2.1.

DP-OCT Instrumentation

The fiber-based dual-channel DP-OCT system used in this study has been described previously.²⁸ Briefly, partially polarized light from an optical semiconductor amplifier (AFC Technologies, central wavelength ${\lambda }_0=1.31\mu \mathrm{m}$ , $\mathrm{FWHM}=60\mathrm{nm}$ ) is polarized and coupled into fast and slow axes of a polarization-maintaining (PM) fiber. Light is split by a $2\times 2$ PM coupler into reference and sample paths. The phase of light in the reference path is modulated by a $\mathrm{Li}\mathrm{Nb}{\mathrm{O}}_3$ waveguide phase modulator while group delay is varied by a rapid scanning optical delay line. Additionally, the rapid scanning optical delay line (RSOD) was used to balance dispersion imbalance between reference and sample paths introduced by the $\mathrm{Li}\mathrm{Nb}{\mathrm{O}}_3$ waveguide phase modulator and birefringent calcite wedges. Polarization modes of light in the sample path undergo a user-selectable differential delay when propagating through a pair of birefringent calcite wedges with orthogonal fast axes. The differential delay between polarization modes may be varied by adjusting the thickness of the calcite wedges. The differential delay is set to provide equal pathlengths of light reflecting from the bottom surface of a $1\mathrm{mm}$ -thick glass window providing a reference air-glass interface and the atherosclerotic tissue specimen. A rubber o-ring positioned underneath the $1\text{-}\mathrm{mm}$ -thick glass window provided an adjustable distance between the reference air-glass interface and the atherosclerotic tissue specimen. Unpolarized thick surface of light was focused on the specimen with a 0.2 numerical aperture so that the Rayleigh range included both the specimen and reference air-glass interface. Light returning from the sample and reference path recombine and interfere in the PM coupler. A Wollaston prism separates interfering light in fast and slow axes of the PM fiber into channel 1 (ch1) and channel 2 (ch2).

The dual-channel Michelson interferometer was used to measure differential phase between light backscattered from the reference air-glass interface and atherosclerotic tissue specimen surface when applying a sinusoidal focused magnetic flux density. Detected fringe signals may be written

The two signals recorded from channels 1 and 2 by the DP-OCT system were used to measure magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ due to movement of magnetic nanoparticles taken up by macrophages. Forward and reverse bandpass filtering was used for zero-phase distortion and to eliminate noise from interferometric fringe data. After data processing, a second-order low-pass Chebyshev type II digital filter was utilized to eliminate noise from the optical path length change signal with a stopband ripple $20\mathrm{dB}$ down from the peak passband and with cutoff frequency of $30\mathrm{Hz}$ . After processing, the root-mean-square level of phase noise of signals recorded from arterial test specimens was reduced from $15\mathrm{nm}\text{to}<5\mathrm{nm}$ . Noise in unfiltered signals was predominantly at $60\mathrm{Hz}$ and believed to originate in mechanical vibration.

2.2.

Magnetic Flux Density Generator

A finite element method (FEM) was used to design the magnetic flux density generator and evaluate spatial variation of field strength. The magnetic field generator consists of a solenoid (Ledex 6EC, Saia-Burgess Inc., USA), a function generator (HP 33120A, Hewlett Packard Inc., USA), a current amplifier, and power supply. FEM calculations (Maxwell SV, Ansoft Inc., USA) and teslameter (Magnetometer, AlphaLab Inc., USA) measurement indicated that maximum magnetic flux density at a distance of $1.5\mathrm{mm}$ from the tip of the iron core was $1.8\text{Tesla}$ . The FEM simulation demonstrated that an iron core positioned along the centerline of the solenoid substantially increased $(10\times )$ magnetic flux density at the target specimen compared to without the core. Magnetic flux density distributions from the FEM simulation showed the maximal and principal direction of the magnetic field strength was along the $z$ direction. The conical iron core provided focusing and substantially increased the magnetic flux density.

2.3.

Iron Nanoparticles

Monocrystalline iron oxide nanoparticles (MIONs) composed of a monocrystalline iron core with a dextran coating was used as a magnetic label and were taken up by inflammatory macrophages in atherosclerotic lesions. The core diameter³⁰ of these particles is $\sim 5\mathrm{nm}$ and with the dextran coating is $<40\mathrm{nm}$ . The reduced diameter of these particles reduces competitive uptake by the reticulo-endothelial system, increases blood circulation half-life, and allows for increased phagocytosis by atherosclerotic macrophages.²⁹

2.4.

