SEARCH

SEARCH BY CITATION

Keywords:

  • flavohemoglobin;
  • mycobacteria;
  • electron-transfer;
  • phylogeny;
  • nitric-oxide dioxygenase

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Flavohemoglobins (flavoHbs) constitute a distinct class of chimeric hemoglobins in which a globin domain is coupled with a ferredoxin reductase such as FAD- and NADH-binding modules. Structural features and active site of heme and reductase domains are highly conserved in various flavoHbs. A new class of flavoHbs, displaying crucial differences in functionally conserved regions of heme and reductase domains, have been identified in mycobacteria. Mining of microbial genome data indicated that the occurrence of such flavoHbs might be restricted to a small group of microbes unlike conventional flavoHbs that are widespread among prokaryotes and lower eukaryotes. One of the representative flavoHbs of this class, encoded by Rv0385 gene (MtbFHb) of Mycobacterium tuberculosis, has been cloned, expressed, and characterized. The ferric and deoxy spectra of MtbFHb displayed a hexacoordinate state indicating that its distal site may be occupied by an intrinsic amino acid or an external ligand and it may not be involved in nitric oxide detoxification. Phylogenetic analysis revealed that mycobacterial flavoHbs constitute a separate cluster distinct from conventional flavoHbs and may have novel function(s). © 2011 IUBMB IUBMB Life, 63(5): 337–345, 2011


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Flavohemoglobins (flavoHbs) are monomeric proteins having a classical 3/3 helical globin domain linked with a flavin adenine dinucleotide (FAD) -binding reductase domain that may transfer electrons to the heme iron and allow its rapid re-reduction to maintain it in the ferrous state for the rebinding of oxygen (1–3). Rapid electron transfer via redox domain may allow these two-domain globins to carry out diverse redox reactions that may be vital for their native host. Recent upsurge in availability of genomic data from various organisms has indicated that gene encoding for flavoHb may be widely distributed in unicellular organisms, both prokaryotes and lower eukaryotes (1, 2, 4); however, none has been found so far in higher organisms suggesting that flavoHbs may be specifically relevant to the cellular metabolism of microbes. Physiological function(s) of flavoHbs has been the subject of intense debate over the last decade. Several putative functions have been proposed for flavoHbs (5–8). However, accumulating experimental evidences support its role in protection of microorganisms from the deleterious effects of nitric oxide (NO) by detoxifying it to nitrate via the NO-dioxygenation reaction (1, 2, 7, 8). The hydrophobic distal pocket of flavoHbs has been found much more flexible and expandable than conventional Hbs (3, 9) that allow them to accommodate large apolar molecules including phospholipids and perform other cellular functions. Existence of more than one flavoHb in some bacteria, yeasts, and fungi (10, 11) suggest that these hemoglobins (Hbs) may be involved in diverse cellular functions.

FlavoHbs usually function as an integral part of stress response and virulence of several pathogenic bacteria and fungi by maintaining the cell redox homeostasis at the aerobic/anaerobic interface when cells are exposed to various environmental stresses (7, 12). Although importance of flavoHbs for the parasitic life has been established in several cases (13, 14), none has been reported from pathogenic or nonpathogenic mycobacteria so far. When genomic data of various mycobacterial species was examined, occurrence of multiple flavoHb encoding genes has been detected in many of them. After amino-acid sequence alignment and comparison of mycobacterial flavoHbs, we identified a novel class of flavoHb, exhibiting unconventional heme and reductase domains, in mycobacteria apart from conventional flavoHb. Occurrence of two distinct classes of flavoHbs in mycobacteria is interesting as well as intriguing. Notably, the new class of flavoHb is present in majority of mycobacterial species and it also coexists along with a conventional flavoHb in some of the mycobacterial species. At present, nothing is known about mycobacterial flavoHbs and their role in cellular metabolism. This study provides the first report on a new class of flavoHbs that have been identified in mycobacteria and may not be widespread among microbes unlike conventional flavoHbs. Primary characterization of one of the flavoHbs of this group, present in Mycobacterium tuberculosis, was done, and phylogenetic analysis was conducted to understand functional and evolutionary correlation between two classes of mycobacterial flavoHbs.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Bioinformatics and Phylogenetic Analysis

BLAST searches within the mycobacterial protein data bank were done by using the sequences of different flavoHbs, for example, E. coli, Ralstonia eutropha, and Saccharomyces cerevisiae. Sequences retrieved from these searches were aligned using CLUSAL W sequence alignment tool from EMBL-EBI server. Individual FlavoHb of mycobacteria were identified by paralogs and ortholog searches using E. coli HMP as primary template. Multiple sequence alignment and conservation and divergence within the heme and reductase domains of various mycobacterial flavoHbs were analyzed. A phylogenetic tree was built up using the neighbor-joining method (15) focusing on mycobacterial flavoHbs and their close paralogs/orthologs and conventional flavoHbs of E. coli, Ralstonia eutropha, Bacillus subtilis, and Saccharomyces cerevisiae. The evolutionary distances were computed using the Poisson correction method (16) and are in the units of the number of amino-acid substitutions per site. All positions containing gaps and missing data were eliminated from the dataset (complete deletion option). Phylogenetic analyses were conducted in MEGA4 (17).

