Determination of the amounts and oxidation states of hemoglobins M Boston and M Saskatoon in single erythrocytes by infrared microspectroscopy.

The reduced abnormal subunits of two M-type hemoglobins, Boston (His alpha 58-->Tyr) and Saskatoon (His beta 63-->Tyr), have been determined in the presence of normal human hemoglobin A by measurement of C-O stretch bands in infrared spectra of carbon monoxide complexes. Use of an infrared microscope coupled to a Fourier transform infrared spectrometer of high sensitivity permitted measurements to be made on as small a hemoglobin mixture as is contained in a single erythrocyte. The abnormal subunits of both Hbs M exhibit bands near 1970 cm-1 compared with bands near 1951 cm-1 for the normal subunits. The increase in 1970 cm-1 band intensity upon erythrocyte reduction with dithionite provided a measure of the extent of abnormal subunit oxidation; in cell suspensions about 60% of the abnormal subunits of Hb M Boston and 80% for Hb M Saskatoon remained reduced. The amount of Hb present as abnormal Hb averaged about 25% for Hb M Boston cells and about 50% for Hb M Saskatoon cells. However, the ratio of Hb M to Hb A in individual cells varied markedly, with the ratio expected to decrease as the cell ages. These results demonstrate the unique utility of infrared microspectroscopy for the study of differences in abnormal Hb status among individual erythrocytes.

Red blood cells of individuals with hemoglobin M disease contain mutant Hb molecules in which the ability of abnormal subunits to transport 0, is compromised since they are unstable and unable to be maintained as a reduced species under physiological conditions (1)(2)(3). Clinically, the presence of Hb M is accompanied by chronic cyanosis caused by methemoglobinemia (4). Five M-type Hb variants have been identified, and two of them, namely Hb M Boston (a!2His-58 "p,) and Hb M Saskatoon (a!2p2His-63 %), have the distal histidine replaced by a tyrosine residue (5). The structural and functional properties of Hb M Boston and Hb M Saskatoon have been extensively studied (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15). The x-ray crystallographic analysis of Hb M Boston in the natural valence hybrid (a2+Mpzd""y) has been carried out by Pulsinelli et al. (16). The ferric iron atoms in the * This work was supported in part by United States Public Health W. S. C.), and by Grants-in-aid 05209210 and 05670116 from the Min-Service Grant HL-15980, by a gift from Apex Bioscience, Inc. (to istry of Education, Science, and Culture (Japan) (to M. N.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  abnormal a! subunits were found bonded to the tyrosines that replaced distal histidines. Both proteins were also studied by using various spectroscopic techniques including infrared (171, circular dichroism (18), nuclear magnetic resonance (191, electron paramagnetic resonance (20,211,and resonance Raman (14,22). A high frequency infrared C-0 stretching band, which arises from the abnormal subunits, was observed near 1970 cm-l in both reduced Hb M Boston and Hb M Saskatoon carbonyls (17,23).
A homozygous Hb M carrier has never been found, presumably because the absence of at least some Hb A is incompatible with fetal survival. The percentage of Hb M molecules in the total Hb molecule population for an individual heterozygous carrier was found to be about 20-30% for Hb M Boston (6,7,24) and 2 0 4 0 % for Hb M Saskatoon (8,21,25,26). However, little is known about the distribution and oxidation state of abnormal Hb molecules within individual erythrocytes. In order to fully understand the genetic expression, biological synthesis, degradation, and oxidation state of mutant Hb molecules in circulating red cells, it is important and highly desirable to study them in a single cell under physiologically relevant conditions. Recent developments in infrared instrumentation, in which a microscope with infrared optics is coupled to a Fourier transform infrared spectrophotometer, have greatly enhanced the sensitivity of infrared spectroscopy and enable us to study the individual cells under physiologically relevant conditions (27). In the present study, we have examined the IR spectra of erythrocytes that contain Hb M in single cells as well as in cell suspensions. Two discrete C-0 stretch bands near 1970 and 1950 cm" from CO bound to reduced heme iron were observed in both Hb M Bostonand Hb M Saskatoon-containing erythrocytes exposed to carbon monoxide. The proportion of functional (Fez+) heme relative to oxidized heme in the abnormal subunits of Hb M molecules in individual erythrocytes varied widely. Hb M Boston erythrocytes averaged only about 25% of total Hb molecules as abnormal Hb, whereas Hb M Saskatoon cells contained about 50% Hb M and 50% Hb A. About 60% of abnormal a subunits of Hb M Boston and about 80% of abnormal p subunits of Hb M Saskatoon were fully reduced, as demonstrated by the ability to bind CO without addition of reductant.
