Locked chromophore analogs reveal that photoactive yellow protein regulates biofilm formation in the deep sea bacterium Idiomarina loihiensis

Article abstract of DOI:10.1021/ja9057103

Idiomarina loihiensis is a heterotrophic deep sea bacterium with no known photobiology. We show that light suppresses biofilm formation in this organism. The genome of I. loihiensis encodes a single photoreceptor protein: a homologue of photoactive yellow

Full text of DOI:10.1021/ja9057103

Published on Web 11/05/2009
Locked Chromophore Analogs Reveal That Photoactive
Yellow Protein Regulates Biofilm Formation in the Deep Sea
Bacterium Idiomarina loihiensis
Michael A. van der Horst,† T. Page Stalcup,‡ Sandip Kaledhonkar,§
Masato Kumauchi,‡ Miwa Hara,‡ Aihua Xie,§ Klaas J. Hellingwerf,† and
Wouter D. Hoff*,‡
Department of Molecular Microbial Physiology, Swammerdam Institute for Life Sciences,
UniVersity of Amsterdam, The Netherlands, Department of Microbiology and Molecular Genetics
and Department of Physics, Oklahoma State UniVersity, Stillwater, Oklahoma 74078
Received July 13, 2009; E-mail: wouter.hoff@okstate.edu
Abstract: Idiomarina loihiensis is a heterotrophic deep sea bacterium with no known photobiology. We
show that light suppresses biofilm formation in this organism. The genome of I. loihiensis encodes a single
photoreceptor protein: a homologue of photoactive yellow protein (PYP), a blue light receptor with
photochemistry based on trans to cis isomerization of its p-coumaric acid (pCA) chromophore. The addition
of trans-locked pCA to I. loihiensis increases biofilm formation, whereas cis-locked pCA decreases it. This
demonstrates that the PYP homologue regulates biofilm formation in I. loihiensis, revealing an unexpected
functional versatility in the PYP family of photoreceptors. These results imply that I. loihiensis thrives not
only in the deep sea but also near the water surface and provide an example of genome-based discovery
of photophysiological responses. The use of locked pCA analogs is a novel and generally applicable
pharmacochemical tool to study the in vivo role of PYPs irrespective of genetic accessibility. Heterologously
produced PYP from I. loihiensis (Il PYP) absorbs maximally at 446 nm and has a pCA pKa of 3.4.
Photoexcitation triggers the formation of a pB signaling state that decays with a time constant of 0.3 s.
FTIR difference signals at 1726 and 1497 cm-1 reveal that active-site proton transfer during the photocycle
is conserved in Il PYP. It has been proposed that a correlation exists between the lifetime of a photoreceptor
signaling state and the time scale of the biological response that it regulates. The data presented here
provide an example of a protein with a rapid photocycle that regulates a slow biological response.
1. Introduction
by genome projects indicates that many additional unanticipated
photobiological responses in chemotrophic bacteria remain to
be discovered.
Living organisms use a limited number of types of photore-
ceptor proteins to obtain information about their ambient light
climate. These photoreceptor proteins can be grouped into
different families, on the basis of their structure and light-
absorbing cofactor.1 Until very recently, such photoreceptor
proteins had been found exclusively in phototrophic microor-
ganisms. However, genome sequencing projects have uncovered
many examples of photoreceptor proteins in chemotrophic
microorganisms.2,3 Bioinformatics analyses showed that mem-
bers from most known photoreceptor protein families are present
in chemotrophic Bacteria. Most of these have not been
characterized regarding their in vivo function or in vitro
biochemical and photophysical characteristics. The widespread
occurrence of as of yet unstudied photosensory proteins revealed
One example of a chemotrophic bacterium with a photore-
ceptor encoded in its genome is the deep-sea bacterium
Idiomarina loihiensis. This organism belongs to the group of
γ-proteobacteria and was isolated at 1 km depth in the Hawaiian
ocean.4 The genome of I. loihiensis contains a cluster of 32
genes that encode proteins involved in the production and
secretion of polysaccharides.5 This prompted us to study biofilm
formation in I. loihiensis. In parallel, we examined the genome
of I. loihiensis for the presence of photoreceptor proteins from
all five established photoreceptor families that can be unambigu-
ously recognized based on their sequence,1 and a single
photoreceptor gene was identified.2 This gene encodes a protein
that shows clear homology to photoactive yellow protein
(PYP).2,6 This blue-light photoreceptor protein was first found
in Halorhodospira halophila,7,8 and its physicochemical proper-
† University of Amsterdam.
‡ Department of Microbiology and Molecular Genetics, Oklahoma State
University.
§ Department of Physics, Oklahoma State University.
(1) van der Horst, M. A.; Hellingwerf, K. J. Acc. Chem. Res. 2004, 37,
(4) Donachie, S. P.; Hou, S.; Gregory, T. S.; Malahoff, A.; Alam, M. Int.
J. Syst. EVol. Microbiol. 2003, 53, 1873–1879.
13–20.
(2) van der Horst, M. A.; Key, J.; Hellingwerf, K. J. Trends Microbiol.
(5) Hou, S.; et al. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 18036–18041.
(6) Kumauchi, M.; Hara, M. T.; Stalcup, P.; Xie, A.; Hoff, W. D.
Photochem. Photobiol. 2008, 84, 956–969.
2007, 15, 554–562.
(3) Purcell, E. B.; Crosson, S. Curr. Opin. Microbiol. 2008, 11, 168–
178.
(7) Meyer, T. E. Biochim. Biophys. Acta 1985, 806, 175–183.
9
10.1021/ja9057103 CCC: $40.75  2009 American Chemical Society
J. AM. CHEM. SOC. 2009, 131, 17443–17451 17443

A R T I C L E S
Horst et al.
ties have been well characterized.9-11 Studies on the properties
of 6 additional PYPs have been initiated (see ref 6 for a review).
This work has led to a detailed understanding of PYP photo-
chemistry. Light activation of the initial pG state of PYP (λmax
446 nm) involves the photoisomerization of the vinyl bond of
its p-coumaric acid (pCA) chromophore from trans to cis,
resulting in the short-lived, red-shifted pR state. Subsequent
protonation of the pCA by Glu4612,13 causes formation of the
pB′ state, and subsequently leads to structural changes in the
protein that result in the formation of the blue-shifted pB state
(λmax 355 nm). The pB state is the presumed signaling state of
PYP. The protein then thermally recovers to its pG groundstate
on a time scale that ranges from 1 ms for the PYP from
Rhodobacter sphaeroides,14 to ∼0.5 s for the PYP from H.
halophila (Hh PYP),8,15 to ∼1 h for the PYP from Salinibacter
ruber.6,16
above. After 24 h, nonattached cells were removed by washing the
plates three times with demi water.
Attached cells were stained for 60 min with 1% crystal violet.
