Development of a test strip for rapid detection of Gymnodinium catenatum

Harmful algal blooms (HABs) are major ecological and environmental problems in China’s coastal waters and seriously threaten the stability of the marine ecosystem and human health. Gymnodinium catenatum is a toxic red tide dinoflagellate. It can produce paralytic shellfish toxins (PSP), which cause serious hazards to marine organisms, public health, and safety. In this paper, a test strip based on colloidal gold immunochromatography (GICG) was developed for the rapid detection of Gymnodinium catenatum. The experimental results showed that the test strip has good specificity and sensitivity. It not only detects the different components of Gymnodinium catenatum but also may detect algal toxins. The lowest density of Gymnodinium catenatum that can be detected by this test strip is approximately 120 cells/mL. Cross-reaction indicated that the test strip had a high specificity for Gymnodinium catenatum. This test strip provides a rapid method for in situ detection of Gymnodinium catenatum and a reference method for the monitoring of other harmful algae to serve as an early warning of upcoming red tides. It also provides a new way to prepare more detection methods for toxic algal toxins.


Introduction
Harmful algal blooms are a major ecological and environmental problem in China's coastal waters and seriously threaten the stability of marine ecosystems, the healthy development of the mariculture industry, and human health (Gu et al., 2022). Gymnodinium catenatum is a harmful algae that can produce paralytic shellfish toxins. The toxins can accumulate in a variety of commercially harvested seafood species, particularly in bivalve shellfish by suspension-feeding on toxic dinoflagellates (Costa et al., 2015;Liu et al., 2020). Poisonous shellfish usually appear, taste, and smell no different than unaffected animals. Poisoning occurs when people accidentally eat contaminated seafood (Halstead & Schantz, 1984). The toxins can cause numbness of the fingers and extremities, tingling, nausea, and vomiting at low levels of exposure, and also can result in muscular paralysis and death by respiratory paralysis and cardiovascular shock at higher doses (Costa et al., 2015;Narahashi & Moore, 1968). Therefore, it is very important to establish an accurate and rapid detection method of harmful microalgae for the emergency response to red tide.
Abstract Harmful algal blooms (HABs) are major ecological and environmental problems in China's coastal waters and seriously threaten the stability of the marine ecosystem and human health. Gymnodinium catenatum is a toxic red tide dinoflagellate. It can produce paralytic shellfish toxins (PSP), which cause serious hazards to marine organisms, public health, and safety. In this paper, a test strip based on colloidal gold immunochromatography (GICG) was developed for the rapid detection of Gymnodinium catenatum. The experimental results showed that the test strip has good specificity and sensitivity. It not only detects the different components of Gymnodinium catenatum but also may detect algal toxins. The lowest density of Gymnodinium catenatum that can be detected by this test strip is approximately 120 cells/mL. Cross-reaction indicated that the test strip had a high specificity for Gymnodinium catenatum. This test strip provides a rapid method for in situ detection of Gymnodinium catenatum and a reference method for the monitoring of other harmful algae to serve as an early warning of upcoming red Vol:. (1234567890) Currently, the monitoring and identification methods of red tide algae mainly include analysis methods based on morphological characteristics, cytochromes, molecular probes, and the intersection of different technologies, such as flow cytometry (Pearson et al., 2021;Romero-Martinez et al., 2017). Although these technologies can detect red tide algae, they have common shortcomings in that they need instruments or professionals to assist in detection (Liu et al., 2022). For the sudden occurrence and rapid growth of red tides, these methods cannot be adapted for use in daily large-scale monitoring, especially high-frequency detection during the peak period from April to June every year in China's coastal cities, such as Fujian Province and Qingdao Province (Gu et al., 2022). For example, a massive bloom of Gymnodinium catenatum broke out in the Taiwan Strait in June 2017, causing serious PSP events and economic losses Zhang et al., 2020).
Colloidal gold immunochromatography (GICA) is a rapid detection technology for solid-phase immunolabeling based on immunochromatography, monoclonal antibodies, and colloidal gold labeling. This technology uses colloidal gold as a tracer marker in antigen antibody-specific immune responses (Wan et al., 2022). Colloidal gold is a gold particle formed by HAuCl 4 under the action of reducing agents, such as trisodium citrate, ascorbic acid, and tannic acid. Colloidal gold particles have a high electron density and are negatively charged in a weakly alkaline environment (Gao et al., 2005). They can quickly absorb biopolymers, such as proteins with positively charged groups, through electrostatic interactions to form stable complexes (Bendayan, 2000). Compared with other detection methods, GICA has the advantages of fast and sensitive results. It also does not need any instrument and can be directly observed with the eyes (Lin et al., 2020;Sun et al., 2005). Therefore, it is widely used in many fields, such as food security and medicine (Shim et al., 2006). However, the application of GICA is rare in the detection of microalgae and algal toxins.
In this study, a test strip based on the application of the colloidal gold immunochromatography assay method was developed for the rapid detection of Gymnodinium catenatum. The test strip has the advantages of a simple operation process and high sensitivity and specificity. The test results can be observed with the eyes directly within 3-10 min.
Moreover, the toxins produced by Gymnodinium catenatum can also be detected by this test strip. Therefore, it is very suitable for the on-site rapid detection of large quantities of samples. The test strip does not require professionals to operate. In the monitoring of daily aquaculture water, the measured seawater can be directly dropped on the test strip without treatment, and the existence of Gymnodinium catenatum in the seawater can be judged by the color of the test strip. It is easy to be mastered at the grassroots level and can be popularized in a large area.

