In Vivo Monitoring of Fluorescent Nanosphere Delivery in

Division of Medical Engineering Research and Center for Nanomedicine Research, National Health Research Institutes, Zhunan 350, Taiwan, Department of ...
0 downloads 0 Views 420KB Size
Download PDF
Anal. Chem. 2005, 77, 1125-1131

In Vivo Monitoring of Fluorescent Nanosphere Delivery in Anesthetized Rats Using an Implantable Fiber-Optic Microprobe Leu-Wei Lo,† Pi-Ju Tsai,‡,§ Sam Hong-Yi Huang,† Wen-Ying Chen,‡ Yun-Tse Wang,‡ Chia-Hua Chang,† and Chung-Shi Yang*,|,⊥

Division of Medical Engineering Research and Center for Nanomedicine Research, National Health Research Institutes, Zhunan 350, Taiwan, Department of Education and Research, Taichung Veterans General Hospital, Taichung 400, Taiwan, Department of Applied Chemistry, Providence University, Shalu 433, Taiwan, and Department of Applied Chemistry, National Chi-Nan University, Puli 545, Taiwan

An implantable needle-type fiber-optic microprobe was constructed to monitor in vivo fluorescent substances in anesthetized rats. This fiber-optic microprobe was composed of coaxial optical fibers that were catheterized using a thin-wall tube of stainless steel (o.d. ∼400 µm; i.d. ∼300 µm). When the fiber-optic microprobe was placed in solutions containing various concentrations of fluorescent nanospheres (20 nm), either in the presence or in the absence of 10% Lipofundin acting as an optical phantom, we observed nanosphere concentration-dependent responses of the fluorescence intensity. The microprobe was then implanted into the livers and brains of anesthetized rats to monitor the in situ extravasation of preadministered fluorescent nanospheres from vasculature following the hepatic and cerebral ischemic insults. Both types of ischemic insults showed immediate increases in fluorescent intensities when 20-nm fluorescent nanosphere were administered, but neither ischemic insult induces such an increase when we administered 1000-nm fluorescent nanospheres. Additional experiments can be performed to further narrow the size range of the increase in blood vessel permeability following ischemic insult; such “size” information may be valuable when formulating drugs for optimal local delivery. Although a wide variety of fluorescent substances are used intensively for in vitro biological studies, the in vivo and in situ monitoring of these substances is studied much less often, probably because of difficulties in the efficient assembly of miniaturized fiber optics to detect the relatively weak fluorescence signal arising within such a turbid medium as tissue. To our knowledge, the use of our implantable fiber-optic microprobe is the first minimally invasive technique capable of investigating the “size * To whom the correspondence should be addressed. Tel: +886-49-2910960, ext 4891. Fax: +886-49-2917114. E-mail: [email protected]. † Division of Medical Engineering Research, National Health Research Institutes. ‡ Taichung Veterans General Hospital. § Providence University. | Center for Nanomedicine Research, National Health Research Institutes. ⊥ National Chi-Nan University. 10.1021/ac0487552 CCC: $30.25 Published on Web 01/13/2005