Nonbound crystal violet was removed by washing three times with
demi water. Crystal violet was resolublized in 80% ethanol/20%
acetone, and cell attachment was quantified by measuring the
absorbance at 595 nm in a plate reader (Spectramax plus, Molecular
devices). To test the effect of pCA and its analogs on cell
attachment, concentrated solutions of the chromophores (pCA, trans-
locked, and cis-locked) in DMF were added to the cultures in the
wells to a final concentration of 1 mM. The final concentration
and solvent used for adding these three chromophores to I. loihiensis
cell cultures were identical. The values for biofilm formation in
the dark and light, and in the dark in the presence of pCA and the
two pCA analogs as determined by crystal violet binding were
averaged over 6 wells in a 96 wells microtiter plate, and the
reproducibility of the resulting values was confirmed in two
independent experiments. Attachment of the cells to the wall of
the wells prevented exact quantitation of growth by OD600
measurements. Visual inspection of the 96-well plates indicated
that there is only limited growth during overnight incubation in
the wells. Addition of chromophores did not noticeably affect cell
growth.
Examination of the I. loihiensis genome5 indicates that its
pyp gene is located in the immediate vicinity of a gene encoding
a diguanylate cyclase. Since c-di-GMP, a recently discovered
bacterial signaling molecule, is a central regulator for the
synthesis of the exopolysaccharides that are involved in biofilm
formation,17-20 we examined the role of light in regulating
biofilm formation in I. loihiensis. This revealed that photoac-
tivation of PYP suppresses biofilm formation in this organism.
Synthesis of pCA Analogs. The cis-locked chromophore analog
2-(4-Hydroxy-phenyl)-cyclopent-1-enecarboxylic acid was synthe-
sized from ethyl-2-oxocyclopentane carboxylate as described in the
Supporting Information. The identity and purity of the intermediate
and final products was determined using FTIR and NMR spectros-
copy and mass spectrometry.
2. Experimental Section
Cloning of the Il pyp Gene. On the basis of the genome
sequence of I. loihiensis,5 the oligonucleotides 5′-AGGTAACAC-
CATGGAGATTGTTCAATT-3′ and 5′-TGAGTCTGGATCCT-
TATAGTCGCTTAACGAATA-3′ were used as the sense primer
and antisense primer, respectively, for PCR amplification of the
pyp gene from this organism, thereby introducing NcoI and BamHI
restriction sites. Purified I. liohiensis genomic DNA was used as
the template in the PCR reaction. The amplified gene was cloned
into the pET-16b vector (Novagen) using the NcoI and BamHI
restriction sites. The nucleotide sequence of the cloned Il PYP was
confirmed by DNA sequence analysis (BigDye-terminated reaction
analyzed on ABI Model 3730 DNA Analyzer, Recombinant DNA
and Protein Core Facility at Oklahoma State University). Plasmid
DNA was transformed into E. coli BL21 (DE3) for apoPYP
overproduction.
Overproduction and Purification of Hh PYP and Il PYP. PYP
from H. halophila was produced and isolated as described in22 as
hexa-histidine tagged apoprotein in Escherichia coli. The apoprotein
was reconstituted with the 1,1-carbonyldiimidazole derivative of
the respective chromophore (i.e., 4-hydroxy-cinnamic acid (pCA),
7-hydroxy-coumarin-3-carboxylic acid (trans-locked; Molecular
Probes, Invitrogen), 2-(4-Hydroxy-phenyl)-cyclopent-1-enecarboxy-
lic acid (cis-locked), see Figure 1C) as described in.23 Hh PYP
was used in 10 mM Tris-HCl, pH 8.0 without removal of the
N-terminal hexahistidine tag.
Culturing and Attachment of I. loihiensis. I. loihiensis strain
L2-TR(T) was cultured in Marine Broth (Difco) at 30 °C with gentle
shaking. Biofilm assays were performed at 20 °C, in a well-
thermostatted room, using a method adapted from ref 21. An
overnight culture of I. loihiensis was diluted ten times and incubated
in 96-well flat-bottom PVC microtiter plates (Costar 3595). One
plate was covered with black cloth, another with transparent plastic
foil. Plates were illuminated with a white light bulb (50 W, Philips)
at ∼80 cm distance, resulting in a light intensity of ∼80 microe-
insteins m-2 s-1. Care was taken to minimize effects of illumination
on other physical parameters in the experimental setup, such as
temperature and humidity. To this end, the experiment was also
carried out in an incubator/shaker, thermostatted at 20 °C. The
incubator was not rotating and illumination was carried out from
the outside, through the glass shield in the lid of the incubator. To
ensure that the position in the 96-wells plate did not alter the
behavior of the cells, the addition of pCA and its analogs was
randomized over the plate, and plates were incubated as described
(8) Meyer, T. E.; Yakali, E.; Cusanovich, M. A.; Tollin, G. Biochemistry
1987, 26, 418–423.
(9) Cusanovich, M. A.; Meyer, T. E. Biochemistry 2003, 42, 4759–4770.
(10) Hellingwerf, K. J.; Hendriks, J.; Gensch, T. J. Phys. Chem. A 2003,
107, 1082–1094.
(11) Imamoto, Y.; Kataoka, M. Photochem. Photobiol. 2007, 83, 40–49.
(12) Xie, A.; Hoff, W. D.; Kroon, A. R.; Hellingwerf, K. J. Biochemistry
1996, 35, 14671–14678.
Overproduction of Il apoPYP was induced by the addition of 1
mM IPTG, extracted from E. coli BL21 (DE3) using 8 M urea and
reconstituted with p-hydroxycinnamic anhydride (Sigma-Aldrich)24
following the procedure described form the PYP from H. halo-
phila.25 The apoPYP from I. loihiensis readily reconstituted upon
the addition of the pCA anhydride, yielding a bright yellow protein
(13) Xie, A.; Kelemen, L.; Hendriks, J.; White, B. J.; Hellingwerf, K. J.;
Hoff, W. D. Biochemistry 2001, 40, 1510–1517.
(14) Haker, A.; Hendriks, J.; Gensch, T.; Hellingwerf, K.; Crielaard, W.
FEBS Lett. 2000, 486, 52–56.
(15) Hoff, W. D.; van Stokkum, I. H.; van Ramesdonk, H. J.; van
Brederode, M. E.; Brouwer, A. M.; Fitch, J. C.; Meyer, T. E.; van
Grondelle, R.; Hellingwerf, K. J. Biophys. J. 1994, 67, 1691–1705.
(16) Memmi, S.; Kyndt, J.; Meyer, T.; Devreese, B.; Cusanovich, M.; Van
Beeumen, J. Biochemistry 2008, 47, 2014–2024.
(22) Kort, R.; Hoff, W. D.; van West, M.; Kroon, A. R.; Hoffer, S. M.;
Vlieg, K. H.; Crielaand, W.; van Beeumen, J. J.; Hellingwerf, K. J.
EMBO J. 1996, 15, 3209–3218.
(17) Hickman, J. W.; Tifrea, D. F.; Harwood, C. S. Proc. Natl. Acad. Sci.
U.S.A. 2005, 102, 14422–14427.
(18) Paul, R.; Weiser, S.; Amiot, N. C.; Chan, C.; Schirmer, T.; Giese, B.;
Jenal, U. Genes DeV. 2004, 18, 715–727.
