A Novel, Near Infrared, Long Stoke's Shift Dye for Applications Ranging from Gels, to Cells, to Living Mice

Tao Ji,^a Alinafe Mpande,^a M. Catherine Muenker,^a Sean Orton,a Douglas Vizard,^a Gilbert Feke,^a Kerryn Medeiros,^a Stephanie Chiang,^a Victoria Jacobs,^a William McLaughlin,a W. Matthew Leevy,^a,^†

^aCarestream Health Molecular Imaging, 4 Science Park, New Haven, CT 06511

^†Correspondingauthor contact information:  E-mail: warren.leevy@carestreamhealth.com

Introduction

Experimental methods utilizing fluorescence are commonplace in the modern biological laboratory.  Fluorescent probes are now routinely used for imaging gels and blots for in vitro assays detecting proteins, DNA, RNA, and other biological macromolecules.  Further, fluorescent probes are used in a number of cell biology techniques like microscopy, and flow cytometry. Within the last 10 years, fluorescence based methods have become popular for non-invasive imaging of disease models in living mice.  Without question, fluorescence techniques are a cornerstone of biological experiments ranging from in vitro to in vivo studies. 

  A number of fluorophores with many different wavelengths have been commercialized for specific biological assays.  The vast majority of these reagents contain fluorochromes with green (~525 nm) or red (~600 nm) emission. Blue emitters (<475 nm) are less common, while near infrared (NIR) (>650) probes are increasing in prevalence. However, it is rare that a single fluorophore is suitable for use across different experimental platforms.  Thus, we leveraged the synthetic dye expertise at Kodak to develop a large Stoke's shift, NIR fluorophore capable of spanning the in vitro and in vivo space, and accommodating a multitude of filter sets that are common to most equipment.

RESULTS/DISCUSSION

A NIR fluorophore was synthesized to have an excitation maximum of 670 nm and a long Stoke's shift (LSS) of 85 nm.  This compound was named LSS670, and its structure is the subject of a pending patent.

Figure 1 shows the broad excitation spectrum of LSS670, and its wide emission at 755 nm.  LSS670 was prepared with an NHS-ester side group to enable convenient bioconjugation to antibodies or other bioactive ligands.  Here, we present the spectral properties of the free dye, the in vitro applications of its antibody conjugates, and experiments in living mice that demonstrate its usefulness in vivo.

Table 1 provides a comparison of LSS670 to other commonly used fluorophores in the same spectral region.  Several unique characteristics of the LSS670 fluorochrome are noted in Table 1.  First, the excitation and emission maxima of LSS670 are separated by 80 nm, while the difference with the other dyes is in the range of 20 nm.  When the widths of these spectra are also factored, the excitation and emission curves of LSS670 have far less spectral overlap than the other listed fluorochromes.  This is important since the excitation and emission components of filter sets are not permitted to overlap due to spectral cross-talk, or bleed through.  Thus, one must either sacrifice a considerable fraction of the excitation spectrum or capture a reduced portion of the emission to accommodate a given filter set.  Meanwhile the reduced overlap of excitation and emission spectra of LSS670 permits better signal detection.  A second important feature of LSS670 arises when considering the breadth of its excitation and emission spectra.  The range of suitable excitation and emission wavelengths for LSS670 overlaps with those of Texas Red, Cy5, Cy5.5, and numerous other fluorochromes.  Thus, LSS670 will remain a functional fluorochrome over a number of different filter sets ranging from red to near-infrared.  This flexibility is crucial when considering the variety of filter sets that are incorporated into a range of experimental fluorescence instruments.

The LSS670 molecule was conjugated to monoclonal antibodies (mAbs) to assess its compatibility as the reporter group of a targeted probe.  First, LSS-670 was attached to a goat anti-rabbit mAb.  A direct comparison of LSS670 and Alexa 680 (Invitrogen Corp.) secondary antibodies was performed on two identical western blots.  The blots were prepared by first taking two gels and injecting two individual lanes with 3T3 cell lysate (Rockland Immunochemical, Gilbertsville, PA).  In addition, one lane of ECL Plex Rainbow fluorescent marker (GE Healthcare Ltd., Piscataway, NJ) was added as a size standard.  After electrophoresis, the gels were transferred onto a membrane (Millipore, Billerica, MA) and blocked with 5% milk and PBS for 1 hour.  Both blots were then probed with a Rabbit Anti-HSP primary antibody (Santa Cruz, Santa Cruz, CA) at a 1:100 dilution in blocker for 1 hour.  After washing with PBS, one blot was incubated in goat anti-rabbit LSS670, and the second blot in goat anti-rabbit Alexa 680.  Both secondary antibodies had a concentration of 2mg/ml, and the final dilution in blocker for both blots was 1:2000.  The blots were then washed again, and 16-bit images acquired on a Kodak Image Station using a 650 nm wide-bandpass excitation filter and a 750nm emission filter.  The images were psuedocolored green and are shown in Figure 2.  The bands are clearly visible, indicating the LSS670 chromophore functioned effectively as a fluorescent reporter for this in vitro assay.  Furthermore, the presence of LSS670 did not prohibit the mAb from binding its target.  This particular experiment also revealed that the LSS670 conjugate is approximately 60% brighter than Alexa 680 at equal concentrations, in the given filter set.

