Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 May 7.
Published in final edited form as: Biochemistry. 2013 Apr 24;52(18):3025–3027. doi: 10.1021/bi400382p

Heme degradation by Staphylococcus aureus IsdG and IsdI liberates formaldehyde rather than carbon monoxide

Toshitaka Matsui †,*, Shusuke Nambu , Yukari Ono , Celia W Goulding , Kouhei Tsumoto ¶,§, Masao Ikeda-Saito †,*
PMCID: PMC3672231  NIHMSID: NIHMS472302  PMID: 23600533

Abstract

IsdG and IsdI from Staphylococcus aureus are novel heme degrading enzymes containing unusually non-planar (ruffled) heme. While canonical heme degrading enzymes, heme oxygenases, catalyze heme degradation coupled with the release of CO, in this study we demonstrate that the primary C1 product of the S. aureus enzymes is formaldehyde. This finding clearly reveals that both IsdG and IsdI degrade heme by an unusual mechanism distinct from the well-characterized heme oxygenase mechanism as recently proposed for MhuD from Mycobacterium tuberculosis. We conclude that heme ruffling is critical for the drastic mechanistic change for these novel bacterial enzymes.


Heme degradation in pathogenic bacteria is crucial to acquire iron, an essential nutrient required for its survival and infection, from host heme molecules.1-3 An enzyme family termed heme oxygenase (HO) is known to degrade heme into iron and biliverdin with the release of a meso carbon atom CO (Scheme 1).4-7 The HO catalysis proceeds through three successive mono-oxygenation reactions where the substrate heme activates oxygen molecules for a series of self-oxidations. Until recently, the HO family was thought to be the only heme degrading enzymes, however, a novel family of heme degrading enzymes has been identified in Staphylococcus aureus.8 IsdG and its paralogue, IsdI, are heme degrading enzymes of S. aureus composing of a heme-uptake system called the Isd system.8 The IsdG-type proteins have distinct structures compared to the HO-family enzymes.9,10 Especially, heme bound to the IsdG-type proteins show unusual non-planarity, best described as ruffled. This heme ruffling is expected to significantly modulate the O2 activation chemistry on the heme molecule, and in fact, a novel tetrapyrrole, staphylobilin, having an additional oxidation at a meso carbon is produced (Scheme 1).10 Nevertheless, the unique reaction mechanism of the IsdG-type enzymes remains unclear.

Scheme 1.

Scheme 1

Heme degradation products of the HO- and IsdG-type enzymes*

We have recently found that MhuD, an IsdG-type enzyme from Mycobacterium tuberculosis, catalyzes the heme degradation without generating CO (Scheme 1).11 The carbon atom at the site of ring cleavage is not released but retained as an aldehyde group, resulting in formation of a novel chromophore termed mycobilin. This unprecedented observation indicates a unique reaction mechanism of MhuD which is distinct from the well-understood mechanism of canonical HO. We have attributed the drastic mechanism change to the heme ruffling as observed in the S. aureus IsdG and IsdI heme bound structures. The S. aureus enzymes, however, have been assumed to generate CO despite lack of any experimenl evidence.10 This assumption is based on the structural similarities of staphylobilin and biliverdin at their ring cleavage site and removal of the meso carbon (Scheme 1). To this end, we have experimentally determined the C1 products of S. aureus IsdG and IsdI for the first time.

Single-turnover heme degradation reactions by IsdG and IsdI as well as MhuD and rat HO-1 were conducted using ascorbate as an electron donor. HPLC and MS analyses confirm that the four enzymes afford their characteristic tetrapyrrole products (Figure S1). The amount of CO produced was determined by trapping CO by a H64L variant of myoglobin (Mb) which has an extremely high CO affinity.12 Figure 1 shows the Soret region difference spectra of the ferrous H64L Mb induced by CO generated by the single turnover heme degradation. As reported earlier,11 while the reaction solution of rHO-1 exhibited a significant difference spectrum, essentially no spectral change was observed for MhuD. The CO yields summarized in Table 1 show almost stoichiometric formation of CO in the HO-1 reaction and its absence in the MhuD reaction. IsdG and IsdI exhibited the CO-induced difference spectra, but their low intensities indicate low yields of CO (~14%, Figure 1 and Table 1). This finding clearly demonstrates that CO is not the major C1 product of the S. aureus enzymes.

Figure 1.

Figure 1

CO quantification as difference spectra of H64L Mb upon formation of its ferrous CO form. Spectral changes of ferrous H64L Mb were calculated from the spectra taken before and after the heme degradation by 5 μM heme complexes of rat HO-1 (green), MhuD (purple), IsdG (blue) and IsdI (red), respectively. The black broken line represents difference spectrum of 5 μM ferrous H64L Mb upon saturation with CO.

