Kinetics of hemoglobin-carbon monoxide reactions

Proc. Natl. Acad. Sci. USA

Vol. 74, No. 7, pp. 2620-2623, July 1977 Chemistry

Kinetics of hemoglobin-carbon monoxide reactions measured with a superconducting magnetometer: A new method for fast reactions in solution (magnetic susceptibility/flash photolysis/metalloproteins)

JOHN S. PHILO Department of Physics, Stanford University, Stanford, California 94305

Communicated by William M. Fairbank, April 25,1977

ABSTRACT A new technique for measuring fast reactions in solution has been demonstrated. The changes in magnetic susceptibility during the recombination reaction of human hemoglobin with carbon monoxide after flash photolysis have been measured with a new high-sensitivity instrument using cryogenic technology. The rate constants determined at 200 (pH 7.3) are in excellent agreement with those obtained by photometric techniques [Gray, R. D. (1974) J. Biol. Chem. 249, 2879-2885]. A unique capability of this new method is the determination of the magnetic susceptibilities of short-lived reaction intermediates. The magnetic moment of the intermediate species Hb4(CO)3 was found to be 4.9 ± 0.1 jtB in 0.1 M phosphate buffer by partial photolysis experiments. This value agrees with the predictions of two-state allosteric models of cooperativity in hemoglobin. Possible applications and improvements in this technique are discussed.

a superconducting quantum interference device (SQUID) magnetometer that can sense very small magnetic fields. It is

primarily the very low noise and fast response of the SQUID magnetometer that make kinetic experiments possible. The instrument may be operated in two modes. To measure the total susceptibility of a sample, the change in magnetometer output is recorded as the sample is inserted into the sensing coil. In this mode, the instrument can resolve a change in volume susceptibility* of 9 X 1012, which is equivalent to a change of 0.0001% of the susceptibility of a typical diamagnetic sample (such as a protein in solution). This is an improvement of about two orders of magnitude over vibrating sample, force, or other SQUID systems. The system is calibrated with a sample whose susceptibility is known. In the second operating mode, the sample remains fixed within the sensing coil, and changes in susceptibility due to chemical reactions, temperature changes, etc., are measured. This mode was used for these kinetic experiments. The system response time is 300 us. For kinetic experiments, the susceptometer is characterized by a white noise spectrum above 5 Hz with a rms level of 1.1 X 10-11 (Hz)-1/2, or a concentration of a spin 'k species of 0.70,uM(Hz)-1/2. The present sample Dewar configuration permits sample temperatures in the -5 to 100° range; low-temperature experiments are possible with slight modifications. Sample temperature may be regulated to +0.0010. This instrument will be described in detail elsewhere. Flash Photolysis Experiments. For these experiments, the HbCO solution is placed in a cylindrical quartz bulb (0.30 cm3) which is placed within the sensing coil for the duration of the experiment. A Lucite rod serves as a light pipe to carry the photolytic flash to the sample. A commercial photographic flash unit was used with a flash input energy of 50 J and a pulse duration of 1 ms. Up to 80 mj of visible photons reached the sample. The susceptibility changes after the flash were recorded on a digital signal averager. The interval between flashes was 20 s, and usually 32 or 64 transients were averaged. The system time constant could be varied upward from its lower limit of 300 jss and was normally set at 1 ms. All experiments were done at pH 7.3 and 20' ± 0.1'. Data Analysis. Because the iron in HbCO is diamagnetic and in Hb it is paramagnetic (S=2), the changes in magnetic susceptibility are directly proportional to the change in Hb concentration.t In all these experiments, at equilibrium the Hb is

