Previous Article | Next Article ![]()
Journal of Bacteriology, July 2007, p. 4960-4963, Vol. 189, No. 13
0021-9193/07/$08.00+0 doi:10.1128/JB.00272-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Division of Bacteriology, Department of Infection and Immunity, Jichi Medical University, Tochigi 329-0498,1 Department of Bacteriology, Graduate School of Medicine, Dentistry and Pharmacology, Okayama University, Okayama 700-8558, Japan2
Received 19 February 2007/ Accepted 12 April 2007
|
|
|---|
|
|
|---|
In this study, we have identified a qualitative alteration of FldA in H. pylori: FMN-bound FldA holoenzyme (holo-FldA) is converted to FMN-free FldA apoprotein (apo-FldA). The decrease in holo-FldA levels appears to correlate with the accumulation of apo-FldA levels during long-term culture. We have also managed to purify and identify the agent involved in the conversion from holo- to apo-FldA by using the class C acid phosphatase (HppA), a nonspecific acid phosphatase (10). This report is the first to describe the role of HppA in FldA catabolism.
H. pylori NCTC 11638 was used as the bacterial strain for this study. The organisms were cultured with brain heart infusion broth containing 5% horse serum in an atmosphere of 5% O2, 10% CO2, and 85% N2 at 37°C in a microaerobic chamber (Concept Plus; Ruskinn Technology, Leeds, United Kingdom). Numbers of CFU were determined by the conventional method (Fig. 1A). Organisms suspended in 50 mM Tris buffer (pH 7.5) were sonicated to obtain cell lysates, and then the proteins in the cell lysates of the organisms recovered at each culture time were analyzed by native 12.5% polyacrylamide gel electrophoresis (PAGE). Two protein bands that had clearly undergone significant alterations during culture were detected at the side of the positive pole in a native PAGE gel stained by Coomassie brilliant blue (CBB) R-250 (Fig. 1B). The slow-migrating protein band (Px) was still undetectable by CBB staining on day 0.5 of the culture (106.7 CFU/ml). By day 3 (108.3 CFU/ml), however, it was strongly detectable, and thereafter it steadily increased in intensity up to day 14 (104.6 CFU/ml). In contrast, the fast-migrating protein band (Py) was strongly detectable on both day 0.5 and day 3 of the culture but was undetectable by day 14.
![]() View larger version (32K): [in a new window] |
FIG. 1. Alterations in CFU during microaerobic culture and native PAGE profiles of bacterial proteins. (A) H. pylori NCTC 11638 was cultured in brain heart infusion broth containing 5% horse serum in a microaerobic chamber. Numbers of CFU at the indicated culture times were determined by the conventional method. (B) Proteins (80 µg/lane) in cell lysates prepared from organisms recovered at the indicated culture times were analyzed by native PAGE followed by CBB staining.
|
To characterize the two forms of FldA (Px and Py) in detail, the purified Px and Py proteins were then analyzed by matrix-assisted laser desorption ionization-time-of-flight mass spectrometry (MALDI-TOF MS) (Shimadzu Co.). The mass spectrum of an FldA protein with a molecular mass of about 17.5 kDa and the mass spectrum of FMN (identified with the mass spectrum of an authentic FMN) were both detected in the Py protein, whereas only the former was detected in the Px protein (Fig. 2A). Thus, the Px protein was determined not to contain FMN. When the gel was irradiated by UV light at a 254-nm wavelength after native PAGE of the purified Px and Py proteins, a fluorescent yellow band of FMN was detected at the position corresponding to the Py protein band stained by CBB but not at the position corresponding to the Px protein band (Fig. 2B). The Py protein was thus confirmed to contain FMN. These results indicate that the fast-migrating FldA (Py protein) is a holoenzyme (holo-FldA) containing FMN, while the slow-migrating FldA (Px protein) is an FMN-free apoprotein (apo-FldA).
