[⁶⁴Cu]Cu-Albumin Clearance Imaging to Evaluate Lymphatic Efflux of Cerebrospinal Space Fluid in Mouse Model

Abstract

Purpose

Clearance of brain waste in the cerebrospinal fluid (CSF) through the meningeal lymphatic vessels (mLV) has been evaluated mostly through the fluorescent imaging which has inherent limitations in the context of animal physiology and clinical translatability. The study aimed to establish molecular imaging for the evaluation of mLV clearance function.

Methods

Radionuclide imaging after intrathecal (IT) injection was acquired in C57BL/6 mice of 2–9 months. The distribution of [^99mTc]Tc-diethylenetriamine pentaacetate (DTPA) and [⁶⁴Cu]Cu-human serum albumin (HSA) was comparatively evaluated. Evans Blue and [⁶⁴Cu]Cu-HSA were used to evaluate the distribution of tracer under various speed and volume conditions.

Results

[^99mTc]Tc-DTPA is not a suitable tracer for evaluation of CSF clearance via mLV as no cervical lymph node uptake was observed while it was cleared from the body. A total volume of 3 to 9 μL at an infusion rate of 300 to 500 nL/min was not sufficient for the tracer to reach the cranial subarachnoid space and clear throughout the mLV. As a result, whole-body positron emission tomography imaging using [⁶⁴Cu]Cu-HSA at 700 nL/min, to deliver 6 μL of injected volume, was set for characterization of the CSF to mLV clearance. Through this protocol, the mean terminal CSF clearance half-life was measured to be 123.6 min (range 117.0–135.0) in normal mice.

Conclusions

We established molecular imaging to evaluate CSF drainage through mLV using [⁶⁴Cu]Cu-HSA. This imaging method is expected to be extended in animal models of dysfunctional meningeal lymphatic clearance and translational research for disease-modifying therapeutic approaches.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13139-022-00746-6.

Introduction

In the brain, a significant amount of waste products are generated as a by-product of metabolic activity and discharged in the cerebrospinal fluid (CSF) which may promote the development of brain diseases [1, 2]. It was known that the drainage of CSF mainly occurs through the arachnoid villi, but due to the nature of the arachnoid villi composed of vascular structures, it is difficult to drain the macromolecule, and it is assumed that there are alternative routes [3]. Recently confirmed existence of the peri-vascular lymphatic system in the dura matter that drains in the cervical lymph node (LN) suggests a functional implication of brain waste clearance from CSF to meningeal lymphatic vessels (mLV) [4–8].

Brain waste after its origin from the neurons and glial cells is disposed into the interstitial fluid and then via the paravascular space to the CSF in subarachnoid space (SAS) [9–11]. Then, they cross the arachnoid barrier cells for a transit through the arachnoid granulation-like dural gap to enter the mLV [12–15]. The cranial mLV route around the dural sinuses in the dorsal and basal aspect of the cranium and drain in the cervical LN [16–18]. Dysfunctional excretion of CSF through mLV causes a backlog of macromolecules, mediators, and immune cells in the CSF and brain interstitial fluid in aging, Alzheimer’s disease, and other neurological disorders [16, 19–21]. On the contrary, alleviation of macromolecular backlog was documented in the animal models undergoing novel experimental treatments [16, 22].

