Stimuli‐Responsive Polymer Nanoprobes Intended for Fluorescence‐Guided Surgery of Malignant Head‐and‐Neck Tumors and Metastases

Nano‐sized carriers are widely studied as suitable candidates for the advanced delivery of various bioactive molecules such as drugs and diagnostics. Herein, the development of long‐circulating stimuli‐responsive polymer nanoprobes tailored for the fluorescently‐guided surgery of solid tumors is reported. Nanoprobes are designed as long‐circulating nanosystems preferably accumulated in solid tumors due to the Enhanced permeability and retention effect, so they act as a tumor microenvironment‐sensitive activatable diagnostic. This study designs polymer probes differing in the structure of the spacer between the polymer carrier and Cy7 by employing pH‐sensitive spacers, oligopeptide spacers susceptible to cathepsin B‐catalyzed enzymatic hydrolysis, and non‐degradable control spacer. Increased accumulation of the nanoprobes in the tumor tissue coupled with stimuli‐sensitive release behavior and subsequent activation of the fluorescent signal upon dye release facilitated favorable tumor‐to‐background ratio, a key feature for fluorescence‐guided surgery. The probes show excellent diagnostic potential for the surgical removal of intraperitoneal metastasis and orthotopic head and neck tumors with very high efficacy and accuracy. In addition, the combination of macroscopic resection followed by fluorescence‐guided surgery using developed probes enable the identification and resection of most of the CAL33 intraperitoneal metastases with total tumor burden reduced to 97.2%.


Introduction
Treatment of malignant tumors usually requires a multi-modality approach including surgery, chemotherapy, and radiation therapy.Indeed, surgical removal still represents the primary treatment modality for most solid tumors. [1]Most tumor cells should be removed during surgery as any remaining malignant cells could cause disease relapse but it is desirable to spare healthy tissues and to remove only the necessary margin from the immediate vicinity of the tumor. [2]Therefore, it is essential to accurately determine the boundary between healthy and tumor tissue to enable precise and safe resection.This could be achieved by using a fluorescent contrast agent that accumulates in the tumor to a greater extent than in the healthy tissue, thus increasing the signal-to-noise ratio between malignant and healthy tissue.To date, some lowmolecular-weight contrast agents for optical imaging, such as indocyanine green (ICG) and 5-aminolevulinic acid, have been approved for clinical use. [3,4]However, their tumor accumulation is not specific and their body half-life is very low, leading to an unsatisfactory target-to-background ratio.
It has been repeatedly reported that polymer conjugates possess several significant advantages in the controlled delivery of various low-molecular-weight molecules for the treatment of malignant or inflammatory diseases. [5]The nanosystems exhibit prolonged blood circulation, preferentially accumulating in the target tumor tissue due to the so-called enhanced permeability and retention (EPR) effect. [6]Consequently, the main advantage of polymer-bound anticancer drugs over traditional chemotherapy is the dramatic reduction in adverse side effects as due to the fact that the circulating drug is inactive while bound to the carrier and the accumulation within the tumorous tissue is much higher, thus, it does not damage healthy tissues and vital organs.[9] Similarly, fluorescent dyes are stabilized against biodegradation when bound to the polymer carrier. [10]Indeed, due to the proximity of the fluorophores located on the polymer chain, the fluorescence intensity of the polymer-bound dye is usually weaker than that of the free fluorophore [11,12] due to fluorescence quenching of the nearby fluorophore molecules bound to the same polymer chain.The absorbed light energy is not dissipated from the dye only in the form of via fluorescence but also through non-radiant processes.As a fluorescent contrast agent, either the conventional fluorescent dyes emitting in red visible and near-infrared region (NIR-I, 700-900 nm), but also some substances emitting in the second near-infrared region (NIR-II, 1000-1700 nm) can be used. [13]The advantage of using NIR-II region is its ability to penetrate deeper into tissue (up to 10 mm) compared to NIR-I, which only penetrates up to 3 mm.As many endoscopy systems used in medicine are still equipped with NIR-I cameras only, NIR-II remains a less used alternative.
In this work, we developed long-circulating stimuli-sensitive fluorescent diagnostic probes based on synthetic biocompatible water-soluble N-(2-hydroxypropyl)methacrylamide (HPMA)based polymers that preferentially accumulate in solid tumors due to the EPR effect.Moreover, a series of polymer probes differing in the spacer structure between the polymer carrier and the fluorescent dye were designed and synthesized by employing three pH-sensitive spacers cleavable under mildly acidic conditions, two oligopeptide spacers susceptible to cathepsin B-catalyzed enzymatic hydrolysis, and one control spacer nondegradable under physiological conditions.The stimuli-sensitive release behavior of developed polymer probes and the activation of the fluorescent signal after the dye release from the nanosystem was investigated and confirmed.Moreover, the determination of fluorescence intensity after nanoprobe application showed an increased tumor-to-normal tissue signal contrast, which is fundamental for successful fluorescence-guided surgery.

