and binding) measurements suggests a qualitative abnormaltiy of betaNGF in dysautonomia. Alternatively, elevation of betaNGF antigen can be regarded as secondary to disease. This alternative seems less likely since we must then suppose that the normalcy of functional assays in spurious.

Sanfilippo disease type B: presence of material cross reacting with antibodies against alpha-N-acetylglucosaminidase

1. alpha-N-Acetylglucosaminidase, the enzyme deficient in Sanfilippo disease type B (mucopolysaccharidosis III B) was purified from normal human urine. An antiserum was raised in rabbits against the purified enzyme. Preincubation of the antiserum with crude alpha-N-acetylglucosaminidase from normal human urine, followed by centrifugation, led to a marked reduction of the enzyme activity in the supernatant. Formation of the antibody-enzyme complex had no influence on the activity. The thermal stability of the enzyme was markedly enhanced by complex formation with the antiserum. 2. In the urine from three patients with Sanfilippo disease type B the presence of cross-reacting material could be demonstrated by incubating the antiserum with alpha-N-acetylglucosaminidase in the presence of Sanfilippo B urine or by pretreatment of the antiserum with Sanfilippo B urine. 3. Immunodiffusion and immunoelectrophoresis of crude normal or Sanfilippo B urine gave rise to up to four precipitation lines, only one of which exhibited alpha-N-acetylglucosaminidase activity in the case of normal urine. Purified alpha-N-acetylglucosaminidase yielded only a single precipitation line. After adsorption with the purified enzyme the antiserum did not cross react with any of the urinary proteins. 4. On a quantitative determination of cross-reacting material using Sepharose immobilized antibodies in the urine from two Sanfilippo B patients the amount of cross-reacting material appeared to be less than one fourth of the amount of alpha-N-acetylglucosaminidase protein in an age-matched control urine. The cross-reacting material present in the urine of Sanfilippo B patients had a significant lower binding affinity for antibodies against alpha-N-acetylglucosaminidase than preparations from normal human urine. Taking into account this lower binding affinity, it can be calculated that the amount of cross-reacting material in the urine of Salfilippo B patients exceeds that of normal controls. 5. It is concluded that Sanfilippo disease type B is due to a mutation of a structural gene coding for alpha-N-acetylglucosaminidase. The mutation affects the catalytical and immunological properties of the enzyme protein.