Pyroelectric Catalysis-based “Nano-lymphatic” Reduces Tumor Interstitial Pressure for Enhanced Penetration and Hydrodynamic Therapy

Owing to deciency of lymphatic reux in the tumor, the retention of tumor interstitial uid causes the aggravation of tumor interstitial pressure (TIP), which leads to unsatisfactory tumor penetration of nanomedicine. It is the main inducement of tumor recurrence and metastasis. Herein, we design a pyroelectric catalysis-based “Nano-lymphatic” to decrease the TIP for enhanced tumor penetration and treatments. It realizes photothermal therapy and decomposition of tumor interstitial uid under NIR-II laser irradiation after reaching the tumor, which reduces the TIP for enhanced tumor penetration. Simultaneously, reactive oxygen species generated during the pyroelectric catalysis can further damage deep tumor stem cells. The results indicate that the “Nano-lymphatic” relieves 52% of TIP, leading to enhanced tumor penetration, which effectively inhibits the tumor proliferation (93.75%) and recurrence. Our nding presents a novel strategy to reduce TIP by pyroelectric catalysis for enhanced tumor penetration and improved treatments, which is of great signicance for drug delivery. react with negative charges to form ROS. The above results revealed the excellent cellular uptake and cytotoxicity of M/CdS-HA under 1064-nm laser irradiation.


Introduction
Nowadays, drug delivery has become one of the most controversial issues, which is also the top concern of all anticancer therapies, with the rapid development of nanomedicines [1][2][3][4][5] . The major obstacle in achieving better delivery e ciency is the abnormally high tumor interstitial pressure (TIP) [6][7][8][9] . It forms a barrier to the e cient uptake of the therapeutic agents from blood capillaries. In human capillaries, the venous pressure is nearly 20 mmHg, while the interstitial pressure of normal tissues is negative (about -1 to -3 mmHg) 6 . This outward pressure assures the penetration out from capillaries to the surrounding tissues. Whereafter, the lymphatic promotes the excess tissue uid to ow back to the blood system, completing the blood circulation [10][11][12] . However, the interstitial pressure in the tumor is as high as 30~130 mmHg due to the retention of interstitial uid caused by de cient lymph-vessel and lymphatic re ux, as well as dense extracellular matrix (ECM) [13][14][15] . The much smaller net outward pressure or even net inward pressure seriously limits the normal ow from vessels into the tumor 3 , which leads to unsatisfactory delivery e ciency of therapeutic agents. Among the mounting works on drug delivery, most efforts focus on reducing tumor interstitial solid pressure (TISP) caused by dense ECM, proteins, and broblasts via enzymolysis, which left the mass deposition of interstitial uid unsolved [16][17][18] . At present, little research can effectively reduce the tumor interstitial uid pressure (TIFP). Therefore, it is still a big challenge to decrease the TIFP for e cient tumor penetration and treatments of nanomedicines.
Because the basis of tumor interstitial uid is water (over 90%), the decomposition of water can effectively reduce the volume of tumor interstitial uid, meaning that decrease of TIFP 19 . Our group has designed a photocatalytic nanomedicine system to reduce TIFP, inspired by the catalytic water splitting employed in the eld of new energy 20 . This nanomedicine mimics lymphatic in releasing tumor interstitial uid and is termed as "Nano-lymphatic". In the previous report, we have preliminarily showcased the feasibility of the tumor interstitial uid decomposition and decrease of TIFP by photocatalysis 19 .
Nevertheless, the catalytic nanomedicine system could only be excited by the high-energy ultraviolet-visible light, which highly restricted its application in vivo due to the weak tissue penetration of ultravioletvisible light [21][22][23] . Therefore, we proposed an ideal and unexplored strategy to address the problem: Reducing TIP by pyroelectric catalysis. Different from photocatalysis, pyroelectric catalysis occurs under temperature variation instead of light excitation 24,25 . The pyro-generated positive and negative charges initiate the redox reactions, including water splitting and reactive oxygen species (ROS) generation [26][27][28] . Excitingly, the smart combination of photothermal agents and pyroelectric materials decomposes the tumor interstitial uid to reduce TIFP via pyroelectric catalysis-based water splitting and ablates tumor ECM, proteins, and broblasts to reduce TISP via hyperpyrexia. Releasing of TIP in this way can improve the permeability of nanomedicine into the center of tumors signi cantly. Furthermore, the ROS generated during the pyroelectric catalysis can further damage the deep tumor stem cells 29,30 . We identify this tumor-penetrating therapeutic strategy based on water splitting and ROS generation as "Hydrodynamic therapy".
