Fluorescent Proteins and Their Use in Marine Biosciences, Biotechnology, and Proteomics

Abstract

This review explores the field of fluorescent proteins (FPs) from the perspective of their marine origins and their applications in marine biotechnology and proteomics. FPs occur in hydrozoan, anthozoan, and copepodan species, and possibly in other metazoan niches as well. Many FPs exhibit unique photophysical and photochemical properties that are the source of exciting research opportunities and technological development. Wild-type FPs can be enhanced by mutagenetic modifications leading to variants with optimized fluorescence and new functionalities. Paradoxically, the benefits from ocean-derived FPs have been realized, first and foremost, for terrestrial organisms. In recent years, however, FPs have also made inroads into aquatic biosciences, primarily as genetically encoded fluorescent fusion tags for optical marking and tracking of proteins, organelles, and cells. Examples of FPs and applications summarized here testify to growing utilization of FP-based platform technologies in basic and applied biology of aquatic organisms. Hydra, sea squirt, zebrafish, striped bass, rainbow trout, salmonids, and various mussels are only a few of numerous instances where FPs have been used to address questions relevant to evolutionary and developmental research and aquaculture.

Appendices

Appendix A: Further Examples of Hydrozoan FPs

AceGFP was cloned from Aequorea coerelescens, the Pacific belt jellyfish (Gurskaya et al., 2003). This variant stands as a monomeric green-emitting (505 nm) mutant of a colorless GFP-like protein. Historically AceGFP served as N-terminal or C-terminal fusion tag for screening and FRET-based applications and had relatively low toxicity when expressed in vertebrate or invertebrate cell lines. More recently, AceGFP was further developed into a photoswitchable dual color (green and cyan) FP termed PS-CFP2 (Chudakov et al., 2004). Xia and co-workers (2006) found that the new homolog can be productively employed as a photostable blue chromophore in tricolor fusion protein imaging with EGFP and the coral-derived monomeric red FP, mRFP, using widefield fluorescent microscopy with common filter sets and inexpensive xenon light source.

TagGFP is, or was formerly, also known as macGFP and embodies a green-emitting (505 nm) enhanced version of GFP from the leptomedusa Aequorea macrodactyla commonly found in the East China Sea (Xia et al., 2002). In comparison with its wild-type parent, monomeric TagGFP comprises at least 14 amino acid substitutions and attains sufficient photo and pH stability. It is optimized for protein labeling and stable expression in long-term cultures.

JRed was engineered from the anthomedusan jellyfish purple chromoprotein, anm2CP, and provides a rare example of a radically red-shifted FP from a hydrozoan (Shagin et al., 2004). Owing to its real red (610 nm) emission, JRed can be combined with cyan, green, and yellow FPs in one multicolor experiment. This monomeric mutant was designed for protein localization studies and stable cell line generation. Even at high levels of overexpression, fusion proteins of JRed do not undergo aggregation. However, its relatively high photobleaching rate may limit the sensitivity of this probe in some applications.

PhysRFP is featured by a patent claim as a putative red FP from the Pacific bluebottle jellyfish Physalia utriculus (Yanagihara, 2002). This chromophore reportedly has excitation and bimodal emission maxima located in near-ultraviolet and green/far-red, respectively, and it does not undergo ultraviolet-induced green-to-red photoconversion. No other FPs to date are portrayed with such a red-shifted emission from this aquatic class; hence this largely uncharacterized protein may be a potential area for future exploration and substantiation.

Appendix B: Further Examples of anthozoan FPs

dsCFP or dsFP483 is an unusual cyan FP at the blue end of the color spectrum (Matz et al., 1999). It was recently crystallized from the Indo-Pacific soft mushroom coral Discosoma striata (Wang et al., 2005). dsCFP distinguishes itself with violet (443 nm) excitation and cyan (483 nm) emission maxima, as well as a hexameric organization in the crystal structure. Owing to its spectral properties, dsCFP may come into view as potentially efficacious FRET probe in combination with yellow emitting FPs.