Results and Discussion
Image-guided surgery provides better control of the surgical procedure with accurate resection of tumor margins, as well as reducing the probability of incomplete tumor and less healthy tissue removal.Herein, we present novel tumor stimuli-sensitive near-infrared polymer probes intended for the fluorescenceguided surgery of solid tumors, especially head-and-neck carcinomas.The structure of the polymer probes was designed to ensure tumor accumulation and subsequent pH-sensitive or enzymatic activation of the fluorescent signal.The efficacy of the polymer probes was validated during the surgical removal of intraperitoneal metastases or orthotopic head and neck carcinomas.Figure 1

Synthesis of the Reactive Derivatives of Cy7
Fluorescent dyes were modified before conjugation to the polymer precursors to introduce the reactive carbonyl group required for subsequent reaction with the hydrazide groups of the polymer precursors to form pH-sensitive hydrazone bonds.Three different keto acids, namely 5-cyklohexyl-5-oxopentanoic acid (COP), 4-(2-oxopropyl)benzencarboxylic acid (OPB) and 4-oxo-4-(2-pyridyl)butanoic acid (PYR) were selected as differences in the structure of the surrounding of the carbonyl group should result in different stability of the corresponding hydrazone bond and consequently, different rates of hydrolytic release of the fluorophore from the polymer carrier.The keto derivatives COP-Cy7, OPB-Cy7, and PYR-Cy7 were prepared by a two-step reaction.First, activated derivatives of the keto acids were prepared by a carbodiimide coupling reaction with 2thiazoline-2-thiol (TT) in presence of 4-(dimethylamino)pyridin (DMAP) in THF as described previously. [11]In the second step, activated keto acids were reacted with amino group-containing Cy7.
The synthesis of derivative Az-VCit-PAB-Cy7 containing an enzymatically cleavable linker was based on the reaction of the corresponding reactive 4-nitrophenyl carbonate derivative of dipeptide valine-citrulline (VCit) with the amino derivative of Cy7.The purity and identity of the products were investigated by HPLC and LCQ Fleet mass analyzer with electrospray ionization (ESI-MS) confirming the synthesis of Cy7 derivatives with high purity and yield.

