at low pH: (i) the iron-proximal histidine (F8) bond in ligand- free myoglobin is broken and replaced by a weak-field ligand,. (ih) the distal pocket in MbCO is ...
Proc. Natl. Acad. Sci. USA Vol. 87, pp. 205-209, January 1990 Biophysics a
Metastable intermediates in myoglobin at low pH SANGHWA HAN*, DENIS L. RoUSSEAU*, GIORGIo GIACOMETTIt,
AND
MAURIzIo BRUNORIt
*AT&T Bell Laboratories, Murray Hill, NJ 07974; tDepartment of Biology, University of Padova, Via Trieste 75, 35100 Padova, Italy; and tDepartment of Biochemical Sciences and Consiglio Nazionale delle Ricerce Center of Molecular Biology, University of Rome "La Sapienza," Piazzale Aldo Moro, 5, 00152 Rome, Italy
Communicated by Hans Frauenfelder, October 5, 1989
ABSTRACT Resonance Raman and optical absorption spectra of ligand-free (deoxy) myoglobin and CO-bound myoglobin (MbCO) at pH 2.6 have been measured by using continuous-flow/rapid-mixing techniques. The spectra of deoxy myoglobin at low pH within 6 ms of the pH drop demonstrate that the iron-histidine bond has been ruptured but that the heme is still five-coordinate. Comparison with data from model complexes indicates that a weak-field ligand, such as a water molecule, is coordinated at the fifth position. The Raman spectrum of MbCO at low pH has an Fe-CO stretching mode that is characteristic of a six-coordinate heme with an unhindered Fe-CO moiety. Immediately following the pH drop in this case, there is no indication that the iron-proximal histidine bond is broken. Three different structural changes are detected at low pH: (i) the iron-proximal histidine (F8) bond in ligandfree myoglobin is broken and replaced by a weak-field ligand, (ih) the distal pocket in MbCO is opened, and (iii) protein constraints on the heme group in MbCO are relaxed. Previous conclusions that the kinetics of CO-binding in hemoproteins at low pH is modified by rupturing the iron-proximal histidine bond are supported by these new results which, however, demand a more complete reevaluation of the phenomenon. Since heme proteins carry out diverse biological functions, a great deal of effort has gone into determining the molecular basis for their properties. Myoglobin (Mb) is one of the simplest heme proteins and has been extensively studied in order to understand the factors that control the binding and release of oxygen, its physiological ligand, as well as CO, which is a convenient oxygen substitute for laboratory studies (1, 2). In several reports, the binding of CO to Mb (3, 4) and other heme proteins (5, 6) has been monitored after a rapid drop in pH. For most Mbs and monomeric hemoglobins (Hbs) (3, 4) and for human tetrameric Hb (5), it was found that at low pH (-2.5) a large increase in the CO-binding rate occurs. This was interpreted as a consequence of protonation of the proximal histidine and the associated rupture of the iron-histidine bond, thereby allowing the iron atom to adopt an in-plane position and thus lower the barrier for CO binding. The rupture of the iron-histidine bond was supported by the absorption spectra, which, in the case of the ferrous ligand-free protein at low pH, are similar to those of fourcoordinate model compounds (4). Resonance Raman scattering is a well-suited technique to determine ligand coordination and the structure of the active site in heme proteins. Modes in the high-frequency region are known to be sensitive to axial coordination, porphyrin macrocycle core size, and heme doming (7). Axial ligand modes are often present in the resonance Raman spectrum and may be used to determine the heme coordination and the structure of the ligands (8, 9). In addition, the properties of the iron-histidine bond may be monitored, since the ironhistidine stretching mode is present in five-coordinate ferrous The publication costs of this article were defrayed in part by page charge
payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
205