© 2005 American Chemical Society

window” of vascular permeability for the in vivo delivery of nanospheres in both ischemic livers and brains. Fluorescent dyes and markers play important roles in biological studies. Fluorescent nanospheres, including quantum dots, are used intensively for in vitro investigations such as tissue and cell staining, imaging, and tracing.1-4 For in vivo and continuous monitoring of these substances, the type II quantum dots were engineered, with a maximal absorption cross section and fluorescence emission in the near-infrared (NIR) region, to allow sentinel lymph nodes 1 cm deep to be imaged.5 The NIR imaging, though noninvasive, is not suitable to real-time monitor the local distribution and delivery dynamics of fluorescent nanospheres at a very specific depth in tissue. Magnetic resonance imaging (MRI) has been reported to analyze nanoparticle efflux and localization with superparamagnetic iron oxide (USPIO).6 The coregistration of positron emission tomography and MRI can be combined to determine changes in blood-brain barrier (BBB) transport characteristics.7,8 Nevertheless, these techniques are expensive and too slow to detect changes in nanoparticle delivery in response to dynamic variations in physiological conditions. The analytical capability provided by continuous monitoring of in vivo and in situ fluorescent nanospheres, particularly those prepared in a variety of sizes, is crucial for carrier formulation and therapeutic efficacy of drug delivery to specific tissues or organs. In vivo microdialysis perfusion is one of the most powerful techniques for analyzing substances in extracelluar samples. (1) Chan, W. CW.; Maxwell, D. J.; Gao, X.; Bailey, R. E.; Han, M.; Nie, S. Curr. Opin. Biotechnol. 2002, 13, 40-46. (2) Jaiswal, J. K.; Mattoussi, H.; Mauro, J. M.; Simon, S. M. Nat. Biotechnol. 2003, 21, 47-51. (3) West, J. L.; Halas, N. J. Annu. Rev. Biomed. Eng. 2003, 5, 285-292. (4) Wu, X.; Liu, H.; Liu, J.; Haley, K. N.; Treadway, J. A.; Larson, J. P.; Ge, N.; Peale, F.; Bruchez, M. P. Nat. Biotechnol. 2003, 21, 41-46. (5) Kim, S.; Lim, Y. T.; Soltesz, E. G.; Grand, A. M. D.; Lee, J.; Nakayama, A.; Parker, J. A.; Mihaljevic, T.; Laurence, R. G.; Dor, D. M.; Cohn, L. H.; Bawendi, M. G.; Frangioni, J. V. Nat. Biotechnol. 2004, 22, 93-97. (6) Muldoon, L. L.; Varallyay, P.; Kraemer, D. F.; Kiwic, G.; Pinkston, K.; WalkerRosenfeld, S. L.; Neuewlt, E. A. Neuropathol. Appl. Neurobiol. 2004, 30, 70-79. (7) Lin, K. P.; Huang, S. C.; Baxter, L.; Phelps, M. E. IEEE Trans. Nucl. Sci. 1994, 41, 2850-2855. (8) Yang, C. S.; Chang, C. H.; Tsai, P. J.; Chen, W. Y.; Lo, L. W. Anal. Chem. 2004, 76, 4465-4471.

Analytical Chemistry, Vol. 77, No. 4, February 15, 2005 1125

Microdialysis perfusion has been used, in conjunction with analysis using an ex vivo fluorometer, to sample extravasated fluorescent nanospheres in an anesthetized rat brain.8 Cerebral ischemia and reperfusion induced transient accumulations of extracellular nanospheres in the brain. Nevertheless, in addition, it is an ex vivo and not situ monitoring method; continuous sampling of the biological fluids may wash out the accumulating nanospheres, which may affect the accuracy when estimating the concentrations of these nanospheres. The sampling of extracellular fluids per se may also induce undesired biological complications on the locus. Furthermore, the selective range of sampling size constrains the in vivo microdialysis from being a “one-size-fitsall” analytical technique. To surpass the limitations inherent in microdialysis, we have used miniaturized fiber optics to investigate the delivery of fluorescent nanospheres in vivo. Optical fibers can be fabricated to function as biosensors for the analysis of different biologically active substances; for example, enzyme-modified optical fiber tips can be used for the analyses of cellular nitric oxide9 and glutamate in a polycarbonate membrane hole that mimics a single biological cell.10 A fluorescence affinity hollow fiber sensor has been developed for continuous monitoring of transdermal glucose.11 Luminescence-based imaging-fiber oxygen sensors have been utilized for the in situ measurement of oxygen consumption in an isolated perfused mouse heart.12 An implantable fiber-optic sensor has been designed for the continuous monitoring of the concentrations of drugs, such as adriamycin, in the blood of the carotid body.13 In combination with organically modified silicate (ormosil) nanoparticles, a real-time ratiometric method has been validated for measuring dissolved oxygen inside live cells.14,15 To further apply fiber optics to in situ and in vivo measurements of intrinsic or extrinsic fluorescent compounds located at depth in tissues, not only must the miniaturization provide minimal invasion but we must also take into account the optical complications that arise from the turbid tissue, such as tissue absorption and scattering that may impose significant interference to measured signals, prior to the design of the fiber-optic sensors. Accordingly, we designed and constructed an implantable fiberoptic microprobe that consisted of a central gold-coated optical fiber for excitation delivery and the surrounding coaxial optical fiber bundle for collecting the fluorescence emission. The multifiber configuration can avoid the Fresnel reflection that otherwise occurs on the single fiber as a result of the different reflective index at the fiber-tissue interface.16 The signal-to-noise ratio of the fiber-optic microprobe was further validated by using a tissuelike optical phantom, a 10% Lipofundin suspension. Similar to Intralipid, Lipofundin is an intravenous nutrient comprising an emulsion of phospholipid micelles and water.17 Because it is turbid (9) Barker, S. L. R.; Zhao, Y.; Marletta, M. A.; Kopelman, R. Anal. Chem. 1999, 71, 2071-2075. (10) Cordek, J.; Wang, X.; Tan, W. Anal. Chem. 1999, 71, 1529-1533. (11) Ballerstadt, R.; Schultz, J. S. Anal. Chem. 2000, 72, 4185-4192. (12) Zhao, Y.; Richman, A.; Storey, C.; Radford, N. B.; Pantano, P. Anal. Chem. 1999, 71, 3887-3893. (13) Lu, W.-X.; Chen, J. Anal. Chem. 2003, 75, 1458-1462. (14) Xu, H.; Aylott, J. W.; Kopelman, R.; Miller, T. J.; Philbert, M. A. Anal. Chem. 2001, 73, 4124-4133. (15) Koo, Y.-E. L.; Cao, Y.; Kopelman, R.; Koo, S. M.; Brasuel, M.; Philbert, M. A. Anal. Chem. 2004, 76, 2498-2505. (16) Hecht, E. Optics, 3rd ed.; Addison-Wesley Longman: Reading, MA, 1998; pp 345-346.

