Introduction

Nano metal–organic frame works (NMOFs) has been attracting constant attention in anticancer nanomedicine. Theoretically, nano drug delivery system (NDDS) can achieve targeted delivery and controlled release of anticancer components, dramatically improving the drug uptake and therapeutic efficacy of general chemotherapeutic formulations [1, 2]. More importantly, enhanced targeting uptake and diminished toxicity originated from NDDS are favorable for overcoming the limitations of traditional medication [3, 4]. Among all NDDS, NMOFs constructed from inorganic metal and organic ligands via coordination bond are regarded one of the most versatile ones. Large specific surface area and highly accessible porosity enable NMOFs efficient loading ability towards various chemotherapeutic drugs [5,6,7]. Rich functional groups available for surface modification can introduce enhanced targeting and boosted biocompatibility to NMOFs [6, 8]. All these merits make relevant NDDS draw growing interests in biomedicine [7], with sharply increasing publications emerged since the first case in 2007 (Figure S1) [9].

UiO-66 is one of the most famous NMOFs in biomedical fields [10, 11]. Formed by the assembly between metal node Zr6O4(OH)4 featuring as hard Lewis acid and twelve carboxylic acid ligands, UiO-66 is famous for excellent thermal and chemical stability [12]. Intrinsic porosity stemming from crystalline porous framework and abundant aromatic skeleton of organic linkers endow UiO-66 with ideal channels and intermolecular interactions (C–H···π bond, π-π stacking, ion–dipole interactions) to accommodate drug guests [13]. Side-chain groups such as –NH2, –N3 and –COOH pre-installed into the organic linkers are available for surface modification and targeting decoration [14]. Based on these, there are plenty of studies applying nano UiO-66 derivatives as NDDS carrier for nanomedicine (Figure S2 and Table S1), making UiO-66 almost become a paradigm of NMOFs with medical applications [15,16,17,18,19]. However, there are also some inconsistency or disparity in the prospective about biomedical application of UiO-66. For instance, although most of the reported researches claim that UiO-66 exhibits excellent stability both in vitro and in vivo, there are still some severe concerns about the rapid degradation of Zr-based NMOFs under physiological conditions [20,21,22]. Surface modification may be a vital alternative approach to addressing this problem, and some work has found that silica coating, PEGylation and polydopamine modification can enhance the physiological stability [23,24,25]. However, we found that most of previously work utilize powder X-ray diffraction (PXRD) to weigh the stability of the prepared NMOFs and they regard the carriers as stable providing that the PXRD patterns showed no obvious changes before and after contact with external factors [17, 26,27,28,29]. Although this kind of validation can shape the crystallinity of selected samples, it cannot reflect the stability exactly since partial decomposition is not easy to observe in PXRD patterns. Besides, plenty of the targeting ability of related UiO-66 and many other NDDS is validated via fluorescence imagining (FI) based on the chemical probe constructed by loading specific fluorescent components into corresponding nanocarrier [30, 31]. Of course, optical imaging with high temporal and spatial resolution is a good choice to visually point the cell uptake and distribution, but the validity of biodistribution given by FI probe should be taken cautiously. Since the ectopic xenografts and normal organs, particularly mononuclear phagocytosis system (MPS) with high uptake of nanoparticles, are not in the same spatial depth, even small amount of tumor uptake of fluorescence probe might be overestimated by FI [32]. All these confusions motivated us to perform a quite comprehensive investigation using more reliable approaches to value the perspective of UiO-66 derived NDDS in biomedical field.

89Zr (T1/2 = 78.41 h) related radiochemistry may provide a powerful tool to weigh the pharmaceutic value of UiO-66 even some other Zr-based NMOFs. Obviously, one can utilize 89Zr doped natZr metal resources to assemble with relevant organic ligands, easily obtaining correspondingly radiolabeled NMOFs available for radioactive tracing [24]. As is well known, radioactive tracer method can work well even below 37 Bq (10–14 ~ 10–13 g), while common chemical analysis is feasible over 10–9 g. This means that one can assess the stability of Zr-based NMOFs, accurately monitoring even tiny radioactivity release from related nanocarriers. Meanwhile, 89Zr is one of the most promising radionuclides in positron emission tomography-computed tomography (PET–CT), with suitable half-life and the relatively low translational energy suitable for noninvasive biodistribution over a relatively long term [33]. Undoubtedly, nuclear imaging featuring with much higher tissue penetration and detection sensitivity can better reckon the accumulation of NDDS in whole-body, especially those underestimated by FI in MPS. Despite a preliminary attempt in PET–CT imaging in murine model by Chen et al. [24], there is still a lack of reasonable evaluation of 89Zr radiolabeled UiO-66 NMOFs in anticancer effect.

