RESULTS AND DISCUSSION

Design and synthesis

To design a unimolecular dual-reporter probe that enables crosstalk-free multiplexed signal output, three specific attributes are considered. (1) It exhibits good water solubility in the intrinsic state, enabling its accumulation in the kidney. (2) It can be quickly activated by ROS to produce an accumulated signal that will not be cleared easily in a short time, thereby reducing the influence of frequent urination. (3) It releases a multiplexed signal with completely distinct metabolic behaviors, avoiding crosstalk between the same type of signal output to ensure signal retention in the kidneys and output in urine. Based on these considerations, we performed a retrosynthesis analysis before the synthesis of the probes DHU-AKI-1, 2, 3 (see Schemes S1 and S2). For example, DHU-AKI-3 included three main moieties: a PEG cluster, a CA chromophore, and a ROS-responsive fragment 7 that contained leucomethylene blue (LMB) and (2-amino-1,3-phenylene) dimethanol conjugated through a urea bond. The PEG cluster and CA were symmetrically linked to the benzyl alcohol of the ROS-responsive fragment. First, we synthesized CA as a main block in DHU-AKI-3 using compound 1 as the starting material and introduced a benzyl group using a classic substitution reaction to increase its hydrophobicity (Scheme S3). We obtained the highly ROS-sensitive key precursor 7 with multiple modifiable sites (benzyl alcohol) from our previous work [29,30]. To improve the reaction yield, 7 was symmetrically protected by silicon to give compound 8. Subsequently, a substitution reaction took place between the amino group of the CA dye and unprotected benzyl alcohol of 8 after it was activated with 4-nitrophenyl chloroformate to offer 15, in which the silane-protected group was subsequently removed to provide DHU-AKI-1 (Scheme S4). Subsequently, to improve the kidney-targeting performance, probes were conjugated to hydrophilic PEG with varying molecular weights, yielding DHU-AKI-2 and DHU-AKI-3. DHU-AKI-4 without CA was also synthesized as a reference compound (Scheme S5). The intermediates and final compounds were characterized by nuclear magnetic resonance (NMR) spectroscopy and high-resolution mass spectrometry (HR-MS) (Figs S20–S38).

In vitro characterization

We investigated the optical properties and response capabilities of the prepared probes by means of ultraviolet-visible (UV-vis) absorption, fluorescence spectroscopy, and PA methods in phosphate buffer (PB, pH 7.4, Fig. 2a). The ROS selectivity of various probes was initially evaluated by studying the fluorescence changes in reporters in the presence of various ROS (•OH, ONOO−, t-BuOO•, H2O2, TBHP, NO, O2•−, ROO•, and HOCl). As shown in Fig. 2b and Fig. S1, both DHU-AKI-1 and DHU-AKI-2 showed significantly enhanced fluorescence intensity at 686 nm only in the presence of hypochlorous acid (HOCl), and the fluorescence intensity gradually increased with increasing HOCl concentration. DHU-AKI-3, however, gave an enhanced fluorescence signal after a reaction with multiple ROS (ONOO−, H2O2, TBHP, NO, O2•−, and HOCl). The broad-spectrum ROS-activatability of DHU-AKI-3 improved its bioavailability for application in complex organisms. Using HOCl as the representative activator, the time-dependent response curves of these probes were tested. As shown in Fig. S2, the fluorescence intensity at 686 nm of DHU-AKI-2 and DHU-AKI-3 rapidly increased to a plateau within ∼10 s upon the addition of HOCl, which was far faster than that of DHU-AKI-1 (∼60 s). This rapid response was beneficial to detect ROS with a short half-life in living systems. Under pseudo-first-order conditions, the rate constants were 0.0349, 0.1942, and 0.1867 s−1 for DHU-AKI-1, DHU-AKI-2, and DHU-AKI-3, respectively (Fig. 2c). In addition, the calculated logP values of DHU-AKI-1, 2, and 3 are 7.03, 1.39, and − 1.43, respectively (see Table S2 in SI). These results demonstrated that DHU-AKI-3 exhibits good water solubility and is expected to promote its accumulation in the kidney. Therefore, considering the broad-spectrum ROS activation, rapid response, and excellent water solubility properties, we selected DHU-AKI-3 as the probe candidate for the following study.