Bacterial Strains, Plasmids, Gene Cloning, and Protein Purification

E. coli strains, JM 109 and BL21DE3, were used for the cloning and expression of recombinant proteins. E. coli cells were grown in Luria Bertani or terrific broth (containing 24 g of yeast extract, 12 g of Bacto-Tryptone, 12.3 g of K2HPO4, 2.3 g of KH2PO4) at 37 °C at 180 rpm unless mentioned otherwise. MtbFHb were retrieved from the genomic DNA of M. tuberculosis H37Rv and expressed in E. coli using standard polymerase chain reaction (PCR) techniques. Authenticity of PCR-amplified gene was checked by nucleotide sequencing. Recombinant genes were cloned on pET 15b at NdeI –BamHI sites and expressed in E. coli BL21DE3, cultured in Terrific broth supplemented with δ-aminolevulinic acid (500 μM) and FeCl3 (20 μM) at 37 °C. Recombinant MtbFHb when expressed in E. coli, appeared associated with cell membranes and was retrieved after treating the cell pellet with 1% sarcosyl. Recombinant MtbFHb was purified from the cell lysate using nickel NTA column (Quiagen) following manufacturer's instructions. It resulted in approximately 70–75% pure preparation of protein exhibiting distinct reddish pink color. These fractions were further purified by ion-exchange column (DEAE-Sepharose CL4B, Pharmacia), equilibrated with 10-mM TrisHCl (pH 8.0) and eluted using 0.12 M NaCl gradient. The protein and hemoglobin profile was monitored at 280 and 414 nm. Absorption and CO difference spectra of whole cells or purified protein preparation were recorded using Carry 100 Spectrophotometer as described previously (18). Gel-filtration chromatography was done on Superdex 75 column, equilibrated with 50 mM TrisHCl (pH 7.0) and protein was eluted in 0.2 M NaCl after calibration with standard molecular weight markers (Carbonic anhydrase, 29 kD; Egg albumin, 40 kD and BSA 66 kD).

NO Consumption and Oxygen Uptake

NO consumption activity of MtbFHb expressing cells was measured polarographically in a 2-ml reaction chamber with ISO-NO (World Precision Instruments) essentially as described previously (19). NO uptake was measured from the slope of curving traces recorded in the specified concentration of NO (30 μM) and corrected for background rate of NO decomposition recorded for the control in the absence of any protein. Nitric oxide dioxygenase (NOD) activity of MtbFHb was determined following the published procedure (20). Oxygen consumption by cells or cell extract was measured with an oxygen monitor (Yellow Spring Instrument model 55) at 37 °C in 3 mL of air saturated 0.1 M potassium phosphate buffer (pH 7.5). The electrode was calibrated with air equilibrated water. Cell culture (1 mL) was concentrated, washed with 0.1 M phosphate buffer, and then added quantitatively to 3-mL air-saturated buffer to check the oxygen consumption by monitoring the change in the oxygen concentration of the buffer containing the cells. Three independent experiments were performed for each set.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Two Distinct Classes of FlavoHbs in Mycobacteria

Computational and sequence analysis of available mycobacterial genome and protein data indicated that more than one flavoHb may be present in many species of mycobacteria. An interesting pattern in the occurrence of flavoHb encoding genes was found in both slow- and fast-growing mycobacteria. Opportunistic pathogens and fast-growing mycobacteria displayed presence of two flavoHb encoding genes (Table 1); a conventional flavoHb (type I) similar to other microbial flavoHbs (1, 2) and a new class (type II) showing crucial differences within the functionally conserved regions of heme and reductase domains. Interestingly, the new class of type II flavoHb appeared in majority of mycobacteria including both virulent and avirulent species. Presence of type II flavoHb in microbial genome was found restricted to certain bacterial species, mainly belonging to actinomycetes. Existence of two different flavoHbs in the same organism suggests that they may be playing different functions in cellular metabolism of mycobacteria.

