Download PDF - SAGE Journals

C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
207
Structural Modifications of Methanococcus Jannaschii Flagellin Proteins
C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
Structural modifications of Methanococcus jannaschii
flagellin proteins revealed by proteome analysis
Carol S. Giometti,* Sandra L. Tollaksen and Gyorgy Babnigg
Biosciences Division, Argonne National Laboratory, 9700 South Cass Avenue, Building 202, Room B117, Argonne, IL 60439, USA.
E-mail: [email protected]
Claudia I. Reich and Gary J. Olsen
Department of Microbiology, University of Illinois at Urbana-Champaign, Urbana, IL, USA
Hanjo Lim and John R. Yates, III
Mass Spectrometry Group, The Scripps Institute, La Jolla, CA, USA
Methanococcus jannaschii is an autotrophic archaeon originally isolated from an oceanic thermal vent. The primary metabolic pathway for energy production in this hyperthermophilic microbe is methanogenesis from H2 and CO2. As an autotroph, M. jannaschii
requires only CO2 as a carbon source for synthesizing all necessary biomolecules. Changes in the environmental availability of these
molecules can be expected to activate regulatory mechanisms manifested as the up and down regulation of specific genes and the concomitant increase and decrease in abundance of the corresponding proteins. In our analysis of the proteome of M. jannaschii, we have
observed significant changes in the abundance of a common subset of predominant proteins in response to reduced H2 concentration,
limited ammonium availability and the stage of cell growth (exponential compared with stationary). The masses of tryptic peptides
from these proteins match those predicted by M. jannaschii genome open reading frames annotated as flagellin B1 (MJ0891) and
flagellin B2 (MJ0892). Multiple proteins with different isoelectric points and molecular weights match each of these proteins and the
abundance of these protein variants changes with growth conditions. These data indicate that structural modifications altering both
the isoelectric point and size of the M. jannaschii flagellin B1 and B2 proteins occur in response to growth conditions and growth stage
of M. jannaschii and further suggest the regulation of M. jannaschii motility through structural modifications of the building blocks
of the flagella.
Keywords: proteomics, flagellin, archaea, Methanococcus jannaschii
Introduction
M. jannaschii is an autotrophic hyperthermophile originally isolated from a deep sea hydrothermal vent.1 An obligate anaerobe, M. jannaschii derives energy by using H2 to
reduce CO2 to methane (CH4). Analysis of the complete
genome sequence of M. jannaschii reveals approximately
1785 proteins, 70% of which have been assigned tentative
functions based on DNA sequence similarity with known
archaeal, bacterial or eukaryotic proteins.2
This paper is being published jointly in European Journal of Mass Spectrometry and in Proteomics (ISSN 1615-9853).
Access to the complete genome sequence and associated annotations for M. jannaschii provides the opportunity
to study the proteome of this microbe in detail. In addition to
identifying all of the detectable proteins with relative ease,
due to the small size of the genome, the metabolism of this
microbe is amenable to investigation through manipulation
of its growth environment. The metabolic dependency of M.
jannaschii on H2 for energy metabolism, for example, suggests that deprivation of H2 should activate a cellular
response designed to protect existing energy stores and to
seek sources of H2. In experiments designed to grow M.
jannaschii in the presence of lower than normal H2 partial
pressure, we have observed quantitative changes in a number of proteins. The masses of tryptic peptides from these
208
Structural Modifications of Methanococcus Jannaschii Flagellin Proteins
proteins have revealed changes in the abundance, not only of
proteins known to be involved in the methanogenesis pathway, but also of a group of proteins annotated as flagellin
proteins.
In bacteria, the flagellar filaments are composed of
polymers of one or more protein subunits referred to as
flagellin.3 The genome annotation for M. jannaschii indicates genes encoding three flagellin proteins (FlaB1, FlaB2
2
and FlaB3) in an operon-like arrangement, although the
flagellum-like structures on M. jannaschii have been sug4
gested to be more similar to type IV pili (Tfp) than flagella.
5
In a recent paper, Mukhopadhyay et al. reported that two
proteins with tryptic peptide masses similar to those predicted by the FlaB2 and FlaB3 open reading frames (ORFs:
MJ0892 and MJ0893, respectively) were altered in abundance in response to cell growth under different hydrogen
partial pressures. Thus, the expression of multiple flagellin
proteins in M. jannaschii is regulated in response to changes
in the cellular environment.
We have compared the two-dimensional electrophoresis (2DE) patterns of M. jannaschii proteins from cells
grown under different hydrogen partial pressures, different
concentrations of ammonium and harvested at different
stages of growth. From these patterns, we have determined
the observed isoelectric points and molecular weights for the
multiple protein spots identified as flagellin B1 and B2 and
compared them with the predicted isoelectric points and
molecular weights from the M. jannaschii ORFs for FlaB1
and FlaB2. Based on a thorough examination of the predicted protein sequences from these ORFs and the tryptic
peptide fragments revealed by mass spectrometry analysis of
proteins extracted from the 2DE gels, we conclude that posttranslational modifications of the flagellin B1 and B2 proteins occur as part of the cellular response to these different
growth conditions producing dramatic structural changes in
the final protein products that could affect flagellum assembly and, therefore, cell motility.
