1.

Introduction

Breast cancer is one of the most diagnosed and leading causes of cancer deaths in women. Clinically, breast cancers are divided based on its prognostic and therapeutic implications. Triple-negative breast cancers (TNBC) do not express estrogen receptor or progesterone receptor, and do not have HER-2/Neu amplifications. TNBC’s suffer from poor prognosis and have limited response to targeted therapy techniques.¹ 10% to 15% of all breast cancers are reported to be TNBC.² While TNBC is treatable by chemotherapy, a noninvasive and nonradioactive imaging modality to track the occurrence of apoptosis is needed for an effective treatment plan.

The common targets for specific imaging probes for apoptosis include activated effector caspases^3,4 and probes that bind to components of the scrambled plasma membrane such as phosphatidylserine (PS) that are transiently exposed on the outer surface of the cell during apoptosis.^4–6

Caspases are cysteine-aspartate proteases that play a pivotal role in the regulation of apoptosis. The intrinsic and extrinsic pathways of apoptosis lead to a transient increase of a small number of intracellular executioner caspases: caspase-3 and caspase-7. To successfully reach the caspase-3 and caspase-7 targets, the apoptosis probes have to penetrate the cell membrane. If there is no apoptosis, the probes do not have the ability to leave the cells resulting in false positives. Despite such drawbacks, there have been several developments on optical and nuclear imaging of apoptosis based on fluorescent probes and caspase inhibitors such as isatins^7–9 for in vivo applications. Although these caspase inhibitors have been proposed for imaging of the activated caspases, its rapid and transient increase renders imaging of this apoptosis biomarker particularly challenging.¹⁰ In fact, every patient may show a different ideal temporal window for imaging, which occurs between 6 and 24 h after treatment. This requirement to choose the best time to image may limit the clinical application of these agents.

On the other hand, PS is a major anionic phospholipid, representing 2% to 10% of total cell phospholipids. Its externalization occurs within a few hours of an apoptotic stimulus by an active transport mechanism that leads to presentation of 50 to 100 million molecules per apoptotic cell on its surface. This makes PS an abundant and accessible target for molecular imaging. During apoptosis, PS is exposed on the cell surface facilitating redistribution of phospholipids across the bilayer, making it an accessible target for molecular imaging.^5,6 Currently, annexin V labeled with radioisotope ${}^{99\mathrm{m}}\mathrm{Tc}$ is the most widely used imaging probe for PS-based detection of apoptosis that shows an elevated uptake of the tracer in some lesions that respond to chemotherapy.^4,5,11 Annexin V is also one of the few cell death imaging agents that has reached phase II/III clinical trials. However, annexin V has a low PS binding affinity and the binding requires the presence of calcium. In many cases, the endogenous Ca++ level is not high enough to facilitate the binding between annexin V and PS; hence, the localization of radiolabeled annexin in apoptotic cells is suboptimal. Also, this technique is limited by radioisotope half-life, which may require frequent injection and repeated imaging.

We propose to address these limitations by labeling, indocyanine green (ICG), a nonradioactive FDA approved near-infrared (NIR) absorbing probe with a long plasma half-life PS monoclonal antibody (mAb) and image the apoptotic tumor site with a relatively new imaging modality known as multispectral optoacoustic tomography (MSOT). Apart from anatomical information, multiwavelength optoacoustic imaging provides functional information of the tissue noninvasively with high contrast and good spatial resolution based on optoacoustic or photoacoustic effect.^12–14 The use of exogenous NIR contrast agents such as ICG further improves the capabilities of MSOT to provide additional deep functional information of tissues.

2.

Materials and Methods

2.1.