Table 1. Occurrence and distributions of flavoHbs in mycobacteria
Mycobacterium SpeciesNo. of flavoHbsStatusFlavoHbs
Type I (E. coli type)Type II (M. tuberculosis type)
  • a

    Pseudogene.

  • FlavoHbs were identified in different species of mycobacteria using E. coli flavoHb as template followed by its paralogs and orthologs searches. Based on sequence alignment and comparison, two different classes of flavoHbs were identified. Type I designates for flavoHbs similar to flavoHbs of bacteria and yeasts. Type II designates for the new class of flavoHbs present in M. tuberculosis and other mycobacteria.

M. laprae0aVirulent
M. tuberculosis (H37 Rv)1VirulentYes (Rv0385)
M. BovisAF2122/971VirulentYes (BCG_0393)
M. abcesseus1VirulentYes (MAB_4269)
M. paratuberculosis1VirulentYes (MAP_3851)
M. avium1VirulentYes (MAV_4795)
M. marinum1VirulentYes (MMAR_0644)
M. ulcerans1VirulentYes (MUL_128)
M. gilvum2AvirulentYes (Mflv_4884)Yes (Mflv_0255)
M. smegmatis2AvirulentYes (MSMEG_1336)Yes (MSMEG_0719)
M. vanvallani2AvirulentYes (Mvan_1541)Yes (Mvan_639)
M. sp. JLS2AvirulentYes (Mjls_0944)Yes (Mjls_457)
M. sp. MCS2AvirulentYes (Mmcs0916)Yes (Mmcs_470)
M. sp. KMS2AvirulentYes (Mkms_0933)Yes (Mkms_481)

Structural Characteristics of Type II flavoHbs of Mycobacteria: Unusual Features of Heme and Reductase Domains

Sequence comparison of mycobacterial type I and type II flavoHbs with other bacterial flavoHbs (Fig. 1A) indicated that structural features for adopting a three over three globin fold and signature sequences of typical microbial globins, for example, B10-Tyr, CD1-Phe, E7-Gln, F8-His (3, 21), are present in type II flavoHbs, but there are several differences within the functionally conserved regions of their heme and reductase domains. The most notable differences within the globin domain of type II flavoHb is the lack of conserved hydrogen bonding and disruption of tetrainteractions between HisF8-GluH23-TyrG5 and contact between TyrG5 and TyrH12 within the proximal site due to mutation at GluH23 and TyrH12 residues. These observations indicated that the peroxidase such as catalytic site, present in conventional flavoHbs (22), is absent in this class of flavoHbs. The reductase domain of type II flavoHbs of mycobacteria is also modified within the cofactor binding sites, for example, the RXYS motif of the FAD binding site and GXGXXP motif of the nicotinamide adenine dinucleotide (NADH) binding sites (23, 24). A RKY/F sequence motif, known as high affinity lipid-binding motif (25), appeared conserved within the proximal site of heme in type II flavoHbs. Interestingly, mycobacterial type II flavoHbs exhibited very high (>70%) overall sequence conservation among them but displayed less than 25% homology with conventional (type I) flavoHbs. These unusual structural features indicated that type II flavoHbs of mycobacteria might be structurally and functionally different from conventional flavoHbs.

thumbnail image

Figure 1. (A) Structure-based sequence alignment of mycobacterial flavohemoglobins (type I and type II) with E. coli and Ralstonia flavoHbs. Conserved residues in heme and reductase domains of flavoHbs are highlighted in light gray, and the residues, which are different from conserved and present in MtbFHb (type II), are shown in dark gray. FAD and NADPH binding motifs, present in conventional flavoHbs, are shown in vertical boxes. E. c, Escherichia coli; R. e, Ralstonia eutropha; M. v, Mycobacterium vanvallani; M. g, Mycobacterium gilvum; M. j, Mycobacterium ssp. jls; M. s, Mycobacterium smegmatis; M. u, Mycobacterium ulcerans; M. m, Mycobacterium marinum; M. p, Mycobacterium paratuberculosis; M. av, Mycobacterium avium; M. tb, Mycobacterium tuberculosis; M. a, Mycobacterium absessceu. Type I, denotes protein sequence resembling with conventional E. coli type flavoHbs and type II, sequence designates for the new class of flavoHbs identified in M. tuberculosis and other mycobacteria. (B) Phylogenetic analysis of mycobacterial flavoHbs. The phylogenetic analysis was performed by using MEGA4 (17) program. Mycobacterial genome was searched extensively for the homologs/paralogs and orthologs for flaoHbs and then the first blast hit of different genome was selected and analyzed by sequence alignment and comparisons for making the evolutionary tree. Mycobacterial type II flavoHbs make separate cluster from type I FlavoHbs that form cluster along with conventional flavoHbs of bacteria and yeasts. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