Materials and methods
Cell growth
M. jannaschii cells were grown in a 16-L stainless steel
fermenter (New Brunswick Scientific, operated by the Fermentation Facility of the University of Illinois School of
Life Sciences) under control conditions (defined as growth
at 83°C in a minimal mineral salts medium containing
476 mM NaCl, 14 mM MgCl2, 14 mM MgSO4, 22 mM
NH4Cl, 4.6 mM KCl, 2 mM Na2S, 0.8 mM KH2PO4, 10 mM
NaHCO3, 8.4 µM CoCl2, 3.8 µM FeCl3, 7.3 µM ZnCl2,
5.1 µM MnCl2, 5 mM CaCl2, 4.2 µM NiCl2, 3.1 µM CuSO4,
2 µM Na2MoO4, 2 µM Na2SeO4, 2 µM Na2WO4 and 73.4 µM
nitrilotriacetic acid, pH 6.0) pressurized with 100 kPa
hydrogen and 25 kPa CO2 and harvested in the exponential
phase of growth.
To study the effect of growth under reduced hydrogen
pressure, the partial pressure of hydrogen was reduced from
100 kPa to 5 kPa with the total gas pressure maintained by
supplementing with N2. Cells were harvested 2 hours later,
still in exponential growth phase. For comparison of proteins
in M. jannaschii cells in exponential and stationary growth
phases, cells were grown under control conditions, harvested at different time points and counted with a
Petroff–Hauser cell counter. Protein changes related to the
concentration of ammonium in the growth media were analyzed by inoculating a fermenter containing control media
(except NH4Cl concentration was 0.44 mM instead of
22 mM), allowing the cells to reach exponential growth
phase and harvesting an aliquot. The concentration of NH4Cl
in the fermenter was increased to 22 mM and an aliquot of
cells (still in exponential growth) was collected three hours
later. All cells were washed twice with 385 mM
NaCl + 28 mM MgCl2 after harvest, collected by centrifugation and rapidly frozen.
Sample preparation
Frozen M. jannaschii cells were thawed and immediately mixed with five volumes of a solution containing 9 M
urea, 2% (v / v) 2-mercaptoethanol, 2% (v / v) ampholytes
(Bio-Rad pH 8–10) and 4% (v / v) Nonidet P40 in the presence of protease inhibitors (Boeringer Mannheim Complete
Mini protease inhibitor cocktail). The cell lysate was incubated at room temperature for 10 min and then centrifuged at
435,000 g for 10 min in a Beckman TL100 tabletop ultracentrifuge. Protein concentrations of the resulting clarified
lysates were determined by using the Ramagli modification
of the Bradford protein assay.6
Electrophoresis
Isoelectric focusing (IEF) was done using a mixture
(50 : 50) of pH 3–10 and pH 5–7 carrier ampholytes (BioRad) in 7-inch tube gels with a diameter of 1.5 mm for
14,000 V–h essentially as described by Anderson and
Anderson.7 Aliquots of M. jannaschii whole-cell lysates
containing 20 µg of protein (for silver staining) or 300 µg of
protein (for Coomassie Blue staining) were loaded onto each
IEF gel. After IEF, first-dimension gels were equilibrated in
a buffer containing sodium dodecyl sulfate (SDS) as originally described by O’Farrell8 and then loaded onto 10–17%
9
polyacrylamide gradient gels cast as slabs. The seconddimension separation was done using the Laemmli buffer
8,9
system as previously described. Proteins were then fixed
and stained in the gels using 0.2% (w / v) Coomassie Blue
R250 in 2.5% (v / v) phosphoric acid and 50% (v / v) etha10
nol, or fixed in 50% (v / v) ethanol with 0.1% (v / v) formaldehyde and 1% (v / v) acetic acid for subsequent staining
11
with silver nitrate.
Digitization and quantitative analysis of 2DE patterns
The 2DE protein patterns were digitized using an
Eikonix 1412 CCD scanner interfaced with a VAXstation
C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
4000-90. Parameter files, referred to as spot files, were pro12
duced as described previously. A master pattern of the M.
jannaschii proteins detected with silver stain was generated
by making a copy of a spot file from cells grown under control conditions and harvested in exponential phase. Proteins
in all of the spot files representing the 2DE protein patterns
from experiments were matched to the master pattern so that
each protein was numbered and the integrated density was
determined. The software used for image acquisition and
data analysis has been previously described.12–14
209
ter (Thermo Finnigan, San Jose, CA, USA). Tandem mass
spectra were automatically collected under computer control
during the 30 min LC-MS/MS runs. Recorded MS/MS spectra were then subjected directly to SEQUEST database
15
without need for manual interpretation.
searches
SEQUEST was used to identify proteins in a spot by correlating experimental MS/MS spectra to protein sequences in
the M. jannaschii open reading frame database.
