Pretargeted Imaging beyond the Blood–Brain Barrier—Utopia or Feasible?

Pretargeting is a promising nuclear imaging technique that allows for the usage of antibodies (Abs) with enhanced imaging contrast and reduced patient radiation burden. It is based on bioorthogonal chemistry with the tetrazine ligation—a reaction between trans-cyclooctenes (TCOs) and tetrazines (Tzs)—currently being the most popular reaction due to its high selectivity and reactivity. As Abs can be designed to bind specifically to currently ‘undruggable’ targets such as protein isoforms or oligomers, which play a crucial role in neurodegenerative diseases, pretargeted imaging beyond the BBB is highly sought after, but has not been achieved yet. A challenge in this respect is that large molecules such as Abs show poor brain uptake. Uptake can be increased by receptor mediated transcytosis; however, it is largely unknown if the achieved brain concentrations are sufficient for pretargeted imaging. In this study, we investigated whether the required concentrations are feasible to reach. As a model Ab, we used the bispecific anti-amyloid beta (Aβ) anti-transferrin receptor (TfR) Ab 3D6scFv8D3 and conjugated it to a different amount of TCOs per Ab and tested different concentrations in vitro. With this model in hand, we estimated the minimum required TCO concentration to achieve a suitable contrast between the high and low binding regions. The estimation was carried out using pretargeted autoradiography on brain sections of an Alzheimer’s disease mouse model. Biodistribution studies in wild-type (WT) mice were used to correlate how different TCO/Ab ratios alter the brain uptake. Pretargeted autoradiography showed that increasing the number of TCOs as well as increasing the TCO-Ab concentration increased the imaging contrast. A minimum brain concentration of TCOs for pretargeting purposes was determined to be 10.7 pmol/g in vitro. Biodistribution studies in WT mice showed a brain uptake of 1.1% ID/g using TCO-3D6scFv8D3 with 6.8 TCO/Ab. According to our estimations using the optimal parameters, pretargeted imaging beyond the BBB is not a utopia. Necessary brain TCO concentrations can be reached and are in the same order of magnitude as required to achieve sufficient contrast. This work gives a first estimate that pretargeted imaging is indeed possible with antibodies. This could allow the imaging of currently ‘undruggable’ targets and therefore be crucial to monitor (e.g., therapies for intractable neurodegenerative diseases).