Magnetic susceptibility measurements have often been used by chemists and biochemists to probe the electronic states of transition metal ions. Such measurements can yield information on the symmetry and strength of ligand fields and the oxidation state of the ion, as well as on the strength of interactions between clusters of ions. Magnetochemical methods have been particularly helpful in studies of metalloproteins and of synthetic analogues of their active sites. The power of many physical techniques has been greatly extended by their use in kinetic studies to measure reaction rates and the properties of intermediate species not present in equilibrium. However, conventional magnetic susceptibility instruments generally lack both the sensitivity and time resolution necessary for measuring fast reactions in solution. In our laboratory, we have been applying superconducting technology to construct a very-high-sensitivity magnetic susceptibility instrument for research in biophysics and chemistry. In addition to higher precision and sensitivity for equilibrium measurements, we hoped through this approach to achieve sufficient time resolution for kinetic measurements. This paper reports measurements of the kinetics of the reaction of human adult hemoglobin with carbon monoxide and a measurement of the magnetic moment of an intermediate in this reaction. MATERIALS AND METHODS Magnetic Susceptibility Instrument. The essential elements of the magnetic susceptometer are indicated in Fig. 1. The samples are placed within a room temperature Dewar flask and are magnetized in the homogeneous field (0.43 T) of a superconducting electromagnet. The sample's magnetization produces changes in the magnetic field at a superconducting sensing coil. These magnetic field changes are transmitted to

Note that SI units are used here; to convert to cgs units, divide by 4ir. t Strictly speaking, the proportionality holds only if the magnetic moment of each heme is independent of the ligation state of the others. This is not always true (see Results), but this is a small effect and may be ignored for most purposes.

*

Abbreviation: SQUID, superconducting quantum interference de-

vice.

2620

Proc. Natl. Acad. Sci. USA 74 (1977)

