(479a) DNA Aptamer-Decorated Gold Nanoparticles for In Vivo Detection of Human Matrix Metalloproteinase-9 Via Molecular Photoacoustic Imaging

Matrix metalloproteinase-9 (MMP-9) plays major roles in extracellular matrix (ECM) remodeling and membrane protein cleavage, suggesting a high correlation with cancer cell invasion and tumor metastasis. However, because of the structural and functional similarity within the MMP family of proteins, it has been quite challenging to detect/target MMP-9 with high selectivity and sensitivity, especially in complex in vivo environments. In this study, we are introducing a contrast agent based on a DNA aptamer that can selectively target MMP-9 with picomolar (pM) detection range. By inducing programmed hybridization/dehybridization of DNA aptamers on the surface of plasmonic gold nanoparticles, we enable the controlled aggregation of nanoparticles, and therefore, the detection of MMP-9 using photoacoustic (PA) imaging with high specificity and sensitivity.

To develop the sensor (Fig. A), the surface of 15 nm-diameter gold nanospheres (AuNSs) were modified with two different sets of DNA strands, termed AuNS-1 and AuNS-2. For AuNS-1, double strands of MMP-9 aptamer and its partial –two base mismatches– complementary sequence (comp-1) were grafted on the surface of AuNSs. To create AuNS-2, single strands and full-match complementary sequences of comp-1 (comp-2) were decorated on the particle surface. In the absence of MMP-9, the binding of comp-2 to comp-1 is blocked by the pre-bound MMP-9 aptamers, maintaining the surface plasmon resonance (SPR) of the mixture of AuNS-1 and AuNS‑2 at 532 nm wavelength. However, in the presence of MMP-9, the MMP-9 aptamer DNA strand preferentially binds to the MMP-9 protein instead of comp-1 and forms an MMP-9–aptamer complex. The release of the MMP-9 aptamer sequence from AuNS-1 then enables the binding of comp-2 to comp-1 to allow coupling of AuNS-2 and AuNS-1. The resulting aggregation of AuNS‑1 and AuNS-2 in the presence of MMP-9 causes an optical absorption shift for detection via ultrasound (US)-guided spectroscopic PA (sPA) imaging in the ideal tissue optical window (NIR-I optical window), above 700 nm wavelength.

Construct behavior was first assessed qualitatively by observing colormetric changes in the presence of MMP-9. In comparison to buffer controls, only a solution containing both the MMP sensor and MMP-9 changed color, indicating plasmon coupling induced by the MMP-9–aptamer complex formation and subsequent binding of comp-1 and comp-2 to aggregate AuNS-1 and AuNS-2 (Fig. B). UV-Vis-NIR spectrophotometry further confirmed the red-shift of the MMP sensor only in the case of incubation with the target protein, MMP-9 (Fig. C). To evaluate the potential of our MMP sensor in sPA imaging, we embedded the MMP sensor with and without MMP-9 in a dome-shaped 8% gelatin phantom. The phantom was imaged using the combined US/PA imaging system (Vevo 2100/LAZR, Visualsonics, Inc.). The PA signal of our MMP sensor pre-treated with MMP-9 was very strong and distinct at 750 nm, whereas the MMP sensor without MMP-9 did not exhibit apparent PA signal at the same wavelength range (Fig. D).

Finally, we assessed MMP-9 detection with our sensor in vivo using US/sPA imaging and photoacoustic tomography (TriTOM, Photosound Technologies, Inc.) in a xenograft mouse model of breast cancer (MDA-MB-231). MDA-MB-231 cells were stimulated to secrete MMP-9 by the pre-treatment of phorbol 12-myristate 13-acetate (PMA), followed by the injection of MMP sensor 24 h before imaging. To visualize the volumetric distribution and detectability of the MMP sensors, the photoacoustic tomography (PAT) system was used to generate 3-dimensional maps (Fig. E) of the skin and surface vasculature (532 nm, yellow) and the MMP sensor (710 nm, red). The aggregated MMP sensor was distinguished from endogenous absorbers using PAT, with the bulk of the 710 nm signal distributed on the periphery of the tumor (blue dashed circle, Fig. E). The results illustrate the viability of the developed MMP sensor as a contrast agent in a mouse model of cancer by exhibiting PA signal in the NIR-I optical window.

In summary, our results demonstrate a proof-of-concept study for developing a PA contrast agent as a biosensor that detects MMP-9 by exploiting plasmonic AuNSs and DNA aptamers. The use of DNA aptamers enabled sensitive and selective detection of MMP-9, while the plasmon coupling between AuNSs by programmable DNA hybridization resulted in the selective PA signal amplification in the NIR-I optical window. Furthermore, oligomeric DNA sequences’ versatility and customizability will enable the design of other AuNS-based sensors for the detection and capture of biomolecules by utilizing the associated aptamer sequence. By incorporating multiple aptamers that target different biomolecules onto the nanoparticles, our approach provides unlimited opportunity to develop a contrast agent for multiplex imaging.