(23) Hendriks, J.; Gensch, T.; Hviid, L.; van der Horst, M. A.; Hellingwerf,
K. J.; van Thor, J. J. Biophys. J. 2002, 82, 1632–1643.
(24) Imamoto, Y.; Ito, T.; Kataoka, M.; Tokunaga, F. FEBS Lett. 1995,
374 (2), 157–160.
(19) Ryjenkov, D. A.; Tarutina, M.; Moskvin, O. V.; Gomelsky, M. J.
Bacteriol. 2005, 187, 1792–1798.
(20) Cotter, P. A.; Stibitz, S. Curr. Opin. Microbiol. 2007, 10, 17–23.
(21) O’Toole, G. A.; Kolter, R. Mol. Microbiol. 1998, 30, 295–304.
(25) Mihara, K.; Hisatomi, O.; Imamoto, Y.; Kataoka, M.; Tokunaga, F.
J. Biochem. 1997, 121, 876–880.
9
17444 J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009

Locked Chromophores Reveal PYP Function
A R T I C L E S
denatured Il PYP to that of native and SDS-denatured wild type
H. halophila PYP, which has a known extinction coefficient at 446
nm (45 500 M-1 cm-1).27 A 150 W halogen quartz light source
(Cuda) with a broadband blue filter (band-pass filter 59855 Oriel,
CT) was used to initiate the photocycle of Il PYP in 10 mM Tris-
HCl pH 7.5. After approximately 15 s of illumination the actinic
light was shuttered, and the thermal recovery of the dark state was
measured with a time resolution of 100 ms. The photocycle recovery
kinetics at 446 and 355 nm was biexponential. A mixed buffer
(Glycine, Succinate, MES and MOPS, 100 mM each) was used
for pH titration measurement in the pH range from 2.0 to 12.0.
The pH was adjusted using 1 N NaOH or HCl. The acid titration
was carried out in a darkened room using the minimal intensity of
red light needed for handling of the samples and reading of the pH
values. The pH dependence of the sample absorbance at its absorp-
tion maximum (446 nm) was described using the Henderson-
Hasselbalch equation.
FTIR Spectroscopy. Il PYP and Hh PYP samples for FTIR
spectroscopy were prepared at 8 mM in 50 mM phosphate buffer
in D2O at pH* 6.6 as described.12 For each sample, 2.7 µL PYP
was sandwiched between two CaF2 windows (15 mm diameter and
2 mm thickness) with a 12 µm spacer. Steady state and time-
resolved infrared measurements were measured using a Bruker IFS
66v FTIR spectrometer with a liquid nitrogen-cooled Mercury
Cadmium Telluride (MCT) detector. Water vapor in the optical
path was removed by vacuum, except for the sample compartment,
which was purging with dry nitrogen gas (2.2 L/min). Sample
temperature was controlled at 300 K using a water circulator
(Neslab, RTE- 111) in all the FTIR measurements. FTIR absorption
spectra of steady state PYP samples were measured in the spectral
range of 4000-850 cm-1 at 2 cm-1 spectral resolution and 40 kHz
scanning velocity, and averaged over 175 scans, using CaF2
windows as a reference. The second derivative of the absorbance
was calculated using the Bruker OPUS software based on Savitzky-
Golay algorithm with 13 smoothing points.
For time-resolved FTIR spectroscopy, the photocycle of PYP
was triggered with blue laser pulses (475 nm, 4 ns pulse duration,
and ∼3 mJ per pulse) from a Quantel, Brilliant laser with an Opotek
OPO. The repetition rate of the pump laser was set at 0.125 Hz,
for Hh PYP and Il PYP. Light-induced infrared absorption changes
of PYP were measured in rapid-scan FTIR mode at 4.5 cm-1
spectral resolution with 200 kHz scanning velocity, and averaged
over 256 repetitions. A quadruple splitting technique was employed
so that the time resolution for the first spectrum was improved from
46 ms (without splitting) to 8 ms. For proper synchronization
between the scanning mirror of the Bruker FTIR system and the
laser flashes, a triggering system was designed and utilized where
the FTIR system served as the master clock, and two delay
generators (SRS, DG535) that were externally triggered by the FTIR
system (coupled to the scanning mirror), were used to trigger the
laser flash lamps (∼10 Hz), the laser Q-switch, and rapid-scan FTIR
data acquisition.
Figure 1. Regulation of biofilm formation in I. loihiensis by light and
p-coumaric acid analogs. Error bars indicate standard deviations. (A) Biofilm
formation in growing I. loihiensis cultures was measured in the light (open
bars) and in the dark (gray bars), revealing that light suppresses biofilm
formation. The absorbance of crystal violet at 595 nm is shown, providing
a measure for the amount of cells attached to the PVC microtiter plate well.
(B) Effect of the addition of pCA, cis-locked pCA, and trans-locked pCA
on biofilm formation in I. loihiensis cultures grown in the dark. The error
in panels (A) and (B) were calculated for values of 6 wells in a 96-well
microtiter plate. (C) pCA chromophore analogs used to lock PYP into its
pG and pB states. The structures of the chromophore used in this study are
shown: 4-hydroxy-cinnamic acid (pCA, 1), (2) 7-hydroxy-coumarin-3-
carboxylic acid (trans-locked, 2), and 2-(4-hydroxy-phenyl)-cyclopent-1-
enecarboxylic acid (cis-locked, 3). (D) UV/vis absorbance spectra of Hh
PYP containing pCA (solid), trans-locked pCA (dotted), and cis-locked pCA
(dashed).
with an absorbance maximum at 446 nm. The reconstituted holo-
Il PYP was purified by anion-exchange chromatography using
DEAE-sepharose CL-6B (Pharmacia) and sodium chloride as the
eluent. The purified protein had a purity index (ratio of absorbance
at 280 and 446 nm) of 0.48. Possible proteolytic damage of the
purified protein was examined by mass spectrometry (Applied
Biosystems Voyager-DE Pro MALDI-TOF MS, Recombinant DNA
and Protein Core Facility at Oklahoma State University). This
yielded a mass of 14,121 Da, indistinguishable within the error of
the instrument from the theoretical mass of 14 131 Da.
UV-Vis Absorbance Spectroscopy. UV-visible absorption
spectra were measured at room temperature using an HP-8453
(Hewlett-Packard) diode array spectrophotometer and a monochro-
mator-based Cary300Bio (Varian) spectrophotometer. The relative
value of the molar extinction coefficient of the Il PYP was
determined based on the effect of denaturing the protein in 10 mM
Tris-HCl pH 7.5 by the addition of 2% SDS on its absorbance
spectrum.26 The extinction coefficient of Il PYP was derived by
comparing the amplitude of the absorbance band at 345 nm of the
A macro program was designed and employed to compute the
time-resolved infrared difference spectra, ∆A(t,ν) ) CpB(t)·d·[εpB(ν)
- εpG(ν)], using the infrared absorption of the pG state (preflash
spectrum) as the baseline. The peak-to-peak signal of the first
infrared difference spectrum of Il PYP was 10 mOD in the Amide
I region, with a noise level of 0.03 mOD (3 × 10-5 OD).