A well-designed pyroelectric catalysis-based "Nano-lymphatic" is reported in this study for enhanced tumor penetration by reducing TIP, which combines photothermal therapy (PTT) and "Hydrodynamic therapy". In detail, the pyroelectric material CdS is grown on an ultrathin Nb 2 C nanosheet (MXene) in situ to provide M/CdS, which is then modi ed with tumor-targeted hyaluronic acid (HA) to the nal "Nanolymphatic" (M/CdS-HA). When the "Nano-lymphatic" reaches the tumor region through HA-mediated tumor target and under NIR-II (1064 nm) laser irradiation, MXene reduces the TISP via PTT rst. The temperature variation then triggers CdS to decompose the tumor interstitial uid via pyroelectric catalysis, leading to the decrease of TIFP. Bene t from the decrease of both TISP and TIFP, the "Nano-lymphatic" can highly penetrate the center of the tumor. Concurrently, the ROS generated during the pyroelectric catalysis breaks the ROS level threshold in the deep tumor stem cells, resulting in oxidative damage of cellular constituents to cell apoptosis or necrosis (Scheme 1). Our ndings reveal an innovative way to enhance the tumor penetration of nanomedicine, which is hugely critical in drug delivery. It is the rst demonstration of pyroelectric catalysis-based water splitting strategy for enhanced tumor penetration and improved treatments to the best of our knowledge. Furthermore, the scanning TEM high-angle annular dark-eld (STEM-HAADF) image and elements mapping of M/CdS were revealed in Fig. 1c. The elements C and Nb corresponded to MXene, and Cd and S corresponded to CdS. The above results con rmed the successful synthesis of M/CdS. The atomic force microscope (AFM) was used to characterize the height of M/CdS. As revealed in Fig. 1d When the temperature goes up (dT/dt > 0), the electric dipole moments decrease, which reduces the polarization intensity of pyroelectric material. The balance between the screening charges and polarization charges can be partially broken. Therefore, owing to the weak constraint, some screening charges is driven to the material surface to establish a new equilibrium. The pyro-generated free charges on the surface facilitate a pyroelectric potential through the electrode connected with an open circuit. If a short circuit alters the open circuit, the pyroelectric current will emerge 25 . On the contrary, when temperature decreases (dT/dt < 0), the heat dissipation from pyroelectric material leads to an increase of polarization intensity. Thus, the free charges from the surrounding electrolyte are adsorbed and redistributed to compensate for dipoles change. Likewise, an opposite pyroelectric current appears. When the temperature keeps invariant (dT/dt = 0), the electric balance in pyroelectric material is unbroken. Thus, none of pyroelectric potential and current generated 31,32 .
The electrochemical properties and pyroelectric effect of M/CdS were characterized by the electrochemical workstation. First, the electrochemical impedance spectra (EIS) Nyquist plots of CdS, MXene, and M/CdS in the dark and with 1064-nm laser irradiation were tested respectively, to evaluate the pyroelectric-induced electric charge transfer. As shown in Fig. 2b, the radii of semicircle for CdS showed no signi cant change in response to laser irradiation, which was reasoned that the CdS could not be excited by the low-energy 1064-nm laser (1.17 eV). Meanwhile, the radii of semicircle for MXene exhibited a slight reduction under laser irradiation, illustrating the ultrafast hot electrons transfer (Fig. 2c). However, the radii of the semicircle of M/CdS showed an obvious reduction under laser irradiation (Fig. 2d), suggesting the excellent charge transfer with the temperature variation. To investigate the advantages of composite, the EIS Nyquist plots of CdS, MXene, and M/CdS under laser irradiation were further studied.
As presented in Fig. 2e, the radii of semicircle for M/CdS possessed a smaller radius than CdS, which revealed that MXene signi cantly accelerated the separation of the pyroelectric-induced electric charges and resulted in a reduced resistance.
