A hemoglobin from plants homologous to truncated hemoglobins of microorganisms -- Data Supplement - Supplemental Data -- Proceedings of the National Academy of Sciences

Supplemental data for Watts et al. (2001) Proc. Natl. Acad. Sci. USA 98 (18), 10119–10124. (10.1073/pnas.191349198)

Recombinant Hb Expression and Purification.
Flanking BamHI restriction sites and a ribosome-binding site were attached to the ARAth GLB3 coding sequence in a PCR with primers 5'-CGG GAT CCT AAC TAA CTA AAG GAG AAC AAC AAC AAT GCA ATC GCT GCA AGA TAA GGC A-3' and 5'-CGG GAT CCT ATT ATT CTT CTG CTG GTT TAT TGG CT-3' on expressed sequence tag clone N37398 as template. This construct was subcloned into the BamHI site of the pET24(+) (Novagen) expression vector. Preliminary experiments showed that GLB3 was toxic to Escherichia coli, lysing most cells within hours of induction with 1 mM isopropyl b-D-thiogalactoside (IPTG) at 37°C, so a slower induction was used. E. coli strain BL21 (DE3) containing the expression construct was grown at 28°C to an OD₆₀₀ of 0.6 and then recombinant protein expression induced with 300 mM IPTG. The following morning, the bacterial cells were harvested by centrifugation, resuspended in 100 mM Tris·Cl (pH 7.6), sonicated, and cleared by centrifugation. The 30-90% ammonium sulfate fraction was obtained, dialyzed twice into 4 liters of 20 mM Tris·Cl (pH 7.6), and then reduced with sodium dithionite in the presence of CO to form the carbonmonoxy species of the recombinant Hb. Subsequent steps took place at 4°C with the protein being maintained in the carbonmonoxy form to prevent oxidation. Immediately after converting the Hb to the carbonmonoxy species, the extract was loaded onto DEAE-sepharose ion-exchange resin in 20 mM Tris·Cl (pH 7.6), washed with 20 mM Tris·Cl/30 mM NaCl, and then eluted with 50 mM Tris·Cl, pH 7.6/50 mM NaCl. Red fractions were pooled, adjusted to 1 M ammonium sulfate, and loaded onto phenolsepharose hydrophobic interaction resin. The column was washed with 50 mM Tris·Cl, pH 7.6/1 M ammonium sulfate and then eluted with 50 mM Tris·Cl, pH 7.6/500 mM ammonium sulfate. Red fractions were dialyzed into 20 mM sodium phosphate buffer (pH 6.1) and then passed over CM-sepharose resin. Red fractions, containing purified Hb, were concentrated and left at 4°C to reequilibrate to the oxy form. The quaternary structure of purified ferric GLB3 was measured by analytical ultracentrifugation by using samples of 3 and 30 mM concentration. Spectral and Kinetic Analysis.
Absorbance spectra were collected with a Cary50 spectrophotometer from Varian. Oxygenated samples were prepared by reducing the protein with sodium dithionite and then desalting over a G25 column equilibrated in air and 0.1 M phosphate buffer at pH 7.0. Spectra of the deoxygenated (ferrous) Hb were generated by adding sodium dithionite to either a ferric or oxy sample. Carbonmonoxy samples were prepared by exposing deoxy Hb to CO. Kinetics of O₂ and CO binding were measured by laser-flash photolysis and stopped-flow rapid mixing. The flash photolysis apparatus and the procedure for measuring CO binding at different ligand concentrations have been described previously. Time courses for CO rebinding after flash were measured between 10 ns and 0.1 s. These data were concatenated into a single time course, as shown in Fig. 5 B and C. Rate constants for CO binding after rapid mixing were measured by using the procedure described by Trent et al. (1). These time courses were fitted to single exponential decays, and these values are reported at different CO concentrations in Fig. 5A. The method for measuring O₂ binding initiated by flash photolysis has been described previously. To measure O₂ binding by stopped flow, one gas-tight syringe filled with 0.1 M potassium phosphate, pH 7.0, was bubbled with nitrogen and another with 100% O₂. Other O₂ concentrations were generated by mixing 100% O₂ with nitrogen purged buffer in a new syringe. Protein samples were made as follows: 30 ml of a 1-mM solution of GLB3 was titrated to 1 mM with sodium dithionite. The protein was then diluted into a 5-ml gas-tight syringe that had been previously purged with nitrogen. This protein was reacted with various O₂ concentrations, as reported in Fig. 5A. Absorbance changes were positive at 410 nm and negative at 432 nm, as expected for the conversion of GLB3 from the deoxy to the oxy state. Oxygen dissociation and CO dissociation rate constants were measured by using previously described methods. Phylogenetic Analyses.
As the 2-on-2 (or truncated) Hbs analyzed are from a wide range of organisms, we used a number of methods to infer relationships between groups of these sequences. Amino acid sequences were used exclusively, as coding bias differs markedly between some of these organisms. Sequences of 2-on-2 and 3-on-3 Hbs were aligned by eye by using the crystal structures of sperm whale myoglobin, lupin leg Hb (LUPlu GLB2S;II), and Chlamydomonas eugametos and Paramecium caudatum 2-on-2 Hbs as a guide. Sequences used in our analysis are listed in Table 4. An alignment of all the 2-on-2 and 3-on-3 Hb amino acid sequences used in the paper is available below (Table 6). Additional information regarding the three types of methods used is indicated below.

