Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Polymer-chlorosome nanocomposites consisting of nonnative combinations of self-assembling bacteriochlorophylls Gregory Scott Orf, Aaron M. Collins, Dariusz M. Niedzwiedzki, Marcus Tank, Vera Thiel, Adam Kell, Donald A. Bryant, Gabriel A. Montano, and Robert E. Blankenship Langmuir, Just Accepted Manuscript • Publication Date (Web): 06 Jun 2017 Downloaded from http://pubs.acs.org on June 9, 2017
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Polymer-chlorosome nanocomposites consisting of non-native combinations of self-assembling bacteriochlorophylls Gregory S. Orf1, Aaron M. Collins2, Dariusz M. Niedzwiedzki3, Marcus Tank4,5, Vera Thiel4,5, Adam Kell6, Donald A. Bryant4,7, Gabriel A. Montaño2,*, and Robert E. Blankenship1,3,* 1
Departments of Chemistry and Biology, Washington University in St. Louis, St. Louis, MO 63130
2
Center for Integrated Nanotechnologies, Los Alamos National Laboratory, Los Alamos, NM 87545
3
Photosynthetic Antenna Research Center (PARC), Washington University in St. Louis, St. Louis, MO 63130
4
Department of Biochemistry and Molecular Biology, The Pennsylvania State University, University Park, PA
16802, USA 5
Department of Biological Sciences, Tokyo Metropolitan University, Tokyo, Japan 192-0397
6
Department of Chemistry, Kansas State University, Manhattan, KS 66506, USA
7
Department of Chemistry and Biochemistry, Montana State University, Bozeman, MT 59717
*
To whom correspondence should be addressed: Gabriel A. Montaño (
[email protected]) or Robert E. Blankenship
(
[email protected])
Abstract Chlorosomes are one of the characteristic light-harvesting antennas from green sulfur bacteria. These complexes represent a unique paradigm: self-assembly of bacteriochlorophyll pigments within a lipid monolayer without protein influence. Due to their large size and reduced complexity, they have been targeted as models for the development of bio-inspired lightharvesting arrays. We report production of biohybrid light-harvesting nanocomposites mimicking chlorosomes, composed of amphiphilic diblock copolymer membrane bodies that incorporate thousands of multiple types of natural self-assembling bacteriochlorophyll pigments derived from green sulfur bacteria. The driving force behind the assembly of these polymer1 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
chlorosome nanocomposites is the transfer of the mixed raw materials from organic to aqueous phase. We incorporated up to five different, self-assembling, pigment types into single nanocomposites that mimic chlorosome morphology. We establish that the copolymer-BChl selfassembly process works smoothly even when non-native combinations of BChl homologs are included. Spectroscopic characterization revealed that the different types of self-assembling pigments participate in ultrafast energy transfer, expanding beyond single chromophore constraints of the natural chlorosome system. This study further demonstrates the utility of flexible short-chain, diblock copolymers for building scalable, tunable light-harvesting arrays for technological use, as well as allows for an in vitro analysis of the flexibility of natural selfassembling chromophores in unique and controlled combinations.
Keywords Diblock copolymers, bacteriochlorophylls, chlorosomes, light-harvesting, ultrafast energy transfer, self-assembly
2 ACS Paragon Plus Environment
Page 2 of 40
Page 3 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Introduction Photosynthesis, the chlorophyll-based biological process by which light energy is transduced into chemical energy, is achieved in vivo by a tightly-regulated process of lightabsorption and photochemistry, controlled by light-harvesting (LH) antennas and reaction center complexes, respectively.1,2 LH antennas are highly diverse in form and function across disparate biological taxa.1,3 This diversity allows different photosynthetic organisms to adapt to wideranging photon fluxes and light quality.1 Efforts to develop artificial, bio-inspired LH molecules depend on understanding how natural LH antennas function.4 One of the most intriguing natural LH antennas is the chlorosome, which is found in green sulfur bacteria (GSB), some filamentous anoxygenic phototrophs, and one species of Acidobacteria.5,6 The chlorosome is a megadalton ellipsoidal complex in which a lipid monolayer surrounds tens-to-hundreds of thousands of individual bacteriochlorophyll (BChl) c, d, or e pigments that non-covalently self-assemble into large aggregate structures without influence of structural proteins. This contrasts with all other known natural LH systems, which rely on intricate protein coordination of pigment molecules. The local concentration of BChl inside each chlorosome can reach up to 1 Molar,7 yet these LH antennas have high energy transfer efficiencies and can support photosynthesis at extremely low light intensities.8 Their large dimensions (>100 nm on the long axis), high light absorption cross section, and programmed self-assembly attributes have earned chlorosomes increased attention from biomaterials scientists as a model to construct practical artificial LH complexes.5 BChl c, d, and e are unique among natural photosynthetic pigments in that they simultaneously contain a C-31 hydroxyethyl group and lack a sterically bulky C-132 methoxycarbonyl group, allowing for self-assembly through C-31 hydroxyethyl®central 3 ACS Paragon Plus Environment
Langmuir
magnesium coordination.5,9 BChl f, another member of this pigment group, has been generated recently via mutagenesis and assembles in the same manner as BChls c, d, and e.10–12 Selfassembly of the pigments alters their absorption properties, broadening and red-shifting their Qy (S0®S1) absorption band into the near-IR, as well as increasing the absorption bandwidth of the complex (Fig. 1). On one side of the chlorosome, BChl a molecules bind with multiple copies of the small hydrophobic protein CsmA to form a two-dimensional lattice structure called the baseplate, which functions as an energy sink for the self-assembled BChls c, d, e, or f.13–16
400
500
600
700
Wavelength (nm)
800
900
657
400
500
600
700
800
900
Wavelength (nm)
720
BChl e
660 400
500
600
700
Wavelength (nm)
800
BChl f
Absorbance (a.u.)
