*2.1. Delivery of Hydrogel-Encapsulated Anti-CTLA-4 by Peri-Tumor Injection Preferentially Targets TDLN*

A whole-animal optical imaging technique was adapted to investigate whether a peritumor injection of hydrogel-encapsulated anti-CTLA-4 creates a subcutaneous (SQ) mAb reservoir that preferentially targets regional TDLN. A near-infrared anti-CTLA-4/DyLight 800 conjugate was developed in order to allow both in vivo live imaging and ex vivo quantitation of mAb trafficking. When tumor-bearing mice were treated with 100 µg of anti-CTLA-4/DyLight 800 by intraperitoneal (IP) injection, the labeled mAb rapidly dispersed throughout the peritoneal cavity as expected (Figure 1a). In contrast, a peri-tumor SQ injection of low-dose hydrogel-encapsulated anti-CTLA-4/DyLight 800 (25 µg) resulted in a localized reservoir of mAb at the site of injection. These primary distribution patterns remained relatively constant when serially imaged over the course of 48 h. Expression of the label within lymph nodes was not observed in live animals, likely due to the overwhelming intensity of fluorescence from the primary injection and the deeper location of lymph nodes. In order to improve the sensitivity and specificity for detecting the micro-distribution of anti-CTLA-4 DyLight 800, sets of animals that had been treated in an identical manner were sacrificed at 24 and 48 h after injection. Bilateral axillary and inguinal lymph nodes (LN) were surgically excised and simultaneously examined by ex vivo optical imaging (Figure 1b). No fluorescent signal was detected at 24 h in any LN regardless of whether the animals received IP or SQ (hydrogel-based) dosing with anti-CTLA-4/DyLight 800. However, at 48 h, a fluorescent signal was always observed in axillary TDLN from animals that had received low-dose hydrogel-encapsulated anti-CTLA-4/DyLight 800. No signal was detected in animals that had received IP dosing even though they had received fourtimes the mAb dose. In addition, no fluorescent signals were detected from inguinal or contralateral LN in either group. These findings directly support our primary hypothesis that encapsulating anti-CTLA-4 within the hydrogel matrix and delivering it by peri-tumor injection produces a localized reservoir of mAb that preferentially targets TDLN. Minimizing systemic exposure to anti-CTLA-4 is particularly important due to the role of CTLA-4 in autoimmune-related toxicity [16]. As demonstrated previously, high-dose systemic administration of anti-CTLA-4 results in high serum concentrations that trigger/enhance the production of autoimmune antibodies [13]. These results also explain why targeted low-dose anti-CTLA-4, as delivered by the hydrogel, can produce equal/greater anti-tumor efficacy than does systemic dosing [13].

**Figure 1.** Biodistribution of fluorescent-labeled anti-CTLA-4 to tumor draining lymph nodes (TDLN) following either systemic (intraperitoneal; IP) or hydrogel (subcutaneous; SQ) injection. C57BL/6 mice bearing palpable MC-38 tumors implanted in the right posterior flank were treated with either 100 µg of anti-CTLA-4/DyLight 800 by IP injection or 25 µg of hydrogel-encapsulated anti-CTLA-4/DyLight 800 by peri-tumor SQ injection. Biodistribution of the injected anti-CTLA-4 was determined (**a**) immediately following administration by whole-animal in vivo optical fluorescence imaging and (**b**) at 24 and 48 h by ex vivo optical fluorescence imaging performed on surgically resected axillary TDLN and inguinal LN. **Figure 1.** Biodistribution of fluorescent-labeled anti-CTLA-4 to tumor draining lymph nodes (TDLN) following either systemic (intraperitoneal; IP) or hydrogel (subcutaneous; SQ) injection. C57BL/6 mice bearing palpable MC-38 tumors implanted in the right posterior flank were treated with either 100 µg of anti-CTLA-4/DyLight 800 by IP injection or 25 µg of hydrogel-encapsulated anti-CTLA-4/DyLight 800 by peri-tumor SQ injection. Biodistribution of the injected anti-CTLA-4 was determined (**a**) immediately following administration by whole-animal in vivo optical fluorescence imaging and (**b**) at 24 and 48 h by ex vivo optical fluorescence imaging performed on surgically resected axillary TDLN and inguinal LN.

results also explain why targeted low-dose anti-CTLA-4, as delivered by the hydrogel, can

produce equal/greater anti-tumor efficacy than does systemic dosing [13].

