Jonathan Glass

Transferrin-binding and iron-binding proteins of rabbit reticulocyte plasma membranes: Three distinct moieties

1. Transferrin-membrane complexes and iron-binding membrane complexes were solubilized with sodium dodecyl sulfate from the plasma membranes of reticulocytes that had been incubated with (59Fe,125I)-labeled transferrin. Gel filtration of solubilized material demonstrated 125I-labeled transferrin complexed to two moieties, a minor component (Peak I) of apparent molecular weight 435,000 and a major component (Peak II) of apparent molecular weight 200,000. Most of the membrane 59Fe was located in Peak I. 2. Sepharose-bound anti-transferrin was used to purify the 125I-labeled transferrin-membrane complexes. The 59Fe/125I ratio in the transferrin complex purified from Peak I was the same as in the original transferrin and thus contained membrane-bound transferrin to which the 59Fe was still attached. The 59Fe/125I ratio in the purified Peak II transferrin complex was 0.33 times that of the original transferrin, indicating that more than 60% of its 59Fe had been delivered to the reticulocyte. 3. The purified transferrin complexes analyzed by SDS-polyacrylamide gel electrophoresis demonstrated a single band of apparent molecular weight 78,000 both by Coomassie blue stain for protein and by 125I radioactivity. The specific activity of this material was 0.27 and 0.56 times that of the original transferrin for Peak I and Peak II, respectively, indicating that transferrin in Peak I and II was bound to a membrane component with a molecular weight similar to that of transferrin. 4. The isoelectric focusing pattern of the Peak II transferrin complex showed isoelectric points of pH 6.7 and 6.2 compared to pH 5.4 for transferrin. 5. On the basis of these studies we propose that transferrin is first bound to a membrane protein and then delivers iron to a membrane component distinct and separate from the transferrin-binding moiety. Prior to its release, transferrin markedly depleted of iron is still bound to a component in the plasma membrane.

The cross-linking of 125I-labelled transferrin to rabbit reticulocytes

125I-labelled transferrin has been covalently linked to proteins of rabbit reticulocyte membranes using three different cross-linking reagents: (a) bis(methyl) suberimidate (b) 4-methyl mercaptobutirimidate and (c) Cu-o-phenanthroline. Analysis of the products of the cross-linking reactions by SDS polyacrylamide gel electrophoresis revealed three radioactively labelled components. The major component corresponds to non-cross-linked transferrin while a second smaller component represents a complex (125 000 daltons) of membrane protein and transferrin. A third component (210 000 daltons) was observed when 4-methyl mercaptobutirimidate was used as a cross-linking reagent. The possibility that the two higher molecular weight complexes contain the receptor protein for transferrin is discussed.

The stability in various detergents of transferrin-transferrin receptor complexes from reticulocyte plasma membranes

Transferrin-membrane protein complexes were solubilized either with 0.4% sodium dodecyl sulfate (SDS), 1% Triton X-100 or 0.5% sulfobetaine 3-14 from the plasma membranes of rabbit reticulocytes previously labeled with 125I and then incubated with 131-labeled transferrin. When the solubilized membranes were analyzed by gel filtration fractionation, marked variation in the preservation of transferrin-transferrin receptor interaction was noted between the three detergents. After SDS solubilization, more than 80% of the 131I-labeled transferrin remained associated with membrane proteins with apparent molecular weight of the transferrin-receptor complexes of 1400 000 and 240 000. In contrast, after Triton X-100 solubilization only 40% of the transferrin was still complexed to membrane proteins with an apparent molecular weight of the complex of 450 000. Dissociation of transferrin from its receptor was most marked following sulfobetaine solubilization, with less than 30% of the transferrin still complexed. Following gel filtration 131I-labeled transferrin-125I-labeled membrane protein complexes were immunoprecipitated with goat specific anti-rabbit transferrin antibodies. The immunoprecipitates were analyzed under stringent dissociating conditions by two SDS-polyacrylamide gel electrophoretic techniques. In a linear 5-25% polyacrylamide gradient the 125I-labeled receptor obtained after membrane solubilization with all three detergents had an apparent molecular weight of 80 000. In contrast, in a different system using 10% polyacrylamide gel two 125I-labeled receptor components were detected wih apparent molecular weights of 90 000 and 80 000. These results demonstrate that estimates of the molecular weight of the transferrin receptor depended on the conditions of electrophoresis and suggest that the transferrin receptor is partially modified, perhaps by glycosylation.