730
Absorbance (a.u.)
670
BChl d
Absorbance (a.u.)
750
BChl c
Absorbance (a.u.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 40
900
705
645
400
500
600
700
800
900
Wavelength (nm)
Figure 1: The structures of the self-assembling pigments from chlorosome-bearing bacteria. BChls c, d, and e are found in natural (or wild type) strains, while BChl f is currently found only in laboratory-derived mutants.5,10,11 Insets: the absorption spectrum of the pigment in monomeric, fully-solvated form (dotted line) and oligomeric, chlorosome form (solid line). Note that the chlorosome spectra also contain some BChl a absorbing around 800 nm and carotenoid absorbing around 450-500 nm. Data reproduced with permission from Ref. 5. Copyright 2013, Springer.
While knowledge of the in vivo biogenesis pathway of the chlorosome would aid in efforts to copy chlorosome-like self-assembly in vitro, this information is scarce.5,17 The hydrophobic interactions between groups of BChl c, d, e, or f molecules and the chlorosome lipid 4 ACS Paragon Plus Environment
Page 5 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
monolayer envelope are thought to initiate self-assembly of the BChls. Capturing and replicating this self-assembly process in vitro has been attempted for over 30 years using various strategies such as solid film deposition, binary solvent mixtures, addition of natural lipids, addition of detergents, addition of carotenoids, and de novo synthesis of pigment analogs.18–29 These attempts have produced varying levels of success, and the likely self-assembled molecular coordination structure in natural chlorosomes has only been demonstrated in the past few years.30,31 A methodology has recently been developed to create artificial, chlorosome-like complexes using a short-chain, amphipathic, diblock copolymer.32,33 The copolymer is a mixture of two polymers built from repeating units of butadiene and ethylene oxide. When slowly infused with aqueous medium, copolymer amphiphiles spontaneously assemble into small micelle-like structures.34 In the additional presence of BChl c, chlorosome-like structures are formed whereby thousands of the hydrophobic pigment molecules are encapsulated inside of a copolymer monolayer. The pigments within the micelles show high degrees of long-range order and energy transfer, as probed by linear dichroism and fluorescence.33 These structures, called polymer-chlorosome nanocomposites (PCNs), mimic natural chlorosomes in morphology and function.33 PCNs contain a fundamentally different organization scheme than detergent-BChl aggregates in that the polymer is able to adjust conformation to readily mimic a lipid monolayer and allow for ordering of BChl into long-range aggregates.33 Chlorosomes usually contain only one type of self-assembling BChl (i.e. only BChl c, or d, or e, but not mixtures).1,5 However, there are a few known exceptions in which two types of these molecules occupy the same chlorosome.35–40 In principle, having many species of pigment within a single chlorosome would expand the spectral range for light harvesting and therefore 5 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
enhance photosynthetic output. However, as such a situation does not seem to be widespread in nature, it is difficult to ascertain whether such a situation is energetically functional or optimal. The development of multi-pigment PCNs allows for such questions to be addressed. This study seeks to utilize the PCN methodology to investigate PCNs composed of multiple types of self-assembled BChls that overcome constraints of the natural chlorosome system, namely producing chlorosomes of tailored heterogeneous composition. There were two goals: (1) to create tunable PCNs using a combination of natural pigment types, and (2) to investigate whether the enhanced PCNs still maintain self-assembled chlorosome morphology and ultrafast energy transfer character. Addition of BChl a, the pigment found universally in the chlorosome baseplate complex, into the PCNs (even though it cannot self-assemble in the same way as BChls c, d, e, or f) was also investigated. Lastly, the PCNs are compared to atypical natural chlorosomes containing a mixture of BChls c and d.
