Cardiac troponin detection in one drop of saliva or urine for 10-minute assessment of acute myocardial infarction

Abstract

Acute myocardial infarction (AMI) can be fatal with sudden death, heart failure, and cardiac rupture. However, it is a challenge to achieve point-of-care testing (POCT) for cardiac troponin T (CTnT), an acute myocardial infarction biomarker, using conventional ELISA because it requires professional expertise and involves time-consuming steps. Herein, we propose a printed immunosensing photonic crystal biochip (PCB), which enables the sensitive, convenient, and non-invasive detection of CTnT. The fluorescence enhancement property of the PCB enables detection of CTnT within 10 minutes, with a limit of detection of 0.25 pg mL⁻¹, compared to the conventional ELISA (2.5 h, 30 pg mL⁻¹). Owing to the hydrodynamic enrichment of the PCB, 10 μL of sample can be measured, minimizing interference from impurities in body fluids and human sampling errors. We used 100 clinical samples, including serum, saliva, and urine, to verify the reliability of the PCB assay; compared with ELISA, the PCB assay exhibited consistencies of 0.998, 0.999, and 0.999, respectively. Notably, monitoring the saliva of a patient with an AMI history using the PCB assay enables rapid therapeutic intervention within 30 minutes. The PCB was first used for non-invasive body fluid testing to provide reliable results for clinical diagnosis, with the potential to provide AMI health promotion.

---

New concepts

We introduce a printed immunosensing photonic crystal biochip (PCB) that enables rapid, non-invasive detection of cardiac troponin T (CTnT), a critical biomarker for acute myocardial infarction (AMI), by integrating photonic crystal fluorescence enhancement and microfluidic enrichment to create a novel point-of-care testing platform. Unlike conventional ELISA assays requiring 2.5 hours, professional operation, and a detection limit of 30 pg mL⁻¹, the PCB achieves CTnT quantification in 10 minutes with 0.25 pg mL⁻¹ sensitivity, leveraging hydrodynamic enrichment to minimize sample volume (10 μL), reduce impurity interference, and support non-invasive testing using saliva and urine, with clinical validation across 100 samples showing 0.998–0.999 consistency with ELISA. This design introduces a bioresponsive material system that bridges optical signal amplification and fluidic engineering, offering a template for next-generation biosensors demanding high sensitivity, miniaturization, and non-invasive sampling while expanding materials science into interdisciplinary territory by merging photonics, nanotechnology, and clinical diagnostics to address unmet medical needs.

Introduction

Acute myocardial infarction (AMI) is a serious health threat due to its high morbidity, mortality, and recurrence rates. AMI is one of the leading causes of morbidity and mortality in Europe, Asia, Australia, and North America, with more than 15 million deaths per year worldwide.¹ A prospective study of Chinese patients with acute myocardial infarction centered on the assessment of adverse cardiac events showed that the recurrence rate of infarction in such patients within one year of discharge was 2.5%, with mortality rates in patients with early recurrence being as high as 53.5% within one year.² The duration of 120 minutes after the onset of acute myocardial infarction, especially the first 60 minutes, is the prime time for treatment. Treatment beyond 12 hours has limited efficacy.³ Even with systematic interventions, the readmission rate of patients with acute myocardial infarction within 30 days of discharge is still as high as 6.3%,⁴ and the mortality rate within one year ranges from 28.0% to 32.1%.⁵ From 2008 to 2020, the cause-of-death surveillance system of the China disease surveillance system (CDSS) reported a total of 710 187 deaths of cardiovascular diseases, and the ratio of out-of-hospital deaths to total deaths was 77.13%.⁶ Therefore, timely and accurate long-term monitoring of acute myocardial infarction at home or POCT is important.

Studies have shown that myocardial cells suffer irreversible damage 20 to 40 minutes after an acute myocardial infarction when specific enzymes or proteins like cardiac troponin (CTn), myoglobin (Myo), and creatinine kinase (CK) appear in the blood. Changes in the concentration of markers of myocardial injury have been significantly associated with the extent and severity of myocardial necrosis and its prognosis.⁷ In 2007, the UDMI, a consortium of the European Society of Cardiology, the American College of Cardiology Foundation, the American Heart Association, and the World Heart Federation, was formed to reach a consensus on diagnosing myocardial infarction. The UDMI defines myocardial injury based on an elevation of the 99th percentile upper reference limit of cardiac troponin in a normal reference population.⁸ To determine whether CTn concentrations are increasing or decreasing, serial CTn assays are required to differentiate between acute and chronic myocardial injuries. Chronic myocardial injury is characterized by a steady increase or minimal change in the CTn concentration during serial measurements, whereas acute myocardial injury is characterized by a dynamic increase in CTn concentration, which, with and without the features of myocardial ischemia (i.e., acute nonischemic myocardial injury), is referred to as acute myocardial infarction.^9,10 CTnT, a type of troponin CTn, is considered a more specific and concrete clinical marker of acute myocardial infarction.¹¹ Therefore, the introduction of CTn in clinical practice can help in the early diagnosis and care of myocardial infarction in patients with chest pain,¹² and in the rapid elimination of myocardial infarction.¹³

