Experimental proof for a signal peptidase I like activity
in Mycoplasma pneumoniae, but absence of a gene
encoding a conserved bacterial type I SPase
Ina Catrein, Richard Herrmann, Armin Bosserhoff and Thomas Ruppert
Zentrum fu
¨
r Molekulare Biologie Heidelberg, Universita
¨
t Heidelberg, Germany
Mycoplasma pneumoniae is a human pathogenic bac-
terium [1,2], characterized by a small genome of 816
kbp [3], the lack of a bacterial cell wall and a parasitic
lifestyle [4].
Some species of the genus mycoplasma, e.g. Myco-
plasma genitalium, Mycoplasma gallisepticum and
Mycoplasma pneumoniae exhibit a flask-like shape,
which is believed to be formed and maintained by a
cytoskeleton-like structure [5–10]. This flask-like shape
is caused by the attachment organelle, an asymmetric
extension of the cell composed of an assembly of
unique proteins [11]. M. pneumoniae interacts with its
host cell by adhering with the attachment organelle to
specific receptors. This interaction takes place only if
the P1 protein, the bacterial main adhesin, is inserted
correctly into the attachment organelle [12]. The
proper insertion depends, among others, on the pro-
teins P40 and P90. Absence of these proteins causes a
random insertion of the P1 protein and a cytadher-
ence-negative phenotype [13]. P1 is encoded by
MPN141 and both, P40 and P90 by MPN142. These
genes are organized, together with MPN140 which

Germany
Fax: +49 6221 545891
Tel: +49 6221 546895
E-mail:
(Received 15 December 2004, revised
11 March 2005, accepted 7 April 2005)
doi:10.1111/j.1742-4658.2005.04710.x
Although the annotation of the complete genome sequence of Mycoplasma
pneumoniae did not reveal a bacterial type I signal peptidase (SPase I) we
showed experimentally that such an activity must exist in this bacterium, by
determining the N-terminus of the N-terminal gene product P40 of MPN142,
formerly called ORF6 gene. Combining mass spectrometry with a method
for sulfonating specifically the free amino terminal group of proteins, the
cleavage site for a typical signal peptide was located between amino acids 25
and 26 of the P40 precursor protein. The experimental results were in agree-
ment with the cleavage site predicted by computational methods providing
experimental confirmation for these theoretical analyses.
Abbreviations
CID, collision induced fragmentation; SPase I, signal peptidase I.
2892 FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS
experimentally. The P1 protein can be cross-linked
with P40 and P90 under in vivo conditions, indicating
that they form a larger protein complex embedded into
the cytoplasmic membrane [20,21].
To get insight into the function of this protein com-
plex, the subunits had to be characterized in detail.
The N-terminal end of P90 was determined by Edman
degradation. This protein begins with an arginine at
amino-acid position 455 of the proposed precursor
[22]. The predicted molecular mass of P40 is therefore

To test experimentally whether there is a type I
SPase activity in M. pneumoniae, we determined the
N-terminus of P40 as it appears in protein extracts of
M. pneumoniae.
Determination of the N-terminus of P40 by Edman
degradation failed due to the limited amount of start-
ing material. An alternative method is the tryptic
digestion of the purified protein and subsequent ana-
lysis of the derived peptides by mass spectrometry. If
a candidate mass is detected and peptide sequencing
proves, that there is no trypsin cleavage site at the
amino terminus of this peptide, as read from the gene
sequence, then this peptide is taken as the N-terminal
peptide of this protein [27]. This is, however, not an
exact proof, as such a cleavage may also result from
proteolytic contaminants of the trypsin being used like
chymotrypsin or due to pseudotrypsin formed from
trypsin by autolysis [28]. These possibilities can be
excluded, when the amino terminus of the intact pro-
tein is specifically labeled before tryptic digestion.
Liminga and colleagues [29] introduced an N-
hydroxysuccinimide ester of 3-sulfonic-propionic acid
(CAF reagent) to modify peptides after tryptic diges-
tion for enhanced peptide sequencing using matrix-
assisted laser desorption ⁄ ionization time of flight mass
spectrometry (MALDI-TOF-MS) with chemical assis-
ted fragmentation (CAF) [30]. We modified this
method in a way so that only the N-terminus of the
mature P40 was labeled with the CAF reagent. After
tryptic digestion, the N-terminal sulfonated peptide of