Results
Protein identification by peptide mass spectrometry
Capillary liquid chromatography-electrospray tandem
mass spectrometry (µ-LC-ESI-MS/MS) was used to identify
M. jannaschii proteins. Approximately 300 µg of protein
was loaded onto replicate gels. After second-dimension separation, the proteins were detected by staining with
Coomassie Blue followed by five cycles of destaining in
20% (v / v) ethanol. Protein spots to be identified were cut
from replicate gels and the proteins within 2–5 replicate
spots were digested in-gel with trypsin (Promega sequencegrade trypsin, 12.5 ng µg–1). The peptides were eluted from
the gel pieces by extracting three times, first with equal parts
of 25 mM ammonium bicarbonate and acetonitrile and then
twice with equal parts of 5% (v/v) formic acid and
acetonitrile. The tryptic digests were loaded onto a 365 µm
o.d. × 100 µm i.d. fused-silica capillary column packed with
either 10 µm POROS 10 R2 (PE Biosystems) to a length of
10–15 cm or 5 µm Zorbax XDB-C18 packing material
(Agilent Technologies) to a length of 7–9 cm. The tryptic
peptides were separated by using a 30 min linear gradient of
2–60% solvent (80% acetonitrile / 0.5% acetic acid) for
POROS 10R2 and of 0–60% solvent (80% acetonitrile / 0.02% heptafluorobutyric acid) for Zorbax-C18 and
were then introduced into an LCQ ion trap mass spectrome-
Comparison of proteins from M. jannaschii grown under
different conditions
Comparison of 2DE patterns of proteins from M.
jannaschii grown with control (100 kPa) and with deficient
(5 kPa) hydrogen partial pressures revealed numerous quantitative protein changes (Figure 1). Among these changes
was the almost complete disappearance of a relatively
intense protein spot at molecular weight 30 kDa (master spot
number [MSN] 1223) and a triplet of protein spots at molecular weight 27 kDa directly below (MSN 1224, 1225 and
1226) in cells grown with reduced hydrogen. Along with the
disappearance of these four proteins, four other proteins
(MSN 81, 119, 166 and 307) were observed consistently to
appear or increase significantly in abundance (Table 1).
In a comparison of M. jannaschii cells grown with control hydrogen partial pressure but harvested in either exponential or stationary phase, proteins MSN 81, 119, 166 and
307 were missing or just barely detectable in patterns from
exponential cells, but were significantly expressed in stationary cells [Figures 2(c) and (d)]. Unlike the hydrogendepleted cells [Figure 2(b)], however, the stationary cells
contained MSN 1223, 1224, 1225 and 1226 in amounts comparable to exponential cells. Cells harvested in exponential
Figure 1. The two-dimensional gel electrophoresis patterns of proteins from M. jannaschii cells grown under control and low
hydrogen partial pressure conditions. (a) Control; (b) low hydrogen (see Materials and Methods for details). Arrows indicate proteins present or absent under specific growth conditions. These proteins are shown in detail in Figure 2.
210
Structural Modifications of Methanococcus Jannaschii Flagellin Proteins
Table 1. Quantitative differences in M. jannaschii proteins after cell growth with control and deficient hydrogren concentrations. Master spot number, the protein identification number designated defined by matching x- and y-coordinates relative to the M. jannaschii
protein master pattern; average integrated density, the average of values from four to five replicated 2DE gels of the same protein sample; low hydrogen, 5 kiloPascals (kPa); control hydrogen, 100 kPa; ND, not detectable; lost, protein was detectable in patterns from
control cells but not in patterns from low hydrogen cells; induced, protein was not detectable in control cells but was detectable in low
hydrogen cells. The numbers in parentheses are the coefficients of variation (standard deviation divided by the mean times 100) for the
average integrated density values.
Master spot no.
Average integrated density: control hydrogen/low hydrogen
Ratio of control/low hydrogen
MSN0081
11450 (4%) / 119400 (8%)
0.96
MSN0119
4480 (12%) / 56480 (7%)
0.79
MSN0166
ND / 16680 (4%)
induced
MSN0307
ND / 17250 (25%)
induced
MSN1223
114020 (5%) / ND
lost
MSN1224
7220 (13%) / ND
lost
MSN1225
92100 (16%) / ND
lost
MSN1226
18870 (32%) / ND
lost
phase after growth with low ammonium (0.44 mM) [Figure
2(e)] contained minimal amounts of MSN 81, 119, 166, 307,
1223, 1224, 1225 and 1226. When cells still in exponential
growth were harvested 3 hours after the concentration of
ammonium was raised to 22 mM [Figure 2 (f)], all of these
proteins were detectable.
Identification of M. jannaschii proteins
The sequences of tryptic peptides from proteins MSN
81, 119, 166, 307, 1223, 1224, 1225 and 1226 were obtained
by µLC-MS/MS (Table 2). Using SEQUEST, the M.
jannaschii ORFs predicting tryptic peptide sequences
matching, with the highest confidence level, those observed
for the proteins analyzed were MJ0891 and MJ0892,
flagellin B1 and flagellin B2 precursors, respectively. The
M. jannaschii flagellin B1 and B2 proteins have a high
degree of sequence similarity with the Methanococcus
voltae flagellin B1 and B2 sequences, especially at the
amino and carboxy terminal ends (Figure 3). The amino termini include a sequence of 12 amino acids that could serve
as a leader sequence in the precursor proteins.16, 17 Multiple
amino acid sequences that could serve as N-glycosylation
17
sites (for example, Asn-X-Thr/Ser ) are also present in both
M. voltae and M. jannaschii flagellin B1 and B2 proteins.
M. jannaschii MSN 81 and MSN 1223 contained similar tryptic peptides matching those predicted for flagellin
B2, whereas MSN 1224, 1225 and 1226 (1225 and 1226
were analyzed by mass spectrometry as one protein due to
their proximity on the gels), 119, 166 and 307 all contained
tryptic peptides matching those predicted for flagellin B1
(Figure 4). The peptide of flagellin B1 containing the N-
glycosylation motif at residue 113 was detected in MSN 119,
but not in MSN 166, 307, 1224, 1225 or 1226. Similarly, the
flagellin B2 peptides with glycosylation sites at residues 95
and 171 were detected in tryptic digests of MSN 81, but not
in digests of MSN 1223. The observed molecular weights for
MSN 81 and MSN 1223 were 24.8 kDa and 30.2 kDa,
respectively, whereas the predicted molecular weight for M.