Experimental Section Bacterial strains, growth conditions, and pigment purification Each type of BChl was extracted and purified from a different organism. Chlorobaculum (Cba.) tepdium TLS (for BChl c), Cba. tepidum ΔbchU (for BChl d), Pelodictyon phaeum (for BChl e), and Cba. limnaeum ΔbchU (for BChl f) were grown anaerobically as previously described.10,35,41 Rhodobacter capsulatus SB1003 (for BChl a) was grown semi-aerobically in the dark as previously described.42 BChls were extracted and purified by HPLC as previously described,43 except that the isocratic flow rate was increased to 1.5 mL min−1. The four main peaks for BChl c (4.0–6.25 6 ACS Paragon Plus Environment
Page 6 of 40
Page 7 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
min), BChl d (3.75–6.0 min), BChl e (3.25–5.25 min), and BChl f (3.0–5.25 min) corresponding to spectroscopically identical homologs differing in alkylation at the C-8 and C-12 positions were collected. Only one BChl a peak (10.5–12.0 min) was collected. Multiple HPLC replicates were pooled until 1 mg total of each pigment was collected, as determined by absorption spectroscopy. Pure pigment was dried under a stream of argon gas, transferred to an anaerobic chamber, sealed under positive pressure, and stored at –20 °C until needed. Chlorosomes from the GSB Prosthecochloris sp. HL-130-GSB that include both BChl c and d simultaneously were also purified for comparison. This strain was isolated in 2012 from a cyanobacteria-dominated mat in Hot Lake,44 a magnesium sulfate-rich, meromictic salt lake near Oroville, WA, USA.45 The strain was grown as reported previously,44 and the chlorosomes were isolated as reported previously.10 The chlorosomes were reduced by addition of sodium dithionite to 10 mM final concentration and tightly sealed from air prior to any spectroscopic characterization. Self-assembly of natural BChls with diblock copolymer Hydroxyl-terminated poly(ethylene oxide)-block-poly(butadiene) (PEO-b-PBD) with block weights of 1.3 and 1.2 kDa and a polydispersity index of 1.1 (product # P6723-BdEO) was purchased from Polymer Source, Inc. (Quebec, Canada). The PCNs were prepared by slowly infusing polymer and pigments dissolved in tetrahydrofuran (THF) with Tris-HCl buffer.33 Briefly, PEO-b-PBD, BChl c, d, e, f, or a (or combinations thereof) were mixed from stock solutions in THF to give final molar ratios of 20:1 pigment:polymer (total PEO-b-PBD concentration was 1.5 µM and total pigment concentration was 30 µM). 20 mM Tris-HCl buffer, pH 8.0, was added at 2 mL h–1 to samples with stirring. The final concentration of THF did not exceed 5% of the total solution volume. The assembled PCNs were then purified on continuous 7 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
sucrose (20-40%, v/v) gradients containing 20 mM Tris-HCl buffer, pH 8.0, at 250,000 ´g for 15 h at 4 °C. Finally, sucrose was removed via buffer exchange to fresh Tris-HCl buffer, resulting in fully-purified PCNs. These samples could be stored at 4 °C in the dark for least 180 days without noticeable degradation. Steady-state optical spectroscopy and spectropolarimetry All samples were adjusted with 20 mM Tris-HCl buffer, pH 8.0, to a Qy band absorbance value of 0.2. Steady-state absorption measurements were performed using a Lambda 950 UVVis-NIR spectrophotometer (Perkin-Elmer Inc., Waltham, MA). 5 K steady-state absorption spectra were measured by a HR125 Fourier transform spectrometer (Bruker Corp., Billerica, MA) with a spectral resolution of 4 cm-1. Prior to cooling, samples were mixed 1:1 (v/v) with a glass-forming solution (55:45 glycerol:ethelyene glycol, v/v) and placed in a gelatin capsule with a path length of 0.8 cm. Samples were cooled to 5 K inside a 10-DT Super Vari-Temp liquid helium cryostat (Janis Research, Woburn, MA). The sample temperature was read and controlled with a Model 330 temperature controller (Lake Shore Cryotronics, Inc., Westerville, OH). Steady-state fluorescence was collected using a customized PTI fluorimeter (Photon Technology International Inc., Birmingham, NJ). Excitation (bandwidth: 4 nm) was performed at 515 nm (to excite BChl e or f selectively, if they were present) or at 450 nm (to excite BChl c or d selectively, if e and f were absent), ensuring that the excitation priority was given in the following order: BChl f ; e ; d ; c. Circular dichroism (CD) spectra were collected using a J-815 spectropolarimeter (JASCO Inc., Easton, MD), scanning from 850 nm to 350 nm. The detector sensitivity was set to 200 mdeg, sampling speed was set to 50 nm min-1, bandwidth set to 2 nm, and integration time set to 1 s. Five scans of each sample were averaged to obtain noise correction and the averaged spectra 8 ACS Paragon Plus Environment
Page 8 of 40
Page 9 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