In this work, we proposed to integrate 89Zr into UiO-66 NMOF decorated with anticancer cargo (doxorubicin, DOX) and targeting motif (PEG-FA), aiming to constructed a typical NDDS (DOX@89Zr-UiO-66-PFG-FA) for a PET–CT guided systematic elevation of the anticancer effect of the obtained NMOF carrier. It must be pointed out that the initial synthesis, surface modification, guest encapsulation involving UiO-66, together with the anticancer efficacy evaluation of related NDDS both in vitro and in vivo, had been performed according to classical methods, excepting that we chose to dope 89Zr from the beginning of NMOFs assembly. Our goal is to take a deep insight of the actual potentiality of UiO-66 in anticancer NDDS, with the aid of high sensitivity of radioactive tracing and high performance of nuclear imagining.

Results and discussion

Synthesis and characterization of DOX@UiO-66-PEG-FA

The microscopic morphology of the prepared UiO-66 based NDDS has been well regulated in this case. SEM analysis indicates that raw UiO-66-NH2 is slightly rough and quasi-spherical nanoparticle with an average diameter of about 85 nm (Fig. 1a). No significant changes in the micromorphology of UiO-66 NDDS had occurred after DOX encapsulation and surface PEGylation, save that the average particle size increased slightly to 106 nm (DOX@UiO-66-NH2) and 111 nm (DOX@UiO-66-PEG-FA), respectively. The average hydrodynamic diameter of UiO-66-NH2, DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA were 105.7, 190.1 and 163.1 nm, respectively, which is slightly larger than that from SEM analysis (Figure S4). It is worthy to note that DOX@UiO-66-PEG-FA could suspense in deionized water homogeneously for a long period of time, indicating that surface PEGylation is beneficial to increasing the dispersibility of UiO-66 (Figure S5). From the PXRD results, the position of the characteristic peak of the synthesized UiO-66-NH2 was 7.3° consistent with the simulated pattern obtained from the crystal information file (CIF) (Fig. 2a) [34]. In the FTIR spectrum (Fig. 1d) of UiO-66-NH2, the characteristic bands of N–H stretching vibration (3434 cm−1), C–H group stretching vibration (2853 cm−1), C–N stretching vibration (1433 cm−1), C–H bending vibration (1250 cm−1), Zr-O stretching vibration (766 cm−1), and symmetric and asymmetric stretching vibration of –COO– on BDC-NH2 (1384, 1575 cm−1) can be observed [35,36,37]. Compared with UiO-66-NH2, DOX@UiO-66-NH2 showed an additional new absorption peak at 1116 cm−1 which is attributed to the stretching vibration of –C–O–C–, thus evidencing that DOX was successfully loaded. The UV–Vis absorption spectra (Fig. 1e) showed a new absorption peak at 434 to 581 nm (DOX) for DOX@UiO-66-NH2, which further confirmed that DOX was successfully loaded. Since the characteristic peaks of the FTIR and UV–Vis spectra of FA-PEG-NHS are prone to overlap with DOX, it was impossible to determine whether FA-PEG-NHS was modified on the carrier by these two characterization results. Hence, it was necessary to ascertain this with other characterizations. Obviously, the surface zeta potentials of UiO-66-NH2 (+ 38.5 mV), DOX@UiO-66-NH2 (+ 31.5 mV), and DOX@UiO-66-PEG-FA (+ 13.0 mV) change enough to further confirm cargo accommodation and surface PEGylation of the MOF carrier (Fig. 1f), the incorporation of deprotonated DOX and PEG linker can diminish the surface potential of UiO-66 stepwise.