Synthesis of the Polymer Precursors and Polymer Probes
Here, HPMA-based copolymers were employed as they were recognized as suitable polymer carriers for advanced delivery of various bioactive low-molecular-weight compounds. [14][17] Here, polymer precursors were prepared by controlled radical reversible addition-fragmentation chain-transfer (RAFT) polymerization ensuring the low dispersity (≈1.1) of the final polymers, indicating the controlled mechanism of the polymerization.The weight-average molecular weight of all polymer precursors was set in the range of 28-37 kg mol −1 , thus ensuring the elimination of the polymers after fulfilling their role as the carrier. [18]All the polymer precursors contained enough reactive groups required for the following reactions.The DBCO groups intended for the metal-free click chemistry were introduced into the polymer precursor P-A-TT by the amidic reac-tion with 3-amino-1-(11,12-didehydrodibenzo[b,f ]azocin-5(6H)yl)propan-1-one (DBCO-amine); no significant changes in molecular weight and dispersity of the P-DBCO polymer precursor after modification were observed (see Table 1).Importantly, all polymer precursors had a hydrodynamic diameter in the range of 6-9 nm guaranteeing prolonged blood circulation and subsequent elimination of the polymer carrier via glomerular filtration once the role of the carrier is over.The keto derivatives of the fluorescent dye Cy7 (COP-Cy7, OPB-Cy7 and PYR-Cy7) were attached to polymer precursors containing hydrazide groups via a hydrolytically labile hydrazone bond yielding polymer probes P-H-COP-Cy7, P-H-OPB-Cy7, and P-H-PYR-Cy7 as shown in Figure 2. The synthesis was performed in methanol with acetic acid according to the procedure published previously. [11]The attachment of the dye derivative did not significantly change the molecular weight and dispersity, see Figure S1 (Supporting Information).In addition, the hydrodynamic radii of prepared labeled probes remain like ones of precursors being in the range of 8.3-10.0nm.The polymer probes contained 0.7 to 1.7 wt.% dye, which is sufficient for the given surgical application.
The polymer probe P-GFLG-Cy7 containing an oligopeptide spacer degradable by lysosomal enzymes, and non-degradable polymer probe P-A-Cy7 with an amide bond between the polymer carrier and the fluorophore were prepared via the aminolytic reaction of the corresponding polymer precursors containing reactive TT groups (P-GFLG-TT and P-A-TT) with Cy7-amine in the presence of ethyldiisopropylamine (DIPEA).The polymer probe P-VCit-Cy7 containing a valine-citrulline spacer susceptible to cathepsin B-catalyzed hydrolysis was prepared via strainpromoted cycloaddition of the azide derivative Az-VCit-PAB-Cy7 to the polymer precursor P-DBCO bearing DBCO groups.The attachment of dye did not affect the dispersity or molecular weight of the resulting polymer probes, see Figure S2 (Supporting Information).The hydrodynamic size for enzymatically degradable probes varied from 8.8 to 9.4 nm.Thus, all the polymer probes showed the size, which allows prolonged circulation with subsequent body elimination via urine.Importantly, the amount of the dye bound to the enzymatically degradable and non-degradable polymer probes was higher than that obtained for the pH-sensitive polymer probes and thus also sufficient for the following experiments.Moreover, the fluorescence intensity of the bound Cy7 was significantly lowered for all the polymer probes due to the quenching caused by the interaction of nearby fluorophores in the single polymer coil.The structures of the polymer probes are shown in Figure 3.

