Molecular imaging of small animals with fluorescent proteins: From projection to multimodality

Abstract

Fluorescent proteins (FPs) have been widely adopted in cell research for protein trafficking and reporter gene expression studies, as well as to study other biological processes. However, biological tissue has high light scattering and high absorption coefficients of visible light; hence, using FPs in small animal imaging remains a challenge, especially when the FPs are located deep in the tissue. In small animals, fluorescence molecular imaging could potentially address this difficulty. We constructed fluorescence molecular imaging systems that have two modes: a planner mode (projection imaging) and a multimodality mode (fluorescence molecular tomography and micro-CT). The planner mode can provide projection images of a fluorophore in the whole body of a small animal, whereas three-dimensional information can be offered by multimodality mode. The planner imaging system works in the reflection mode and is designed to provide fast imaging. The multimodality imaging system is designed to allow quantification and three-dimensional localization of fluorophores. A nude mouse with a tumour targeted with a far-red FP, which is appropriate for in vivo imaging, was adopted to validate the two systems. The results indicate that the planner imaging system is probably suitable for high throughput molecular imaging, whereas the multimodality imaging system is fit for quantitative research.

Introduction

Fluorescent proteins (FPs) have had a significant impact on research in the fields of cell biology and molecular biology [1]. They can be adopted in research to reportor gene expression and to study protein trafficking and other biological processes in live cells noninvasively [2], [3]. In the last decade, FPs have mainly been employed in live cell imaging ex vivo. Some types of fluorescence microscopy, such as wide-field fluorescence microscopy [4], confocal microscopy and multi-photon microscopy have been used widely all over the world in conjunction with FPs [5]. However, in ex vivo research, the original internal environment and various molecular feedback pathways cannot be kept intact [6]; the fact that in vivo imaging of the biological environment can be maintained makes these techniques invaluable to drug discovery and fundamental disease research [6].

However, two barriers have obstructed the progress of in vivo imaging of FPs. FPs themselves are the first challenge. The excitation and emission wavelengths of most FPs used in practice are in the visible band [3]. Unfortunately, the extinction coefficient of haemoglobin, which is ubiquitous in living tissue, is so high that the light emitted from FPs cannot easily penetrate the tissue of a living small animal with a modest noise level [7]. Furthermore, autofluorescence from small animal tissue decreases the signal-to-noise ratio (SNR) of fluorescence. The far-red FPs developed in recent years can partially fulfil the requirements of in vivo imaging [8]. Near infrared FPs will be the ultimate solution because the absorption coefficient and autofluorescence of tissue are quite low in the near infrared band, although the quantum efficiency of such FPs will still need to be enhanced [9]. The characteristics of fluorescence imaging systems are the other obstacle. Although fluorescence microscopy is appropriate for imaging FPs in vivo in some instances, the imaging depth of such systems is limited to about 1 mm [10]. In recent years, many research groups have pursued deeper imaging in living small animals. Different types of fluorescence imaging systems for whole-body small-animal imaging have been proposed to address this issue [11].

Two different modes of fluorescence molecular imaging have been proposed: planner mode and tomographic mode (fluorescence molecular tomography, FMT) [7]. Planner mode means that the projection image of fluorophores in small animals is acquired, and this can be acquired via two modalities: reflection mode and transmission mode. The excitation light source and the detector, which is used to collect fluorescence, are located on the same side of the animal in reflection mode. In contrast, they are located on different sides of the animal in transmission mode. FMT is different in planner mode, in which the fluorophore can be quantitatively imaged in vivo in three-dimensions. There are three types of FMT, and they differ in terms of the laser used in the system [11]: time domain, frequency domain and continuous wave domain FMT. Picosecond or femtosecond pulsed lasers, light modulated in time or space, and a continuous wave laser are adopted in the aforementioned three FMT systems, respectively [12], [13], [14]. However, there are still some barriers that make FMT difficult to employ in practice. First, due to the absence of any structural information, it is difficult to locate the fluorophore in the bodies of small animals. Furthermore, the highly light scattering nature of biological tissue leads to poor image quality and unreliable quantification results [15]. In recent years, multimodality imaging approaches have been introduced in optical molecular imaging. X-ray computed tomography [16], magnetic resonance imaging [17], and photoacoustic tomography [18] have been combined with FMT to improve imaging capabilities. X-ray computed tomography has the advantages of low cost and high imaging speed with high resolution [19]. Several dedicated multimodality systems combining FMT with micro-CT have been constructed in recent years [20], [21], [22] that, due to the combined system could perform with full capacity of sub-systems [15].

In this paper, we will focus on the continuous wave based fluorescence imaging systems for small animal imaging. The imaging principle and instrumentation of a planner imaging system and a multimodality system will be introduced. The application of the systems to small animal imaging using a far-red fluorescent protein will be exhibited to demonstrate the feasibility of the systems.