jannaschii flagellin B2 precursor (MJ0892) is 22.5 kDa. The
isoelectric points of MSN 81 and MSN 1223 were measured
as 4.98 and 4.38, respectively, compared with the predicted
isoelectric point of 5.4. The triplets of proteins identified as
flagellin B1 had observed molecular weights of 27.3, 27.5
and 27.6 kDa (MSN 1224, 1225 and 1226, respectively) or
23.4, 23.6 and 23.4 kDa (MSN 119, 166 and 307, respectively, although the predicted molecular weight of the
flagellin B1 precursor is 22.7 kDa. The isoelectric points of
the two triplet clusters were 4.31-4.46 (MSN 1224, 1225,
1226) and 4.63–4.73 (MSN 119, 166, 307), slightly less than
the predicted isoelectric point of 4.91.
Discussion
The flagella of most bacteria contain only one type of
protein subunit, with a few exceptions, such as Caulobacter
crescentus, which expresses two flagellin proteins (flagellins A and B), apparently assembled in tandem.18 In contrast, multiple genes encoding flagellin-like proteins have
2,17,19
been consistently revealed in archaeal genomes.
The
Halobacterium halobium genome contains five genes
encoding five structurally related flagellin proteins, two at
one locus (flagellins A1 and A2) and three at a separate locus
C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
211
Figure 2. The region of the two-dimensional gel electrophoresis patterns of M. jannaschii proteins from cells grown and harvested
under various conditions that shows dramatic differences in specific proteins. (a) Cells grown in 100 kPa hydrogen and harvested
in exponential phase [region of pattern in Figure 1(a)]; (b) cells grown in 5 kPa hydrogen and harvested in exponential phase
[region of pattern in Figure 1(b)]; (c) cells grown in 100 kPa hydrogen and harvested in exponential phase (different experiment
than cells for Figure 1); (d) same as (c), but harvested in stationary phase; (e) cells grown in 0.44 mM ammonium and harvested in
exponential phase; (f) cells grown in 22 mM ammonium and harvested in exponential phase. Numbers refer to the master spot
numbers for M. jannaschii proteins referred to in Table 1. The difference in the relative abundance of protein 1225 and protein
1226 in (a) compared with (c), both cell preparations harvested in exponential phase during two different experiments, remains to
be characterized. This difference could indicate variability in the modification process responsible for producing the isoelectric
point difference that is independent of growth conditions.
(flagellins B1, B2 and B3). Flagellins B1, B2 and B3 are sulfated glycoproteins with molecular weights of 26, 30 and
36 kDa according to SDS polyacrylamide gel electrophoresis migration patterns,20 although the gene sequences predict
molecular weights of approximately 20 kDa. High-resolution, one-dimensional SDS gels revealed a stair-step pattern
21
of protein bands when purified flagella were analyzed.
Deglycosylation of the H. halobium flagellin proteins produced proteins migrating as if they were 7 kDa smaller, still
20
inconsistent with their genome sequences. Thus, a protein
modification in addition to glycosylation has been suggested
for the H. halobium flagellin B proteins.
M. voltae has a multigene family encoding four flagellin
proteins, all with N termini that have a high degree of simi17,19
larity with the H. halobium flagellins. The conservation
of the flagellin amino terminus across the diverse branches
of the archaeal domain represented by Halobacterium and
Methanococcus has led to the hypothesis that these proteins
possess leader peptides,16 hence their annotation in some
sequence databases as flagellin precursors. The four M.
212
Structural Modifications of Methanococcus Jannaschii Flagellin Proteins
Table 2. Peptides detected by µLC-MS/MS of tryptic digests of M. jannaschii proteins. Master spot number, as described for Table 1; Peptide sequence matched, the peptide best matched by SEQUEST database search; Xcorr, the cross correlation score; Mj ORF, open reading frame in the M. jannaschii sequenced genome; pI, predicted isoelectric point; MW, predicted molecular weight; Protein name,
identification based on genome annotation.
a,b
Master spot no.
Peptide sequence matched
Xcorr
Mj ORF
pI
MW
Protein name
MSN 81
K.AINYLAIYITPNAGSAAIDLNQTK.I
L.AIYITLPNAGSAAIDLNQTK.I
N.ASAVGLNLVPR.T
5.4056
4.7986
MJ0892
5.40
22577
Flagellin B2
2.9426
3.2058
5.4866
4.1721
3.3848
3.4061
5.1703
6.1497
4.0556
MJ0891
4.91
22700
Flagellin B1
3.5357
3.9475
3.2181
4.3844
6.1616
3.6614
MJ0891
4.91
22700
Flagellin B1
MJ0891
4.91
22700
Flagellin B1
K.LAIYITPNAGSAPIDLK.N
3.4110
3.8005
4.0745
R.SEVSGQFQPEFGAPAVIQFTTPAAY.T
R.SEVSGQFQPEFGAPAVIQFTTPAAYTQTVIELQ.