1126

Analytical Chemistry, Vol. 77, No. 4, February 15, 2005

and has no strong absorption in the visible region of the electromagnetic spectrum, we used the 10% Lipofundin suspension as a tissue-simulating phantom to understand the characteristics of the designed microprobe in collecting fluorescence signal against the high light scattering property of tissue. Vascular permeability is a determinant parameter for implementing successful drug delivery and controlled release. The fact that many chemotherapeutic agents are very effective in vitro but less so in vivo is attributed mainly to the inability of the agent to cross the vascular wall effectively. Vascular permeability is increased following many types of physiological disorders, including inflammation and cancer. Collecting in vivo information regarding vascular permeability following these injuries or disorders is important because it provides information on the sizes of nanoparticle-containing therapeutic drugs available for extravasation from the vasculature to specific sites within the tissue or organ. There have been a number of studies in which tumorspecific nanoparticle delivery has been investigated using rodent dorsal skin window chambers to visualize the tumor’s vasculature.18,19 Additionally, the vascular permeability increase is particularly important in the brain because it provides an in vivo evaluation of damage to the BBB. The BBB controls the composition of the extracellular fluid in the central nervous system and buffers against changes in the systemic circulation; in addition, it accounts for the low extracellular concentrations of amino acids and proteins in the brain relative to the blood and also restricts access to the immune system and systemically administered drugs.20 In earlier studies, we demonstrated that nanoparticlebased in vivo investigations can be used to analyze BBB permeability following ischemia and reperfusion.8 In the present study, we used the aforementioned fiber-optic microprobe, in combination with the systemic administration of 20-nm and 1-µm fluorescent nanospheres, to further study the vascular permeability during in vivo hepatic and cerebral ischemia. To our knowledge, this implantable fiber-optic microprobe provides itself as the first minimally invasive technique capable of investigating the “size window” of vascular permeability for in vivo nanosphere delivery to different tissues or organs, such as the liver and brain, under ischemic insult. EXPERIMENTAL SECTION Construction of the Fiber-Optic Microprobe. The fiber-optic microprobe (Figure 1A) was composed of a central optical fiber having a 100-µm core diameter and 20-µm cladding, coated with a 30-µm layer of gold (Fiberguide Industries, Inc., Stirling, NJ). This central fiber was surrounded by 10 fibers with 50-µm cores (Fiberguide Industries, Inc., Stirling, NJ) (Figure 1B). The central fiber carries light from a 488-nm argon laser to the tissue while the surrounding fibers collect the emitted fluorescence and return it to the detector. The gold-coated central fiber prevented crosstalk between the excitation light and the surrounding fiber bundle. The complete fiber bundle was 300 µm in diameter, and it was sheathed (Figure 1C) in a 3-cm-long, thin-walled stainless steel (17) Flock, S. T.; Jacques, S. L.; Wilson, B. C.; Star, W. M.; van Gemert, M. J. C. Laser Surg. Med. 1992, 12, 510-519. (18) Kong, G.; Dewhirst, M. W. Int. J. Hyperthermia 1999, 15, 345-370. (19) Kong, G.; Braun, R. D.; Dewhirst, M. W. Cancer Res. 2000, 60, 44404445. (20) Edwards, R. H. Nat. Neurosci. 2001, 4 (3), 221-222.