hemes (10, 11). The previous optical absorption experiments on heme proteins at low pH were done with stopped-flow techniques because most of the proteins are not stable for a long period of time under these conditions (3-6). It is difficult to measure the resonance Raman spectrum with a stoppedflow apparatus because irradiation of a fixed sample with a laser can give photochemical effects such as ligand dissociation (11) and photoreduction (12). Thus, we have developed a rapid-mixing continuous-flow apparatus in which samples are mixed and passed into an observation cell where their resonance Raman spectrum or optical absorption spectrum may be obtained with a dead time of only about 5 ms. In this paper we report the results of experiments on Mb at low pH obtained with this apparatus. Both the ligand-free and the CO-bound ferrous forms of the protein were examined. In ligand-free (deoxy) Mb at low pH, the iron-histidine bond ruptures within 6 ms, but the heme remains five-coordinate by binding to a weak-field ligand, probably water. At low pH the CO-bound protein retains coordination to the histidine and is six-coordinate but has a more open distal pocket. METHODS AND MATERIALS Sperm whale Mb (Sigma) at 1 mM was buffered in 10 mM phosphate (pH 7.0), filtered, and reduced with Na2S204. This stock solution, either oxygen-free or CO-saturated, was mixed with an equal volume of either degassed or COsaturated buffer at 300 mM in the rapid-mixing apparatus to reach the final pH. The pH values we report are those measured directly after mixing. The apparatus used for transient resonance Raman and optical absorption measurements was manufactured by Update Instruments (Madison, WI) and modified to be suitable for the resonance Raman measurements. The samples were mixed in Wiskind four-grid mixers. From the mixer the samples flow through the 2-cm-long Raman capillary cell. This cell is constructed from quartz and has a central hole with a cross section of 0.0625 mm2. For the Raman experiments, the incident laser and the gathered scattered light are at right angles to each other and to the flow in the capillary. Optical absorption measurements were made by using the same cell (path length, 0.25 mm). Several mixing modes were used. In some experiments two samples were combined in a single mixer, and the resulting mixture was passed directly into the optical cell. In other experiments, three samples were combined in a double mixing arrangement by first mixing two samples in one mixer and then passing the resulting mixture into the second mixer, where it was combined with a third sample prior to passage into the sample chamber. The flow rate of the samples in the optical cell was 0.63 ml/s for a single mixing experiment, resulting in a 6-ms dead time. In the Raman measurements, the scattered light was dispersed with a 1.25-m (Spex Industries, Metuchen, NJ) monochromator and detected with an intensified linear photodiode array (Princeton Applied Research), which was interfaced to a personal computer (AT&T). With a 1200 line per mm
206
Biophysics: Han et al.
Proc. Natl. Acad. Sci. USA 87 (1990)
grating, a spectral range of about 600 cm-' is available, Incident laser frequencies of 406.7, 413.1, and 441.6 nm were used in the Raman measurements. The same apparatus was used for the absorption measurements, except that the grating was changed to one with 30 lines per mm, allowing a range of 400 nm. A xenon arc lamp was used as the light source in the optical absorption measurements. There are some small systematic errors in the absorbance values of strong transitions (e.g., the Soret band) due to small light leaks. The only correction applied to the raw data was a subtraction of the electronic background of the detector.
RESULTS In Fig. 1 the absorption spectra of deoxy-Mb and CO-bound Mb (MbCO) at neutral and low pH are presented. In the spectra of the deoxy-Mb, the single peak in the visible range at 556 nm splits into a doublet at 563 and 538 nm, and the peak of the Soret transition shifts toward the violet by about 8 nm upon decreasing the pH from 7.0 to 2.6. These changes are qualitatively consistent with those reported previously for Mb at low pH and thereby serve as a confirmation that the mixing apparatus and conditions used in the experiments reported here result in the same intermediates as those studied previously at this pH (3, 4). In the spectra of MbCO, the Soret maximum is at approximately the same position in the neutral pH and in the low pH samples. In the visible region the a maximum shifts toward the blue by about 7 nm.
1700
1600 1500 1400 RAMAN SHIFT (cn
