A probe for NIR-II imaging and multimodal analysis of early Alzheimer’s disease by targeting CTGF

To date, earlier diagnosis of Alzheimer’s disease (AD) is still challenging. Recent studies revealed the elevated expression of connective tissue growth factor (CTGF) in AD brain is an upstream regulator of amyloid-beta (Aβ) plaque, thus CTGF could be an earlier diagnostic biomarker of AD than Aβ plaque. Herein, we develop a peptide-coated gold nanocluster that specifically targets CTGF with high affinity (KD ~ 21.9 nM). The probe can well penetrate the blood-brain-barrier (BBB) of APP/PS1 transgenic mice at early-stage (earlier than 3-month-old) in vivo, allowing non-invasive NIR-II imaging of CTGF when there is no appearance of Aβ plaque deposition. Notably, this probe can also be applied to measuring CTGF on postmortem brain sections by multimodal analysis, including fluorescence imaging, peroxidase-like chromogenic imaging, and ICP-MS quantitation, which enables distinguishment between the brains of AD patients and healthy people. This probe possesses great potential for precise diagnosis of earlier AD before Aβ plaque formation.

1. Authors need to provide direct evidence that the observed signal differences in the WT and APP/PS1 mouse brain result from varying CTGF concentrations, for example using ELISA.In suggested experimental condition, is the amount of CTGF in the AD model brain samples sufficiently higher than in the WT? Do these values correlate quantitatively or qualitatively with the fluorescence response obtained using DGC? 2. Authors explained that the developed DGC nanoparticle has high BBB penetrability.However, the lack of experimental results characterizing this BBB penetrability hinders the evaluation of the study's interpretations and main claims.Merely citing previous literatures is inadequate for a proper assessment of the experiments and their outcomes.Therefore, direct evidence providing BBB penetrability, such as PAMPA assay is necessary.In this context, is there no difference between BBB penetrability in the in vivo imaging of WT and APP/PS1 mice?Typically, the BBB in AD model transgenic mouse is expected to be compromised.3.In vivo stability and/or pharmacokinetic studies of DGC as a potential NIR-II imaging agents should be included in this manuscript.Additionally, as an initial study of CTGF measurement in early AD brain samples, authors should demonstrate results from narrow time points, not only 3and 9-month-old, to provide robust evidence that CTGF-induced NIR signals from DGC are detectable at the early stages of AD. 4. In the human brain sample imaging results, portions stained by DGC are shown in unspecified extracellular areas.Authors should clarify why this observation contrasts with the results from cell experiments.5.The manuscript briefly compares DGC with Aβ in Figure 5, showing significant co-localization in the areas of Aβ deposition.This could potentially interfere with accurate CTGF concentration measurements.Although the authors assert that the DGC assay is suitable for early AD diagnosis, the study only includes non-demented (healthy control) and advanced AD pathology patient samples, which do not represent early-onset disease.To substantiate claims of early diagnosis using patient samples, imaging studies should focus on stages where CTGF is expressed prior to Aβ accumulation.6.The introduction is too brief to provide adequate context for the study.Authors should expand on the background information to explain the development of their strategy and its relevance to early AD diagnosis.Furthermore, the results and discussion sections are overly concise.A more detailed description of the finding and comprehensive discussion is required.
Reviewer #3: Remarks to the Author: In this work, the authors developed a CTGF-targeting peptide-coated gold nanocluster (DGC) that can emit both NIR-II and red fluorescence, allowing non-invasive NIR-II imaging of CTGF when there is no appearance of Aβ plaque deposition.Basically, the DGC probes were well designed and performed with a variety of characterizations.The authors also demonstrated the imaging potential of DGC probes in vitro and in vivo by some studies with analyzed data.The work has been solidly done, below is a list of technical issues for the authors to consider when revising their manuscript: 1.For the preparation of DGC probes, how to control the ratio of HAuCl4 and DAG peptide? 2. The stability and quantum yield of DGC probes should be characterized.3. The co-localization coefficient should be calculated in Figure 2d.4. The scale bar should be provided in Figure 3a and Figure 4b.In addition, the semiquantitative analysis should be provided in Figure S7. 5. Since the authors claimed that the DGC probes can imaging CTGF in NIR-II window, what is the advantage of NIR-II imaging in this work?More data on NIR-II imaging should be provided.6.The in vivo biosafety of DGC probes should be characterized by H&E staining and biochemical analysis.7. How accurate or threshold can DGC be in detecting CTGF compared with currently reported probes?