Figure 1. Miniaturized fiber-optic microprobe constructed for in vivo monitoring of the delivery of fluorescent nanospheres. (A) a, the central excitation fiber having a 100-µm core diameter and a 20-µm cladding; b, a 30-µm coating layer of gold; the scale bar has a length of 100 µm. (B) c, 10 emission fibers, each having a 50-µm core, surround the central excitation fiber coaxially. (C) The complete fiber bundle has a 300-µm diameter and is encased into a 3-cm-long stainless steel tube (i.d. ∼300 µm; o.d. ∼400 µm); the black bar has a length of 1 cm. (D) The tip of the stainless steel tube was finished with a bevel of 30° to facilitate insertion of the microprobe into tissues.

tube (Small Parts, Inc., Miami Lakes, FL) having an outside diameter of 0.016 in/ (∼400 µm) and an inside diameter of 0.012 in. (∼300 µm). The stainless steel tubing provided sufficient structural rigidity; its tip was beveled by 30° to facilitate the insertion of the fiber-optic bundle into tissues (Figure 1D). The microprobe tip was doped with Sigmacote, a silicone solution in heptane (Sigma), to protect the tip and suppress blood clotting. Apparatus and Reagents. Nanospheres. Carboxylated polystyrene nanospheres, FluoSpheres, modified with fluorescent dye (20 nm: F8787, the coefficient of variation for size distribution is 20% and fluorescein equivalents per nanosphere is 1.8 × 102. 1 µm: F8823, the coefficient of variation for size distribution is within 5% and fluorescein equivalents per nanosphere is 1.3 × 107) were purchased from Molecular Probes, Inc. (Eugene, OR). The size distribution of the nanospheres is with a coefficient of variation of ∼20%, as determined by electron microscopy. Stock solutions (particle/mL) were diluted with 50 mM phosphate buffer solution, and then absorption and emission spectra (absorption 505 nm; emission 515 nm) were measured using a Cary Eclipse spectrophotometer (Varian Inc., Palo Alto, CA). The nanosphere solutions were administered intravenously (as a bolus) into the animals. The fluorescence intensity of the nanoparticles in the interstitial fluid, which was detected by the fiber-optic microprobe, was monitored using a photomultiplier tube (PMT; Hamamatsu R928) equipped with a front 510-nm long-pass filter (Chroma Technology Corp., Rockingham, VT). A Melles Griot argon laser (488 nm, 40 mW; Carlsbad, CA) was used as the excitation light source for nanosphere fluorescence (Figure 2). Lipofundin MCT/LCT Solution as the Phantom Medium. In this study, we used stock 10% solids Lipofundin MCT/LCT (B. Braun

Figure 2. Schematic representation of the experimental setup. The microprobe was catheterized as a bifurcated microlight guide. The central fiber carried the excitation light from a 488-nm argon laser to the tissue; the surrounding fibers collect the emitted fluorescence from the nanospheres at the locus and return it to a Hamamatsu R928 PMT equipped in the front with a 510-nm long-pass filter.