![]() View larger version (22K): [in a new window] |
FIG. 2. Analysis of holo-FldA (Py protein) and apo-FldA (Px protein). (A) Purified Py protein (upper), purified Px protein (middle), and an authentic FMN (lower) were subjected to MALDI-TOF MS with -cyano-4-hydroxycinnamic acid used as a matrix. (B) After subjection of the Py (10 µg/lane) and Px (10 µg/lane) proteins to native PAGE, the proteins in the gel (lanes 1, Py; lanes 2, Px) were exposed to UV light at a 254-nm wavelength (right) and stained by CBB (left). (C) The fluorescence intensity of FMN was assayed with an authentic FMN (15 µg/ml, peak 1), purified holo-FldA (500 µg/ml, peak 2), and DTT-treated holo-FldA (500 µg/ml, peak 3) by scanning at an excitation wavelength of 536 nm.
|
The following experiments were carried out to examine the difference in migration between holo- and apo-FldA in native PAGE. The fluorescence intensity of FMN was scanned at an excitation wavelength of 536 nm, and the absorption bands of FMN were detected in purified holo-FldA with three peaks of the absorption curve, which was identical to the absorption curve of an authentic FMN (Fig. 2C, peaks 1 and 2). The FMN absorption bands disappeared when the purified holo-FldA was treated with dithiothreitol (DTT) at a final concentration of 50 mM, leaving only one weakly detectable FMN absorption band at the short wavelength (Fig. 2C, peak 3). This meant that the FMN in the purified holo-FldA was in an oxidized form and was reducible by DTT treatment. The FMN absorption bands were undetectable in the purified apo-FldA and in the holo-FldA from which FMN had been removed by 10% trichloroacetic acid (TCA) precipitation. These results confirm the absence of FMN in the purified apo-FldA and TCA-treated holo-FldA (an artifact apo-FldA). The native PAGE profiles of FldA were analyzed with purified holo-FldA (containing oxidized FMN), DTT-treated holo-FldA (containing reduced FMN), purified apo-FldA (without FMN), and TCA-treated holo-FldA (without FMN). The electrophoretic position of DTT-treated holo-FldA corresponded with that of the holo-FldA containing oxidized FMN. Thus, the reduction of the isoalloxazine ring in the FMN with DTT treatment did not influence the migration of holo-FldA in native PAGE. In contrast, the TCA-treated holo-FldA (an artifact apo-FldA) had been electrophoretically switched to a position identical to the position of the purified apo-FldA. These results confirmed the utility of native PAGE in analyzing the conversion from holo- to apo-FldA and the conversion from apo- to holo-FldA.
Holoflavodoxin of Chondrus crispus, a macroalga, has also been found to migrate toward the positive pole faster than apoflavodoxin in native PAGE (9). The electrophoretic behaviors of holo- and apo-FldA in H. pylori were thus found to be compatible with those of two forms of flavodoxin in C. crispus in native PAGE. These results suggest that holo-FldA carries a stronger negative electric charge than does apo-FldA because of the phosphate group of FMN in the former. It is necessary, however, to carry out further investigations to clarify the differences in migration between holo- and apo-FldA in the native PAGE gel.
The conversion from apo- to holo-FldA was examined by native PAGE analysis. A purified apo-FldA or an artifact apo-FldA (prepared by TCA precipitation of a purified holo-FldA) was incubated in the presence or absence of FMN for 2 h at room temperature in 150 mM Tris buffer (pH 7.5). After incubation, the specimens were subjected to native PAGE with a control holo-FldA. Native PAGE analysis detected binding of FMN to the purified apo-FldA and the artifact apo-FldA: the apo-FldA incubated with FMN migrated toward the positive pole faster than the apo-FldA incubated without FMN in the gel, and the electrophoretic position of the apo-FldA with FMN was identical to the position of the control holo-FldA. These results indicate that apo-FldA spontaneously joins with FMN and thereby converts to holo-FldA.