Introduction
Positron emission tomography (PET) is a non-invasive nuclear molecular imaging technique that allows for the quantification and visualization of biological processes in vivo with high sensitivity, resolution, and relatively low patient radiation burden [1][2][3][4][5]. PET relies on the administration of radioligands and is commonly used for diagnosis, drug monitoring, or to gain insights into a certain disease pathology [3,4,[6][7][8]. The development of new PET radioligands is essential to develop therapies for diseases with respect to the quantification of a drug's target engagement or to monitor treatment success [9][10][11]. Antibodies (Abs) have high target specificity, selectivity, and typically low non-displaceable binding, making them highly suitable for nuclear PET imaging [12][13][14][15][16].
Particularly for brain diseases, the need for PET radioligands is high [17][18][19]. Neurodegenerative disorders are the second most common cause of death, and the number of affected individuals is rapidly growing due to aging societies [20][21][22][23][24]. Developing radioligands for brain targets, however, is not easy and in contrast to tracers imaging targets outside of the brain, the development of central nervous system (CNS) PET tracers is complicated as these tracers have to cross the blood-brain barrier (BBB) [21,25]. This protective barrier prevents toxins from entering and damaging the brain, but also hampers the uptake of pharmacological agents, especially large molecules such as Abs [26][27][28][29][30]. Monoclonal Abs have been revolutionizing PET imaging within oncology, but the application within CNS remains limited due to their poor BBB penetration [12,13]. In order to increase the brain uptake, several mechanisms for facilitated brain delivery are exploited [20,26,[31][32][33][34]. The most commonly exploited strategy is receptor-mediated transcytosis (RMT), which involves the conjugation of an Ab to a transporting moiety. This moiety binds to a transporter (carrier) and consequently delivers the construct into the brain ( Figure 1A) [35][36][37]. The transferrin receptor (TfR) is the most commonly used transporter thus far [38][39][40][41].  (RTM). A bispecific ligand binds to its carrier protein and is transported into the brain by transcytosis. (B) A schematic overview of pretargeted imaging beyond the BBB. The TCO-Ab is injected first and accumulated at its target within the brain. In a second step, a radiolabeled Tz is administered. This Tz crosses the BBB and reacts with the TCO-Ab. This ligation results in a radiolabeled Ab that can be imaged. (C) An overview of the factors contributing to pretargeting in the brain. Previous work describes the identification of animal models and BBB-permeable Tzs. In this work, parameters regarding BBB-permeable TCO-Abs were investigated.
Abs often have slow pharmacokinetics and long circulation times and therefore it can take several days before an Ab has accumulated at its target and is excreted from the rest of the body. As such, Abs have to be labeled with long-lived radioisotopes (e.g., zirconium-89 or iodine-124). These radionuclides typically result in relatively high patient radiation burden [42,43]. A strategy to circumvent the use of long-lived radioisotopes is pretargeting. In this strategy, a two-step approach is used where a non-radioactive, tagged Ab is administered first and allowed to accumulate at its target and at the same time eliminated from the blood. In a second step that is usually carried out a couple of days later, a radiolabeled molecule-often a small molecule with rapid clearance from blood-is administered. Subsequently, this radiolabeled molecule reacts rapidly and selectively in vivo with the tagged Ab ( Figure 1B) [44,45]. Thus, pretargeting allows for the use of short-lived isotopes such as fluorine-18 [46]. The most attractive reaction for pretargeting is tetrazine ligation [45]. This ligation is highly selective and ultra-fast and is based on the reaction between a tetrazine (Tz) and a trans-cyclooctene (TCO). Successful pretargeting in the periphery was shown by Rossin et al. for the first time using a 111 In-labeled Tz in combination with a TCO-modified CC49 mAb [47]. The in vivo performance of the Tz ligation depends on the properties of the Tz as well as on those of the TCO. Suitable Tzs should be stable, highly reactive, clear rapidly from the blood stream, and easy to radiolabel. We have recently developed robust methods to radiolabel highly reactive Tzs with fluorine-18 as it is the most relevant radionuclide for clinical PET studies [48][49][50][51]. For brain pretargeted imaging, the Tz must also enter the brain. Such Tzs have also been developed by us [52]. Usually, TCOs are conjugated to the Ab as this increases the TCO stability by reducing the isomerization rate to its unreactive cis-cyclooctene (CCO) form [45,53,54]. For pretargeting, the TCO concentration per Ab should be as high as possible without significantly interfering with the properties of the Ab. A higher TCO concentration increases the chances of a Tz to react in vivo with the Ab when administered in a second step. Finding the right balance between TCO-loading, stability, and reactivity is therefore crucial when preparing TCO-Ab conjugates [45,53,54].
In this study, we investigated whether the TCO-concentrations required for pretargeted imaging are feasible to reach within the brain using a BBB shuttle bispecific Ab. For this purpose, we modified the monospecific anti-amyloid beta (Aβ) Ab 3D6 and bispecific anti-Aβ anti-transferrin receptor (TfR) Ab 3D6scFv8D3 [55,56] with different concentrations of TCOs and quantified the minimum TCO concentration to achieve a suitable contrast between the high and low binding regions. The contrast was determined using pretargeted autoradiography on brain sections of an Alzheimer s disease mouse model. Additionally, biodistribution studies in wild-type mice were carried out to correlate how different TCO/Ab ratios alter the brain pharmacokinetics of our bispecific Ab. Figure 1C displays the objectives of this study.

Theoretical Considerations to Determine the Minimum Required TCO Concentration
As with any imaging approach, a reasonable contrast between target-rich and targetpoor (high and low binding) regions is needed. The contrast is dependent on the tracer itself-specific binding vs. non-displaceable binding-as well as on the maximum available concentration of the target. In pretargeted approaches, TCOs are the target of interest and as such, it is interesting to estimate the minimum TCO concentration required to distinguish the high from low binding regions. To estimate this minimum concentration, we assumed that the in vitro contrast determined by autoradiographic studies can be extrapolated to the in vivo situation when a Tz is used that is known to result in clear in vitro and in vivo contrast. 111 In-labeled 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA)-PEG 11 -tetrazine ( 111 In-Tz) has been successfully used for in vivo pretargeted imaging and has low non-specific binding due to its high hydrophilicity [47,57]. Therefore, we chose 111 In-Tz as a model Tz for this study. We decided to set the threshold for the high/low binding region ratio to be a minimum of 2. This value appeared reasonable as such a ratio can be visually distinguished and is representative of the performance of PET tracers used to diagnose neurodegenerative diseases in the clinic such as the beta-amyloid imaging tracer [ 11 C]PIB [58].
In the in vitro pretargeted autoradiography setup, the minimum TCO concentration (TCO min ) required to distinguish between high and low binding regions can be expressed through the following parameters: the minimum amount of Abs (n Ab min ) applied to the tissue slice that provides a sufficient imaging contrast (high/low binding region ratio ≥2), the percentage of applied Abs that bind to the tissue (η b ), the number of reactive TCO moieties per Ab (N TCO ), and the volume of the tissue sample that was tested (v tiss ). The relationship between these parameters is shown in Equation (1). Once the parameters in the right part of the equation are determined experimentally, the TCO min can be calculated.
where TCO min = the minimum required TCO-concentration per gram brain tissue (nmol/g); n mAb min = the minimum amount of TCO-Abs that provides an in vitro image; η b = the percentage of Abs that become bound to the tissue (%); N TCO = the number of reactive TCOs per mAb; v tiss = the tissue volume (µL).
TCO min can also be converted into the Ab concentration (Ab min ) by dividing it by the TCO loading, as shown in Equation (2).
where Ab min = the minimum required Ab concentration per gram brain tissue (nmol/g); TCO min = the minimum required TCO-concentration per gram brain tissue (nmol/g); N TCO = the number of reactive TCOs per mAb. Ab min , determined from the in vitro experiments, can be compared with the Ab concentration that can be trafficked over the BBB in in vivo experiments (Ab traffic ). We expect pretargeted imaging over the BBB to be possible when Ab min is the minimum in the same order of magnitude as Ab traffic (Equation (3)).
Success criteria that the brain pretargeting works: Ab traffic = the Ab concentration that can be trafficked over the BBB (nmol/g); Ab min = the minimum required Ab concentration per gram brain tissue (nmol/g).

Antibody Modifications and TCO Quantification (N TCO )
TCO-PEG 4 -NHS was conjugated in a non-site-selective manner to the lysine residues of the Abs. Due to the high prevalence of lysines on proteins, this strategy allowed us to reach a wide range of TCOs conjugated to the Ab [43,59]. Different equiv. of TCOs-ranging from 10-1000 equiv.-were added to 3D6 and mAb3D6scFv8D3. The number of reactive TCOs was determined by titration. TCO-Abs was incubated with an excess of 111 In-Tz and then applied to an SDS-PAGE gel. Exposure of these SDS-PAGE gels to phosphor screens allowed for the quantification of the TCO moieties per Ab ( Figure 2A) (Supplementary Section S2). A maximum of approx. 19 TCOs/Ab could be reached ( Figure 2B). Interestingly, 500 equiv. of TCO-PEG 4 -NHS resulted in a similar TCO-loading compared to 1000 equiv.
However, the ELISA data showed a significant reduction in affinity to both Aβ and TfR when 1000 equiv. was used, and therefore this TCO-Ab was excluded for further examination ( Figure 2C and Supplementary Section S9). TCO-3D6scFv8D3 up until 500 equiv (giving approx. 15 TCO/Ab) was used for further evaluation. The affinity reduction can be explained by the non-site-specific conjugation to the lysine residues [9]. Lysine residues were also present in the complementarity-determining region (CDR), meaning that alterations to this binding domain can occur when adding TCOs. As previously mentioned, higher TCO-load reduced the affinity of the Ab construct toward TfR and therefore, we also determined the influence of the TCO load on the brain accumulation of this construct (biodistribution section).

Pretargeted Autoradiography to Determine the Minimum Amount of Abs (n mAb min ) Required for Sufficient Imaging Contrast
Autoradiography is an imaging technique that allows for the visualization of radioligands on tissues of interest and thus to determine the imaging contrast of novel CNS radioligands, for example, on brain sections [60,61]. In order to determine the imaging contrast that is feasible to reach with pretargeting, we performed pretargeted autoradiography experiments according to our previously published pretargeted autoradiography methodology [52]. Brain sections from Tg-ArcSwe mice were used for these studies. Tg-ArcSwe mice possess high Aβ pathology in the cortex (cor, high binding region), low pathology in the cerebellum (cer, low binding region), and have successfully been used for PET Aβ imaging with [ 124 I]3D6scFv8D3 [62]. Pretargeted autoradiography is a two-step imaging approach and simulates in vivo pretargeting as close as possible ( Figure 3A). Brain sections were first incubated overnight with TCO-3D6 or TCO-3D6scFv8D3. The next day, the excess of mAbs was washed away. In the second step, the brain sections were incubated with 111 In-Tz for one hour. This timeframe was sufficient to achieve complete ligation of the available TCOs as 111 In-Tz is a bispyridyl tetrazine with an expected reaction timeframe within minutes [49]. Afterward, the excess 111 In-Tz was washed away and the sections were quantified and analyzed using the uptake values in the cortex and in cerebellum.
This methodology allowed for the determination of the imaging contrast using TCO-Abs with different TCO loadings at different concentrations. Images with TCO-3D6 and TCO-3D6scFv8D3 revealed that an increased imaging contrast was achieved by increasing the TCO-Ab concentration as well as by increasing the number of TCOs on the Ab ( Figure 3B). An increased imaging contrast of 1.5-fold could be observed when increasing the absolute TCO number per Ab from 7 to 15 for both Abs. To assess the dependence of the imaging contrast on Ab concentration, TCO-3D6 and TCO-3D6scFv8D3 were incubated at different concentrations, ranging from 0 to 0.20 µg/mL (Supplementary Section S10). The cor/cer ratios were determined. A ratio >2 was observed for both TCO-Abs using 0.06 µg/mL. At this concentration, a clear distinct uptake in the high and low binding regions was visually evident. A cor/cer ratio of approx. 1.3 was determined for 0.02 µg/mL. Consequently, only the 0.06 µg/mL concentration fulfilled our contrast requirement (i.e., a ratio of >2 with respect to the high-to-low binding region uptake). A total of 0.06 µg Ab translated to a n mAb min of 0.17 pmol TCO-3D6scFv8D3.

Determination of TCO-Abs (η b ) Bound to Tissue and Tissue Volume (v tiss )
The percentage of TCO-Ab bound to brain sections (η b ) was determined by gamma counting. TCO-Ab conjugate with 3.6 TCOs/Ab was ligated with 111 In-Tz. The resulting 111 In-labeled Ab was applied to the brain sections. Unbound 111 In-labeled Ab was washed off and the brain sections were wiped off the glass plate and measured by a gamma counter. A η b of 7% was determined.
Brain sections of 20 µm were used with four sections per slide. Based on the average brain sizes of the C57BL/6J mice, a total tissue volume (v tiss ) of approx. 4 µL was estimated per slide [63,64].

Minimum Required TCO Concentration In Vitro
Equation (1) was used to calculate the TCO min . n mAb min , η b , N TCO , and v tiss were determined/estimated as described in previous sections. A TCO min of 10.7 pmol/g of TCOs was determined. This amount is needed to reach the minimum high binding/low binding ratio of 2.

Biodistribution TCO-3D6scFv8D3-Determination of Ab traffic
To estimate which Ab concentration of TCO-3D6scFv8D3 can enter the brain in vivo (Ab traffic ), a biodistribution study in wild-type (WT) mice was conducted. We decided to use TCO-Ab conjugates containing approx. 7 TCO/Ab, as this ratio did not result in a significant decrease in the TfR affinity compared to the unmodified Ab (Supplementary Materials). TCO-Abs containing~7 TCO/Ab has also been successfully used for tumor pretargeting and as such, we believe that this TCO loading was a reasonable starting point [47]. TCO-3D6scFv8D3 containing 6.8 TCO/Ab was labeled with an 18 F-labeled Tz ( Figure 4A). An excess of TCO-Ab (1:9 Tz/TCO ratio) was used to avoid an additional purification step. Consequently, the protein losses and changes in the protein concentration could be minimized, and therefore, the TCO-Ab concentration in the solution was precisely known during the experiment. The WT mice with an average weight of 25 g were injected with 50 µg of Ab, corresponding to a dose of approx. 2 mg/kg. After two hours, the animals were euthanized, a total of 16 organs were collected, and the radioactivity amount was quantified. The percentage injected dose per gram tissue (%ID/g) was calculated ( Figure 4B). A brain uptake of 1.1% ID/g was detected, which agreed with the published data for the unmodified 3D6scFv8D3 [65]. Among the peripheral organs, the uptake was mainly observed in the lungs and the liver ( Figure 4C). An overview of the uptake in all organs can be found in Supplementary Section S11.

Is Pretargeting Feasible? Ab min vs. Ab traffic
We estimated that Ab traffic needs to be in the same order of magnitude or higher than Ab min to enable sufficient imaging contrast in pretargeted brain imaging (Equation (3)). Ab min can be calculated according to Equation (2). TCO min was determined to be 10.7 pmol/g and the TCO loading per Ab used determined that Ab traffic was 6.8. As this bispecific Ab can enter the brain with an ID/g of 1.1% and shows an affinity toward Aβ that is comparable to the unmodified Ab version ( Figure 2C), we believe that this TCO amount can be applied for in vivo pretargeted imaging. Consequently, Ab min was determined to be 1.6 pmol/g. Ab min = TCO min N TCO = 10.7 pmol/g 6.8 = 1.6 pmol/g An in vivo brain uptake of 1.1% ID/g using a dosage of 50 µg resulted in a total of 2.6 pmol/g TCO-Ab (Ab traffic ) that was able to enter the brain. This value was higher than the calculated theoretical value, and consequently, the amount of required Ab min and measured Ab traffic were in the same order of magnitude. These estimations indicate that pretargeted imaging beyond the BBB is feasible using the described setup (Table 1). Table 1. A summary of the results. Pretargeting beyond the BBB appears to be feasible as the minimum amount of TCOs required for pretargeted brain imaging can be delivered using the RMT brain delivery of TCO-3D6scFv8D3 (i.e., Ab traffic ≤ Ab min ). An average mouse weight of 25 g was used for the calculations.

.4 Yes
Positive values indicate that BBB pretargeting is possible. * However, the ratio was in the same order of magnitude, which indicates that pretargeting beyond the BBB might nevertheless be feasible.

pH Stability of TCOs Conjugated to 3D6scFv8D3
Substances that are transported via RMT into the brain were exposed to different pH values ranging from approx. 5 to 7.5 [66]. Since TCOs are acid-sensitive, we investigated the stability of TCOs conjugated to 3D6scFv8D3 at different pH values [67]. At all of the tested pH values, TCOs were stable and reactive to 111 In-Tz ( Figure 4D).

Materials
Unless otherwise stated, all reagents and solvents were purchased from commercial suppliers and used without further purification. All of the water used was ultrapure (>18.2 MΩ cm −1 ). Other solvents were of analytical or HPLC grade and used as received. [ 111 In]InCl 3 was purchased from Curium. Automated syntheses were performed in a Scansys Laboratorieteknik synthesis module housed in a hot cell. The analytical-HPLC system consists of a 170U UVD detector, a Scansys radiodetector, and a Dionex system connected to a P680A pump. The system was run by Chromeleon software.

General Information
The radiochemical conversion (RCC) was determined by analyzing an aliquot of the crude reaction mixture by radio-HPLC analysis, integrating the radioactive peaks of the chromatogram. Radiolabeled products were characterized by associating the UV-HPLC traces of the reference compounds with the radio-HPLC chromatogram of the radiotracer. The radiochemical yield (RCY) was determined using the initial activity at the beginning of the synthesis and that of the formulated product at the end of the synthesis (E.O.S.) and corrected for decay (d.c.). The molar activity (Am) was determined by integrating the area of the HPLC-UV absorbance peak of the radiolabeled product in the HPLC chromatogram. The radiochemical purity (RCP), radiochemical yield (RCY), and molar activity (Am) values are given as the mean values.
Radiochemical conversion (RCC) of all radiolabeled compounds was determined by analyzing a labeled aliquot of the reaction mixture by radio-HPLC and analyzed by integrating the radioactive peaks from the reaction solution. The products were characterized by comparing the radio-HPLC trace of the reaction mixtures with the HPLC UV traces of the authentic 19 F-reference samples, respectively. The radiochemical yield (RCY) was determined using the activity of [ 18 F]fluoride received from the cyclotron at the beginning of the synthesis and that of the formulated product at the end of the synthesis. The decomposition was corrected and decay corrected (d.c.). The molar activity (Am) was determined by integrating the area of the UV absorbance peak corresponding to the radiolabeled product on the HPLC chromatogram. This area was converted into a molar mass by comparison with an average of integrated areas (triplicates) of a known concentration for the corresponding reference compounds.
[ 111 In]3D6scFv8D3 was used for the in vitro experiments described in Section S6.

