Mechanism of Insertion of Diphtheria Toxin: Peptide Entry And
Total Page:16
File Type:pdf, Size:1020Kb
Proc. Natl. Acad. Sci. USA Vol. 81, pp. 3341-3345, June 1984 Biochemistry Mechanism of insertion of diphtheria toxin: Peptide entry and pore size determinations (photoreactive probes/membrane penetration/permeability/liposomes/Sendai virus) LEORA SHALTIEL ZALMAN AND BERNADINE J. WISNIESKI* The Department of Microbiology and The Molecular Biology Institute, University of California, Los Angeles, CA 90024 Communicated by Everett C. Olson, February 21, 1984 ABSTRACT Diphtheria toxin (DTx) is an extremely po- formed cation-selective channels in planar lipid membranes tent inhibitor of protein synthesis. It is secreted as a linear and 18-A pores in liposomes (4, 7). In contrast, Donovan et polypeptide, which is cleaved to produce disulfide-linked A al. (8) found that native DTx forms anion-selective channels and B fragments. Fragment A, the inhibitor of protein synthe- under acidic conditions and concluded these pores were too sis, requires fragment B, the recognition subunit, for entry small (5 A) to allow A to cross the membrane. While the into intact cells. Fragment B has been proposed to form a calculated channel size has been questioned, to our knowl- transmembrane channel through which A gains access to the edge no other pore size determinations have been performed cytosol. If it were demonstrated that the B subunit had an ex- with native DTx. clusive association with membrane lipid acyl chains, this might Because direct information about the nature of toxin- indicate that A is secluded in a proteinaceous B channel. How- membrane interactions is critical to solving the entry path- ever, our results from intramembranous photolabeling studies way, we used a photoreactive glycolipid probe (9, 10) to as- show that both subunits of DTx enter the hydrocarbon domain certain which of the two domains of the toxin penetrates the of the bilayer. Toxin cleavage is not required for penetration. membrane bilayer. These experiments were carried out with Decreasing pH leads to increased binding and hence indirectly simple biological and artificial targets and both cleaved and to increased penetration. Parallel permeability studies indicate uncleaved forms of DTx. The effect of low pH on toxin entry that cleaved DTx does indeed form pores (24 A in diameter) was also monitored in light of previous models of lysosomal and they are larger than those previously reported (5 A) with involvement in the toxin entry process (4, 11-13). Sendai vi- native toxin. The data suggest that these are dimeric struc- rus served as the biological target. Our preliminary studies tures. Cleaved DTx is much more effective than intact DTx at showed that DTx binds to paramyxoviruses (Sendai and pore formation. Thus, we conclude that, while pore formation Newcastle disease virus) and the photolabeling of viral enve- is a feature of toxin-membrane interaction, the pore structure lope proteins is independent of temperature and pH in the does not protect A from contact with lipid side chains and may ranges employed. Parallel studies of the rates of diffusion of in fact consist of both the A and B domains in a dimeric config- different-sized solutes through liposomes containing toxin uration, (AB)2. were performed to assess the relative effectiveness of cleaved and uncleaved toxin at channel formation. Use of a Diphtheria toxin (DTx) is produced by Corynebacterium well-established liposomal swelling assay (14, 15) enabled us diphtheriae that contain the bacteriophage P3 (for review of to delineate the size and structure of the resultant channels. DTx see ref. 1). DTx is secreted as a linear polypeptide of Mr 63,000, which is cleaved by cosecreted proteases to pro- duce disulfide-linked A and B fragments, Mr = 21,000 and MATERIALS AND METHODS 40,000, respectively. Fragment A catalyzes the ADP-ribosyl- ation of elongation factor 2, an essential protein component Materials. Native DTx was generously provided by D. B. of ribosomal protein synthesis. The modified form of elonga- Cawley; a-toxin, by A. Bernheimer. Sendai virus was grown tion factor 2 is inactive; protein synthesis stops and the cell in 10-day-old embryonated chicken eggs and isolated 48 hr dies. later by procedures developed for rapid harvesting of New- While fragment A is an active inhibitor of cell-free protein castle disease virus (16). Sendai virus is not as dense as synthesis, it requires fragment B for entry into intact cells. Newcastle disease virus strain HP16 and hence bands at a Fragment B is the binding-recognition subunit. The receptor lighter density in each of the gradients (16) employed for pu- for B has been identified as a glycoprotein in several cell rification. The photoreactive probe N-[12-(4-azido-2-nitro- types (2). However, toxin has been shown to bind to protein- phenoxy)stearoyl][1-14C]glucosamine (12APS-[14C]GlcN), free membranes under low pH conditions (3) and the func- specific activity 60 ,uCi/,umol (1 Ci = 37 GBq), was synthe- tional insertion of A has been detected in liposome targets sized essentially as described (16), using dicyclohexyl carbo- (J. J. Donovan, personal communication). Conductance diimide instead of N-ethoxycarbonyl-2-ethoxy-1,2-dihydro- studies with a segment of DTx called cross-reacting material quinoline. 