Fluorescent Immunological Paper-based Assay for Exosome detection


 This paper reports the development of fluorescent-linked immunosorbent paper-based assay for exosome detection. The paper-based device was fabricated with sandwich lamination for easy handling and was coated with exosome-specific antibody as a biosensing platform to detect exosome sample from the cell culture media. This assay employed fluorescent detection which is followed by tagging fluorophore-conjugated detecting antibody on exosome samples. The fluorescent readout was evaluated and quantified from image processing software. This assay can detect high concentration of exosome samples (~ 1010 exosome/mL). However, this assay has encountered various challenges. First, the exosome concentration prepared from cell culture media from cancer-derived ovarian and mesothelial cell lines may be insufficient to reach detectable range. Second, chemical contamination from exosome isolation kits may affect assay sensitivity. Therefore, assay optimization and minimizing chemical contamination are required which could enhance assay specificity and sensitivity.

Ordinarily, exosomes are circulated in biological uids such as saliva, blood, and urine. Many studies recommended that these exosomes play a signi cant role in cell-to-cell interaction and some pathological processes such as cancers [6], pregnancy complications [7,8] and neurodegenerative diseases [9]. As such, during pregnancy, exosomes are one of signal pathways to communicate between the mother and the fetus [10]. During gestation, the human placenta releases exosome into the maternal circulation in the rst trimester. Oxygen tension and glucose concentration including placental mass regulated the release of placental exosomes [11,12]. There has been reported that the expected mothers, who have a higher concentration of placental-derived exosomes, face the risk of pregnancy complications such as gestational diabetes [8] and preeclampsia [7] than that of normal pregnancy. Therefore, detecting unusual level of placental-derived exosomes early as rst trimester of pregnancy can help expectant mothers at risk to be treated properly before the complication develops.
Thus, the accurate puri cation and detection of exosomes from bodily uids is still challenging [13]. Conventional methods such as gradient density centrifugation, ultra ltration or immunological separation have been reported [14,15]. However, these centrifugal isolations are time-consuming, and require expensive instrument and facilities, which may be inaccessible in poor-resource settings.
Moreover, the commercial exosome isolation kit is also available. These kits shorten the process by precipitating exosome with polyethylene glycol or related chemicals, but the product may be contaminated with chemicals from the kits which possibly impact the downstream analysis. [16] Alternatively, ow cytometry [17,18], nanoparticle tracking analysis (NTA) [19,20], electrochemical-based method [21,22], Enzyme-linked immunosorbent assay (ELISA) [23,24] are successfully demonstrated to quantify and detect exosome. ELISA is known as a conventional method to detect and analyse various biomolecules, based on immunological a nity between antibodies and protein on exosome membranes. However, the standard protocol is time consuming with laborious sample loading, washing, and incubation step.
Paper-based platform enables a number of micro uidic sample preparation tasks such as preconcentration and heating. [25,26] Paper-based ELISA is a promising platform for diagnostic application due to the low cost, friendly use and small sample needed. This platform was successfully demonstrated for detecting antibodies and exosomes [27][28][29][30]. In limited-resource environments, colorimetric assay is useful and easily observed by naked eyes, however minor colour change may be di cult to differentiate due to the broad range of colour spectrum. [31] Fluorescent-linked immunosorbent assay (FLISA) promotes narrow detection range due to speci c excitation and emission wavelength. Therefore, the present study developed FLISA paper-based assay for exosome detection. The exosome target was captured with exosome-speci c antibody and formed immunological complex on the paper. The uorescent readout relies on detecting emission photon excited by a speci c wavelength, which results from immunological a nity between exosomes and uorophore-conjugated antibody.

Materials and chemicals
Paper materials used as a platform device in this study were chromatography lter paper (CHR). Gloss laminate pouch with 80-µm thickness was used to make paper hydrophobic. For immunological assay, Rabbit monoclonal anti-human CD9 and CA125 monoclonal Anti-human (Anti-human CA125, MA1-24603, Invitrogen, USA) were used for coating and detecting antibody. AlexaFlour488-conjugation kit (ab236553, Abcam, UK) was used to conjugate AlexaFlour488 onto an antibody. Bovine Serum Albumin (BSA, A1595, Sigma Aldrich, USA) and Non-fat milk (NFM, Coles, Australia) were used to block the assay to minimize background signal due to non-speci c binding. For exosome preparation, we used the same chemicals as described in the previous study. [32] 2.2 Paper-based device fabrication Adapting from the previous study, we fabricated laminated CHR paper (L-CHR), Figure 1. Brie y, the device design was drawn by CorelDraw software and cut into 5-mm diameter piece using laser cutting machine including a laminated lm which were cut with the hole of 4-mm diameter. Before the aligned paper-based device was fed into a laminator at a speed of 10 mm/s of 120 o C, the paper was aligned into the centre of laminated lm hole. Finally, the paper-based device was cut into individual devices.