were then manually blank-corrected using the spectrum of pure buffer. All samples were adjusted to an identical Qy band absorbance in a 1-cm path before measurement. Linear dichroism (LD) data were acquired by casting gels containing the PCNs and 10% acrylamide, 0.4% bis-acrylamide, 0.05% ammonium persulfate, and 0.03% N,N,N',N’tetramethylethylenediamine into 1 cm × 1 cm cuvettes. The gels were then uniaxially compressed to 50% of the original width and measured with light polarized normal and parallel to the stretching axis.46 Picosecond time-resolved fluorescence and global fitting Time-resolved fluorescence (TRF) measurements of the quaternary PCN, quinary PCN, and chlorosome sample from GSB isolate HL-130-GSB were carried out using an instrument described previously.47 Excitation pulses were generated at 512 nm with a final frequency of 8 MHz (125 ns between subsequent excitations) and photon flux of 1 × 1010 photons cm–2 per pulse. In the streak camera, a fast speed M5676 single sweep unit and a slow speed M5677 single sweep unit (Hamamatsu Photonics USA, Middlesex, NJ) were interchanged depending on the necessary temporal resolution. The TRF spectra were cleared from random noise by recomposing the data from the dominant principal components using singular value decomposition. Global fitting of the TRF dataset was accomplished with the program ASUfit (provided by Dr. Evaldas Katilius and the School of Molecular Sciences at Arizona State University), which calculates a combination of monoexponential spectral-kinetic components following an unbranched decay model convoluted by Gaussian approximation of the instrument response function (IRF).47 The IRF approximation was confirmed by measuring the profile of light scattered onto the detector with no sample present in the system. The global fitting spectral-kinetic profiles, termed evolution-associated 9 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 40
fluorescence spectra (EAFS), were obtained for all three of the samples in which TRF was measured. Dynamic Light Scattering Dynamic light scattering (DLS) was performed on the PCNs at 633 nm on a Zetasizer Nano (Malvern Instruments Ltd, Malvern UK). DLS measures the diffusivity (D) of the PCNs (and can be used for chlorosomes as well) and calculates their hydrodynamic radii, dH, using Einstein’s equation: dH = kT/3πηD, where k is the Boltzmann constant, T is the absolute temperature in Kelvin, and η is the viscosity of the solution. Because PCNs are not spherical, DLS estimates the diameter of a solvated hypothetical solid sphere having the same diffusion coefficient as the PCN; this is the same approach taken with native chlorosomes.48 While this information does not provide all size dimensions of the PCNs, it allows for a comparison between PCNs of differing pigment content as well as size homogeneity.
Results and Discussion Mimicking natural photosynthetic LH antennas in vitro has been a difficult challenge. In this study, we have used an established method for synthesizing PCNs32–34 to generate heterogeneous nanocomposites that consist of mixed populations of self-assembling, naturally derived BChl c/d/e/f molecules. Our new collection of PCNs represents a systematic study of the effect of incorporating different combinations of the naturally derived pigments into each nanocomposite. We show that (1) the conformational flexibility of diblock copolymers and selfassembling BChl molecules allows for production of tunable, light absorbing nanocomposites capable of ultrafast excitation energy transfer, and (2) simple synthesis routines can produce nanocomposites with tailored pigment combinations. Lastly, because our PCNs mimic natural 10 ACS Paragon Plus Environment
Page 11 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
chlorosomes in morphology but expand upon their natural constraints, we will briefly address the biological question as to why individual species of chlorosome-bearing bacteria usually contain only one type of self-assembling BChl. Self-assembly of nanocomposites Sixteen PCN samples were synthesized using the self-assembly method described above. These represent all possible unary, binary, ternary, and quaternary combinations of BChl c, d, e, and f in equimolar amounts, plus one variable stoichiometry example, and one quinary combination that incorporated a small amount of BChl a. The PCNs are expected to mimic natural chlorosomes in morphology, except that the natural lipid monolayer is replaced with selfassembled diblock copolymer and no true analog baseplate is present because the CsmA protein was not added. Fig. 2A shows an example of sucrose gradient purification of the ternary, quaternary, and quinary PCNs. Interestingly, all PCNs containing BChl e showed two populations that differed in density. In the PCNs lacking BChl e, only one, denser population was observed. In the PCNs that contain two density populations, the absorption spectra of the two populations were nearly identical (data not shown). However, we studied only the denser populations of each PCN type, as these were previously shown to be optimized structures.33 The dense populations of the PCNs were examined by dynamic light scattering (Supplementary Information, Table S1), revealing hydrodynamic radii between 100-200 nm with a unimodal size distribution. From their size, we estimate that each individual PCN contains 105-106 BChls in its center.7 This is very similar to the size range for natural chlorosomes.48
11 ACS Paragon Plus Environment
Langmuir
Energy transfer BChl f
A
BChl e
BChl d
BChl c
B
20 - 40% sucrose
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 40
Diblock copolymer Heterogeneous BChl c,d,e,f assemblies BChl a aggregates
c+d+e
c+e+f c+d+f
c+d+e+f d+e+f
c+d+e+f+a
Figure 2: Separation of higher-rank PCNs, showing two main populations. (A) Sucrose density gradients (20-40% (w/v) sucrose) of the ternary, quaternary, and quinary PCNs. (B) Cross-sectional model of the quinary PCN (long axis running toward and away from the reader) with heterogeneously-distributed BChl homologs from the denser population separated in (A).
