Local detection of pH-induced disaggregation of biocompatible micelles by fluorescence switch ON

Fluorogenic nanoparticles (NPs) able to sense different physiological environments and respond with disaggregation and fluorescence switching OFF/ON are powerful tools in nanomedicine as they can combine diagnostics with therapeutic action. pH-responsive NPs are particularly interesting as they can differentiate cancer tissues from healthy ones, they can drive selective intracellular drug release and they can act as pH biosensors. Controlled polymerization techniques are the basis of such materials as they provide solid routes towards the synthesis of pH-responsive block copolymers that are able to assemble/disassemble following protonation/deprotonation. Ring opening metathesis polymerization (ROMP), in particular, has been recently exploited for the development of experimental nanomedicines owing to the efficient direct polymerization of both natural and synthetic functionalities. Here, we capitalize on these features and provide synthetic routes for the design of pH-responsive fluorogenic micelles via the assembly of ROMP block-copolymers. While detailed photophysical characterization validates the pH response, a proof of concept experiment in a model cancer cell line confirmed the activity of the biocompatible micelles in relevant biological environments, therefore pointing out the potential of this approach in the development of novel nano-theranostic agents.


Introduction
Nanomedicine exploits dened nanostructures for the development of new systems with improved selectivity and efficiency with respect to existing diagnostic and therapeutic methods. 1,2 As an example, nanoparticles (NPs) are presently used for the delivery of mRNA in many anti-COVID-19 vaccines. 3 In this framework, "smart NPs", able to respond to either internal or external stimuli, have offered great advantages in terms of versatility, controllability and specicity. [4][5][6][7][8][9] Amphiphilic block co-polymers have emerged as one of the most promising approaches for the design of these stimuli-responsive nanotheranostic agents, as they can assemble into NPs of controlled size and surface properties. 10 Additionally, the stability of these self-assembled NPs can be controlled 11 by tuning parameters like the block co-polymer composition, 12,13 the chain-length or the polymerization method. 14 As a consequence, stimuli-induced changes in the block co-polymer composition can be easily exploited to decrease NP stability inducing disassembly.

Stimuli-responsive NPs
To obtain stimuli-responsive NPs, the polymeric units should include structure specic blocks sensitive to a selected stimulus that can be exploited to trigger their conversion from stable to unstable and if possible their disaggregation. 15 Control over the polymeric composition is, in this context, fundamental since the overall response, and hence NP disaggregation, arises from the individual response of the single polymeric unit that has to be the same in order to achieve a sharp effect. Between the different stimuli that can be exploited to induce NP disaggregation, pH decrease is particularly important in nanomedicine. Cancer tissues, for example, possess a particularly acidic environment, which can be exploited as the stimulus for the selective response of pH-sensitive NPs. 16 Intracellular vesicles such as endosomes and lysosomes are known to undergo progressive acidication, 17 a process which can be exploited to trigger NP disassembly and intracellular drug release. Nevertheless, although this pH-induced disassembly process of NPs is very important for therapeutic application, the actual visualization of the disaggregation of the NPs in the target tissue is not trivial and it can be achieved by a tailored functionalization of the polymers with specic uorescent molecules through which uorescence switching OFF/ON upon NP disassembly can be achieved. Fluorescence imaging is, in fact, a powerful tool for the visualization of biological events thanks to its high speci-city and sensitivity, 18 while various organic uorescent molecules have been proposed for the development of diagnostic and sensing platforms. 19 For the design of uorogenic NPs, although non-covalent integration of the emissive component in the system is synthetically easy and convenient, 20 the covalent binding of the uorophore to the NP guarantees higher speci-city preventing uorophore leaking and unspecic responses.