Figure 2.: Evaluation of the optical properties and response capability of probes in vitro. (a) The chemical structure of probes. (b) The fluorescence intensity (FI) of reporters (10 μM in PB) at 686 nm after addition of various ROS, A–J: only reporter, H2O2 (80 μM), O2•− (80 μM), NO (80 μM), ROO• (80 μM), TBHP (80 μM), •OH (80 μM), t-BuOO• (80 μM), ONOO− (80 μM), and HOCl (40 μM). (c) Pseudo-first-order kinetic plot of reporters (10 μM) in the presence of equilibrium concentrations of HOCl. (d) The changes in fluorescence spectra of 10 μM DHU-AKI-3 before and after adding various concentrations of HOCl in PB (pH 7.4). (e) The changes in absorption (Abs) spectra of DHU-AKI-3 (10 μM in PB) upon adding different concentrations of HOCl. (f) PA intensity of DHU-AKI-3 with different concentrations (0, 5, 10 and 20 μM in PB) and 20 μM DHU-AKI-3 after incubation with HOCl (20 μM, +ROS group represents 20 μM HOCl) at 805 nm. Data are the mean ± s.d. for n = 3. (g) Stability test: the absorption intensity changes of DHU-AKI-3 (10 μM in PB) at 664 nm in different solutions of the time series (Cyan: MEM; Brown: 10% FBS; Pink: pH 7.4 PBS; 1–9: 0, 1, 2, 4, 6, 12, 24, 36, and 48 h; 10: after incubation with 25 μM HOCl. These data are the mean ± s.d.; n = 3). (h) Schematic illustration showing the molecular mechanism of DHU-AKI-3 through ROS activation. (i and j) TEM images of (i) DHU-AKI-3 (50 μM) without the addition of HOCl, (j) DHU-AKI-3 (50 μM) in the presence of HOCl (50 μM), and the insets show the color changes (TEM images were acquired at different magnifications to optimally visualize structural features; scale bars are indicated individually. Scale bars = 1 μm).

The fluorescence, absorption, and PA spectra in PB solution were measured to evaluate the properties of DHU-AKI-3. As shown in Fig. 2d, DHU-AKI-3 showed a negligible fluorescence signal in the absence of ROS. Upon the addition of various concentrations of HOCl; however, the maximum fluorescence emission at ∼686 nm increased in a concentration-dependent manner. The detection limit of the DHU-AKI-3 response to HOCl was 28.5 μM based on the 3_σ_/k method (Fig. S3). In addition, DHU-AKI-3 showed double absorption peaks of CA at 718 and 805 nm, but a new absorption peak at ∼664 nm appeared with the addition of HOCl, which indicates the release of the MB fluorophore (Fig. 2e). The release efficiency of MB was measured to be ∼49% (Fig. S4). Meanwhile, we observed a gradually increasing PA signal at 805 nm with the increasing concentration of DHU-AKI-3, and the PA intensity remained consistent before and after responding with HOCl (Fig. 2f). This property was favorable for real-time tracking of DHU-AKI-3 in vivo. To further evaluate the potential of DHU-AKI-3 for application in vivo, we performed anti-interference experiments by incubating DHU-AKI-3 with various physiologically relevant cations, anions and amino acids. The results showed that these species caused negligible interference with the optical response of the probe, confirming its strong resistance to biological interference under physiological conditions (Fig. S5). Even under high concentrations of various reductive species (glutathione, N-Acetyl-L-cysteine, glucose and formaldehyde), the probe still retained a reliable response (Fig. S6). The probe’s response performance under different pH conditions indicated that DHU-AKI-3 exhibited good intrinsic stability in varying pH solutions, and could produce significantly enhanced fluorescence at 686 nm over a physiological pH range (4–8) after the addition of HOCl (Fig. S7). The result further confirms the reliability of the probe under physiologically relevant conditions. Furthermore, we evaluated its stability by detecting its absorbance at 664 nm in different medium including phosphate buffer saline (PBS, pH 7.4), 10% fetal bovine serum (FBS), and FBS-free culture medium (MEM). As illustrated in Fig. 2g, DHU-AKI-3 alone displayed negligible absorption in those mediums even up to 48 h, indicating its good stability for biological applications. However, the addition of HOCl induced a pronounced absorption enhancement at 664 nm. These results demonstrated the desirable properties of DHU-AKI-3 in ROS response and crosstalk-free multiplexed signal output (Fig. 2h).

