Porcine haemin (Cl-Fe(III)PPIX) was from Fluka (98%). All lipids and other reagents were obtained from Sigma-Aldrich (Vorna Valley, South Africa). Solutions of haematin (HO-Fe(III)PPIX) were prepared by dissolving 2 mg of Cl-Fe(III)PPIX in 0.400 ml of 0.1 M NaOH. These solutions were vortexed and sonicated for 3 – 5 min and then were made up to 1 ml with a 1:9 v/v mixture of acetone/methanol. Lipid solutions (3.31 mM) were prepared by dissolving MPG, MSG, DOG, DPG, DLG or NLB in 1:9 v/v acetone/methanol. Citric buffer was prepared at 50 mM concentration from citric acid, pH adjusted to 4.8 with NaOH. Acetate and MES buffers (50 mM, pH 4.8) were prepared from anhydrous sodium acetate and 2-(N-morpholino)ethanesulfonic acid (MES) sodium salt by pH adjustment with perchloric acid.
Kinetics of β-haematin formation were performed following methods previously reported [15, 16]. Briefly, this was as follows: 50 ml of citric buffer (pH 4.8, 50 mM) was pre-incubated for 30 min in a water bath at 37°C in a 9 cm internal diameter Schott-Duran crystallization dish. HO-Fe(III)PPIX solution (0.5 ml) in acetone/methanol (1:9) was mixed with 1.0 ml of 1:9 acetone/methanol (control) or a lipid solution (3.31 mM) in the same solvent mixture. For studying the effect of lipid to Fe(III)PPIX ratio, the lipid solution was diluted with acetone/methanol to the desired ratio before mixing with Fe(III)PPIX. The resulting mixture was carefully layered on the surface of the pre-incubated citric acid using a 1 ml syringe. Incubation was allowed to proceed for varying lengths of time from 1 to 60 min. After the given incubation time, the solution was centrifuged at 10,000 rpm for 15 min. The supernatant, which contained no measureable Fe(III)PPIX concentration, was dicarded and the pellet washed with 1 ml of 5% pyridine prepared by mixing 5 ml pyridine, 50 ml acetone, 10 ml HEPES (0.2 M, pH 7.5) and 35 ml of water. Pyridine forms a low-spin complex with the Fe(III) centre in free Fe(III)PPIX which has an absorption maximum at 405 nm, but does not react with β-haematin under these conditions and is a particularly reliable method of quantitation . The yield of β-haematin formed was obtained by measuring the absorbance of the supernatant from the pyridine washed pellet at 405 nm wavelength using a Varian Cary 100 UV–VIS spectrophotometer following dilution of 0.05 ml of the solution into 1 ml of water. The absorbance provides a measure of the unreacted Fe(III)PPIX remaining at each given time point. Percent conversion to β-haematin was then obtained by difference, using the control as the measure of 0% conversion. Data were fitted to an exponential equation, in keeping with previous studies using lipids [14–16].
To study the effect of lipid:Fe(III)PPIX ratio at fixed lipid concentration, the same procedure described above was adopted, but the concentration of Fe(III)PPIX dissolved in acetone/methanol was increased using a fixed lipid concentration. The effect of buffers on kinetics was investigated by replacing the aqueous citric buffer with the same volume of either actetate buffer or MES (both 50 mM at pH 4.8). To observe the effect of ions and other cellular components on kinetics of β-haematin formation, these substances were included in the buffer. This involved dissolution of NaCl, Na2HPO4, NaHCO3, KCl, adenosine 5’-triphosphate disodium salt or glutathione in the citric buffer solution prior to pH adjustment. Each of these salts was added at both red blood cell cytoplasmic and serum concentrations. Glutathione was prepared in argon purged solutions and the reaction performed under an Ar atmosphere. For Ca2+ and Mg2+, CaCl2·2H2O and MgCl2·6H2O were prepared in the same way, but acetate buffer was used instead of citric buffer because of the potential of citrate to coordinate these alkaline earth metal ions.
Yields of β-haematin were determined as described above for kinetics experiments, except that samples were incubated for 30 min and a reading taken at the end of the experiment to determine the conversion to product. The effect of 2,3-DPG on β-haematin formation was confined to its effect on the overall yield because of the fact that this reagent is not available in quantities needed for kinetics experiments. For this purpose, 2,3-diphospho-D-glyceric acid pentasodium salt was dissolved in the citric buffer prior to pH adjustment.
For Fourier transform infrared (FTIR) spectroscopic and transmission electron microscopic (TEM) characterization of β-haematin, samples were prepared using the method described above and allowed to incubate for 10 min. Material was collected just below the surface of the solution (in the mixing zone formed between the acetone/methanol lipid solution and the citrate buffer). This is clearly visible as a narrow milky emulsion layer at the boundry between the dark Fe(III)PPIX-containing layer and the clear underlying buffer solution.
For FTIR the product was dried over P4O10 in a dessicator. The dried solid was gently crushed to a fine powder using a mortar and pestle. The finely ground powder was mixed with Nujol to form a Nujol mull. The Nujol mull was then used for FTIR spectroscopy to obtain a spectrum of the product using a Perkin Elmer Spectrum 100 FT-IR spectrophotometer.
For TEM, materials collected from the various preparations of β-haematin, from the same boundary layer as described for FTIR experiments, were deposited on a carbon coated grid. The grid was stained for 5 min with uranyl acetate to improve contrast and immediately washed with distilled water. The grid was allowed to dry before viewing using a TECNAI G2 transmission electron microscope.
Confocal microscopy was performed at various time points during the formation of β-haematin. Again, samples were collected as described for FTIR and TEM experiments. For these studies, a 2 μl suspension of the materials was placed on a glass microscope slide. A thin glass coverslip was carefully placed on top of the solution to avoid air bubble formation in the solution between the two slides. The slide was then mounted inverted onto a LSM510-META Zeiss confocal microscope for fluorescence imaging. The excitation wavelength was at 516 nm and emission was imaged between 575 and 630 nm.