In Vitro Release of Free Cy7 Dye from the Polymer Probes
Since the primary goal of this study was to compare fluorescent probes with various mechanisms and rates of fluorophore release, the fluorophore release from the polymer probes was monitored by HPLC equipped with UV-vis and fluorescent detectors.First, we evaluated the dye derivative release rates from the polymer probes with pH-sensitive hydrazone spacers in aqueous media at different pH.While the polymer probes P-H-OPB-Cy7 and P-H-PYR-Cy7 containing an acid-labile hydrazone bond between the dye derivative and the polymer carrier released ≈50% of Cy7 at pH 6.5 within 10 h, 95% of the dye was hydrolyzed from the analogous polymer conjugate P-H-COP-Cy7 under the same conditions (Figure 4A).Importantly, no significant release of Cy7 was observed for all the polymer probes over 10 h at physiological pH 7.4 (Figure 4B).
Polymer probes with the enzymatically cleavable fluorophore P-GFLG-Cy7 and P-VCit-Cy7 were incubated (in phosphate buffer at pH 6.0) with cathepsin B, a representative lysosomal protease.The polymer probe P-GFLG-Cy7 released ≈86% of the dye after 48 h, then the enzymatic reaction rate gradually slowed down due to the decreasing activity of cathepsin B. In contrast, the polymer probe P-VCit-Cy7 rapidly released ≈60% of Cy7 in 5 h reaching a plateau, with very slow further fluorophore release until the end of the experiment at 48 h (Figure 4C) where it reached almost 69%.It is hypothesized that the observed kinetics are related to the formation of 4-aminobenzyl alcohol derivative of the fluorophore (or drug) that is released by lysosomal enzymes and then spontaneous 1,6-elimination of 4-aminobenzyl alcohol leads to the formation of the original fluorophore and carbon dioxide and 4-aminobenzyl alcohol may inhibit the active site of the enzyme resulting in the observed phenomenon.The hypothesis is in accordance with literature [19,20] describing the inhibition of proteases with similar aminobenzyl derivatives.The major advantage of this self-immolating VCit spacer compared to similar enzymatically triggered systems is the increased enzymatic release of the fluorophore as the VCit spacer is a better substrate for cathepsin B. [21] This behavior was confirmed by the release experiment shown in Figure 4C.It was hypothesized that this might result in a higher release of the fluorophore in the tumor tissue followed by an increase in the fluorescence and, consequently, better contrast between the healthy and malignant tissue during fluorescence-guided surgery.Importantly, no release of the Cy7 dye was observed for both P-GFLG-Cy7 and P-VCit-Cy7 incubated in plasma and for the non-degradable P-A-Cy7 polymer probe (data not shown).
All degradable polymer probes exhibited stimuli-sensitive behavior of Cy7 release.The pH-sensitive probes were stable in conditions mimicking the bloodstream, while the dye was progres-sively released at lower pH simulating the extracellular space of the tumor tissue.Similarly, the enzymatically degradable probes were tumor-tissue-responsive nanoprobes, as they released significant amounts of the fluorescent dye after incubation with cathepsin B. Thus, all the synthesized polymer probes showed suitable stimuli-sensitive behavior, so could be used in the surgical experiment.
The Cy7 fluorophores attached to the polymer chains were obstructed in the quantum yield of the fluorescence as the nearby fluorophores tend to dissipate accepted energy from the light irradiation in the form of non-radiant processes.Indeed, such quenching could be advantageous to decrease the fluorescence signal of the polymer probe from the healthy tissue.Once the tumor stimuli-sensitive probe is accumulated in the tumor, the dye is released, and the fluorescence is fully activated.We have proved such behavior for all the developed stimuli-sensitive probes.The pH-sensitive probes showed a sharp increase in fluorescence on incubation at low pH ensuring dye release.Figure 4D shows more than a 3-fold increase in fluorescence when the P-H-OPB-Cy7 probe is incubated at low pH.Similar behavior under acidic conditions was observed for P-H-COP-Cy7 and P-H-PYR-Cy7 probes.P-H-OPB-Cy7 showed a relative fluorescence increase of 3.6-fold with maximal fluorescence intensity of 566 RFU, while P-H-COP-Cy7 showed a relative fluorescence increase of only 2fold, therefore the polymer probe H-COP-Cy7 was not used in the following in vitro experiments.Although P-H-PYR-Cy7 showed the highest relative fluorescence increase of 5.8-fold, the overall maximum intensity was only 47.2 RFU, see Table 2.
The increase in free Cy7 fluorescence of the enzymatically cleavable probes could not be measured immediately as in the case of hydrolytically cleavable probes after the addition of acid   because the enzyme needs some time for the cleavage reaction.Therefore, we attempted to measure the increase in fluorescence of the respective polymer probes after incubation with cathepsin B in phosphate buffer (pH = 6.0, 0.001 EDTA) and compare the initial and maximal fluorescence.Unfortunately, we were unable to measure the fluorescence increase of the free Cy7 released from polymer probes with enzymatically cleavable fluorophore P-GFLG-Cy7 and P-VCit-Cy7 using a fluorometer or by HPLC as the released free dye has very low water-solubility and precipitated from the buffer in comparison with the more water-soluble Cy7derivatives released from pH-sensitive probes.The HPLC release measurement showed a significant decrease in polymer-bound Cy7 but there was no increase in free Cy7 fluorescence due to its insolubility.On the other hand, the following in vitro experiment showed the increase of fluorescence in dynamic model after 24 h incubation with cells.

Internalization-Mediated Activation of Fluorescent Polymer Probes
Stimuli-mediated activation of the fluorescent probes was studied in vitro using laser scanning confocal microscopy (LSCM).
The squamous cell carcinoma cells (FaDu) were incubated with fluorescently labeled polymer probes at an equal concentration of fluorophore Cy7 (5 μg mL −1 ) and the potential increase in fluorescence under intracellular conditions (specifically lower pH or presence of lysosomal enzymes) of stimuli-sensitive polymer probes was investigated at 37 °C over time.
The polymer probe P-A-Cy7 with a non-cleavable spacer was used as a control.Cy7 is shown in the "fire" color scheme for easily observable concentration differences and cell nuclei were labeled with Hoechst 33 342 (shown in yellow in Figure 5).Unfortunately, it was not possible to distinguish between the polymerbound and released dye, therefore, the Cy7 in Figure 5 as "fire" corresponds to the total fluorescence intensity.After 6 h of incubation at 37 °C, the initial Cy7 fluorescence was slightly more pronounced in cells incubated with P-H-OPB-Cy7 and both enzymatically cleavable polymer probes, but the fluorescence intensity increased significantly after 24 h of incubation.Specifically, polymer probes P-H-OPB-Cy7 and P-VCit-Cy7 showed a dramatic increase in fluorescence, indicating intracellular-stimuli dye release followed by an increase in observed fluorescence.The polymer probe P-GFLG-Cy7 also showed a considerable increase in the fluorescent intensity when compared to the control polymer probe P-A-Cy7.Polymer probe P-H-PYR-Cy7 showed only a slight increase in fluorescence over time consistent with the lowest absolute fluorescence intensity shown in Table 2.
The Cy7 signal image analysis was applied to quantify the LSCM images as shown in Figure 6.The highest Cy7 signal at 6 h was observed for the P-H-OPB-Cy7 probe, which is consistent with the release rate of this probe.After 24 h, the highest fluorescent signal was observed for the probes P-H-OPB-Cy7 and P-VCit-Cy7, showing an 85-96-fold increase in fluorescence intensity compared to the free cells incubated without probes.The signal of the control polymer probe P-A-Cy7 did not significantly change over time and was less than one-sixth of the stimuli-sensitive probes.In summary, P-GFLG-Cy7 and P-VCit-Cy7 show excellent potential for the fluorescence labeling of tumor cells due to their stimuli-sensitive properties tailored for tumor cell-driven fluorescence activation.