Disaggregation induced emission (DIE)
The strategy we propose here to achieve a uorogenic response to disassembly is based on disaggregation induced emission (DIE). 21 This approach presents, for the visualization of NP disassembly, an advantage with respect to alternative aggregation induced emission (AIE) since it offers the possibility to observe a switch ON of uorescence upon disaggregation instead of uorescence quenching, which would be less specic. 22,23 Perylene diimides (PDIs) constitute an important class of organic dyes that present ideal features for the design of systems whose response is based on the uorescence signal switch ON or OFF upon aggregation-disaggregation processes. [24][25][26][27][28] In fact, PDI dyes, which are very strongly uorescent, exhibit a very low uorescence quantum yield in the aggregated form. 29 As one of the main prerequisites for bio-application is the high biocompatibility of the probe, organic assemblies, consisting of a uorescent core made by an organic emitter and a shell made by biocompatible/biodegradable polymers, are very promising candidates. Here, we describe the preparation and characterization and we also demonstrate the bio-application of NP (micelles) that: (i) result from the selfassembly of block-polymers with a well-dened composition and bear a pH-responsive block, (ii) are stable at pH 7.4, but disaggregate upon moderate acidication (pH 5.0), (iii) their disassembly can be detected since it is associated to a clear switch ON of the uorescence (DIE), (iv) are different from similar previously reported systems, 23,30 and whose uorescence change is considerable and is demonstrated in a signicant biological model. The micelle constituting units are amphiphilic block co-polymers terminating with a single PDI unit prepared by exploiting ring opening metathesis polymerization (ROMP) of functionalized norbornene derivatives. ROMP was the synthetic approach of choice due to the high control that it endows to the system, leading to uniform populations. 31,32 Moreover, ROMP recently emerged as a powerful polymerization technique in nanomedicine as both natural motifs and targeting moieties or therapeutics can be easily polymerized. [33][34][35][36][37] Therefore, in this study we aimed to exploit such a polymerization technique to prepare micellar NPs that are able to respond to relevant physiological environments with uorescence "switch ON".

pH-responsive micelles
Herein, we constructed amphiphilic block copolymers capable of being formulated into NPs, utilizing three different norbornene monomers selected from the four m1, m2, m3 and m4 shown in Scheme 1a. More in detail, the rst polymer R is constituted by a PEGylated hydrophilic block (m3 block) and a tertiary amine pH-responsive hydrophobic block (m2 block) efficiently functionalized by a uorescent perylene derivative (m1). The polymer's intrinsic tendency to self-assemble was exploited to obtain spherical micelles (RM), where the variation of the pH from 7.4 to 5.0 led to the uorescence "switch ON". As shown in Scheme 1b, emission is triggered due to the protonation of the amine of the hydrophobic block m2, causing a structural change in the aggregation properties of the amphiphiles and the PDI derivative. A second polymer C constituted by a PEGylated hydrophilic block (m3 block) and a pH insensitive hydrophobic block (m4 block) efficiently functionalized by a uorescent perylene derivative (m1) was prepared as the control. The copolymer C was assembled into micelle CM which did not undergo disassembly in acidic conditions, but could be disaggregated in the presence of a surfactant such as sodium dodecyl sulphate (SDS).
This work provides a careful optical characterization of the obtained micelles at different pH levels of interest and demonstrates a simple approach to generate novel responsive nanomaterials for the development of nano-theranostic and drug delivery tools. Herein, as a proof of concept, in vitro experiments in a model cancer cell line are reported to prove the biocompatibility and cellular internalization and to validate the effective responsiveness of RM to relevant physiological environments.

Results and discussion
Fluorogenic micelles were prepared via the self-assembly of polymeric amphiphilic units R and C (synthesized by ROMP polymerization as described in the ESI †) as shown in Scheme 1a. In particular, polymer R contains a section of amino terminated units (pK a ¼ 6.9) that switch from being hydrophobic to hydrophilic upon protonation. Both polymers R and C were characterized by size exclusion chromatography (SEC) showing a M W of 21 060 g mol À1 (Đ ¼ 1.069) and 18 070 g mol À1 (Đ ¼ 1.013) respectively (Table S1 †).
For the preparation of the micelles, the polymers were dissolved in dimethylformamide (DMF) and the solvent was slowly replaced by dialysis into PBS for 72 h. The micelles resulting from the assembly of R and C will be referred to as RM and CM respectively (Scheme 1b) and their formation was demonstrated both by dynamic light scattering (DLS) and by cryo-TEM as shown in Fig. 1. Both types of micelles showed a hydrodynamic diameter d $20 nm. 38,39 We would like to emphasize that RM contains a block of the amino terminated monomer m2 which is mostly unprotonated and hydrophobic at physiological pH (7.4), but undergoes protonation at acidic pH becoming hydrophilic. Hence, RM stability is expected to be strongly pH-dependent and to decrease at low pH leading to disassembly triggered by the formation of a hydrophilic protonated R polymer (according to Scheme 1b). In contrast, micelles CM do not contain protonatable sites and the m4 block remains hydrophobic independent of the pH. As a consequence, the stability of CM is not expected to be pH-dependent.
As shown in Scheme 1a, both the polymeric chains R and C possess a starting monomeric unit m1, which contains a perylene diimide (PDI) chromophore. PDI derivatives are known to give strong p-p interactions that lead to the formation of aggregates with altered photophysical properties with respect to the starting monomer. In particular, the formation of H aggregates leads to a distortion of the absorption band and to considerable quenching of the uorescence. 29 Hence, both the formation of the micelles and possible disaggregation processes can be studied by photophysical measurements.