Table 1.

Yields of C1 products in heme degradationa

Enzyme COb HCHOc HCOOHc
rHO-1 106 ± 4 4 ± 2 1 ± 1
MhuD 1 ± 1 4 ± 1 5 ± 1
IsdG 14 ± 2 82 ± 1 16 ± 1
IsdI 14 ± 1 81 ± 5 15 ± 1
a

Yields are given in percent (%)

b

n=2

c

n=3

Since staphylobilin actually lacks one meso carbon atom, IsdG and IsdI should afford alternative C1 products. The aldehyde group retained in mycobilin compared to staphylobilin (Scheme 1) suggests formation of formaldehyde (HCHO) and/or its oxidized form, formic acid (HCOOH), by the S. aureus enzymes. Formaldehyde in the reaction solution was converted by an acetylacetone method to 3,5-diacetyl-1,4-dihydrolutidine,13 which was quantitated by HPLC analysis (Figure 2A and Table 1). While rHO-1 and MhuD afford negligible amount of formaldehyde, the prominent signals observed for IsdG and IsdI reveal high yields of formaldehyde in their reactions (approximately 80 %). Quantitation of formic acid was based on its enzymatic oxidation.14 Formic acid dehydrogenase consumes formic acid to generate an equimolar amount of NADH. HPLC quantitation of NADH (Figure 2B and Table 1) indicates negligible (rHO-1 and MhuD) or very low yields (IsdG and IsdI) of formic acid. These observations firmly identify formaldehyde as the primary C1 product of both IsdG and IsdI.

Figure 2.

Figure 2

HPLC chromatograms for quantitation of C1 products. (A) Formaldehyde as 3,5-diacetyl-1,4-dihydrolutidine and (B) formic acid as NADH produced by rat HO-1 (green), MhuD (purple), IsdG (blue), and IsdI (red). Product signals are marked with asterisks. Concentrations of enzymes and standards were 20 μM (formaldehyde) and 50 μM (formic acid).

This study shows that S. aureus IsdG and IsdI degrade heme by an unusual mechanism where one meso carbon atom of the porphyrin ring is liberated primarily as formaldehyde (Table 1). While canonical HO releases the meso carbon as CO,6,7 only small amounts of CO are produced by the S. aureus enzymes. The absence of CO production is first reported for heme degradation by MhuD which retains the meso carbon at the cleavage site as the aldehyde group (Scheme 1).11 The similar aldehyde formation by IsdG and IsdI are suggestive of their mechanistic similarity to MhuD catalysis. The low CO yields indicate that the three IsdG-type enzyme reactions do not involve verdoheme, a key intermediate of ring cleavage by HO (Scheme S1). Among three successive oxygenations in canonical HO catalysis, CO is generated at the second step, meso-hydroxyheme to verdoheme.6,7 Hydroxyheme is highly reactive with O2 to release CO due to its radical character,6,15 and this rapid auto-oxidation does not require the assistance by the HO enzyme. Thus, the unique mechanisms of the three IsdG-type enzymes appear to be induced by drastic changes in the intrinsic reactivity of hydroxyheme (or its equivalent). Since the most remarkable structural feature common to hemes in MhuD and IsdG/IsdI is their severe non-planarity,9-11 we propose that the heme ruffling is critical to suppress CO production and to promote aldehyde formation. The ruffling can modulate reactivity of hydroxyheme through the large steric distortion and/or by changing its electronic configuration.16 The major difference between the MhuD and IsdG/IsdI reactions is the fate of the aldehyde group; retention in the tetrapyrrole by MhuD or release as formaldehyde by the S. aureus enzymes (Scheme S1). While deformylation is known to be mediated by a peroxoheme complex, such a nucleophilic deformylation normally liberates formic acid,17 which is a minor product of IsdG and IsdI (Table 1).

Great attention should also be paid with regards to the biological significance of the unusual products of the IsdG-type enzymes. Considering the common absence of CO in the M. tuberculosis and S. aureus enzymes, the most probable scenario is that microbes employing the IsdG-type enzymes do not favor endogenous production of this toxic gas. Exogenous CO is shown to activate the dormancy regulon of M. tuberculosis at biologically relevant concentrations18 and to have a significant bactericidal effect on S. aureus at higher concentrations.19 Moreover, the unique linear tetrapyrroles and formaldehyde may also have some biological functions.