Ex Vivo Biodistribution in Mice Bearing FaDu Head and Neck Carcinoma (HNc)
Ex vivo biodistribution was performed in animals bearing the subcutaneously implanted FaDu HNc carcinoma, showing remarkable tumor accumulation of all polymer probes (Figure 7A).A significant fluorescent signal was also visible in the kidneys indicating urinary excretion of the probe and released dye, and in the liver demonstrating partial hepatobiliary excretion.Importantly, the increased tumor-to-muscle ratio indicating increased tumor accumulation for all the probes was highest for P-H-OPB-Cy7 and P-VCit-Cy7 (Figure 7B).These two probes also showed the highest absolute fluorescence intensity in the tumor area, which was significantly higher than that obtained for the nondegradable probe P-A-Cy7.
Since the probes are identical in size, the increased fluorescence signal could be attributed to the activation of the signal within the tumor area.These results are also in concordance with in vitro FaDu cell results, where these two probes produced the highest signals after cell internalization.

Fluorescence-Guided Detection and Resection of Peritoneal Carcinomatosis
As the polymer probes accumulate extensively in tumor tissue, we then evaluated the benefit of fluorescence-guided surgery in the detection and resection of intraperitoneal metastases.Based on the in vitro internalization and ex vivo biodistribution results, two probes were selected for these studies: P-H-OPB-Cy7 as the most effective probe both in vitro and ex vivo, and P-A-Cy7 as a control.
Metastases of 1 mm and less in diameter were clearly visible and distinguishable from healthy tissue using the optical system (Figure 8A).The combination of macroscopic resection followed by fluorescence-guided surgery using P-H-OPB-Cy7 or P-A-Cy7 probes enabled the identification and resection of most of the CAL33 intraperitoneal metastases, as the total tumor burden reduced to 97.2% and 80.0%, respectively (Figure 8B).Furthermore, the P-H-OPB-Cy7 probe was more effective compared to the control P-A-Cy7 probe as it identified more residual metastases that were not visible under normal light (80.5% vs. 56%) and had a lower false negative rate (2.8% and 20%, respectively).
We further compared the efficacy of P-H-OPB-Cy7 or P-A-Cy7 probes for the detection of FaDu peritoneal carcinomatosis using the clinically approved Storz 4k ICG imaging system.The Red Fluorescent Protein (RFP) fluorescence positivity is the gold standard for the identification of tumor-positive tissue, see Figure 8C, simultaneously enabling the evaluation of the colocalization of Cy7 and RFP fluorescence.In the P-H-OPB-Cy7 group, 59 nodules were resected and of the 50 RFP-positive nodules, 45 were Cy7 positive, whereas 63 nodules were resected in the P-A-Cy7 group and 38 of the 52 RFP-positive nodules were Cy7 positive.Thus, the P-H-OPB-Cy7 probe showed higher sensitivity and positive predictive value for the identification of intraperitoneal tumors compared to P-A-Cy7 (Table 3).The Cy7 fluorescence colocalized well with the RFP fluorescence signal for both probes but the higher diffuse accumulation of the control P-A-Cy7 probe in healthy tissues made the orientation in the operation field and detection of malignant tissue more difficult compared to P-H-OPB-Cy7 probe.This is attributed to the favorable tumor-to-healthy tissue ratio of the P-H-OPB-Cy7 probe as shown in Figure 7B.The presence of tumor cells in the resected specimen was further confirmed by histopathological examination.Interestingly, lymph node tissue was identified in a high number of false positive specimens (Cy7 positive, RFP negative).