Subsequently, the protein was digested by trypsin
and the supernatant was analyzed by ESI-QTOF MS
(Fig. 3) in the positive ion mode, to get high quality
fragmentation pattern for peptide sequencing. Fifteen
I. Catrein et al. Signal peptidase I activity in M. pneumoniae
FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS 2893
signals of various m ⁄ z-values from the peptide mass
fingerprint were subjected to collision induced frag-
mentation (CID). P40 was identified (probability
based Mowse score of 276) by sequence tags of six
different peptides covering the N-terminal region
(position 26–203) of P40 (Table 1). The fourfold
charged signal with m ⁄ z ¼ 850.2 (rmm: 3396.8) corre-
lated with the mass of the peptide 26–55 (calculated
rmm: 3260.8), if an increase in the molecular
mass by 136 Da, the mass of the sulfonic acid modi-
fication, is taken into account. The identity of this
peptide was confirmed by collision induced fragmen-
tation (CID). The complex fragmentation pattern of
the fourfold protonated peptide shows single, double
and triple charged fragment ions, which can be easily
distinguished from the isotopic pattern. After
deconvolution, the peptide was identified by the
almost complete series of y-fragments and a long
series of b-fragments (Fig. 4). From the difference of
the precursor mass to the y-28 fragment and from
the b1 fragment, which both represent the N-terminal
part of the peptide, it is evident that the N-terminal
amino acid is asparagine (position 26), but increased
in mass by 136 Da. Because this modification is only

In a first reaction lysine residues are specifically converted to
homoarginine by O-methylisourea. Then, the free a-amino group at
the N-terminus is sulfonated by the CAF reagent.
12 3
kDa
250
150
100
75
50
37
25
15
Fig. 1. Enrichment of P40 for mass spectrometric analysis. The pro-
teins were separated by SDS ⁄ PAGE (12.5%). Lane 1, molecular
mass marker; lane 2, proteins (1.5 lg) enriched by immunoprecipita-
tion; lane 3, total cell extract of M. pneumoniae (7 lg). The gel was
stained with colloidal Coomassie blue. The protein band, which was
cut out for mass spectrometric analysis is indicated by an arrow.
Signal peptidase I activity in M. pneumoniae I. Catrein et al.
2894 FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS
with m ⁄ z of 436.3, threefold charged. It turned out
that this peptide 71–81 (VGDTKLVALVR) contained
a lysine in the middle of the sequence, also modified
by sulfonation (Fig. 6).
Other amino acids were hardly, if at all, influenced
by this two step labeling procedure. The oxidation of
tryptophan or the conversion of asparagine to aspartic
acid, as observed for peptide 146–152 (ATWVFER)
and peptide 204–226 (VNGVAQDTVHFGSGQESSW

(+136 Da). Most of the unlabelled fragments in the spectrum are y- and b-ions after neutral loss of H
2
O from the side chains of aspartic acid
and threonine.
I. Catrein et al. Signal peptidase I activity in M. pneumoniae
FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS 2895
proteolytic processing steps, are necessary. Edman de-
gradation is a widely used technique for determining
the N-terminus of a protein. This method, however, is
not only limited by its inability to deal with amino ter-
minal modified proteins, but also by its poor sensitivity
compared to mass spectrometry. On the other hand,
mass spectrometry is still not very efficient for analyz-
ing whole proteins despite the progress made during
the last years. Such an analysis normally begins
with the proteolytic degradation of a protein, followed
by the analysis of the cleavage products. It is, how-
ever, difficult to prove which of the identified peptide
represents the amino terminus of the protein. To over-
come this problem, we labeled the amino terminus of
the protein after electrophoretic separation within the
excised gel piece.
As shown, the N-terminal amino group of P40 could
be sulfonated by the CAF reagent. After tryptic diges-
tion, the N-terminal sulfonated peptide 26–54 was
identified by collision induced dissociation of the four-
fold protonated peptide ion. The charge state of this
peptide is unexpected: a peptide containing two basic
amino acids (arginine and histidine) and a blocked
amino terminus should not acquire more than two