HETEROGENEITY OF MESSENGER RNA DEFECTS IN THE THALASSEMIA SYNDROMES

The aand p-thalassemia syndromes are characterized by reduced or absent synthesis, respectively, of the aor 8-globin subunits of normal adult hemoglobin (Hb A: a,B,).’ A common feature of these syndromes is reduced or absent messenger RNA coding for the affected g l ~ b i n . ~ ~ Elegant studies, reviewed elsewhere in this Annal, have shown that in some conditions (SF thalassemia, hereditary persistence of fetal hemoglobin, the “classical” Asian forms of a thalassemia, the a thalassemia encountered in Blacks, and a few Indian and Pakistani patients with “classical” high Hb A,+ thalassemia) , absence of mRNA coding for the affected globin is due to partial or complete deletion of the appropriate structural genes. However, the most common forms of /3 thalassemia appear to be associated with reduced mRNA accumulation despite the presence of grossly intact B-globin genes. Moreover the presence of apparently complete gene sequences coding for mRNAs accumulating in reduced amounts can be inferred from studies of globins, globin messenger RNAs, and globin genes of at least some patients with the Mediterranean forms of Hb H disease (a thalassemia), the Hb Lepore syndrome, and the Hb Constant Springs syndrome. Thus, in a substantial portion of the patients with

Targeted Genetic Screen in Amyotrophic Lateral Sclerosis Reveals Novel Genetic Variants with Synergistic Effect on Clinical Phenotype

Amyotrophic lateral sclerosis (ALS) is underpinned by an oligogenic rare variant architecture. Identified genetic variants of ALS include RNA-binding proteins containing prion-like domains (PrLDs). We hypothesized that screening genes encoding additional similar proteins will yield novel genetic causes of ALS. The most common genetic variant of ALS patients is a G4C2-repeat expansion within C9ORF72. We have shown that G4C2-repeat RNA sequesters RNA-binding proteins. A logical consequence of this is that loss-of-function mutations in G4C2-binding partners might contribute to ALS pathogenesis independently of and/or synergistically with C9ORF72 expansions. Targeted sequencing of genomic DNA encoding either RNA-binding proteins or known ALS genes (n = 274 genes) was performed in ALS patients to identify rare deleterious genetic variants and explore genotype-phenotype relationships. Genomic DNA was extracted from 103 ALS patients including 42 familial ALS patients and 61 young-onset (average age of onset 41 years) sporadic ALS patients; patients were chosen to maximize the probability of identifying genetic causes of ALS. Thirteen patients carried a G4C2-repeat expansion of C9ORF72. We identified 42 patients with rare deleterious variants; 6 patients carried more than one variant. Twelve mutations were discovered in known ALS genes which served as a validation of our strategy. Rare deleterious variants in RNA-binding proteins were significantly enriched in ALS patients compared to control frequencies (p = 5.31E-18). Nineteen patients featured at least one variant in a RNA-binding protein containing a PrLD. The number of variants per patient correlated with rate of disease progression (t-test, p = 0.033). We identified eighteen patients with a single variant in a G4C2-repeat binding protein. Patients with a G4C2-binding protein variant in combination with a C9ORF72 expansion had a significantly faster disease course (t-test, p = 0.025). Our data are consistent with an oligogenic model of ALS. We provide evidence for a number of entirely novel genetic variants of ALS caused by mutations in RNA-binding proteins. Moreover we show that these mutations act synergistically with each other and with C9ORF72 expansions to modify the clinical phenotype of ALS. A key finding is that this synergy is present only between functionally interacting variants. This work has significant implications for ALS therapy development.