MATERIALS AND METHODS Cell Preparation-Freshly drawn blood from individual donors with Hb M disease was saturated with CO by passing CO gas over the top of a blood sample for 30 min. The cell sample for the single cell infrared (SCIR)' analysis was prepared as described previously (23,27). Hbs M Boston and M Saskatoon were isolated and purified as described previously (28).
Measurement of the CO-IR Spectrum of a Single Erythrocyte Exposed to CO-The CO-IR spectra were measured by placing the cell suspen-The abbreviation used is: SCIR, single cell infrared spectroscopy. sion on the stage of an infrared microscope (IR-PLAN, Spectra-Tech, Inc.) coupled to a Perkin-Elmer model 1800 Fourier transform infrared spectrometer equipped with a Hg/Cd/Te detector. The cell suspension was mntained between two BaF, discs (1 mm thick, 13 mm in diameter). The microscope was equipped with a 15x m a~~~c a t i o n objective and a lox condenser for visual and infrared use. Two variable image masking apertures were used to select and isolate the target cell and to reduce stray light and other spurious signals. The viewing area containing the single cell chosen was about 10 x 10 pm. A similar size of cell-free area adjacent to the target cell was chosen for recording the spectrum of the medium. Each spectrum was obtained as the average of 1000 scans with a 4 or 8 em" resolution and 2 or 4 cm" interval at room temperature. The spectrum of single erythrocyte was obtained by an appropriate digital subtraction of the medium spectrum from the cell plus medium spectrum. Final difference spectra were subjected to a five-point smoothing with a Savitsky-Golay function. 10 erythrocytes from each Hb M group were examined using the infrared microscope.
Measurement of the CO-IR Spectra of Erythrocyte Suspensions Exposed to CO-Infrared spectra of erythrocyte suspensions in Krebs-Ringer phosphate-dext~se buffer (119 m NaCl, 4.7 IriM KCl, 1.2 m M MgSO,, 0.7 m M CaCl,, 3.7 m M NaH,PO,, 11.2 m M NaHPO,, and 5.5 r n M dextrose) at 20 "C were measured in a Beckman FH-01 cell with CaF, windows and 0.05-mm spacer. A 1000-scan interferogram was collected in single beam mode with a 4 cm" resolution and a 2 cm" interval. A reference spectrum was collected under identical conditions with only the medium in the infrared cell. The difference spectrum was obtained by digital subtraction of the reference spectrum from the spectrum of the cell suspension as described previously (23). represented about 15% of total integrated band intensity. The percentage of total Hb that is Hb M Boston can be estimated from the integrated intensity of C-0 stretch bands that remains after subtraction of the C-0 band caused by Hb A under fully reduced conditions (Fig. 2 A ) . The integrated intensity under the 1970 cm-l band of a typical CO-saturated fully reduced cell suspension amounted to 12.5% of the total integrated C-0 band area, which indicates that 25% of the Hb present is Hb M Boston. Since CO molecules bind only to re- The spectra of nearly all single erythrocytes examined by SCIR exhibited two C-0 stretch bands. The integrated intensity under the 1970 cm-I band in cells where this band was most intense represented -30% of total CO-IR band area. For cell suspensions under fully reduced conditions, the 1970 cm" band area consisted of -25% of total C-0 band areas (Fig. 2B), which indicated that Hb M Saskatoon represented -50% of the Hb in the erythrocytes. Comparisons of the CO-IR spectra of cell suspensions saturated with CO as freshly drawn with spectra after sodium dithionite reduction suggested that about 80% of the abnormal /3 subunits of Hb M Saskatoon in fresh suspensions are in the functional reduced form (Fez+) with only 20% in the nonfunctional oxidized form (Fe3+), which corresponds to the finding that -5% of the total Hb is present as metHb.

CO-IR
M. Nagai et al., manuscript in preparation.

DISCUSSION
Infrared spectroscopy has been very useful in studies of the structure and reactions of hemoglobin (23). Earlier studies were limited to the purified proteins (29,30) and cell suspensions (31,32). Recent developments in infrared microspectroscopy make possible the extension of these studies t o the individual erythrocyte (27). In this way information on differences between erythrocytes in areas such as mutant gene expression, abnormal protein distribution over a cell population, the oxidation state of subunits, and the binding of infrared-active ligands such as CO can be obtained.