Conjugation and Purification

In this work, PS mAb (Cell Signal Solutions, New Jersey) was tagged with succinimide activated ICG (AAT Bioquewst Inc., California). The desired concentration of antibody solution was prepared by diluting the PS mAb stock ($400\mu \mathrm{g}/\mu \mathrm{L}$) in $600\mu \mathrm{L}$ of phosphate buffer solution. For conjugation, 1 mg of ICG-sulfo-EG4-OSu was dissolved in $50\mu \mathrm{L}$ of dimethyl sulfloxide (DMSO) and added to the antibody solution to keep the final concentration of DMSO to $\le 5\%$. The mixture was incubated at room temperature for 1 h and subjected to gel filtration purification using a G-25 gel filter column. The eluted solution contained the conjugated products while the unbound ICG was trapped in the gel column. The conjugation and purification process is summarized in the schematic as shown in Fig 1. NIR spectrometry and multispectral optoacoustic scans (iThera Medical GmbH, Munich, Germany) were performed on the eluted ICG-PS mAb and chemotherapeutic drug doxorubicin (Sigma-Aldrich, St Louis, Missouri) to determine the maximum absorption peaks in the NIR wavelength range of 680 to 980 nm.

Fig. 1

Schematic of conjugation and purification process of ICG and PS mAb.

3.

Results

The greenish appearance of the eluted product was an indication of successful conjugation of ICG to PS mAb while the unconjugated ICG remained in the gel column. NIR spectroscopy of the conjugates later confirmed the binding of ICG to antibody. Like ICG-PS mAb, doxorubicin also has a longer plasma half-life that may contribute to the optoacoustic signal generation during MSOT scan resulting in a false-positive signal in the tumor region. To overcome this, NIR absorbance of both doxorubicin and ICG-PS mAb were performed to check for absorption maximum in the NIR region. NIR absorption spectrum [Fig. 2(a)] of the conjugates showed peaks at 715 and 780 nm due to ICG while doxorubicin did not show any absorbance. The results not only confirm the binding of ICG to PS mAb but also showed that the peaks fall within 680 to 980 nm wavelength range of the MSOT system. Prior to injecting ICG-PS mAb and doxorubicin into the animal, a phantom study was conducted to investigate the optoacoustic effect of the compound. The resultant spectra of the compound from the reconstructed MSOT data matched closely with that of NIR absorption spectra. Due to the negligible absorbance of doxorubicin in the NIR region, it did not show any optoacoustic signal generation as compared to ICG as shown in Fig. 2(b).

Fig. 2

NIR spectrum of ICG-PS mAb and doxorubicin. (a) Absorption spectrum obtained from spectrometer. (b) NIR spectrum obtained from MSOT phantom study.

Pre- and posttreatment MSOT scans were conducted on control and treated tumor groups. The axial images showing the cross section of mice along the tumor are as shown in Fig. 3(a). The tumor site in this figure is highlighted with white circles. The first and second columns represent the multispectral optoacoustic images of control and treatment groups obtained on days 0, 1, 3, 5, 7, and 10, respectively. As shown in Fig. 3(a), the control tumor showed no sign of ICG signal in the tumor region on all scan days despite the presence of ICG-PS mAb in circulation as seen in the images. In the treatment group, there is a gradual increase in the accumulation of ICG-PS mAb over time in the tumor site. To estimate the percentage change in MSOT signal intensity due to the uptake of ICG-PS mAb, an ROI was drawn to include the entire volume of the tumor. The percentage change in intensity with respect to the baseline (day 0) was calculated. As shown in Fig. 3(b), the treatment group showed highest increase in intensity on day 5, while the control group showed very minimal change in signal intensity throughout the study. Tomographic slices showing the distribution and accumulation of ICG-PS mAb mostly in the periphery of the tumor on day 5 are presented in Fig. 3(c).

Fig. 3

(a) MSOT images showing an increased uptake of ICG-PS antibody in the treatment group tumor region indicating the cells have undergone apoptosis. The control group shows no uptake of ICG-PS antibody in the tumor region. (b) The percentage change in the MSOT signal intensity in the tumor region on different scan days as compared to the baseline for both control and treatment groups. (c) Tomographic slices of the scanned tumor area on day 5 (treatment group) with each slice separated by 1 mm.