REVIEWER COMMENTS
Reviewer #1 (Remarks to the Author): In this manuscript, Lu et al. reported a novel gold cluster-based probe modified with peptides that specifically bind to connective tissue growth factor (CTGF).Owing to the elevated expression of CTGF is an upstream regulator of amyloid-beta (Aβ) plaque, the probe holds a promise for potential application in the early diagnosis of Alzheimer's disease (AD).The unique feature of this probe is their capacity to emit near-infrared (NIR) light, rendering them suitable for in vivo imaging.Moreover, their ability to target CTGF could potentially make them superior to probes that target Aβ plaques.
The article is well-crafted, and the experiments convincingly highlight the distinctive properties of the newly developed probe.I would suggest accepting it after the following concerns are addressed.Considering that the elimination half-life (t1/2z) of DGC in mice is about 20 h, we infer that DGC maintains a stable structure throughout the diagnostic testing time window.
These results demonstrate that DGC has good stability in vitro and in vivo.Response: BIAcore 8K+ instrument (Pharmacia Biosensor AB) was used to detect the affinity between DGC probe or DAG peptide and recombinant human CTGF protein, and BIAcore Insight Evaluation 3.0 software was used for the real-time binding and kinetic analysis, as described in the METHODS section.The concentration of DGC was initially determined by ICP-MS quantification of Au.However, the affinity to CTGF should be compared as a complete molecule.Result of MALDI-TOF-MS indicated the precise molecular formula of DGC can be derived as Au26DAG8.Thus, the accurate concentration of DGC molecules is obtained through dividing the Au concentration by 26.The obtained KD of DAG and DGC was 2.23 × 10 -5 M and 2.19 × 10 -8 M, respectively, which is almost 1000 times.The much stronger affinity of DGC may be due to the multivalent (eight) display of DAG peptides on the surface of the probe, which is a well-established theory for enhancing target affinity (J.Am.Chem.Soc. 2020, 142, 4800-4806;Bioconjugate Chem. 2022, 33, 1922-1933;Adv. Sci. 2022, 9, 2103098;ACS Chem. Biol. 2023, 18, 1066-1075).Response: According to previous reports, the expression pattern of CTGF was slightly different in several AD mouse models.The homing of DAG peptide was mainly seen in the cortex of the Tg2576 mice and in the hippocampus of the hAPP-J20 mice (Nat. . 2017, 8, 1403).A marked increase of CTGF levels in the brains of APP/PS1 transgenic mice were also observed, and is mainly located in the cortical (Mol. Neurobiol. 2012, 45, 440-454;Hum. Mol. Genet. 2017, 26, 3909-3921).In this study, DGC signals were observed in both the hippocampus and the cortex of 9-month-old APP/PS1 mice, and showed more obvious in the cortex than the hippocampus, as shown in Figure 3a  Reviewer 1: Does the sensitivity of in vivo imaging in the NIR-II region differentiate between 3-month-old and 9-month-old AD mice?

Commun
Response: To answer this question, we performed in vivo and ex vivo NIR-II imaging of 3-month-old and 9-month-old APP/PS1 mice after intravenous injection of DGC, simultaneously.Result showed that the overall NIR-II fluorescence intensity in brain of 9-month-old APP/PS1 mice is obvious higher than that of 3-month-old mice, both in vivo or ex vivo (Fig. R9), indicating that DGC probe has the ability and sensitivity to distinguish between early and late-stage AD mice.Response: We speculate that large lesions may be more easily labeled by both antibodies and probes, so there will be better co-localization.From the overall fluorescence pattern, it can be found that the 9-month-old brain sections have more extensive CTGF expression than the 3-month-old brain sections, and the labeling efficiency of antibody and DGC probe may be affected, that resulting in slightly less effective of co-localization in small lesions.However, it was clear that CTGF expression was higher in the brain of 9-month-old AD mice than 3-month-old ones.
Reviewer 1: The authors emphasized the uniqueness of the NIR-II emission, I think it is necessary to add some discussions of the emission mechanism.
Response: Understanding the exact origins of NIR-II photoluminescence (PL) of gold nanoclusters (AuNCs) is of great significance for the development of highly effective NIR-II probes.However, the mechanism is very complicated and still under debate (Chem. Soc. Rev. 2019, 48, 2422-2457).So far, only Au25 clusters have been systematically studied for the correlation of PL origins to their structures, through experimental characterizations and theoretical calculations (J.Phys.Chem.Lett. 2021Lett. , 12, 1514Lett. -1519)).By examining different surface ligands, it was shown that the NIR emission of Au25 cluster was likely originated from the Au13 icosahedral core state (J. Phys.Chem.Lett. 2021Lett. , 12, 1514Lett. -1519)).It was found that the NIR-II PL of AuNCs can be enhanced by increasing the rigidity of surface ligands (J.Am.Chem. Soc. 2015, 137, 8244-8250), and the intrinsic structural rigidity also strongly affect the NIR PL quantum efficiency of thiolated AuNCs (J.Am.Chem. Soc. 2020, 142, 12140-12145).
Recent studies indicated the great significance of the central Au atom and the surface "lock rings" as well the surface "lock atoms" in enhancing the NIR-II PL of AuNCs by suppressing the nonradiative decay (ACS Nano. 2021, 15, 16095-16105;Small. 2021, 17(11), 7).In summary, the NIR-II PL emission of AuNCs is related to the Au core and the surface ligands as well the charge transfer states, which leads to the complexity of their fluorescence origin (Coord.Chem.Rev. 2021, 448, 214184).
Reviewer #2 (Remarks to the Author): The manuscript titled "A probe for NIR-II imaging and multimodal analysis of early Alzheimer's disease by targeting connective tissue growth factor" presents a novel peptide-coated gold nanocluster, refereed to as DGC, which serves as a NIR-II emissive probe for detecting CTGF in the cortex of early AD mice and human brains.Authors well demonstrated the required analyses and biological experiments using the proposed materials.However, additional critical experiments are essential to substantiate the suitability of DGC nanoparticle for early diagnosis of AD, to be published in Nature Communications.Thus, I do not agree with acceptance of the current manuscript with the following reasons.
Reviewer 2: Authors need to provide direct evidence that the observed signal differences in the WT and APP/PS1 mouse brain result from varying CTGF concentrations, for example using ELISA.In suggested experimental condition, is the amount of CTGF in the AD model brain samples sufficiently higher than in the WT? Do these values correlate quantitatively or qualitatively with the fluorescence response obtained using DGC?
Response: Thanks for the valuable suggestion.According to previous reports, the elevated CTGF in the brain of APP/PS1 mice was mainly secreted by the reactive astrocytes and was located primarily in the cortex rather than uniformly distributed throughout the brain (Mol. Neurobiol. 2012, 45, 440-454;Hum. Mol. Genet. 2017, 26, 3909-3921).In order to more accurately characterize the expression pattern in AD brain, we used immunofluorescence staining to detect CTGF expression in the brain sections of APP/PS1 mice from 1-month to 9-month-old according to the editor's and Reviewer's suggestion, and the age-matched WT mice were used as controls.The results of immunofluorescence staining of CTGF antibody showed that the expression level of CTGF in the cortex of APP/PS1 mice gradually increased with the increase of age, with a significant linear correlation, while the expression level in the brain of agematched WT mice was significantly lower (Figure R1).To evaluate whether the amount of CTGF in the AD brain samples sufficiently higher than that in age-matched WT brain in the same experimental condition, we compared the expression level of CTGF in the brain lysates of early-stage (1-and 2-month-old) mice by ELISA.The ELISA quantitative results showed that the expression of CTGF in APP/PS1 mice brain was significantly higher than that of WT mice at the same age (Fig. R10).To accurately evaluate the correlation between DGC probe and CTGF expression, DGC was simultaneously labeled on the brain sections that immunofluorescence stained by CTGF antibody (Fig. R1).Results showed that DGC and CTGF antibody were well colocalized in the brain sections, and there was a linear correlation with increasing CTGF expression.In addition, we performed ex vivo NIR-II imaging and luminescence semiquantification of the brains of APP/PS1 mice aged from 1-to 9-month after DGC probe injection.Results showed that with the increase of age of APP/PS1 mice, the NIR-II luminescence intensity of DGC in the brain became stronger and stronger, and the semiquantitative results showed a linear correlation (Fig. R11), which was consistent with the results of immunofluorescence analysis of brain slices.Reviewer 2: Authors explained that the developed DGC nanoparticle has high BBB penetrability.However, the lack of experimental results characterizing this BBB penetrability hinders the evaluation of the study's interpretations and main claims.