Distribution of Hb M Molecules among Erythrocytes-In the present study, the erythrocytes from donors with Hb M Boston or Hb M Saskatoon disease have been studied individually as well as collectively by using Fourier transform infrared spectroscopy. For Hb M Boston-containing erythrocytes, an average of about 25% of H b molecules in the cell population are found to be M-type Hb on the basis of the infrared spectrum of COsaturated fully reduced erythrocyte suspension.
The C-O stretch band a t 1970 cm" from the abnormal a subunits of Hb M Boston was observed only in two-fifths of the erythrocytes that were examined by SCIR. However, it is not likely that the remaining cells truly contain no mutant Hb. There are several possible explanations for the absence of an abnormal a-CO band in some of the erythrocytes examined by SCIR. First, the abnormal a subunits of Hb M Boston in those cells may be essentially all in the oxidized form (Fe3+), which is resistant to enzymatic reduction by metHb reductases (28). Second, the 1970 cm" band may be too weak to be clearly detected in view of the limited sensitivity and signal-to-noise ratio of infrared microspectroscopy. Third, the inherent instability of the Hb M may have resulted in modifications in protein structure that prevent CO binding. The variations in band intensity of the 1970 cm" band of single cells may reflect differences in cell age; the abnormal Hb M molecules are expected to be synthesized as reduced species in newly matured erythrocytes. Because of exposure to oxidizing conditions during the erythrocyte life span Hb M molecules will be gradually oxidized and will remain so because of their resistance to enzymic reductions (28). Therefore, it is not surprising that, as found, the most intense 1970 cm" bands observed in some erythrocytes are actually more intense than that observed in a fully reduced cell suspension.
For Hb M Saskatoon-containing erythrocytes, an average of about 50% of the Hb molecules are found to be the M-type Hb.
The C-0 stretch band at 1970 cm" from abnormal p subunits has been observed in almost all of the erythrocytes examined by SCIR. However, the intensities of these bands varied widely among individual cells. The most intense 1970 cm" band observed in a single erythrocyte represented -30% of integrated intensity of the CO-IR bands (suggesting a Hb " 3 A ratio of 3\21, which is much more intense than the -25% of total intensity observed in a reduced cell suspension. It should be noted that the single Hb M Saskatoon-containing erythrocytes were examined at a resolution of 8 cm" rather than the 4 cm" resolution used to examine the cell suspension, thereby making the single cell determination less accurate. However, these findings may indicate that the mutant Hb M Saskatoon gene can be expressed at a higher rate than Hb A in the cells. Oxidation State of H b M Molecules in Circulation-Treatment of a freshly drawn blood sample with CO gas allows CO molecules to replace the 0, ligands from heme iron (Fez+) of hemoglobins and to prevent further autoxidation. Therefore, the oxidation state of Hb M molecules in circulating erythrocytes can be determined from CO-IR spectra. About 5% metHb was detected in both Hb M Boston-and Hb M Saskatooncontaining blood samples, and almost all of the metHb comes from the abnormal subunits. Two major metHb reductase systems have been found in human erythrocytes: NADH-dependent metHb reductase identified as NADH-cytochrome b, reductase (33, 34) and NADPH-dependent metHb reductase identified as NADPH-flavin reductase (35). The most important pathway of metHb reduction utilizes NADH-cytochrome b, reductase for the transfer of electrons from NADH to heme irons (33). The level of metHb in circulating blood of normal adults is maintained under 0.6% of total Hbs by the metHb reductases (4, 36, 37). The reduction of Hb M Boston, Hb M Saskatoon, and other M-type H b s by various metHb reductases purified from human erythrocytes and by chemical reductants has been extensively studied (28,38). It has been found that metHb M Saskatoon can be reduced by NADH-cytochrome b, reductase at a rate comparable with that of metHb A (28). On the other hand, no enzymatic reduction of metHb M Boston was observed in the presence of NADH-cytochrome b, reductase (38). The results of our present study show that about 80% of Hb M Saskatoon molecules in erythrocytes are maintained as reduced species that are able to bind CO ligand, whereas only 60% of Hb M Boston molecules in the cells exist as reduced species. The higher percentage of mutant Hb usually found in the patients with Hb M Saskatoon appears related to the greater ability of the erythrocyte to keep this type of Hb mutation in a stable reduced state.