Fig. 1
figure 1

SEM images of UiO-66-NH2 (a), DOX@UiO-66-NH2 (b) and DOX@UiO-66-PEG-FA (c) (scale bar: 200 nm). d FTIR spectrum of DOX, UiO-66-NH2, DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA. e UV–vis absorption spectrum. f Zeta potentials

Full size image
Fig. 2
figure 2

a PXRD patterns of UiO-66-NH2, DOX@UiO-66-NH2, DOX@UiO-66-PEG-FA and the latter two treated with saline, FBS, PBS for 3 days. b Stability assessment of DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA in different media. c Cumulative release profiles of DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA at different pH of DOX

Full size image

Structure analysis and radiochemistry alike manifest that DOX@UiO-66-PEG-FA has acceptable in vitro stability. As shown in Fig. 2a, the main peaks of as-synthesized UiO-66-NH2 in PXRD diagram was relatively wide. This is due to that the small grain size of the prepared NMOF can induce X rays to produce diffusion, hence broadening the diffraction pattern. Compared with the raw UiO-66-NH2, the diffraction peak of DOX@UiO-66-NH2 became wider and shifted to higher 2theta degree, which may be caused by the reduction of crystal face distance and the poor crystal consistency due to DOX load. DOX@UiO-66-PEG-FA possesses similar PXRD pattern to those of DOX@UiO-66-NH2, indicating that the surface modification of FA-PEG-NHS did not affect the crystallinity of the NMOF carrier. However, the PXRD pattern of DOX@UiO-66-NH2 changed dramatically even disappeared after 3 days of immersion in FBS and PBS. Actually, the in vivo instability of Zr-based NMOFs is due to that phosphates in the blood have much stronger affinity towards Zr4+ than general organic ligands. This can well explain why DOX@UiO-66-NH2 exhibited better stability in saline without phosphates than in FBS and PBS. In contrast, the PXRD patterns of DOX@UiO-66-PEG-FA had maintained well in all selected media. From this point, UiO-66 itself should be regarded as unstable or susceptible to physicochemical conditions, and surface PEGylation is necessary for enhancing its stability. It is worth noting that even the PEGylated DOX@UiO-66-PEG-FA still present broadened diffraction peaks relative to the as synthesized one in PXRD patterns. This suggests that partial decomposition may undergo for the designed NDDS, which can be consolidated by radiochemistry investigation. In this case, we chose to prepare 89Zr radiolabeled UiO-66 via radioisotope doping: 89Zr + natZr was applied as zirconium resource to assemble with BDC-NH2, leading to radioactive NMOF carrier with a radiochemical yield of 98.2%. Theoretically, radioactivity release from the solid-phase carrier can well reflect the damage of UiO-66 based NDDS. However, over 90% of 89Zr had retained in both DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA after 3 days of immersion in different media (Fig. 2b). Different from the results of PXRD where DOX@89Zr-UiO-66-NH2 presented severe structure damage in simulated physiological media, UiO-66 based NMOF without PEGylation appeared to be able to retain most of the radioactivity in the solid-phase materials. This should be attributed that phosphates can destroy the crystallinity of UiO-66 by replacing the organic linkers, but still form Zr-containing nanoprecipitate due to the extremely low solubility product constant of Zr3(PO4)3 (10–132). As a result, no obviously radioactivity release can be detected in the liquid phase in this situation, as 89Zr had been fixed in the amorphous solids. It must be clarified that the radiochemistry investigation in this section does not mean that UiO-66 without PEGylation has enough stability but only reveal that the nanocarrier can hold the radionuclides even after structure damage.

DOX loading and release studies

Efficient anticancer cargo load and controllable release behavior have been observed on UiO-66 based NDDS. DOX, a classical chemotherapeutic compound was loaded onto UiO-66-NH2 through electrostatic interactions [26]. As shown in Table S2, the drug loading capacity (DLC) of DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA was 79.94% and 79.28%, respectively, which are quite comparative with the results in many previously-reported cases [26, 38]. To assess the stimulus responsiveness of DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA to the tumor microenvironment, the cumulative release of DOX from the nanomedicine in PBS (pH 5.0 and 7.4) was investigated using dialysis. After 8 d, DOX@UiO-66-NH2 released 29.1% and 16.7% DOX in PBS pH 5.0 and 7.4, those for DOX@UiO-66-PEG-FA were 27.2% (pH 5.0) and 14.9% (pH 7.4), respectively (Fig. 2c). From the results, the DOX release of DOX@UiO-66-PEG-FA and DOX@UiO-66-NH2 in PBS at pH 5.0 were about 13% higher than that in PBS at pH 7.4, proving beyond any doubt that DOX@UiO-66-PEG-FA and DOX@UiO-66-NH2 have a pH-responsive release property. DOX release rate of DOX@Uio-66-NH2 was slightly higher than that of DOX@UiO-66-PEG-FA, probably because the PEG modification protected part of DOX@UiO-66-PEG-FA from phosphate attack. And it can also be observed that both drugs basically reached the maximum release of DOX after 6 d, indicating that both DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA can achieve slow release of chemotherapeutic drugs.