Fluorescence-Guided Surgery of Cal33-luc Orthotopic Tumor-Bearing Mice
Finally, we investigated the utilization of selected polymer probes for the guided surgery of orthotopic Cal33-luc HNc tumors.Mice were randomized into two groups that underwent surgery with or without optically guided surgery (OGS).As shown in Figure 9, 3 out of 9 mice operated without OGS had important local tumor reapportions after 3 weeks, whereas the local tumors reappearing in the 3 mice operated with OGS were barely detectable by luciferase at week 5, and reappeared after 10 weeks, indicating the advantage of the OGS in the surgery outcome.This highlights the advantage of optically guided surgery with a positive impact on reducing the amount of tumor cells left locally in the invaded margins.
The individual examples of tumor recurrence in mice are shown in Figure S3 (Supporting Information).The difference in the relapsed tumor size after 1, 3, 5, and 10 weeks after surgery shows that the removal of tumors with optically guided surgery using the polymer probe P-H-OPB-Cy7 significantly contributed  to the removal of most tumor cells or at least to prolong the time between tumor recurrence.

Conclusion
Here, novel stimuli-activated optical contrast agents suitable for high-resolution fluorescence-guided surgery of difficult-tooperate head and neck tumors were developed.Hydrolytically and enzymatically activatable polymer-based nanoprobes, stimuli-sensitive either to the acidic tumor microenvironment or to enzymes present in the endosomes of tumor cells, were compared, and evaluated to serve as nanoprobes with high tumorto-healthy tissue contrast for fluorescence-navigated surgery.The probes with enzymatically degradable ValCit-PAB spacer or hydrolytically degradable hydrazone bond showed the highest potential determined in vitro on the tumor cells.Those nanoprobes took advantage from fluorescent quenching occurred after nanoprobe synthesis and tumor milieu-associated
Methods: The synthesis and purity of the monomers and reaction manners were monitored by reversed-phase HPLC using Chromolith Performance RP-18e columns (100 × 4.6 mm, Merck, Germany) with a linear gradient of water-acetonitrile (0%-100% acetonitrile) in the presence of 0.1% TFA with a UV-vis diode array detector (Shimadzu, Japan).The same HPLC method was used with a linear gradient of water-acetonitrile (0%-100% acetonitrile) without TFA for the determination of hydrolytically released Cy7.
The molar mass of the monomers was determined using mass spectrometry performed on an LCQ Fleet mass analyzer with electrospray ionization (ESI-MS; Thermo Fisher Scientific, Inc., MA, USA).The determinations of the molecular weights and polydispersity of the copolymers and the enzymatic dye release from the polymer conjugates were performed by size-exclusion chromatography (SEC) on an HPLC system (Shimadzu, Japan) equipped with UV, differential refractive index (RI), and multi-angle light scattering (LS) detectors (Wyatt Technology Corp., USA)  penicillin-streptomycin (Gibco, Waltham, MA USA) in a humidified incubator at 37 °C with 5% CO 2 .The FaDu cell line expressing Red Fluorescent Protein (RFP) was a kind gift from Dr. Trzil (BIOPHARM, Research Institute of Biopharmacy and Veterinary Drugs, Liběchov, Czech Republic).
The CAL-33 cell line derived from a tongue carcinoma was a kind gift from the Antoine Lacassagne Cancer Center Oncopharmacology laboratory (Nice, France).Stable expression of the luciferase gene was obtained by lentiviral transfection, thus generating the Cal33-luc derivative.The cells were maintained in a monolayer culture in DMEM (Dulbecco's Modified Eagle's medium) supplemented with 10% of heat-inactivated FBS (Fetal Bovine Serum) (v/v) at 37 °C in the presence of 5% CO 2 .
Laser Scanning Confocal Microscopy of Fluorescently Labeled Polymer Probes: Laser scanning confocal microscopy (LSCM) was used to visualize the time-dependent intracellular uptake of selected polymer conjugates.FaDu cells were detached with 0.05% trypsin (Thermo Fisher Scientific, Czech Republic) at 37 °C for 3 min and seeded (2×10 4 cells per well) into confocal chambers (1 μ-dishes providing a 35 mm 2 growth area with four chambers, Bio-Port Europe, Prague, Czech Republic) in 500 μL cell culture medium 24 h before the experiment.The following day, the medium was replaced by a fresh medium with the dissolved fluorescentlylabeled probe in a concentration of 5 mg L −1 related to Cy7.Cells were incubated at 37 °C either for 6 or 24 h and the cell nuclei were visualized with Hoechst 33 342 (5 mg L −1 , Thermo Fischer Scientific, Czech Republic, 10 min).The cells were washed three times with Dulbecco′s Phosphate Buffered Saline and fixated with 4% paraformaldehyde (Sigma-Aldrich, Czech Republic).