Experimental design and assay optimization
We employed sandwich FLISA assay on paper-based platform using immunoa nity interactions. In this study, we used Anti-human CD9 as a primary antibody to capture all exosomes as CD9 was expressed and generic tetraspanin family of mammalian exosome membrane [33]. Anti-human CA125 was commonly used as a biomarker to diagnose ovarian cancer [34,35]. Thus, we selected Anti-human CA125 as a secondary antibody to capture ovarian cancer exosome. OVCAR3 exosome as of ovarian cancer cell lines was considered to be positive case and Met5a exosome as of mesothelial cell lines was considered to be negative case. PBS was used for control experiment by obtaining background signal and con rming the validity of the assay.

Secondary antibody concentration and incubation time optimization
According to our previous colorimetric immunological assay study, we selected 2% BSA for blocking [29]. However, concentration of uorophore-conjugated anti-human CA125 and incubation time need to be optimized. In this study, 2% NFM was also evaluated. Brie y, 5 µL of 2% NFM blocking solution was added to incubate for 10 min and subsequently dry for 10 min. Then, we dropped anti-human CA125 tagged with AlexaFlour488 for 5 min onto the paper and washed subsequently. Subsequently, we put the paper device under uorescent microscope and excite with light at a wavelength of 488 nm. We anticipated that the optimal condition for blocking could result in the lowest intensity of bright green colour.

Analytical performance of paper-based FLISA for exosome detection
In this study, we employed paper-based FLISA for exosome detection. Assay preparation and exosome preparation was adapted as described in the previous study [32]. Brie y, we dropped 5 µL of 0.1 mg/mL Rabbit anti-human CD9 on L-CHR paper and incubate for 10 min, then dry at 37 o C for 10 min. Next, a volume of 5 µL of 2% NFM solution was added to block and prevent non-speci c absorption. In the presence of ovarian cancer exosome, AlexaFlour488-conjugated anti-human CA125 was left on exosome membranes as well as on the paper platform and was excited with light at the wavelength of 488 nm. The uorophore molecules will emit green light. The intensity of the green uorescence signal is proportional to the amount of excited photon. However, in the absence of ovarian cancer exosome, there was no AlexaFlour488-conjugated anti-human CA125 was left on the paper platform.
Furthermore, sample loading procedure was also investigated. We proposed two methods in the loading step. First, we performed the assay as our previous study which separates the step between loading sample and loading secondary antibody, Figure 2a. The exosome sample was incubated on L-CHR paper for 10 min, and the secondary antibody was loaded onto the paper for 20 min, respectively. Second, we incubated our exosomes with uorophore-conjugated antibody solution before loading the mixture onto the paper, Figure 2b. The exosome sample was incubated in a secondary antibody solution for 30 min at room temperature. The mixture was loaded onto the paper and incubated for additional 10 min. Subsequently, the paper device was washed with PBS. Then, we placed the paper-based device under the uorescent microscope and observed the emission of bright green signal when it was excited by the light at wavelength 488 nm.

Equipment setup and data acquisition
Due to the experimental result based on intensity of change from uorophore, the paper device was observed under uorescent microscope (Nikon Eclipse Ti-2, Nikon, Japan) with an excitation wavelength of 488 nm from light source (pE-4000, CoolLED, UK) and exposure time at 300 milliseconds. For image acquisition, the image was observed and recorded by an imaging software (NIS-Elements AR 5.00.00, Nikon, Japan). In the software, we can adjust LookUp Table (LUTs) which affects the colour output value. We set the range of LUTs value from 500 to 2,000 and 500 to 5,000 to show the best result of green signal observed under the microscope (see Supplementary Information section S.1). Image processing and quanti cation were performed with MATLAB.