In conclusion, this study shows the unusual formaldehyde formation in heme degradation by S. aureus IsdG and IsdI probably due to heme ruffling. It has been postulated that heme ruffling in the IsdG-type enzymes enhances its reactivity while controlling the regioselectivity.10,16 Our results suggest that heme ruffling opens a new path for its degradation.

Supplementary Material

1_si_001

Acknowledgments

Funding

This work has been supported by Grants-in-Aid for Scientific Research (M.I.-S., 2412006, 24350081; T.M., 23550186; K.T., 24000011, 21360398) from JSPS and MEXT, Japan, by the Strategic Alliance Project from MEXT, Japan, and by the National Institutes of Health Grant (C.W.G., AI081161).

Footnotes

Supporting Information. Experimental procedures, Figure S1, and Scheme S1. This material is available free of charge via the Internet at http://pubs.acs.org.

The authors declare no competing financial interest.

REFERENCES

  • 1.Schmitt MP. J. Bacteriol. 1997;179:838–845. doi: 10.1128/jb.179.3.838-845.1997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Wandersman C, Stojiljkovic I. Curr. Opin. Microbiol. 2000;3:215–220. doi: 10.1016/s1369-5274(00)00078-3. [DOI] [PubMed] [Google Scholar]
  • 3.Genco CA, Dixon DW. Mol. Microbiol. 2001;39:1–11. doi: 10.1046/j.1365-2958.2001.02231.x. [DOI] [PubMed] [Google Scholar]
  • 4.Yoshida T, Kikuchi G. J. Biol. Chem. 1978;253:4230–4236. [PubMed] [Google Scholar]
  • 5.Tenhunen R, Marver HS, Schmid R. J. Biol. Chem. 1969;244:6388–6394. [PubMed] [Google Scholar]
  • 6.Ortiz de Montellano PR. Acc. Chem. Res. 1998;31:543–549. [Google Scholar]
  • 7.Matsui T, Unno M, Ikeda-Saito M. Acc. Chem. Res. 2010;43:240–247. doi: 10.1021/ar9001685. [DOI] [PubMed] [Google Scholar]
  • 8.Skaar EP, Gaspar AH, Schneewind O. J. Biol. Chem. 2004;279:436–443. doi: 10.1074/jbc.M307952200. [DOI] [PubMed] [Google Scholar]
  • 9.Lee WC, Reniere ML, Skaar EP, Murphy ME. J. Biol. Chem. 2008;283:30957–30963. doi: 10.1074/jbc.M709486200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Reniere ML, Ukpabi GN, Harry SR, Stec DF, Krull R, Wright DW, Bachmann BO, Murphy ME, Skaar EP. Mol. Microbiol. 2010;75:1529–1538. doi: 10.1111/j.1365-2958.2010.07076.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Nambu S, Matsui T, Goulding CW, Takahashi S, Ikeda-Saito M. J. Biol. Chem. 2013;288:10101–10109. doi: 10.1074/jbc.M112.448399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Rohlfs RJ, Mathews AJ, Carver TE, Olson JS, Springer BA, Egeberg KD, Sligar SG. J. Biol. Chem. 1990;265:3168–3176. [PubMed] [Google Scholar]
  • 13.Nash T. Biochem. J. 1953;55:416–421. doi: 10.1042/bj0550416. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Schaller K-H, Triebig G. Formate. In: Bergmeyer HU, Bergmeyer J, Grassl M, editors. Methods of Enzymatic Analysis. 3rd ed. Verlag Chemie; Weinheim: 1985. pp. 668–672. [Google Scholar]
  • 15.Morishima I, Fujii H, Shiro Y, Sano S. Inorg. Chem. 1995;34:1528–1535. [Google Scholar]
  • 16.Takayama SJ, Ukpabi G, Murphy ME, Mauk AG. Proc. Natl. Acad. Sci. U. S. A. 2011;108:13071–13076. doi: 10.1073/pnas.1101459108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Wertz DL, Sisemore MF, Selke M, Driscoll J, Valentine JS. J. Am. Chem. Soc. 1998;120:5331–5332. [Google Scholar]
  • 18.Kumar A, Deshane JS, Crossman DK, Bolisetty S, Yan BS, Kramnik I, Agarwal A, Steyn AJ. J. Biol. Chem. 2008;283:18032–18039. doi: 10.1074/jbc.M802274200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Nobre LS, Seixas JD, Romao CC, Saraiva LM. Antimicrob. Agents Chemother. 2007;51:4303–4307. doi: 10.1128/AAC.00802-07. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1_si_001

RESOURCES