(Top) Linear diagram of the whole diphtheria toxin (DT) molecule (protease nicked), with an expanded view of the channel-forming T domain. Black regions denote hydrophobic stretches within the T domain. (Bottom) Sequence of the H6-tagged T domain, slightly modified from Senzel et al. 1998 and edited to correct two errors in that sequence.

Free H6 peptide, added to the trans side, blocks T domain channel conductance at negative voltages. Before the start of each record, T domain lacking an amino terminal H6 tag was added to the cis compartment to a concentration of ∼200 (A) or 1 (B) ng/ml. (A) T domain–induced conductance is seen rising at +60 mV; when the voltage was switched to −60 mV, the conductance remained on. When the voltage was switched back to +60 mV, it continued to slowly rise. Approximately 7 min into the ∼15-min break, free H6 peptide was added to the trans solution to a concentration of 200 μg/ml; during the break, T domain–induced conductance continued to rise, reflecting the continual insertion of channels in the membrane. In the presence of trans H6 peptide, conductance turned off rapidly, but incompletely, at −60 mV. When the voltage was switched back to +60 mV, conductance rapidly recovered. (B) In the absence of H6 peptide, a single T domain channel stayed open at +60 and −60 mV, with unresolvably brief flickers to a zero-conductance closed state. During the 4-min break, free H6 peptide was added to the trans solution to a concentration of 100 μg/ml. In the presence of trans H6 peptide, the channel, at +60 mV, behaved the same as before H6 addition; at −60 mV, it flickered rapidly between the open state and a prolonged zero-conductance blocked state. The solutions on both sides of the membrane were 1 M KCl, 2 mM CaCl₂; the cis solution contained 30 mM Mes, pH 5.3, and the trans contained 50 mM HEPES, pH 7.2. The records were filtered at 100 Hz by the chart recorder.

H6 peptide chemically linked at residues 293 or 320 blocks the T domain channel at positive voltages. Before the start of each record, T domain with H6 peptide attached at residue 293 (A) or 320 (B) was added to the cis compartment to a concentration of ∼300 (A) or 400 (B) ng/ml. (A) After addition of the T domain, conductance rose over time to the level seen at the start of the record, with the voltage held at +20 mV. When the voltage was switched to −20 mV, the conductance nearly doubled. Subsequent voltage steps to +30, +40, and +50 mV with intervening pulses to corresponding negative voltages demonstrated two facts. (a) The conductance at each positive test voltage was less than that at the corresponding negative voltage; the magnitude of the rectification increased with larger voltages. (b) Positive 10-mV increments produced disproportionately small current increases. (B) A single channel flickered between the open state and a zero-conductance blocked state at +60 mV, spending almost all its time in the zero-conductance state. When the voltage was switched to −60 mV, the channel became unblocked. Similar behavior was seen at ±80 mV after a 1-min break. (Records similar to those in A and B were obtained with T domain having the H6 peptide attached at residue 376, 291, or 294.) The solutions on both sides of the membrane contained 1 M KCl, 2 mM CaCl₂, 1 mM EDTA; the cis solution contained 30 mM Mes, pH 5.3, and the trans contained 50 mM HEPES, pH 7.2. The records were filtered at 100 Hz by the chart recorder.

Trans streptavidin reversibly increases the conductance and noise induced by T domain channels biotinylated at residue 261. Before the start of the record, T domain biotinylated at residue 261 was added to the cis compartment to a concentration of ∼200 ng/ml. During the entire time, the voltage was held at +30 mV. Conductance increased linearly for several minutes, to ∼3.7 nS (∼90 channels) at the start of the record. At the first arrow, streptavidin was added to the trans solution to a concentration of 25 μg/ml. Subsequently, both the conductance and the noise increased significantly. At the second arrow, TCEP was added to the trans solution to a concentration of 20 mM, to remove the biotin with its attached streptavidin. After a transient increase in noise (caused by stirring) and conductance, the noise decreased to prestreptavidin levels, and the conductance returned to the pattern of linear rise seen before streptavidin addition. (Similar records were obtained with the biotinylated 235 mutant.) Solutions and filtering were as given in Fig. 5.