Photophysical properties
In order to understand the effect of micelle formation on the photophysical properties of the PDI chromophore, we rst investigated the absorption and uorescence properties of the reference precursor m1 in dichloromethane (CH 2 Cl 2 , DCM), a solvent where PDI molecules aggregate very poorly. The absorption spectrum of m1 in CH 2 Cl 2 (1 Â 10 À6 M), shown in Fig. 2a, clearly shows the typical vibration structure of the lowest energy absorption band of the PDI monomer with a maximum at 526 nm (3 $80 000 M À1 cm À1 ). 26,27 A similar vibrational structure is observed for the uorescence band upon excitation at 470 nm as shown in Fig. 2b. This is the expected emission of the monomer, with a maximum at 535 nm and a uorescence quantum yield F ¼ 90%. 29 The attribution of the uorescence to the monomeric species was conrmed by the excitation spectrum at 575 nm (not reported) that matched the absorption spectrum of the monomer shown in Fig. 2. The excited state lifetime was measured by time correlated single Following this, we investigated the photophysical properties of the micelles RM and CM in PBS at pH 7.4. The absorption spectra of the micelles RM and CM (in a concentration of the polymeric components R and C of 1.5 Â 10 À6 M) are shown in Fig. 2a and they clearly show a distortion of the absorption spectrum that results from the p-p interaction of the PDI units at the core of the micelles. In both cases, the absorption maximum is shied to lower wavelengths with respect to the monomer (l max ¼ 469 nm and l max ¼ 492 nm for RM and CM respectively) and it is consistent with the formation of H-aggregates in the cores of the micelles. 29 This conclusion is also supported by the quenching of uorescence of the micelles with respect to the monomer. The uorescence spectra, upon excitation at 470 nm, are shown in Fig. 2b. For the CM micelles, two uorescence bands are present at 536 and 661 nm, the latter being more intense than the former. For RM, in contrast, the main uorescence band, upon 470 nm excitation, is at 541 nm and only a weak emission can be observed at 668 nm. While the bands at 536 and 541 nm (green uorescence) present the typical energy and structure of the PDI isolated chromophore, the bands at 661 and 668 nm (red uorescence) correspond to emissions that have been attributed, for similar systems, to the presence of aggregates, and they arise from delocalized excited  states with lower energy and unstructured uorescence. 26 It is noteworthy that for both RM and CM the average quantum yield was quite low being 5% and 12% respectively (see 1.3.2 of the ESI †). The origin of high energy green uorescence was conrmed using the excitation spectra (l em ¼ 580 nm), shown in Fig. 2c, that presented the typical structured band of the monomer. For a more detailed investigation, the uorescence maps of RM and CM were acquired and are shown in Fig. 3 (a  and b respectively). These contour maps represent the uorescence intensity as a function of the excitation and emission wavelength and they allow us to correlate the excitation and emission spectra of the different species present in a sample. The emission map of RM clearly shows the dominant presence of the green structured emission independent of the excitation wavelength while, in the case of CM, two different peaks can be clearly identied, corresponding to the green and red emissions. In addition, TCSCP was used to acquire the time resolved emission spectra (TRES) of RM and CM as shown in Fig. 3 (c and  d respectively). TRES maps allow us to follow the kinetics of excited state deactivation at different wavelengths. It is interesting to note that both in the case of RM and CM the green emission decays can be tted with a mono-exponential model with lifetimes (5.16 and 5.36 ns for RM and CM respectively), which are comparable to the one measured for the nonaggregated m1 in CH 2 Cl 2 . These results suggest that the green emission is indeed originated from a fraction of the polymeric chains R and C which is not assembled into the micelles, and hence in the micellar solution the structured emission band is due to a minor fraction of the unquenched, free polymeric moieties. Furthermore, the presence of large disassembled highly emitting molecules was conrmed by anisotropy uorescence measurements (shown in Fig. 4). This technique provides information about the kinetics of the reorientation processes occurring between the excitation and the emission of a uorophore and, in particular, about the rotation rate. Fluorescence anisotropy in the 530-540 nm range was as high as r ¼ 0.27 and r ¼ 0.28 for RM and CM respectively, suggesting that the rotation of the PDI uorophore is slow since it is bound to a high molecular weight polymer.
Having claried that the residual green emission was due to un-associated polymeric units, it was possible to use the uorescence quantum yield of this emission to quantify the fraction of free polymeric chains that was about 6% in the case of RM and about 10% in the case of CM (see 1.3.3 of the ESI †). As previously mentioned, in the case of CM a second broad and intense uorescence band in the red region was detected. This uorescence is typical of PDI aggregates and, hence, it is originated by the CM micelles themselves. This is also demonstrated by the excitation spectrum at l em ¼ 660 nm (shown in Fig. 2c) that presents the prole of the PDI aggregates. We would also like to stress that, unlike what is observed for the dynamic formation of an excimer where a delay time for the formation of an excimer species is detected (since it requires the encounter of an excited and a ground state chromophore), in the case of CM the TRES map shown in Fig. 3d reveals that the formation of the red emitting species is fast (<100 ps). In particular, this formation is much faster than monomer excited state decay, demonstrating that the two processes (monomer excited state decay and exciplex formation) are not correlated and hence conrming that the two emissions can be attributed to two different species. As shown in Fig. 3d the decay of the red uorescence of CM could be tted with a biexponential model to give two lifetimes as long as 5.7 and 17.0 ns. In contrast, in the case of RM, the red emission is not only much weaker than the green one, but as shown in Fig. 3c, it is also much shortlived than that of CM. In fact, in this case, it could be tted with a bi-exponential model giving two lifetimes as short as 0.58 and 3.64 ns.
This behaviour can be easily explained by the presence of unprotonated m2 units in the micelles. In fact, it is well known that amino derivatives are able to quench excited PDI molecules via an electron transfer mechanism. Therefore, as shown in Scheme 1b, we can assume that the unprotonated amino group in the micelles indeed produces fast deactivation of the PDI excited state leading to complete quenching of the uorescence of the micelles. In summary, we can conclude that at pH 7.4: (i) both R and C are, to a large extent (more than 90% in molar fraction), assembled into micelles; (ii) a residual green uorescence can be observed in both cases due to the presence of a minor fraction of disassembled polymers; (iii) micelles CM show a broad uorescence band in the red region with a long excited state lifetime; (iv) micelles RM show only a weak emission in the red region with a much shorter lifetime with respect to CM. This is due to quenching of the emission due to electron transfer processes involving unprotonated amino groups present in the micelles.