The exploration of the CSF to mLV and the brain interstitial fluid to CSF clearance routes have depended heavily upon ex vivo fluorescence imaging which is incapable of collecting data over multiple time points from the same animal. Indeed, in vivo longitudinal analysis is necessary for the verification of group differences and individual variations [23]. Furthermore, the reported ex vivo profiles of CSF to mLV clearance were based upon tracer infusion volume and rates which were significantly higher than that of the physiological volume and rates of CSF production [5, 11, 16, 24–33], which may raise the intracranial pressure (ICP) resulting in an increased CSF outflow through both the lymphatic and arachnoid villi [34] leading to erroneous quantification. Meanwhile, the injection of gadolinium contrast into CSF reportedly caused neurotoxicity in animals with an overall adverse event rate of 13% in humans [35]. The safety recommendations of keeping the contrast infusion at the lower end may affect the image interpretability because the “% change in signal unit ratio” in the cervical LN was indistinct in comparison to that in the CSF [36]. On the contrary, radionuclide imaging is inherently capable of producing images with a high target-to-background ratio with an injection of tracer at sub-pharmacological dose but with a high specific activity. The translational research on neurodegenerative disease relies heavily on mouse models, albeit confining the use of radionuclides within ex vivo scintillation counting while the use of radionuclide imaging remained underutilized in mice [37–39].

The current study aimed to establish the radionuclide molecular imaging method for the evaluation of brain-CSF clearance through mLV and optimize the imaging protocol, to be as much as possible within the physiological range.

Methods

Animals

Male C57BL/6 mouse of 2–9 months of age was used. They were housed 2–5 per cage with free access to standard food and potable water. The housing room was maintained at a constant temperature of 22–24 °C with a 12/12-h light and dark cycle.

Tracer Injection Under Anesthesia

The animal was first induced with 3% isoflurane with oxygen flow 500 mL/min for ~3 min, and then placed in a prone position and head away from the operator on a heating pad with its nose inside a gas mask providing 2.8% isoflurane with oxygen flow 500 mL/min during the injection. The L₄ spine was palpated for localizing the site for intrathecal (IT) injection. About 1 × 1-cm area of local skin was shaved, iodine solution was applied, and then, a 1-cm-long sagittal incision was made with scissors. Dissection of muscles was avoided with the idea that it helps to reduce the backflow. The needle was inserted through the muscles adjacent to the L₄ spine, up to an appropriate length and watching for the tail to flick [40]. CSF injection was made by a syringe pump (Harvard Apparatus), calibrated for the rates and volumes using a 0.5-cc (500 μL) syringe with major graduations of 50 μL and minor graduations of 10 μL. Injection of [⁶⁴Cu]Cu-HSA was done at the desired rate to deliver a certain volume of total tracer. The needle was then kept in place until the first imaging session was over. Thereafter, the needle was removed, and the wound was closed with 6-0 silk. The animal had remained under continuous anesthesia during the intrathecal access, injection, first session of image acquisition, and wound closure. Thereafter, the animal was placed within a warm cage having access to food and water, and allowed to wake up spontaneously. The animal was induced with anesthesia for 3 min before each of the upcoming sessions of image acquisition.

Selection of Radiotracer

The [^99mTc]Tc-diethylenetriamine pentaacetate (DTPA) with a molecular weight of 0.39 kDa and the [⁶⁴Cu]Cu-human serum albumin (HSA) with a molecular weight of 66.5 kDa were used for single-photon emission computed tomography (SPECT) and positron emission tomography (PET) imaging respectively. Both tracers were injected initially at an infusion rate of 2 μL/min for 10 min.

Selection of Rate and Volume of Infusion for IT Injection

Evans Blue dye is known for avid binding with albumin, thus capable of representing the kinetics of albumin in the SAS [41]. Evans Blue dye was suspended in artificial CSF for IT injection for the ex vivo experiment. The infusion rate and volume were set in consideration of the physiological CSF production rate of 350 nL/min in mice [32]. Three different experimental sets of infusion rates and volumes were evaluated: (1) 300 nL/min and 3 μL, (2) 300 nL/min and 9 μL, and (3) 500 nL/min and 3 μL. The distribution of Evans Blue dye along the spinal and cranial SAS was inspected ex vivo at 25 min post-injection. To further confirm the optimal infusion rate and volume in PET imaging, two different sets of infusion rate and volumes were evaluated: (1) 500 nL/min and 3 μL and (2) 700 nL/min and 6 μL.