Mouse toxicity test
Five female mice (specific pathogen free, SPF) purchased from the Experimental Animal Center of Xiamen University were numbered 1-5 and weighed 20-25 g. The mice were intraperitoneally injected with different concentrations of Gymnodinium catenatum. After the injection, timing was conducted to observe the state of the mice after injection.

Preparation of polyclonal and monoclonal antibodies
When Gymnodinium catenatum grew to the exponential growth stage, it was fixed with 2% formaldehyde overnight and centrifuged for 10 min at 106 × g. The algal cells were collected and washed twice with PBS, and the algal cells were collected by centrifugation at 106 × g for 10 min each time. Then, the collected algal cells were resuspended in 1 mL PBS and mixed with equal volumes of Freund's complete adjuvant and Freund's incomplete adjuvant(F5506-10 × 10 mL, USA) (Li et al., 2022). After that, Balb/c mice were immunized with the prepared adjuvant, and the immune dose was approximately 2 × 10 6 cells/mL. The algae cells emulsified with Freund's complete adjuvant were used for the initial immunization. After 10 days of immunization, the algae cells emulsified with Freund's incomplete adjuvant were used to immunize the mice 4 times, with a time interval of 7 days/time. One week after the last immunization, whole blood was collected, placed at room temperature for 2 h, and then centrifuged at 425 × g for 10 min; then, the serum was collected. After that, the titer of polyclonal antibody in the serum was determined by indirect ELISA.
A total of 1 × 10 5 cells/mL Gymnodinium catenatum was coated in an ELISA plate overnight. Then, the ELISA plate was washed 3-5 times with PBST. After washing, the 5% BSA solution was sealed in a 37 °C water bath for 1 h, and then, the plate was washed again. The primary antibody (serum of immunized mice) was added with gradient dilution and blocked in a 37 °C water bath for 1.5 h. After washing the plate, the secondary antibody (HRP enzyme-labeled antibody IgG) was added, and the plate was sealed in a 37 °C water bath for 1.5 h. After washing the plate, TMB substrate was added and the light reaction was avoided for 10-15 min, and then 2 mol L −1 H 2 SO 4 was added to terminate the reaction. Finally, the OD 450 per well was determined by a microplate reader.
When the serum titer met the requirements, the spleen cells of the mice were taken, and monoclonal antibodies were prepared using in vitro hybridoma technology (Hnasko & Stanker, 2015;Zhang, 2012).