Chemistry: Philo

2621

AK= 5x101 C,,

VACUUM ROOM TEMPERATURE SPACE

z 040

0 40

80

100

~~~~~~~~~160

100 TIME AFTER FLASH (ims)

FIG. 2. Magnetic susceptibility changes after flash photolysis of HbCO. The scale in SI volume susceptibility units is indicated by the vertical segment. The initial negative offscale change is an experimental artifact (see Results). Total [Hb] = 38.5 MM; [CO] = 130 AM; 0.1 M phosphate, pH 7.3; 20°. Average of 64 transients. System time constant = 1 Ms. FIG. 1. Schematic cross section of magnetic susceptometer.

from the error matrix and the estimated uncertainties Xm and fully saturated with CO and is therefore diamagnetic. The change in susceptibility from its equilibrium value was converted directly to Hb concentration by using the known molar susceptibility difference between Hb and HbCO in the buffer used (1). The rate of recombination of Hb with CO after photolysis was analyzed to find an effective value for e', the rate constant for the forward reaction. Such an analysis is complicated by the dissociation of 20-30% of the Hb tetramers into dimers at the concentrations used in these experiments (2). Further, as a consequence of heme-heme interaction, the apparent value of e' increases with fractional saturation. However, the primary purpose of these experiments was to demonstrate the validity of the new technique. Therefore, for purposes of comparison with the rate constants obtained by spectrophotometric techniques by Gray (2), a constant value for over the final 70% of the (tetramer) reaction has been assumed. Over this range, the fast dimer reaction has already gone essentially to completion. Because the pseudo-first-order criterion [CO] >> [Hb] was not always met in these experiments, the more general second-order formulation was used. Differences between the and f# subunits have been neglected. The rate of HbCO dissociation is very slow and therefore the back reaction may be neglected. The kinetic equations therefore become: [1] [Hbk [Hb]o ff'[CO]eq t} a

exp

[Co]t

[Co]o

in which the subscripts refer to concentrations at times 0 and and at equilibrium. If CO is in great excess, its concentration remains essentially constant, the denominators in Eq. 1 become t

equal and the kinetics become pseudo first order. If we define K(t) to be the change in volume susceptibility from its equilibrium value, then K(t) = Xm[Hb]t. With the fact that [COlt = [Hb]t + [CO°eq and a constant defined as A Xm[CO]eq, Eq. 1

becomes:

K(t)

K(t)

+ A

K(O)

K(0)

+

Jt1Atl

A

exp

[-XM

[21

The values of et' were determined from least-squares fits of the susceptibility data to Eq. 2, et' and K(0) being treated as adjustable parameters. Standard deviations for e' were determined

[CO]eq.

For the partial photolysis experiments the entire reaction record was used, and it was assumed that the dimers and Hb4(CO)3 intermediates react at the same rate. These experiments were conducted at very low [CO]eq (small A), and K(0) and the fraction e'/Xn were treated as adjustable parameters. The dimers were assumed to have the same molar susceptibility as do free chains. Sample Preparation and Reagents. Human Hb A was purified by the toluene procedure (3) from whole blood and was stripped of organic phosphates on a Sephadex G-25 gel filtration column. The fraction of MetHb was initially less than 1%. 2,3-Diphosphoglycerate was obtained from Sigma Chemical Co. as the pentacyclohexammonium salt and was converted to the free acid on a Dowex 50W X 8 (H form) column. 2,2Bis(hydroxymethyl)-2,2',2"-nitrilotriethanol (Bis-Tris) was obtained from Aldrich Chemical Co., purified sodium dithionite from J. T. Baker Chemic4l Co., and CO from Liquid Carbonic Corp. Solutions containing CO were prepared by diluting buffer saturated with CO at atmospheric pressure at 200 with deoxygenated buffers, assuming a CO solubility of 1.0 mM. A few grains of sodium dithionite were added directly to the syringe used to fill the sample bulbs to remove residual dissolved 02Heme concentrations were measured in a Beckman DK2-A spectrophotometer by using the extinction coefficients of Banerjee et al. (4). RESULTS The recombination reaction after photolysis of the HbCO was studied in 0.05 M 2,2-bis(hydroxymethyl)-2,2',2"-nitrilotriethanol/0.1 M NaCl buffer both with and without added 1 mM diphosphoglycerate, as well as in 0.1 M phosphate. The susceptibility changes during a typical experiment are shown in Fig. 2. The vertical segment represents a fractional change of only 0.0055% of the total susceptibility of the sample and is equivalent to a change in lfb concentration of 3.3 1sM. Fig. 2 clearly illustrates the feasibility of experiments at micromolar concentrations with millisecond time resolution when signal averaging is possible, and at somewhat higher concentrations for nonrepetitive experiments. The initial negative offscale change is due to a spurious (and unexpected) signal from the

2622

Chemistry:

Philo

Proc. Natl. Acad. Sci. USA 74 (1977) Table 1. Rate constants for combination of Hb with CO

0.6. In

et (AM-1 s'1) Magnetic Photometrict 20.00 + 0.10 230 - 20 pH 7.3 pH 7.0

A B C

33.6 41.3 38.5

0.28 + 0.01 0.211 + 0.006 0.232 ± 0.004

0.4-

w 0.3-

aIx

Buffer*

Total [Hb], AM

0.2-

o.

45

5

0.100.08-

0. tX 0.04OslI

25

45

65 85 TIME (ms ) -

105

FIG. 3. Pseudo-first-order kinetic plot of data of Fig. 2, normalized to 15 ms after the flash. The solid line represents a second-order rate constant of 0.23 1AM-1 s1.

