Lack of allosterically controlled intramolecular transfer of nitric oxide from the heme to cysteine in the β subunit of hemoglobin

In 1996, Stamler's group¹  published a revolutionary work stating that, rather than simply destroying NO activity through the deoxygenation reaction, hemoglobin (Hb) is capable of preserving, transporting, and exporting NO activity. It has been proposed that these novel activities of Hb are mediated by SNO-Hb, which is stabilized in the (high-oxygen affinity) R-state and destabilized in the (low-oxygen affinity) T-state.^1-3  NO binds rapidly to the heme of deoxygenated Hb to form Fe-NO Hb. According to the SNO-Hb hypothesis, when Fe-NO Hb is oxygenated (and undergoes a T to R transition) some of the NO on the heme is transferred to the β-93 cysteine to form SNO-Hb.^2-4  Upon re-deoxygenation, some of the NO (actually NO⁺) on the cysteine is transferred back to heme and some is exported by the red blood cell to effect vasodilation.^2-5 

The amount of heme-bound NO that undergoes allosterically controlled intramolecular transfer is claimed to be dependent on the amount of NO bound per hemoglobin tetramer.^2,3  Previously, a decrease in yield of SNO-Hb upon oxygenation of an initially deoxygenated Fe-NO Hb sample was attributed to loss of allosteric control when the NO/Hb ratio was too high.²  More recently, the SNO-Hb hypothesis has been refined wherein it is stated that only NO bound to hemes of the β subunits (to form β-nitrosyl Hb) can efficiently undergo allosterically controlled intramolecular transfer.^3,6  Due to the relative dissociation rates, α-nitrosyl Hb is more stable than β-nitrosyl Hb.^7-9  In the T-state, a percentage of the α-nitrosyl proximal histidine bonds break forming pentacoordinate α-nitrosyl Hb (as opposed to the hexacoordinate form) that has a characteristic hyperfine splitting signature in electron paramagnetic resonance (EPR).^8,10  The efficiency of transfer from heme to thiol was predicted to be β-nitrosyl > hexacoordinate α-nitrosyl > pentacoordinate α-nitrosyl.⁶  It has recently been suggested that inability of Hb to facilitate allosterically controlled intramolecular transfer contributes to the pathology of certain diseases. The mutation in sickle cell Hb was suggested to impede intramolecular transfer.¹¹  In addition, sustained hypoxia purported to be associated with pulmonary hypertension was suggested to result in the build up of α-nitrosyl species with concomitant loss of β-nitrosyl Hb and SNO-Hb.⁶ 

Working together with other laboratories on oxygen cycling-induced changes in heme nitrosylation/S-nitrosation, we did not find evidence for allosterically controlled intramolecular transfer.¹²  Our observations refuting the idea of allosterically controlled intramolecular transfer were criticized based on there being (at least in some cases) unequal distribution of NO among Hb tetramers and prolonged incubations, which would tend to result in a buildup of α-nitrosyl species.³  The current study was designed to test whether allosterically controlled intramolecular transfer occurs when samples are prepared to maximize the amount of β-nitrosyl Hb and to equally distribute small amounts of NO among Hb tetramers.

Normal adult Hb from red blood cells was obtained as described previously.^13,14  Hb samples were deoxygenated by gentle flow of an inert gas and stirring. All experiments were performed at room temperature. Determination of percentages of oxygenated, deoxygenated, and methemoglobin was performed using near infrared spectroscopy.¹³  SNO-Hb standards for validation studies were prepared as described previously.¹⁵  The CO/CuCl/cysteine (3C) assay was also used for validation studies as described.¹⁶ 

The NO donor ProliNO (Cayman Chemical, Ann Arbor, MI) was prepared in 0.01 M NaOH and a small aliquot was rapidly mixed with the deoxygenated Hb for a final concentration of NO of about 5 μM. The half-life of ProliNO is 2 seconds at 37°C, which allowed NO to be evenly distributed among Hb tetramers in our experiments. After 1 minute, aliquots of the sample were frozen for EPR analysis and prepared for measurements of SNO-Hb using SNO stabilization solution and the tri-iodide chemiluminescence method as described previously.^12,17  EPR using a Bruker EMX 10/12 (Bruker, Billerica, MA) was performed as described previously.¹⁸  Samples were oxygenated by exposure to air and gentle stirring for 5 minutes after which they were assayed for oxygen saturation, Fe-NO Hb, and SNO-Hb. The amount of Fe-NO Hb and its distribution among hexa/pentacoordinate species and subunits was analyzed by fitting to basis spectra and performing a double integral as described previously.¹⁹ 