Data processing and quanti cation
Qualitative data was observed as a bright green colour under excitation at the wavelength of 488 nm. All photos were processed and quanti ed with MATLAB. However, the uorescent image emitted uorescent signal at a wavelength of 525 nm which is green. Therefore, only the mean grey value (G) from the green channel was used to quantify the uorescent signal. To con rm that the colour originates from the immuno uorescent assay, the value reported in this study was subtracted by the background signal from negative samples as: Where G is the mean gray value of the sample of interest, and G neg is the mean grey value from negative sample. All experiments were individually performed with a sample number of at least n = 3. The error bar is determined by standard deviation.

Optimization of secondary antibody concentration and incubation time
Before performing the analytical assay for exosome detection by paper-based FLISA, we need to optimize concentration of uorophore-conjugated antibody and incubation time as well as blocking solution. These two parameters are the main factors, which affect the assay performance. High uorophoreconjugated antibody concentration and longer incubation time help promotes sensitivity, but could cause high background signal [36,37]. We designed the experiment by varying the concentration of uorophoreconjugated antibody from 1, 5 and 10 µg/mL and varying concentration time from 5, 10 and 20 minutes.
Subsequently, all paper-based devices were observed under the uorescent microscope and quanti ed by MATLAB. Figure 3 showed the uorescent signal from various concentration of uorophore-conjugated antibody under incubation at 5 min. The highest concentration up to 10 µg/mL can be used as it produced relatively low green uorescent intensity. However, it may leave some contaminant small particles resulting from un ltered PBS. In addition, Figure S.2 depicts the uorescent signal from different incubation time at a concentration of uorophore-conjugated antibody of 10 µg/mL. The results showed that longer incubation up to 20 min still produced relatively low green uorescent intensity. Therefore, based on this optimization, for analytical performance assay of exosome detection assay, we can select the uorophore-conjugated antibody concentration at 10 µg/mL and incubation time up to 20 min as they produce relatively low background signal which could promote the immunological complex to form properly and produce the detectable signal for exosome detection. In addition to optimal conditions, LUTs value set by image acquisition software was selected with the scale of 500-2,000 as providing the widest range of green uorescent signal.

Analytical performance assay
FLISA assay for exosome detection employs the presence of uorophore-conjugated detecting antibody excited by a speci c wavelength of the light source. To validate the exosome uorescent immunological assay, we performed the analytical assay with a sample containing exosomes collected from ovarian cancer cell lines. On the contrary, a negative control sample is mesothelial, which is non-ovarian-cancer cell line. We investigated sample loading steps. Exosome concentration used in this investigate was 10 9 exosome/mL. Figure 5a depicts the uorescent photos taken from the rst method which is separately loading samples and uorophore-conjugated secondary antibody. The G value indicates that there is no signi cant difference and the negative control provide a bit higher signal comparing to the positive.
Nevertheless, Figure 5b shows the uorescent images taken from the second method, where exosome sample was incubated with uorophore-conjugated secondary antibody before loading the mixture onto the paper. The G value indicates that positive value provided a higher signal which was 5.746 (positive) vs 0.978 (negative). Therefore, we prepared the exosome sample by incubating with AlexaFlour488conjugated anti-human CA125 for 30 min before loading onto the paper.
For analytical assay, we investigated concentration variation, which was 10 3 , 10 6 , 10 9 , 10 10 exosome/mL ( Figure 6). The higher concentration of exosome sample can generate higher signal determined by G value , while uorescent readout seems not to produce signi cant difference by naked eye. Nevertheless, the result seems to be inconsistent due to sample and assay batch-to-batch variation. We repeated the experiment in each condition to obtain the average of G value (Figure 7). However, the variation within the same exosome concentration may be too large to differentiate from other concentrations. Moreover, the exosome concentration may also be insu cient for FLISA method to produce green signal higher than a background signal.

Conclusion
We have developed and investigated a paper-based FLISA assay for exosome detection. Fluorescent readout facilitates high speci city of the readout due to speci c wavelength emission and can be quanti ed by image processing. With the proof-of-concept, we optimized the assay and faced many challenges in assay development for detecting exosomes. First, exosome concentration from cell culture may be insu cient for generating uorescent emission signal in the detectable range. Second, the high background signal from non-speci c binding and PEG contamination from exosome isolation kits may disturb the interaction between proteins [38]. In the future, this assay may require extensive optimization on assay operation to obtain better results. Schematic diagram of assay preparation and optimization: (a) separately loading exosome sample and secondary antibody solution; (b) incubating exosome sample with secondary antibody before loading the mixture onto the paper.