In vitro cellular uptake and killing studies

The in vitro anticancer investigations indicate that DOX@UiO-66-PEG-FA should work as a good therapeutic formulation toward breast cancer. Due to the autofluorescence property of DOX, the uptake of DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA by 4T1 cells can be detected using confocal laser microscopy [39]. Both drugs were tracked using DOX (red fluorescence), DAPI (blue fluorescence) locating the nucleus and Dio (green fluorescence) available for shaping the cell membrane (Fig. 3). It was observed that DOX@UiO-66-NH2 or DOX@UiO-66-PEG-FA was internalized in almost every cell even after incubation for only 4 h (Figure S6). Moreover, with the increase of incubation time, the red fluorescence intensity of each cell kept constant, and more red fluorescence was gather around the nucleus, directing that the cells further internalized the nanomedicine. In order to better understand the uptake of the two drugs by 4T1 cells, DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA was used as radioactive tracer to further investigate the cell uptake of the designed NDDS quantitively. As shown in Fig. 4a, the binding rate of 4T1 cells to DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA exceeded 30% at only 4 h, which increased with time gradually. The results were highly consistent with the results of fluorescence imaging. It was also observed that the binding rate of DOX@89Zr-UiO-66-NH2 to 4T1 cells was slightly higher than that of DOX@89Zr-UiO-66-PEG-FA, which is abnormal with general recognition. It is believed that PEGylation and targeting modification can elevate the biocompatibility and tumor affinity of nanocarrier, which should lead to higher cancer cell uptake of DOX@89Zr-UiO-66-PEG-FA. The paradox may be due to that DOX@89Zr-UiO-66-NH2 has a high surface positive charge (+ 31.5 mV), which enable the this NDDS to bind to the cell membrane composed of anionic phospholipids via electrostatic interactions [40]. Although the 4T1 cell uptake of DOX@89Zr-UiO-66-PEG-FA with lower surface potential (+ 13.0 mV) is little weaker than that of DOX@89Zr-UiO-66-NH2, it can be well blocked by folate (Figure S7). Therefore, despite the diminished electrostatic force with the cell membrane, surface PEGylation and targeting modification can bring up specific tumor affinity to UiO-66 based nanocarrier. The results of the cell internalization experiments showed that both nanomedicines stayed stably in the cells and the amount of internalization increased with time, which was mutually confirmed with the fluorescence imaging and cell binding experiments (Fig. 4b). The efflux of DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA decreased with time and were not significantly different from each other (Fig. 4c). All these results indicate that electrostatic interactions between the positively-charged NMOFs and negatively-charged cell membrane are very strong, which can explain why lots of previously-reported work has obtained good results in cell uptake investigation evaluating tumor cell affinity of corresponding NDDS. In this part, one may find that PEGylation and targeting modification of UiO-66 based carrier is not necessary if cell uptake and internalization are taken in account only. However, since specific tumor affinity is more desirable in actual therapeutic process and biocompatibility is also a very important factor, we thought that related surface decoration should still be necessary.

Fig. 3
figure 3

Fluorescence images of 4T1 cells incubated with DOX@UiO-66-NH2 (I) and DOX@UiO-66-PEG-FA (II) for 24 and 48 h respectively. Scale bars: 50 µm\(.\)

Full size image
Fig. 4
figure 4

a Binding rates of DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA to 4T1 cells. b Internalization of DOX@89Zr-UiO-66-NH2 and DOX@89Zr-UiO-66-PEG-FA into 4T1 cells after incubation for 4, 8, 12, 24 and 48 h. c Efflux profile of 4T1 cells after incubation with the drug for 2 h. d Cell survival of 4T1 cells after treatment with UiO-66-PEG-FA, 89Zr-UiO-66-PEG-FA, Free DOX, DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA for 24 h, n = 6