Image Acquisitions and Reconstruction

Whole-body SPECT/computed tomography (CT) was done using NanoSPECT/CT plus (Mediso). A total of 24 projections into an 80 × 80 acquisition matrix were obtained with the frame time being 15 s for SPECT acquisition. The image acquisition time points were at 9 min post-injection followed by 2, 4, 8, 12, and 24 h post-injection. The SPECT reconstruction used a 3-dimensional ordered-subsets expectation maximum (OSEM) algorithm. The whole-body CT used 55-kVp x-rays with 180 projections, 500 ms of exposure time, and a 1.5 pitch. Genisys PET box (Sofie Biosciences) was used for whole-body PET acquisition of static PET in list mode for 6 min at each of the image acquisition time points. Reconstruction of PET images was automatically done by the vendor-provided software that produced the DICOM files. Imaging of a tube containing PET tracer with a known activity was acquired for calibration and percent injected dose (%ID) calculation.

Image Analysis

Analyses are done using the MIM software. A separate 3D region of interest (ROI) was drawn over SAS and the other organs, e.g., lymph nodes, heart, and liver. Additional whole-body ROI was drawn over the 9-min post-injection image to derive the %ID. The radioactivity within the needle at the injection site was excluded in all cases. Estimated ROI counts were corrected for the decay of radioisotope and thereafter normalized as %ID. The %ID was then plotted against the corresponding time points.

Statistical Analysis

Generation of plots and calculation of clearance half-life was done using the GraphPad Prism. CSF clearance was fitted to an exponential function, using the least-squares method, and the coefficient of determination R² was used as indicator goodness of fit, with R² > 0.95 considered to be a good fit. Values were reported as median and range.

Results

Selection of Radiotracer

[^99mTc]Tc-DTPA was injected with the infusion rate of 2 μL/min for 10 min. [^99mTc]Tc-DTPA distributed along the spinal to cranial SAS. However, no cervical lymph node uptake was visualized while [^99mTc]Tc-DTPA was cleared from the body through the urinary tract (Fig. 1a). Next, an IT injection of [⁶⁴Cu]Cu-HSA was made with the same infusion rate of 2 μL/min for 10 min. The spinal and cranial SAS was seen and the cervical LNs were visualized as proof of clearance of [⁶⁴Cu]Cu-HSA from CSF through mLV (Fig. 1b).

Fig. 1.

Image from fusion [^99mTc]Tc-DTPA SPECT-CT in a 2-month-old C57BL/6 mice, at 4 h post-injection showing the needle and the tracer inside the spinal subarachnoid space. There is a visualization of the urinary bladder, white arrow (a). Example image from [⁶⁴Cu]Cu-HSA PET in a 9-month-old C57BL/6 mice at 4 h post-injection showing the cranial as well as the spinal subarachnoid space. There is a visualization of the cervical lymph node on both sides, arrowheads (b). The tracer was infused at a rate of 2 μL/min for 10 min on both occasions; the color bar indicates the activity concentration of the tracer

Optimal Rate and Volume of Infusion by Ex Vivo Experiment

After Evans Blue dye infusions at a rate of “300 nL/min for 10 min,” the dye was confined within the thoracolumbar SAS and did not reach the cervical or cranial level (Fig. 2a). Continuing Evans Blue dye infusion at a rate of “300 nL/min for 30 min” was also found to be within the thoracolumbar SAS (Fig. 2b). The Evans Blue dye infusion reached the basal cistern of the other mouse when the infusion was made at a rate of “500 nL/min for 6 min” (Fig. 2c).

Fig. 2.