Disaggregation experiments
Micelles RM were designed to present pH-responsive disaggregation attributed to the presence of the alkyl-amine moieties that are protonated at a pH below neutral and similar to the physiological pH (pK a ¼ 6.9). In order to check this pH response, two different experiments were performed: (i) Fig. 4 Fluorescence anisotropy spectra of RM at pH 7.4 (black circles) and pH 5.0 (black triangles). Fluorescence anisotropy spectra of CM in the absence (red circles) and in the presence of SDS (red triangles). titration with an HCl solution to quickly decrease the pH and validate the system (Fig. S6 and S7 †), and (ii) dialysis against an acetate buffer solution at pH ¼ 5.0, to better mimic the pH found in physiological environments. In the latter case, acidi-cation of RM was achieved via a dialysis experiment using a membrane with low permeability (cutoff ¼ 3500 kDa) in order to avoid the out-diffusion of both micelles and polymeric units. A solution of RM in PBS was dialysed against acetate buffer at pH 5.0. In Fig. 5a the uorescence spectra of RM before and aer dialysis (2 h, 24 h and 48 h) are shown demonstrating that the pH decrease produces an increase in the uorescence intensity by about 9 times aer 48 h dialysis. We would like to stress that much of the uorescence increase in the RM sample occurs within the rst 2 hours of dialysis against acidic acetate buffer. In contrast, when the same experiment was performed on the control sample, CM, as shown in Fig. 5b, almost no change in the uorescence spectrum was observed upon decreasing the pH from 7.4 to 5.0 even aer 48 h dialysis.
In order to understand in detail the pH-responsive uorescence of RM, we performed a comprehensive photophysical characterization of the systems at pH ¼ 5.0. In particular, as shown in Fig. 2a, the absorption spectrum of RM greatly changed by changing the pH from 7.4 to 5.0 exhibiting the typical features of non-aggregated PDI in the acidic solution. 24,26 The disaggregation was conrmed by the average uorescence quantum yield of the RM sample, which increased to 46% upon decreasing the pH from 7.4 to 5.0. The TRES map of RM at pH ¼ 5.0 is shown in Fig. 3e. It is worth mentioning that it was possible to t the decay of the green uorescence with a monoexponential model demonstrating that disaggregation did not lead to any signicant change in the uorescence excited state lifetime, that was s ¼ 5.24 ns at pH 5.0. This result clearly shows that acidication indeed produced an increase in the fraction of free R molecules, as a result of the disaggregation of the micelles due to the protonation of the amino groups. Additionally, as shown in Fig. 4, uorescence anisotropy of RM did not change by decreasing the pH from 7.4 to 5.0 demonstrating that at both pH levels, the uorescent species are the same and are the free polymer chains. For the CM system, the pH decrease from 7.4 to 5.0 had no effect on aggregation due to the absence of easily protonatable sites in C. Nevertheless, as a control experiment, micelles CM were treated with a surfactant (SDS) which is known to solubilize the polymers leading to micelle disaggregation. As shown in Fig. 2a, the absorption spectrum of CM completely changed aer the addition of SDS presenting the typical structure of nonaggregated PDI. Correspondingly, as shown in Fig. 5c, the red emission attributed to the micelles completely disappeared while a strong increase of the green uorescence of polymer C could be detected. In the same Fig. 5c, a very similar behaviour can be observed for RM. The nal uorescence quantum yield of CM aer the addition of SDS was as high as 87% and so was compatible with complete disaggregation of the micelles. The TRES map of CM in the presence of SDS is shown in Fig. 3f and it conrms the complete disappearance of the red emission band.
Moreover, the green uorescence band could be tted with a single mono-exponential decay with a lifetime s ¼ 5.10 ns very similar to the one measured in the absence of SDS. As far as uorescence anisotropy is concerned, as shown in Fig. 4, the addition of SDS leads to a decrease of the anisotropy of CM that in any case remains as high as r ¼ 0.22. This result is compatible with the incorporation of more than just one C molecule in the SDS micelles and a partial depolarization of uorescence because of the homo-energy transfer between the PDI units. The disaggregation of RM micelles in this acidic environment (pH ¼ 5.0) was also investigated by dynamic light scattering (DLS) experiments, as well as by TEM (Fig. 1, S8 and S9 †). In this context, we would like to stress that neither DLS nor TEM allowed us to identify clear individual R or C molecules. Nevertheless, in the case of DLS, disassembly of the micelles is expected to produce a decrease in the scattering intensity detected by the instrument as the single disaggregated molecular units exhibit poor scattering. Indeed, in the case of RM, acidication of the solution at pH 5.0 upon dialysis led to a decrease in the scattering intensity of about two orders of magnitude verifying the disaggregation of most of the RM micelles to poorly scattering molecular units. However, as shown in Fig. 1c, this corresponds to only a minor decrease in the hydrodynamic diameter related to the minor fraction of the still aggregated micelles, while a minor fraction of the larger aggregates is observed both before and aer the pH change. Additionally, the residual fraction of undissociated RM could also be detected in TEM at acidic pH, which revealed the presence of residual particles aer the acidication of RM (Fig. S9 †).
A similar result was also observed in the case of acidication of RM by using HCl, where a decrease of about two orders of magnitude of the scattering intensity was noted without a substantial change in the measured hydrodynamic diameter of the few remaining highly scattering micelles (Fig. S7 †). In contrast, no changes in the scattering intensity, in the hydrodynamic diameter and in the TEM images were detected upon acidication of CM with both acetate buffer and HCl, therefore conrming their stability aer pH reduction (Fig. S7-S10 †).