Cloning, Expression, and Characterization of Type II FlavoHb from M. tuberculosis

To gain an insight into the primary characteristics of type II flavoHbs, one of its representatives, encoded by Rv0385 gene in Mycobacterium tuberculosis, was cloned and expressed in E. coli. SDS-PAGE analysis confirmed the presence of a 43-kDa band (Fig. 2A) corresponding to the expected size of M. tuberculosis flavoHb (MtbFHb) protein. Gel filtration analysis of MtbFHb substantiated that it is a monomeric protein of 43.5 kDa (Fig. 2C). Absolute spectra of MtbFHb indicated that protein predominantly exists in the ferric state. The absorption spectra of the ferric species exhibits Soret and visible bands at 414 and 536/570 nm, respectively (Fig. 2B), suggesting a hexacoordinated low-spin (6CLS) heme with an intrinsic amino acid residue or exogenous ligand bound to the distal site of the heme. The absorption spectrum of the ferrous species shows Soret and visible bands at 428 and 533/559 nm, respectively, substantiating the 6CLS configuration of heme, consistent with the presence of a sixth ligand. Exposure of the ferrous protein to CO caused the absorption bands to shift to 423 and 542/572 nm, respectively, typical for CO-bound heme, indicating that the distal ligand is displaced by the CO. This is in sharp variance with conventional flavoHbs that exist in penatcoordinated high spin state (22). These observations indicated that MtbFHb and presumably other mycobacterial type II flavoHbs may be structurally and functionally distinct from conventional type I flavoHbs.

thumbnail image

Figure 2. (A) Overexpression of type II flavoHb of M. tuberculosis in E. coli: (1) Molecular weight marker, (2) E. coli BL21DE3 with control plasmid, pET15b, (3) E. coli BL21DE3 expressing MtbFHb of M. tuberculosis, (4) Purified preparation of MtbFHb. (B) Spectral properties of MtbFHb. Optical absorption spectra of MtbFHb (ferric, deoxy, and CO bound forms). (C) Gel-filtration profile of MtbFHb. Molecular mass of MtbFHb was determined on Superdex G-75 gel-filtration column. Molecular Weight reference standards were taken as 1. Bovine serum albumin, 66 kD, 2. Egg albumin, 40 kD, and 3. Carbonic anhydrase, 29 kD.

Download figure to PowerPoint

As protection against NO and reactive nitrogen species has been found one of the major physiological functions of bacterial and yeast flavoHbs (type I), we tested the NO and oxygen metabolizing properties of MtbFHb by comparing the NO and oxygen uptake of MtbFHb overexpressing cells of E. coli. MtbFHb expressing cells did not show any significant NO uptake when compared with isogenic cells expressing E. coli HMP (Table 2) but displayed moderately improved respiratory activities. NOD activity of MtbFHb was estimated as 12 μM−1 s−1, which was several folds lower than the HMP of E. coli and trHbN of M. tuberculosis (26). Overall observations, thus, suggested significant differences in structural and functional properties of type II flavoHb of M. tuberculosis (MtbFHb) when compared with conventional type I flavoHbs.

Table 2. NO and O2 uptake properties of E. coli expressing MtbFHb
StrainsNOa consumption n mole of NO (heme−1 s−1)O2a uptake (μmol min−1 10−9 cells)
  • a

    NO and oxygen uptake measurements of cells, expressing MtbFHb, were done essentially as mentioned in Experimental procedures and also described earlier (19). Before oxygen uptake measurements, the 0.1 M potassium phosphate buffer was stirred vigorously to make it air saturated (DO = 250 μM). Cell number in the culture was determined by simultaneously plating the cells on Luria Bertani (LB) plate and counting the colonies. Data represent average values obtained after three independent experiments.

