ICA was purchased from Welch Materials, Inc. (Shanghai, China; purity of 95%). Sodium periodate was obtained from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). Bovine serum albumin (BSA), ovalbumin (OVA) and Freund's complete and incomplete reagents were obtained from Sigma-Aldrich; Merck Millipore (Darmstadt, Germany). All other chemicals and reagents were of analytical grade and were purchased from Sinopharm Chemical Reagent Co., Ltd.

The composition of the four TCM detected were as follows: i)≈Zhuanggu guanjie wan (bolus) (China Resources Sanjiu Medical and Pharmaceutical, Shenzhen, China): Cibotii Rhizoma, Epimedii Folium, Dipsaci Radix, Psoraleae Fructus, Spatholobi Caulis, Olibanum, Myrrha, Aucklandiae Radix, Drynariae Rhizoma, Taxilli Herba, Angelicae Pubescentis Radix, Rehmanniae Radix Praeparata; ii) Pishen Shuangbu decoction: Taxilli Herba (30 g), Corn Stigma (30 g), Polygoni Multiflori Radix (24 g), Chuanxiong Rhizoma (9 g), Eucommiae Cortex (9 g), Magnetitum (3 g), Fossil Fragments (30 g) and Epimedii Folium (9 g); iii) Erxian decoction: Curculiginis Rhizoma (9 g), Epimedii Folium (9 g), Morindae officinalis Radix (9 g), Phellodendri Chinensis Cortex (4.5 g), Angelicae Sinensis Radix (9 g) and Anemarrhenae Rhizoma (4.5 g); iv) Tongbi Prescription: Aconiti Kusnezoffii Radix (15 g), Aconiti Radix (10 g), Asari Radix Et Rhizoma (6 g), Ephedrae Herba (10 g), Taxilli Herba (15 g), Zingiberis Rhizoma (10 g), Glycyrrhizae Radix Et Rhizoma (6 g), Epimedii Folium (15 g), Eucommiae Cortex (10 g) and Dipsaci Radix (15 g). All aforementioned Chinese herbal medicine was purchased from Tongrentang (Beijing, China).

The compositions of the buffers and solutions used in the present study were as follows: Phosphate-buffered saline (PBS; pH 7.4): NaCl (137 mmol/l), Na₂HPO₄•12 H₂O (10 mmol/l), KCl (2.68 mmol/l) and KH₂PO₄ (1.47 mmol/l); carbonate buffer solution (CBS; pH 9.6): Na₂CO₃ (15 mmol/l) and NaHCO₃ (35 mmol/l); washing buffer: PBS with 0.05% Tween-20 (PBST); blocking buffer: 10 mg/ml gelatine in PBS (GPBS); tetramethylbenzidine (TMB) substrate solution: Combination of part A (0.5 ml of stock comprising 24.3 ml of 0.1 mol/ml citric acid, 25.7 ml of 0.2 mol/ml Na₂HPO₄ and 50 ml deionized water), part B (10 ml of stock comprising 2 mg of TMB dissolved in 1 ml of methanol) and part C (32 µl of 0.75% H₂O₂); stopping solution: 2 mol/ml H₂SO₄ and 1 mM HCl (pH 4.0; 20 ml/g); hypoxanthine-aminopterin-thymidine (HAT); polyethylene glycol (PEG); and hypoxanthine-thymidine (HT).

The conjugates were synthesized using a periodate oxidation procedure according to a previously reported protocol with modifications (22,23). Briefly, the ICA was dissolved in CBS at 1 mg/ml, following which 1 ml of freshly prepared sodium periodate solution (8 mg) was added drop-wise to 5 ml of the ICA solution. The mixture was stirred at 25°C for 1 h, following which 5 mg of BSA dissolved in 1 ml of CBS was added, and the final pH was adjusted to 9.0 using 0.05 M carbonate buffer (pH 9.6). Following stirring at 25°C for 6 h, the mixture was dialyzed six times against PBS. The dialysate of the ICA-BSA conjugate was stored at 4°C for detection and immunization. The ICA-OVA conjugate was synthesized using the same method described above.

A total of 5 female BALB/c mice (6 weeks old) were purchased from Vital River Laboratories (Beijing, China). The mice were fed a standard rodent diet (Keaoxieli Animal Feed Co., Ltd., Beijing, China) ad libitum and housed in an environmentally controlled (23±2°C; 12 h light/dark cycle) animal facility. Mice were sacrificed by cervical dislocation. The present study was performed according to the Guidelines for the Care and Use of Laboratory Animals and was approved by the Joint Ethical Review Committee of the Beijing University of Chinese Medicine (Beijing, China; approval no. 2013 BZHYLL00106).