Merely citing previous literatures is inadequate for a proper assessment of the experiments and their outcomes.Therefore, direct evidence providing BBB penetrability, such as PAMPA assay is necessary.In this context, is there no difference between BBB penetrability in the in vivo imaging of WT and APP/PS1 mice?Typically, the BBB in AD model transgenic mouse is expected to be compromised.
Response: According to the suggestion, PAMPA assay was performed and the test concentration of DGC was determined by reference to the blood concentration.Results showed that the permeability Mean Pe of DGC is higher than 1.5 * 10 -6 cm/s at several concentrations, and the Pe value of 40 μM DGC is 2.15 ± 0.44 * 10 -6 cm/s, indicating a high BBB permeability (Table R2).On the other hand, before the ex vivo NIR-II imaging of AD brains, DGC probes that did not cross BBB were removed by cardiac perfusion with saline.Therefore, the significant NIR-II luminescence within the brains come from the probes that have already crossed the BBB, demonstrating that DGC can effectively penetrate the BBB.
Indeed, studies have shown that when the disease is fully developed (9 months of age) in AD mice, the permeability of the BBB increases, that perhaps increasing DGC leakage.However, there was no significant difference in BBB permeability between AD and WT mice in early stage (Nat. Commun., 2017, 8, 1403).Therefore, we compared the brains of APP/PS1 mice from 1-to 3-months of age with the age-matched WT mice using ex vivo NIR-II imaging.Results showed that AD brains exhibited significantly stronger DGC NIR-II luminescence than that of age-matched WT brains, as early as 1 month of age, when BBB had not yet been significantly leaked in this context (Fig. R2).These results suggest that the increase of DGC in AD brains is not caused by BBB leakage, but is related to CTGF expression.The pharmacokinetic study of DGC was performed according to the request.Detailed parameters are attached below (Table R3).Results showed that the elimination half-life (t1/2z) of DGC in male and female mice was 21.38 h and 19.80 h, respectively, indicating a longer metabolic time than small molecule probes, which is more suitable for in vivo diagnostic applications.To demonstrate the detectable of DGC in brains at the early stages of AD, APP/PS1 mouse in more narrower time points (1-, 2-and 3-month-old) were detected by in vivo and ex vivo NIR-II imaging, and the age-matched WT mouse were used as controls.In vivo imaging results showed that the brain fluorescence intensity of AD mice was higher than that of age-matched WT mice (Fig. R13).Considering that black hair on the body surface and the residual probe in cerebrovascular may affect the accuracy of in vivo imaging, ex vivo analysis of brains after cardiac perfusion with saline can provide more reliable evidence.The ex vivo NIR-II imaging clearly showed that DGC fluorescence in AD brain was significantly enhanced compared with that in age-matched WT brains, and the difference was significant even at 1 month of age (Fig. R2).Response: CTGF is a secretory protein that overexpressed by reactive astrocytes in the context of AD, which will be secreted to the extracellular region proximity to reactive astrocytes.Therefore, DGC staining in extracellular areas can be observed in the brain section, in addition to astrocyte cell bodies (Nat. Commun., 2017, 8, 1403;BMC Neurosci., 2003, 116(1), 1-6).However, staining in cells may lose extracellular secretory portions due to cell fixation and rinsing steps, similar results can be found in previous reported work (Nat. Commun., 2017, 8, 1403).Response: According to the suggestion, we obtained another two brain slices of AD patients at a very early stage according to the Thal standard and two healthy control samples from the Department of Neurology, Xiangya Hospital, Central South University (Table R1).The two early AD brain slices only observed very little Aβ accumulation in the neocortex and were therefore clinically identified to be Thal I stage by Xiangya Hospital, Central South University (Neurology. 2002(Neurology. , 58, 1791(Neurology. -1800;;Mol. Neurodegener. 2019, 14, 32).What's exciting is that the fluorescence signal of DGC probe and CTGF antibody in these two brain slices was strong and significantly different from that of the healthy controls.The quantitative results of ICP-MS also verified the significant difference.Moreover, the DGC probe (red) can co-locate well with CTGF immunofluorescence staining (green) on the AD brain slices (Fig. R3).
These results suggest that DGC probe can identify AD brain prior to obvious Aβ accumulation, with a potential for early diagnosis of AD patient.Response: To establish and optimize of the synthesis procedure of DGC, we explored the ratio of precursor and reaction condition based on the ratio of Au atom to DAG peptide, and determined the optimal ratio and reaction kinetic parameters by detecting the fluorescence intensity of the product (Fig. R14).The optimal ratio of Au atom to DAG peptide was finally determined to be 1:1 in this study.The absolute quantum yield (QY) of DGC probe was measured to be ~0.7% using an integrated sphere technique, which was comparable to other reported gold nanoclusters for biological applications.Although this QY is not high, but sufficient to produce a bright signal in the AD mouse brain for detection in vivo in this study.
Reviewer 3: The co-localization coefficient should be calculated in Figure 2d.
Response: Thanks for the suggestion, and the co-localization coefficient in Figure 2d was calculated by Image J software.The co-localization coefficient (Pearson's correlation coefficient) of DGC and CTGF antibodies in Figure 2d was 0.946 (Fig. R15).