#### *2.2. Incorporation of HAse into the Hydrogel Matrix Enhances Antibody Delivery to TDLN 2.2. Incorporation of HAse into the Hydrogel Matrix Enhances Antibody Delivery to TDLN*

We had previously reported that incorporating HAse into the hydrogel mixture results in a number of beneficial effects. The reactive moieties on thiolated carboxymethyl hyaluronic acid (CMHA-S) and poly-(ethylene glycol)-diacrylate (PEG-DA), designed to promote spontaneous cross-linking and formation of the hydrogel matrix, can also interact with incorporated proteins with denaturing effects. The presence of HAse reduces this effect, likely by acting as an alternative protein target or impairing the interaction of these reactive moieties with incorporated proteins [11]. This protective effect from HAse was also observed for incorporated anti-CTLA-4. In an in vitro antibody release assay, total recovery of anti-CTLA-4 from the standard hydrogel preparation was 86.8% and this increased to 93.6% with addition of HAse 50 U and 97.8% with HAse 250 U (Figure S1). In addition, incorporated HAse breaks down the hyaluronic acid backbone of the matrix in a dose- and time-dependent manner. This can be used to fine-tune the rate of mAb release from the hydrogel matrix and assures complete release of encapsulated mAb payloads [14,17] (Figure S1). HAse also promotes more rapid resorption of the hydrogel matrix, promoting repeated injections at the same site for future cycles of immunotherapy. We had previously reported that incorporating HAse into the hydrogel mixture results in a number of beneficial effects. The reactive moieties on thiolated carboxymethyl hyaluronic acid (CMHA-S) and poly-(ethylene glycol)-diacrylate (PEG-DA), designed to promote spontaneous cross-linking and formation of the hydrogel matrix, can also interact with incorporated proteins with denaturing effects. The presence of HAse reduces this effect, likely by acting as an alternative protein target or impairing the interaction of these reactive moieties with incorporated proteins [11]. This protective effect from HAse was also observed for incorporated anti-CTLA-4. In an in vitro antibody release assay, total recovery of anti-CTLA-4 from the standard hydrogel preparation was 86.8% and this increased to 93.6% with addition of HAse 50 U and 97.8% with HAse 250 U (Figure S1). In addition, incorporated HAse breaks down the hyaluronic acid backbone of the matrix in a doseand time-dependent manner. This can be used to fine-tune the rate of mAb release from the hydrogel matrix and assures complete release of encapsulated mAb payloads [14,17] (Figure S1). HAse also promotes more rapid resorption of the hydrogel matrix, promoting repeated injections at the same site for future cycles of immunotherapy.

HAse is clinically approved for patient use based on its ability to break down SQ tissue barriers and facilitate the access of fluids, medications, and mAb into tissue lymphatic pathways [15,18]. Existing data suggest that this promotes faster and more efficient trafficking to draining LN [19]. We therefore hypothesized that addition of HAse to the

hydrogel formulation could enhance targeting and exposure of TDLN to anti-CTLA-4. To test this hypothesis, tumor-bearing mice were injected with hydrogel-encapsulated anti-CTLA-4/DyLight 800 (50 µg) in the presence or absence of 250 U HAse and then imaged for trafficking of the fluorescent label (Figure 2). In vivo live imaging shows similar localization and concentration of the fluorescent label in both groups at the time of injection. However, from 24 through 72 h, there is a significant and time-dependent reduction in the anti-CTLA-4/DyLight 800 signal at the injection site when HAse was incorporated into the injected hydrogels. Our conclusion is that anti-CTLA-4/DyLight 800 is released faster and/or more effectively in vivo when HAse is present. Consistent with this, when implanted hydrogels were recovered by surgical excision at 2 weeks after injection (Figure 2b), the hydrogels containing HAse were dramatically smaller and appeared to be infiltrated by host cells, consistent with accelerated degradation. hydrogel formulation could enhance targeting and exposure of TDLN to anti-CTLA-4. To test this hypothesis, tumor-bearing mice were injected with hydrogel-encapsulated anti-CTLA-4/DyLight 800 (50 µg) in the presence or absence of 250 U HAse and then imaged for trafficking of the fluorescent label (Figure 2). In vivo live imaging shows similar localization and concentration of the fluorescent label in both groups at the time of injection. However, from 24 through 72 h, there is a significant and time-dependent reduction in the anti-CTLA-4/DyLight 800 signal at the injection site when HAse was incorporated into the injected hydrogels. Our conclusion is that anti-CTLA-4/DyLight 800 is released faster and/or more effectively in vivo when HAse is present. Consistent with this, when implanted hydrogels were recovered by surgical excision at 2 weeks after injection (Figure 2b), the hydrogels containing HAse were dramatically smaller and appeared to be infiltrated by host cells, consistent with accelerated degradation.