Over the past decade, cardiac troponin (CTn) assays have become increasingly sensitive, identifying more patients with previously undetected acute myocardial infarction (AMI) than ever before¹⁴ and playing a crucial role in risk prediction, prognostic assessment, cardiac-specific diagnosis, and therapeutic decision-making in AMI.¹⁵ Currently, enzyme-linked immunosorbent assay (ELISA) and radioimmunoassay kits belong to the two categories of technological tools for detecting biomarkers associated with acute myocardial infarction. However, this method has many drawbacks, as it relies heavily on the multi-step processing of a large number of samples, the use of large hematology analyzers, and a long diagnostic time to ensure the stability and accuracy of the analysis, which is costly and requires the operation of trained healthcare professionals, as well as complex instrumentation.¹⁶ In addition, other traditional methods such as electrochemiluminescence,¹⁷ electromagnetic methods,¹⁸ and colorimetric analysis¹⁹ have solved some of the obvious problems of ELISA and radioimmunoassay; however, in addition to the shortcomings of these techniques, they are not well suited for integrated microfluidic device fabrication, poor stability, and chemical modifications.²⁰ In addition, these techniques have some limitations, such as the need for the invasive collection of blood samples, the psychological stress of frequent sampling on patients, and the need for the assistance of healthcare professionals, which undoubtedly increases the cost of the test and the administrative workload. Nowadays, body fluids such as saliva,²¹ sweat,^22,23 urine,²⁴ and tears²⁵ are emerging as non-invasive alternatives for clinical self-diagnosis as well as routine testing.²⁶ In addition, previous studies have reported that the levels of cTn in serum are higher than those in saliva, and there is a strong consistency.^27,28 A significant correlation has been found between the expression level of cardiac troponin T (cTnT) in saliva and myocardial injury in athletes.^29,30 Currently, there is no reported evidence whether cTn exists in urine samples. However, the concentration of CTnT in saliva is lower than that in serum, and the ability to concentrate samples by traditional assays is limited, as is their sensitivity. Therefore, the development of a noninvasive assay that is both sensitive and specific, and that can rapidly track CTnT concentrations over time, would be beneficial in improving the clinical diagnosis of acute myocardial infarction.

Recently, a fully printed immunosensing biochip with hydrodynamic enrichment and photocrystal-enhanced fluorescence has been applied for biomarker detection.^31,32 Based on previous studies, the present study used a fluorescent labeling method based on quantum crystal PC290 in combination with a double-antibody sandwich immunoassay, which was able to enrich the samples and amplify the fluorescence signals by about 20-fold. The biomarker biochip detection process is shown in Fig. 1a–e. The biochip can detect CTnT in less than 10 minutes during an episode of chest pain using less than 10 μL of body fluid samples. The CTnT concentrations in serum, saliva, and urine from 100 subjects were analyzed using the PCB with liquid multiplexing. The results were consistent, and the detection limit was much lower than that of traditional ELISA kits (30 pg mL⁻¹). CTnT was detected separately: the serum (blood) analytical range is 0.24–3858.00 pg mL⁻¹ with a reliability (R²) of 0.939, the saliva analytical range is 0.39–387.00 pg mL⁻¹ with a reliability (R²) of 0.930, and the urine analytical range is 0.80–205.00 pg mL⁻¹ with a reliability (R²) of 0.906. In terms of clinical applications, long-term monitoring of saliva in subjects with a history of acute myocardial infarction using the PCB method allows for rapid therapeutic interventions in less than 10 minutes. The ability of the PCB to rapidly, sensitively, and noninvasively detect CTnT in real time allows for the early and long-term monitoring of the cardiac status of patients with myocardial injuries, inhibits the worsening of chest pain due to misdiagnosis, and ultimately improves the quality of life of patients.

Fig. 1 Photonic crystal (PC)-based biochip for detecting CTnT in blood, saliva, and urine of patients with chest pain. This figure was created using https://BioRender.com. (a) The patient has chest pain symptoms. (b) Patient's blood, saliva, and urine were extracted. (c) Detection antibody cy5-cAb (40 μg mL⁻¹) was mixed with the samples, respectively, and added to the PCB immobilized with the capture antibody cAb (15.20 μg mL⁻¹). (d) After 10 min, the fluorescence intensity of different CTnT concentrations was detected by the PCB. (e) Concentrations and fluorescence differences of CTnT in body fluids from patients with acute myocardial infarction and acute myocardial injury were measured every 10 minutes using the PCB, where the fluorescence images are those of samples diluted 2–100 times with PBS. The scale bar size is 500 μm.

Results and discussion

Mechanism of the PCB for the detection of the AMI biomarker CTnT

We designed and fabricated a fully printed PCB with pro-hydrophobic enrichment and PC-enhanced fluorescence properties combined. The fully printed PCB consists of PET plastic substrates and a PCB to detect the AMI biomarker CTnT. The selectivity for CTnT in the PCB array is achieved through two mechanisms: first, leveraging the wettability disparity, a hydrophilic–hydrophobic pattern comprising hydrophilic PC arrays and hydrophobic substrates drives trace samples to concentrate in the targeted areas. Second, an antibody-driven process occurs, where the antibody cAb immobilized on the photonic crystal array captures the cy5-dAb bound to CTnT via a double-antibody sandwich immunoassay. This portion of CTnT, released into bodily fluids, can be detected and used as a molecular marker in clinical and experimental studies. The serum, saliva, and urine samples were collected and mixed with CTnT cy5-dAb in brown tubes, which were specifically bound to the CTnT protein to form antigen–antibody complexes. The fabrication process of the PCB can be referred from Lian et al.³¹ The surface of PC microspheres had abundant carboxyl groups, which can be easily modified by biological materials to achieve specific recognition of antigen-antibodies. The activation solution (mixtures of EDC and NHS) and CTnT cAb were then sequentially added to the biochips. The function of the activation solution was to activate the carboxyl groups on the surface of PC nanoparticles to form amide bonds with the amino groups of the antibody, thus immobilizing the antibodies on the PCB. To minimize non-specific adsorption, a blocking solution was used to block the carboxyl sites on the surface of the photonic crystal that were not bound to the antibodies. Fig. 2a illustrates the procedure using the PCB. For patients with chest pain, samples were collected, mixed with cy5-dAb (40 μg mL⁻¹), placed in centrifuge tubes, and turned up and down 10 times to mix thoroughly. Then, 10 μL of the mixture was pipetted onto the prepared PCB. The CTnT cAb immobilized on the surface of the biochip recognized the antigen–antibody complex and formed a double-antibody sandwich structure. After a short reaction at room temperature, the samples were rinsed with deionized water and dried, and their fluorescence intensities were measured. cy5-dAb bound specifically to the PCB surface, and the PC amplified the fluorescence signal to quantify CTnT. The higher the amount of CTnT captured by the cAb, the higher the fluorescence intensity of cy5-dAb.