Preparation of colloidal gold test strip
Colloidal gold was prepared by the trisodium citrate reduction method (Li et al., 2009;Shi et al., 2008). Fifty milliliters of 0.01% chloroauric acid solution was heated to 110 °C under 1 × g magnetic stirring, and 2 mL of 1% trisodium citrate aqueous solution was quickly added. During the reaction, the temperature and stirring speed remained unchanged, boiling until chloroauric acid turned red. Then, the monoclonal antibody was labeled with colloidal gold and purified according to a previously reported method (Xu et al., 2006). The monoclonal antibody of Gymnodinium catenatum and the antibody obtained by immunizing sheep with this antibody were scribed on the NC membrane and dried.
Next, the test strip was assembled. The strip consisted of four parts, including a glass fiber membrane sample pad, a glass fiber membrane antibody labeled bonding pad, the nitrocellulose membrane detection layer (NC membrane), and a glass fiber membrane water absorbent pad. These four parts were successively pasted onto a plastic carrier plate. The assembly of the test strip was carried out as previously reported (Guo et al., 2015;Lin et al., 2020).

Verification of test strip
Tests of the effectiveness, specificity, and sensitivity of the test strips and different algae components solutions were carried out. Different algae solutions were treated as follows.

Effectiveness test
Standard solution of Gymnodinium catenatum: the algae cell precipitate was collected, and the algae cell masses were cleaned once with distilled water and twice with PBS buffer. Then, the algae cells were collected each time by centrifugation at 4 °C, 38 × g, 5 min, and suspended again with PBS buffer. Culture medium of algae cells: Gymnodinium catenatum in the stable growth stage were centrifuged at 38 × g at 4 °C for 5 min. The supernatant was filtered with a 0.21-μm filter. Algal cell extract: Gymnodinium catenatum were ultrasonically broken in an ice water bath (15 min, 22%, 3 s, stop 2 s) and centrifuged at 23 × g at 4 °C for 5 min. The supernatant was filtered by a 0.21μm filter, and the collected filtrate was put into a cryo-storage tube.

Sensitivity test
To determine the detection limit of test strip, concentration about 3 × 10 4 cells/mL of Gymnodinium catenatum algal solution was diluted by PBS to different concentration: 15,000 cells/mL, 6000 cells/mL, 3000 cells/mL, 600 cells/mL, 120 cells/mL, 60 cells/mL. Each concentration was repeated three times.
To further verify the specificity of the test strip, we mixed the algae. Five common red tide algae were prepared. There were Alexandrium tamarense (2.5 × 10 4 cell/mL), Prorocentrum lima (3.1 × 10 4 cell/mL), Alexandrium catenella (1.9 × 10 4 cell/mL), Skeletonema costatum (6.8 × 10 4 cell/mL), and Gymnodinium catenatum (2.3 × 10 4 cell/mL). Dilute these five algae 10 times and 100 times respectively. Mix 1 mL of each algae at three concentration gradient separately. The mixture was divided into two groups: one was the mixture of five algae and the other was the mixture of four algae except Gymnodinium catenatum.
All algae were provided by Center for Collections of Marine Algae, Xiamen University. A total of 60 ~ 80 μL of the sample solution was directly dropped onto the test strip, and it was incubated for 3 ~ 15 min. The experimental results were observed after the color was stable.

Preparation and identification of monoclonal antibodies
The color developing hole with a final detection absorbance value of P (serum to be tested)/N (negative control serum) ≥ 2.1 and P (serum to be tested) > 0.1 was positive. The experimental results are shown in Table 1. The experimental results show that when the dilution ratio was 64,000, P/N = 3.25 ≥ 2.1. Therefore, the titer of the polyclonal antibody against Gymnodinium catenatum was 64,000, which laid a foundation for the preparation and screening of subsequent monoclonal antibodies.