    The LSS670-mAb conjugate was next used for fluorescence microscopy experiments.  We chose to stain Staphylococcus aureus NRS11 bacterial cells because their small size (< 1 µm diameter) presented a challenge for microscopic detection.  These cells (5 x 107 CFU) were incubated with a rabbit mAb with affinity for S. aureus surface antigens (ab20920, Abcam Inc).  After incubation, the cells were rinsed with PBS, and then resuspended and incubated with the LSS670 goat-anti-rabbit bioconjugate for 10 minutes (See Experimental section for detailed protocol).  Following another rinse, the cells were resuspended and 10 µL placed on a glass slide for fluorescence imaging on a Nikon TE2000 microscope with Cascade 512B camera.  Figure 3 shows a field of view of S. aurues cells from this experiment at 500X magnification.  The bacteria cells were readily detected, and appeared as halos due to localization of the fluorophore to the surface of these spherical cells.  Another interesting aspect of this experiment is that the cells were imaged in a Cy3 filter set (ex 540 +/- 13 nm, em 600 +/- 28 nm).  These results highlight the flexibility and functionality of LSS670 to give an appreciable fluorescent signal in a sub-optimal filter set during a challenging fluorescence microscopy experiment.

NIR fluorescent probes have emerged as a viable route to perform non-invasive imaging in living mice.  NIR photons exhibit ideal tissue penetration properties compared to other wavelengths in the optical range.  Thus, researchers have focused on developing probes that incorporate fluorophores with emission in the NIR.  In these studies, the basic goal was to detect and measure biological phenomena occurring within tissues and organs. 

  Since LSS670 has NIR emission, it was tested for its ability to provide a detectable optical signal in tissues during whole animal fluorescence imaging.  A cohort of three mice (BLB/C, Taconic Inc) was intravenously injected with 200 µL of a mixture of unconjugated LSS670 (75 µM) and Visipaque (272 mg/ml Iodine final).  Assuming a blood volume of 1 mL, the final concentration of LSS670 in circulation was approximately 13 µM.  Since LSS670 is a water soluble organic fluorophore with MW below 1 kD, it will be excreted through the renal pathway and stain the kidneys and bladder.  The same principle applies to the Visipaque, although it has a slightly higher molecular weight.  In this fashion, we could assess whether a 13 µM blood concentration of LSS670 would provide sufficient fluorescent signal to image structures like the kidneys and bladder, or other anatomical regions at depths ( < 1 cm) associated with these organs.  The Visipaque was used to enhance the contrast of the kidney and bladder during X-ray imaging, thus yielding the precise anatomical location of these organs in the animal.

  Multimodal imaging was performed ten minutes after injection of the LSS670 Visipaque mixture.  The cohort of mice was anesthetized and imaged on the Kodak Multispectral Image Station where each animal was given an X-ray, followed by a multispectral fluorescence image acquisition (see Experimental section).  This protocol was performed with the animals placed in both the dorsal and ventral positions.  Figure 4 presents images of one mouse from the cohort.  Frames A-C display the mouse as imaged from the ventral side, while D-F present the dorsal view.  Frames A and D present the X-ray of the animal in each position.  The Visipaque provided excellent contrast of the kidneys and bladder in the X-ray image.  Next, frames B and E show the LSS670 fluorescence images from both sides.  Strong signal was detected from the bladder in the ventral, or front view of the mouse (Frame B).  The anatomical origin of this fluorescent signal was confirmed by overlaying the fluorescence image on the X-ray (Frame C).  Meanwhile, when imaging the animal from the back, or ventral side, fluorescence from LSS670 was detected from the kidneys.  Once again, the anatomical origin of this fluorescent signal was confirmed by overlay on the X-ray image (Frame F).  Fluorescence signal was not observed from the kidneys in the ventral view, and vice versa with the bladder.  This was due to attenuation of the fluorescence signal as it passed through greater tissue depths noted at suboptimal viewing angles.  Without question, the LSS670 proved a useful fluorophore for in vivo applications at tissue depths approximating those of the kidneys or bladder.