(i) A similarity matrix of Hbs based on amino acid sequence was generated by using the olddistances program in the gcg (version 10.0-UNIX) sequence analysis package (Genetics Computer Group, Madison, WI). The blosum62 matrix was used with a scoring matrix threshold score of 1.0, and the denominator value set as the length of the shortest sequence without gaps. The sequences compared were grouped into the categories GLB3, GLBO, GLBN, or 3-on-3 Hbs and were a subset of the sequences listed for phylogenetic analyses. Fig. 6 shows similarity estimates between selected sequences from all major groups of 2-on-2 Hbs and some 3-on-3 Hbs. A complete similarity matrix for all the 2-on-2 Hbs and some 3-on-3 Hbs is available in Table 7. Fig. 7 is a graphical representation of groups of sequences based on their similarity to the plant 2-on-2 Hb from Arabadopsis thaliana (GLB3). Similarity scores can be used to group the Hbs into four types: the GLB3 sequences have 82% or higher similarity; GLBO sequences are most similar to GLB3 (31-48%); GLBN sequences are less similar to GLB3 (23-35%); and 3-on-3 Hbs are the least similar to GLB3 sequences (>25%).

(ii) In addition to generating a similarity matrix, blocks of the amino acid sequence alignment were compared by eye by using five different shaded alignments generated by the BOXSHADE program (K. Hoffman and M. Baron; http://www.ch.embnet.org/software/BOX_form.html). This analysis revealed a number of blocks of homology and distinct residues that could be used to distinguish the different 2-on-2 Hb groups. A subset of the alignment that consists of some sequences from all the major groups aligned across the E, F, and G helices and their interconnecting loops (based on homology to Paramecium globin) is provided in Fig. 8. The GLB3 and GLBO sequences share blocks of homology or conserved length that are not found in GLBN sequences.

(iii) Cladistic analyses of the alignments were carried out with the paup (version 4.0.0d55 for Unix) program incorporated in the gcg (version 10.0-UNIX) sequence analysis package (Genetics Computer Group). The portion of the data matrix used for phylogeny reconstruction was from residue 33 of the alignment (F in ARAth GLB1; a residue preceding the N terminus of most 2-on-2 Hbs in the alignment; Fig. 10) to residue 215 (the end of the longest bacterial 2-on-2 Hb, that of Deinococcus radiodurans). All trees were generated by using the amino acid sequence alignment and the heuristic search mode. Bootstrap (1,000 replicates) minimum evolution and maximum parsimony trees were generated with randomized input of sequences and random starting seeds. Remaining optional parameters were set according to the default settings. The outgroup used for the analyses presented in Figs. 2 and 9 was the globin domain of the oxygen sensing protein HtB from Halobacterium salinarum (U75436). In further analyses, relationships between the three 2-on-2 Hb groups (GLB3, GLBO, and GLBN) were also tested by repeating the analysis 19 times (100 bootstrap replicates) with different combinations of 3-on-3 Hbs or Hb-like proteins as outgroups (ARAth GLB1 was used alone twice). Table 5 lists the outgroup combinations used and the partition frequencies observed for GLBO/GLB3, GLBN/GLB3, and GLBO/GLBN clades. The clade that has the most support is presented in bold. The GLBO/GLB3 clade has the most support in the majority of cases.

1. Trent, J., Hvitved, A. N. & Hargrove, M. S. (2001) Biochemistry, in press.