Structure of the test strip
As shown in Fig. 1a, the test strip was based on a PVC board with glass fiber membrane water absorbent pad on the right and the nitrocellulose membrane detection layer (NC membrane) in the middle. A glass fiber membrane sample pad and a glass fiber membrane bonding pad were sequentially glued and fixed to one side of the PVC board. The bonding pad contained a monoclonal antibody against Gymnodinium catenatum labeled with colloidal gold. The NC membrane coated with two kinds of antibodies at two positions is called the test line (T-line) and control line (c-line). The test line contained the monoclonal antibody of Gymnodinium catenatum, while the control line contained the sheep anti-mouse IgG antibody.
After adding the sample, it would diffuse through the filter membrane. If the sample contained the antigen, in the process of diffusion, the antigen would first combine with the antibody marked with colloidal gold coated in the glass fiber membrane sample layer to form an antigen-colloidal gold-labeled antibody complex. Then, the complex continued to spread forward and combined with the monoclonal antibody of Gymnodinium catenatum to form a double antibody sandwich complex of monoclonal antibody-antigencolloidal gold-labeled antibody, resulting in the color of the T-line. The excess colloidal gold-labeled antibody continued to spread forward and bound to sheep anti-mouse IgG antibody, developing color in the C-line. Therefore, the final result was positive (Fig. 1a, c). If the sample did not contain the antigen, the colloidal gold-labeled antibody only contacted the sheep anti-mouse IgG antibody, and only the c-line was colored. The final result was negative (Fig. 1b,  d). If neither T-line nor C-line was colored, the test strip was invalid (Fig. 1e). The similar detection principle was also widely used in the detection of viruses and toxic residues in food and water (Lin et al., 2020;Yu et al., 2019;Zhu et al., 2008).

Test of different algae components
Eighty microliters of different components of Gymnodinium catenatum were dropped directly onto the sample pad of the colloidal gold strip. Then, the cells were incubated for 3 min to 15 min to observe the results. After stable color rendering, we observed whether the detection line (T-line) and quality control line (C-line) displayed color and color depth. The results of the test strip reaction are shown in Fig. 2. The four kinds of algal components including standard solution of algae cells, culture medium of algae cells, algal cell extract, and algal cell stock solution could make the test strip react positively. And, the higher the concentration was, the more obvious the color was (Fig. 2g, h). The test strip could detect various components of Gymnodinium catenatum. The algal cell standard solution (Fig. 2a, b) could make the test strip react. The color effect of algal cell culture medium (Fig. 2c, e) was equal to or even more obvious than that of untreated algal solution dropped on the test strip (Fig. 2g, h). This result showed that the substances secreted by physiological activities of Gymnodinium catenatum might also be used as the antigen of the test strip to make the test strip react. The algal cell Fig. 1 Structure of the test strip: a, b Side view of the test strip; c, d, e front view of the test strip extract also could make the test strip have a color reaction, but the color reaction degree was lower than that of the algal cell standard solution (Fig. 2a, b, f). Two possible explanations for this finding were that ultrasonic fragmentation might have an impact on the structure of the antigen, resulting in a poor detection effect, or the loss of some algal cells after PBS buffer cleaning could have resulted in a decrease in concentration.
To verify the validity of the test strip, different components of algae including standard solution of algae cells, culture medium of algae cells, algal cell extract, and algal cell stock solution were prepared in the experiment. It was to verify whether the test strip can detect other antigen components related to Gymnodinium catenatum in addition to detecting complete algae cells. The standard solution of algae cells only contained Gymnodinium catenatum, mainly surface antigens of G. catenatum. The culture medium of algae cells mainly included the antigens secreted by algal cells, which may be toxins. The algal cell extract were mainly internal and external antigens of G. catenatum. Algal cell stock solution was pure cultured algal solution. In addition to algal cells, there were also culture media and antigens produced during algal culture.
The results of the mouse toxicity test Mice were injected with algal cells and supernatant of different concentrations, and the activity of mice was observed and recorded (Table 2). Mice 1, 2, and 3 were intraperitoneally injected with 3.6 × 10 4 , 6.8 × 10 3 , and 2.04 × 10 3 cells of Gymnodinium catenatum respectively. The results showed that with the increase in the concentration of injected algal cells, the mice died faster. The results of mice 4 and 5 which were injected supernatant and supernatant filtrate of Gymnodinium catenatum respectively suggested that the algal cell culture medium was toxic.
In the bloom of Gymnodinium catenatum, research had found that saxitoxin (STX), decarbamoyl saxitoxin (dcSTX), the N-sulfocarbamoyl gonyautoxins (GTX1-GTX5), and the less toxic N-sulfocarbamoyl-11-hydroxysulfate C-toxins (C1-C4) were the main components of PST in Gymnodinium catenatum (Negri et al., 2007). These toxins are tetrahydropurine derivative with high water solubility and could exist in the supernatant of algae culture liquid. In the analysis of Gymnodinium catenatum toxin, the supernatant was also collected by centrifugation for toxin isolation and identification (Lin et al., 2022).
Combined with the color development results of the above test strip (Fig. 2c, e), it was shown that the strip not only detects toxic Gymnodinium catenatum but possibly also algal toxins. However, whether the test strip detected the toxin secreted by Gymnodinium catenatum still needs to be further verified by experiments such as HPLC-MS/MS. It provided a research basis for the development of the test strip for the detection of algal toxins in the future.