Lucite light pipe. This rapidly decaying signal is independent of the presence of a sample and is believed to be mechanical in origin. In effect, it increases the dead time of the instrument from 1 to 15 ms. We expect to eliminate this problem in future experiments. Fig. 3 is a pseudo-first-order plot of the normalized kinetics of the experiment in Fig. 2. Experiments were conducted in each buffer system at various CO concentrations and were analyzed to find values for the rate constant, e'. The lowest uncertainties for e' were obtained in experiments at low CO concentrations, in which the estimated ±10% uncertainty in equilibrium [CO] had little effect. The root-mean-square deviation of the data from the fits to Eq. 2 was generally less than 1% of the total susceptibility change. Comparison of Rates from Magnetic and Spectrophotometric Methods. In Table 1, the best values for the rate constant et' from these experiments are listed. Also listed are values obtained by Gray (2) from measurements of optical absorption changes. Although the intent of this comparison was primarily to show that this new magnetic technique yields kinetic data of quality comparable to that of optical techniques, the excellent numerical agreement for the rate constants is very encouraging. In fact, because of the slightly different temperatures and pH values, the rates would not be expected to agree exactly. On the basis of known values for the heat of activation (5), Bohr effect (6), and pH dependence of the dissociation rate (7), the rates for our experimental conditions should be about 5% smaller than those from Gray's data. However, this difference is much smaller than the uncertainty arising from the +20 temperature uncertainty in Gray's experiments. Magnetic Moment of Hb4(CO)3. An important feature of kinetic studies is the possibility of measuring properties of intermediate species not present in equilibrium. To demonstrate this capability, the magnetic moment of Hb4(CO)3 was measured. As a result of the cooperative interactions in the Hb tetramer, the affinity for the fourth ligand is very high, and very little Hb4(CO)3 exists at equilibrium in partially saturated solutions. Partial photolysis experiments in 0.1 M phosphate were conducted in which only 7.0% of the CO was removed. The predominant species produced is Hb4(CO)3, and the recombination occurs very rapidly with a rate constant corresponding to e'4 in the Adair formulation (8). The reaction records were analyzed to determine the molar susceptibility (and therefore the magnetic moment) of Hb4(CO)3. Technical difficulties during these first experiments prevented measurements over a sufficient range of CO concentration to permit independent

0.29 0.01 0.21 0.00 0.23 + 0.00

* Buffer: A, 0.05 M 2,2-bis(hydroxymethyl)-2,2,',2'-nitrilotriethanol/0.1 M NaCl; B, A + 1 mM diphosphoglycerate; C, 0.1 M phosphate. t Data from Gray (2); [Hb] = 10.2 ,M.

determination of both e'4 and the magnetic moment. However, with the value e'4 = 7.0 ,M-M so1 determined by Antonini and Gibson (9), a moment of 4.9 I 0.1 MB was found.* This value for the magnetic moment of Hb4(CO)3 is relevant to an understanding of cooperativity in Hb. The magnetic moment per heme of Hb4 is 5.3 AB under these conditions (1). The lower moment for Hb4(CO)3 shows clearly the effect of cooperative interactions on the electronic state of the iron. According to two-state allosteric models of Hb, by the time three ligands are bound the quaternary structure has shifted to the R (high-affinity) conformation with properties essentially identical to those of isolated a and (3 chains (10). Because the magnetic moment of a or # chains is 4.9 MB (1), these new magnetokinetic data support such models. CONCLUSIONS AND DISCUSSION The present study has established the capability of this new method to determine reaction rates and the unique ability to measure the magnetic susceptibilities of short-lived intermediates. It may prove particularly valuable in metalloenzyme research when the electronic states of intermediate species are unknown. It is important to note that this method can give information on paramagnetic species that are not detectable by electron paramagnetic resonance. The range of applications will, of course, depend critically on experimental constraints. The use of this type of system for stopped-flow experiments seems feasible, although the amount of noise introduced by the flow system will have to be determined experimentally. Temperature jump experiments should also be possible. The present instrument in no way represents the best performance attainable. Both microsecond response times and higher sensitivities are possible with existing technology. This work would not have been possible without the continued support of W. M. Fairbank and the earlier work of E. P. Day. The author also wishes to thank C. Yen for her help in purifying the hemoglobin and W. Little for the use of his spectrophotometer. This work was funded in part by the National Science Foundation, Air Force Office of Scientific Research, Office of Naval Research, and the Center for Materials Research, Stanford University. The costs of publication of this article were defrayed in part by the payment of page charges from funds made available to support the research which is the subject of the article. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C. §1734 solely to indicate this fact. 1. Alpert, Y. & Banerjee, R. (1975) Biochim. Biophys. Acta 405, 144-154. t This value is proportional to (e'4)"2.

Chemistry: Philo 2. Gray, R. D. (1974) J. Biol. Chem. 249, 2879-2885. 3. Drabkin, D. L. (1946) J. Biol. Chem. 164, 703-723. 4. Banerjee, R., Alpert, Y., Leterrier, F. & Williams, R. J. P. (1969) Biochemistry 8,2862-2867. 5. Gibson, Q. H. & Roughton, F. J. W. (1957) Proc. R. Soc. London Ser. B 146,206-224. 6. Antonini, E., Wyman, J., Brunori, M., Bucci, E., Fronticelli, C.

Proc. Natl. Acad. Sci. USA 74 (1977)

2623

& Rossi-Fanelli, A. (1963) J. Biol. Chem. 238,2950-2957. 7. Sharma, V. S., Schmidt, M. R. & Ranney, H. M. (1976) J. Biol. Chem. 251, 4267-4272. 8. Adair, G. S. (1925) J. Biol. Chem. 63,529-541. 9. Antonini, E. & Gibson, Q. H. (1960) Biochem. J. 76,534-538. 10. Monod, J., Wyman, J. & Changeux, J. P. (1965) Biochem. J. 76, 534-538.