In Vitro Evaluation of the Squaramide-Conjugated Fibroblast Activation Protein Inhibitor-Based Agents AAZTA5.SA.FAPi and DOTA.SA.FAPi

Recently, the first squaramide-(SA) containing FAP inhibitor-derived radiotracers were introduced. DATA5m.SA.FAPi and DOTA.SA.FAPi with their non-radioactive complexes showed high affinity and selectivity for FAP. After a successful preclinical study with [68Ga]Ga-DOTA.SA.FAPi, the first patient studies were realized for both compounds. Here, we present a new squaramide-containing compound targeting FAP, based on the AAZTA5 chelator 1,4-bis-(carboxylmethyl)-6-[bis-(carboxymethyl)-amino-6-pentanoic-acid]-perhydro-1,4-diazepine. For this molecule (AAZTA5.SA.FAPi), complexation with radionuclides such as gallium-68, scandium-44, and lutetium-177 was investigated, and the in vitro properties of the complexes were characterized and compared with those of DOTA.SA.FAPi. AAZTA5.SA.FAPi and its derivatives labelled with non-radioactive isotopes demonstrated similar excellent inhibitory potencies compared to the previously published SA.FAPi ligands, i.e., sub-nanomolar IC50 values for FAP and high selectivity indices over the serine proteases PREP and DPPs. Labeling with all three radiometals was easier and faster with AAZTA5.SA.FAPi compared to the corresponding DOTA analogue at ambient temperature. Especially, scandium-44 labeling with the AAZTA derivative resulted in higher specific activities. Both DOTA.SA.FAPi and AAZTA5.SA.FAPi showed sufficiently high stability in different media. Therefore, these FAP inhibitor agents could be promising for theranostic approaches targeting FAP.


Introduction
Fibroblast activation protein (FAP) is a post-prolyl proteolytic enzyme that belongs to the S9 family of serine proteases [1]. In addition to FAP, this S9 family includes other proline-specific serine proteases, such as prolyl oligopeptidase (PREP) and the dipeptidyl peptidases 4, 8, and 9 (DPP4, DPP8, and DPP9). Targeting fibroblast activation protein (FAP), overexpressed selectively in cancer-associated fibroblasts (CAFs), has recently become an attractive goal for diagnostic imaging and first therapeutic trials. FAP is involved in the promotion and development of tumor growth and is typically overexpressed in activated fibroblasts in the tumor stroma, whereas it is absent in most normal healthy tissues. Furthermore, FAP is overexpressed in several pathological tissue sites that are characterized by active remodeling [2][3][4][5]. Expression of FAP is found in CAFs in approximately 90% of epithelial carcinomas such as breast, pancreatic, colon, and prostate tumors [6][7][8]. These properties make FAP a very interesting and universally applicable tumor target for a variety of tumor types.

In Vitro Inhibition Measurements
The IC50 values for FAP, PREP, and the DPPs of the hybrid chelator conjugate AAZTA 5

In Vitro Inhibition Measurements
The IC 50 values for FAP, PREP, and the DPPs of the hybrid chelator conjugate AAZTA 5 .SA.FAPi compared to those of DOTA.SA.FAPi are shown in Table 1. The IC 50 values of AAZTA 5 .SA.FAPi as well as those of its non-radioactive complexes [ nat Sc]Sc-AAZTA 5 .SA.FAPi and [ nat Lu]Lu-AAZTA 5 .SA.FAPi for FAP appeared to be in the low nanomolar range (0.55-0.57 nM), whereas the IC 50 values for PREP resulted in the low micromolar range (2.4-3.6 µM). Screening against DPP4 and DPP9 for both SA.FAPi complexes revealed that the remaining activity was more than 50% at a final concentration of 1 µM. Hence, the IC 50 values for the DPPs were reported as >1 µM. The absence of a basic amine in the FAP inhibitor is known to result in an enormous increase of selectivity for the target molecule FAP, whereas the affinity for the DPPs can be drastically reduced. [12,46]. The IC 50 values for FAP and PREP were in the same order of magnitude of those for the previously reported SA.FAPi compounds, i.e., indicating high inhibition potency and excellent FAP-to-PREP selectivity indices. In addition, high selectivity towards DPP4 and DPP9 was achieved.