Then, the pyroelectric current and potential were further studied. By means of the COMSOL nite element simulation, the simulated pyro-potential distribution across CdS was characterized, as shown in Fig. 2f, in which the pyro-potential changes from 0 to 11.8 mV with a temperature variation of 50℃. Subsequently, the photothermal effect, pyroelectric current, and potential response of CdS and M/CdS under 1064-nm laser irradiation were measured. As presented in Fig The conductive band (CB) of CdS was calculated to be -0.935 V because the CB of n-type semiconductors is always 0 ~ 0.1 V more negative than the at-band potential 33 . Therefore, the CB potential of CdS is more negative than the redox potential First, the O 2 generation of M/CdS under 1064-nm laser irradiation was studied using a dissolved oxygen meter. After 5 min of 1064-nm laser irradiation, M/CdS could produce a signi cant amount of O 2 in contrast to MXene and CdS, which proved the occurrence of water splitting by pyroelectric catalysis (Fig. 3b). Then, the ROS generation of M/CdS under 1064-nm laser irradiation was detected using the 1, 3-diphenylisobenzofuran (DPBF) bleaching method (Fig. 3c) Besides, the M/CdS + lactic acid (LA) generated more ROS than M/CdS, which was reasoned that the LA as a sacri cial agent could react with positive charges (Eq. 4), leading to the decrease of charge recombination for enhanced catalysis 27,36 .
LA + q + CO 2 + H 2 O. (4) Next, the pyroelectric catalysis-mediated water splitting of M/CdS was studied, as shown in Fig. 3d. We designed a visual experiment, which proved the water splitting through the decrease of solution volume. Under 2 min of 1064-nm laser irradiation, the MXene solution volume did not change, which indicated the evaporation and expansion by heat could be neglected. Accordingly, the M/CdS showed an apparent decrease in solution volume and gas generation. Thus, the result revealed the M/CdS could realize the water splitting under 1064-nm laser irradiation through pyroelectric catalysis. Subsequently, we used the Matrigel to simulate the ECM for the penetration of MXene and M/CdS. As shown in Fig. 3e, the M/CdS exhibited better penetration in contrast to MXene under 1064-nm laser irradiation and laser alone. The reason might be that the M/CdS could produce heat and decompose water simultaneously, leading to excellent penetration.
The photothermal stability of M/CdS was tested as well. As presented in Fig. 3f, M/CdS could keep the immobile photothermal effect during three laser irradiation cycles, which was better than MXene alone. Furthermore, after 5 min of laser irradiation, the morphology of M/CdS showed no signi cant difference, suggesting its good photothermal stability ( Supplementary Fig. 2). It is known that the NIR-II laser exhibits superior tissue penetration [37][38][39] . Thus, we covered a chicken breast (5 mm) on the samples to further measure the photothermal effect. According to Fig. 3g, the temperature rise presented no signi cant change after the cover of chicken breast, which indicated that the excellent tissue penetration → of NIR-II (1064 nm) laser and potentials of M/CdS for further in vivo PTT. The above results proved that M/CdS showed excellent photothermal stability, ROS generation, water splitting, and ECM penetration, which con rmed the pyroelectric-catalytic process we proposed.
Cellular uptake and cytotoxicity. The cellular uptake and cytotoxicity of M/CdS-HA were studied using HeLa and LO2 cells. To localize the position of M/CdS-HA in cells, the uorescent rhodamine B was modi ed with HA to form M/CdS-HA/RB. As shown in Fig. 4a, the M/CdS-HA/RB exhibited better cellular uptake in HeLa cells than LO2 cells. The reason was that the sur cial HA of M/CdS-HA/RB could combine with the CD44 receptor on HeLa cells to improve the cellular uptake [40][41][42] . Furthermore, quanti cationally, the ow cytometry (FCM) analysis proved that the M/CdS-HA/RB could mostly be taken into HeLa cells (Fig. 4b). Subsequently, the ROS generation in vitro was measured by a 2′, 7′-dichloro uorescein diacetate (DCFH-DA) probe. Under 1064-nm laser irradiation, the HeLa cells treated with PBS, MXene, and CdS presented hardly any uorescence. However, the M/CdS-HA group showed noticeable green uorescence of 2′, 7′-dichloro uorescein (DCF), which testi ed the ROS generation during the pyroelectric catalysis ( Fig. 4c).