The immunizations were performed at 2-week intervals. The mice were subcutaneously injected with a 50 µg volume of the ICA-BSA conjugate in PBS, emulsified with an equal volume of Freund's complete adjuvant in the initial immunization. The second and third immunizations, which contained 50 µg of the ICA-BSA conjugate in Freund's incomplete adjuvant, were injected subcutaneously 2 and 4 weeks following the initial injection. Blood was obtained from the tail vein of the mice and centrifuged at 4°C and 2,227 × g for 10 min following the third immunization, and a titre of sera was examined by indirect ELISA, using ICA-OVA as the solid-phase antigen. After 2 weeks the fourth immunization involved injection with a solution of ICA-BSA (100 µg) in PBS without adjuvant.

At 3 days following the final immunization the spleen from immunized mice was removed using a cell strainer. Using the head of a clean syringe, the spleen was ground to dissociate the splenocytes. The cells were suspended following filtering and splenocytes were isolated and fused with the hypoxanthine-aminopterin-thymidine (HAT)-sensitive SP2/0 mouse myeloma cell line (ScienCell Research Laboratory; Carlsbad, CA, USA), according to the PEG method (24,25). Briefly, following centrifugation at room temperature at 180 × g for 10 min of the blended splenocytes and myeloma cells (at a ratio of 5:1), 1 ml of PEG was added drop-wise to the cell pellet and then incubated for 1 min at 37°C. The HAT medium (Sigma-Aldrich; Merck Millipore) was then added. The hybridoma was transferred to 96-well plates for cell culture. The cells producing MAbs reactive to ICA, as identified using indirect ELISA, were cloned according to the limiting dilution method (26). The established hybridoma was then cultured in the HT medium with 5% CO₂ at 37°C for 10 days.

A total of 20 mice (10-week-old, male) were purchased from Vital River Laboratories (Beijing, China). The mice were fed a standard rodent diet (Keaoxieli Animal Feed Co., Ltd., Beijing, China) ad libitum and housed in an environmentally controlled (23±2°C; 12 h light/dark cycle) animal facility, which had been injected with 300 µl Freund's incomplete adjuvant on the previous day. The hybridomas were transplanted into the abdominal cavity of 10-week-old male BALB/c mice, which had been injected with Freund's incomplete adjuvant on the previous day. The mice were sacrificed prior to the injection by cervical dislocation. After 5–7 days, fluid was drained from the resulting ascites. This fluid was then purified by centrifugation at 4°C and 5,702 × g for 10 min, followed by caprylic acid precipitation and protein quantification (27).

The reactivity of the anti-ICA MAbs with ICA-OVA was determined using indirect ELISA (iELISA). The ICA-OVA was dissolved in CBS, following which 100 µl was added to each well of a 96-well maxisorp immunoplate and incubated for 1 h. Each well was then treated with 200 µl of GPBS for 1 h to inhibit non-specific absorption. The plate was washed three times with PBST prior to the addition of 100 µl of anti-ICA MAbs (1:10,0000). Following incubation for 1 h, the plate was washed three times and incubated with 100 µl of peroxidase-labelled goat anti-mouse IgG solution (1:10,000; cat. no. C1308; Applygen Technologies Inc., Beijing, China) for 30 min. The plate was then washed three times, and 100 µl of TMB substrate solution was added to each well, followed by incubation for 15 min. The reaction was terminated by adding 50 µl of 2 M H₂SO₄, and the absorbance was measured at 450 nm using a BioTek ELx 800 microplate reader (BioTek China, Beijing Chin). All reactions were performed at 37°C.

The reactivity of the anti-ICA MAbs was determined using an icELISA. The protocol for this assay was identical to that used for the iELISA, with the exception that the primary antibodies used were 50 µl of ICA and 50 µl of anti-ICA MAb, and the subsequent incubation duration was 1 h.

The icELISA was established with ICA-OVA (1:2,000) as the solid-phase antigen and ascites induced by the transplant of the hybridoma (1:200,000). Various quantities of ICA were added to compete with the coated antigen, by which the standard curve of inhibition and measuring range were established.

Cross-reactivity (CR) is the most important factor in phytochemical investigations, as there are several structurally-related compounds. In the present study, the assay specificities were examined using icELISA with various associated compounds; the CR% of ICA and the associated compounds were determined according to Weiler's equation (28).

Intra-assay variation was assessed by evaluating the relative standard deviation (RSD%) of the ICA samples of varying concentrations (20, 40, 160 and 640 ng/ml) plated in six replicates across a microtitre plate. Inter-assay variation was determined by evaluating the ICA samples on three different microtitre plates for 3 days consecutively.