Bioinformatics Analysis. The genome of I. loihiensis was
examined for the presence of pcl and tal, and for genes encoding
photoreceptor proteins using NCBI genomic BLAST tools at: http://
www.ncbi.nlm.nih.gov/sutils/genom_table.cgi. Domain architecture
was examined using SMART (http://smart.embl-heidelberg.de/) and
Pfam (http://pfam.sanger.ac.uk/) databases.
(26) Imamoto, Y.; Koshimizu, H.; Mihara, K.; Hisatomi, O.; Mizukami,
T.; Tsujimoto, K.; Kataoka, M.; Tokunaga, F. Biochemistry 2001, 40,
4679–4685.
(27) Meyer, T. E.; Tollin, G.; Hazzard, J. H.; Cusanovich, M. A. Biophys.
J. 1989, 56, 559–564.
9
J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009 17445

A R T I C L E S
Horst et al.
3. Results
Light-Regulated Cell Attachment. During the culturing I.
loihiensis in glass flasks or tubes, cells were observed by eye
to stick together in aggregates or attach to the glass culturing
tube in a ring at the air-liquid interface. These aggregates were
several millimeters in size and appeared after overnight cultur-
ing. In PVC 96-well microtiter plates, cells tended to stick more
to the wall of the well, instead of to each other. To examine if
light regulates this biofilm formation process, we determined
the effects of white light on the attachment of I. loihiensis to
96-well microtiter plates. Bacterial cell attachment to the PVC
was quantified using a crystal violet staining method.21 This
revealed that the presence of light reproducibly suppresses
attachment of I. loihiensis cells to the PVC surface by ∼35%
(Figure 1A).
Since the only photoreceptor protein identified in the genome
of I. loihiensis is a PYP homologue,2 we designed an approach
to study the role of this photoreceptor in the light regulation of
biofilm formation. Because no genetic system is available for
I. loihiensis, we utilized available knowledge on the effects of
locked pCA analogues on PYP photochemistry (Figure 1C).28,29
To block PYP photoexcitation we used 7-hydroxy-coumarin-
3-carboxylic acid as a trans-locked pCA to essentially lock the
protein in its pG dark state. The remaining photochemistry of
PYP reconstituted with this pCA analog28 occurs with very low
quantum efficiency. The cis-locked pCA analog 2-(4-hydroxy-
phenyl)-cyclopent-1-enecarboxylic acid would be expected to
permanently activate PYP by locking it in its pB state.30 Both
chromophores readily attach to Hh PYP. As expected, trans-
locked pCA results in an absorbance spectrum resembling the
pG state (λmax 443 nm), while cis-locked pCA resembles the
absorbance spectrum of the pB state (Figure 1D). PYP
reconstituted with cis-locked pCA exhibits a characteristic peak
at ∼350 nm, but in addition displays a peak at 310 nm. The
free form of this chromophore has absorption maxima at 267,
283, and 340 nm at pH 3, 7, and 10, respectively (data not
shown). The 310 nm peak may arise from a combination of
increased scattering and the n-to-π transition of the carbonyl
group of the chromophore. The increased scattering of the
solution suggests partial aggregation of PYP reconstituted with
cis-locked pCA.
Figure 2. Comparison of the spectroscopic properties of the pG dark state
of the PYPs from Idiomarina loihiensis and Halorhodospira halophila. (A)
UV-vis absorbance spectra of Il PYP (dotted line) and Hh PYP (solid
line). FTIR spectroscopic comparison of the structure of the pG dark state
of Il PYP (red) and Hh PYP (blue) based on infrared absorbance spectra
(B) and second derivative spectra (C) in D2O at pH* 6.6.
loihiensis. We therefore decided to examine the spectroscopic
properties of Il PYP.
Spectroscopy of Il PYP. The UV/vis absorbance spectrum of
Il PYP is essentially identical to that of Hh PYP (Figure 2A).
The extinction coefficient of the characteristic 446 nm absor-
bance band was determined to be 43 350 M-1 cm-1, very similar
to the value of 45 500 for H. halophila PYP.27
Incubation of I. loihiensis with these two pCA analogs
significantly affected biofilm formation in the dark. The addition
of trans-locked pCA resulted in a 45% increase in cell
attachment, whereas cis-locked pCA resulted in a 75% reduction
of attachment, when compared to the addition of pCA or no
chromophore addition (Figure 1B). In the presence of both
chromophore derivatives the cells no longer exhibited the light-
induced suppression of biofilm formation. The addition of
exogenous pCA to the cells did not alter the photoresponse
(compare Figure 1A in the dark and 1B in the presence of pCA).
This result shows that I. loihiensis cells synthesize pCA, and
that the addition of the small amount of DMF used as solvent
for the chromophores to the cell culture does not affect biofilm
formation. These data provide strong evidence that PYP is the
photoreceptor for the light regulation of biofilm formation in I.
The structure of the pG state of Il PYP was compared to that
of the PYP from H. halophila using FTIR absorption and second
derivative spectroscopy (Figure 2B, C). The FTIR absorbance
spectrum of a protein in the 1800-1350 cm-1 region contains
contributions from the protein backbone and from all polar and
charged side chains. The peak positions of these overlapping
bands are better resolved in the second derivative of the FTIR
spectrum.31 Thus, the second derivative of the FTIR absorbance
of a protein provides a sensitive “fingerprint” for the structure
of a protein.31,32 The second derivative FTIR spectra of the pG
states of the PYPs from I. loihiensis and H. halophila are highly
similar, demonstrating that they share a very similar three-
dimensional structure. This is in line with the 75% sequence
identity shared by these two PYPs.
The pKa of the pCA chromophore in Il PYP was determined
by acid titration using UV/vis absorbance spectroscopy. At pH
values below 4 the native pG state was converted to a pBdark
(28) Cordfunke, R.; Kort, R.; Pierik, A.; Gobets, B.; Koomen, G. J.;
Verhoeven, J. W.; Hellingwerf, K. J. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 7396–7401.
(29) van der Horst, M. A.; Arents, J. C.; Kort, R.; Hellingwerf, K. J.
Photochem. Photobiol. Sci. 2007, 6, 571–579.
(31) Dong, A. C.; Huang, P.; Caughey, W. S. Biochemistry 1992, 31, 182–
189.
(30) van der Horst, M. A. Structure/function relations in Photoactive Yellow
Protein. Ph.D. thesis, University of Amsterdam, Amsterdam, 2004.
(32) He, W. Z.; Newell, W. R.; Haris, P. I.; Chapman, D.; Barber, J.
Biochemistry 1991, 30, 4552–4559.