Melsungen AG, Melsungen, Germany) as the scattering medium. It consists of about 90% water and 10% fatty micelles having indices of refraction only slightly higher than that of water. The fluorescent nanospheres were mixed with the 10% Lipofundin MCT/LCT solution to reach final concentrations of 4, 8, 16.2, 33.5, 65, 130, and 260 (× 1010 particles/mL), respectively; we evaluated the fluorescence intensities of these nanosphere/lipofundin solutions using the fiber-optic microprobe. Analytical Chemistry, Vol. 77, No. 4, February 15, 2005

1127

Animal Preparations and Models of Ischemic Insult. Male Sprague-Dawley rats (260-350 g) were deprived of food and had free access to water for 12 h before the experiments. All animal experiments were conducted under the Taichung Veterans General Hospital and National Health Research Institutes guidelines for the care and use of laboratory animals. The animals were anesthetized with urethane (1.2 g/kg, ip), and their body temperature was maintained at 37 °C by using a heating pad. Polyethylene catheters were inserted into the femoral artery for the administration of the fluorescent nanospheres (2.6 × 1014 particles/mL × 1 mL). Hepatic Ischemia. After a transverse upper abdominal laparotomy, the liver was mobilized and isolated by dividing all of its attachments with negligible bleeding. Temporal ischemia was introduced using a procedure similar to the portal triad clamping model (PTC model) that used total ligation of the hepatic pedicles (hepatic artery, portal vein, and bile duct) in the hepatoduodenal ligament. Cerebral Ischemia. The rat’s head was mounted on a stereotaxic apparatus (Davis Kopf Instruments, Tujunga, CA) with the nose bar positioned 3.3 mm below the interaural line. Following a midline incision, the skull was exposed and one burr hole was drilled on the skull for inserting the fiber-optic microprobe. The fiber-optic microprobe was implanted into the cortex (-0.5 mm anterior and 5.5 mm lateral to the bregma and 4.0 mm from the brain surface). Cerebral ischemia was induced by ligation of the bilateral common carotid arteries and unilateral middle cerebral artery. The intracranial cerebral blood flow was monitored using a laser Doppler system. RESULTS AND DISCUSSION The jump from in vitro to continuous in situ and in vivo monitoring of fluorescent substances, such as fluorescent nanoparticles, is a challenge, but it is an important issue necessary for advancing targeted drug delivery and controlled release. We have demonstrated that implantable microdialysis is a minimally invasive method for the interstitial sampling and analysis of nanospheres delivered in vivo.8 The specific molecular weight cutoff (MWCO) for a dialysis membrane, however, prevents the identical microdialysis probe from being utilized for in situ sampling of fluorescent substances that have different molecular sizes; such an application may provide important information regarding in vivo size-related transport. The necessity of sampling biological fluids for ex vivo analysis while using microdialysis may disturb the local microenvironment and thus the accuracy of measured fluorescence substances. In this report, we designed and constructed an implantable fiber-optic microprobe that possesses a size comparable to that of a clinically approved microdialysis perfusion fiber. The configuration of the fiber-optic bundle of the microprobe is illustrated in the Experimental Section and in Figure 1. This bundle was designed to be suitable for the continuous in situ monitoring of fluorescent nanosphere delivery in anesthetized rats, whereby the “permeating window” of vasculature could be determined via analyzing the fluorescence intensity of extravasating nanospheres having various sizes. Within the fiber-optic bundle, the central optical fiber, which delivers the excitation light from a laser, was coated with a layer of gold. This metal coating not only increases the internal reflection of the light but it also prevents any signal contamination 1128