R.TTPAAYTQTVIELQ.
4.9564
6.6269
3.7849
K.ESTEQVASGLSTLQVIGIHDNK.A
K.ILITDGEK.K
3.2328
2.0457
MJ0892
5.40
22577
Flagellin B2
K.ILITDGEKK.A
K.STTPVINK.G
R.TTVTGSVIPEFGAPAVIEFTTPAAYLSTQEVIQLQ.
2.7042
1.5499
2.1959
K.GDIAVLLVNANAVFNK.A
R.SEVSGQFQPEFGAPAVIQFTTPAAY.T
4.6035
5.0167
MJ0891
4.91
22700
Flagellin B1
R.SEVSGQFQPEFGAPAVIQFTTPAAYTQTVIELQ.
4.0840
K.ESTEQVASGLSTLQVIGIHDNK.A
G.FLQQKAMATGKESTEQVASGLSTLQVIGIHDNK.A
K.GDIVALTINASAVGLNLVPR.T
K.ILITDGEKKAVLR.Y
L.STLQVIGIHDNK.A
K.STTPVINKGDIVALTINASAVGLNLVPR.T
L.TINASAVGLNLVPR.T
R.TTVTGSVIPEFGAPAVIEFTTPAAY.L
R.TTVTGSVIPEFGAPAVIEFTTPAAYLSTQEVIQLQ
MSN 119
#
K.ESTEQVASGLM CIGVTGHYDK.T
#
K.ESTEQVASGKM CIGVTGHYDK.T
K.GDIAVLLVNANAVFNKAIPTR.S
K.LAIYITPNAGSAPIDLK.N
K.LFLIYDGESHVLN.Y
K.LFLIYDGESHVLNYSTVTTA.T
R.SEVSGQFQPEFGAPAVIQFTTPAAY.T
R.SEVSGQFQPEFGAPAVIQFTTPAAYTQTVIELQ.
R.TTPAAYTQTVIELQ.
MSN 166
#
K.ESTEQVASGLM C*IGVTGHYDK.T
#
K.ESTEQVASGKM CIGVTGHYDK.T
R.SEVSGQFQPEFGAPAVIQF.T
R.SEVSGQFQPEFGAPAVIQFTTPAAY.T
R.SEVSGQFQPEFGAPAVIQFTTPAAYTQTVIELQ.
R.TTPAAYTQTVIELQ.
MSN 307
#
K.ESTEQVASGLM CIGVTGHYDK.T
#
K.ESTEQVASGKM CIGVTGHYDK.T
MSN 1223
MSN 1224
3.3900
4.2388
3.3634
4.5540
3.3351
3.4069
5.5184
3.6628
3.7208
4.9032
C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
213
Table 2. Peptides detected by µLC-MS/MS of tryptic digests of M. jannaschii proteins. Master spot number, as described for Table 1; Peptide sequence matched, the peptide best matched by SEQUEST database search; Xcorr, the cross correlation score; Mj ORF, open reading frame in the M. jannaschii sequenced genome; pI, predicted isoelectric point; MW, predicted molecular weight; Protein name,
identification based on genome annotation (continued).
a
Master spot no.
Peptide sequence matched
MSN 1225 and
1226
a,b
Xcorr
Mj ORF
pI
MW
Protein name
K.GDIAVLLVNANAVFNK.A
K.GDIAVLLVNANAVFNKAIPTR.S
3.5855
2.6567
MJ0891
4.91
22700
Flagellin B1
R.SEVSGQFQPEFGAPAVIQF.T
R.SEVSGQFQPEFGAPAVIQFTTPAAY.T
R.SEVSGQFQPEFGAPAVIQFTTPAAYTQTVIELQ.
5.4398
4.6284
4.0281
#
M , oxidized methionine; during SEQUEST database searches, these peptides were matched using the assumption of a mass shift of 16 on
each methionine residue due to oxidation occurring during sample preparation and processing.
b *
C , iodoacetamide-modified cysteines; during SEQUEST database searches, these peptides were matched using the assumption of a mass
shift of 57 on each cysteine residue due to alkylation by iodoacetamide.
voltae flagellin genes encode flagellin A and flagellin B1,
B2 and B3, homologous to the H. halobium flagellin A1, A2,
B1, B2 and B3 genes, respectively. Of the four types of
flagellins, FlaB1 and FlaB2 are the major flagellin proteins
expressed in M. voltae.21 As in H. halobium, the genes
encoding the M. voltae flagellins B1 and B2 predict proteins
with molecular weights of 22.4 kDa and 22.8 kDa, respectively, which is considerably smaller than the 33 kDa and
31 kDa molecular weights observed on SDS gels for purified
22
flagellar filaments from M. voltae. All of the M. voltae
flagellin proteins contain internal glycosylation signals
(Asn-X-Thr/Ser) (at positions 38, 71, 77, 115 and 136 in
flagellin B1 and positions 38, 72, 77, 113, 172 and 208 in
flagellin B2) analogous to those in H. halobium, although
there is no evidence that the M. voltae flagellins are actually
glycosylated.19 As in H. halobium, removal of glycosyl
Figure 3. Protein sequence alignments of flagellins B1 and B2 from M. jannaschii and M. voltae. MJ0891, amino acid sequence predicted for M. jannaschii flagellin B1; FLA1_METVO, amino acid sequence predicted for M. voltae flagellin B1; MJ0892, amino acid
sequence predicted for M. jannaschii flagellin B2; FLA2_METVO, amino acid sequence predicted for M. voltae flagellin B2. The box
indicates the 12 amino-terminal amino acids thought to represent a leader sequence in the precursor form of the archaeal flagellin
B proteins. The putative glycosylation sites (Asn-X-Thr/Ser) are underlined in each sequence. Dashes are used to provide optimal
alignment of the M. voltae and M. jannaschii sequences.