The physical change in the tumor volume on all scan days was measured for both groups to confirm the treatment outcome. Percentage change in tumor volume was then calculated for each tumor using the following equation:

The MSOT scan results showed good correlation with the treatment results as the tumor in the treated mouse gradually reduced in size, while the tumor in the control mouse continued to grow as shown in Fig. 4(a). After 20 days, the tumor in the treated mouse completely regressed but the tumor in the control mouse continued to grow as shown in Fig. 4(b). The maximum reduction in tumor volume also occurred between 5 and 7 days as shown in Fig. 4(a) which correlates well with the MSOT signal.

Fig. 4

(a) Percentage change in control and treated tumor volume. (b) Photographic images of the control and treatment mouse showing the tumor on last scan day (day 10) and on day 20. The treatment mouse shows a completely regressed tumor while the tumor in the control mouse continues to grow.

4.

Discussion

The ICG-PS mAb conjugates were purified by a simple gel filtration technique, which separates the unconjugated ICG from the reaction solution. In this method, the ICG-PS mAb and the unconjugated PS mAb remain in the eluted solution. When injected, the unconjugated PS mAb will compete with the ICG-PS mAb for the PS target on the apoptotic cell surface, reducing specific binding sites for ICG-PS mAb. To overcome this, the compound has to be further purified to remove unconjugated PS mAb. High-pressure gel filtration purification is one such technique that can be used to separate the ICG-PS mAb conjugates from the unconjugated antibody, thus removing the competition from unbound PS mAb. This will increase specific binding of ICG-PS mAb which can in turn increase the optoacoustic signal from the targeted apoptotic tumor site during MSOT scan.

The commonly used positron emission tomography (PET) and single photon emission tomography (SPECT) employ short half-lived radioisotopes and ligands of short plasma half-life. Thus, it is not possible to capture occurrence of apoptosis that may occur 1 to 10 days postchemotherapy. The present antibody-based study takes advantage of long circulatory half-life of PS antibodies to localize in the apoptotic site over a long period of time. As shown in Fig. 4(b), the highest localization of ICG-PS mAb was observed on the fifth day and started to decrease gradually thereafter.

The spectral stability of ICG varies with different aqueous conditions. In tissues and cells, the NIR absorption peak is red shifted due to binding of ICG to cell/blood proteins,^15,16 which may lead to missing the absorption maximum of ICG if a single-wavelength photoacoustic tomography is used. Multiwavelength optoacoustic tomography can overcome this issue by covering the entire wavelength range in a single scan for imaging ICG that are bound to tissues and thus eliminating the need for multiple scans from a single-wavelength photoacoustic tomography.

Noninvasive in vivo imaging of apoptosis employing multiwavelength optoacoustic tomography was successfully used to capture the occurrence of apoptosis in TNBC tumors. This technology can be a valuable tool in monitoring therapy response and strategizing for an efficient treatment of TNBC. As demonstrated in this work, the conjugation of antibody to ICG is relatively easy and can be extended to other apoptosis indicating antibodies and proteins, such as phospatidylethanolamine¹⁷ and lactadherin.¹⁸ This technique has the potential to be used in clinical applications since ICG is the only optical contrast agent approved by the FDA and has been used clinically in many areas that include the assessment of lymphatic vascular structure and function,¹⁹ vascular repair,²⁰ measurement of liver function,^21,22 and retinal angiography.^23,24

5.

Conclusion

PS mAb conjugated with ICG is nonradioactive and has a longer plasma half-life. Hence, imaging of apoptosis can be performed over a period of time without the limitation of radioisotope half-life, while taking advantage of the high sensitivity and high resolution of MSOT. We have demonstrated that MSOT images show the occurrence of apoptosis in the treatment group with a steady increase in signal intensity in the tumor region as compared to the control group. Combined with suitable probes, MSOT is of potential in imaging of apoptosis and can be a good tool for evaluating the therapeutic effect of various chemotherapy drugs.