Release of precipitated CA indicator

To verify the precipitation strategy of ROS stimulated CA release, we performed HR-MS, transmission electron microscopy (TEM), and dynamic light scattering (DLS) analysis in an aqueous solution. The free DHU-AKI-3 could completely dissolve in water, benefiting from the modification of the PEG moiety (Fig. 2i and Fig. S8), which facilitated CA delivery to the kidneys and significantly enhanced its bioavailability in vivo. After DHU-AKI-3 reacted with ROS, HR-MS data indicated the formation of MB (m/z [M + H]+ calcd for C16H18N3S+: 284.1216, found: 284.1195) and CA (m/z [M + H]+ calcd for C28H28N4O3S2 : 533.1681, found: 533.1671) (Fig. S9). As expected, we observed the larger aggregates of CA upon ROS activation of DHU-AKI-3, and the hydrodynamic diameter significantly increased to 0.3–4 μm (Fig. 2j and Fig. S8). Furthermore, the threshold concentration of DHU-AKI-3 required for CA precipitation was systematically examined. As shown in Fig. S10, noticeable precipitation was observed when the concentration of DHU-AKI-3 was 20 μM, demonstrating the rapid signal accumulation capability of DHU-AKI-3 within a short timescale. These results confirm that DHU-AKI-3 is capable of preventing the rapid renal clearance of in situ signals, highlighting its promise for long-term monitoring of kidney injury in situ.

Recognition of renal tubular epithelial cell apoptosis

After confirming the ideal performance of DHU-AKI-3 in vitro, we investigated its capacity to diagnose injured renal cells using the human kidney proximal tubular epithelial cell line (HK-2) as the cell model. Initially, we evaluated the cytotoxicity of DHU-AKI-3 through the cell counting kit-8 assay. The results showed that the survival rate of HK-2 cells remained at approximately 100% even after incubating with 300 μM DHU-AKI-3 for 24 h, indicating that DHU-AKI-3 is nontoxic for HK-2 cells and thus suitable for biological applications (Fig. S11). Next, the cellular uptake of DHU-AKI-3 by HK-2 cells was investigated via confocal laser scanning microscope (CLSM) imaging. As shown in Fig. S12, CLSM images displayed no fluorescence signal in normal HK-2 cells after incubating with DHU-AKI-3 (10 μM) for 3 h. Upon treatment with HOCl (50 μM) for different times (15, 30, and 60 min), we observed a statistically significant increase in fluorescence signal, which was observed mainly in the cytoplasm, demonstrating that DHU-AKI-3 could be effectively internalized into HK-2 cells.