The cytotoxicity of all tested groups (PBS, MXene, CdS, M/CdS, and M/CdS-HA with/without 1064-nm laser irradiation) to HeLa cells was tested using MTT assay (Fig. 4d). Under laser irradiation, the MXene Penetration and cytotoxicity to multicell spheroids (MCSs). HeLa cells were seeded into a culture dish with an ultra-low attachment surface to form MCSs. To study the relationship between diffusion and transcytosis, an endocytosis inhibitor genistein (Gen) was used to pre-treated with MCSs to prevent the passage of M/CdS across the cell monolayer (Fig. 5a). We rst revealed the MCSs penetration of M/CdS-HA/RB with laser irradiation from Day 0 to Day 6 using uorescent imaging. As shown in Fig. 5b, the blue uorescence meant the cell nucleus, and the red uorescence localized in the center of MCSs, which suggests the M/CdS-HA penetrated MCSs under laser irradiation on Day 0. Following the laser irradiation on Day 2, Day 4, and Day 6, the MCSs collapsed stepwise, which indicated the excellent penetration and damage to MCSs. Furthermore, we researched the e cacy of different samples (MXene + L, M/CdS, M/CdS + L, and M/CdS + L + Gen) to MCSs by bright imaging (Fig. 5c). On Day 0, the sizes of MCSs in all the tested samples were consistent. After 6-day treatment, the M/CdS could not inhibit the proliferation of MCSs. The MXene + L group showed slight ablation of MCSs, which suggested the hyperpyrexia alone improved the MCSs penetration limitedly. However, the M/CdS + L exhibited complete disintegration of MCSs, which proved better cytotoxicity to MCSs resulting from the enhanced MCSs penetration through hyperpyrexia and TIP decrease. It was noted that the addition of Gen could not affect the damage to MCSs of M/CdS + L, which revealed that the enhanced diffusion by the combination of hyperpyrexia and decrease of TIP was the main approach of M/CdS delivery rather than transcytosis. The results of MCSs viability by MTT assay were consistent with bright images (Fig. 5d) It is necessary to evaluate the biosafety of M/CdS-HA in vivo to assure its further clinical and translational potentials. Several biochemical indexes were measured to assess the side effect of therapeutic agents. Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were used to assess the liver functions, and urea nitrogen (BUN) was used to assess the kidney functions. According to Fig. 6e, the M/CdS-HA + L group exhibited no signi cant difference on functions of liver and kidney compared with the Saline group for the experimental mice. Additionally, none of tested groups showed weight loss for the treated mice (Fig. 6f). Furthermore, the M/CdS-HA + L group showed no obvious organ injury or in ammatory lesions in the major organs (liver, spleen, kidney, heart, and lung) by H&E staining compared with the Saline group ( Fig. 6g and Supplementary Fig. 4). The above results indicated the superior antitumor e cacy and biosafety of M/CdS-HA + L for further clinical transfer.
Tumor penetration in vivo. The hyperpyrexia and pyroelectric catalysis-based decomposition of tumor interstitial uid by M/CdS-HA + L could decrease the TIP, which enhances the tumor penetration of nanomedicine and blood perfusion. The enhanced blood perfusion could bring more nanomedicine to the tumor site, leading to a circle of improved drug delivery. Moreover, the over-expressed LA in the tumor could serve as a sacri cial agent to consume the positive charges generated in pyroelectric catalysis for inhibiting the charge recombination, which promoted the negative charge-based reduction for the enhancement of ROS generation. Therefore, the LA in the tumor microenvironment enabled the M/CdS-HA to exhibit enhanced catalytic performance and antitumor e cacy under laser irradiation (Fig. 7a). Because of the decrease of TIP, blood perfusion was improved, leading to more accumulation of nanomedicine in the tumor. We detected the blood perfusion in the same tumor region without and with 2 min of 1064-nm laser irradiation at different time points (0 min, 15 min, and 30 min). According to Fig. 7d, after 24-h i.v. injection of M/CdS-HA, the blood perfusion was very weak without laser irradiation. After 2 min of 1064-nm laser irradiation, the blood perfusion enhanced gradually with time. It was reasoned that the reduced TIP could rapidly regulate the pressure difference between the blood and tumor interstitial uid, which provided a strong driving force for drug delivery from blood to the tumor. The regions pointed by the white arrows were the bleeding spots, which were formed by the increased blood ow and damage by hyperpyrexia. Simultaneously, the enhancement of blood perfusion mediated by M/CdS-HA + L could increase the intratumoral O 2 content, which improved the therapeutic effect of "Hydrodynamic therapy".
As shown in Fig. 7e, the Saline group exhibited hypoxia and no signi cant change under laser irradiation. In contrast, after the treatment of M/CdS-HA + L, the O 2 content increased obviously, which attributed to the enhanced blood perfusion. The experiments provided forceful evidence for the improvement of the drug delivery following the enhanced blood perfusion.