Fig. 2 Determination of parameters during biochip fabrication and validity analysis of biochip assays. (a) Schematic of the biochip sample detection process. This figure was created using https://BioRender.com. (b) Fluorescence intensity analysis corresponding to different concentrations of cAb (5.20–55.20 μg mL⁻¹) on the chips after photocrystal activation. (c) Fluorescence intensity analysis corresponding to different concentrations of cy5-dAb antibody (20.00–60.00 μg mL⁻¹) on the chips. (d) Fluorescence intensity of CTnT at a concentration of 100 pg mL⁻¹ over 5–40 minutes was measured using biochips. (e) Ratio of fluorescence intensity change to time for 100 pg mL⁻¹ CTnT was detected using biochips. (f) Fluorescence images of different concentrations of CTnT detected on the biochips. Scale bar: 500 μm. (g) Standard curve of fluorescence intensity versus standard antigen CTnT concentration on the chips. Data are expressed as mean ± standard deviation, n = 10 replicates. (h) Standard curve of optical density (OD) value versus CTnT concentration in samples in ELISA. Data are expressed as mean ± standard deviation, n = 10 replicates. (i) Standard curve of fluorescence intensity on biochips versus CTnT concentration in human serum. Data are expressed as mean ± standard deviation, n = 10 replicates. (j) Standard curve of fluorescence intensity on biochips versus CTnT concentration in human saliva. Data are expressed as mean ± standard deviation and n = 10 replicates.

Characterization of the PCB

Materials with photonic band gap (PBG) property having an artificial periodic dielectric structure can produce better fluorescence enhancement effects by matching with the emission wavelength of cy5.^31,32 The PCB in this experiment exhibited the PBG property, and we finally found that the PCB with a microsphere size of 290 nm (named PC290) has the best fluorescence enhancement effect. We measured the reflectance of PC290 at different incidence angles and emission wavelengths and found the reflectance of PC290 peaks at 670–690 nm emission wavelength (Fig. S1, ESI†). The incidence angle of 0° had a little effect on the fluorescence enhancement (Fig. S1, ESI†), and it corresponds to the fluorescence wavelength of the cy5 dye, enabling PC290 to enhance the fluorescence intensity of the dye. The PC layer thickness was positively correlated with the detection sensitivity.³¹ We performed thickness measurements of arbitrary PC layers using a stepper meter, which showed that the average layer thickness of PCB was >10 μm with better fluorescence enhancement (Fig. S2, ESI†). Scanning electron microscopy (SEM) was performed to examine the morphology of the PCB after burning at 110 °C for 10 min. Fig. S3 (ESI†) illustrates the SEM images of the periodic nanostructures of PC after burning. It is clear that the PC can be firmly adhered to the PET substrate while maintaining its original periodic nanostructure. The excitation wavelength generated by the higher electric field strength provides a stronger excitation power, which can enhance the fluorescence effect of the biochip. Referring to previous experiments, PCB with a 0–20 nm antibody layer chose the strongest electric field strength at 640 nm excitation wavelength for the best fluorescence enhancement.³¹ To determine the fluorescence enhancement of PC290 to cy5 dye, cy5 dye-labeled CTnT was added dropwise to PC290 and PET substrates, respectively. The cy5-cAb was excited by confocal microscopy at an injection wavelength of 640 nm. The analysis results of images with NIS-Elements are as follows (Fig. S4 and S5, ESI†). The fluorescence enhancement of cy5-dAb by PC290 was as high as about 20-fold, which greatly improved the sensitivity of CTnT detection. The difference in wet aptitude^31,32 enabled the sample to be enriched around the PCB, allowing for highly sensitive detection of substances measured using 10 μL of sample.

Optimization of cAb and cy5-dAb concentrations

The amount of cAb located on the surface of the PCB directly determines its detection ability, and the cy5-dAb paired with cAb also affects sample detection. First, to determine the optimal concentration of the cAb for encapsulation, 5.20–55.20 μg mL⁻¹ cAb solution was added dropwise to the activated PC microarrays. Meanwhile, 100 pg mL⁻¹ of CTnT protein standard was used to simulate the sample to be tested, and the fluorescence intensity was measured by observing the fluorescence images using a confocal microscope. The results show that the fluorescence intensity increased with the concentration of the cAb, reaching its maximum at 15.20 μg mL⁻¹ before beginning to diminish (Fig. 2b). When different concentrations of cy5-dAb were matched with 15.20 μg mL⁻¹ cAb, the 20.00–60.00 μg mL⁻¹ concentrations of cy5-dAb exhibited different linear relationships. Except that the fluorescence effect detected by cy5-dAb at a concentration of 60 μg mL⁻¹ fluctuated, the fluorescence effects of the antibodies at other concentrations increased linearly with the increase in the concentration of the CTnT standard. Among them, 40 μg mL⁻¹ showed a better matching effect (Fig. 2c). By maintaining the concentration of the cAb constant under various dAb concentration ratios, a low concentration of dAb may not adequately label all the captured CTnT owing to an insufficient dose, resulting in a fluorescence intensity similar to the nonspecific adsorption of the negative background. A dAb with excessively high concentration may correspond to a higher concentration of the cAb, leading to nonspecific adsorption that obscures the variation in fluorescence signals resulting from the increase in the CTnT concentration. The results indicate that the best matched concentration combination for the CTnT antibody pair was 15.2 μg mL⁻¹ cAb and 40 μg mL⁻¹ cy5-dAb.