Animal Preparation and Specimen Preparation

All experimental procedures were performed in accordance with protocols approved by the University of Texas Institutional Animal Care and Use Committee. Four hyperlipidemic Watanabe rabbits and two double-balloon injured fat-fed New Zealand white rabbits were included in the study. The Watanabe is a spontaneous mutation that results in atherosclerotic disease and does not require fat feeding for plaque development. Atherosclerotic plaques may be induced in New Zealand white rabbits by double-balloon injury and fat feeding, as previously reported by Feldman and coworkers.³¹ Two Watanabe and one New Zealand white rabbits were injected with MIONs at a dose of $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ body weight. This iron dosage is $\sim 200$ times the recommended FDA dosage for humans; however, this dose is well tolerated in rabbits. The other three rabbits (two Watanabe and one New Zealand white) were injected with saline and served as controls.

Each aortic specimen was evaluated ex vivo using DP-OCT measurement of tissue surface displacement in response to magnetomotive excitation. After three days following MION injection, an intravascular bolus of heparin was injected to prevent clot formation and animals were then euthanized with a dose of Phenobarbital. Aortic specimens were removed, flushed with saline, and cut into $2\times 3\text{-}{\mathrm{mm}}^2$ -sized specimens for DP-OCT measurement.

2.5.

Histology Analysis

After completion of DP-OCT measurements, the rabbit aorta specimens were fixed in 4% paraformaldehyde for $24\mathrm{h}$ , and embedded in paraffin. Arterial cross sections were cut in to $5\text{-}\mu \mathrm{m}$ -thick sections and stained with Prussian blue to identify iron deposition and RAM-11 for macrophage cell identification in the atherosclerotic lesions. In aortic specimens taken from New Zealand white rabbits, in addition to Prussian blue and RAM-11, alpha smooth muscle actin was used to identify smooth muscle cells. Positive staining for alpha smooth muscle actin is a mirror image of RAM-11 and Prussian blue stains and thus serves as a control.

3.1.

Watanabe Atherosclerotic Aortic Tissue Surface Displacement

In Watanabe atherosclerotic rabbit tissue, measurements of peak-to-peak magnetomechanical tissue surface displacement $\left(\Delta z_s\right)$ are illustrated (Fig. 1 ) in ex vivo aorta specimens following intravenous MION (0.2 mmol Fe/kg body weight) or saline injection. While applying a $2\text{-}\mathrm{Hz}$ , $0.25\text{-}\text{Tesla}$ input magnetic flux density to aorta specimens, $\Delta z_s$ is $138\mathrm{nm}$ , while no tissue surface displacement was observed in saline control aorta specimens.

Fig. 1

Magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ in Watanabe aorta specimens in response to focused magnetic flux density excitation. Magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ in $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ of MION-laden aorta specimens ( $2\mathrm{Hz}$ , $B_{\max }=0.47\text{Tesla}$ ) (top), and magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ in $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ of MION-laden aorta specimens by applying a swept frequency excitation $(1–10\mathrm{Hz})$ over a $1\text{-}\mathrm{s}$ time period $(B_{\max }=1.70\text{Tesla})$ (bottom).

Peak-to-peak magnetomechanical tissue surface displacement $\left(\Delta z_s\right)$ in rabbit aorta specimens (0.2 mmol Fe/kg body weight) was measured in response to $1\mathrm{s}$ application of a swept frequency $(1–10\mathrm{Hz})$ high-intensity focused magnetic flux density (Fig. 1, bottom). A swept frequency magnetic flux density was applied to investigate if the aorta specimen responded to a variable excitation of $1–10\mathrm{Hz}$ . With maximum magnetic flux density of $1.7\text{Tesla}$ , $\Delta z_s$ is $677\mathrm{nm}$ . Frequency response of $\Delta z_s\left(t\right)$ in MION aorta specimens was twice the externally applied modulation frequency (Fig. 1).

Peak-to-peak magnetomechanical tissue surface displacement $\left(\Delta z_s\right)$ in MION atherosclerotic Watanabe aorta ( $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ body weight) with frequency of the sinusoidal magnetic flux density increasing from $2\text{to}5\mathrm{Hz}$ in $1\text{-}\mathrm{Hz}$ increments is (a) 123, (b) 156, (c) 174, and (d) $177\mathrm{nm}$ , respectively (Fig. 2 ). Increasing the modulation frequency of the magnetic flux density from $2\text{to}4\mathrm{Hz}$ , increased $\Delta z_s$ by $30\mathrm{nm}∕\mathrm{Hz}$ . Further increases in frequency of the sinusoidal magnetic flux density did not substantially increase magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ .

Fig. 2

Magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ in MION-laden Watanabe aorta specimens with input frequency ranging from $2\text{to}5\mathrm{Hz}$ in $1\text{-}\mathrm{Hz}$ increments with MION dose ( $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ body weight). $B_{\max }=0.47\text{Tesla}$ .