9
17446 J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009

Locked Chromophores Reveal PYP Function
A R T I C L E S
Figure 4. Comparison of the last photocycle step in the PYPs from I.
loihiensis and H. halophila at pH 7.5. Kinetics of pB decay in Il PYP (solid
line and closed circles) and Hh PYP (dotted line and open circles) detected
at 446 nm. In both PYPs the decay is biexponential. The time constants for
Il PYP were 230 ms (80%) and 5.7 s (20%), and for Hh PYP 380 ms (75%)
and 1.6 s (25%). The inset depicts the light-dark UV/vis difference spectra
of Il PYP (solid) and Hh PYP (dotted) calculated using the absorbance
spectra a t ) 0 s and t ) 10 s.
reduction in the unassigned side chain signal at 1590 cm-1. This
indicates that the light-induced conformational changes in Il PYP
are similar but somewhat smaller compared to those in Hh PYP.
To complement the detection of the kinetics of pB decay in
Il PYP by visible absorbance spectroscopy (Figure 4), we
extracted the time dependence of the amplitude between the
negative amide I difference signal at 1642 cm-1 and the positive
amide I peak at 1624 cm-1 (Figure 5C). This revealed a biphasic
decay with a major fast phase of 0.30 s and a minor slow phase
of 3.6 s, very similar to the values obtained by visible absorbance
spectroscopy. The kinetics of the amide I signals of Il PYP
during pB decay were compared to those obtained for Hh PYP,
confirming that the kinetics of the last photocycle step in these
two proteins is very similar (Figure 5C).
Figure 3. Low pH titration of the PYP from I. loihiensis. (A) UV-vis
absorbance spectra of Il PYP as a function of pH. Arrows indicate the
direction of the change in the amplitudes of the bands at 446 and 350 nm
at increasingly lower pH values. (B) pH dependence of the absorbance at
446 nm as a function pH, fit to the Henderson-Hasselbalch equation.
species7,33 with a protonated pCA chromophore and an absor-
bance maximum at 352 nm (Figure 3A). The transition was
described by a single pKa value of 3.4, with an n value of 1.3
(Figure 3B). In Hh PYP this transition occurs with a pKa of
2.8.7,33
The photochemical activity of the protein was examined by
exposing it to blue light at pH 7.5. This revealed the formation
of a pB photocycle intermediate with an absorbance maximum
at 350 nm (Figure 4). The pB state decayed to the initial pG
state in a biphasic process with time constants of 0.3 s (79%)
and 5 s (21%) (Figure 4). Thus, the pyp gene in the I. loihiensis
genome encodes a genuine, photochemically active PYP. The
kinetics of pB decay in the PYP from I. loihiensis is very similar
to that of the PYP from H. halophila.8,15
The molecular events that occur upon pB formation in Il PYP
were examined using rapid-scan FTIR difference spectroscopy
in D2O (Figure 5A). The pB - pG FTIR difference spectrum
for Il PYP is quite similar to that of Hh PYP (Figure 5B). Thus,
the published assignments of FTIR signals in Hh PYP12,13 can
be applied to Il PYP. The negative signal at 1726 cm-1 is caused
by the deprotonation of Glu46, and matches the signals at 1497
and 1515 cm-1 caused by pCA protonation. The signal at 1726
cm-1 also shows that hydrogen bonding between the side chain
of Glu46 and the phenolic oxygen of the pCA in the pG state
is essentially identical in the two PYPs. The main differences
in the pB - pG difference spectrum for Il PYP are a ∼20%
reduction in the amide I difference signals at 1624 cm-1, and a
Identification of TAL and pCL in I. loihiensis. The in ViVo
functioning of PYP requires the biosynthesis of its pCA
chromophore and covalent attachment to apoPYP. In Rhodo-
bacter capsulatus this is achieved by the activity of two proteins:
tyrosine ammonia lyase (TAL) and p-coumaryl:CoA ligase
(pCL).34 TAL converts tyrosine into pCA and pCL activates
the pCA by attachment to coenzyme A. The coumaryl CoA
then reacts with apoPYP, resulting in the formation of a thioester
bond between pCA and the conserved Cys side chain at the
active site of PYP. Since the I. loihiensis genome encodes a
functional pyp gene, we examined the presence of possible tal
and pcl genes in its genome.
The I. loihiensis genome encodes two proteins with significant
sequence similarity to the known TAL from Rb. capsulatus
(Supplemental Figure 1, Supporting Information). The product
of gene IL135 shares 31% sequence identity and 46% sequence
identity with the TAL from Rb. capsulatus; for gene IL2450
these number are 28 and 43%. The two TAL candidates in I.
loihiensis are fairly divergent: they share 28% sequence identity
and 49% similarity. In H. halophila the pyp gene is present in
an apparent three-gene operon, together with a tal and a pcl
gene.22 The TAL from H. halophila is 32% identical to product
(34) Kyndt, J. A.; Vanrobaeys, F.; Fitch, J. C.; Devreese, B. V.; Meyer,
T. E.; Cusanovich, M. A.; van Beeumen, J. J. Biochemistry 2003, 42,
965–970.
(33) Hoff, W. D.; van Stokkum, I. H. M.; Gural, J.; Hellingwerf, K. J.
Biochim. Biophys. Acta 1997, 1322, 151–162.
9
J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009 17447

A R T I C L E S
Horst et al.
His89 is of particular importance, because hydrogen bonds with
the phenolic oxygen of the Tyr substrate, and mutagenesis has
shown that this residue is critical for its substrate specificity.36
In view of this high level of conservation of functionally critical
residues, the productes of I. loihiensis genes IL135 and IL2450
are very likely to encode functional TALs. Both of these genes
are not in the immediate vicinity of the pyp gene (I. loihiensis
gene IL2385). This is in contrast to the situation in H.
halophila,22 and in Rb. capsulatus,37 where tal is in the located
very close to the pyp gene.
A single gene with high sequence similarity to the pcl genes
from Rb. capsulatus and H. halophila is present in the I.
loihiensis genome (Supplemental Figure 2, Supporting Informa-
tion). This gene (IL2385) encodes a protein with ∼28%
sequence identity and ∼43% sequence similarity to the known
pCLs from Rb. sphaeroides and H. halophila;22,34 all three
proteins are ∼400 residues in length. The crystal structure of
the 4-chlorobenzoate:CoA ligase from Alcaligines sp. AL2007
has been reported.38 This structure revealed functionally im-
portant residues in the substrate binding pocket. Comparison
with the protein encoded by gene IL2385 reveals that key
residues in the substrate binding pocket are conserved: His204,
Ile303, and Gly305 (numbering for 4-chlorobenzoate:CoA
ligase). In CoA ligases for small acyl substrates Gly305 is
substituted with Trp, supporting the conclusion that the I.
loihiensis enzyme binds the larger p-coumaryl substrate. Finally,
Val205 in the 4-chlorobenzoate:CoA ligase is substituted with
Ile in the CoA ligase in I. loihiensis. In plant p-coumaryl:CoA
ligases this position is important for substrate specificity, and
contains Ile,39 as does the I. loihiensis enzyme. Interestingly,
gene IL2385 in I. loihiensis is located immediately downstream
of the pyp gene in an apparent operon (Figure 6A). This is strong
independent support for its functional assignment as a pCL.