45 (CRM45) led to the proposal that B forms a membrane Photolabeling of Native DTx. Probe 12APS-[14C]GlcN channel through which A travels to the cytosol (4, 5). (60,000 cpm) in ethanol was added to the bottom of a test CRM45 is produced by C. diphtheriae cells lysogenic for the tube and partially dried. Sendai virus (50 ,g of protein per mutant phage /345 and unlike DTx has an exposed hydropho- sample) was added in 200 Al of phosphate-buffered saline/ bic domain on a shortened B segment (5). CRM45 is relative- ethylenediaminetetraacetic acid (Pi/NaCl/EDTA; 15 mM ly nontoxic to intact cells, apparently lacking the cell surface sodium phosphate buffer/150 mM NaCl/1 mM EDTA, pH recognition domain (6). Kagan and co-workers observed that 7.0). After 15 min at 37°C, native DTx (40 ,ug of protein) in at low pH the CRM45 B segment as well as CRM45 itself Abbreviations: DTx, diphtheria toxin; CRM45, cross-reacting mate- rial 45, form of diphtheria toxin lacking part of the B domain; The publication costs of this article were defrayed in part by page charge 12APS-[14C]GlcN, N-[12-(4-azido-2-nitrophenoxy)stearoyl][1- payment. This article must therefore be hereby marked "advertisement" 14C]glucosamine. in accordance with 18 U.S.C. §1734 solely to indicate this fact. *To whom reprint requests should be addressed. 3341 Downloaded by guest on September 28, 2021 3342 Biochemistry: Zalman and Wisnieski Proc. NatL Acad Sci. USA 81 (1984) P,/NaCl/EDTA (150 pl) was then added, the pH was adjust- RESULTS ed, and the 370C incubation was continued for 15 min. The The photoreactive probe used to monitor membrane inser- samples were then irradiated at 366 nm for 30 s with a high- tion of DTx was the glycolipid 12APS-[14C]GlcN. Approxi- intensity mercury lamp (75-100 W), layered onto 300-,ul pads mately 95% of this probe spontaneously partitions into the of 15% (wt/vol) Ficoll 400 in P,/NaCI/EDTA, and spun for 1 membrane bilayer within 1 min at 370C. When used with hr at 40,000 rpm (180,000 x g) in a Beckman SW 50.1 rotor membrane systems, it will photolabel only integral mem- equipped with 0.7-ml tube adaptors. The pellets were then brane components (10), as demonstrated here (Fig. 1, lane solubilized in sodium dodecyl sulfate reducing sample buffer 2') by the photolabeling of some Sendai virus proteins (HN, (17) and electrophoresed on 10% polyacrylamide gels (18). F1, and M) but not others (P, NP, F2). Photolabeling of Cleaved DTx. The cleaved form of DTx As shown in Fig. 1, the binding of DTx to Sendai virus was was prepared by incubating whole toxin (usually 0.3-0.6 strongly dependent on the pH. Binding is defined as an asso- mg/ml) with trypsin (Sigma) at 1 jig/ml for 1 hr at room tem- ciation that is maintained after target flotation (vesicles) or perature. Naturally cleaved DTx behaves identically to tryp- sedimentation (viruses) through Ficoll 400 step gradients. At sin-cleaved DTx, and the addition of trypsin inactivators af- neutral pH, the native toxin (lane 1; partially cleaved) bound ter cleavage had no effect on the results. Vesicles were pre- very weakly to the viral membrane (lane 3). As the pH was pared from dimyristoyl phosphatidycholine by reverse phase decreased, more toxin bound and at pH 4 binding was at evaporation (19). Each sample contained 90-100 Ag of phos- least 10-fold higher than at pH 7 (lane 4). Photolabeling from pholipid in 200 pl of Pi/NaCl/EDTA and was incubated with within the viral envelope revealed that the toxin not only probe (60,000 cpm) for 15 min at 370C. DTx (50 tug in 60 ,1 of bound to but inserted into the membrane bilayer at pH 4 Pi/NaCl/EDTA) was then added. After 5 min at 37°C, the (lane 4'). Surprisingly, fragments A and B both inserted into pH of the samples was adjusted, and the incubation was con- the membrane lipid domain as determined by accessibility to tinued for 1 hr. Samples were then irradiated at 366 nm for 30 s with a high-intensity mercury lamp (75-100 W). The sam- ples were mixed with 2 vol of 35% (wt/vol) Ficoll in Ps/Na- Cl/EDTA, overlayed with 16% Ficoll in Pi/NaCl/EDTA and then with Pi/NaCI/EDTA. Samples were centrifuged for 1.5 hr at 40,000 rpm (180,000 x g) in an SW 50.1 rotor equipped with 0.7-ml tube adaptors.