(Fig. 3).
In the peptide mass fingerprint, the N-terminal sulfo-
nated peptide 26–54 was detected not only in its pro-
tonated form, but also as a sodium and potassium
adduct ion (Fig. 3). Interestingly, after collision
induced dissociation of these peptide ions, only y-frag-
ments were observed. The sodium and potassium ions,
respectively, were most likely bound at the deproto-
Fig. 5. SIGNAL P predicted and experimentally verified cleavage site
for P40. The cleavage site for the P40 precursor protein was
predicted by
SIGNAL P (Gram-positive network) to be located
between amino acid 25 (alanine) and 26 (asparagine). The values of
the C- (output from cleavage site networks), S- (output from signal
peptide network) and Y-scores (combined cleavage site score) are
shown for each position in the sequence. The data were generated
by feeding the first 50 amino acids of the gene products of
MPN142 to the publicly available web server http://www.
cbs.dtu.dk/services/SignalP/. For more details see [24,25]. The
experimentally defined N-terminus of the mature P40 agreed with
this prediction (black arrow).
Fig. 6. Fragmentation spectrum of the sulfonated peptide 70–80.
The doubly charged peptide ion with m ⁄ z ¼ 653.9 showed intense
sodium and potassium adducts as observed for the sulfonated
N-terminal peptide 26–54. From the fragmentation pattern (y-frag-
ments are indicated by arrow) the peptide was identified as
70-VGDTKLVALVR-80 containing a lysine in the middle of the
sequence, which is probably modified by sulfonation (+136 Da).
Signal peptidase I activity in M. pneumoniae I. Catrein et al.
2896 FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS

about 9 kDa between the calculated and the actually
observed masses (Fig. 7). It seems reasonable to
assume, that additional proteolytic cleavage took place.
Without further information, we can presently not
decide which were the true intermediates in this pro-
cess. There could be a premature P40 with an extended
C-terminal region, but a premature P90 with an exten-
ded N-terminal region can not be excluded (Fig. 7).
The best way to analyze this precursor-product rela-
tionship would be through puls-chase experiments
with radioactively labeled amino acids. This approach,
however, is hampered by the lack of a defined minimal
medium for M. pneumoniae.
Several predictive methods have been published
[23,24], which facilitate the identification of prokaryotic
signal peptides. A recent study [23] of bacterial and
archaeal proteomes proposed that the fraction of puta-
tive exported or secreted proteins ranges from 8%
(Methanococcus jannaschii) to 37% (M. pneumoniae).
This means that in M. pneumoniae 254 from 688 predic-
ted proteins contain a signal peptide. Establishing and
improving such prediction methods depend on the
availability of experimental data. Although Nielsen
et al. [24] explicitly excluded members of the cell wall-
less mollicutes from their analysis of signal peptides,
applying their program signal p [25,33] precisely pre-
dicted the cleavage site for P40 (Fig. 5). This result was
confirmed by the program exprot [23]. Both methods
failed to confirm the cleavage site of the main adhesine
P1 of M. pneumoniae. This is the only other M. pneu-

MPN294 were proposed as possible candidates enco-
Fig. 7. Schematic model for processing of the gene product of
MPN 142. Cleavage into P40 and P90 takes place after amino acid
454. The N-terminus of the mature P40 starts with amino acid 26.
The molecular mass of about 36 000 Da of P40, as determined by
SDS ⁄ PAGE would correspond to a protein reaching from amino
acid 26–365 (P40
340
). According to this scheme a premature P40
(amino acid 26–454) could be an intermediate as well as a pre-
mature P90 (amino acid 366–1218).
I. Catrein et al. Signal peptidase I activity in M. pneumoniae
FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS 2897
ding a SPase I like activity [16]. Using recombinant
P40 preprotein or suitable synthetic peptide sub-
strates [26] subfractions of protein extracts from
M. pneumoniae can now be assayed for signal pepti-
dase I activity facilitating the identification of the
corresponding gene ⁄ protein.
Experimental procedures
Bacteria
M. pneumoniae reference strains M129 (ATCC 29342; sub-
type 1; broth passage no. 31) and FH (ATCC 15531; sub-
type 2 [41]; broth passage no. 5) were cultivated as
described previously [42] and stored at )80 °C. For sample
preparation, M. pneumoniae cells were grown adherently at
37 °C in 100 mL of modified Hayflick medium [43] in
150 cm
2
cell culture flasks (Greiner, Flacht, Germany).