HAse is clinically approved for patient use based on its ability to break down SQ tissue barriers and facilitate the access of fluids, medications, and mAb into tissue lymphatic pathways [15,18]. Existing data suggest that this promotes faster and more efficient trafficking to draining LN [19]. We therefore hypothesized that addition of HAse to the

*Gels* **2022**, *8*, x FOR PEER REVIEW 4 of 9

**Figure 2.** Impact of hyaluronidase (HAse) on the delivery of hydrogel-encapsulated anti-CTLA-4 to TDLN. C57BL/6 mice bearing palpable MC-38 tumors were treated with 50 µg of hydrogel-encapsulated anti-CTLA-4/DyLight 800 by peri-tumor SQ injection. Anti-CTLA-4 biodistribution was compared in animals receiving a standard hydrogel formulation to one that incorporated HAse (250 U). (**a**) Whole-animal in vivo optical fluorescence imaging of the injection site immediately following injection (0 h) and at 24, 48, and 72 h. (**b**) Hydrogels from the same animals were surgically recovered after 2 weeks for visual inspection. (**c**) Axillary TDLN and ipsilateral inguinal LN were surgically resected at 24, 48, and 72 h from a cohort of animals treated in the same manner and subjected to ex vivo optical fluorescence imaging. **Figure 2.** Impact of hyaluronidase (HAse) on the delivery of hydrogel-encapsulated anti-CTLA-4 to TDLN. C57BL/6 mice bearing palpable MC-38 tumors were treated with 50 µg of hydrogelencapsulated anti-CTLA-4/DyLight 800 by peri-tumor SQ injection. Anti-CTLA-4 biodistribution was compared in animals receiving a standard hydrogel formulation to one that incorporated HAse (250 U). (**a**) Whole-animal in vivo optical fluorescence imaging of the injection site immediately following injection (0 h) and at 24, 48, and 72 h. (**b**) Hydrogels from the same animals were surgically recovered after 2 weeks for visual inspection. (**c**) Axillary TDLN and ipsilateral inguinal LN were surgically resected at 24, 48, and 72 h from a cohort of animals treated in the same manner and subjected to ex vivo optical fluorescence imaging.

The fundamental question addressed in the next study was whether this results in more effective perfusion of TDLN by anti-CTLA-4. As previously described, sets of tumorbearing animals that had been treated in an identical manner were sacrificed at 24, 48, and 72 h after injection of anti-CTLA-4/DyLight 800. Recovered axillary and inguinal LN were simultaneously examined for the presence of labeled anti-CTLA-4 by ex vivo optical imaging (Figure 2c). As hypothesized, a higher level of fluorescence was detected in axillary TDLN at 24 h when the hydrogel contained HAse. At 48 h, the fluorescence level was similar in both groups while the fluorescent signal was only detected at 72 h from animals that received hydrogel containing HAse. This visual trend was confirmed by measuring total fluorescent emission from LN recovered at these time points (Table S1). Ex vivo imaging of recovered inguinal LN showed a similar—but less intense—pattern in animals that received hydrogels containing HAse but no uptake by inguinal LN at any time point in the absence of HAse (Figure 2c). This interesting finding suggests that the increased tissue permeability resulting from HAse can spread anti-CTLA-4 to a wider tissue distribution. Given the size difference between mice and humans, it is not yet clear whether this would have a meaningful effect in a clinical setting. We did not detect anti-CTLA-4 uptake by other organs by whole-body live imaging, nor was a signal detected in isolated spleens. This likely reflects the imaging threshold of our approach and it is very possible that more sensitive approaches—such as immunohistochemistry or flow cytometry—might yield findings that are beyond the focus of our targeted studies.

Taken together, these results suggest that incorporating HAse into the hydrogel formulation enhances the release of functional anti-CTLA-4 from the hydrogel matrix; opens lymphatic barriers in surrounding SQ tissues; and produces faster, higher, and more prolonged binding to target sites in TDLN while still sparing mAb accumulation in distant (e.g., contralateral) LN.