Pretargeted Autoradiography
Pretargeted autoradiography was performed according to the previously published procedure [52]. A schematic overview is provided in Section S5.
Frozen brains of old (~18 months of age; 25-40 g body weight) male C57BL/6 mice and the tg-ArcSwe model harboring the Arctic (AβPP E693G) and Swedish (AβPP KM670/671NL) mutations (provided by Uppsala University) were cut into halves along the sagittal symmetry plane. Each half was mounted, lateral side up, on the cryostat sample holder pretreated with the Tissue-Tek OCT Compound, and fixed by freezing in a cryostat. After fixing, brains were cut at −22 • C into sagittal slices 20 µm thick using a Leica microtome, and the slices were thaw-mounted on Superfrost (70 × 22 mm, Fischer) adhesive slides. Only slices containing both cortex (specific binding) and cerebellar regions (representing nonspecific binding) were used. The slices were allowed to dry and were then put into storage boxes and kept at −80 • C until they were used.
On the day of the experiment, the frozen slides with the mounted brain sections were allowed to come to room temperature for 30 min. Afterward, a PAP pen was applied around the sections and 600 µL 1% BSA in PBS (pH 7.4) with 0.05% Tween 20 was added and incubated at room temperature for 30 min in incubation boxes. Afterward, the BSA solution was removed and 600 µL of PBS (pH 7.4) with 0.05% Tween 20 was applied and incubated at room temperature for 5 min. The preincubation buffer was removed and 600 µL of TCO-3D6 or TCO-3D6scFv8D3 (0.06 µL/mL and 0.006 µL/mL) in PBS (pH 7.4) with 0.05% Tween 20 was applied and incubated overnight at room temperature in a moist incubation box. On the second day, the TCO-mAb solution was removed and the slides were washed in cold PBS (pH 7.4) (3 × 15 min) and cold demineralized water (1 × 30 s). Then, 600 µL 20 nM [ 111 In]2 solution was applied to all sections and incubated for 60 min. Afterward, the radioactive solution was removed and the slides were washed in cold PBS (pH 7.4) (3 × 5 min) and demineralized water (1 × 30 s) and dried by a flow of compressed air. A calibration curve with dilutions of [ 111 In]2-40 nM, 20 nM, 6.7 nM, 2.2 nM, 0.74 nM, 0.25 nM, and 0.12 nM was prepared. After drying the plates, the slides were exposed on phosphor storage screens for 1-12 h, depending on the amount of radioactivity on the slide. Afterward, the phosphor storage screens were read by a Cyclone Storage Phosphor System (Packard Instruments Co., Meriden, CT, USA). Quantification of the plate readings was conducted with Optiquant software (version 3.00, Packard Instruments Co.) and ImageJ by drawing regions of interest on the cortex and cerebellum manually. Data were processed in Excel and the cortex/cerebellum imaging contrast ratios were calculated.

Determination of TCO-Ab Bound to Tissue
Slides with brain sections from the tg-ArcSwe mice were prepared as described in Section S6. Frozen brain slides were allowed to acclimatize to room temperature for 30 min. A PAP pen was applied around the sections and 600 µL 1% BSA in PBS (pH 7.4) with 0.05% Tween 20 was added and incubated at room temperature for 30 min in incubation boxes. Afterward, the BSA solution was removed and 600 µL of PBS (pH 7.4) with 0.05% Tween 20 was applied and incubated at room temperature for 5 min. The preincubation buffer was removed and 600 µL of [ 111 In]3D6scFv8D3 (0.06 µL/mL and 0.006 µL/mL) in PBS (pH 7.4) with 0.05% Tween 20 was applied and incubated overnight at room temperature in a moist incubation box. [ 111 In]3D6scFv8D3 was prepared as described in Section S4. On the second day, the [ 111 In]3D6scFv8D3 solution was removed and the slides were washed in cold PBS (pH 7.4) (3 × 5 min) and cold demineralized water (1 × 30 s). After drying the plates, the brain tissue was wiped off with tissue paper and placed in γ-counting tubes and measured by counting with the γ-counter Cobra II Auto-Gamma Counter (Model D5005, Packard BioScience Company). Quantification of the γ-counts was conducted in Excel to determine the TCO-Abs bound to the tissue.

