Figure 1.
Assay for interactions between αB crystallin and HbA or HbS.
Samples of Hemoglobin A or Hemoglobin S were heated in the presence and absence of αB crystallin and then the interactive aggregates or bound complexes were separated by centrifugation. At low speed (2500 g) only the largest aggregates were in the insoluble (I) fraction, and the Hb bound to αB crystallin remained soluble in the supernatant (soluble) fraction. The soluble fraction was centrifuged again at 14,000 g using a 100 kDa spin-filter to separate the HbA or HbS bound to αB crystallin in the soluble concentrate (Sc) from the soluble HbA or HbS in the filtrate (Sf). With thermal destabilization, aggregation was expected to increase the amount of αB crystallin bound HbA or HbS in the insoluble (I) and soluble concentrate (Sc) fractions and decrease the amount in the soluble filtrate (Sf) fraction. The centrifugation experiments were conducted under conditions of minimal thermal stress to determine the sensitivity of αB crystallin to the earliest stages of protein unfolding and aggregation.
Figure 2.
Minimal differences in the conformation of HbA and HbS.
Top Left: The molecular structure of normal Hemoglobin A (PDB:4HHB). Top Right: The molecular structure of Hemoglobin S (PDB:2HBS), containing the glu6val mutation in the β chain of hemoglobin that causes sickle-cell disease. The α chain of hemoglobin is light grey. The β chain of hemoglobin is dark grey. The heme groups are yellow. The total solvent exposed surface area of the β subunit in Hb A is 23476 Å2 and 23246 Å2 in HbS which is a difference of approximately 1%. The structure of the mutant HbS and wild type HbA are very similar as determined by X-ray diffraction. The two beta-6 residues in hemoglobin make up less than 1% of the total surface area in both proteins and a difference in pI of only 0.11. In the absence of conformational changes, a difference in the surface area and pI appears to account for the difference in the attractive interactions responsible for the self assembly of HbS into filaments. Bottom: The UVCD spectra for both HbA and HbS are characterized by a prominent minimum at 220 nm for alpha helix. The spectra are nearly identical despite the glu6val mutation which results in the self-assembly of HbS fibrils that produce sickling of red blood cells. The substitution of a hydrophobic valine (green) for a charged glutamate (red) at amino acid (β6) had no measurable effect on secondary, tertiary, or quaternary structure.
Figure 3.
Destabilization of HbA or HbS with increasing temperature.
Ellipticity at 220 nm was measured using UVCD every 5 minutes at 37, 50, or 55°C and the ΔEllipticity at 220 nm was recorded as the hemoglobin unfolded over time. The ΔEllipticity for HbA and HbS was similar when recording at 37°C and 55°C. At 50°C, the difference between HbS and HbA was statistically significant, but only after 10 minutes. At 5 minutes, thermal destabilization of HbS or HbA at 37°C, 50°C and 55°C was similar, as measured using ΔEllipticity. The 5 minute time point was chosen to evaluate the sensitivity of αB crystallin to HbA or HbS under thermal stress, prior to measurable unfolding.
Figure 4.
Thermal unfolding and aggregation of HbA or HbS in the presence or absence of αB crystallin.
Aggregation was initiated by heating the sample to 50°C (A) or 55°C (B). A. In the absence of αB crystallin, light scattering was measured as optical density (O.D.) in milliabsorbance units (mAU) and normalized to the maximum observed at 55°C. The light scattering increased progressively with increasing protein unfolding and aggregation at 50°C. There was a measurable inhibition of aggregation of unfolding HbA or HbS in the presence of αB crystallin. B. At 55°C the protective activity of αB crystallin on thermal unfolding and aggregation of HbA and HbS was more obvious. The difference in light scattering for HbA and HbS in the presence or absence of αB crystallin was not statistically significant. No increase in light scattering was observed at 37°C (not shown) which was consistent with the absence of a change in the UVCD (Fig. 3). Aggregation and light scattering of unfolding HbA and HbS were minimal at 5 minutes which was consistent with the conformation measured using UVCD in Fig. 3. At all temperatures, αB crystallin inhibited the aggregation of unfolding HbA and HbS measured using light scattering.
Figure 5.
Interactions between HbA or HbS and αB crystallin increased after only five minutes of heating. A.
SDS-PAGE of insoluble (I), soluble filtrate (Sf), and soluble concentrate (Sc). In the absence of αB crystallin (20 kD), the amounts of insoluble(I) aggregates were barely detectable after only 5 minutes at 37°C, 50°C or 55°C. Nearly all the HbA and HbS (17 kD) were in the soluble fraction (Sf). In the presence of αB crystallin, minimal interactions at 37°C resulted in a small amount of αB+HbA or αB+HbS in the soluble concentrate (Sc). At 55°C the interactions with αB crystallin were stronger and nearly all the HbA or HbS co-sedimented with αB in the soluble concentrate (Sc). B–D. Densitometry quantified the amounts of αB crystallin-bound and unbound HbA or HbS in the SDS-PAGE gels at each temperature. The X-axis shows the sample and the insoluble fraction: I (Insoluble), Sc (soluble-concentrate), or Sf (soluble-filtrate). The Y-axis is the measured percentage of protein in each fraction. B. At 37°C, the soluble unbound HbA and HbS were in the soluble filtrate (Sf) in the absence of αB crystallin. In the presence of αB crystallin, 27% of the HbA and 31% of the HbS was bound to αB crystallin in the soluble concentrate (Sc). The results demonstrated the weak interactions between αB crystallin and HbS or HbA even at 37°C for only 5 minutes. C. At 50°C, all destabilized HbA and HbS were in the soluble filtrate (Sf) in the absence of αB crystallin. In the presence of αB crystallin, the interactions between αB and destabilized HbS or HbA increased. The result was that 54% of the HbS-αB crystallin and 33% of the HbA-αB crystallin were in the soluble concentrate (Sc) at 50°C, an increase of 23% HbS and 6% HbA relative to 37°C after only 5 minutes. D. At 55°C, nearly all the unfolding HbA and HbS were in the soluble filtrate (Sf) in the absence of αB crystallin. In the presence of αB crystallin, the HbA or HbS bound to αB crystallin increased. Relative to 50°C and 37°C HbA-αB crystallin and HbS-αB crystallin were largely in the soluble concentrate (Sc) at 55°C The filtration assay confirmed an increase in interactions between αB crystallin and HbA or HbS with thermal destabilization after only 5 minutes when the differences in conformation were not significant. The interactions were stronger for HbS than HbA.