Verifying the detection time of PCB

We used a double-antibody sandwich method, using a pair of specific antibodies, namely, cAb and cy5-dAb that bind to CTnT. The pair of antibodies recognize and dual-target different surface sites on CTnT, forming antigen–antibody complexes. This method significantly enhances the specificity of the PCB and prevents the interference of other proteins. We optimized the fluorescence detection time of CTnT at 100 pg mL⁻¹. As shown in Fig. 2d, within the detection time range of 40 minutes, before 25 minutes, the fluorescence intensity increased with the increase in time, and after 25 minutes, the fluorescence intensity decreased. The ratio of fluorescence change versus time is shown in Fig. 2e. The detection performance is best after 10 minutes.

Standard curve of the PCB for CTnT

The biochip was developed on the basis of the PC as a solid-phase carrier, combined with a fluorescent labeling method and the double-antibody sandwich method. As the above simulation results, the electric field strength at a height of 0–20 nm on the surface of the PC arrays became most powerful at 640 nm. Its component parameters were explored during the construction of the chip. The standard CTnT (0.25–250 pg mL⁻¹) to be measured was mixed with cy5-dAb and added dropwise to the surface of the chips, where they bound to cAb attached to the surface of the PC. By optimizing the concentration of two antibodies (Fig. 2b and c), we established an excellent detection standard curve. Fig. 2f shows the fluorescence images of the biochip detecting different concentrations of CTnT. The best detection interval was obtained from 0.25 pg mL⁻¹ to 250.00 pg mL⁻¹, and the R² value of the standard curve was 0.9420 (Fig. 2g). The fluorescence intensity values were regressed to the standard curve to obtain their corresponding concentrations of standard CTnT. We can estimate CTnT concentrations in serum, saliva, and urine from the fluorescence intensity values.

The unique optical properties of PCB make ultra-sensitive biological analysis in solution possible. To assess the sensitivity and limit of detection (LTD) of the PCB in serum, a series of human serum samples at concentrations ranging from 0.24 to 3858.00 pg mL⁻¹ were mixed with cy5-dAb (40 μg mL⁻¹), and then 10 μL of the mixture was added to the PC microarray as a detection sample. After 10 minutes of reaction, the PC microarray was rinsed with deionized water to prevent non-specific absorption. In order to ensure the consistency of experimental results, we used ELISA kits with the same paired antibodies for testing. A standard curve was constructed using the ELISA kit for comparison (Fig. 2h). The detection range of CTnT by ELISA was 0.25–80 pg mL⁻¹, and the R² value was 0.9974. The detection range of CTnT from serum using the PC microarray is 0.24–3858.00 pg mL⁻¹, and the R² value is 0.9389 (Fig. 2i). When AMI occurs, the optimal critical point for CTnT is 100 ng mL⁻¹.³¹ The obtained body fluid samples were serially diluted 10-fold until the CTnT concentrations in 2 to 3 samples fell within the range of 0.24 to 3858.00 pg mL⁻¹, which can cover this cutoff value. Using the same method, we assessed the detection range of CTnT in saliva for PC microarray to be 0.39–387.00 pg mL⁻¹, with an R² value of 0.9297 (Fig. 2j). Similarly, the detection range of CTnT in urine was assessed for the PC microarray to be 0.80–205.00 pg mL⁻¹, with an R² value of 0.9056 (Fig. S6, ESI†). The ELISA method demonstrates better reproducibility and higher system linearity, while the PCB method shows relative inadequacies in detecting body fluid samples. We attribute this to the following: the antigen sources in the ELISA method are consistent, whereas for the PCB method, which uses body fluid samples as antigen sources, there is issue of sample heterogeneity. CTnT can be detected in just 10 minutes using PC microarrays, compared with 2–3 hours for traditional ELISA. Compared with ELISA, our products have shorter detection times, lower costs, and a wider detection range.

Consistency testing of PCB and ELISA using serum, saliva, and urine samples

We performed PCB and ELISA on human serum samples and compared the results. The fluorescence signal intensities of 20 serum samples were obtained using PCB. The serum standard curve of CTnT was used to convert the fluorescence intensity into the corresponding detection concentration. The correlation test found that the value of r was 0.9980, which was a significant positive correlation. The R² value of the two methods was 0.9960, p < 0.0001, indicating a strong correlation between the results of the two methods (Fig. 3a and d). At the same time, the fluorescent signal intensities of 20 saliva samples were obtained using a PC microarray. The saliva standard curve of CTnT was used to convert the fluorescence intensity into the corresponding detection concentration. The correlation test found that the value of r was 0.9984, which was a significant positive correlation, with an R² value of 0.9969, p < 0.0001 (Fig. 3b and e). Finally, urine samples were subjected to PC microarray and ELISA, and the results were compared. The fluorescence signal intensities of 15 urine samples were obtained using a PC microarray. The urine standard curve of CTnT was used to convert the fluorescence intensity into the corresponding detection concentration. The correlation test found that the value of r was 0.9991, which was a significant positive correlation, with an R² value of 0.9982, p < 0.0001 (Fig. 3c and f). The results showed that the quantitative analysis of serum,³³ saliva, and urine samples by the PC microarray was comparable to the conventional ELISA.