The design flexibility of self-assembling BChls to create tunable, efficient light absorbing nanocomposites The steady-state absorption and fluorescence emission spectra of each PCN sample are shown in Fig. 3. A table containing the wavelengths of maximum Qy absorption in each PCN sample is also provided (Table 1). In each fluorescence experiment, BChl e or f (the most blueshifted pigment in the Qy region and most red-shifted pigment in the Soret region) is preferentially excited at 515 nm. If BChl e or f was not present in the complex, then BChl d or c was excited at 450 nm.
12 ACS Paragon Plus Environment
Page 13 of 40
Absorbance
Intensity (a.u.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
Fluorescence
A
c only d only e only f only
B
1:1 c:d 1:1 c:e 1:1 c:f 1:1 d:e 1:1 d:f 1:1 e:f
C
1:1:1 c:d:e 1:1:1 c:d:f 1:1:1 c:e:f 1:1:1 d:e:f
D
1:1:1:1 c:d:e:f 1:1:1:1:0.2 c:d:e:f:a
350
450
550
650
750
850
650
750
850
Wavelength (nm)
Figure 3: Steady-state absorption and fluorescence spectra of all PCN samples. (A) unary nanocomposites, (B) binary nanocomposites, (C) ternary nanocomposites, (D) quaternary and quinary nanocomposites. Absorption spectra, normalized at Qy band maximum, are shown at left and fluorescence emission spectra are shown at right. The wavelengths of the maxima are collected together in Table 1.
13 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 14 of 40
Table 1: Summary of absorbance and emission maxima and corresponding FWHM from the steady-state absorption and fluorescence spectra shown in Fig. 3. The PCNs are grouped by red-most pigment in the following order: BChl f; e; d; c; a. The maxima and FWHM for all peaks have been solved via Gaussian multi-peak fitting in Origin 2015 (OriginLab Corp., Northampton, MA). aThere is an absorption and emission shoulder in the quinary PCN due to the inclusion of BChl a. Terminal Acceptor BChl f BChl e
BChl d
BChl c
BChl a
f only
704
Qy Abs. FWHM (nm) 41.8
e only
705
44.0
730
52.0
1:1 e:f
701
41.5
731
44.7
d only
743
58.3
764
50.4
1:1 d:e
713
56.4
763
59.4
1:1 d:f
715
52.7
751
45.7
1:1:1 d:e:f
711
44.9
749
52.6
c only
751
56.0
775
51.4
1:1 c:d
745
49.1
774
49.0
1:1 c:e
730
68.2
769
49.3
1:1 c:f
716
62.5
774
61.6
1:1:1 c:d:e
730
52.7
769
56.3
1:1:1 c:d:f
736
62.4
773
47.8
1:1:1 c:e:f
714
55.8
764
60.3
1:1:1:1 c:d:e:f
721
46.5
764
Pigment ratio
1:1:1:1:0.2 c:d:e:f:a
Qy Abs. Max. (nm)
725, 798
a
47.5, 45.2
a
Em. Max. (nm) 726
Em. FWHM (nm) 48.2
771, 816
62.6 a
53.5, 80.0a
Each unary PCN sample approximates the absorption and fluorescence spectrum (both in maximum and FWHM) of a natural chlorosome composed of its pigment minus baseplate contribution.5 Therefore, we believe that pattern seen in these PCNs is similar to that in natural chlorosomes.33 Interestingly, the BChl e unary PCNs’ absorption maximum in the Qy region (705 nm) appears at nearly the same wavelength as that from the BChl f unary PCNs (704 nm). In Chlorobium species that contain BChl e (e.g. wild-type C. limnaeum), the chlorosome absorbs maximally in the Qy region at about 720 nm, about 15 nm to lower energy than that in our PCN.10 But, the BChl e used in this study comes from P. phaeum, which has a chlorosome that maximally absorbs ~705 nm.49 The only difference between the BChl e from C. limnaeum and P. phaeum is the distribution of homologs differing in methylation at the C-8 and C-12 positions; C. 14 ACS Paragon Plus Environment
Page 15 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
limnaeum10 (in descending order of prevalence): [8-Propyl, 12-Ethyl], [8-Isopropyl, 12-Ethyl], [8-Ethyl, 12-Ethyl], and [8-Neopentyl, 12-Ethyl]; P. phaeum49–51 (probable, in descending order of prevalence); [8-Ethyl, 12-Methyl], [8-Ethyl, 12-Ethyl], [8-Propyl, 12-Ethyl], [8-Isopropyl, 12Ethyl], and [8-Neopentyl, 12-Ethyl]. It appears that this distribution can tune the absorption maximum of the nanocomposites to a similar level as the methylation at the BChl C-20 position or the formylation at the BChl C-7 position.5,50,52–54 This success encouraged us to synthesize nanocomposites that contain mixtures of pigment not observed in nature. Mixtures of two, three, and four different types of selfassembling BChls resulted in higher-rank PCNs (“binary,” “ternary,” and “quaternary,” respectively) with similarly-broad or slightly-broader absorption peaks, with mixtures incorporating BChls c and d generally having wider absorption peaks than those that did not (Fig. 3B-D). For the higher-rank PCNs, the Qy absorbance and fluorescence maxima do not