Reviewer 1 :
How long can the fluorescence last in the cells and AD brain, or how stable is the fluorescence of the probe?Response: We monitored the fluorescence changes of DGC probes in vitro, in U87MG cells and in brain of APP/PS1 mice, respectively.Results showed that the fluorescence intensity of DGC decreased only slightly in vitro after two weeks storage (Fig R4).In cultured U87MG cells, DGC probe could be well internalized, and obvious fluorescence could still be observed in the cells after withdrawal of DGC-containing medium for 72 h (Fig R5).In APP/PS1 mice, we detected temporal changes of luminescence from DGC in the brain by in vivo NIR-II imaging.Results showed that the NIR-II fluorescence was still observed in the brains 24 h after injection (Fig R6).

Figure R4 .
Figure R4. in vitro stability of DGC after two weeks storage in PBS.

Figure R6 .
Figure R6.Stability of DGC in AD brains.Monitoring fluorescence intensity of DGC in brain of WT mice and 3-month-old APP/PS1 mice for 24 h (left).In vivo and ex vivo NIR-II imaging of brains after DGC injected for 24 h (right).

Reviewer 1 :
How are DGC or DAG and CTGF dissociation equilibrium constants (KD) calculated and compared?If the protein of peptide DAG targets CTGF, what is the reason for the 1000-fold increase in the KD values after the formation of probe DGC?

Reviewer 1 :
In Fig 2c, the fluorescence intensity of U87 and SY5Y seems similar.Please clarify this and provide how to calculate the FI in the methods.Response: The analysis of fluorescence intensity in cells in Fig S2a is the average fluorescence intensity in each cell, rather than the fluorescence density at a single site, and the detailed analysis methods have been added to the revised methods section.The fluorescence intensity of each cell was measured by Image J software, the red channel was extracted by clicking Split Channels, and the red fluorescent region except the nucleus was selected by Freehand selections.Then the average fluorescence intensity in this region was measured and recorded for statistical analysis.In Fig 2c, the fluorescence intensities of U87MG and SY5Y seems similar, possibly due to the different morphological sizes of the two cell lines.U87MG cell has a larger nucleus and cell body than SH-SY5Y cell, and also showed more extensive CTGF expression, but the number of cells in the same field of view is significantly less than SH-SY5Y cell, maybe resulting in a similar perception of fluorescence intensity.Reviewer 1: The ICP-MS of the whole brains from DGC-treated normal or APP/PS1 mice should be quantified, not only brain sections.Response: According to the suggestion, the whole brain samples collected after ex vivo NIR-II imaging in Fig. 4 were frozen grinded and digested, and the content of Au in the whole brain was measured by ICP-MS.The results showed that the content of Au in the brain of APP/PS1 mice was more than twice that of the age-matched WT mice (Fig R7), which was consistent with the results of the NIR-II imaging.
. However, in the brains of 3-month-old APP/PS1 mice, we only observed significant DGC signals in the cortex, but almost none in the hippocampus, as shown in below (Fig R8, hippocampus with DAB color).Therefore, we only showed DGC staining in the cotex in Figure 3b.