Animals
Ex vivo distribution experiments were performed with 12-week old male C57BL/6JRj mice purchased from Janvier Labs. Mice were maintained at Panum, AEM Unit 16.2, University of Copenhagen and arrived 4 weeks prior to the experiments for habituation in the new environment. Mice were housed together in individually ventilated cages in a specific pathogen-free, humidity and temperature-controlled facility (12 h light/dark cycle) with free access to water and chow. All procedures were conducted following the European Commission's Directive 2010/63/EU, FELASA, and ARRIVE guidelines for animal research and were approved by The Danish Council for Animal Ethics (license number 2020-15-0201-00581).

Animal Injection and Ex Vivo Biodistribution Procedure
For the ex vivo biodistribution of [ 18 F]3D6scFv8D3, mice (n = 3) were anaesthetized with the injectable anesthesia mix of ketamine (100 mg/kg) and dexmedetomidine (0.5 mg/kg) and kept in anesthesia during the whole experiment. To maintain a physiological body temperature, mice were kept on a warming pad. All animals were tested for pain response by toe pinching. 26G venous catheters (26G BD Neoflon™, VWR) were inserted in the lateral tail veins and fixed with tape. A total of 50 µg of 18 F-3D6scFv8D3 (equaling a volume of 120 µL and between 0.5 and 0.7 MBq per mouse) was injected followed by a saline flush of 80 µL. Two hours after antibody injection, mice were euthanized by an injection of a lethal dose of ketamine/dexmedetomidine (300 mg/kg and 1.5 mg/kg). Blood was collected via cardiac puncture and separated into a plasma and a blood cell fraction using EDTA tubes centrifuged at 4 • C at 2000× g for 10 min.
After blood collection, mice were perfused with ice-cold saline and the following tissues were collected: heart, lungs, bladder, liver, stomach, gall bladder, kidneys, bone, muscle, scull, spleen, right brain hemisphere, left cerebellum, and left cerebrum. All tissues were wet-weighed and counted on an automated γ-counter Cobra II Auto-Gamma Counter (Model D5005, Packard BioScience Company). The concentration of radioactivity in the blood and tissues was determined as percentage injected dose per gram (%ID/g). Data were expressed as mean ± SD (n = 3).

Conclusions
We investigated whether pretargeting beyond the BBB is feasible. Based on our theoretical considerations and experimental data, we believe that it is. The required TCO concentrations can most likely be achieved without altering the current application strategies. However, pretargeting beyond the BBB is not only dependent on the TCO concentration that can be reached within the brain, but also on the ability of the tetrazine to reach its target and be rapidly cleared from the brain tissue and the rest of the body. This is of special importance for Tzs labeled with fluorine-18 with a half-life of approximately 110 min, which we have developed previously [52]. Finally, the imaging contrast of pretargeted imaging is dependent on the efficient clearance of unbound TCO-modified antibody still circulating in the blood. Unbound TCO-Ab has more time to clear from the brain in pretargeted strategies compared to imaging with directly labeled antibodies, and, if needed, clearing agents can be used to reduce unbound TCO-modified Abs circulating in the blood. These agents can be based on tetrazine ligation, where columns loaded with Tzs can be connected to the blood stream to trap the unbound TCO-Ab prior to the administration of the radiolabeled Tz. The imaging contrast was increased 125-fold by removing the circulating antibody in a cancer case [68]. This increase in imaging contrast could be needed to image brain targets with low abundancy. In conclusion, we predict that the TCO concentration within the brain achievable by conventional single-bolus i.v. injection of TCO-Abs is high enough for pretargeting beyond the BBB. Previously, we have developed Tzs that can cross the BBB, can successfully ligate the deposited targets in vivo and is able to image these targets within the brain [52]. In light of this, our next steps are directed toward the study of the possibility of using TCO-3D6scFv8D3 for pretargeted imaging in the Tg-ArcSwe mice in vivo. Additionally, brain uptake with TCO-Abs containing more than 6.8 TCO/Ab could be of interest to test, since this could increase the imaging contrast and lower the required dose. We believe that pretargeting beyond the BBB with antibodies will allow for the imaging of currently 'undruggable' targets and would therefore be crucial to monitor the therapies for currently untreatable neurodegenerative diseases.