Trans streptavidin reversibly affects conductance induced by T domain channels biotinylated at residue 347. Before the start of each record, T domain biotinylated at residue 347 was added to the cis compartment to a concentration of ∼400 (A) or 10 (B) ng/ml. (A) Macroscopic record. Conductance increased for several minutes to the level seen at the beginning of the record. (During the entire experiment, the voltage was held at +40 mV, except for occasional pulses to 0 mV.) At the first arrow, streptavidin was added to the trans solution to a concentration of 25 μg/ml. Over the next 2 min, conductance decreased fourfold. During the 9-min break, the pHs of the cis and trans solutions were brought to 6.2 with Mes (final concentrations 45 mM cis, 25 mM trans), compounding the streptavidin-induced decrease in conductance. At the second arrow, TCEP, pH 6.2, was added to the trans solution to a concentration of 50 mM. Over the next 1.5 min, conductance rose back to prestreptavidin levels. (B) Several channels record. Channels inserted for several minutes, to a level of about six channels. (The voltage was held at +60 mV, except for occasional brief pulses to 0 and −60 mV during the breaks.) Before the first break, streptavidin was added to the cis compartment to a concentration of 30 μg/ml to prevent further incorporation of new channels in the membrane. During the first (1.2-min) break, streptavidin was added to the trans compartment to a concentration of 12 μg/ml, without stirring, and one channel closed. At the first arrow, streptavidin was added (with stirring) to the trans compartment to a final concentration of 25 μg/ml. Subsequently, channels closed to the level of a single open channel. During the second (2-min) break, TCEP was added to the trans compartment to a concentration of 47 mM, and immediately channels rapidly opened until the number exceeded the prestreptavidin level, and then reached a steady state. The solutions on both sides of the membrane contained 1 M KCl, 2 mM CaCl₂, 1 mM EDTA; the cis solution contained 5(A) and 30 (B) mM Mes , pH 5.3; the trans contained 5 A) and 50 (B) mM HEPES, pH 7.2. The records were filtered at 10 (A) or 100 (B) Hz by the chart recorder.

Our model of the distribution of T domain residues in the open channel state. Residues whose positions were determined in this study are shown in black with white numbers.

H6 peptide chemically linked at residue 235 blocks the T domain channel at negative voltages; trans nickel inhibits this block. Before the start of each record, T domain with H6 peptide attached at residue 235 was added to the cis compartment to a concentration of ∼240 (A) or 480 (B) ng/ml. (A) T domain–induced conductance is seen rising at +30 mV; when the voltage was switched to −80 mV, the conductance decreased rapidly. When the voltage was switched back to +30 mV, the conductance recovered immediately. At the first arrow, NiSO₄ was added to the trans solution to a concentration of 25 μM. Subsequently, when the voltage was switched to −80 mV, the conductance turned off very slowly. At the second arrow, EDTA was added to the trans solution to a concentration of 1.8 mM. When the voltage was switched to −80 mV, the conductance again decreased rapidly. (B) A single T domain channel stayed open at +30 mV, with unresolvably brief flickers to a zero-conductance closed state, but flickered between the open state and a prolonged zero-conductance blocked state at −60 mV. During the 15-s break, NiSO₄ was added to the trans solution to a concentration of 80 μM. Subsequently, the channel was “unblocked” at −60 mV. During the 2-min break, EDTA was added to the trans solution to a concentration of 2.4 mM, and the channel then resumed flickering to a zero-conductance blocked state at −60 mV. (Records similar to those in A and B were obtained with T domain having the H6 peptide attached at residue 261 and 267.) The solutions on both sides of the membrane were 1 M KCl, 2 mM CaCl₂; the cis solution contained 8 mM Mes, pH 5.3, 1 mM EDTA, and the trans solution contained 50 mM HEPES, pH 7.2, and 5 μM EDTA. The records were filtered at 10 (A) or 100 (B) Hz by the chart recorder.