214
Structural Modifications of Methanococcus Jannaschii Flagellin Proteins
Figure 4. Protein sequence alignments of M. jannaschii flagellins B1 and B2 based on open reading frames MJ0891 and MJ0892
with the putative sequences of MSN 81, 119, 166, 307, 1223, 1224, 1225 and 1226. The peptides corresponding to the peptide
masses observed for each protein (Table 2) are shown in bold italicized type. The box indicates the 12 amino-terminal amino acids
thought to represent a leader sequence in the precursor form of the archaeal flagellin B proteins. The putative glycosylation sites
(Asn-X-Thr/Ser) are underlined in each sequence. Numbers indicate the amino acid residue location within the sequence.
groups, if they were present, would be insufficient to explain
the size discrepancy between the predicted and observed
SDS molecular weights for the M. voltae flagellin B1 and B2
proteins, suggesting that other protein modifications contribute to the molecular weights of these molecules.
The M. jannaschii genome encodes three flagellin proteins (FlaB1, B2 and B3) with a significant overall similarity
to the respective M. voltae flagellins and with amino terminal sequence homology to the H. halobium flagellins.2,17,19
Our 2DE analysis of the M. jannaschii proteome has
C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
revealed that M. jannaschii cells can express multiple forms
of both flagellin B1 and B2 proteins, depending, in a seemingly complex way, on cellular growth conditions. The
flagellin B2 proteins (seen in 2DE patterns as MSN 81 or
1223) appear as relatively abundant, well-resolved protein
spots differing in molecular weight by 5 kDa and in
isoelectric point by 0.6. In the presence of control levels of
hydrogen, exponentially growing cells express the higher
molecular weight form (MSN 1223) in approximately 10fold more abundance than they express the lower molecular
weight and more alkaline form (MSN 81) [Figures 2(a) and
(c)]. Cells growing exponentially in 20-fold less hydrogen,
in contrast, had no detectable MSN 1223 and 10-fold more
MSN 81 relative to the control cells. The M. jannaschii proteins identified as flagellin B1, based on their tryptic peptide
masses, varied in abundance in concert with the changes
observed in the flagellin B2 proteins. Resolved as a triplet of
proteins with slightly different isoelectric points, the higher
molecular weight flagellin B1 proteins (MSN 1224, 1225
and 1226) in the control cells were abundant whereas the
lower proteins were barely detectable (MSN 119) or undetectable (MSN 166 and MSN 307). The reverse was true in
cells grown with a decreased hydrogen partial pressure; the
lower molecular weight flagellin B1 proteins were abundant
whereas the higher molecular weight proteins were undetectable. Mukhopadhyay and co-workers observed the
appearance and disappearance of two M. jannaschii proteins
identified as B-type flagellins in response to cell growth at
different hydrogen partial pressures.5 These proteins were of
the same apparent molecular weights as MSN 1223, 1224,
1225 and 1226 in our 2DE patterns and they were expressed
most strongly under the high-hydrogen growth condition, as
observed in our experiments. The separation of the proteins
in their report was insufficient to resolve the charge heterogeneity seen in the 2DE patterns reported here and the background staining precluded the observation of the lower
5
molecular weight forms of flagellins B1 and B2.
M. jannaschii cells harvested in stationary phase [Figure 2 (d)] and in exponential phase after a shift from limited
to control levels of ammonium [Figure 2 (f)] showed expression of approximately equal amounts of all forms of
flagellins B1 and B2, indicating either that expression of the
high- or low-molecular weight forms is not mutually exclusive within the cells or that a mixed population of cells was
harvested in the latter experiments. A third group of triplet
protein spots, as yet unidentified but similar in configuration
to MSN 119, 166 and 307 as well as MSN 1224, 1225 and
1226, appeared approximately halfway between these two
sets of triplets in cells from both the stationary and ammonium-enriched cultures. This triplet could represent an additional form of the flagellin B1 protein having a molecular
mass intermediate between that observed for proteins MSN
1224, MSN 1225 and MSN 1226 and that seen for proteins
MSN 119, MSN 166 and MSN 307. After growth on limited
ammonium, M. jannaschii cells were observed to express
only minimal amounts of flagellin B1 or B2 proteins [Figure
215
2(e)], suggesting that flagellin synthesis can be essentially
shut down under specific growth conditions.