Images of ex vivo inspection for extension of Evans Blue dye at 25 min post-injection in 2-month-old C57BL/6 mice showing Evans Blue dye does not extend beyond the lumbar subarachnoid space at an infusion rate of 300 nL/min for 10 min (a) and thoracic subarachnoid space at an infusion rate of 300 nL/min for 30 min with the cervical subarachnoid space remaining unstained (b). The infusion rate of 500 nL/min for 6 min could push Evans Blue dye up to the cranial subarachnoid space (c). Example images from [⁶⁴Cu]Cu-HSA PET in 3-month-old C57BL/6 mice at 4 h post-injection using infusion rate of 500 nL/min for 6 min showing non-visualized cervical lymph node, cranial, and cervical subarachnoid space. Tracer was seen within lumbar subarachnoid space and lower part of thoracic subarachnoid space; colorbar indicates activity concentration of the tracer (b)

Thereafter, [⁶⁴Cu]Cu-HSA PET was done with an infusion rate of “500 nL/min for 6 min.” The cranial SAS and the cervical LNs were not visualized until 4 h post-injection. The heart, reflecting the systemic circulation, was visualized (white arrow) indicating an alternative route for clearance of tracer (Fig. 2d).

CSF Clearance to the Lymphatics

With the infusion rate further increased to 700 nL/min and the total infused volume being 6 μL, all five mice showed a similar pattern of distribution of the tracer in the cranial and spinal SAS, with visualization of cervical (superficial and deep) lymph nodes, iliac and sacral lymph nodes, and liver. Figure 3 and supplementary video 1 show representative images from 9-min as well as the 1-, 2-, 4-, 6-, 12-, and 24-h image acquisition time points. At 9-min post-injection image, there was intense tracer concentration in the spinal SAS, particularly in the lumbar, thoracic, and cervical segments. The 1-h image showed gradual dispersion of tracer in the cranial and sacral SAS as well as the superficial and deep cervical LN. The faint activity in the blood pool and liver was visualized. The images from 1 to 6 h post-injection showed clearance of activity from the thoracic and cervical SAS to the systemic circulation through cervical, iliac, and sacral LNs and accumulated in the liver. At 12 h and 24 h, [⁶⁴Cu]Cu-HSA was mostly cleared from the SAS with retention of activity in the nasal area; cervical, iliac, and sacral LNs; and sacral SAS.

Fig. 3.

Representative images of whole-body [⁶⁴Cu]Cu-HSA PET, in a 6-month-old C57BL/6 mouse, using the infusion rate of 700 nL/min to deliver a total volume of 6 μL over 8.5 min. The cranial and spinal subarachnoid spaces are seen. The superficial and deep cervical lymph nodes, heart, liver, and sacral and iliac lymph nodes are seen; color bar indicates activity concentration of the tracer

The time vs %ID line plots (Fig. 4) showed an overall temporal decline in the %ID from the entire SAS, representing “CSF to mLV clearance” (n=5). The CSF clearance of the [⁶⁴Cu]Cu-HSA was bi-exponential with a mean terminal half-life of 123.6 min (range 117.0–135.0). A good fit was achieved for all data sets. The time vs %ID line plots for the liver and LNs showed peak activity between 4 and 6 h followed by a slow decline (Fig. 5a, b). The total uptake in the superficial and deep was not significantly different compared to that in the iliac and sacral lymph nodes.

Fig. 4.

The %ID plotted against time from the whole-body [⁶⁴Cu]Cu-HSA PET, in five 6-month-old C57BL/6 mice using the infusion rate of 700 nL/min to deliver a total volume of 6 μL over 8.5 min with the ROI counts taken from the 9-min as well as 1-, 2-, 4-, 6-, 12-, and 24-h post-injection image acquisition time points. Temporal change of %ID in the entire subarachnoid space is shown as line plots with separate lines for each of the mice

Fig. 5.

The %ID plotted against time from the whole-body [⁶⁴Cu]Cu-HSA PET, in five 6-month-old C57BL/6 mice using the infusion rate of 700 nL/min to deliver a total volume of 6 μL over 8.5 min with the ROI counts taken from the 9-min as well as 1-, 2-, 4-, 6-, 12-, and 24-h post-injection image acquisition time points. Temporal change of %ID in the liver is shown as line plots with separate lines for each of the mice (a). Line plots of temporal change of %ID for the pelvic and cervical lymph node (n=5); the error bars are shown for each time point (b)

Discussion

We have established a molecular imaging method for the evaluation of CSF clearance through mLV using [⁶⁴Cu]Cu-HSA with the infusion rate of 700 nL/min and the total infused volume of 6 μL. We confirmed the CSF clearance of the HSA via mLV to LNs and systemic circulation in normal mice through the longitudinal acquisition of PET images at multiple time points.