Free H6 peptide, added to the cis side, blocks T domain channel conductance at positive voltages. Before the start of each record, T domain lacking an amino terminal H6 tag was added to the cis compartment to a concentration of ∼30 (A) or 0.1 (B) ng/ml. (A) T domain–induced conductance before addition of H6 peptide resembled that seen in Fig. 2 A. Approximately 3 min into the ∼23-min break, free H6 peptide was added to the cis solution to a concentration of 100 μg/ml; during the break, T domain–induced conductance continued to rise, reflecting the continual insertion of channels. In the presence of cis H6 peptide, conductance was reduced at +60 mV, compared with that at −60 mV. When the voltage was switched to −60 mV, conductance recovered rapidly; when it was switched back to +60 mV, conductance rapidly decreased. (B) In the absence of H6 peptide, a single T domain channel stayed open at +60 and −60 mV, with unresolvably brief flickers to a zero-conductance closed state. During the 1.5-min break, free H6 peptide was added to the cis solution to a concentration of 100 μg/ml. In the presence of cis H6 peptide, when the voltage was held at +60 mV, the channel flickered rapidly between the open state and a prolonged zero-conductance blocked state. When the voltage was switched to −60 mV after a 1-min break, the channel behaved the same as before H6 peptide addition. Solutions and filtering were as given in Fig. 2.

Cis streptavidin increases the conductance and noise induced by T domain channels biotinylated at residue 291. Before the start of the record, T domain biotinylated at residue 291 was added to the cis compartment to a concentration of ∼100 ng/ml. Channel conductance increased for several minutes. Approximately 2.5 min before the start of the record, HEPES was added to the cis solution (final concentration 70 mM) to bring the cis pH to ∼6.8. During the entire record, the voltage was held at +60 mV. At the arrow, streptavidin was added to the cis compartment to a concentration of 25 μg/ml. Subsequently, both the conductance and the noise increased. (Similar records were obtained with the biotinylated 376 mutant.) Solutions and filtering were as given in Fig. 5.

The nature of the streptavidin-induced noise. Before the start of the record, T domain biotinylated at residue 261 was added to the cis compartment to a concentration of ∼50 ng/ml. During the entire experiment, the voltage was held at +60 mV, except for occasional pulses to 0 mV during the breaks. Conductance increased linearly for several minutes. Before the first arrow, it was 467 pS, corresponding to ∼12 channels. At the first arrow (during a 25-s break), streptavidin was added to the trans compartment to a concentration of 30 μg/ml. Immediately after streptavidin addition, the amount of noise increased significantly, and the conductance increased slowly to ∼1.4 nS, corresponding to 35 channels. Note that the noise represents fluctuations to higher, not lower, conductances. During the second break, lasting 4.5 min, conductance continued to rise slowly, to a level of 1.5 nS, corresponding to ∼37 channels. TCEP, pH 7.1, was then added to the trans compartment to a concentration of 20 mM. Subsequently, the conductance fell to the level seen after the break, and the noise decreased drastically. The connected arrows show points in the record of equal conductance (∼32 channels) before and after TCEP addition; note the difference in noise at these points. The solutions on both sides of the membrane contained 1 M KCl, 2 mM CaCl₂, 1 mM EDTA; the cis solution contained 5 mM Mes, pH 5.3, and the trans contained 5 mM HEPES, pH 7.2. The records were filtered at 100 Hz by the chart recorder.

Double dagger model of Choe et al. 1992. See text for discussion of incompatibilities between this model and ours (Fig. 10).

T domain channels having both an H6 peptide chemically linked at residue 293 and an amino-terminal H6 tag are blocked at positive voltages by the former and at negative voltages by the latter. Before the start of each record, T domain with H6 peptide attached at residue 293, as well as an amino-terminal H6 tag, was added to the cis compartment to a concentration of ∼100 (A) or 1 (B) ng/ml. (A) After addition of T domain, conductance rose over time to the level seen at the start of the record, when the voltage was held at +60 mV. When the voltage was switched to −60 mV, the conductance was initially much larger, and then decreased rapidly to nearly zero. (This was consistent with initial unblocking by the H6 tag at residue 293, followed by amino-terminal H6 tag–mediated blocking.) When the voltage was switched back to +60 mV, the conductance was only a fraction of its expected value, namely the initial conductance at −60 mV before channel turn-off. (This was consistent with very fast blocking by H6 tag linked at residue 293.) (B) A single T domain channel flickered between the open state and a zero-conductance blocked state at +60 mV, spending most of its time in the zero-conductance state. When the voltage was switched to −60 mV, the channel became unblocked for a brief time, and then closed. When the voltage was switched back to +60 mV, the channel reopened immediately and resumed its flickering to the zero-conductance state. Similar behavior was seen at ±40 mV. Solutions and filtering were as given in Fig. 5.