4. Discussion
Functional Diversity in the PYP Family. Hh PYP and Il PYP
share 75% sequence identity. This level of sequence similarity
ensures that the two proteins have very similar three-dimensional
structures. However, this sequence-based conclusion does not
transfer to their functional properties. For example, the absor-
bance spectra of the red and green pigments in human color
vision differ by 30 nm, while these proteins share 96% sequence
identity.40 Similarly, the absorbance peaks of the green and blue
variants of proteorhodopsin differ by 35 nm, but their amino
acid sequences are >78% identical.41 The lifetime of the pB
state in the PYP family varies from milliseconds to hours, and
single point mutations can greatly alter pB decay kinetics in
Hh PYP.6 The results reported here demonstrate that the
properties of Hh PYP and Il PYP are highly similar despite
their very different biological functions.
Figure 5. Time-resolved FTIR difference spectroscopy of the PYP from
I. loihiensis. (A) Three-dimensional plot of the time-resolved rapid-scan
FTIR difference absorbance signals during the photocycle of Il PYP in D2O
at pH* 6.6. (B) Light-induced FTIR difference spectra between the long-
lived pB intermediate and the initial pG state for the PYPs from I. loihiensis
(red) and H. halophila (blue). In the inset the difference spectra in the
spectral region 1780-1760 cm-1 are enlarged. (C) Kinetics of the last step
in the PYP photocycle of Il PYP (red) and Hh PYP (blue) monitored by
the peak to peak difference between the amide I signals at 1642 and 1624
cm-1. The solid lines are a biexponential fits of the data, with time constants
of 0.30 s (89%) and 4 s (11%) for Il PYP, and 0.26 s (92%) and 0.8 s (8%)
for Hh PYP.
of gene IL2450 and 26% identical to product of gene IL135.
These four proteins all are approximately 500 amino acids in
length.
In Rhodocista centenaria, a PYP-phytochrome hybrid con-
taining a histidine kinase domain regulates the activity of
A highly conserved Ala-Ser-Gly fragment of ammonia lyases
undergoes spontaneous cyclization to form the 4-methylene-
imidazole-5-one cofactor of the enzyme.35 This ASG sequence
is conserved in the two TAL candidates in I. loihiensis. The
crystal structure of the TAL from Rb. sphaeroides36 has revealed
9 residues that form the substrate binding pocket. Of these 9
residues 8 are conserved (Tyr60, Phe66, His89, Leu153, Tyr300,
Arg303, Asn435, and Gln436) in the two I. loihiensis proteins.
(37) Kyndt, J. A.; Hurley, J. K.; Devreese, B.; Meyer, T. E.; Cusanovich,
M. A.; Tollin, G.; van Beeumen, J. J. Biochemistry 2004, 43, 1809–
1820.
(38) Gulick, A. M.; Lu, X.; Dunaway-Mariano, D. Biochemistry 2004, 43,
8670–8679.
(39) Schneider, K.; Hovel, K.; Witzel, K.; Hamberger, B.; Schomburg, D.;
Kombrink, E.; Stuible, H. P. Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
8601–8606.
(40) Nathans, J.; Thomas, D.; Hogness, D. S. Science 1986, 232, 193–
202.
(35) Baedeker, M.; Schulz, G. E. Structure 2002, 10, 61–67.
(36) Louie, G. V.; Bowman, M. E.; Moffitt, M. C.; Baiga, T. J.; Moore,
B. S.; Noel, J. P. Chem. Biol. 2006, 13, 1327–1338.
(41) Man, D.; Wang, W.; Sabehi, G.; Aravind, L.; Post, A. F.; Massana,
R.; Spudich, E. N.; Spudich, J. L.; Beja, O. EMBO J. 2003, 22, 1725–
1731.
9
17448 J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009

Locked Chromophores Reveal PYP Function
A R T I C L E S
Figure 6. Model for the regulation of biofilm formation in I. loihiensis. (A) Genetic context of the Il pyp gene. Genes are represented to scale as arrows:
pyp is photoactive yellow protein (125 residues, IL2385), pcl is p-coumaryl:CoA ligase (422 residues, IL2386), isp is intracellular signaling protein, see
panel B (392 residues, IL2387). The length (number of basepairs) of the intergenic regions is also indicated. (B) Domain architecture of the intracellular
signaling protein (IL2387) immediately downstream of the pcl gene. The numbers of the residues at the start and end of each domain are indicated. (C)
Model for light-regulated biofilm formation in I. loihiensis. The proposed effect of PYP photoexcitation on the synthesis of c-di-GMP is indicated, together
with the resulting physiological changes that result in biofilm formation.
chalcone synthase, which is involved in the biosynthesis of
photoprotective pigments.42 The PYP from H. halophila func-
tions as the photoreceptor for negative phototaxis.43 The
biological function of the PYP from Thermochromatium tepidum
is not known, but it is part of a larger protein that also contains
a diguanylate cyclase domain.44 The data reported here reveal
that the biological function of the PYP from I. loihiensis is to
photoregulate biofilm formation. This indicates a high level of
functional versatility in the PYP family. The difference in
function between Il PYP and Hh PYP is in striking contrast
with the high level of conservation in both amino acid sequence
and a range of functional properties: the pCA absorbance
maximum and pKa, pB lifetime, and photocycle-associated active
site proton transfer and conformational changes. These data
suggest a recent divergence in the biological function of these
two PYPs.
for the output domains connected to the photosensors and in
the biological responses that they trigger.
Approach to Study in WiWo PYP Function. For 9 of the 12
bacteria known to contain PYP,6 no genetic tools are available.
This presents an important barrier to studies on the in ViVo
function of PYP, and has precluded genetic confirmation of the
function of Hh PYP in negative phototaxis. Here we develop
and apply the use of locked pCA analogs to examine the
involvement of PYP in light-regulated biofilm formation in I.
loihiensis. This provides a generally applicable tool for studies
on the in ViVo role of PYP, even in organisms that are not
genetically accessible. Thus, while the biological function of
the highly studied PYP from H. halophila as the photoreceptor
for negative phototaxis still has not been conclusively deter-
mined using a genetic approach, the results reported here provide
compelling biochemical evidence for the biological function of
the PYP in I. loihiensis.
Since PYP is the only photoreceptor known to contain pCA
as a cofactor, the detection of opposing biological effects upon
the addition of trans-locked and cis-locked pCA provides strong
evidence for a role of this photoreceptor in the photoresponse
studied. Coumaric acids are known to pass the microbial cell
envelope relatively easily.49 Presumably, after passing the cell
membrane, the pCA analog will be incorporated into PYP by
p-coumaryl:CoA ligase (pCL),34 depending on the substrate
specificity of the pCL. De noVo biosynthesis of PYP and PYP
turnover will thus result in the incorporation of pCA analogs
into PYP. The difference between cells exposed to trans-locked
The diversity of output signals generated by PYP photoex-
citation (phototaxis, pigment synthesis, and biofilm formation)
appears to be a general theme in photobiology. Also in the case
of LOV,45 BLUF,46,47 and phytochromes48 great diversity if seen
(42) Jiang, Z.; Swem, L. R.; Rushing, B. G.; Devanathan, S.; Tollin, G.;
Bauer, C. E. Science 1999, 285, 406–409.