CONCLUSIONS

  The LSS670 dye provides a flexible fluorescent reporter for use across different experimental platforms.  It is readily conjugated to biomolecules like antibodies, some of which were used to contrast the protein bands of a western blot.  Additional experiments showed that conjugates could be used for the detection of S. aureus bacteria during fluorescence microscopy.  Finally, untargeted LSS670 gave more than enough signal to image tissues like the kidneys and bladder in living mice.  The LSS670 fluorophore is unique in that it is both near infrared and has a long Stoke's shift.  This compound will permit the consolidation of disparate fluorescent probes in research labs into one probe that may be used across the varying filter sets incorporated into fluorescence research equipment.    

EXPERIMENTAL

In vitro imaging of western blots.  For western blot analysis, a direct comparison of LSS670 and Alexa 680 (Invitrogen Corp.) secondary antibodies was preformed on two identical gels.  Each gel contained two lanes with 12 µg of 3T3 cell lysate (Rockland Immunochemical, Gilbertsville, PA) along with one lane of ECL Plex Rainbow fluorescent marker (GE Healthcare Ltd., Piscataway, NJ) which had Cy3 and Cy5 bands for a size standard.  After electrophoresis, the gels were transferred onto Immobilon-FL PVDF membrane (Millipore, Billerica, MA) and blocked with 5% milk and PBS for 1 hour.  Both blots were then probed with a Rabbit Anti-HSP primary antibody (Santa Cruz, Santa Cruz, CA) at a 1:100 dilution in blocker for 1 hour.  After washing four times in PBS, Blot 1 was incubated in goat anti-rabbit LSS670 (Carestream Health Inc., Rochester, NY) and Blot 2 was incubated in goat anti-rabbit Alexa 680 (Invitrogen).  Both secondary antibodies have a concentration of 2mg/ml, and the final dilution in blocker for both blots was 1:2000. The blots were then washed in PBS for 3 x 5min, and imaged on a Kodak Image Station using a 655nm wide-bandpass excitation filter and a 750nm emission filter (Figure 2), then psuedocolored green. 

Microscopy.  Staphylococcus aureus NRS11 cells were grown in LB Miller broth to a concentration of 5 x 107 CFU/ml.  Then, 1 mL of cells were pelleted by centrifugation (17,000 RPM, 30 sec) and resuspended in 200 µL of TES buffer (5mM TES, 145 mM NaCl, pH 7.4).  Then, 50 µL of monoclonal antibody to S. aureus (4 mg/ml, ab20920-100, Abcam Inc.) was added to the cells, which were then incubated for 10 minutes.  The cells were then pelleted by centrifugation, washed twice with TES buffer, and resuspended in 200 µL of the same buffer.  Next, 50 µL of LSS670 conjugated Goat-anti-Rabbit mAb (1 µM stock) was added.  After two more rinse steps in equal volumes of TES buffer, the cells were resuspened and imaged on a Nikon TE2000 microscope with a Cy3 filter set from Chroma Inc.  16-bit images were acquired using a Cascade 512B camera, then false colored red and converted to an RGB color image in ImageJ v1.40g.

In vivo multimodal imaging.  One male BALB-Swiss & Webster mouse (10 weeks old, 25 grams) was anesthetized through gaseous administration of Isofluorane (IsoSolTM), and 0.5 mL of saline was subcutaneously injected in the shoulder region to prevent dehydration.  A 200 µL solution consisting of VisipaqueTM (272 mg Iodine/mL) and LSS670 dye (75 µM final conc) was injected intravenously via the tail vein.

  Ten minutes after injection, the animal was placed in a prone position on a commercially available animal tray and imaged using the Kodak Multispectral Image Station.  A digital X-ray image (90 s, no binning, f-stop 2.8, FOV 99.44, focal plane 5.04. 0.4mm X-ray filter) was acquired to show anatomical orientation and iodine-dependent renal X-ray contrast resulting from injection of VisipaqueTM.  A 60 second optical image (2x2 binning, f-stop 2.8, FOV 99.44, focal plane 11.19) was also taken by illuminating the specimen with an excitation wavelength of 670 nm and capturing the emitted light with a 790 nm emission filter.   These digital capture settings were then repeated as the mouse was positioned supine for a dorsal view image capture.  After dorsal view capture, the mouse was returned to its cage, and an additional two mice were subjected to this experimental protocol. After image acquisition, multi-modal co-registration of the fluorescence and x-ray images was accomplished through image overlay in ImageJ v1.40g software.