2.1.

Experimental Mice

Female ND4 Swiss Webster mice were purchased from Harlan Laboratory. CHY mouse embryos were purchased from the Medical Research Council (MRC Harwell, UK) on a mixed C3H/HeH and 101/H background. A colony was established in our animal facility at Washington University School of Medicine in St. Louis. Since one mutant allele is sufficient to confer loss of skin lymphatics,³¹ all control mice were littermates bearing no copies of the mutations (wild-type mice). Experiments were conducted in compliance with approval of the Animal Studies Committee at Washington University.

2.2.

Animal Preparation for Imaging

During the experiments, the animals’ temperature was kept at 37°C by a water-circulating heating pad. Mice were anesthetized by inhalation of isoflurane with a concentration of 1.0% to 2.0% and air flow rate of $0.8\mathrm{l}/\min$, and taped to a lab-made animal holder, which was mounted to the OR-PAM system. The mouse ear and tail were carefully positioned and fixed in place, thus limiting the potential motion artifacts that could be caused by the breathing during image acquisition. To image the lymphatic vasculature in the mouse ear or tail, intradermal injection of Evans blue (EB) dye (3% w/w) was carefully performed in the ear pinnae ($10\mu \mathrm{l}$) or the tail skin ($50\mu \mathrm{l}$), respectively, upstream of the imaged area, after the baseline images were acquired. Static areas of interest were imaged 10-min postinjection, whereas dynamic lymphatic flow was continuously imaged in a time span of 2 to 4 min.

2.3.

Imaging of the Lymphatic Vessels In Vivo

The traditional OR-PAM³² was used to image the LV in vivo [Fig. 1(a)]. Briefly, the nanosecond-pulsed laser beam from a dye laser (CBR-D, Sirah) pumped by an Nd:YLF laser (INNOSLAB, Edgewave) is tightly focused into the tissue surface by an optical objective (NA: 0.1 in air). The optical-acoustic beam combiner composed of a layer of silicone oil sandwiched by two prisms is used for the coaxial and confocal alignment of the optical illumination and acoustic detection. The resultant ultrasound waves are first focused by a concave acoustic lens (NA: 0.5 in water) ground into the bottom of the combiner, and then detected by a wideband ultrasonic transducer (V214-BB-RM, Olympus-NDT). A correction lens offsets the optical aberration of the prisms. Although each laser pulse generates a one-dimensional time-resolved signal, a 3-D image can be obtained by raster scanning the animal. The system has a lateral resolution of ca. $5\mu \mathrm{m}$ determined by the tight optical focusing, an axial resolution of ca. $15\mu \mathrm{m}$ determined by the transducer bandwidth, and a penetration depth of ca. 1 mm in highly scattering soft tissues. The cross-sectional imaging speed is ca. 1 Hz over a 3-mm scanning range.

Fig. 1

Optical-resolution PA microscopy (OR-PAM) systems. (a) Schematic of traditional OR-PAM system. AH, animal holder; BS, beam splitter; ConL, condenser lens; CorL, correction lens; M, mirror; MS, motorized stage; OL, objective lens; PH, pinhole; RAP, right angle prism; RhP, rhomboid prism; SOL, silicone oil layer; WT, water tank. (b) Schematic of fast-scanning MEMS-OR-PAM system. AC, aluminum coating; AL, acoustic lens. The MEMS-OR-PAM shares a similar light path outside the dashed box in (a). (c) Molar extinction spectra of oxy-hemoglobin (${\mathrm{HbO}}_2$), deoxy-hemoglobin (HbR), and Evans blue (EB). While 560 and 595 nm were used for two-wavelength measurement on traditional OR-PAM system, 532 nm was used for dynamic measurement on MEMS-OR-PAM. (d) The average PA signals of blood- and 0.3% EB-filled plastic tubes at 560 and 595 nm. (e) The identification of the blood-filled (shown in red) and EB-filled (shown in green) tubes by comparing the PA signals at two wavelengths.

To validate the two-wavelength identification of LV, the traditional OR-PAM was also used to image the blood- and EB-filled plastic tubes. Two tubes (diameter: $300\mu \mathrm{m}$) were filled with oxygenated whole bovine blood and 0.3% EB solution, respectively, and imaged by OR-PAM in water at 560 and 595 nm with an excitation pulse energy of 100 nJ.

2.4.

Lymphatic Flow Dynamics

Lymphatic flow dynamics were assessed with our recently developed wide-field fast-scanning OR-PAM with a water-immersible MEMS scanning mirror (MEMS-OR-PAM).³³ As shown in Fig. 1(b), a pulsed laser beam at 532 nm from a $\mathrm{Nd}:{\mathrm{YVO}}_4$ laser is focused by a condenser lens, then spatially filtered by a pinhole. The filtered laser beam is focused by an optical objective lens. A beam combiner provides acoustic-optical coaxial alignment. The focused laser beam and the generated PA waves are both directed by an MEMS scanning mirror plate. The PA waves are then focused by an acoustic lens and detected by an ultrasonic transducer. Driven by electromagnetic force, the whole MEMS scanning mirror can be submerged in water. Volumetric imaging is provided by fast angular scanning of the MEMS mirror along the $x$-axis and slow linear step-motor scanning of the sample along the $y$-axis. A $3\text{-}\mu \mathrm{m}$ diffraction-limited lateral resolution and $26\text{-}\mu \mathrm{m}$ axial resolution have been achieved, with a maximum penetration depth of 1.0 mm in soft tissue. The maximum in-focus scanning range is $\sim 3.0\mathrm{mm}$ along the $x$-axis, with a cross-sectional frame rate of 400 Hz. By simultaneously scanning both the excitation laser beam and resultant acoustic beam, MEMS-OR-PAM maintains confocal alignment and thus high detection sensitivity over a large field-of-view.