Fig. 3 Collected samples tested by ELISA and biochip: (a) comparison of CTnT concentrations in serum quantified by ELISA and biochip. Data are expressed as mean ± standard deviation and n = 10 replicates. (b) Linear relationship between detected CTnT concentrations in saliva using ELISA and biochip. Data are expressed as mean ± standard deviation and n = 10 repeated tests. (c) Comparison of CTnT concentrations in urine quantified by ELISA and biochip. Data are expressed as mean ± standard deviation and n = 10 replicates. (d) Linear relationship between detected CTnT concentrations in serum using ELISA and biochip. Data are expressed as mean ± standard deviation and n = 10 repeated tests. (e) Comparison of CTnT concentrations in saliva quantified by ELISA and biochip. Data are expressed as mean ± standard deviation and n = 10 replicates. (f) Linear relationship between detected CTnT concentrations in urine using ELISA and biochip. Data are expressed as mean ± standard deviation and n = 10 repeated tests.

Clinical applications of PCB

In practice, continuous detection of CTnT during the general cycle of myocardial injury is an important indicator for determining the extent of myocardial injury and monitoring myocardial infarction.³⁴ Therefore, we followed the changes in CTnT in patients over a one-week period, as shown in Fig. 4a. CTnT peaked 12 hours after the onset of myocardial injury and reached three times its pre-procedure level. The CTnT levels decreased and remained normal from day 1 to day 7, consistent with the pattern of CTnT release during acute myocardial infarction.³⁴ Moreover, we compared the changes in the concentration of CTnT in the body fluids: serum, saliva, and urine of patients with myocardial injury. It was found that as the CTnT concentration in serum increased, the CTnT concentration in saliva and urine also increased, with the mid-CTnT concentration in serum being 1.82–2.21 times higher than that in saliva, and the mid-CTnT concentration in serum being 4.91–5.68 times higher than that in urine (Fig. 4a). In clinical applications, we used the PCB to detect CTnT in saliva and tracked the development of myocardial injury. Timely intervention treatment was carried out within 30 minutes of the occurrence of myocardial injury, and the changes in CTnT were continuously monitored within 3 days, which can effectively prevent the occurrence of AMI (Fig. 4b). Thus, this PCB demonstrated superiority over traditional ELISA tests for the quantitative analysis of serum, saliva, and urine CTnT, and has great potential for on-site detection and home self-testing diagnosis.

Fig. 4 Relationship between CTnT concentrations in body fluids and acute myocardial infarction: (a) CTnT concentrations in patients were quantified over a one-week period using ELISA and the biochip. Data are expressed as mean standard deviation and n = 10 repeated tests. (b) Schematic of AMI development. Concentrations of CTnT were monitored in saliva during the course of AMI, with therapeutic intervention initiated 0.5 hours after onset.

Conclusions

Briefly, the PCB is based on immunoassay principles and can be used to detect the markers of cardiomyocyte damage with a sensitivity as low as 0.25 pg mL⁻¹. In addition, we have evaluated the practical feasibility of the biochip in real human serum, saliva, and urine samples, and the results are consistent (Fig. S7, ESI†). We are also developing a portable device for reading the fluorescence intensity values, which in the future holds promise for the real-time accurate quantification of chest pain biomarkers at home (Fig. S8, ESI†). The PCB provides a rapid, sensitive, real-time, noninvasive, multibody fluid assay for the detection of biomarkers of myocardial injury, the detection of CTnT, and early and long-term cardiac monitoring of patients with myocardial injury, and provides a useful tool for the identification of patients at high risk of acute myocardial infarction, providing a timely and effective method.

Materials and methods

Materials

All reagents are used directly without further purification. Polystyrene microspheres (290 nm) were purchased from Huge Biotechnology. Surfactant (BYK3455) was supplied by Shanghai Macklin Biochemical Co., Ltd. CTnT cAb and dAb were purchased from Xiamen Tongrenxin Biotechnology. cy5-dAb: the dAb was labelled with the fluorescent dye cy5 (Beijing Biotyscience Technology, China). The ELISA kits and CTnT antigen were both purchased from Jiangsu Meimian Industrial Co., Ltd. Trehalose was obtained from Aladdin Holdings Group. Deionized water was purified using a Milli-Q Plus 185 water purification system (Millipore) with a resistivity of 18.2 MΩ cm⁻¹. N-Hydroxysuccinimide (NHS), N-(3-dimethylaminopropyl)-N′-ethyl carbodiimide hydrochloride (EDC), and casein sodium salt (CSS) were obtained from Aladdin. Bovine serum albumin (BSA) and phosphate-buffered saline (PBS) were bought from SINOPEC Chemical Reagent Company. The protease inhibitor (100 μg mL⁻¹) was acquired from Sigma-Aldrich.

Measurement of CTnT concentrations in serum, saliva, and urine samples by ELISA

First, the collected serum, saliva, and urine were centrifuged using a 4000 centrifuge (BY-R18 by Beijing Baiyang Medical Instrument Co., Ltd) at −4 °C for 30 minutes. The centrifuged supernatant was then transferred to a sterile microcentrifuge (NEST) and stored in an ultra-low temperature refrigerator (Beijing Nodong Company). The CTnT levels were measured using two different methods over one month. Specifically, the human CTnT ELISA kit (MEIMIAN) was used to measure the CTnT levels according to the manufacturer's instruction, and each sample was tested twice to ensure the accuracy and reliability of the results.