Considering that nephrotoxic drugs such as cisplatin are the dominant factors of kidney injury [33,34], we investigated whether DHU-AKI-3 could recognize drug-induced renal cell apoptosis. We treated HK-2 cells with different doses of cisplatin for 12 h and determined the apoptosis rate by flow cytometry using an Annexin V-FITC kit (Fig. 3a). As shown in Fig. 3b, the apoptosis rate of the HK-2 cells gradually increased with the increasing concentration of cisplatin (0, 300, 600, 1000 μM), which reached up to 81% apoptosis compared with 1.7% in normal HK-2 cells. We further studied the performance of DHU-AKI-3 to recognize apoptotic HK-2 using CLSM imaging. For normal HK-2 cells, no fluorescence signal was observed after incubating DHU-AKI-3 for 3 h (Fig. 3c and e). By contrast, apoptotic cells showed a gradually increased fluorescence signal with prolonged incubation time of DHU-AKI-3 (0, 0.5, 1, 2, 3, 5 h), reaching a plateau at 3 h. In addition, cells with different degrees of apoptosis showed obviously enhanced intracellular fluorescence signals after incubation of DHU-AKI-3 for 3 h (Fig. 3d and f). These results indicated that DHU-AKI-3 could effectively recognize apoptotic renal tubular epithelial cells. We further verified the CLSM imaging results by TEM images of apoptotic HK-2 cells (Fig. S13). Compared with the control cells treated with FBS-free medium, the DHU-AKI-3-treated cells showed precipitated CA dye in the cytoplasm, demonstrating that the MB fluorophore and CA indicator were released simultaneously. These results robustly verified that DHU-AKI-3 is a promising tool for two-pronged diagnosis of early AKI by integration of in situ and in vitro detection based on multiplexed signal output.

Figure 3.: Imaging of cisplatin-induced apoptosis of HK-2 cells. (a) Schematic diagram of a cellular experiment. (b) Flow cytometer analysis of HK-2 apoptosis using Annexin V-FITC apoptosis detection kit (b1: cells without any treatment; b2: cells stained with Annexin V-FITC/PI; b3–b5: cells treated with 300 μM, 600 μM, and 1000 μM cisplatin for 12 h, respectively, before staining with Annexin V-FITC/PI; b6: the percentage of apoptosis and dead cells in b1–b5, these data are the mean ± s.d.; n = 3). (c) CLSM imaging of apoptosis (treated with 300 μM cisplatin) and normal HK-2 cells after incubation of DHU-AKI-3 (10 μM) for different times. (d) CLSM imaging of HK-2 cells at different apoptosis levels (treated with 0 μM, 300 μM, 600 μM, and 1000 μM cisplatin) after incubation of DHU-AKI-3 (10 μM) for 3 h. (e) and (f) were the average fluorescence intensity in (c) and (d), respectively. Scale bars = 20 μm, λex = 633 nm. Values are the mean ± s.d. for n = 3, *p < 0.05, ***p < 0.001, NS: not significant.

Biodistribution and biosecurity evaluation

For further applications in vivo, we investigated the biodistribution and biosecurity of DHU-AKI-3. We constructed a cisplatin-induced AKI mouse model according to our previous report [35], and performed the fluorescence imaging of major organs in healthy and AKI mice after intravenous administration. At 45 min postinjection, the heart, liver, spleen, lung, kidney, and muscle were harvested for ex vivo imaging. For healthy mice, we did not observe any significant differences in fluorescence intensity after intravenous (i.v.) injection of saline and DHU-AKI-3, respectively (Fig. S14a and b). For AKI mice, however, we detected a significantly increased fluorescence intensity in the kidneys only after i.v. injection of DHU-AKI-3, compared with the saline group (Fig. S14c–e). To more intuitively visualize the biodistribution of the probes, 3D-PA imaging was further performed to compare the real-time distribution of DHU-AKI-3 and the control compound DHU-AKI-1 (lacking the PEG chain) in AKI mice. As shown in Fig. 4a and c, DHU-AKI-3 exhibited a significant PA signal enhancement in the renal region at 30 min post-injection, while the PA signal in the liver was barely detectable. Movies S1 and S2 recorded the real-time 3D-PA imaging of the above process, clearly presenting the spatial distribution of DHU-AKI-3. In contrast, strong PA signal accumulation of DHU-AKI-1 was observed in the left and right lobes of the liver, while the kidney exhibited negligible signal enhancement (Fig. 4b, c, and Movies S3, S4). It indicated that lipophilic DHU-AKI-1 was prone to being captured by the reticuloendothelial system (RES) [36,37]. These imaging results indicated that DHU-AKI-3 was accumulated and metabolized primarily in kidneys, which is due to its excellent water solubility.