The ICA stock solution was spiked into PBS solutions at different volumes (0, 100, 200, 400 and 800 ng/ml), and recovery was determined using icELISA, as described above. The ratio of the value obtained for the known concentration was used to evaluate the matrix effect.

HPLC analysis was performed according to a previously reported protocol (29) with modifications. The HPLC system used in the present study was an Agilent 1260 Infinity with an Agilent ZORBAX SB-C18 column (Agilent Technologies, Inc., Santa Clara, CA, USA; 5 µm; 0.46×150 mm), maintained at room temperature. The components were separated by gradient elution using water (solvent A) and acetonitrile (solvent B) at a constant flow rate of 1.0 ml min⁻¹. The isocratic profile was as follows: 30% solvent B for 0–20 min. Each sample (10 ml) was injected and monitored at 270 nm. The column was held at 30°C.

Dried samples were pulverized and extracted with 50 ml of 70% aqueous ethanol solution in an ultrasonic bath at 50°C for 30 min prior to filtering. The processed samples were analyzed using HPLC. The samples were diluted in the effective measuring range of 10 to 1,000 ng/ml analyzed using icELISA.

In the present study, four types of TCM prescription were prepared in accordance with the traditional method (15). The Folium Epimedii extract and other herb extracts were prepared, as follows: Powdered herbs (30 g) were extracted with boiling water (60 ml) for 30 min and filtered through gauze. Following the addition of ethanol to 70%, the solution was stored at 4°C overnight and filtered. Subsequently, the solvents were vaporized to obtain dry powders, which were diluted with distilled water (10 ml). The samples were diluted in the effective measuring range of 10 to 1,000 ng/ml analyzed using icELISA.

Six-week-old female BALB/c mice were immunized with the ICA-BSA conjugate, and the desired hybridoma-secreting MAb against ICA was cloned using the limiting dilution method following screening using icELISA. The anti-ICA MAb was successfully obtained for further experiments.

As shown in Fig. 1, competitive inhibition occurred between the MAb and ICA-OVA with various concentrations of ICA, resulting in a calibration curve for ICA, analyzed using icELISA. The concentrations of the standard solutions were 9.77, 19.53, 39.06, 78.12, 156.25, 312.50, 625.00 and 1,250.00 ng/ml. Under these conditions, a linear regression coefficient of 0.9828 and linear regression equation, y = −0.119Lg(x) + 1.3644 (9.77 ng/ml - 1.25 µg/ml), were achieved with a half maximal inhibitory concentration of 156 ng/ml. The full measuring range of the assay extended between 10 and 1,000 ng/ml.

As shown in Table I, several compounds associated with ICA were found to have cross-reactivity with the anti-ICA MAb, including polygala acid (0.20%) and quercetin (0.26%). The MAb had no cross-reactivity with other compounds (<0.09%).

Reproducibility and precision are important criteria for an immunoassay. Standard curves for the ICA analyzed using icELISA on 3 days (consecutive) were compared and the variations were calculated. The variations between the well-to-well (intra-assay) and plate-to-plate (inter-assay) replicates were measured. As shown in Table II, the intra-assay RSDs were <4% and the inter-assay RSDs were <10%, which indicated that this assay was accurate and stable.

Various quantities of ICA were added to the Folium Epimedii extract, following which the spiked-in samples were dissolved and mixed evenly, and ICA content was calculated using icELISA. For each level, six samples were analyzed. The levels of ICA in the Folium Epimedii extract were determined using icELISA. ICA recovery ranged between 93.80 and 118.80%, with an average of 105.00% (Table III).

In the present study, the four TCM prescriptions, which varied in their composition ratios, were determined using icELISA without pre-treatment. The concentrations of ICA in the four TCM prescriptions are shown in Table IV.

Powdered samples were extracted as described above. As shown in Table IV, the contents determined using ELISA were consistent with those determined using HPLC.

The results of the experiments in the present study showed that the anti-ICA MAb had high sensitivity and specificity, and that the icELISA method had its own unique advantages, compared with HPLC. It is typical that intra-assay variations are generally lower, compared with inter-assay variations. The factors contributing to these variations are considered to include the hapten quality, coating, plate wells and multichannel pipettor, edge effects due to evaporation, uneven temperature during incubation, and day-to-day variation in the preparation of the reagents. Thus, the creation of a novel standard curve is required every time to reduce the variation.

To the best of our knowledge, the present study developed the first MAb against ICA and established the subsequent ELISA method, which provided a simpler, more efficient and sensitive approach for determining the ICA content in drug materials and biological samples. This method can also serve as a useful tool for investigating the pharmacokinetics and targets of ICA.