E. coli RB9060 (Δ hmp)0.04 ± 0.04.8 ± 0.64
E. coli BL21DE3 (WT)1.14 ± 0.085.3 ± 0.49
E. coli BL21DE3 (E. coli HMP)36.31 ± 7.95.6 ± 0.54
E. coli BL21DE3 (MtbFHb)5. 32 ± 0.077. 1 ± 0.65

Phylogenetic Analysis of Type II FlavoHbs of Mycobacteria

Coexistence of type I and type II flavoHbs in mycobacteria led us to speculate that function(s) of these two flavoHbs may be different from each other. Occurrence of type II flavoHbs in majority of mycobacteria and their presence in limited number of bacterial species (mainly actinomycetes, data not shown) indicated that their function may be novel and specific to their host. To gain an insight into evolutionary corelation between type I and type II flavoHbs of mycobacteria, phylogenetic analysis of two classes of flavoHbs was done. BLAST search within the microbial genome and protein data bank, using E. coli HMP and MtbFHb, retrieved FMN reductase of E. coli and cytochrome b5 reductase of Saccharomyces cerevisiae as orthologs of MtbFHb (type II flavoHbs), whereas benzoate 1,2, dioxygenase appeared one of the closest orthologs of type I flavoHbs of mycobacteria. Therefore, a phylogenetic tree was developed by focusing on type I, type II flavoHbs of mycobacteria and their first orthologs present in different groups (Fig. 1B). Topology of evolutionary tree, thus, developed, separated type II flavoHbs of mycobacteria from type I flavoHbs that formed a separate group along with conventional flavoHbs of bacteria and yeasts. Phylogenetically, type II flavoHbs appeared related to electron-transfer proteins such as FMN-reductase of E. coli and cytochrome b5reductase of Sacchromyces cerevisiae, whereas type I flavoHbs of mycobacteria and other conventional flavoHbs exhibited phylogenetic closeness with dioxygenases. Overall structure of phylogenetic tree and duplication nodes suggest an ancestral duplication and diversion of type I and type II mycobacterial flavoHbs.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

FlavoHbs are widely spread among bacteria, yeast, and fungi (1, 2) and represent a unique example of multidomain protein, where two domains having different functional properties, interact and perform entirely new function(s) by carrying out diverse redox reactions. Heme and reductase domains of flavoHbs display significant sequence conservation within their cofactor binding sites (3, 21). A novel flavoHb, displaying crucial differences within the functionally conserved regions of heme and reductase domain, has been identified in many species of mycobacteria. Occurrence of this unconventional class (type II) of flavoHbs appeared limited to few microbes, mainly belonging to actinomycetes group. Coexistence of type I and type II flavoHbs in several mycobacterial species suggest that they may be playing different functions in cellular metabolism of their host.

Type II flavoHbs may constitute a distinct class within the flavoHb family and may have novel function(s). Functionally conserved regions of heme and redox domains of type I flavoHbs are modified in type II flavoHbs of mycobacteria suggesting that the interactions of heme and redox domain and mode of redox reactions are different in these proteins. This is supported by the observation that type II flavoHb of M. tuberculosis exists in hexacoordinate state and lacks NO metabolizing activity unlike majority of type I flavoHbs. When MtbFHb (type II) was expressed in E. coli, it appeared strongly associated with cell membranes and bioinformatics analysis indicated that its orthologs in E. coli and S. cerevisiae are FMN reductase and Cytochrome b5 reductse, respectively. Structural and functional properties of type II flavoHbs are not obvious at present; however, their occurrence in limited group of microbes suggests that they may have novel function(s) in the cellular metabolism of their host.

To understand the need for two different classes of flavoHbs and their evolutionary corelation, phylogenetic analysis of mycobacterial flavoHbs were conducted. Topology of the evolutionary tree, developed after taking into consideration the paralogs and orthologs of type I and type II flavoHbs of mycobacteria, indicated that type II flavoHbs form a separate cluster from type I flavoHbs and may be functionally different. However, both may have originated from a common ancestor and evolved after duplication and diversion. Although precise origin and functional roles of mycobacterial flavoHbs remains to be elucidated, respective spacing of mycobacterial type I and type II flavoHbs and their paralogs suggest that type II flavoHbs may be phylogenetically more related to electron-transferring proteins (FMN reductases, cytochrome b5 reductases, etc.), whereas type I flavoHbs are closer to dioxygenases.

The diversity in the number and type of flavoHb-encoding genes in the genome of mycobacteria might reflect the difference in requirement of these proteins in individual species. Type II flavoHbs, thus, may be required for the cellular metabolism of both pathogenic and nonpathogenic mycobacteria as it has been found in majority of analyzed mycobacterial genome (except M. leprae, where it has been identified as pseudogene). Phylogenetic studies on invertebrate globins (27) suggested that flavoHbs and flavoHb like single-domain Hbs (FHb/SDSgb) family diverged much more in their structure-function than truncated and globin-coupled sensors. Unconventional flavoHbs of mycobacteria may add another dimension on diversification of microbial two-domain globins.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The authors thank financial support from Council of Scientific and Industrial Research under the SIP10 grant and NSF grant 095635 (SRY) for carrying out this work.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. EXPERIMENTAL PROCEDURES
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES