The reaction center (RC) complex of the green sulfur bacterium Chlorobaculum tepidum is composed of the Fenna-Matthews-Olson antenna protein (FMO) and the reaction center core (RCC) complex. The RCC complex has four subunits: PscA, PscB, PscC, and PscD. We studied the FMO/RCC complex by chemically cross-linking the purified sample followed by biochemical and spectroscopic analysis. Blue-native gels showed that there were two types of FMO/RCC complexes, which are consistent with complexes with one copy of FMO per RCC and two copies of FMO per RCC. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of the samples after cross-linking showed that all five subunits of the RC can be linked by three different cross-linkers: bissulfosuccinimidyl suberate, disuccinimidyl suberate, and 3,3-dithiobis-sulfosuccinimidyl propionate. The interaction sites of the cross-linked complex were also studied using liquid chromatography coupled to tandem mass spectrometry. The results indicated that FMO, PscB, PscD, and part of PscA are exposed on the cytoplasmic side of the membrane. PscD helps stabilize FMO to the reaction center and may facilitate transfer of the electron from the RC to ferredoxin. The soluble domain of the heme-containing cytochrome subunit PscC and part of the core subunit PscA are located on the periplasmic side of the membrane. There is a close relationship between the periplasmic portions of PscA and PscC, which is needed for the efficient transfer of the electron between PscC and P840.
The reaction center (RC) complex of the green sulfur bacterium Chlorobaculum tepidum is composed of the Fenna-Matthews-Olson antenna protein (FMO) and the reaction center core (RCC) complex. The RCC complex has four subunits: PscA, PscB, PscC, and PscD. We studied the FMO/RCC complex by chemically cross-linking the purified sample followed by biochemical and spectroscopic analysis. Blue-native gels showed that there were two types of FMO/RCC complexes, which are consistent with complexes with one copy of FMO per RCC and two copies of FMO per RCC. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of the samples after cross-linking showed that all five subunits of the RC can be linked by three different cross-linkers: bissulfosuccinimidyl suberate, disuccinimidyl suberate, and 3,3-dithiobis-sulfosuccinimidyl propionate. The interaction sites of the cross-linked complex were also studied using liquid chromatography coupled to tandem mass spectrometry. The results indicated that FMO, PscB, PscD, and part of PscA are exposed on the cytoplasmic side of the membrane. PscD helps stabilize FMO to the reaction center and may facilitate transfer of the electron from the RC to ferredoxin. The soluble domain of the heme-containing cytochrome subunit PscC and part of the core subunit PscA are located on the periplasmic side of the membrane. There is a close relationship between the periplasmic portions of PscA and PscC, which is needed for the efficient transfer of the electron between PscC and P840.
The photosynthetic
apparatus
of the anoxygenic photosynthetic green sulfur bacterium Chlorobaculum
(C.) tepidum consists of the reaction center core (RCC) complex,
the Fenna–Matthews–Olson antenna protein (FMO) and chlorosome
antenna complexes, and the menaquinol/cytochrome c oxidoreductase (cytochrome bc complex).[1] The light energy collected by the chlorosome
is transferred to the RCC complex through the chlorosome baseplate
and FMO.[1] The chlorosomes dominate the
absorption spectrum of intact cells in the region of 720–750
nm. The FMO trimer contains 24 molecules of bacteriochlorophyll a (BChl a) with Q absorption bands in the region of 790–830 nm.[2−4] The RCC complex, which is embedded in the cytoplasmic membrane,
contains 16 molecules of BChl a, 4 molecules of chlorophyll a (Chl a), and 2 carotenoids.[1,5,6] The RCC complex in C.
tepidum is an FeS-type (type I) reaction center, generally
similar to the RCC complex of photosystem I (PSI) in oxygenic photosynthetic
organisms. Unlike the heterodimeric PsaA/PsaB core of PSI, the RCC
in C. tepidum exhibits a homodimeric core structure
formed by two 82 kDa PscA proteins. The other three gene products
in the RCC are the 24 kDa PscB Fe–S protein, a 23 kDa cytochrome c551 (PscC) protein, and a 17 kDa PscD protein.[5] The PscA protein carries the primary donor P840
(a special pair of BChl a molecules), the primary
electron acceptor A0 (Chl a 670), a possible
secondary electron acceptor A1 (menaquinone), and Fe–S
center X (FX).[1,5,7,8] The PscB protein binds two 4Fe–4S
centers called FA and FB as the terminal electron
acceptors. The PscC protein, which mediates the transfer of an electron
from the menaquinol/cytochrome c oxidoreductase to
P840, has three membrane-spanning regions at the N-terminal end and
a soluble domain that binds a single heme group at the C-terminal
end on the periplasmic side of the membrane.[9] The PscD subunit of RCC shows some similarities in the amino acid
sequences with PsaD in PS I of plants and cyanobacteria.[10] PscD is loosely bound to the RCC and is not
essential for photosynthetic growth.[10] In
addition, the lack of the PscD subunit does not induce any serious
defect in the kinetics of electron transfer reactions.[10,11]Two biochemical preparations can be made of the C.
tepidum reaction center consisting of FMO and RCC (FMO/RCC)
or a minimal
complex containing only PscA and PscC. These two preparations have
been used extensively for spectroscopic and biochemical analysis,
but relatively few successful structural studies have been reported.[12−16] Both complexes have been studied by scanning transmission electron
microscopy (STEM). STEM predicts one or two FMO copies per RCC for
the FMO/RCC complex with a predicted mass of 454 kDa.[15] In the STEM images containing a single FMO, there is an
apparent place for a second FMO, suggesting that some FMO is lost
during preparation or the possibility of two distinct populations
of the FMO/RCC complex. A small knob protrudes from the RCC that is
likely PscB and PscD. For PscA/PscC, STEM analysis suggests two copies
of the PscA subunit and at least one copy of the PscC subunit with
a mass of 248 kDa.[16] High-resolution crystal
structures are available for only FMO and the soluble heme-containing
domain of PscC.[9,17] The lack of an FMO/RCC complex
crystal structure and the low-resolution nature of STEM leave significant
gaps in our understanding of the subunit organization of the FMO/RCC
complex.Structural mass spectrometry provides useful tools
for characterizing
protein organization.[18−20] Previously, our lab revealed the orientation of FMO
between the baseplate and the RCC.[21,22] We found that
the side of FMO containing BChl a 3 contacts the
cytoplasmic membrane using glycine ethyl ester (GEE) labeling.[22] Additional cross-linking data using zero-length
1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide (EDC) suggest that
FMO directly interacts with the CsmA protein, which is located in
the chlorosome baseplate.[23] The combination
of protein cross-linking and mass spectrometry in studies of native
proteins and protein complexes has become a popular tool in structural
mass spectrometry.[24−28] The previous studies have demonstrated the application of protein
cross-linking in studies of protein complexes in photosynthetic systems.[29,30] In this paper, we purified the intact FMO/RCC complex, which is
composed of FMO, PscA, PscB, PscC, and PscD subunits. The spatial
interaction between FMO and the RCC was studied by chemically cross-linking
the purified sample with three different cross-linkers: bissulfosuccinimidyl
suberate (BS3), disuccinimidyl suberate (DSS), and 3,3-dithiobis-sulfosuccinimidyl
propionate (DTSSP). The interaction sites of the cross-linked reaction
center are revealed by liquid chromatography coupled to tandem mass
spectrometry (LC–MS/MS). The results also indicated that FMO,
PscB, PscD, and part of PscA are exposed on the cytoplasmic side of
the membrane. The short distance between the soluble heme domain of
PscC and PscA facilitates the transfer of an electron between PscC
and P840.
Materials and Methods
Purification of the FMO/RCC Complex
Green sulfur bacterium C. tepidum strain TLS was
grown anaerobically at 45 °C
for 2 days. The cells were harvested by centrifugation at 8000g for 15 min. The FMO/RCC complex was purified by a method
reported previously with minor modifications.[31] The cells were resuspended in 20 mM Tris-HCl buffer (pH 8.0) and
broken by being sonicated. The supernatant was collected after low-speed
centrifugation and then ultracentrifuged at 150000g for 1 h to pellet the membranes. After being washed in 20 mM Tris-HCl
buffer containing 150 mM NaCl and 1 mM EDTA, the pellet was resuspended
in 20 mM Tris-HCl buffer to an OD810 of 6 cm–1; 10% DDM was added to the suspension to a final DDM concentration
of 2%, and the mixture was left at 4 °C for 1.5 h. The solution
was loaded onto step sucrose density gradients from 10 to 50% sucrose
and ultracentrifuged at 160000g for 13 h. The dark
green band from the sucrose gradient was then loaded onto a DEAE-cellulose
column of with a bed volume of ∼50 mL, which was equilibrated
with 20 mM Tris-HCl buffer (pH 8.0) and 0.05% DDM. The sample was
eluted with a linear gradient from 0 to 1 M NaCl in the same buffer.
Fractions containing both FMO and the RCC complex determined from
the shoulder at 807 and 835 nm were collected and concentrated for
future use.
Chemically Cross-Linked FMO/RCC Complex
The purified
FMO/RCC complex as described above was washed with 20 mM phosphate
buffer and cross-linked by BS3 (11.4 Å), DSS (11.4
Å), and DTSSP (12.0 Å). The mixture was incubated for 30
min at room temperature and then loaded onto the desalting column
(Zeba Spin Desalting Columns, 7K molecular weight cutoff, Thermo Fisher
Scientific Inc.). For BS3 and DSS, both isotopic [1:1 mixture
of deuterated (d12) and nondeuterated
(d0), Creative Molecules Inc.] and nonisotopic
linkers are used. Sodium dodecyl sulfate–polyacrylamide gel
electrophoresis (SDS–PAGE) and Blue-native gel electrophoresis
were performed as described previously.[32,33]
LC–MS/MS
and Data Analysis
The stained bands
of the SDS–PAGE gel were excised and digested with trypsin.
The samples were analyzed by LC–MS/MS using both a Waters Synapt
G2 Q-IM-TOF instrument and a Thermo LTQ Orbitrap (Thermo Scientific,
San Jose, CA) as described in the published protocol.[34] The data from the Waters Synapt G2 Q-IM-TOF instrument
were submitted to the ProteinLynx Global Server (version 2.5, Waters
Inc., Milford, MA) to identify the peptide sequence. The data for
cross-linked peptide identification obtained from a Thermo LTQ Orbitrap
were analyzed with xQuest.[34,35] The cross-linked peptides identified by xQuest were
further manually validated.
Results and Discussion
Purification
and Identification of the FMO/RCC Complex
The purified FMO/RCC
complex exhibits a BChl a absorption
band at 809 nm with a slight shoulder at 835 nm, which is consistent
with previous work, as shown in Figure 1.[31] The Q band of BChl a at 600 nm and the Q band
of Chl a at 670 nm are also observed.[15]
Figure 1
Absorbance spectrum of the FMO/RCC complex.
Absorbance spectrum of the FMO/RCC complex.
Cross-Linking by BS3, DSS, and
DTSSP
Five
bands on the SDS–PAGE gel at 60, 40, 30, 19, and 16 kDa were
identified as PscA, FMO, PscB, PscC, and PscD, respectively, by in-gel
digestion and subsequent LC–MS/MS analysis. The purified FMO/RCC
complex was cross-linked with different concentrations of BS3 as shown in Figure 2A. After being cross-linked
by BS3, the complex could not be denatured by SDS buffer,
which resulted in a bright band on top of the separating gel with
a mass of >250 kDa. With lower concentrations of the cross-linker,
some FMO/RCC complexes were not fully cross-linked, resulting in faint
bands of the five subunits. The cross-linked complex was further evaluated
by Blue-native gels. The faint band at 145 kDa could be assigned to
some free FMO trimers in the sample. The higher-mass region showed
two bands, 450 and 600 kDa. The mass difference is similar to that
of an FMO trimer. The data suggest that there are two types of FMO/RCC
complexes in the sample, which is consistent with the STEM data reported
previously.[15] The mass around 600 kDa can
be explained by a complex with a 2(FMO)3(PscA)2(PscB)(PscC)2(PscD) composition with 48 molecules of BChl a in FMO, 16 molecules of BChl a, and 4
molecules of Chl a in RCC.[1] The predicted mass is ∼560 kDa, which is close to the mass
of 600 kDa as shown in Blue-native gels.
Figure 2
(A) SDS–PAGE of
the FMO/RCC complex: markers (lane 1), non-cross-linked
control sample (lane 2), and FMO/RCC complex cross-linked by different
concentrations of BS3 at room temperature [lanes 3–5
(10, 5, and 0.5 mM, respectively)]. (B) Blue-native gel of the FMO/RCC
complex cross-linked by 10 mM BS3 (lane 1) and markers
(lane 2).
(A) SDS–PAGE of
the FMO/RCC complex: markers (lane 1), non-cross-linked
control sample (lane 2), and FMO/RCC complex cross-linked by different
concentrations of BS3 at room temperature [lanes 3–5
(10, 5, and 0.5 mM, respectively)]. (B) Blue-native gel of the FMO/RCC
complex cross-linked by 10 mM BS3 (lane 1) and markers
(lane 2).The purified complex was then
cross-linked with three different
cross-linkers, including DTSSP, DSS, and BS3. The disulfide
(S–S) bonds in DTSSP can be broken by reducing agents such
as β-mercaptoethanol. DSS is a hydrophobic cross-linker and
BS3 hydrophilic, but they have the same linking length.
Panels A and B of Figure 3 show the results
of the cross-linking experiments. All three cross-linkers resulted
in a bright band on the top of the separating gel, which was identified
as the cross-linked FMO/RCC complex. In Figure 3B, very faint FMO bands are visible in lanes 7 and 8 for DSS-cross-linked
samples that are absent in lanes 3 and 4 for BS3-cross-linked
samples, indicating the hydrophilic cross-linker BS3 works
slightly better for this experiment. Lane 5 indicates the FMO/RCC
complex cross-linked with BS3 on ice, and a faint FMO band
is observed like those with DSS. In Figure 3A, β-mercaptoethanol was added to DTSSP-cross-linked samples
to break S–S bonds in DTSSP, giving a band pattern similar
to that of the control sample. The presence of β-mercaptoethanol
had no effect on BS3- and DSS-linked samples. The appearance
of a bright band on top of the separating gel and the reversibility
of DTSSP cross-linking (lane 1 in Figure 3A)
clearly indicate that all five subunits of the FMO/RCC complex can
be successfully cross-linked.
Figure 3
Chemical cross-linking of the FMO/RCC complex
by different cross-linkers
(10 mM) and identification of subunit interactions by LC–MS/MS:
(A) DTSSP-cross-linked sample treated with (lane 1) and without (lane
2) a reducing agent, markers (lane 3), and non-cross-linked control
sample (lane 4). (B) BS3-cross-linked sample treated with
(lane 3) and without (lane 4) a reducing agent and a sample cross-linked
on ice and treated without a reducing agent (lane 5) and DSS-cross-linked
sample treated with (lane 7) and without (lane 8) a reducing agent,
markers (lanes 1 and 6), and non-cross-linked control sample (lane
2). (C) MS/MS spectrum of the interlinked peptide between PscA and
PscC induced by BS3.
Chemical cross-linking of the FMO/RCC complex
by different cross-linkers
(10 mM) and identification of subunit interactions by LC–MS/MS:
(A) DTSSP-cross-linked sample treated with (lane 1) and without (lane
2) a reducing agent, markers (lane 3), and non-cross-linked control
sample (lane 4). (B) BS3-cross-linked sample treated with
(lane 3) and without (lane 4) a reducing agent and a sample cross-linked
on ice and treated without a reducing agent (lane 5) and DSS-cross-linked
sample treated with (lane 7) and without (lane 8) a reducing agent,
markers (lanes 1 and 6), and non-cross-linked control sample (lane
2). (C) MS/MS spectrum of the interlinked peptide between PscA and
PscC induced by BS3.
Structural Analysis by Chemical Cross-Linking and LC–MS/MS
The LC–MS/MS results were grouped into monolinked peptides,
intralinked peptides, and interlinked peptides.[27] The MS/MS spectrum in Figure 3C
shows the interlink between PscA and PscC by BS3. As the
cross-linker targets solvent accessible lysine side chains, all those
linked peptides should be located on the solvent accessible surfaces.