Sensitivity test
Cultured Gymnodinium catenatum at a density of 3 × 10 4 cells/mL was diluted to determine the lowest concentration that could be detected. Each concentration was repeated three times. The results of this test Fig. 3 The result of the sensitivity test: a 30,000 cell/mL, b 15,000 cell/mL, c 6000 cell/mL, d 3000 cell/ mL, e 600 cell/mL, f 120 cell/mL, g 60 cell/mL were judged within 5 min after the reaction started. The experimental results showed that when the algal cell solution with a concentration of 3 × 10 4 cells/mL was used for detection, the color of the T-line was the deepest, and the color of the T-line gradually became lighter with the decrease in the concentration of the detected algal solution. The algal cell solution with a final concentration of 60 cells/mL could not make the T-line of the test strip have a color reaction. Therefore, the minimum concentration of Gymnodinium catenatum detected by this test strip was approximately 120 cells/mL.
Compared with the lowest density, the outbreak of Gymnodinium catenatum was approximately 500 cells/ mL, and the results indicated that the strip had high sensitivity (Hong, 2018). Therefore, with rapid detection and high sensitivity, the test strip could play a role in the early warning of the outbreak of Gymnodinium catenatum red tide and is suitable for on-site red tide monitoring (Fig. 3).

Specificity test
Some common kinds of red tide algae and algae similar to Gymnodinium catenatum were used to verify the specificity of the test strip. Sixty microliters of the sample solution was directly dropped on the test strip. The results showed that the T-line of the test strip could be colored only when the algal solution contained Gymnodinium catenatum but could not be colored by other algal solutions or PBS of the control group. Moreover, Karenia mikimotoi and Gymnodinium catenatum belong to the same order, but they could not make the T-line colored when testing. This result indicated that the test strip has high specificity (Fig. 4).
In the experiment of mixed algae, the results showed that the T-line of the test strip could not be colored by PBS only (Fig. 5g). In the detection of Gymnodinium catenatum only (Fig. 5h−j), the T-line could be colored when the concentration were 2.3 × 10 4 cell/mL and 2.3 × 10 3 cell/mL (Fig. 5h, i) and could not be colored with the concentration of 2.3 × 10 2 cell/mL. The mixed algal solution of five algae (including Gymnodinium catenatum) with a concentration of about 1 × 10 4 cells/mL could make the T-line color (Fig. 5a). However, the mixed algal solution of four kinds of algae (except Gymnodinium catenatum) with approximately 1 × 10 4 cell/mL could not make the T-line color (Fig. 5b). Moreover, the mixed algal solution of five algae (including Gymnodinium catenatum) with a concentration of about 1 × 10 3 cell/mL could make the T-line color (Fig. 5c), but the color was lighter. Similarly, the four mixed algal solutions without Gymnodinium catenatum could not make the color of T-line develop at a concentration of about 1 × 10 3 cells/mL (Fig. 5d). When Fig. 4 The result of the specificity test: a Gymnodinium catenatum; b Karenia mikimotoi; c Alexandrium tamarense; d Prorocentrum donghaiense; e Alexandrium catenella; f Thalas-siosira weissflogii; g Skeletonema costatum; h Phaeodactylum tricornutum Bohlin; i Nitzschia closterium; j PBS only; k mixing of the above algae the algal mixed solution was about 1 × 10 2 cell/mL, T-line also did not develop color (Fig. 5e, f) because the cell number of G. catenatum was too low after mixing.