(43) Sprenger, W. W.; Hoff, W. D.; Armitage, J. P.; Hellingwerf, K. J. J.
Bacteriol. 1993, 175, 3096–3104.
(44) Kyndt, J. A.; Fitch, J. C.; Meyer, T. E.; Cusanovich, M. A.
Biochemistry 2005, 44, 4755–4764.
(45) Crosson, S.; Rajagopal, S.; Moffat, K. Biochemistry 2003, 42, 2–10.
(46) Gomelsky, M.; Klug, G. Trends Biochem. Sci. 2002, 27, 497–500.
(47) Jung, A.; Domratcheva, T.; Tarutina, M.; Wu, Q.; Ko, W. H.; Shoeman,
R. L.; Gomelsky, M.; Gardner, K. H.; Schlichting, I. Proc. Natl. Acad.
Sci. U.S.A. 2005, 102, 12350–12355.
(49) Baranowski, J. D.; Nagel, C. W. Appl. EnViron. Microbiol. 1983, 45,
(48) Sharrock, R. A. Genome Biol. 2008, 9, 230.
218–222.
9
J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009 17449

A R T I C L E S
Horst et al.
and cis-locked pCA would be expected to reveal the maximal
physiological response triggered by PYP. This presents a
generally applicable chemical approach to examine the involve-
ment of PYP in a photoresponse.
stomatal opening, also exhibit slow photocycling times (from
minutes to hours).
Here, we present an example of a photosensory protein that
does not follow this trend. Il PYP exhibits a fast photocycling
time but triggers a slow biological response: biofilm formation.
Strikingly, despite the highly distinct biological functions of
the PYPs from I. loihiensis and H. halophila, their photocycling
times are essentially identical. Thus, the overall kinetics of a
photocycle cannot be directly used to infer information regarding
the time scale of the biological response that it triggers. This
result also has implications for modeling of signaling pathways
in the living cell based on kinetic information for the individual
signal transduction components.
Biofilm formation in I. loihiensis exposed to trans-locked
versus cis-locked pCA differs by a factor ∼5, while incubation
in the dark versus light results in a difference by a factor ∼1.6.
The smaller effect of light exposure is in line with the fairly
short lifetime of the pB state of Il PYP. The difference between
the two locked pCA analogs provides an estimate of the maximal
amplitude of this physiological response. Apparently, the light
intensities used in the experiments reported here do not trigger
this maximal response. In view of the pB lifetime of Il PYP,
this light intensity would be expected to yield a photostationary
state in which 5-10% of the PYP molecules in the cell are in
the pB state, which may yield a submaximal response. This
analysis suggests that Il PYP functions as a photoreceptor for
relatively high light intensities, providing an explanation for
the apparent mismatch between the photocycling time of Il PYP
and the time scale of the biological response that it triggers.
Light Regulation of Biofilm Formation. Biofilm formation is
a complex process. Two key external stimuli known to regulate
biofilm formation are nutrient availability and quorum sensing.61,62
Here we report a novel external stimulus that affects biofilm
formation: light.
Recently it was reported that cell attachment in Caulobacter
crescentus is regulated by blue light via the LOV-histidine
kinase photoreceptor protein LovK.63 Key differences with the
results reported here are that (i) attachment of C. crescentus
occurs not via expolysaccharides but through a specialized stalk,
(ii) in I. loihiensis a pCA-based PYP and not a flavin-based
LOV photoreceptor regulates cell attachment, (iii) blue light
inhibits biofilm formation in I. loihiensis but promotes cell
attachment in C. crescentus, and (iv) the photocycling time of
the LovK is on the hours time scale, while that of Il PYP occurs
on the subsecond time scale. For LovK it has been proposed
that light provides information on the position of the organism
in the water column.63 A similar biological role is plausible for
light-regulated biofilm formation by PYP in I. loihiensis.
A number of features in the I. loihiensis genome suggest a
model for the signal transduction chain in PYP-directed light
regulation of biofilm formation. The Il pyp gene is present in
an apparent operon with a downstream gene encoding a pCL
(Figure 6A). The two-gene pyp-pcl operon is immediately
adjacent to a gene encoding a multidomain signaling protein
that is transcribed convergently with the pcl gene and overlaps
with it by three base pairs. This signaling protein consists of
two N-terminal PAS domains and a C-terminal GGDEF domain
The degree of biofilm formation by I. loihiensis in the
presence of trans-locked pCA would be expected to be identical
to that for cultures grown in absolute darkness. However, the
experiments (Figure 1) showed a somewhat lower degree of
biofilm formation for cultures grown in darkness. Two possible
explanations for this observation are (i) that small amounts of
light reached the cultures during the course of the experiments,
or (ii) that a small amount of dark noise occurs in PYP
reconstituted with native pCA so that even in the absence of
light a basal level of signal is generated. The current data do
not allow these two effects to be distinguished.
Correlating the Time Scales of Receptor Activation and
Biological Response. The photocycling rates of photosensory
proteins cover a wide range of time scales, from milliseconds
to hours. In the PYP family the lifetime of the pB state varies
from minutes and hours for the PYPs from Rhodocista cente-
naria42 and Salinibacter ruber6,16 to hundreds of milliseconds
for the PYP from H. halophila.8,15 The photocycles for the PYPs
from Rb. sphaeroides and Rb. capsulatus are even faster.14,37,50
Interestingly, the lifetime of the signaling states51 in a
photocycle appears to match the typical time scale of the
biological response that they trigger.52,53 A correspondence
between the lifetime of an activated signaling component and
the time scale of the biological response that it triggers has also
been reported for two-component regulatory systems. In this
case the lifetime of phosphorylated response regulators varies
according to the duration of the response.54,55 Hh PYP and the
archaeal sensory rhodopsins both exhibit a photocycling time
of ∼0.5 s and both trigger phototaxis responses, which occur
on a subsecond to second time scale.43,56 The slower photocy-
cling time of the PYP from R. centenaria (∼50 s) corresponds
to its function in the regulation of gene expression.42 A similar
correspondence is seen in BLUF domain proteins.46 The BLUF
domain protein AppA functions as an antirepressor, and exhibits
a photocycle of 15 min.57 In contrast, the BLUF protein PixD
is a photoreceptor for phototaxis in Synechocystis, with a
photocycle time of ∼10 s.58,59 The LOV-based phototropins in
plants,60 which trigger phototropism, chloroplast movement, and
(50) Haker, A.; Hendriks, J.; van Stokkum, I. H.; Heberle, J.; Hellingwerf,
K. J.; Crielaard, W.; Gensch, T. J. Biol. Chem. 2003, 278, 8442–
8451.
(51) Yan, B.; Takahashi, T.; Johnson, R.; Spudich, J. L. Biochemistry 1991,
30, 10686–10692.
(52) Jung, K. H.; Trivedi, V. D.; Spudich, J. L. Mol. Microbiol. 2003, 47,
1513–1522.
(53) van der Horst, M. A.; Laan, W.; Yeremenko, S.; Wende, A.; Palm,
P.; Oesterhelt, D.; Hellingwerf, K. J. Photochem. Photobiol. Sci. 2005,
4, 688–693.