Quantitative analysis of the experiments in Watanabe MION aorta specimens using box-and-whisker plots of peak-to-peak magnetomechanical tissue surface displacement $\left(\Delta z_s\right)$ versus applied sinusoidal $\left(2\mathrm{Hz}\right)$ input voltage $(2–10{\mathrm{V}}_{\mathrm{pp}})$ with (a) $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ MION dose $(n=60)$ and (b) control specimens $(n=20)$ are shown in Fig. 3 . Corresponding magnetic flux density for sinusoidal input voltages was 0.25, 0.47, 0.67, 0.87, and $1.10\text{Tesla}$ , respectively. Compared to MION aorta specimens, magnetomechanical tissue surface displacement of Watanabe control specimens did not vary with increasing magnetic field; with maximum tissue surface displacement of $<20\mathrm{nm}$ corresponding to the background noise level.

Fig. 3

Quantitative analysis of MION Watanabe aorta specimens using box-and-whisker plot of peak-to-peak magnetomechanical tissue surface displacement $\left(\Delta z_s\right)$ versus sinusoidal $\left(2\mathrm{Hz}\right)$ input voltage with $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ MION dose $(n=60)$ (left) and control specimens $(n=20)$ (right). Edges of rectangular box are the 25th percentile, and 75th percentile, respectively. The central small box indicates the mean, and whiskers show minimum and maximum values.

3.2.

New Zealand White Aortic Tissue Surface Displacement

In double-balloon injured fat-fed New Zealand white rabbits, measurements of peak-to-peak magnetomechanical tissue surface displacement $\left(\Delta z_s\right)$ are illustrated (Fig. 4 ) in ex vivo aorta specimens following intravenous MIONs ( $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ body weight). While applying a $2\text{-}\mathrm{Hz}$ , $0.25\text{-}\text{Tesla}$ input magnetic flux density to aorta specimens, $\Delta z_s$ is $75\mathrm{nm}$ and decreases to $51\mathrm{nm}$ with filtering.

Fig. 4

Magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ in New Zealand white rabbit aortic specimens in response to focused magnetic flux density excitation. Input magnetic field (top); magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ in $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ of MION-laden aorta specimens ( $2\mathrm{Hz}$ , $B_{\max }=0.25\text{Tesla}$ ) (middle), and filtered magnetomechanical tissue surface displacement $\left[\Delta z_s\left(t\right)\right]$ (bottom).

3.3.

Histology

Histologic images of aortic tissue sections from a Watanabe heritable hyperlipidemic (WHHL) rabbit injected with MIONs at a dose of $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ body weight and interrogated with DP-OCT ( $\Delta z_s=80\mathrm{nm}$ with $2\mathrm{Hz}$ , $B_{\max }=0.25$ Tesla) are shown in Fig. 5 . Prussian blue stain (a) demonstrates MION engulfed by superficial macrophages (arrows) which also stains positive with RAM-11 [(b), arrow], confirming these cells are macrophages in intimal hyperplasia.

Fig. 5

Histologic images of atherosclerotic Watanabe MION-laden aorta tissue sections where magnetomechanical tissue surface displacement was measured with DP-OCT ( $\Delta z_s=80\mathrm{nm}$ with $2\mathrm{Hz}$ , $B_{\max }=0.25\text{Tesla}$ ). Prussian Blue (a) and RAM-11 stains (b) at a MION dose ( $0.2\mathrm{mmol}\mathrm{Fe}∕\mathrm{kg}$ body weight) are shown. Adjacent histological sections indicated in (a) and (b) are $5\text{-}\mu \mathrm{m}$ apart and demonstrate presence of MIONs in macrophages (arrows).

Histologic images of aortic tissue sections from a double-balloon injured fat-fed New Zealand white rabbit injected with MIONs at a dose of 0.2 mmol Fe/kg body weight and interrogated with DP-OCT ( $\Delta z_s=75\mathrm{nm}$ with $2\mathrm{Hz}$ , $B_{\max }=0.25\text{Tesla}$ ) are shown in Fig. 6 . The section stained with Prussian blue (left) demonstrates iron oxide present in the deep intima. The section stained with RAM-11 confirms that MIONs are in the macrophages. The section stained with $\alpha$ -smooth muscle (SM) actin demonstrates SM cells in adjacent areas and represents a mirror image of the sections stained with Prussian blue and RAM-11.

Fig. 6

Histologic images of double-balloon injured fat-fed abdominal aorta from a New Zealand white rabbit. Left: section stained with Prussian blue for iron; Middle: section stained with RAM-11 for macrophages; Right: section stained for $\alpha$ SM actin for SM cells ( $i=\text{intima}$ ; $m=\text{media}$ ).