We determined the optimal infusion rate and volume for PET imaging by considering the physiological production rate and volume of CSF in mice. The CSF volume and CSF production rate in mice generally known through studies are 36.6 μL and 370 nL/min [32, 33]. Accordingly, the infusion rate in this study in the 6-month-old mice was twice higher than the physiological rate of CSF production and the total injected volume reached 16% of the total CSF volume. On the contrary, some of the investigators have used infusion rates of 1–2 μL/min that is 2.9–5.7 times higher than the physiological rate of CSF production and infusion volume of 10–15 μL that reached 25–37.5% of the total CSF volume [5, 11, 16, 24–30, 32] (Supplementary Table 1).

This was of particular concern because infusion at a rate twice that of the physiological rate of CSF production had doubled the ICP in the rat model [42, 43], while a twofold rise of ICP in the ovine model caused the CSF drainage rate to rise several-fold higher through both the lymphatic and arachnoid villi routes [34]. However, a 2.5-mmHg rise of ICP in mice during cisterna magna infusion using rate and volume of 1 μL/min and 10 μL resulted in about 160% of the ICP at 10 weeks, which was claimed to be mild and transient based on the observation of the absence of pathological reflux to ventricles during infusion [31, 44]. In addition to the dependence on rate and volume of infusion, the rise of ICP during an intrathecal infusion is subjected to homeostatic pressure-accelerated drainage of CSF into blood [45]. Since the immediate post-injection images in our study do not show any evidence of CSF to lymphatic outflow, we may assume that our infusion rate did not raise the ICP to a point that would require compensation through pressure-accelerated drainage of CSF. Thus, the infusion rate and volume used in this study would at least induce less perturbation of the ICP, and consequently, physiological changes are expected to be transient and insignificant.

The in vivo PET imaging found the average terminal half-life of HSA from CSF in the entire SAS to be 123.6 min when injected at a rate of 700nL/min to deliver 6 μL of tracer. Published data suggest a clearance half-life of 48 min using bovine serum albumin radiolabelled with I-125 which was injected IT at a rate of 2.26 μL/min delivering a total volume of 12 μL in 2-month-old mice by the ex vivo scintillation counting of blood-free CSF [32]. This 2.5 times higher clearance half-life in our study is attributable to the 70% lower infusion rate and 50% lower infusion volume despite the similar molecular characteristic of the tracer delivery through the same route of administration. The time-%ID curves in our study show accumulation of activity in the liver while there was the appearance of activities in the urinary bladder and colon indicating separation of HSA in liver and kidneys followed by renal and hepatic excretion of [⁶⁴Cu]Cu [46, 47]. The combined %ID in the pelvic and cervical LNs remained similar throughout the study period and the %ID was not significantly different in the cranial and caudal parts of SAS based on injection site (Supplementary fig 2a-b). Accordingly, it can be assumed that the proportion of clearance of CSF through the cranial and pelvic mLV may be similar. The higher uptake in superficial cervical LN in comparison to that in deep cervical LN at all time points may reflect the difference in their received amount of lymph from their corresponding territories in addition to the difference in their sizes (Supplementary fig 3a). A continuous increase in iliac LN uptake, while uptake of sacral LN reached an early peak and decrease, reflects the fact that sacral mLV is draining CSF to iliac LN via sacral LN (Supplementary fig 3b).

We have chosen HSA rather than DTPA as the radiotracer for radionuclide molecular imaging of mLV clearance. The exploration started with the use of DTPA inspired by its established utility in clinical radionuclide cisternography [9]. Radionuclide cisternography is a well-established nuclear medicine imaging technique that uses intrathecally injected radiolabeled DTPA to visualize the flow of CSF. However, in this study, no evidence was found for DTPA to pass through mLV until it is cleared from the CSF into the systemic circulation. The clearance of DTPA without visualization of the lymph node is likely attributable to the molecular size of the DTPA (393 Da) which may allow the passage of tracer via the mLV through LN without sequestration, in addition through the venous clearance to reach the systemic circulation [48, 49]. This also matches with the fact that cervical LN uptake of radiolabelled-DTPA has never been reported in humans [50, 51], whereas HSA in addition to being biocompatible [52] has a molecular weight (66.5 kDa) nearly similar to that of 55–62-kDa tau proteins [53]. The hydrodynamic diameter of the [⁶⁴Cu]Cu-HSA was 8.8 ± 1.4 nm, which falls within the range of tau that can reach up to 14 nm [54, 55]. Surface charge also affects the lymphatic clearance of the molecule and a denser negative surface charge can confer faster lymphatic clearance [56, 57]. [99mTc]Tc-DTPA, HSA, and phosphorylated tau molecules bear negative surface charges at physiological pH [58–62]. Therefore, the size effect is expected to be dominant, and the effect by surface charge is expected to be similar among the molecules.

HSA labeling with [⁶⁴Cu]Cu was done using click-based technology (supplementary fig. 1) which is known for high radiochemical purity and serum stability [63]. Availability of [⁶⁴Cu]Cu-HSA in a high specific activity facilitated a high-quality image despite the low volume of tracer injection in this study. The 12.7 h of half-life of [⁶⁴Cu]Cu allowed longitudinal observation of clearance through multiple time points over 24 h.

Indeed, the HSA is known to be immunogenic in mice but the rapid catabolism of HSA in mice results in a rapid decline in the amount of immunogenic HSA after a single injection [64, 65]. Given an HSA concentration of 1 μg/μL in our administered tracer, the total administered HSA could have reached 6 μg for one mouse whereas the dose-dependent anaphylactic response leading to death in mice due to xenogeneic albumin required re-injection of 25 μg or higher amount through intravenous route after 3 weeks of an initial subcutaneous immunization [66]. Since we have made a single injection, the issue of immunological reaction at a later time point is beyond the scope of this study. Furthermore, our final goal is the clinical application of this imaging method, and [64Cu]Cu-HSA may be a feasible candidate tracer.

Our results show that an infusion rate that was either lower or up to twice higher than the physiological rate was insufficient to facilitate the distribution of tracer throughout the SAS after a single injection through the intrathecal route. Concordantly, some investigators who used infusion rates of 0.5 or 1 μL/min with infusion volumes less than 10 μL had a partial observation which was confined either around the cranial or the spinal SAS [4, 37, 67–71] (Supplementary Table 1).

Contrary to the intra-cisterna magna, the widely used route for tracer delivery, the intrathecal route, delivers the tracer at a site where the turbulence of CSF is known to be slow unlike the fast turbulence in the vicinity of cisterna magna owing to the vigorous speed of CSF in the fourth ventricle, ventral surface of the brain stem, and cervical SAS [72–74]. Therefore, the slow speed of delivery in our experiment could have pushed the tracer steadily towards the rostral direction without any significant regurgitation through the punctured dura at the injection site. Moreover, the resemblance of intrathecal route with that for clinical radionuclide cisternography may favor the translatability of our protocol.

We preferred to use multiple time points of static image acquisition for short durations under intermittent anesthesia over a longer duration of dynamic imaging under continuous anesthesia primarily because the CSF flow in the spinal SAS, the CSF clearance from spinal SAS to peripheral lymphatic, and the contractility of peripheral lymph vessels are significantly inhibited by isoflurane in comparison to the awake condition [75, 76] which altogether can explain the 9-min post-injection image (Fig. 3) showing faint visualization of cranial and sacral SAS along with non-visualization of peripheral lymph nodes when the animal had been under continuous anesthesia for intrathecal access followed by the injection and then the image acquisition. Moreover, a 60-min period of isoflurane inhalation without thermal support can cause hypothermia [77] that may lead to potentially fatal cardio-respiratory dysregulation [78].

Taken together, this PET imaging protocol may be applied for verifying the claimed coincidence of both glymphatic dysfunction and dysfunctional “CSF to mLV clearance” [16] on the experimental models of glymphatic dysfunction in traumatic brain injury and ischemic stroke [25, 67]. This protocol may also be applied for in vivo evaluation of dysfunctional “CSF to mLV clearance” which is documented as a hallmark for progression or exacerbation of neurodegenerative manifestations in the transgenic animals or the experimental animal models of lymphatic dysfunction or the models of other neurodegenerative and chronic systemic diseases [16, 22, 79]. Since the clearance of CSF macromolecules through mLV has been a target for therapeutic strategies intended for alleviation of neurodegenerative conditions by locally delivered lymphangiogenic factor [80] or disruption of the blood-brain barrier [22, 81], the protocol reported in the current study may find a role for response evaluation after those novel therapeutic approaches. Click chemistry-based radiolabeling of anti-CD4-antibody or cell surface protein of NK-cell, with [⁶⁴Cu]Cu followed by their intrathecal delivery using our protocol, can be a theranostics implication to target the autoreactive CD4⁺ T-effector cells known for exacerbating Alzheimer’s disease [82] or to study the interaction of exogenous NK cells in aging and transgenic mouse models [83]. Our protocol with appropriate modification can also have an extended application for the clinical diagnosis of spinal CSF leaks [84], which has diagnostic implication in spontaneous intracranial hypotension. The diagnosis of CSF leak in spontaneous intracranial hypotension has been inadvertently dependent upon digital subtraction myelography [85, 86] that requires two consecutive injections of contrast injection and eight exposures of dual-energy CT that imparts a significant effective radiation dose [87, 88].

Limitations of this study include the inability to establish a protocol that uses tracer infusion at a rate that is lower than or equal to the physiological rate of CSF production and can claim zero perturbation of homeostasis. Since the ICP in mice was not measured, the change of ICP during infusion and the duration of ICP remaining changed cannot be documented. The diurnal variation of CSF to mLV clearance [89] could not be separately demonstrated in this study because the study lasted over 24 h which confers consecutive inclusion of both the dark and light cycles.

Conclusions

We have established PET imaging for the evaluation of mLV clearance of macromolecules in normal mice with IT injection of [⁶⁴Cu]Cu-HSA at an infusion rate of 700 nL/min that delivers a total volume of 6 μL. This protocol may have an extended utility in animal models of dysfunctional clearance from CSF to mLV clearance and translational research for disease-modifying therapeutic approaches.

Author Contribution

The first, second, third, ninth, and last authors were responsible for the study design. The first and second authors were responsible for interpretation of data and manuscript drafting. The first eight authors were also responsible for performing experiments and data documentation.

Data Availability

The data and material can be made be available upon communication with the corresponding author.

Declarations

Competing Interests

Azmal Sarker was supported by SNU Presidential Fellowship. Minseok Suh, Yoori Choi, Ji Yong Park, Hyeyeon Seo, Seokjun Kwon, Hyun Kim, Eunji Lee, and Dong Soo Lee declares that they have no conflict of interest.

Ethics Approval and Consent to Participate

The study and experiment methodology of this study were approved by the Institutional Review Board and Institutional Animal Care and Use Committee (registration number SNU-200513-8) of Seoul National University College of Medicine. All experiments were conducted under relevant guidelines and regulations regarding the care and the use of animals for the experimental procedures.

Consent for Publication

None

Conflict of Interest

This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIP) (No. 2020R1A2C2101069 and No. 2021R1F1A1064340).