Full size image

Cell killing effect of anticancer cargo can be well enhanced by UiO-66 based NDDS. As depicted in Fig. 4d, UiO-66-PEG-FA and 89Zr-UiO-66-PEG-FA exhibit high biosafety in vitro as cancer cells in related groups still maintained more than 90% cell viability. This indicates that both the UiO-66 based nanocarrier and radioactive tracing are quite biocompatible. Of course, free DOX has shown certain cell growth inhibition which can be assigned to its chemical toxicity. Once encapsulated into UiO-66 based carrier, improved anticancer effect of DOX can be observed for DOX@UiO-66-NH2 and DOX@UiO-66-PEG-FA. Overall, DOX@UiO-66-PEG-FA had a slightly stronger killing effect on cells than DOX@UiO-66-NH2. Due to the presence of overexpressed folate receptors (FR-\({\varvec{\upalpha}}\)) on 4T1 cells, nanoparticles modified with FA-PEG motifs are supposed to have better biocompatibility and can approach lysosome closer via folate receptors-mediated cellular internalization [41,42,43]. In the lower pH environment of lysosome, DOX@UiO-66-PEG-FA is able to release large amounts of DOX and induce cancer cell apoptosis. Combining cell uptake and drug killing effect analysis, although the binding rate of DOX@89Zr-UiO-66-NH2 to 4T1 cells was slightly higher than that of DOX@89Zr-UiO-66-PEG-FA, it behaved weaker in cell killing. At this level, surface PEGylation and targeting modification have been demonstrated to be necessary again as aforementioned.

PET/CT imaging and in vivo biodistribution

Systematic investigations about the tumor targeting ability uncover that UiO-66 based NDDS should be viewed cautiously. As in vitro cell experiments pointed out, surface PEGylation and targeting modification may enable NMOFs good tumor targeting ability, but whether this can work in vivo is still needed to verify. In this situation, we first conducted an anatomical biodistribution in murine 4T1model, employing trace amount of DOX@89Zr-UiO-66-PEG-FA as radioactive probe. As shown in Fig. 5a, no observable radioactivity accumulation at tumor sites could be detected from 12 to 24 h, when the radioactive tracer was injected intravenously. On the other hand, MPS consisting of liver, lung and spleen has presented a very strong ability to capture DOX@89Zr-UiO-66-PEG-FA. More badly, obvious heart accumulation of the radioactive probe can be observed at 12 h, although it declines sharply within 24 h. For intravenous injection, nanoparticles loaded with chemotherapy drugs encounter a number of biological barriers that seriously impede effective drug delivery [44]. Meanwhile, due to the large average hydrated particle size as well as positive surface potential, serum albumin, apolipoproteins, immunoglobulins, etc. are easy to adsorb onto the surface of NDDS during blood circulation, and proteins corona can also form around the nanoparticles readily [45, 46]. It is then ingested by phagocytes in MPS, resulted in large accumulating of NDDS in lungs, livers, spleens and other organs [47]. This result seems to be quite different from plenty of previously-reported work where the tumor targeting ability was checked by FI [48]. As discussed above, FI featuring with high temporal and spatial resolution has an inevitable drawback known as limited tissue penetration ability. At whole-body level, the ectopic xenografts and normal organs, particularly MPS with high uptake of nanoparticles, are not in the same spatial depth. As an aftermath, in certain FI analysis, even small amount of tumor uptake of fluorescence probe might be exaggerated due to the “location advantage” while that in the MPS might be discounted by tissue shield. Considering that MPS has abundant macrophages which can capture external matters in superior efficiency, an alternative approach to circumventing this problem is local administration of corresponding therapeutic formulations. With this in mind, we chose to check the tumor retention ability of intratumorally-injected DOX@89Zr-UiO-66-PEG-FA in murine 4T1model. Compared with intravenous injection, DOX@89Zr-UiO-66-PEG-FA is prone to stain at tumor sites after local administration from 12 to 24 h, and there was basically no uptake in other organs. All the above results indicate that DOX@UiO-66-PEG-FA can work as a therapeutic formulation via local but not whole-body administration. Of course, intratumoral injection is not of value in clinical application of diagnostic formulations, but has been demonstrated to be one of the pivotal avenues of therapeutic drugs [49, 50]. First, not every solid tumor can be excised directly in practice in clinic, like those close to aorta or upper esophageal cancer. In this situation, adjuvant therapy before surgery even nonsurgical radical treatment combined chemotherapy with radiotherapy become pivotal. Besides, it has been demonstrated that local administration is much efficient and safer for nonsurgical or presurgical adjuvant cancer treatment where one already knows the specific location of tumor sites, since it can well kill tumor tissue efficiently and precisely sparing healthy tissues and organs. Hence, local administration of anticancer NDDS is still has great meaning in clinical practice.