fer. The concentration of this suspension was determined
by reading the absorbance at 280 nm and using an extinc-
tion coefficient of 1.4 for a 1 mgÆmL
)1
solution of IgG. We
obtained from 2 mL of serum about 10 mg protein, mainly
IgG as revealed by SDS ⁄ PAGE. The IgG-enriched fraction
retained the ability to recognize P40 in western blotting
experiments (I. Catrein, unpublished results). SDS ⁄ PAGE
and western blotting were performed as published recently
[44].
Cross-linking of IgG to magnetic beads
The enriched IgG fraction (10 mg) was bound to 2 mL sus-
pension of magnetic beads, which carried recombinant Pro-
tein A covalently attached (DynabeadsÒ Protein A, Dynal
Biotech, Oslo, Norway) following the instructions of the
manufacturer.
Isolation of P40 from cell extracts
M. pneumoniae M129 from 10 cell culture flasks (150 cm
2
,
100 mL modified Hayflick medium) were collected, washed
twice with phosphate buffer and the pellet suspended in
2 mL lysis buffer (500 mm NaCl, 50 mm Tris ⁄ HCl pH 7.5,
0.5% Triton X-100 and the protease inhibitor cocktail com-
plete, EDTA-free (according to the manufactures recom-
mendation; Roche, Basel, Switzerland). The bacteria were
sonicated with a Branson sonifier for 8 · 15 s at 4 °C with
intervals of 1 min and the suspension separated in pellet
and supernatant by centrifugation (60 min, 75 000 g)ina

ted with water for 15 min. After removal of excess water,
100 lLofO-methylisourea hemisulfate (140 mm in 200 mm
sodium bicarbonate, pH 10) was added and incubated over
night at room temperature. The supernatant was removed,
the gel piece washed twice with water and shrunk in aceto-
nitrile on ice. After removal of acetonitrile, 6 mg CAF rea-
gent (chemical assisted fragmentation reagent; kindly
provided by Amersham Biosciences, Freiburg, Germany)
(Fig. 1) dissolved in 60 lL 250 mm sodium bicarbonate,
pH 9.4, was added. After 10 min on ice the Eppendorf tube
was quickly brought to room temperature for additional
5 min. Then, the reaction was stopped by addition of 2 lL
50% hydroxylamine solution. After 1 h the supernatant
was removed and the gel piece washed twice with water,
25 mm ammonium bicarbonate, pH 8.5. After shrinkage in
acetonitrile, the gel piece was incubated with 15 lL25mm
ammonium bicarbonate, pH 8.5 containing 140 ng modified
porcine trypsin (8 ngÆlL
)1
) (Promega, Madison, WI, USA)
for 4 h at 37 °C. Digestion was stopped by formic acid
Signal peptidase I activity in M. pneumoniae I. Catrein et al.
2898 FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS
(final concentration: 2%) and the sample was stored at
)20 °C. MS analysis was performed on a Q-TOF mass
spectrometer (Applied Biosystems, Darmstadt, Germany)
equipped with a nano-ESI ion source (Protana, Odense,
Denmark) as described previously [27]. If indicated,
MS ⁄ MS spectra were deconvoluted using Bayesian Peptide
Reconstruct (mass tolerance: 0.1 Da; S ⁄ N threshold: 2) pro-

tures of Mycoplasma pneumoniae. J Bacteriol 102, 855–
861.
8 Meng KE & Pfister RM (1980) Intracellular structures
of Mycoplasma pneumoniae revealed after membrane
removal. J Bacteriol 144, 390–399.
9 Wilson MH & Collier AM (1976) Ultrastructural study
of Mycoplasma pneumoniae in organ culture. J Bacteriol
125, 332–339.
10 Hegermann J, Herrmann R & Mayer F (2002) Cytoske-
letal elements in the bacterium Mycoplasma pneumoniae.
Naturwissenschaften 89, 453–458.
11 Krause DC & Balish MF (2004) Cellular engineering in
a minimal microbe: structure and assembly of the term-
inal organelle of Mycoplasma pneumoniae. Mol Micro-
biol 51, 917–924.
12 Feldner J, Gobel U & Bredt W (1982) Mycoplasma
pneumoniae adhesin localized to tip structure by mono-
clonal antibody. Nature 298, 765–767.
13 Baseman JB, Cole RM, Krause DC & Leith DK (1982)
Molecular basis for cytadsorption of Mycoplasma pneu-
moniae. J Bacteriol 151, 1514–1522.
14 Aravind L & Koonin EV (1998) A novel family of pre-
dicted phosphoesterases includes Drosophila prune pro-
tein and bacterial RecJ exonuclease. Trends Biochem Sci
23, 17–19.
15 Inamine JM, Loechel S & Hu PC (1988) Analysis of the
nucleotide sequence of the P1 operon of Mycoplasma
pneumoniae. Gene 73, 175–183.
16 Dandekar T, Huynen M, Regula JT, Ueberle B, Zimm-
ermann CU, Andrade M, Doerks T, Sanchez-Pulido L,