Fabrication of the PCB

For the specific fabrication process of the PCB, please refer to the relevant content by Lian.²⁹ First, 20 μL mixed EDC (40 mg mL⁻¹) and NHS (40 mg mL⁻¹) solution was added to each detection unit (d = 600 μm, 3 × 3, this 3 × 3 dot matrix) and left for 30 min to fully activate the carboxyl terminus on the surface of the PC, which were then rinsed three times with deionized water and dried. Second, 10 μL of CTnT cAb (15.20 μg mL⁻¹) was added to each testing microarray and reacted for 2 h, rinsed three times with deionized water and dried. Third, the unreacted carboxyl groups on the PC surface were blocked with 1% BSA buffer + 0.005% CSS for 30 min, followed by rinsing with deionized water for 1 min to remove the excess blocking solution. All the above-mentioned steps were performed at room temperature (25 °C). The PCBs were stored at 4 °C after preparation.

Measurement of the fluorescence gain effect of the PCB

One droplet of CTnT cy5-dAb solution (10 μL, 40 μg mL⁻¹) was dispensed onto PC spots and the pure PET substrate, respectively. The PC arrays were excited with an excitation wavelength of 640 nm using a confocal microscope. The fluorescence intensities were calculated, and the images were acquired using NIS-Elements, respectively.

Optimization of the cAb antibody concentration

First, 20 μL mixed EDC (40 mg mL⁻¹) and NHS (40 mg mL⁻¹) solution was added to each detection unit and left for 30 min to fully activate the carboxyl terminus on the surface of the PC, which were then rinsed three times with deionized water and dried. CTnT cAb (10 μL, 5.20–55.20 μg mL⁻¹) was added dropwise to the arrays, followed by incubation at room temperature for 2 h. The arrays were then rinsed three times with deionized water and dried. Then, the unreacted carboxyl groups on the PC surface were blocked with 1% BSA buffer + 0.005% CSS for 30 min, followed by rinsing with deionized water for 1 min to remove the excess blocking solution. The fluorescence intensity was measured on the PCB.

Optimization of the cy5-dAb antibody concentration

CTnT protein solutions (0.04–3858.00 pg mL⁻¹) were mixed with cy5-dAb (20–60 μg mL⁻¹). Then, 10 μL of each mixed solution was applied to a microarray, which was subsequently rinsed three times with deionized water and dried. The fluorescence intensity was measured using PCFLISA chips.

Establishment of the standard curves for CTnT detection by the PCB

The CTnT antigen standards were diluted to different concentrations (0.25–250.00 pg mL⁻¹) and mixed with cy5-dAb (40 μg mL⁻¹). Then, 10 μL of the mixture was added dropwise to one detection unit and the reaction was carried out at room temperature in a dark environment for 10 min. The reaction was then terminated by rinsing with deionized water and dried. The fluorescence intensity was measured on the PCB.

Detection of CTnT standard curves of serum, saliva, and urine

A high-concentration CTnT protein solution (38 580.00 pg mL⁻¹) was obtained from serum. Then, a 10 μL drop of CTnT protein solution (0.24–3858.00 pg mL⁻¹) was placed on the PCB unit. After reacting for 10 min, the PCB was washed with deionized water and dried. The fluorescence intensity was measured on the PCB.

A high-concentration CTnT protein solution (387.00 pg mL⁻¹) was obtained from saliva. Then, a 10 μL drop of CTnT protein solution (0.39–387.00 pg mL⁻¹) was placed on the PCB unit. After reacting for 10 min, the PCB was washed with deionized water and dried. The fluorescence intensity was measured on the PCB.

A high-concentration CTnT protein solution (205.00 pg mL⁻¹) was obtained from urine. Then, a 10 μL drop of CTnT protein solution (0.80–205.00 pg mL⁻¹) was placed on the PCB unit. After reacting for 10 min, the PCB was washed with deionized water and dried. The fluorescence intensity was measured on the PCB.

Measurement of CTnT concentrations in serum, saliva, and urine samples using the PCB

Centrifuged serum, saliva, and urine to be tested were diluted to 1/1000, 1/100, and 1/10 of the original concentration, respectively, and mixed with the cy5-dAb (40 μg mL⁻¹). A phosphate-buffered saline solution (pH = 7.2–7.4) was used for the dilution of the samples. Ten microliters of the mixture were added dropwise to each detection unit, and parallel experiments were conducted 10 times to verify the detection accuracy. The reaction was terminated by rinsing with deionized water three times after 10 minutes, which served to prevent nonspecific adsorption. The fluorescence intensity values were read using a confocal microscope within 10 min of the end of the reaction. The fluorescence intensity of one array was calculated by averaging the fluorescence intensity of nine points in 3 × 3 PC arrays, and the final value for each biomarker concentration was determined by averaging the fluorescence intensity of 10 different arrays.

Characterization

Contact angles (CAs) were measured using an OCA20 instrument (Dataphysics) at 25 °C. The reflectance spectra of the PC were acquired using an angular resolution spectrometer (HR4000CG-UV-NIR Ocean Optics, Inc.). Scanning electron microscopic (SEM) images of the PC were acquired using a scanning electron microscope (JEOL-SU8020). Fluorescence images were photographed using a Nikon C2 confocal microscope.