(54) Janiak-Spens, F.; Sparling, J. M.; Gurfinkel, M.; West, A. H. J.
Bacteriol. 1999, 181, 411–417.
(55) Thomas, S. A.; Brewster, J. A.; Bourret, R. B. Mol. Microbiol. 2008,
69, 453–465.
(56) Sundberg, S. A.; Alam, M.; Spudich, J. L. Biophys. J. 1986, 50, 895–
900.
(57) Masuda, S.; Bauer, C. E. Cell 2002, 110, 613–623.
(58) Masuda, S.; Hasegawa, K.; Ishii, A.; Ono, T. A. Biochemistry 2004,
43, 5304–5313.
(59) Okajima, K.; Yoshihara, S.; Fukushima, Y.; Geng, X.; Katayama, M.;
Higashi, S.; Watanabe, M.; Sato, S.; Tabata, S.; Shibata, Y.; Itoh, S.;
Ikeuchi, M. J. Biochem. 2005, 137, 741–750.
(60) Swartz, T. E.; Corchnoy, S. B.; Christie, J. M.; Lewis, J. W.; Szundi,
I.; Briggs, W. R.; Bogomolni, R. A. J. Biol. Chem. 2001, 276, 36493–
36500.
(61) Stanley, N. R.; Lazazzera, B. A. Mol. Microbiol. 2004, 52, 917–924.
(62) Boehm, A.; Steiner, S.; Zaehringer, F.; Casanova, A.; Hamburger, F.;
Ritz, D.; Keck, W.; Ackermann, M.; Schirmer, T.; Jenal, U. Mol.
Microbiol. 2009, 72, 1500–1516.
(63) Purcell, E. B.; Siegal-Gaskins, D.; Rawling, D. C.; Fiebig, A.; Crosson,
S. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 18241–18246.
9
17450 J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009

Locked Chromophores Reveal PYP Function
A R T I C L E S
(Figure 6B). The latter domain presumably exhibits diguanylate
cyclase activity, which catalyzes the conversion of GTP into
c-di-GMP.17-19 Recently c-di-GMP has emerged as a central
regulator for biofilm formation.20 The I. loihiensis genome
encodes 32 genes encoding enzymes involved in exopolysac-
charide synthesis and transport, 34 proteins containing an
GGDEF domain, and 22 proteins containing an EAL domain.5
The EAL domain is involved in the breakdown of c-di-GMP
through its phosphodiesterase activity.64 In addition, the I.
loihiensis genome contains 3 proteins that have been annotated
to contain a PilZ domain, which function as a receptor for c-di-
GMP.20 These observations suggest the possibility that Il PYP
regulates the diguanylate cyclase activity of the downstream
GGDEF protein (Figure 6C). A functional link between PYP
and c-di-GMP is confirmed by the PYP from Thermochroma-
tium tepidum, which contains a diguanylate cyclase domain.44
Genome-Based Identification of Photobiology in Chemotro-
phs. I. loihiensis was identified in samples taken 35 km off the
coast of Hawai’i at 1300 m below the water surface near a
hydrothermal vent and is closely related to an uncultivated
organism found at 11 000 m depth in the Mariana Trench.4 Thus,
I. loihiensis is regarded as a deep sea bacterium. At such depths
no sun light penetrates, and organisms in this environment live
in complete darkness. I. loihiensis is an aerobic heterotroph
capable of amino acid catabolism5 and has not been reported
to exhibit photobiological responses.
The results reported here that I. loihiensis contains a
functional PYP and exhibits light-induced suppression of biofilm
formation are strong evidence that this organism also lives in a
part of the ocean’s water column where its biology is influenced
by light. The inhibition of biofilm formation by light suggests
that planktonic I. loihiensis cells are favored in the light, while
biofilms form at greater depths. The metabolism of I. loihiensis
is centered around amino acid catabolism, and it has been
proposed that the organism grows on the proteinaceous particles
that are present near deep sea hydrothermal vents.5 I. loihiensis
is likely to form biofilms on such particles and may be dispersed
when such particles flow to the ocean’s surface.
it is likely that many novel light responses remain to be
discovered in these organisms.
A novel approach to study the in ViVo function of PYP was
developed based on the use of trans-locked and cis-locked
derivatives of the pCA chromophore in PYP, which results in
constitutively inactive and constitutively active PYP derivatives,
respectively. This approach was used to demonstrate that PYP
is the photoreceptor for light-induced suppression of biofilm
formation in I. loihiensis. The approach is generally applicable,
and is particularly valuable for studies on organisms for which
no genetic tools are available.
The genome of I. loihiensis encodes two enzymes that are
likely to have TAL activity, and in addition it contains a pcl
gene that forms an apparent operon with the pyp gene. This
indicates that I. loihiensis contains the two enzymes for pCA
synthesis and activation that allow holoPYP production.
The identification of Il PYP as the photoreceptor for biofilm
regulation in I. loihiensis broadens the range of biological
processes regulated by PYP. Such diversity in output signals is
an emerging theme from studies on different types of
photoreceptors.
Heterologously overproduced Il PYP is a yellow and pho-
toactive protein. Its visible absorbance maximum, pCA pKa,
active site proton transfer, and pB lifetime are very similar to
the well-studied PYP from H. halophila, despite the very
different biological function of Hh PYP. The genome of I.
loihiensis encodes proteins with clear homology to pCL
(immediately downstream of the pyp gene) and TAL for the
synthesis and activation of the pCA chromophore of PYP.
The fast kinetics of pB decay in Il PYP do not match the
slow biological function that it triggers, providing a counter-
example for a proposed correlation between the time scales for
the photocycle of a photoreceptor and the biological response
that it causes.
Acknowledgment. W.D.H. gratefully acknowledges support
from NIH grant GM063805 and OCAST grant HR07-135S, and
from startup funds provided by Oklahoma State University. K.J.H.
was supported by the “Molecules to cells” program from NWO. A.X.
was supported by OCAST grant HR02-137R and the Oklahoma
State Regents for Higher Education. We thank Dr. Shaobin Hou
and Dr. Maqsudul Alam (University of Hawaii) for providing
genomic DNA of I. loihiensis and Hans Biera¨ugel and Dr. Jan van
Maarseveen for synthesis of chromophore analogs.
5. Conclusions
I. loihiensis can form biofilms, and this process is inhibited
by light. This light regulation of cell physiology is unexpected
for a chemotrophic deep sea bacterium, and indicates that the
organism also occurs in illuminated areas of the water column,
close to the ocean surface. Genome sequencing projects have
revealed photoreceptors in many chemotrophic bacteria. Thus
Supporting Information Available: Synthesis method for
locked chromophores; TAL multiple sequence alignment; pCL
multiple sequence alignment. This material is available free of
charge via the Internet at http://pubs.acs.org.
(64) Schmidt, A. J.; Ryjenkov, D. A.; Gomelsky, M. J. Bacteriol. 2005,
187 (14), 4774–4781.
JA9057103
9
J. AM. CHEM. SOC. VOL. 131, NO. 47, 2009 17451