Although EPR is widely accepted as a sensitive and specific method to measure Fe-NO Hb, there has been some disagreement over the accuracy of the tri-iodide method in determining the concentration of SNO-Hb.²⁰ Figure 1A shows that the tri-iodide method compares well with another recently developed assay.¹⁶  In Figure 1B, we show that the tri-iodide method is capable of accurately measuring SNO-Hb in the range of concentrations and NO/Hb ratios expected in our oxygenation studies as previously demonstrated.¹⁵ 

Figure 2A shows typical EPR spectra from a partially nitrosylated sample before and after oxygenation. The deoxygenated sample (deoxy Hb) presents a triplet hyperfine structure characteristic of pentacoordinate α-nitrosyl Hb. This hyperfine structure disappears upon oxygenation as expected due to a change in the quaternary state of the Hb (T to R). The SNO-Hb hypothesis predicts that, upon the oxygen-induced quaternary change, the total amount of nitrosyl Hb should be substantially reduced. This prediction is not observed.

Table 1 summarizes the characteristics of the partially nitrosylated Hb samples prepared under deoxygenated conditions followed by oxygenation. Because the total amount of NO released is so small (NO/Hb = 0.00085) and the NO donor is mixed before it releases NO, there should be virtually no Hb tetramers with more than one NO bound. As expected by the equal association rate of NO with hemes α and β subunits,²¹  the initial sample contains 50% nitrosylation on the β subunits—the most one can expect for a deoxygenated sample. A small amount of SNO-Hb was detected by chemiluminescence in the initial samples, perhaps due to a small amount of nitrite contamination.²²  Upon oxygenation, contrary to the prediction of the SNO-Hb hypothesis, the concentration of Fe-NO Hb did not decrease. The slight increase observed may have been due to the reaction of deoxygenated Hb with the small amount of nitrite present (the concentration of nitrite was 2.0 ± 1.1 μM before oxygenation and 1.3 ± 0.6 μM afterward).²²  Effective conversion from the T to R quaternary states is expected for the percentages of oxygenation and deoxygenation and low amounts of methemoglobin (MetHb) shown in Table 1. This allosteric transition is confirmed by the disappearance of the pentacoordinate α-nitrosyl species upon oxygenation, which is destabilized in the R-state. Overall, the data shown in Figure 2 and Table 1 show that we did not observe the loss of Fe-NO Hb and gain in SNO-Hb predicted by the SNO-Hb hypothesis.

Figure 2B presents a summary of the measured percentage changes in Fe-NO Hb and SNO-Hb upon oxygenation. In addition, the predicted percentage changes based on an 80% transfer efficiency²  are also shown. It has been almost 10 years since the SNO-Hb hypothesis was originally proposed,¹  yet no conclusive evidence using the widely accepted technique of EPR has been put forth to support the notion of an allosterically controlled intramolecular transfer mechanism. Results have been presented using EPR to support the intramolecular transfer but most of these have included nitrite in the Hb preparations^3,4  and we have previously suggested that the added nitrite was potentially responsible for some of the phenomena attributed to intramolecular transfer.¹²  One set of EPR data has been presented in the absence of nitrite, yet the growth in Fe-NO Hb presented was said to be due to slow loss of SNO-Hb on aging without indication to its relation to allosteric control.³  Other evidence for intramolecular transfer has relied on detection using photolysis chemiluminescence, but this technique reports levels of Fe-NO Hb in human blood that are at least 10 times that measured by EPR.^4,6,23  We conclude that allosterically controlled intramolecular transfer from heme to thiol does not occur under normal conditions and therefore a defect in this processing of NO by Hb does not contribute to pathologic states.

We thank Dr Neil Hogg for critical reading of a draft of this manuscript.