seem to represent simple averages of the maxima of the individual pigments. Additionally, the FWHM of fluorescence of each higher-rank PCN is within ~10 nm of the average FWHM of the unary PCNs. These findings provide evidence for the level of pigment heterogeneity within each individual nanocomposite and the nanocomposite population as a whole (see next section). Linear dichroism (LD) spectra were collected and confirm (Supplementary Information, Figure S1) that the pigment assembly in each type of PCN is similar to that in natural chlorosomes. This includes long-range ordering of the BChls, with the net Qy dipole of BChl aggregates generally aligning along the long-axis of the complex.55,46,56 Circular dichroism (CD) spectra were also collected (Supplementary Information, Figure S1) and are S-shaped in the Qy region, indicative of an exciton-coupled system, but not perfectly conservative, likely due
15 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 40
to intrinsic CD of the pigments themselves or a contribution from interactions with transition dipole moments outside of our measurement range.46 The addition of BChl a to the quaternary PCN already containing BChls c, d, e, and f yielded particularly interesting results. In this “quinary” PCN, a sizeable absorbance shoulder is seen around 800 nm, a signature remarkably similar to the natural chlorosome baseplate.14,57 Additionally, a fluorescence shoulder is observed at 817 nm, red-shifted compared to free BChl a.43 Further, BChl a, which cannot self-assemble in the same manner as BChl c, d, e, or f, does not seem to disrupt the structure of the PCNs as a whole, as probed by low-temperature absorption spectroscopy (Supplementary Information, Figure S2). Even at 5 K, little resolution is observed in the Qy region of the absorption spectrum, indicating BChl a does not disrupt general pigment assembly. Evaluating the 790-810 nm region of the absorption spectrum confirms that the BChl a in our quinary PCN acts similarly to previously reported PCNs,33 suggesting that the BChl a may reside solely in the polymer layer and aggregates into a structure with an absorbance maximum red-shifted from ~770 nm (free BChl a) to 795-800 nm. The lack of a baseplate in the PCN apparently does not necessarily prevent BChl a from forming redshifted aggregates. This opens the intriguing possibility that the proteins of the baseplate may act more as an organizational matrix for the native chlorosomes, rather than purely acting as a facilitator for BChl a spectral tuning. Varying the identity and stoichiometric ratio of pigments used to create the PCNs illustrates the excellent absorptive tunability of the nanocomposites (Figs. 3B-D, 4). The absorption spectrum (bandwidth and wavelength maximum) of the higher-rank PCNs can be finely tuned by simply altering pigment stoichiometry, although equimolar ratios result in an absorption band with larger full-width at half-maximum (FWHM) (Fig. 4). While it is 16 ACS Paragon Plus Environment
Page 17 of 40
understood that the pigments used are from various species of GSB, to date, it has not been possible to generate nanocomposites that demonstrate spectral properties with this granular level of control. The PCN platform affords such control using only a limited set of starting pigment types. This is highly encouraging from an engineering standpoint, as precise absorption tunability while maintaining efficient energy transfer character is a desirable trait for the manufacture of light-absorbing pigment products. Because generating artificial photosynthetic structures that exhibit expanded spectral coverage over native systems has been a long-sought goal,20,21,58–62 the ease with which spectral tuning can be performed using the PCN methodology is an exciting result. Researchers in the field have generally relied upon synthetic organic chemistry to develop new classes of pigments to fill spectral gaps in artificial LH antennas, however, our PCN methodology demonstrates that one can also simply vary the stoichiometry of a limited palette of self-assembling pigments to fill precise spectral gaps.
c only e only 2:1 c:e 1:1 c:e
Qy Normalized Absorbance (a.u.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
400
500
704
600
726 741
700
751
800
900
Wavelength (nm)
Figure 4: Tuning of binary PCN absorption by varying the stoichiometric ratio of the two types of pigment. PCNs containing only BChl c or e are shown for comparison. The wavelengths of maximum Qy absorption are labeled for clarity.