Statistical analysis

GraphPad Prism 8.0 was used to plot the figures and statistical analysis. The Shapiro–Wilk method was used to analyze normality for groups. Data conforming to normal distribution were expressed as mean ± SD and further analyzed using a t-test or analysis of variance. Quantitative data, not conforming to a normal distribution, were expressed using the median and interquartile spacing and further analyzed by the Kruskal–Wallis test, and p < 0.05 was considered statistically significant. Each point was averaged from at least ten independent measurements.

Author contributions

X. W., N. C., F. L., M. S., and Z. L. conceived and designed the experiments. X. W. and Z. T. completed the writing of the manuscript. Z. L., X. W., and N. C. printed and characterized the biochip. X. W., N. C., and F. L. characterized the various detection properties. X. W. and Z. L. calculated the finite element simulation. X. W., Y. Y., J. Z., T. R., H. S., M. S., and N. C. conducted the diagnostics of human body fluids.

Conflicts of interest

The authors declare that they have no conflict of interest.

Compliance with ethics guidelines

This study and all participants followed ethical guidelines.

Ethical approval

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2008 (5). Informed consent was obtained from all patients to be included in the study.

Data availability

The data supporting this article have been included as part of the ESI.†

Acknowledgements

Thanks to the National Key Research and Development Program of China (No. 2023YFE0111500), the National Natural Science Foundation of China (No. 62476286), the Beijing Science and Technology Planning Project (No. Z231100005923039), the Rehabilitation Medicine Association of China (No. KFKT-2024-003), and the Ministry of Science and Technology (No. SQ2022YFB4700129) for providing support for this research.