LSCM measurements were performed by Abberior Instruments (Göttingen, Germany) equipped with excitation lasers,  ex = 405; 485; 561, and 640 nm.Stimulated emission depletion (STED) laser was used for Cy7 visualization (near-infrared spectrum) at  ex = 775 nm.Data were analyzed with the software Inspector 16.1 (Göttingen, Germany).Samples were scanned using Nikon CFI Plan Apo Lambda 60x Oil objective (Nikon, Japan) with 1.4 numerical apertures.An excitation wavelength of 405 nm was used for the detection of Hoechst 33 342 dye-stained nuclei, and emitted light was detected through a 422-467 nm filter.The size of each image was 60×60 μm with a resolution of 1200×1200 px.All samples were measured in duplicate in two independent experiments.Quantitative analysis of fluorescence images was performed using ImageJ software version 1.53t (National Institutes of Health, MD, USA).
Fluorescence-Guided Surgery of Orthotopic Head and Neck Tumors: The animals used were female nude mice aged 5-6 weeks from Janvier Labs (Le Genest, Saint-Isle, France).The experimental protocol was approved by the French Education and Research Ministry (experimentation authorization no 33281-202109291004967).The orthotopic Cal33-luc model was previously described and validated. [29,30]Briefly, a 0.5-1 mm fragment from a tumor developing secondarily to subcutaneous Cal33-luc cell implantation in adult mice was implanted in the inner aspect of the cheek (n = 19).Animals were randomized 21 days after tumor implantation to perform tumor resection under visual macroscopic guidance using NIR-I optical imaging (Optical Guided Surgery, OGS).The mice were randomly split into two groups: 9 mice for surgery without optical guidance (-OGS) and 10 mice with optically guided surgery (+OGS).In both cases, the mice received an i.v.administration of the P-H-OPB-Cy7 probe 24 h before surgery.Three hours after probe administration, the tumors were resected as extensively as possible by the surgeon through an external incision of the cheek with or without OGS.Three mice died during the surgery due to anesthesiologic complications, 2 mice from the -OGS group and 1 mouse from the +OGS group.The mice were regularly scanned for signs of tumor persistence after surgery.
Fluorescence-Guided Surgery of Peritoneal Carcinomatosis: Briefly, mice bearing intraperitoneal FaDu tumors expressing RFP were injected with polymer probes 24 h before imaging (c Cy7 = 40 mм, 2 groups, n = 10 per group).The mice were sacrificed, and the intraperitoneal cavity was examined in white light to resect the macroscopically visible tumor metastases.After the cytoreduction, mice were scanned for Cy7 and RFP fluorescence using the Xtreme In vivo Imaging System (Bruker BioSpin, Ettlingen, Germany).Consequently, the surgeon proceeded with the surgery without information about the localization of metastases.The residual tumor was detected using a 4 K ICG Storz imaging system (Holte, Denmark) in the fluorescence overlay mode and resected.After the surgery, the mice were scanned for residual RFP and Cy7 fluorescence using the Xtreme In vivo Imaging System, and RFP-positive foci were additionally resected.Resected material was examined for RFP and Cy7 fluorescence as well as histopathological evaluation.Sensitivity was defined as the ratio of truepositive results to the total positive samples.
Mice bearing Cal33-luc mouse model of intraperitoneal dissemination were i.v.injected with Cy7-labeled polymer probes 24 h before tumor resection (c Cy7 = 40 mм, 2 groups, n = 9 per group).The mice received an intraperitoneal injection of luciferin (150 mg kg −1 ) 5 min before the bioluminescence imaging using an IVIS Kinetic device (Caliper Life Sciences, Hopkinton, MA, USA) to determine the number of bioluminescent foci in the peritoneal cavity and the total bioluminescence intensity.Subsequently, mice were sacrificed, and the peritoneal cavity was opened using a midline laparotomy incision and systematically explored.Macroscopical resection of nodules visible under normal light followed by resection of fluorescence-positive nodules was performed to investigate whether the use of a fluorescence imaging system improved the detection of intraperitoneal metastases.Macroscopically visible metastases were surgically resected without the help of an optical system.Consequently, the residual tumors were visualized using a surgical camera in fluorescence mode (Fluoptics SAS, FB700 V2.3, excitation 680 nm, emission 700 nm) and resected by a surgeon who was not informed about the probe that was injected.The bioluminescence of the resected specimen was determined.After the surgery, the mice were scanned for the presence of residual bioluminescence-positive nodules and resected tumor specimens were sent for histopathological examination.
All experiments were approved by the ethics committee of the First Faculty of Medicine, Charles University, and by the Ministry of Education, Youth, and Sports of the Czech Republic.Experiments were performed in accordance with Act No. 246/1992 Sb. on the Protection of Animals Against Cruelty and Decree 419/2012 on the Protection of Experimental Animals (both were in accordance with the legislation of the European
weight (M w ) and dispersity (Ð) were determined by size-exclusion chromatography (SEC) using differential refractive index (RI), and multi-angle light scattering (LS) detection; b) Determined spectrophotometrically by UV-vis in methanol; c) Hydrodynamic diameter was determined by dynamic light scattering in PBS at 25 °C.