17 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 18 of 40
Pigment heterogeneity in individual nanocomposites The common theme across all PCNs was general downhill energy transfer, as evidenced by fluorescence. This is especially noticeable in the ternary and quaternary nanocomposites. After selective excitation into BChl e or f around 515 nm, all PCNs maximally fluoresced at wavelengths longer than 750 nm, well beyond the normal fluorescence maximum of BChl e or f (Figs. 3B-D, 5, and 6). This, along with the sharp separation seen in sucrose density fractionation of the PCNs, indicates that each individual nanocomposite should contain all the pigment types that were added into solution during complex formation, producing stochastic heterogeneity (i.e., there are not individual nanocomposites that are homogenously BChl c only, and so on). Although PCN membrane thickness is not yet known, even at a lower membrane thickness such as ~2.5-3 nm, as previously reported for supported monolayers of PEO-b-PBD,34 sufficient coupling would not exist between individual PCNs to result in the observed energy transfer results. In addition, the complete absence of higher energy fluorescence indicates that the observed fluorescence is essentially dominated by intra-PCN energy transfer. For the higher-rank PCNs (i.e. binary, ternary, quaternary, quinary), the Qy absorbance maxima do not represent simple averages of the maxima of the individual pigments. A simple average would be expected if each individual pigment type only interacted with other pigments of its own type. For example, in the 1:1 d:f PCN, a simple average of the d-only and f-only PCNs would predict that this binary PCN would absorb maximally at 723.5 nm (average of 704 nm and 743 nm), however, the observed absorbance maximum for this binary PCN is 715 nm. This indicates that pigments of each individual type are not significantly spatially-separated from those of another type within the same nanocomposites, and are likely assembling and electronically-interacting with one another. Similarly, when considering the steady-state 18 ACS Paragon Plus Environment
Page 19 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
fluorescence data for the higher-rank PCNs, the wavelengths of maximum fluorescence for the higher-rank PCNs do not exactly match what is observed for a unary PCN containing the redmost absorbing BChl (nor are they averages of the fluorescence spectra of unary PCNs of the constituent BChls). If each individual nanocomposite contained all the types of BChls added during synthesis, but each BChl only assembled with its own type and remained completely spatially separated from the others, then we would expect to observe fluorescence, after energy transfer, with the wavelength maximum of the red-most BChl. For example, in the 1:1:1 d:e:f PCN, the observed fluorescence maximum occurs at 749 nm. But, if the BChl d, e, and f only assembled amongst their own type, we would expect to observe a fluorescence spectrum with maximum at 764 nm (the BChl d-only fluorescence maximum) or 740 nm (the average of BChl d, e, and f unary PCNs fluorescence). Based upon these observations, we hypothesize that within each higher-rank nanocomposite, there exist heterogeneous mixed assemblies of pigments of different types which have electronically coupled together to produce unique absorption spectra. Due to this stochastic mixing, there may exist “high energy” and “low energy” pools of aggregates within single nanocomposites; this is supported by TRF observations of fast intra-PCN energy equilibration in the quaternary and quinary PCNs (described in a later section). However, while the data presented here in total suggest that the full diversity of BChls added during synthesis are represented in equal probability within each individual nanocomposite, we cannot determine with certainty the composition or spatial organization of these different “high-energy” or “lowenergy” aggregates within individual nanocomposites.