The number of monolinked peptides exceeds the number of intra- or
interlinked peptides as monolinked peptides indicate the lysines available
for cross-linking, but only a subset of the available lysines that
are spatially close enough can form intra- or interlinks. As is shown
in Figure S1 of the Supporting Information, the distance between intralinked lysines 93K and 247K is 18.8 Å in the crystal structure of the FMO trimer,
and they may be even closer because 93K is located on the
loop.Figure 4A summarizes the interlinks
between the subunits of the FMO/RCC complex. The interlinks were classified
into two groups, confirmed and likely cross-links, on the basis of
the quality of the MS/MS spectra, i.e., the identification of the
peaks in the spectra and the sequence ion coverage of the peptide.
For both of them, almost all the major peaks in the MS/MS spectra
can be assigned. The cross-linked products with high-quality MS/MS
spectra that meet several criteria, for instance, the sequence ions
covering >70% of the sequence of the peptide, are classified as
confirmed
cross-links, and the confirmed cross-links were used as major constraints
in establishing the structural model of the FMO/RCC complex. While
cross-links with relatively lower-quality MS/MS spectra, like the
ones that have sequence ions covering only ∼50% of the sequence
of the peptide, are listed as likely cross-links to support our model. 79K of FMO in the middle of the FMO trimer is found to be linked
with 107K of PscD and possibly 45K of PscA and 36K and 60K of PscB. Because FMO is cytoplasmic,
PscB, PscD, and 45K of PscA should also be cytoplasmic.
As reported previously, 93K and 215K are located
on the upper exterior loops near the chlorosome.[22] Our results show that 215K of FMO is linked
to 46K of PscD and 93K of FMO is likely to be
linked to 30K of PscD. Therefore, the C- and N-terminal
lysines of PscD bind to both the top and middle side of the FMO trimer,
which allows for the proper binding of FMO to the RCC complex. This
is consistent with a previous report that in the PscD deletion strain,
the BChl a/P840 ratio in the FMO/RCC complex is lower,
suggesting that some of FMOs were partially detached from the RCC
without PscD.[10]
Figure 4
(A) Identification of
the interlinks between the subunits of the
FMO/RCC complex (red, confirmed; black, likely). The structure of
the FMO complex is shown with the linkages to the other subunits indicated.
(B) Membrane topological model of PscA with monolinked or cross-linked
lysines indicated. The region of sequence that binds the FX iron–sulfur center is indicated with a red box.
Our results are also
informative with respect to electron transfer
within the FMO/RCC complex on both the donor and acceptor sides. 111K of PscD is a conserved lysine residue thought to be similar
to lysine 106 in PsaD from PSI, which is involved in the direct interaction
of Fd with the iron–sulfur protein PsaC.[10] Our results indicate that 107K of PscD is linked
to 79K of FMO while 36K and 60K of
PscB are probably linked to the same lysine of FMO. At the same time, 107K of PscD is linked to the conserved lysine 111K of PscD (Figure S6 of the Supporting Information). Thus, those lysines should be fairly close, and it is very likely
that the conserved lysine 111K of PscD is close to PscB
to facilitate the transfer of an electron from RC to Fd as the role
of PsaD in PSI.[10](A) Identification of
the interlinks between the subunits of the
FMO/RCC complex (red, confirmed; black, likely). The structure of
the FMO complex is shown with the linkages to the other subunits indicated.
(B) Membrane topological model of PscA with monolinked or cross-linked
lysines indicated. The region of sequence that binds the FX iron–sulfur center is indicated with a red box.P indicates the
periplasmic side
of the membrane.On the
periplasmic side of the complex, we found several cross-links
between PscA and PscC. The N-terminal domain of PscC contains three
transmembrane helices, and the C-terminal, soluble heme domain is
located on the periplasmic side of the membrane.[9,36]132K and 338K of PscA are linked to several lysines
in the soluble domain of PscC. Therefore, 132K and 338K of PscA should also be located on the periplasmic side,
and the soluble domain of PscC should be close to PscA. 315K of PscA is probably linked to PscC and thus is also likely to be
on the periplasmic side of the membrane. The cross-link between PscA
and the soluble heme binding domain of PscC is consistent with the
efficient electron transfer from PscC to P840.[9,36−38]A hydropathy plot prediction of possible transmembrane
helices
of PscA and PscC was constructed by ExPASy ProtScale (Figure S2 of
the Supporting Information). Monolinks
can provide further structural information based on the hydropathy
plot. As shown in Table 1, nine monolinks were
found from the top to bottom sides of FMO because it is a water-soluble
protein. Six monolinks from the soluble domain of PscC were found,
and four of them are shown in the crystal structure. Figure 4B shows the membrane topological model of PscA.