Figure 4.: 3D-PA imaging of probes in vivo. (a) PA imaging of AKI mice after being treated with DHU-AKI-3 (10 mg/kg, 200 μL) at 0, 15, 30, 60, and 120 min post-injection. White scale bar: 10 mm, red scale bar: 5 mm. (b) PA imaging of AKI mice after being treated with DHU-AKI-1 (10 mg/kg, 200 μL) at 0, 15, 30, 60, and 120 min post-injection. White scale bars: 10 mm, red scale bars: 5 mm. (c) The PA intensity of mice’s kidneys and liver in (a) and (b), values are the mean ± s.d. for n = 3.

Accordingly, we further performed the blood routine analysis to evaluate the biosecurity of DHU-AKI-3. The major biochemical indicators of healthy mice after i.v. injection of DHU-AKI-3 for 24 and 48 h showed no significant changes compared with the control mice (Fig. S15). In addition, we did not observe any adverse effects on renal function after administration of DHU-AKI-3 for 24 h (Fig. S16). Histological staining of major organ tissues in the saline and DHU-AKI-3 groups showed no obvious tissue damage or lesions in the pathological sections of those mice (Fig. S17). These results demonstrated that DHU-AKI-3 possessed good biocompatibility and could be safely applied in vivo.

In situ long-term monitoring of early AKI

Encouraged by the excellent performance of DHU-AKI-3 in vitro, we further studied its feasibility for diagnosing early-stage AKI in mouse models. Ischemia-reperfusion (IR) injury is a common cause of AKI, especially in the field of renal transplantation [38]. Accurately diagnosing IR-induced kidney injury within a short time using conventional methods remains challenging, however. We constructed a unilateral kidney IR injury model to evaluate the capability of DHU-AKI-3 for detecting early-stage AKI (Fig. 5a). We subjected the left kidney of mice to ischemia for 45 min by clamping the kidney pedicle using a mini aorta clamp, followed by reperfusion. We then intravenously injected DHU-AKI-3 at 24 h postoperation and conducted a longitudinal ultrasound and PA imaging at different time points after the DHU-AKI-3 injection. As shown in Fig. 5b and c, the PA signal in the kidneys of the sham operation group (b-1) was close to the background (0 h) at different time points postinjection (0.5, 1, 2, 8, 12, 24, 48, 60, and 72 h). In addition, we did not observe any PA signal enhancement in the right kidney (normal) of the IR-induced AKI mice model (b-2). We detected, however, an obvious PA signal in the left kidney (IR injury, b-2) at 1 h postinjection of DHU-AKI-3, which showed a long retention time, approximately 2.6-fold higher than the right kidney (normal, b-2) at 60 h. Significantly, the blood biochemical analysis showed that the SCr value of IR mice was approximately 2-fold higher than the control and sham operation groups, whereas BUN and uric acid (UA) levels did not show any significant differences (Fig. 5d–f). This result indicated that conventional clinical methods were not able to accurately diagnose unilateral kidney IR injury. By contrast, DHU-AKI-3 enabled noninvasive, long-term monitoring of early-stage kidney injury with high specificity.

Figure 5.: Long-term monitoring of kidney injury in situ and the metabolic pathways of DHU-AKI-3 in the kidneys. (a) Schematic illustration of reporters for long-term in-situ monitoring of unilateral kidney (left) ischemia reperfusion injury (n = 3). (b) Representative PA images of mice after i.v. injection with the DHU-AKI-3 (10 mg/kg, 200 μL) at different time points (0, 0.5, 1, 2, 8, 12, 24, 48, 60, and 72 h), scale bars = 2 mm. (c) The dynamic PA intensity of the mice’s kidneys in (b). (d–f) SCr (d), BUN (e) and UA (f) analysis of the mice in different groups of (b), *p < 0.05, NS: not significant, n = 3. (g) Schematic illustration of renal metabolism of DHU-AKI-3 in the normal and injured kidney. (h) H&E staining images of the kidney from normal and AKI mice, scale bar = 50 μm. (i) CLSM images of whole kidney slices from normal and AKI mice with i.v. injection of DHU-AKI-3 (10 mg/kg, 200 μL), scale bars = 2 mm. (j) The enlarged view in R1 and R2 regions, respectively. T: renal tubule, G: glomerulus, scale bars = 20 μm. (k) TEM images of the kidney in normal and AKI mice injected with DHU-AKI-3 (10 mg/kg, 200 μL). White arrows indicate the precipitated CA, scale bars = 5 μm.