Figure R9 .
Figure R9.The in vivo NIR-Ⅱ images of WT mice (left), 3-month-old APP/PS1 mice (middle) and 9-month-old APP/PS1 mice (right) after intravenous injection of DGC for 8 h.And the ex vivo imaging of brains isolated from the mice (cardiac perfusion with saline, λ ex = 808 nm).

Figure R1 .
Figure R1.Immunofluorescence staining of CTGF and DGC probe on APP/PS1 mouse brain sections and the age-matched wild-type (WT) brain sections from 1-to 9-month-old.

Figure R6 .
Figure R6.Stability of DGC in AD brains.Monitoring fluorescence intensity of DGC in brain of WT mice and 3-month-old APP/PS1 mice for 24 h (left).In vivo and ex vivo NIR-II imaging of brains after DGC injected for 24 h (right).

Figure R12 .
Figure R12.Serum fluorescence monitoring and ICP-MS quantification of DGC in serum after tail vein injection.

Figure R13 .
Figure R13.The in vivo NIR-Ⅱ images of APP/PS1 mice (right) and age-matched WT mice (left) from 1-month to 3-month-old after intravenous injection of DGC.

FAM-
DAG labeled human brain sections of AD patient (left) and hiPSC-derived BMECs cells (right) from Nat. Commun., 2017, 8, 1403 Reviewer 2: The manuscript briefly compares DGC with Aβ in Figure 5, showing significant co-localization in the areas of Aβ deposition.This could potentially interfere with accurate CTGF concentration measurements.Although the authors assert that the DGC assay is suitable for early AD diagnosis, the study only includes non-demented (healthy control) and advanced AD pathology patient samples, which do not represent early-onset disease.To substantiate claims of early diagnosis using patient samples, imaging studies should focus on stages where CTGF is expressed prior to Aβ accumulation.

FigureReviewer 2 :
Figure R3.(a) The fluorescence imaging of cortex regions in brain slices from early stage of AD patient (without obvious Aβ accumulation) and Healthy control (HC) after staining with DGC (red fluorescence) and FITC labeled-CTGF antibody (green fluorescence).The merge pictures show the colocalization of DGC and CTGF.Scale bar = 50 μm.(b) The fluorescence intensity of DGC in

Figure R14 .
Figure R14.Optimization of the synthesis procedure of DGC by regulating precursor ratio, reaction temperature and reaction temperature.

Reviewer 3 :
The stability and quantum yield of DGC probes should be characterized.Response: The stability of DGC in vitro, in cells and in APP/PS1 mice were determined respectively.Results showed that the fluorescence intensity of DGC decreased only slightly in vitro for two weeks storage (Fig R4).In cultured U87MG cells, DGC probe could be well internalized, and obvious fluorescence could still be observed in the cells after 72h incubation (Fig R5).In APP/PS1 mice, we detected temporal changes of DGC in the brain by in vivo NIR-II imaging.Results showed that NIR-II fluorescence was still observed in the brains 24h after injection (Fig R6).Considering that DGC has a half-life of about 20h in mice, we infer that DGC maintains a stable structure throughout the diagnostic testing time window.These results demonstrate that DGC has good stability for potential AD diagnostic applications.

Figure R4 .
Figure R4.In vitro stability of DGC after two weeks storage in PBS.

Figure R6 .
Figure R6.Stability of DGC in AD brains.Monitoring fluorescence intensity of DGC in brain of WT mice and 3-month-old APP/PS1 mice for 24 h (left).In vivo and ex vivo NIR-II imaging of brains after DGC injected for 24 h (right).

Table R2 .
The effective permeability coefficient of DGC

Table R3 .
Values of parameters of the Pharmacokinetic of DGC in C57/6J mice.

Table R1 .
The detailed information of brain tissue donors