Figure 2 .
Figure 2. Synthesis of a) polymer precursor P-Hyd; b) activated derivatives with a TT group; c) dye Cy7 derivative and d) polymer probe with hydrolytically cleavable dye derivative.

Figure 3 .
Figure 3.The fluorescently-labeled polymer probes.The fluorescent dye Cy7 is attached via a hydrolytically or an enzymatically cleavable bond.The specific cathepsin B cleavage site is shown with blue arrows, orange arrows indicate the hydrolytically cleavable hydrazone bond attachment, and green arrows indicate where the Cy7 is attached.The last polymer conjugate P-A-Cy7 contains a non-cleavable dye and was used as a control.

Figure 4 .
Figure 4. Release rates of Cy7 from the fluorescently-labeled polymer probes A) at pH 6.5 mimicking the tumor environment B) at physiological pH 7.4 C) in presence of lysosomal protease cathepsin B. D) Release-triggered fluorescence increase for P-H-OPB-Cy7.
fluorometrically after the addition of 0.1м HCl.

Figure 5 .
Figure 5. Confocal microscope images of the FaDu cell line incubated with Cy7-labeled polymer probes from left to right: P-H-OPB-Cy7; P-H-PYR-Cy7; P-GFLG-Cy7; P-VCit-Cy7 and control polymer P-A-Cy7.The first row corresponds to an incubation time of 6 h, and the second to 24 h.The polymer probes are labeled with Cy7 (colored "fire" for sufficient resolution, a color map is displayed on the right), and the nuclei are stained with Hoechst 33 342 (colored yellow).Scale bar = 10 μm.

Figure 8 .
Figure 8. A) Representative bioluminescence images of CAL33 intraperitoneal tumors preoperatively (left) and postoperatively (middle).Intraoperative image of a mouse with a CAL33 intraperitoneal tumor as visualized by the Fluoptics system (right).B) Comparison of P-H-OPB-Cy7 and P-A-Cy7 probes detection of intraperitoneal metastases.C) Representative image of a FaDu intraperitoneal tumor visualized by the Storz system in white light (left), fluorescence mode (middle), and overlay mode (right).

Table 3 .Figure 9 .
Figure 9. Optically guided surgery of Cal33-luc orthotopic tumors.Mice engrafted orthotopically with human Cal33-luc tumor fragments in the cheek (n = 19) were split into 2 groups that were operated on without optical guidance (-OGS, n = 9) or using fluorescence-guided surgery (+OGS, n = 10).The mice were evaluated using bioluminescence 1 week, 3 weeks, 5 weeks, and 10 weeks after surgery A) to detect and measure the presence of tumor recurrence with or without OGS.B) Example of mice with a recurrent tumor after 1 and 5 weeks after surgery with or without OGS.

Table 1 .
Characterization of polymer precursors.

Table 2 .
Fluorescence increases in pH-sensitive probes after Cy7 release at pH 2.

Table 4 .
Characteristics of the prepared polymer nanoprobes.