The flagellin B1 proteins represented by MSN 1224,
MSN 1225 and MSN 1226 and the flagellin B2 proteins represented by MSN 1223 have apparent molecular weights of
approximately 27 and 30 kDa, respectively, in the presence
of SDS, although the predicted molecular weights for both
proteins are approximately 21 kDa. This size discrepancy is
almost identical to that reported for both H. halobium and M.
voltae19,20,22 and cannot be accounted for by the inclusion of
the leader sequence of only 1.4 kDa. Actually, no tryptic
peptides with masses similar to those predicted for the leader
sequence were found for any of the M. jannaschii proteins
analyzed, suggesting that the expressed proteins observed in
2DE patterns lack the precursor leader sequence. Flagellin
proteins, due to their association with the cell membrane as
part of the flagellar structure, have hydrophobic domains
that may bind SDS inefficiently, thus altering their migration
position in the electrophoresis pattern. Alternatively, as yet
undiscovered post-translational modifications, specific to
the archaeal flagellin B1 and B2 structures, may cause the
higher molecular weight migration of these proteins, since
similar discrepancies between the SDS molecular weights
and predicted molecular weights have been reported for H.
halobium20 and M. voltae19,23 and, now, for M. jannaschii.
Like the flagellin B1 and B2 sequences of H. halobium
and M. voltae, the predicted protein sequence for M.
jannaschii flagellins B1 and B2 includes glycosylation signals (Asn-X-Thr/Ser). Whereas M. voltae has five glycosylation sites in the flagellin B1 sequence and six in the
flagellin B2 sequence, M. jannaschii has only three signals
in the flagellin B1 sequence (at residues 38, 113 and 129)
and five in the flagellin B2 sequence (at residues 38, 95, 128,
135 and 171).
It is notable that the glycosylation site at residue 38
appears in both flagellin B1and B2 in both M. voltae and M.
jannaschii, whereas the remainder of the potential glycosylation sites appear to be species- and flagellin-specific.
Similar to M. voltae19 and in contrast to H. halobium,24 there
is no direct evidence for glycosylation of the M. jannaschii
flagellin B1 and B2 proteins. The lack of charge heterogeneity for flagellin B2 (MSN 81 and 1223), with five possible
glycosylation sites, suggests there are no post-translational
modifications of this protein (such as the glycosylation with
sulfated saccharides observed for all H. halobium flagellin
proteins24) that would affect the isoelectric point. The triplets
of flagellin B1 (MSN 1224, 1225 and 1226 and MSN 119,
166 and 307), however, do suggest a post-translational modification that causes isoelectric point heterogeneity and subtle differences in molecular weight.
The tryptic digests from MSN 119 and MSN 81 contained peptides with the N-glycosylation motifs at residues
113 and 95 and 171, respectively, whereas the remainder of
the proteins analyzed did not. This observation is consistent
with these two proteins being the least modified forms, i.e.
having the most unmodified peptides, in the groups of
216
Structural Modifications of Methanococcus Jannaschii Flagellin Proteins
flagellin B1 and B2 proteins separated by 2DE. The more
acidic, higher molecular weight proteins could be expected,
conversely, to have more modified peptides, which would
not match the predicted peptide masses. Further study will
be required to confirm whether the observed charge and
molecular weight heterogeneity is due to sulfated glycoproteins such as those found in H. halobium22 or to some
other type of modification such as deamidation or phosphorylation.
The molecular weight and isoelectric point differences
between the two populations of flagellin B1 and B2 proteins
suggest (a) varying degrees of protein–protein association
(for example, flagellin protein homodimers that remain
together under the denaturation conditions used for sample
preparation) that vary with cell growth conditions, (b) the
occurrence of a proteolytic event that results in the removal
of a peptide sequence from each of these proteins, or (c) the
synthesis of proteins lacking a portion of the predicted protein sequence under specific growth conditions. The same
sample preparation method has been used for all experiments described, suggesting that the dissociation efficiency
of possible flagellin complexes should be consistent. Thus,
if the difference in molecular weight of the flagellins
observed is related to the dissociation of protein complexes,
that dissociation must be a cellular event. The observed
molecular weights do not correlate with any obvious combination of flagellin B1 or flagellin B2 homocomplexes, however. A proteolytic event, i.e. a post-translational modification, is expected to cause the loss of a series of amino acids
from the amino or carboxy terminal end of the flagellin proteins which would total masses of approximately 4 kDa and
5 kDa for flagellin B1 and B2, respectively, based on the
positions of the respective proteins in the 2DE patterns. The
complete amino acid sequences of flagellins B1 and B2 were
not covered by the mass spectrometry analyses done here, so
proteolysis cannot be completely ruled out as a reason for the
appearance of flagellin proteins with different molecular
weights. However, tryptic peptides with masses comparable
to those expected from the carboxy termini of both proteins
were found in preparations from all of the proteins identified
as flagellin B1 and B2, both high and low molecular weight.
Thus, no carboxy terminal peptides were consistently missing from any of the proteins studied. No tryptic peptides with
masses consistent with the predicted amino terminal
sequences of the flagellin B1 and B2 proteins were found in
any of the preparations analyzed, either the high- or the lowmolecular-weight forms. Thus, no differences indicative of
proteolysis from the amino terminal end of the proteins were
observed either. Therefore, our observations thus far suggest
that M. jannaschii cells possess a regulatory mechanism that
directs the synthesis of multiple forms of the flagellin B1 and
B2 proteins having different molecular weights under different growth conditions.