To validate the critical role of the activatable precipitation strategy in the long-term monitoring of kidney injury in situ, we selected the control compound DHU-AKI-4 (lacking the CA moiety) for comparative study. Compared with the saline group, the DHU-AKI-4 group showed weak PA signal enhancement of MB (690 nm, 1.5-fold) in the kidney only at 0.5 h post-treatment, which was rapidly cleared within 2 h (Fig. S18), indicating the rapid metabolism and short half-life of released MB in vivo [39]. In comparison, DHU-AKI-3 showed obvious advantages for in situ long-term kidney monitoring.

The metabolic pathways in the kidneys

Having validated that DHU-AKI-3 was metabolized primarily in the kidneys and could diagnose early-stage AKI, we further investigated the metabolic pathways of DHU-AKI-3 in the kidneys (Fig. 5g). Hematoxylin and eosin (H&E) staining of the renal tissue showed normal renal tubule and glomerulus morphology in the control group without cisplatin treatment (Fig. 5h). After nephrotoxic cisplatin administration, however, the injured renal tubular epithelial cells showed mild atrophy, loss of brush border, and even ballooning degeneration. The renal tubule is the kidney’s first line of defense against nephrotoxins [40,41]. Next, we dissected the kidneys at different time points after injection of DHU-AKI-3 for fluorescence imaging and TEM analysis. The highly hydrophilic DHU-AKI-3 was filtered out by the glomerulus and reabsorbed through renal tubules for kidney accumulation, and eventually metabolized into the urine through the normal kidney without enhanced fluorescence. Compared with the healthy control group, the fluorescence intensity of MB was enhanced only in the renal cortex of AKI mice (Fig. 5i). An enlarged view of the renal cortex (R1 and R2) clearly showed a strong fluorescence signal of MB in the renal tubular but not in the glomerulus, which illustrated that the renal clearable DHU-AKI-3 was effectively activated in the damaged renal tubules (Fig. 5j). Furthermore, TEM images showed large precipitation in the proximal tubular epithelial capillary cavity of the mice after DHU-AKI-3 injection (Fig. 5k). This phenomenon indicated the release of precipitated CA and further explained the in situ long retention time of the PA signal. These results robustly demonstrated that renal clearable DHU-AKI-3 was capable of diagnosing early-stage AKI through the integration of in situ and in vitro detection with crosstalk-free multiplexed output.

Integration of in situ and in vitro diagnosis by crosstalk-free multiplexed signal output

We studied the capability of DHU-AKI-3 for two-pronged diagnosis of early-stage AKI using a cisplatin-induced AKI mouse model. Different stages of AKI mice were intravenously injected with saline (control group) and DHU-AKI-3, respectively (Fig. 6a). We performed ultrasound/PA imaging at different time points after administration of DHU-AKI-3. For the healthy groups (I and II) without the cisplatin treatment, the PA signal in the kidneys remained consistent with the background throughout the tracking process (Fig. 6b). The same was true for the PA signal in the AKI mice after injection of saline (group III). For the early-stage AKI group (group IV), however, the PA intensity of CA rapidly increased in kidneys at 0.5 h after injection, which was ∼1.6-fold higher than the control time (0 h), further showing the long retention time of the CA signal in kidneys (∼60 h). Moreover, the PA signal in the kidneys of late-stage AKI mice (group V) followed a similar trend, but with stronger PA intensity (∼2.5-fold enhancement at 0.5 h postinjection) (Fig. 6c). Because more ROS was produced in the late-stage damaged kidneys, more abundant CA was released and precipitated in situ. These results indicated that DHU-AKI-3 achieved long-term monitoring of early-stage AKI in situ.