Cell culture experiments
Having demonstrated the disaggregation of RM upon acidication from pH 7.4 to pH 5.0 in a buffered solution, we then tested the ability of these micelles to respond to a physiological pH decrease in living cells. We would like to stress that although another pH-responsive PDI-based uorogenic nanosystem has previously been reported, it did not show any disaggregation ability, and its application at the cellular level was not demonstrated. 40 In particular the endo-lysosomal pathway of cells is characterized by a gradual pH decrease, a process that cells exploit for degradation of unnecessary cargo upon internalization. In this pathway, the drop in the pH ranges from 7.4 (extracellular) to 6.0 in the late endosomes and can reach values around 4.5 in lysosomal compartments. 17 Self-assembled delivery systems capable of disassemblying upon such pH change are particularly interesting for pH-activated release of specic cargos. In general, verifying the actual disassembly of nano-vessels in response to acidication is not trivial; however regarding RM, on the basis of the previously discussed disaggregation, this process can be visualized in real-time by following the uorescence response.
HeLa cells, a model cancer cell line oen used for cellular uptake studies, 41 were incubated with either CM or RM and analysed by cytouorimetry, a technique that allows the easy quantication of the uorescence response in a cellular environment. As shown in Fig. 6a, cells treated with the responsive micelles RM show a gradual increase of the uorescence that is, aer 48 h incubation, almost one order of magnitude higher than the one observed in the presence of the control micelles CM. Moreover, in order to further test the biocompatibility of the probe, cell viability was assessed by MTT assay aer up to 48 h incubation at increasing doses of both CM and RM. As shown in Fig. 6b and S12, † no signicant reduced viability was observed in our experimental conditions upon either RM or CM incubation even at a concentration of 200 nM. This result demonstrates the biocompatibility of the developed micelles.
Although the cytouorimetric experiment gave important quantitative information (Fig. 6a), which were in agreement with an increase of the RM uorescence associated with endolysosomal acidication, confocal imaging experiments were needed in order to demonstrate (i) the actual internalization of RM by the cells through the endolysosomal pathway and (ii) the disaggregation of the micelles upon acidication. To this end, with a second experiment, cells were incubated with either RM or CM for three hours and then washed with PBS and observed using a confocal uorescence microscope. In this way, the actual localization of uorescence in the lysosomes and its temporal evolution (lysosome acidication is indeed a gradual process) 17 could be visualized. In order to allow localization, cell lysosomes and nuclei were stained with lysotracker red and Hoechst respectively, and the uorescence was monitored immediately aer washing, as well as aer 2.5 and 24 h incubation of the cells in micelle-free cell media ( Fig. 7 and S13 †). Fluorescence imaging revealed that 24 h incubation led to the appearance of punctuating green uorescence dots in the case of RM. Such uorescent signals were higher in number and intensity as compared to that of cells incubated with the control micelles (CM) and as compared to the initial time points for both RM and CM. In order to visualize the spatial overlap between the two uorescent signals, the red of the lysotracker and green of the disaggregated RM and colocalization analysis of the green and the red channel for each incubation time (time zero, 2.5 h and 24 h) was performed using ImageJ soware ( Fig. 7 and Table S2 †). As expected, in the case of CM, the disaggregation did not occur upon endocytosis even aer 24 h incubation and therefore no overlay of the green and red channels was observed. In contrast, upon 24 h cell incubation with RM, when the channels were combined, a high number of overlapping pixels were represented as white dots. Such results are compatible with the previously described micelle disassembly and therefore indicate that these pHresponsive micelles can be efficiently internalized by cells, that they can sense the different physiological pH values and that their disassembly, as a response to acidication, can be easily monitored owing to the uorescence "switch ON". This response to physiological environments is an interesting feature in nano-theranostics as it can be exploited to specically target diseased tissues and control spatiotemporal drug release from nanocarriers as well as optical responses. 42 3 Conclusions pH-responsive uorogenic NPs are of great interest in the eld of nanomedicine as they can contribute to the development of both novel therapeutics and diagnostic tools. In this work, we demonstrated that ROMP polymerization, which is particularly interesting because it allows the direct polymerization of functional monomers, can be used to generate pH-responsive uorogenic micelles. While such micelles are stable at physiological pH, they rapidly disassemble at pH 5 giving rise to uorescence "switch ON". Such a transition was further conrmed with in vitro experiments in a model cancer cell line, in which the drop in the pH along the endolysosomal pathway resulted in green uorescence activation. This proofof-concept work provides a powerful synthetic platform in which therapeutics of interest can be easily incorporated via direct polymerization using ROMP, allowing the development of novel nano-theranostic agents in which uorescence activation is accompanied by drug release at the targeted site.

Data availability
The data supporting the ndings of this study are available within the article and in the ESI. †

Conflicts of interest
There are no conicts to declare. Fig. 7 Confocal fluorescence microscopy of cells incubated with either RM or CM for 3 h. Cells were washed and the images were acquired immediately, as well as after 2.5 and 24 h. Disassembled micelles are visible in the green channel and lysosomal vesicles were stained with lysotracker red. Cell nuclei were stained with Hoechst (blue). The colocalization analysis of the green and the red channel is also illustrated: colocalization is represented as a white signal.