19 ACS Paragon Plus Environment
Langmuir
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 20 of 40
Structural disorder in individual nanocomposites The low-temperature absorption spectra of the quaternary and quinary PCNs speak to their low amount of internal structural disorder. In the quaternary PCN, the absorption spectrum in the Qy region at low temperatures still smooth (Supplementary Information, Fig. S1). This provides further evidence that each individual PCN in solution contains all four pigments that were included during the self-assembly process and that they are likely evenly distributed in stochastic heterogeneity (as described above). If there were multiple populations of PCNs in which there were true heterogeneous mixtures of different pigment combinations, one would expect the spectrum to resolve into individual absorption peaks at cryogenic (5 K) temperatures; this is clearly not the case. At cryogenic temperature, the FWHM of the quaternary and quinary PCNs also increase; this is especially noticeable for the quinary PCN. It is well-established that excitonically-coupled pigment aggregates exhibit larger absorption bandwidth as the temperature is decreased. Coupled with the fluorescence response of samples containing multiple BChls in which fluorescence is limited to the lowest energy emitter, we conclude that individual PCNs contain all BChls present in each sample. Importantly, it should be noted that in some cases (particularly the 1:1 d:f and 1:1:1 d:e:f PCNs mentioned above), the spectral characteristics are significantly shifted from what one would expect for composite spectra, or simple addition, of the samples. One interpretation of such data is that the BChls, although chemically distinct species, selfassemble in a heterogeneous fashion such that they form new, hybrid excitonic states. Although there is a fair amount of broadening and non-uniform shifts, results indicate that there is an ability to tune energy levels within the range afforded by their constituent pigments by simple alteration of the stoichiometry of those pigments. The prospect of tuning a nanocomposite’s 20 ACS Paragon Plus Environment
Page 21 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
absorption by including (or not including), at any desired stoichiometry, a limited palette of chromophores is still compelling. It is simply not possible with these data to understand all the excitonic-coupling implications of self-assembly in the higher-rank PCNs. Therefore, we propose
further
spectroscopic
characterization,
including
two-dimensional
electronic
spectroscopy or hole-burning spectroscopy, to further explore these conclusions. Rapid energy equilibration within BChls aggregates in the PCNs The quaternary and quinary PCNs were chosen for analysis using picosecond timeresolved fluorescence at room temperature. Both PCNs were excited directly into the BChl f Soret band at 515 nm; BChls c, d, e, and a minimally absorb at this wavelength. Contour plots of streak camera response, as well as the global fitting of the TRF, are shown in Figs. 5 and 6.
21 ACS Paragon Plus Environment
Langmuir
1:1:1:1 c:d:e:f A
Counts Counts
TRF
max
Delay Time (ps)
120 100
0
80 60 40 20 0
720
740
760
780
800
Norm. Fluorescence (a.u.)
140
B
20
820
C
40
60
80
100
Delay Time (ps) D
Representative Spectra
60
30 ps 40 ps 50 ps 60 ps 70 ps
50 40 30 20
EAFS < IRF 35 ps
Amplitude (a.u.)
70
Fitted Kinetics 730 nm 750 nm 770 nm 790 nm Fits IRF
Wavelength (nm)
Fluorescence (a.u.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 40
10 0
720
740
760
780
800
820
720
Wavelength (nm)
740
760
780
800
820
Wavelength
Figure 5: Time-resolved fluorescence of the quaternary (BChl c, d, e, and f) PCN sample. (A) Contour plot of the streak camera image, (B) kinetic traces of fluorescence decay at different wavelengths obtained from global fitting, (C) representative fluorescence spectra taken at different delay times relative to the IRF, and (D) the evolution-associated fluorescence spectra (EAFS), the results of global fitting of the TRF data, showing that two spectral-kinetic components are responsible for the observed fluorescence.
22 ACS Paragon Plus Environment
Page 23 of 40
A
Counts
TRF
122.5
max
Delay Time (ps)
250 200
0
150 100 50 0
720
740
760
780
800
Norm. Fluorescence (a.u.)
1:1:1:1:0.2 c:d:e:f:a 300
B
20
820
C
60
80
720 nm 740 nm 760 nm 780 nm 800 nm Fits IRF
100
120
Delay Time (ps)
Representative Spectra 50 ps 75 ps 100 ps 150 ps 200 ps
100 80 60 40
D
EAFS < IRF 29 ps 122 ps
Amplitude (a.u.)
120
Fitted Kinetics
40
Wavelength (nm) 140
Fluorescence (a.u.)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Langmuir
20 0
720
740
760
780
800
820
720
Wavelength (nm)
740
760
780
800
820
Wavelength (nm)
Figure 6: Time-resolved fluorescence of the quinary (BChl c, d, e, f and a) PCN sample. (A) Contour plot of the streak camera image, (B) kinetic traces of fluorescence decay at different wavelengths obtained from global fitting, (C) representative fluorescence spectra taken at different delay times relative to the IRF, and (D) the evolutionassociated fluorescence spectra (EAFS), the results of global fitting of the TRF data, showing that three spectralkinetic components are responsible for the observed fluorescence.
The TRF spectra for both PCNs (Figs. 5A and 6A) could be satisfactorily globally fit using two or three monoexponentially decaying spectral-kinetic components. The representative spectra (Figs. 5C and 6C) show a gradual weighting of fluorescence to longer wavelengths as delay time increases. Global fitting (EAFS, Figs. 5D and 6D) shows a very short (