References

1. W. H. Organization, World health statistics 2018: monitoring health for the SDGs, sustainable development goals, 2021, https://apps.who.int/iris/handle/10665/272596.
2. M. K. Song JL, S. Hu, Y. Gao, X. Li, H. M. Krumholz, X. Zheng and P. C. G. China, Incidence, predictors, and prognostic impact of recurrent acute myocardial infarction in China, Heart, 2020, 107(4), 313–318.
3. Z. B. Tao NY and Y. G. Ren, et al., Research on detection methods of acute myocardial infarction progress, J. Southeast Univ., 2018, 37(6), 1085–1088.
4. D. K. Li J, X. K. Bai, F. A. Masoudi, J. A. Spertus, X. Li, X. Zheng, H. Zhang, X. Yan, R. P. Dreyer and H. M. Krumholz, Thirty-day hospital readmission after acute myocardial infarction in China, Circ. Cardiovasc. Qual. Outcomes, 2019, 12(5), e005628.
5. Y. Y. Xu HY and C. S. Wang, et al., Association of hospital level differences in care with outcomes among patients with acute ST-segment elevation myocardial infarction in China, JAMA Netw. Open, 2020, 3(10), e2021677.
6. Writing committee of the report on cardiovascular health and diseases in China, Report on Cardiovascular Health and Diseases in China 2021: An Updated Summary, Biomed. Environ. Sci., 2022, 35(7), 573–603,  DOI:10.3967/bes2022.079.
7. C. A. DeFilippis AP, N. L. Mills, J. A. de Lemos, A. Arbab-Zadeh, L. K. Newby and D. A. Morrow, Assessment and treatment of patients with type 2 myocardial infarction and acute nonischemic myocardial injury, Circulation, 2019, 140(20), 1661–1678,  DOI:10.1161/circulationaha.119.040631.
8. A. J. Thygesen K and A. S. Jaffe, et al., Executive group on behalf of the joint European society of cardiology (ESC)/American college of cardiology (ACC)/American heart association (AHA)/world heart federation (WHF) task force for the universal definition of myocardial infarction. Fourth universal definition of myocardial infarction, Circulation, 2018, 138, e618–e651,  DOI:10.1161/CIR.0000000000000617.
9. S. L. Sarkisian L and T. S. Poulsen, et al., Prognostic impact of myocardial injury related to various cardiac and noncardiac conditions, Am. J. Med., 2016, 129, 506–514.e501.
10. S. A. Chapman AR and K. K. Lee, et al., Long-term outcomes in patients with type 2 myocardial infarction and myocardial injury, Circulation, 2018, 137, 1236–1245.
11. T. D. Shen W, H. Cui, D. Yang and Z. Bian, Nanoparticle- based electrochemiluminescence immunosensor with enhanced sensitivity for cardiac troponin I using N-(aminobutyl)-N-(ethylisoluminol)-functionalized gold nanoparticles as labels, Biosens. Bioelectron., 2011, 27(1), 18–24,  DOI:10.1016/j.bios.2011.05.022.
12. M. Mueller, S. Celik and M. Biener, et al., Diagnostic and prognostic performance of a novel high-sensitivity cardiac troponin T assay compared to a contemporary sensitive cardiac troponin I assay in patients with acute coronary syndrome, Clin. Res. Cardiol., 2012, 101, 837–845.
13. T. H. Collet JP and E. Barbato, ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST- segment elevation, Eur. Heart J., 2020, 41, 3495–3497.
14. D. S. Luepker RV and D. R. Jacobs, et al., The effect of changing diagnostic algorithms on acute myocardial infarction rates, Ann. Epidemiol., 2011, 21, 824–829,  DOI:10.1016/j.annepidem.2011.08.005.
15. C. A. Ma H and R. O'Kennedy, The role of antibody- based troponin detection in cardiovascular disease: a critical assessment, J. Immunol. Methods, 2021, 497, 113108,  DOI:10.1016/j.jim.2021.113108.
16. H. T. Hsueh and C. T. Lin, An incremental double-layer capacitance of a planar nano gap and its application in cardiac-troponin T detection, Biosens. Bioelectron., 2016, 79, 636–643,  DOI:10.1016/j.bios.2015.12.105.
17. L. P. Jiang MH and M. C. Yan, et al., Self-accelerated electrochemiluminescence emitters of Ag@SnO₂ nanoflowers for sensitive detection of cardiac troponin T, Electrochim. Acta, 2018, 271.
18. Y. Z. Guo L, S. Zhi, Z. Feng, C. Lei and Y. Zhou, Sensitive detection of cardiac troponin T based on superparamagnetic bead-labels using a flexible micro fluxgate sensor, RSC Adv., 2017,(7), 52327–52336,  10.1039/C7RA10355G.
19. M. Y. Zhang S and Q. Huo, Different interaction modes of biomolecules with citrate-capped gold nanoparticles, ACS Appl. Mater. Interfaces, 2014, 6(23), 21184,  DOI:10.1021/am506112u.
20. A. A. Ulloa-Gomez AM, A. Lucas, S. B. Somvanshi and L. Stanciu, Smartphone-based colorimetric detection of cardiac troponin T via label-free aptasensing, Biosens. Bioelectron., 2023, 222, 114938,  DOI:10.1016/j.bios.2022.114938.
21. C. X. Liao C and Y. Fu, Salivary analysis: an emerging paradigm for non- invasive healthcare diagnosis and monitoring, Interdiscip. Med., 2023, 1, e12051,  DOI:10.1002/inmd.12051.
22. K. Kwon, J. Kim, Y. Deng, S. R. Krishnan, J. Choi, H. Jang, K. Lee, C. J. Su, I. Yoo, Y. Wu, L. Lipschultz, J. H. Kim, T. S. Chung, D. Wu, Y. Park, T. I. Kim, R. Ghaffari, S. Lee, Y. Huang and J. A. Rogers, An on-skin platform for wireless monitoring of flow rate, cumulative loss and temperature of sweat in real time, Nat. Electron., 2021, 4(4), 302–312,  DOI:10.1038/s41928-021-00556-2.
23. H. Y. Nyein, M. Bariya, B. Tran, C. H. Ahn, B. J. Brown, W. Ji, N. Davis and A. Javey, A wearable patch for continuous analysis of thermoregulatory sweat at rest, Nat. Commun., 2021, 12(1), 1823,  DOI:10.1038/s41467-021-22109-z.
24. Z. K. Jordaens S, W. Tjalma, C. Deben, K. Beyers, V. Vankerckhoven, P. Pauwels and A. Vorsters, Urine biomarkers in cancer detection: A systematic review of preanalytical parameters and applied methods, Int. J. Cancer, 2023, 152(10), 2186–2205,  DOI:10.1002/ijc.34434.
25. K. E. Lee S, C. E. Moon, C. Park, J. W. Lim, M. Baek, M. K. Shin, J. Ki, H. Cho, Y. W. Ji and S. Haam, Amplified fluorogenic immunoassay for early diagnosis and monitoring of Alzheimer's disease from tear fluid, Nat. Commun., 2023, 14(1) DOI:10.1038/s41467-023-43995-5.
26. S. R. Villiger M and T. Vetsch, Evaluation and review of body fluids saliva, sweat and tear compared to biochemical hydration assessment markers within blood and urine, Eur. J. Clin. Nutr., 2018, 72, 69–76.
27. I. Mirzaii-Dizgah and E. Riahi, Salivary high-sensitivity cardiac troponin T levels in patients with acute myocardial infarction, Oral Dis., 2013, 19(2), 180–184,  DOI:10.1111/j.1601-0825.2012.01968.x.
28. R. Westreich, Y. Neumann, O. Deutsch, G. Krief and D. Zager, Development of saliva-based cTnI point-of-care test: a feasibility study, Eur. Heart J., 2020, 41(Supplement_2) DOI:10.1093/ehjci/ehaa946.1693.
29. A. Ovchinnikov, Saliva As a diagnostic tool for early detection of elevated cardiac troponin- I in athletes after strenuous exercise, Clin. Chim. Acta, 2024, 558 DOI:10.1016/j.cca.2024.118115 , Meeting Abstract.
30. A. Ovchinnikov and P. Suldin, Quantification of cardiac troponin- I in saliva of student- athletes at rest and following a multistage 20- m shuttle run test, Clin. Chim. Acta., 2024, 558 DOI:10.1016/j.cca.2024.118112 , Meeting Abstract.
31. W. T. Lian ZW, H. D. Wang, J. Chi, L. Cheng, D. Xie, X. Pan, Y. Hu, Z. Tan, S. Chen, X. Yang, Y. Yun, W. Wu, C. Li, M. Su and Y. Song, At-home COVID-19 rapid antigen test down to 0.03 pg mL⁻¹ of nucleocapsid protein, Small, 2023, 19(28), e2301162,  DOI:10.1002/smll.202301162.
32. S. M. Chi JM, B. J. Xue, L. Cheng, Z. Lian, Y. Wang, X. Yang, X. Wang, H. Xie, H. Wang, Y. Yun, J. Du and Y. Song, Fast and sensitive detection of protein markers using an all- printing photonic crystal microarray via fingertip blood, ACS Sens., 2023, 8(4), 1742–1749,  DOI:10.1021/acssensors.3c00029.
33. N. Wettersten and A. Maisel, Role of cardiac troponin levels in acute heart failure, Card. Fail. Rev., 2015, 1(2), 102–106,  DOI:10.15420/CFR.2015.1.2.102.
34. T. Y. Yin HC and P. C. Wang, et al., A review of cardiac troponin I and AMI diagnosis and its detection method, Adv. Mod. Biomed., 2010, 19, 4.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5mh00525f

‡ These authors contributed equally to this work.

---