Fig. 5
figure 5

a In vivo biodistribution studies 4T1 tumor-bearing mice at 12 h and 24 h after intravenous injection of DOX@89Zr-UiO-66-PEG-FA. b In vivo biodistribution studies 4T1 tumor-bearing mice at 12 h and 24 h after intratumoral injection of DOX@89Zr-UiO-66-PEG-FA. PET–CT imaging of 4T1 tumor-bearing mice at 12 h, 36 h, 3 d, 4 d, 7 d and 9 d after intratumoral injection of DOX@89Zr-UiO-66-NH2 c or DOX@89Zr-UiO-66-PEG-FA (d)

Full size image

According to animal ethics, a long-term 89Zr based PET–CT imaging study was conducted. One thing needed to clarify first is that we were still not sure at that time about the necessity of surface modification of the designed NDDS, even after we decided to employ local administration to investigate the anticancer effect of DOX@UiO-66-PEG-FA. In the in vitro experiments, we found that DOX@UiO-66-NH2 without surface modification had better performance in targeting cancer cells, both in cell uptake and internalization but slightly lower cytotoxicity. In anatomic biodistribution study, we demonstrated that DOX@UiO-66-PEG-FA can retain at tumor site very well but made no investigation about the behavior of DOX@UiO-66-NH2. Therefore, both were utilized as PET probes to verify the necessity of surface modification of the prepared NDDS. As shown in Fig. 5c, it is easy to see that under the same radioactivity dose injected intratumorally, signal in DOX@89Zr-UiO-66-NH2 group is significantly weaker than that of the DOX@89Zr-UiO-66-PEG-FA group. Moreover, the signals were concentrated in a small area and did not penetrate other sites even after a long time, indicating that DOX@89Zr-UiO-66-NH2 had poor tumor permeability. DOX@89Zr-UiO-66-NH2 with large hydrodynamic diameter and positive zeta potential (+ 31.5 mV) is highly susceptible to binding with the negatively charged proteins in the tumor environment via electrostatic interactions. As a result, most of the nanomedicine will stay in the injection area for a long time. On the other hand, radioactive signals started presenting in liver, spleen and lungs in murine 4T1 model intratumorally-injected with DOX@89Zr-UiO-66-NH2 after 12 h. With prolonging the time, the signals intensity increased significantly. This means that some nanomedicine had escaped from the tumor area. In contrast, the two-photon signals of DOX@89Zr-UiO-66-PEG-FA injected to tumor-bearing mice were retained dispersedly in the whole tumor area over 9 days, without any radioactive signal found in other parts. Based on the sharp contrast, we can surely conclude that it is necessary to apply FA-PEG-NHS for surface modification, ignoring other existed issues. Surface modification and targeting decoration can not only enable the nanomedicine to bind tumor cells specifically to achieve long-term drug retention, but also improve the tumor penetration ability of related formulation.