Analytical Chemistry, Vol. 77, No. 4, February 15, 2005

from cross-talk between the laser and the relatively weak fluorescence signals. Consequently, this fiber could be used further as a sensor for fluorescence lifetime measurements, where excitation leakage often causes inaccurate readings for short lifetimes.21 A differently design fiber-optics system has been reported for in situ laser fluorometry.22 A single optical fiber was used to perform the excitation of the tissue and to transfer the fluorescence signal back from the tissue to the photoreceiver. Although this design has the advantage of requiring only a single fibertissue interface, it must deal with the Fresnel reflection, which directs the excitation light back into the fiber, that results from the different refraction index at the interface; this process causes excitation leakage to the signal photodetector and degrades the signal measurement. For the multifiber configuration of our constructed microprobe, the excitation light and signal are transmitted by different optical fibers and so there is no Fresnel reflection. The excitation light is absorbed and scattered by the tissue and attenuated quickly from the tip of the fiber. Thus, the scattered excitation light collected by the signal fibers is extremely weak relative to that caused by Fresnel reflection in the singlefiber design. Incorporated with laser scanning multiphoton microscopy, microendoscopes (350-1000 µm in diameter) based on gradient index (GRIN) microlenses that effectively eliminate self-phase modulation within the endoscope can exhibit micrometer-scale resolution for in vivo tissue imaging.23,24 This is an excellent tool for minimally invasive imaging of fluorescent substances deep in the tissue; however, the combination with multiphoton microscopy results in an expensive implementation. To avoid Fresnel reflection, a GRIN microlens-based probe constructed with different photon-tissue interfaces for excitation and emission has been attempted, in which the excitation light coupled through parallel fibers surrounding the perimeter of the lens.25 In this scheme, only light backscattered off tissue at and below the image plane returned through the GRIN lens was detected. Therefore, the signal-to-noise ratio was not appreciable unless with a significant increase of excitation energy. On the contrary, our fiber-optic microprobe was configured using gold-coated fiber at the center to deliver the laser excitation, so that fluorescence emitted from the optical volume beneath the central excitation would be efficiently collected by the surrounding fibers. We established the performance of our fiber-optic microprobe in highly scattering media (Figure 3), a 10% Lipofundin MCT/ LCT suspension, which we used as a tissue-simulating phantom. We prepared samples of the 20-nm fluorescent nanospheres at final concentrations of 4, 8, 16.2, 33.5, 65, 130, and 260 (× 1010 particles/mL) using either Ringer solution or a 10% lipofundin suspension. The fluorescence intensity measured using the fiberoptic microprobe in the nanosphere Ringer solution indicated that (21) Lo, L. W.; Wilson, D. F. In Oxygen Transport to Tissue; Thorniley, Harrison, James, Ed.; Kluwer Academic/Plenum Publishers: New York, 2003; Vol. XXV, pp 117-123. (22) Renault, G.; Raynal E.; Sinet, M.; Muffat-Joy, M.; Berthier, J. P.; Cornillault, J.; Godard, B.; Pocidalo, J. J. Am. J. Physiol. 1984, 246, H491-H499. (23) Jung, J. C.; Schnitzer, M. J. Opt. Lett. 2003, 28, 902-904. (24) Levene, M. J.; Dombeck, D. A.; Kasischke, K. A.; Molloy, R. P.; Webb, W. W. J. Neurophysiol. 2004, 91, 1908-1912. (25) Fisher, J. A. N.; Civillico, E. F.; Contreras, D.; Yodh, A. G. Opt. Lett. 2004, 29, 71-73.

Figure 3. Characterizing the fiber-optic microprobe by measuring the fluorescence of nanospheres against highly light-scattering media. Samples of the 20-nm fluorescent nanospheres were prepared at final concentrations of 4, 8, 16.2, 33.5, 65, 130, and 260 (× 1010 particles/ mL). (A) The fluorescence intensity was proportional to the nanosphere concentration when prepared in the Ringer solution. (B) In a 10% Lipofundin suspension, which was used as a phantom tissue, the concentration dependence of the fluorescence intensity was evident for nanosphere concentrations > 65 × 1010 particles/mL.

it is dependent on the nanosphere concentration (Figure 3A). In contrast, the fluorescence intensity dependence in the 10% lipofundin suspension was pronounced only at nanosphere concentrations of >65 × 1010 particles/mL (Figure 3B). These results suggest that even though the multifiber configuration was used to eliminate Fresnel reflection, the scattered excitation light collected by the signal fibers still interfered with the fluorescence intensity measurements, albeit to a modest degree. This finding is especially true for fluorescence measurements conducted in such a turbid environment as that found in tissue. For an anesthetized rat having a weight of ∼260 g (blood volume: ∼5% of total body weight), an administered sample of fluorescent nanospheres of 2.6 × 1014 particles/mL × 1 mL results in a final concentration of ∼2.0 × 1013 particles/mL in vascular circulation. This value is ∼10 times higher than the highest concentration of nanospheres that we used in the fiber-optic measurements of fluorescence intensity versus nanosphere concentration and is 40 times higher than the threshold concentration (65 × 1010 particles/ mL) above which the scattering interference in Lipofundin phantom could be neglected (Figure 3B). For continuous in situ monitoring of the delivery of fluorescent nanospheres, we validated the use of the fiber-optic microprobe