Figure 6.: Integration of in situ and in vitro detection by crosstalk-free multiplexed signal output. (a) Schematic illustration of a two-pronged strategy. (b) Representative PA images of different mice (I: healthy mice + saline; II: healthy mice + DHU-AKI-3; III: early AKI mice + saline; IV: early AKI mice + DHU-AKI-3; V: later AKI mice + DHU-AKI-3) after i.v. injected with saline or DHU-AKI-3 (10 mg/kg, 200 μL) at different post-treatment time points (0, 0.5, 1, 2, 8, 12, 24, 48, 60, and 72 h), scale bars = 2 mm. (c) The dynamic PA intensity of kidneys in (b). (d) Picture of urine collected at 3 h post-injection in (b). (e) Fluorescent images of the bladder in different groups of mice after i.v. injected with saline or DHU-AKI-3 at different post-injection time points (0, 0.5, 1, 2 and 3 h), and collected urine at 3 h post-injection, followed by fluorescence imaging (n = 6). The fluorescence intensity of (f) bladder and (g) urine in (e). (h) Schematic diagram of smartphone-assisted portable urinalysis in vitro. (i) The photographs of the test strip with different groups of urine were captured by the camera of a smartphone, and automatically translated into RGB values by the Color Recognition software. (j) The color intensity of pictures from test strips with different groups of urine.

To further confirm the integration diagnosis of DHU-AKI-3 for in vivo and in vitro detection with crosstalk-free duplex signal output, we collected the urine samples from each group (I–V) of mice at 3 h post-treatment of DHU-AKI-3. As shown in Fig. 6d, the urine color of the control groups (I–III) all showed yellow, whereas the urine color of groups IV and V changed from yellow to visible light green and dark green, respectively. Further HR-MS detection of urine verified the presence of MB (see Fig. S19 in SI). These results indicated that blue MB was excreted into the urine of groups IV and V, which were consistent with the PA imaging data in situ. This phenomenon of urine color changes verified the feasibility of portable urinalysis in vitro by colorimetric analysis. We then measured the MB signal in the bladder and urine of different groups by in situ fluorescence imaging and in vitro colorimetric methods (Fig. 6e). We found a negligible fluorescence signal of MB in the bladder of control groups I and II at different time points after injection with saline and DHU-AKI-3. Similarly, the fluorescence signal in urine at 3 h post-treatment showed no significant difference (Fig. 6f and g). Compared with the early-stage AKI mice (group III) after injection of saline, we detected a statistically increased fluorescence in the bladder of group IV in a time-dependent manner and showed a ∼3.5-fold enhancement in urine at 3 h postinjection of DHU-AKI-3. This high consistency of urinalysis in vitro and PA imaging in situ ensured the accuracy of urine diagnosis.

To achieve a more convenient detection application, we developed a smartphone-assisted colorimetric urinalysis in vitro for portable diagnosis of early-stage AKI (Fig. 6h). We collected urine samples from all mice of different groups at predetermined times and dripped the samples on the surface of white test strips. The strips showed obvious color changes that were visible to the naked eye, thus facilitating visual real-time monitoring in vitro. We further combined the camera of the smartphone and the Color Recognition software to realize convenient detection. The pictures of the test strip were recognized by software that automatically translated the results into R (Red), G (Green), and B (Blue) values. We calculated the color intensity of pictures according to the formula I = 0.3_R_ + 0.59_G_ + 0.11_B_ [42]. As shown in Fig. 6i and j, groups I–III showed higher color intensity (I), whereas the I value of urine from groups IV and V was statistically significantly decreased, which indicated the excretion of blue MB from the kidney into urine. Accordingly, the mutual authentication of duplex optical signal output (PA signal of CA in situ and fluorescence signal of MB in bladder) validated the accuracy of portable urinalysis in vitro. Thus, the unimolecular dual-reporter probe DHU-AKI-3 could accurately diagnose early-stage AKI by integration of in situ long-term monitoring and portable urinalysis in vitro with crosstalk-free multiplexed output. Overall, this method is a promising tool for visual analysis of the AKI state directly in the urine, matching the demands of future point-of-care testing.