Conclusions
In this experiment, we developed a test strip for the rapid detection of harmful microalgae by preparing monoclonal antibodies against Gymnodinium catenatum, combined with colloidal gold labeling technology and immunochromatography technology. The innovation of this research is to develop a test strip for rapid detection of Gymnodinium catenatum, which can also detect toxins produced by Gymnodinium catenatum. At present, most of the international and domestic studies on the application of immunological methods to detect red tide algae and their toxins are based on the preparation of antibodies and ELISA detection methods, while there are few studies on the use of polyclonal antibodies and monoclonal antibodies to prepare GICA (colloidal gold immunochromatography) for detection of phytoplankton; any products that can be applied to practice have not been reported. The difference between GICA and other immunological methods in the detection of red tide algae and their toxins is that the GICA reaction time is greatly shorter. It does not need other complex instruments or equipment or highly demanding professional technicians. Compared with the ELISA method, the immunochromatography strip method use the colloidal gold particles instead of enzyme. So, there is no need of special facilities, skilled technician, or washing steps by using the test strip. Also, the temperature becomes less influential on this testing method (Grif et al., 2007). The test strip could get the result less than 10 min, whereas ELISA takes Fig. 5 The result of the specificity test of mixed algae: a mixture of five algae (concentration of each kind of algae is about 1 × 10 4 cell/mL; b mixture of four algae except Gymnodinium catenatu (concentration of each kind of algae is about 1 × 10 4 cell/mL); c mixture of five algae (concentration of each kind of algae is about 1 × 10 3 cell/mL); d mixture of four algae except Gymnodinium catenatu (concentration of each kind of algae is about 1 × 10 3 cell/mL); e mixture of five algae (concentration of each kind of algae is about 1 × 10 2 cell/mL); f mixture of four algae except Gymnodinium catenatu (concentration of each kind of algae is about 1 × 10 2 cell/mL); g PBS only; h Gymnodinium catenatum only (2.3 × 10 4 cell/mL); i Gymnodinium catenatum only (2.3 × 10 3 cell/mL); j Gymnodinium catenatum only (2.3 × 10. 2 cell/mL) days to get results. Therefore, the test strip using colloidal gold immunochromatography is an ideal and rapid determination method for monitoring of Gymnodinium catenatum. The pretreatment procedure for samples is simple, even in regard to how to process the samples. Once the color of T-line and c-line in the test strip is developed, it will hardly become light or fade. The test strip has good stability, specificity, and high sensitivity and can realize the rapid detection of red tide on site.
Acknowledgements The authors gratefully acknowledge comments from Kefu Zhou that greatly improved the manuscript.
Author contribution All authors contributed to the study conception and design. Shuyue Li, Xiaoxiao Liu, and Weixin He did main experiments (algal culture, algal antigen treatment, and test strip verification); Lingyue Li and Jiazhao Zhang were involved in data analysis; Kefu Zhou was involved in the preparation of polyclonal antibody; Changgong Zhang was involved in the preparation of monoclonal antibody and assembly of test strip. The first draft of the manuscript was written by Junhua Fang.
Funding This study was supported by the Fujian fishery resources monitoring center offshore toxic (harmful) red tide algae and toxins rapid detection method development and application service procurement project (FZJZ-2020-1).
Data availability Data generated during and/or analyzed during the current study is available from the corresponding author on reasonable request.

Declarations
Ethical approval The ethical code of animal experiment in this paper is XMULAC20220020.

Competing interests
The authors declare no competing interests.