Finally, we explored the tumor recurrence and tumor apoptosis of all tested groups. In the experiment of tumor recurrence, we chose MXene-HA with i.t. injection + L as control. As revealed in Fig. 7f,

Discussion
Although nanomedicines have brought great hope to tumor treatments in recent years, the ine cient penetration into tumor severely limits their application. Previous studies focused on size-shrinking of drugs [43][44][45] , remodeling of tumor microenvironment 46 , and transcytosis [47][48][49] to enhance the tumor penetration. However, none of these strategies overcome the key hinder of drug delivery: TIP, especially TIFP. Therefore, for the rst time, we utilize pyroelectric catalysis-based water splitting to reduce TIFP and hyperpyrexia to reduce the TISP. Because this work simulates the lymphatic function to enhance the blood perfusion, we identify it as "Nano-lymphatic". Furthermore, the hyperpyrexia and ROS generated by NIR-II laser irradiation can e ciently damage the tumor cells. Bene t from the decrease of TIP, the "Nanolymphatic" exhibits enhanced tumor penetration and inhibition of tumor proliferation and recurrence.
In this contribution, the n-type semiconductor CdS and metalloid MXene are constructed to form a Schottky junction to realize further catalysis. Due to its wide energy gap, CdS cannot be excited by NIR-II light (1064 nm, 1.17 eV). Accordingly, we harness the pyroelectric effect of CdS under temperature variation to achieve the water splitting. The temperature variation can be realized through the photothermal effect of MXene under NIR-II light. However, the pyroelectric effect is hard to characterize. We utilize the COMSOL nite element simulation and electrochemical methods to prove the pyroelectric currents and potentials. Furthermore, we validated the photothermal effect, photothermal stability, water splitting, O 2 evolution, and ROS generation of "Nano-lymphatic".
We proved the superior antitumor effect of "Nano-lymphatic" by in vitro and in vivo experiments. Furthermore, the simulated tumor ECM by Matrigel, 3D tumor spheroids, and solid tumor born on the mice were used to prove the enhanced penetration of "Nano-lymphatic". The combination of hyperpyrexia and decomposition of tumor interstitial uid reduces the TIP signi cantly. Thereinto, the TIFP is the major constituent of TIP. In addition, the "Nano-lymphatic" can decrease the amount of LA in tumors, which indicates that the LA is the sacri cial agent to react with positive charges, inhibiting the recombination of charge carriers for enhanced ROS generation. Based on the above results, we identify the tumorpenetrating therapeutic strategy based on water splitting and ROS generation as "Hydrodynamic therapy".
In summary, we designed a pyroelectric catalysis-based "Nano-lymphatic" to reduce TIP for enhanced tumor penetration and treatments. We rst identify the "Nano-lymphatic" and "Hydrodynamic therapy" in this work. "Nano-lymphatic" represents a kind of nanomedicine to reduce the TIFP, and "Hydrodynamic therapy" is a newly emerging deep-penetrating therapeutic strategy with catalytic water splitting and ROS generation. Furthermore, we bring the pyroelectric catalysis into nanotechnology-based tumor treatments.
It is expected that such concepts can be bene cial for future advances in drug delivery, catalytic nanomedicine, and cancer therapy.  6 were dissolved into 29 mL ultrapure water, which added into single-layer MXene solution with one-hour stirring. Then, 1 mL Na 2 S solution (0.4 mg/mL) was rapidly added into above solution. The color of the system converted from black to dark green. The M/CdS was synthesized after a half-hour stirring. Finally, the precipitation of M/CdS after centrifugation was resuspended and incubated with 1 mg/mL of HA solution for 2 hours. The M/CdS-HA was obtained after centrifugal puri cation.

Methods
Characterization of structure and performance. TEM (HT7700, Japan) was used for characterization of the M/CdS. The samples were prepared onto a copper grid coated carbon lm for TEM measurement. Tumor penetration in vivo. The TIP level was measured by the WIN technique. Brie y, the pressure detector was calibrated before measurements. The needle was attached to a pressure transducer via a PE-50 polyethylene tube lled with sterile heparinized saline. The anesthetized mice, whose body temperature was maintained at 37.5°C with a heating pad, were carefully unhaired above the tumor. The needle was inserted into the central area of the tumor to test the TIP. The amount of LA in tumor was tested by ELISA kit. The tumor-bearing mice injected with Saline and M/CdS-HA were anesthetized, and the photoacoustic imaging system and blood oxygen imaging system were used to detect blood perfusion and blood oxygen concentration without and with 2 min of 1064-nm laser irradiation at different time points (0 min, 15 min and 30 min). The cell apoptosis of central tumor was tested by FCM and Ki67 staining section.
Ethics statement.
All animal procedures were performed in accordance with the statute of laboratory animal ethics committee of Yanshan University to ensure the low suffering for the animals, which were approved by the Animal Ethics Committee of Yanshan University.
Statistical analysis. The statistical analysis was performed by using Student's two-tailed t-test with statistical signi cance assigned at P < 0.05 (signi cant), P < 0.01 (moderately signi cant), and P < 0.001