As described above, some enzymatic activity was required for the conversion from holo-FldA to apo-FldA. To identify one of the activities involved in the conversion from holo-FldA to apo-FldA, we conducted experiments using cell lysates obtained from organisms cultured for 2 days (108.0 CFU/ml) and prepared with 50 mM Tris buffer (pH 7.5 or 8.6). In brief, cell lysates of pH 7.5 and 8.6 were applied to a Q-Sepharose (Sigma-Aldrich Co.) anion-exchange column and an SP Sepharose (Amersham Biosciences AB, Uppsala, Sweden) cation-exchange column, respectively, for fractionation with various concentrations of NaCl. After dialysis against 50 mM Tris buffer (pH 7.5), each fraction obtained from anion- or cation-exchange chromatography was incubated with purified holo-FldA overnight at 37°C in 50 mM Tris buffer (pH 7.5). The activity involved in the conversion from holo-FldA to apo-FldA was confirmed by monitoring the generation of apo-FldA from holo-FldA in native PAGE after the incubation. A sodium dodecyl sulfate-PAGE analysis revealed a protein with a molecular mass of 28 kDa at significant levels in the fractions exhibiting the activity involved in the conversion from holo-FldA to apo-FldA, though the expression levels of the same 28-kDa protein, as determined by CBB staining, differed from fraction to fraction. Interestingly, we found that the culture supernatant of the organisms also exhibited the activity involved in the conversion from holo-FldA to apo-FldA. To investigate further, we tried to purify the activity involved in the conversion from holo-FldA to apo-FldA by using a culture supernatant containing far less of the extra proteins than in the cell lysates. Organisms cultured for 2 days in PPLO broth containing 0.2% 2,6-di-O-methyl-ß-cyclodextrin were washed once with phosphate-buffered saline (PBS), suspended in PBS, incubated overnight in a microaerobic chamber, and removed by centrifugation to obtain a supernatant. After concentration of the supernatant and replacement of the PBS with 50 mM Tris buffer (pH 8.6), the buffer-replaced supernatant was subjected to cation-exchange chromatography using SP Sepharose (Amersham Biosciences AB) equilibrated with 50 mM Tris buffer (pH 8.6). The activity involved in the conversion from holo-FldA to apo-FldA was successfully obtained as a single protein with a molecular mass of 28 kDa from the fraction eluted with 0.2 M NaCl. PMF analysis (Shimadzu Co.) identified a 28-kDa protein (the activity involved in the conversion from holo-FldA to apo-FldA) with class C acid phosphatase (HppA) matching the HppA protein (calculated pI value, 9.41) of the H. pylori CZD 115 strain, with a top score of 111 in the Mascot Search (Matrix Science Ltd.).
A confirmatory PMF analysis identified the activity involved in the conversion from holo-FldA to apo-FldA as HppA. We thus inferred that the HppA-induced dephosphorylation of FMN plays an important role in the conversion from holo- to apo-FldA. To investigate further, we examined the generation of riboflavin (Rib) from FMN in holo-FldA reacted with HppA. A MALDI-TOF MS analysis (Shimadzu Co.) was performed to detect the mass spectrum of Rib or FMN in holo-FldA solution incubated in the presence or absence of HppA. The holo-FldA solution incubated without HppA exhibited the mass spectrum of FMN (identified with the mass spectrum of an authentic FMN), whereas the holo-FldA solution incubated in the presence of HppA exhibited the mass spectrum of Rib (identified with the mass spectrum of an authentic Rib) (Fig. 3A). This confirmed that the FMN within the FldA protein was dephosphorylated via the action of HppA, which in turn resulted in Rib. We next examined whether the phosphomonoesterase activity of HppA induced FMN dephosphorylation. As shown in Fig. 3B, holo-FldA was converted to apo-FldA in the presence of HppA without 4-methylumbelliferyl phosphate (MUP), a substrate for the phosphomonoesterase activity of a nonspecific acid phosphatase, whereas the HppA-induced conversion of holo-FldA to apo-FldA was inhibited in the presence of excessive MUP. These results confirm that the FMN in holo-FldA is hydrolyzed by the phosphomonoesterase activity of HppA and that the resulting Rib is released from the FldA protein.
![]() View larger version (18K): [in a new window] |
FIG. 3. Dephosphorylation of FMN in holo-FldA via the phosphomonoesterase activity of HppA. (A) Holo-FldA (1 µg) was incubated in the presence or absence of HppA (0.05 µg) overnight at 37°C in 50 mM Tris buffer (pH 7.5) containing 2 mM MgCl2. After incubation, solutions of holo-FldA, an authentic FMN, a mixture of holo-FldA and HppA, and an authentic Rib were subjected to MALDI-TOF MS with -cyano-4-hydroxycinnamic acid used as a matrix to analyze the mass spectrum of FMN or Rib. (B) Holo-FldA (1 µg) was incubated for 2 h at 37°C with HppA (0.05 µg) in the presence or absence of MUP (50 µg) in 50 mM Tris buffer (pH 5.5) containing 2 mM MgCl2. After incubation, the specimens were analyzed together with control holo-FldA by native PAGE followed by CBB staining.
|
Published ahead of print on 20 April 2007. ![]()
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»