2.5.

Statistics

Data are expressed as the $\mathrm{mean}\pm \mathrm{SD}$, unless otherwise indicated. Statistical differences were assessed using a two-tailed student’s $t$-test. A $p$ value less than 0.05 is considered significant.

3.1.

Multiwavelength OR-PAM of Tube Phantoms

Two plastic tubes were imaged by OR-PAM at 560 and 595 nm [Fig. 1(c)]. The ${\mathrm{HbO}}_2$ concentration in the first tube was 2.3 mM, whereas the EB concentration in the other tube was 3.1 mM. The averaged PA signals’ amplitudes from the two tubes at two wavelengths are shown in Fig. 1(d), after normalization to the PA signal of ${\mathrm{HbO}}_2$ at 560 nm. The results show that the EB and ${\mathrm{HbO}}_2$ have comparable PA signals at 595 nm. The difference of the PA signals at two-wavelengths was used to distinguish the two tubes [Fig. 1(e)]. Although the ${\mathrm{HbO}}_2$ had a stronger PA signal at 560 nm than that at 595 nm (rendered in red), the EB had the opposite trend (rendered in green). The colors in Fig. 1(e) represent different absorbers, and the brightness represents the respective concentration. The averaged signal-to-noise ratio of EB at 595 nm is $\sim 40\mathrm{dB}$, suggesting a noise-equivalent detection sensitivity of $\sim 31\mu \mathrm{M}$.

3.2.

Multiwavelength OR-PAM of Lymphatic Vasculature at a Steady State

EB injected intradermally was uptaken by lymphatic capillaries and provided the imaging contrast for OR-PAM. By using multiwavelength measurements, we could spectrally separate the lymphatic vasculature from the blood vasculature in vivo. Before the injection of EB, the area of interest was imaged by OR-PAM at 595 nm, showing only the blood vessels [Fig. 2(a)]. Ten minutes after the injection, the same area of interest was imaged by the traditional OR-PAM at 560 and 595 nm [Fig. 2(b)]. Although hemoglobin in blood vessels has an absorption four times stronger at 560 nm than that at 595 nm, EB has an opposite absorption trend. Hence, blood vessels have a stronger signal amplitude at 560 nm, whereas LV filled with EB have a much stronger signal amplitude at 595 nm. Since blood vessels and LV are spatially separated, a direct comparison using a simple differential operation between the two images at 560 and 595 nm can be used to distinguish the LV from the blood vessels [Fig. 2(c)]. The colors in the figures represent different absorbers, and the brightness represents the respective concentration. The comparison of the PA images before and after the injection of EB further confirmed the identification of LV by using the two-wavelength method. In addition to the separation of the LV from blood vessels in a mouse ear, we also imaged the blood vessels and LV in a mouse tail [Figs. 2(d) and 2(f)]. The two forms of hemoglobin, deoxy-hemoglobin (HbR), and oxy-hemoglobin (${\mathrm{HbO}}_2$) can also be spectrally differentiated by the two-wavelength measurements, from which the oxygen saturation of hemoglobin (${\mathrm{sO}}_2$) can be computed [Fig. 2(e)].

Fig. 2

Traditional OR-PAM of LV in the skin of a mouse ear and tail. (a) PAM image of a mouse ear at 595 nm before the injection of EB, showing only blood vessels. (b) PAM images of the same mouse ear at 560 and 595 nm after the injection of 3% EB in the ear pinnae. (c) Spectral separation of the blood vessels (shown in red) and LV (shown in blue) in the ear skin. BV, blood vessels; LV, lymphatic vessels. (d) PAM image of a mouse tail vasculature at 560 nm before the EB injection. (e) PAM of the oxygen saturation of hemoglobin (${\mathrm{sO}}_2$) in the mouse tail, where the tail veins and arteries can be separated. (f) Spectral separation of blood vessels and LV in the tail after the EB injection.

3.3.