Radiolabeling and In Vitro Stability in Complexwith Gallium-68, Scandium-44, and Lutetium-177
Gallium-68: DOTA.SA.FAPi complexed with gallium-68 showed very high kinetics in quantitative radiochemical yields (RCYs) in our previous work [24]. Gallium labeling of AAZTA 5    Scandium-44: AAZTA 5 .SA.FAPi demonstrated excellent complexation with scandium-44 even at RT. We tested 5-20 nmol of precursor, which resulted in quantitative labeling already after 5 min for all amounts (Figure 4a, Figure S5). Stability was tested in HS, phosphate-buffered saline (PBS), and NaCl at 37 • C, demonstrating in highly satisfactory       Stability studies of both conjugates were performed in HS, PBS, and saline over a period of 10 days at 37 °C. In PBS and NaCl, very high stability values could be achieved for [ 177 Lu]Lu-AAZTA 5 .SA.FAPi, with >99% after 2 d, >98% after 3 d, and >95% intact conjugates after 10 days. In HS, the 177 Lu-AAZTA complex showed >99% of stability after 1 h, >98% after 3 h, and >96% after 6 h. However, the stability decreased significantly with time. After 1 d, the remaining stability of [ 177 Lu]Lu-AAZTA 5 .SA.FAPi in HS was >83%, after 2 d it was >64%, and after 3 d it was >55% (Figure 7a). Nevertheless, the stability of [ 177 Lu]Lu-AAZTA 5 .SA.FAPi in HS was satisfactory, with >95% intact conjugate after 6 h. If it is assumed that small molecules accumulate in the target tissue within the first few hours, and therefore their stability in HS over a long period is not relevant. [ 177 Lu]Lu-DOTA.SA.FAPi showed very high stability, with >99% of intact conjugate in HS within the measured time period of 10 days. In PBS and NaCl, the stability was high, i.e., >98% after 3 d and still >93% after 10 d (Figure 7b). Stability studies of both conjugates were performed in HS, PBS, and saline over a period of 10 days at 37 • C. In PBS and NaCl, very high stability values could be achieved for [ 177 Lu]Lu-AAZTA 5 .SA.FAPi, with >99% after 2 d, >98% after 3 d, and >95% intact conjugates after 10 days. In HS, the 177 Lu-AAZTA complex showed >99% of stability after 1 h, >98% after 3 h, and >96% after 6 h. However, the stability decreased significantly with time. After 1 d, the remaining stability of [ 177 Lu]Lu-AAZTA 5 .SA.FAPi in HS was >83%, after 2 d it was >64%, and after 3 d it was >55% (Figure 7a). Nevertheless, the stability of [ 177 Lu]Lu-AAZTA 5 .SA.FAPi in HS was satisfactory, with >95% intact conjugate after 6 h. If it is assumed that small molecules accumulate in the target tissue within the first few hours, and therefore their stability in HS over a long period is not relevant. [ 177 Lu]Lu-DOTA.SA.FAPi showed very high stability, with >99% of intact conjugate in HS within the measured time period of 10 days. In PBS and NaCl, the stability was high, i.e., >98% after 3 d and still >93% after 10 d (Figure 7b).

Lipophilicity Measurements
Lipophilicity (logD value) was determined via the "shake-flask" method. For both precursors AAZTA 5 .SA.FAPi and DOTA.SA.FAPi logD (pH = 7.4), values were measured for the 68   Sc complexes. Table 2 shows the logD values for the respective radiotracers.  5 .SA.FAPi showed almost identical logD values of −2.68 and −2.53, respectively. The carboxyl groups and the ionic bonds between chelator and radiometal favor the hydrophilic character of these radiotracers. The logD value of FAPI-04 is reported in the literature as −2.4 ± 0.28, confirming the hydrophilic character of 68 Ga-DOTA complexes [22].

Lipophilicity Measurements
Lipophilicity (logD value) was determined via the "shake-flask" method. For both precursors AAZTA 5 .SA.FAPi and DOTA.SA.FAPi logD (pH = 7.4), values were measured for the 68   Sc complexes. Table 2 shows the logD values for the respective radiotracers. The lipophilicity of the radiolabeled compounds [ 68

Inhibition Assays
Enzymes: Recombinant human FAP and PREP were expressed and purified as published [24]. Recombinant human dipeptidyl 9 (DPP9) was purified as described by De Decker et al. [46]. Human dipeptidyl peptidase 4 was purified from seminal plasma as published [47].
IC 50 measurements and counter-screening: IC 50 -measurements of the probes for FAP and PREP were carried out as published, using, respectively, Z-Gly-Pro-AMC and Suc-Gly-Pro-AMC as the substrate [24]. IC 50 experiments were repeated in triplicate, and the results are presented as mean ± standard deviation. Methods and data fitting were performed as published earlier [24]. Screening against DPP4 and DPP9 was performed at final probe concentrations of 10 µM and 1 µM using Ala-Pro-paranitroanilide (pNA) as the substrate at the respective final concentrations of 25 µM (DPP4) and 150 µM (DPP9) at pH 7.4 (0.05 M HEPES-NaOH buffer with 0.1% Tween-20, 0.1 mg/mL BSA, and 150 mM NaCl). Probes were pre-incubated with the respective enzyme for 15 min at 37 • C; afterwards, the substrate was added, and the velocities of pNA release were measured kinetically at 405 nm for at least 10 min at 37 • C. Measurements were executed using the Infinite 200 (Tecan Group Ltd., Mennendorf, Switzerland), and the Magellan software was used to process the data. If the remaining activity was more than 50% at 1 µM, the IC 50 values for the DPPs were reported as >1 µM.

Radiolabeling and Stability Measurements
Gallium-68: 68 Ge/ 68 Ga generators (ITG Garching, Germany) were used with ethanolbased post-processing evaluated by Eppard et al. [48]. Elution of gallium-68 was performed with 0.05 M HCl trapped on a micro-chromatography CEX column AG 50W-X4. The column was washed with 80% EtOH/0.15 M HCl, and 68   SA.FAPi was determined using the "shake-flask" methodology. After reaction of the precursor with the respective radionuclide, the reaction solution was adjusted to pH 7.4 with NaOH. Aliquots of~5 MBq for the 68 Ga complexes and of~3 MBq for the 44 Sccomplexed were taken and adjusted to a total volume of 700 µL with PBS (n= 4). 700 µL 1-octanol was added, and the solution was shaken for 2 min (1500 rpm). Afterwards, each tube was centrifuged for 2 min. 400 µL of the octanol-and PBS phases were pipetted in new tubes, and aliquots of each phase (3 µL of the PBS phase and 6 µL of the octanol phase) were measured via radio-TLC. The PBS phases were adjusted to 700 µL, and 700 µL octanol was added to each tube. The procedure was repeated twice. LogD values were calculated as the logarithm of the octanol/PBS ratio.

Informed Consent Statement: Not applicable.
Data Availability Statement: The study did not report any data.