by using both hepatic and cerebral ischemia as in vivo models. The microprobe was placed into the liver of an anesthetized rat to analyze the concentration changes of fluorescent nanospheres extravasating from the hepatic vasculature under ischemic insult. As indicated in Figure 4A, the Ringer solution used as a sham, with a volume equal to that of the fluorescent nanosphere solution, was introduced (iv, bolus) prior to the ischemic onset. There was no change in the fluorescence signal detected by the implanted microprobe during the course of ischemia. In contrast, for the 20-nm fluorescent nanospheres, we observed an immediate response of increasing fluorescence intensity. The upward signal manifested the increase of extravasation of circulating nanospheres, probably as a result of the augmentation of vascular permeability subject to ischemia (Figure 4B). The baseline, which depended on the individual animal, sometimes increased slightly in response to administering the fluorescent nanospheres. We infer that under normaxia, the 20-nm nanospheres may still be able to permeate through the blood vessels, but only to a very limited degree. As we mentioned above, because of its capability to measure nanospheres of various sizes, the fiber-optic microprobe is better suited than microdialysis for monitoring the delivery of fluorescent nanospheres in vivo. Therefore, we employed the 1-µm nanospheres to preliminarily verify the size cutoff of vascular permeability in ischemic liver. The rather constant fluorescence intensity measured using the fiber-optic microprobe implies that the 1-µm nanospheres remained in the vasculature even after the onset of ischemia (Figure 4C). When hepatic ligation occurs, the physical operation might cause an instant and transient disturbance in the microprobe readings, but we observed that the disturbance on the signal was insignificant and it restored back to the baseline rapidly. In conclusion, these results suggest that the fiber-optic microprobe can be used for in situ evaluation of the “permeating window” of hepatic vasculature under ischemic insults that are in the range from 20 nm to 1 µm. Because the tight junction formed by the BBB greatly impedes the delivery of therapeutic drugs, information concerning the BBB permeability under various physiological conditions is invaluable for determining the effectiveness of chemotherapeutics toward central nervous system (CNS) disorders. Therefore, we used our constructed fiber-optic microprobe to examine the effect of cerebral ischemia on the extracellular fluorescent nanospheres extravasating from BBB (Figure 5). Similar results were obtained in comparison to those investigated under hepatic ischemia and illustrated in Figure 4. Both types of ischemic insults increased the accumulation of extracellular fluorescent nanospheres immediately and significantly, which suggests a rapid increase in blood vessel permeability following ischemic injuries. The immediate increase of fluorescence intensity after cerebral ischemia, however, was not as prominent as that after hepatic ligation (Figure 5B). This finding suggests that, in comparison to the hepatic vasculature, a BBB tight junction is more difficult for those nanospheres to surpass, even within a “size window” for permeation as low as 20 nm. We have characterized the vascular permeability in ischemic liver and brain on the basis of the concentration (fluorescence intensity) of circulating nanospheres that entered the optical sensing volume of the implanted microprobe via the hepatic vasculature or BBB. Because the excitation light was scattered Analytical Chemistry, Vol. 77, No. 4, February 15, 2005

1129

Figure 4. Hepatic vascular permeability following ischemia. The microprobe was placed into the liver of an anesthetized rat. The change in vascular permeability in ischemic liver was evaluated in terms of the change in the fluorescence intensity of the extravasating nanospheres. (A) The Ringer solution containing no fluorescent nanosphere was administered (R; iv, bolus) prior to the total ligation of hepatic pedicles (L). No change in the fluorescence intensity was observed throughout the course of ischemia. (B) For preadministered 20-nm fluorescent nanospheres (N), an immediate and significant increase in the fluorescence intensity was detected upon the induction of hepatic ischemia. (C) For preadministered 1-mm nanospheres (N), the fluorescence intensity remained at the basal level following the ischemic insult. The transient variation of fluorescence intensity observed immediately after the ligation might be a consequence of the physical disturbance resulting from the ligation surgery.

and attenuated rapidly as it moved away from the tip of the microprobe, the fluorescence was maximal nearest the tip. The collected fluorescence arose primarily from a small volume of tissue (