High-Speed MEMS-OR-PAM of Lymphatic Flow Dynamics

We have then assessed the lymphatic flow dynamics by fast-scanning MEMS-OR-PAM. Although the mouse tail blood vasculature can be easily imaged [Figs. 3(a) and 3(b)], the visualization of the LV requires the intradermal injection of EB dye in the tip of the mouse tail. The dye is thus uptaken by lymphatic capillaries in the mouse tail and subsequently propelled into the collecting LV. Taking advantage of the volumetric imaging capability of MEMS-OR-PAM, we studied the lymphatic flow dynamics for both superficial LV ($<300\text{-}\mu \mathrm{m}$ deep) and deeper collecting LV ($>1\text{-}\mathrm{mm}$ deep). The region of interest was imaged with a volumetric frame rate of 2 Hz and a scanning step size of $5\mu \mathrm{m}$, which was fast enough to assess the flow dynamics of the LV. In LV that were located closer to the epidermis, the overall signal intensity in the field-of-view was averaged to show the flow dynamics [Figs. 3(c) and 3(d), and 4]. In deeper LV, the temporal profile of the signal intensity was used to separate the LV from the blood vessels, since the blood vessel signal was maintained during the process [Figs. 3(e) and 3(f), and 5]. Furthermore, the lymph velocity and direction can be potentially quantified by comparing the temporal profiles of the PA signals from different regions along the LV [Fig. 3(g)].

Fig. 3

MEMS-OR-PAM of lymphatic flow dynamics in a mouse tail. (a) MEMS-OR-PAM of mouse tail blood vasculature. (b) A cross-sectional image of the mouse tail, showing the deep tail vein. (c) A snapshot of the dynamics of the superficial lymphatic capillaries in the mouse tail after the EB injection in the tail skin. (d) Time course of the relative PA amplitude change, where the arrows show the contractile rate of the lymphangions. (e) Snapshots of the dynamics of the collecting lymphatic vessel in a mouse tail after EB injection. (f) Blood vessels (shown in red) and collecting LV (shown in blue) were separated based on the temporal change in their signal strength. (g) Time courses of the relative PA signal changes at three regions of interest (ROI1-3) on the collecting lymphatic vessel and one region of interest (ROI4) on the blood vessel. The speed and direction of lymphatic flow can be extracted from time delays between ROI1 and ROI3.

Fig. 4

MEMS-OR-PAM of the superficial lymph flow dynamics in a mouse tail after Evans blue (EB) injection. The averaged PA signal amplitude is plotted on the right (MOV, 2.96 MB) [URL: http://dx.doi.org/10.1117/1.JBO.19.11.116009.1].

Fig. 5

MEMS-OR-PAM of the flow dynamics in the collecting LV in a mouse tail. The EB was injected at the tip of the tail 50 s after the imaging started (MOV, 3.04 MB) [URL: http://dx.doi.org/10.1117/1.JBO.19.11.116009.2].

We are aware that other physiological factors may affect the EB dye concentration during the flow measurement. For instance, the time window used for the temporal correlation will certainly affect the accuracy of the measurement. A longer time window measurement could potentially improve this accuracy, but the physiological conditions of the LV may change, as does the flow speed. Therefore, we view that an optimal time window should be chosen between the measurement accuracy and the temporal resolution of the measurement.

3.4.

High-Speed MEMS-OR-PAM of Mouse Model with Defected Lymphatic System

In addition to the nonpathological conditions where the lymphatic system was intact (wild-type mice), we also studied the lymphatic dynamics of a genetically engineered mouse model that selectively lacks lymphatic capillaries in the skin.^31,34 This mouse mode has a point mutation in the tyrosine kinase domain of the VEGF-C receptor VEGFR3 that controls lymphatic vessel development.^35,36 Here, the so-called “CHY mice” were imaged by MEMS-OR-PAM and compared with wild-type littermate controls [Figs. 6(a) and 6(b)]. The region of interest, $\sim 10\text{-}\mathrm{mm}$ downstream from the injection site, was imaged with a volumetric frame rate of 2 Hz and a scanning step size of $5\mu \mathrm{m}$, as shown in Figs. 7 and 8. The EB was injected 30 s after the imaging was started. Although we can clearly see the LV that quickly showed up after the EB injection in wild-type mice, we did not observe lymphatic flow in the CHY mice. For each mouse model, the signal intensity before and after the EB injection was quantified [Fig. 6(c)]. The results show that while the wild-type mice (Fig. 7) had a significant signal increase by 31% after the dye injection, the CHY mice (Fig. 8) did not show a noticeable signal increase, thus reflecting the lymphatic network defect in the skin of CHY mice.

Fig. 6

Lymphatic flow dynamics in nonpathological and pathological conditions. Snapshot MEMS-OR-PAM images of lymphatic flow dynamics of (a) wild-type and (b) CHY mice before and after EB injection. (c) Normalized PA signal. Bars: $\mathrm{means}\pm \mathrm{SD}$. LV, lymphatic vessels. $*p\le 0.05$.

Fig. 7

MEMS-OR-PAM of the lymph flow dynamics in a wild-type mouse tail. The EB was injected at the tip of the tail 30 s after the imaging started (MOV, 1.55 MB) [URL: http://dx.doi.org/10.1117/1.JBO.19.11.116009.3].

Fig. 8

MEMS-OR-PAM of the lymph flow dynamics in a CHY mouse tail. The EB was injected at the tip of the tail 30 s after the imaging started (MOV, 2.40 MB) [URL: http://dx.doi.org/10.1117/1.JBO.19.11.116009.4].