Therapeutic Efficacy

Despite all suspicions presented above, DOX@UiO-66-PEG-FA has displayed enough anticancer effect in murine breast cancer model. Compared with PBS group, DOX@89Zr-UiO-66-PEG-FA is able to inhibit tumor growth significantly, as tumor size of corresponding group had even shown a slight decline during the first 10 days (Fig. 6a). Free DOX, as a whole-body therapeutic drug, also presents a certain anticancer effect, but quite limited in tumor growth inhibition. In addition, 89Zr-UiO-66-PEG-FA without anticancer cargo had ignorable effect on tumor growth, which can confirm the biosafety of the designed nanocarrier. This also suggests that present radioactivity utilized for PET–CT would not disturb the metabolism of the murine mode. The median survival in the DOX@89Zr-UiO-66-PEG-FA group is 30 days, a significant prolongation relative to 18 days in the PBS group (Fig. 6b). More importantly, the tumor of one mouse in DOX@89Zr-UiO-66-PEG-FA group had been ablated during the first 30 days (passed away at 60 d suddenly). This surely should be ascribed to that DOX@89Zr-UiO-66-PEG-FA with high loading of anticancer cargo has excellent tumor retention capacity. PET–CT imaging of the 89Zr-UiO-66-PEG-FA group and the DOX@89Zr-UiO-66-PEG-FA group alike was performed during the therapeutic process (Fig. 6e). As shown in Fig. 6e, radioactivity accumulation can only be observed in tumor area over 6 days in both groups. These results confirmed the good tumor retention of DOX@89Zr-UiO-66-PEG-FA again, even under high dose of 89Zr and high concentration of DOX. During the treatment period, the weight of tumor-bearing mice in each group did not decrease significantly, indicating that all treatments did not cause systemic toxicity (Fig. 6c). However, it was observed that the mice in the free DOX group developed ulcers at tumor site (Fig. 6d). In contrast, the use of MOFs to load DOX can avoid direct tissue contact with high concentrations of chemotherapeutic agents and has pH stimulation response, releasing DOX slowly at the tumor site.

Fig. 6
figure 6

a Tumor volume cures of 4T1 tumor bearing mice under four treatment conditions. b The survival curves of 4T1 tumor bearing mice. c Body weight change curves of mice in each group during the treatment period. d Optical pictures of each group of mice after 18 days of treatment. Group I: PBS; Group II: Free DOX; Group III: 89Zr-UiO-66-PEG-FA; Group IV: DOX@89Zr-UiO-66-PEG-FA. e PET–CT imaging of 4T1 tumor-bearing mice at 1, 2, 3, 5 and 6 d after intratumoral injection of 89Zr-UiO-66-PEG-FA and DOX@89Zr-UiO-66-PEG-FA respectively

Full size image

To further evaluate the therapeutic efficacy and toxicity of DOX@89Zr-UiO-66-PEG-FA, heart, liver, spleen, lung, kidney and tumor were selected for hematoxylin and eosin (H&E) staining (Figure S8). Both spleen and liver were found to have lesions of varying degrees, as evidenced by structural irregularities in the hepatocyte cords, hepatocyte enlargement, cytoplasmic vacuole formation (black arrowheads), and local inflammatory cell infiltration (blue arrowheads); there was a decrease in lymphocytes (black arrowheads) and an increase in hematopoietic cells (blue arrowheads). Among the three groups, the free DOX group had the most severe lesions. This may be because that free DOX spreads throughout the body from the tumor site and has adverse side effects on healthy tissues. In addition, tumor tissue in the free DOX group showed a large range of tumor cell necrosis (in the circle), and tumor tissue in the DOX@89Zr-UiO-66-PEG-FA group showed a small range of tumor cell necrosis (in the circle), with more cell division signs (black arrow). This was due to a decrease in the tumor growth inhibiting effect of DOX@89Zr-UiO-66-PEG-FA after 21 days of administration, leading to tumor recurrence.

Conclusion

In this work,the prepared DOX@89Zr-UiO-66-PEG-FA NDDS has excellent stability and shows promising binding and internalization towards 4T1 cells, presenting pH-stimulated response property and achieving controlled release of medical cargo. Although DOX@89Zr-UiO-66-PEG-FA has a slightly lower binding rate to 4T1 cells compared to DOX@89Zr-UiO-66-NH2 due to its lower surface potential, it shows specific affinity to tumor cells after targeted modification. In addition, DOX@89Zr-UiO-66-PEG-FA exhibits stronger cytotoxic effects on 4T1 cells compared to DOX@89Zr-UiO-66-NH2. After intratumoral administration, long-term PET–CT tracer showed that DOX@89Zr-UiO-66-PEG-FA had superior tumor retention, and tumor penetration ability was much stronger than DOX@89Zr-UiO-66-NH2 without surface modification. This suggests that appropriate surface modification can improve the possibility of UiO-66 derived NDDS for tumor treatment. The PET–CT guided anticancer investigation in this case demonstrate that it is necessary to carry out a comprehensive analysis in an objective manner when evaluating the medical potentiality of UiO-66 NDDS in anticancer. Also, it is important to implement the corresponding therapeutic regimen according to the characteristics of the substance.