The scleractinian coral Galaxea fascicularis provided a green-emitting (505 nm) bright FP named Azami Green (AG) which was adapted for use in a wide range of applications (Karasawa et al., 2003). AG absorbs light at bluish green (492 nm) wavelengths and forms a tetrameric complex. Unlike many other FPs, however, this protein is fairly stable in both acidic and basic medium and can be used to identify cells or to report gene expression under suboptimal physiological conditions. A series of mutations were also introduced to produce a monomeric variant with optimized properties.

cmGFP or cmFP512 was cloned from the tentacles of the azooxanthellate coral Cerianthus membranaceus (Wiedenmann et al., 2004a). This green-emitting protein is characterized by an excitation maximum at green (503 nm) wavelengths and a relatively bright emission in the yellowish-green (512 nm) region. The interactions between the subunits of the tetramer appear weak which makes cmGFP a potential candidate for future engineering of monomeric mutants.

Appendix C: Further Examples of Photoactivatable FPs

The so-called blinker protein Dronpa is an anthozoan green FP with photochromic properties whose fluorescence can be reversibly modulated- turned on and off on demand-by exposure to different wavelengths of light. Dronpa was derived from Pectinidae coral and is known as one of the brightest FP (Ando et al., 2004). Usefully, it is a monomeric protein unlike most PAFPs. Dronpa’s emission decreases following excitation with green light and the photoconverted protein regains its fluorescence after illumination with blue light. This pattern of inactivation and activation can be repeatedly executed, which is useful for studying protein movement and trafficking inside cells. The rather fast bleaching rate of this protein, however, requires rapid imaging and proper attenuation. In a string of recent publications, the photoswitching properties of Dronpa were characterized in detail by single-molecule and fluorescence correlation spectroscopy (Habuchi et al., 2005b; Dedecker et al., 2006) and the mechanism responsible for its reversible photoswitching was described as a simple photo-induced protonation/deprotonation process (Habuchi et al., 2006).

Semi-rational mutagenesis of the amino acid residues surrounding the chromophore played a key role in engineering Kikume Green-Red FP (KikGR) from the coral Favia favus (Tsutsui et al., 2005). KikGR emits green fluorescence that can be irreversibly converted to 593-nm red fluorescence by illumination with ultraviolet or violet light or by two-photon excitation in the far red domain for increased resolution. The conversion is highly dependent on the irradiation wavelength. The normal excitation light used to emit red or green fluorescence does not induce photoconversion permitting convenient regional labeling. The relative ratio of green to red fluorescence can be used for real-time tracking of proteins or cells.

KillerRed was developed from the hydrozoan chromoprotein anm2CP (Bulina et al., 2006). It is a dimeric red FP that can generate reactive oxygen species upon green light illumination. In this sense, KillerRed is fundamentally different from the other PAFPs since here it is a chemical reaction that can be activated and not the fluorescence. KillerRed acts as a photosensitizer but unlike chemical photosensitizers, it can be expressed by cells either individually or in fusion with a target protein. The photo-generated oxygen radicals can damage the neighboring cells which allow precise inactivation of selected proteins in chromophore-assisted light inactivation and high-specificity light-induced cell killing. KillerRed may also allow light-induced killing of tumor cells opening new perspectives for photodynamic therapy.

Appendix D: Calculation of Actual and Apparent Brightness

A measure of brightness is frequently calculated in the literature as E ∼ ηɛϕ where E is the actual, or physical brightness, ɛ the molar extinction coefficient (in M ⁻¹ cm ⁻¹), and ϕ the quantum yield (Chudakov et al., 2005; Shaner et al., 2005). η stands for an arbitrary scaling variable, typically set to 1.0 × 10⁻³ for expressing E in mM ⁻¹ cm ⁻¹. While this definition is simple and conveniently linear it is not consistent with the human perception of brightness which, at a given adaptation level, is not linear. To account for this nonlinearity, brightness can be calculated by a power law in the form of B = αL ^p (Bodmann et al., 1980). In this equation B denotes a measure of the apparent, or perceived brightness, in units of nominal brightness values corresponding to a stimulus of luminance under a dark background. The exponent p has a value of 1/3 and α is parameter of observing conditions which can be set equal to 25 cd ^–1/3 m ^2/3 for zero surround luminance. L increases as a linear function of the luminous intensity of the emitted light, i.e. L ∝ I over the unit area. I is given by I = βɛϕ in cd m ⁻². Here ɛ is used in terms of SI units, 10⁻¹ m ² mol ⁻¹ (which equals M ⁻¹ cm ⁻¹). β, a normalization constant, is set to 2.35 × 10⁻³ cd mol m ⁻⁴ to optimally match the physical brightness of FPs upto 100 mM ⁻¹ cm ⁻¹. To display brightness as a gray value, E and B can be straightforwardly converted to standard 8-bit grayscale units in the RGB color space.