signal peptides and prediction of their cleavage sites.
Protein Eng 10, 1–6.
25 Bendtsen JD, Nielsen H, von Heijne G & Brunak S
(2004) Improved prediction of signal peptides: SignalP
3.0. J Mol Biol 340, 783–795.
26 Paetzel M, Karla A, Strynadka NC & Dalbey RE
(2002) Signal peptidases. Chem Rev 102, 4549–4580.
I. Catrein et al. Signal peptidase I activity in M. pneumoniae
FEBS Journal 272 (2005) 2892–2900 ª 2005 FEBS 2899
27 Opitz C, Di Cristina M, Reiss M, Ruppert T, Crisanti
A & Soldati D (2002) Intramembrane cleavage of
microneme proteins at the surface of the apicomplexan
parasite Toxoplasma gondii. EMBO J 21, 1577–1585.
28 Karty JA, Ireland MM, Brun YV & Reilly JP (2002)
Defining absolute confidence limits in the identification
of Caulobacter proteins by peptide mass mapping.
J Proteome Res 1, 325–335.
29 Handley J (2001) Novel MS agents. Analytical Chem-
istry 73, 414A.
30 Bromley E, Leeds N, Clark J, McGregor E, Ward M,
Dunn MJ & Tomley F (2003) Defining the protein
repertoire of microneme secretory organelles in the api-
complexan parasite Eimeria tenella. Proteomics 3, 1553–
1561.
31 Keough T, Youngquist RS & Lacey MP (1999) A
method for high-sensitivity peptide sequencing using
postsource decay matrix-assisted laser desorption ioniza-
tion mass spectrometry. Proc Natl Acad Sci USA 96,
7131–7136.
32 Schlosser A & Lehmann WD (2000) Five-membered

39 Chambaud I, Heilig R, Ferris S, Barbe V, Samson D,
Galisson F, Moszer I, Dybvig K, Wroblewski H, Viari
A, Rocha EP & Blanchard A (2001) The complete gen-
ome sequence of the murine respiratory pathogen Myco-
plasma pulmonis. Nucleic Acids Res 29, 2145–2153.
40 Braun V & Bosch V (1972) Sequence of the murein-lipo-
protein and the attachment site of the lipid. Eur J Bio-
chem 28, 51–69.
41 Su CJ, Chavoya A, Dallo SF & Baseman JB (1990)
Sequence divergency of the cytadhesin gene of Myco-
plasma pneumoniae. Infect Immun 58, 2669–2674.
42 Regula JT, Ueberle B, Boguth G, Go
¨
rg A, Schno
¨
lzer
M, Herrmann R & Frank R (2000) Towards a two-
dimensional proteome map of Mycoplasma pneumoniae .
Electrophoresis 21, 3765–3780.
43 Hayflick L (1965) Tissue cultures and mycoplasmas.
Tex Rep Biol Med 23 (Suppl. 1), 285+.
44 Dumke R, Catrein I, Pirkl E, Herrmann R & Jacobs E
(2003) Subtyping of Mycoplasma pneumoniae isolates
based on extended genome sequencing and on expres-
sion profiles. Int J Med Microbiol 292, 513–525.
Supplementary material
The following material is available from http://www.
blackwellpublishing.com/products/journals/suppmat/ EJB/
EJB4710/EJB4710sm.htm
Fig. S1. (A) Isolation of P40 from cell extracts, (B)