As mentioned above, 45K of PscA was likely cross-linked
to FMO. Monolinked 60K of PscA was found, and there should
not be any transmembrane helix between the two lysines based on the
hydropathy plot; thus, residues 45–60 of PscA should all be
located on the cytoplasmic side of the membrane. In addition, 132K of PscA was found to be cross-linked with the soluble
domain of PscC on the periplasmic side. At the same time, 188K of the PscA monolink was observed, and there should be one transmembrane
helix between the two lysines based on the hydropathy plot; thus, 188K of PscA should be located on the cytoplasmic domain. Residues
315–338 of PscA should be located on the periplasmic domain
as they were both linked to the soluble domain of PscC. Monolinks
of 408K and 414K indicate that residues 408–414
should all be located on the same side of the membrane. Meanwhile,
residues 315–338 are in the periplasmic domain, and there is
one transmembrane helix between those two domains based on the hydropathy
plot. Thus, residues 408–414 should be on the cytoplasmic domain.
As there are two transmembrane helices between 408K and
the monolinked 519K of PscA, the latter lysine should also
be on the cytoplasmic domain. Furthermore, on the basis of the hydropathy
plot, the region containing residues 525–536, which represents
the FX binding motif FPCxGPxxGGTC,
should also be on the cytoplasmic side of the membrane.[39] In analogy with Photosystem I, it is anticipated
that FX, FA, and FB are all on the
cytoplamic side to facilitate electron transfer. Several monolinks
were found on the PscD protein, such as 40K, 46K, 107K, and 111K, some of which were linked
to FMO as mentioned above. It is not surprising to see that 160K of PscB was monolinked and is thus water accessible as it is close
to the iron–sulfur cluster binding peptide with the CxxCxxCxxxCP
motif (residues 140–151).
We propose a structural model
of the FMO/RCC complex as shown in
Figure 5. The crystal structure of FMO from
Protein Data Bank entry 3ENI and the PscC soluble domain from Protein Data Bank
entry 3A9F were
used. The interlinks between the FMO trimer and the PscD protein indicate
that both are located on the cytoplasmic domain. Monolinks of PscA
and the interlinks between PscA and the soluble domain of PscC showed
that PscA is a membrane protein composed of transmembrane helices
(probably 11) and the periplasmic domain is very close to the soluble
domain of PscC. The likely cross-linking between PscA and FMO indicated
that FMO is close to the cytoplasmic domain of the PscA. The likely
cross-linkings of PscB and FMO mean that PscB is spatially close to
the FMO trimer, and thus FMO, PscB, and PscD should all sit on PscA.
The STEM dark field images reported previously showed a knob protruding
from the RCC.[15] Our results suggest that
PscB and PscD should be the knob sitting on PscA. In addition, the
iron–sulfur cluster binding domain of PscB should be water
accessible, and PscD should be close to the chlorosomal side of the
FMO trimer.
Figure 5
Proposed structural model of the FMO/RCC complex.
Proposed structural model of the FMO/RCC complex.
Conclusions
The spatial interaction
between FMO and RCC was studied by chemically
cross-linking the purified FMO/RCC sample. All the subunits of RC
can be linked together by BS3, DSS, and DTSSP. The interaction
sites of the cross-linked complex were studied using LC–MS/MS.
The results showed a short spatial distance between FMO and the RCC.
The PscD subunit is thought to stabilize FMO to the RCC complex and
facilitate the transfer of an electron from the RCC to Fd. The short
distance of the soluble domain of PscC and PscA explains the efficient
electron transfer between PscC and P840. A structural model of the
FMO/RCC complex consistent with these results is proposed.
Authors: C Hager-Braun; D L Xie; U Jarosch; E Herold; M Büttner; R Zimmermann; R Deutzmann; G Hauska; N Nelson Journal: Biochemistry Date: 1995-07-25 Impact factor: 3.162
Authors: Alexander Leitner; Thomas Walzthoeni; Abdullah Kahraman; Franz Herzog; Oliver Rinner; Martin Beck; Ruedi Aebersold Journal: Mol Cell Proteomics Date: 2010-03-31 Impact factor: 5.911