The observation of differential expression of multiple
forms of proteins identified as the flagellin B1 and B2 proteins in M. jannaschii suggests a regulatory mechanism in
the Archaea for the assembly and function of flagella. Further characterization of the structural differences between
these multiple forms of flagellins B1 and B2 will require
additional investigations into the chemical composition of
the post-translational modifications resulting in the charge
heterogeneity of flagellin B1 but not flagellin B2 and into the
amino acid composition of the different molecular weight
variants of the two proteins. Such investigations will provide
critical insight into the regulation of cell motility in this
poorly understood branch of life.
Acknowledgments
The authors thank Eric F. Johnson for his expert assistance with the growth of M. jannaschii in the University of
Illinois Fermentation Facility. The research described in this
paper was supported by the Microbial Genome Program in
the Office of Biological and Environmental Research,
United States Department of Energy, under contract No. W31-109-ENG-38.
References
1. W. Jones, J.A. Leigh, F. Mayer, C.R. Woese and R.S.
Wolfe, Arch. Microbiol. 136, 254 (1983).
2. C.J. Bult, O. White, G.J. Olsen, L. Zhou, R.D.
Fleischmann, G.G. Sutton, J.A. Blake, L.M. Fitzgerald,
R.A. Clayton, J.D. Gocayne, A.R. Kerlavage, B.A.
Dougherty, J.-F. Tomb, M.D. Adams, C.I. Reich, R.
Overbeek, E.F. Kirkness, K.G. Weinstock, J.M. Merrick,
A. Glodek, J.L. Scott, N.S.M. Geoghagen, J.F.
Wiedman, J.L. Fuhrmann, D. Nguyen, T.R. Utterback,
J.M. Kelley, J.D. Peterson, P.W. Sadow, M.C. Hanna,
M.D. Cotton, K.M. Roberts, M.A. Hurst, B.P. Kaine, M.
Borodovsky, H.-P. Klenk, C.M. Fraser, H.O. Smith, C.R.
Woese and J.C. Venter, Science 273, 1058 (1996).
3. S. Koyasu, M. Asada, A. Fukuda and Y. Okada, J. Mol.
Biol. 153, 471 (1981).
4. D.P. Bayley and K.F. Jarrell, J. Mol. Evol. 46, 370
(1998).
5. B. Mukhopadhyay, E.F. Johnson and R.S. Wolfe, Proc.
Natl. Acad. Sci. USA 97, 11522 (2000).
6. L.S. Ramagli and L.V. Rodriguez, Electrophoresis 6,
559 (1985).
7. N.G. Anderson and N.L. Anderson, Anal. Biochem. 85,
331 (1978).
8. P.H. O’Farrell, J. Biol. Chem. 250, 4007 (1975).
9. N.L. Anderson and N.G. Anderson, Anal. Biochem. 85,
341 (1978).
10. N.L. Anderson, S.L. Nance, S.L. Tollaksen, F.A. Giere
and N.A. Anderson, Electrophoresis 6, 592 (1985).
11. C.S. Giometti, M.A. Gemmell, S.L. Tollaksen and J.
Taylor, Electrophoresis 12, 536 (1991).
C.S. Giometti et al., Eur. J. Mass Spectrom. 7, 207–217 (2001)
12. N.L. Anderson, J. Taylor, A.E. Scandora, B.P. Coulter
and N.G. Anderson, Clin. Chem. 27, 1807 (1982).
13. C.S. Giometti and J. Taylor, in Advances in Electrophoresis, Ed by M.J. Dunn. Walter de Gruyter and Co., New
York, USA, pp. 359–389 (1991).
14. C.S. Giometti, J. Taylor, M.A. Gemmell, S.L. Tollaksen,
N.D. Lalwani and J.K. Reddy, Appl. Theoret. Electrophoresis 2, 101 (1991).
15. J.K. Eng, A.L. McCormack and J.R. Yates III, J. Am.
Soc. Mass Spectrom. 5, 976 (1994).
16. M.L. Kalmokoff, T.M. Karnauchow and K.F. Jarrell,
Biochem. Biophys. Res. Comm. 167, 154 (1990).
17. M.L. Kalmokoff and K.F. Jarrell, J. Bacteriol. 173, 7113
(1991).
18. A. Fukuda, M. Asada, S. Koyasu, H. Yoshida, K.
Yaginuma and Y. Okada, J. Bacteriol. 145, 559 (1981).
217
19. M.L. Kalmokoff, K.F. Jarrell and S.F. Koval, J.
Bacteriol. 170, 1752 (1988).
20. L. Gerl and M. Sumper, J. Biol. Chem. 260, 13246
(1988).
21. K.F. Jarrell, D.P. Bailey, V. Florian and A. Klein, Mol.
Microbiol. 20, 657 (1996).
22. M. Alam and D.J. Oesterhelt, Mol. Biol. 176, 459 (1984).
23. M.L. Kalmokoff, K.F. Jarrell and S.F. Koval, J.
Bacteriol. 170, 1752 (1988).
24. F. Weiland, G. Paul and M. Sumper, J. Biol. Chem. 260,
15180 (1985).
Received: 1 February 2001
Revised: 30 March 2001
Accepted: 23 May 2001
Web Publication: 12 July 2001