A novel linear displacement isothermal amplification with strand displacement probes (LDIA-SD) in a pocket-size device for point-of-care testing of infectious diseases

Abstract

Nucleic acid amplification is crucial for disease diagnosis, especially lethal infectious diseases such as COVID-19. Compared with PCR, isothermal amplification methods are advantageous for point-of-care testing (POCT). However, complicated primer design limits their application in detecting some short targets or sequences with abnormal GC content. Herein, we developed a novel linear displacement isothermal amplification (LDIA) method using two pairs of conventional primers and Bacillus stearothermophilus (Bst) DNA polymerase, and reactions could be accelerated by adding an extra primer. Pseudorabies virus gE (high GC content) and Salmonella fimW (low GC content) genes were used to evaluate the LDIA assay. Using strand displacement (SD) probes, a LDIA-SD method was developed to realize probe-based specific detection. Additionally, we incorporated a nucleic acid-free extraction step and a pocket-sized device to realize POCT applications of the LDIA-SD method. The LDIA-SD method has advantages including facile primer design, high sensitivity and specificity, and applicability for POCT, especially for amplification of complex sequences and detection of infectious diseases.

1. Introduction

Nucleic acid amplification is a powerful technique for clinical disease diagnosis, such as early detection of tumour and lethal infectious diseases [16], [9]. In the case of COVID-19, nucleic acid detection proved to be an excellent screening method [11], [3]. PCR-based techniques are commonly used for nucleic acid detection due to facile primer design. However, making PCR meet the requirements of point-of-care testing (POCT) is challenging. PCR is usually performed using precise and expensive thermal cycling equipment, whereas POCT assays tend to be more easily operated and handled [7].

Compared with PCR, isothermal amplification methods have significant advantages for POCT, and novel methods are emerging. For example, loop-mediated isothermal amplification (LAMP) [15], cross-priming amplification (CPA) [25] and isothermal multiple-self-matching-initiated amplification (IMSA) [5] are single enzyme-based reaction strategies. By contrast, strand displacement amplification (SDA) [22], rolling circle amplification (RCA) [10], isothermal helicase-dependent amplification (HDA) [21], nucleic acid sequence-based amplification (NASBA) [14] and recombinase polymerase amplification (RPA) [13] are multiple enzyme-based methods.

LAMP and other single enzyme-based isothermal amplification methods only require Bacillus stearothermophilus (Bst) DNA polymerase and specific primers to initiate a rapid reaction. Therefore, they are low-cost and popular for POCT. However, rigorous target site requirements and complex primer design limit their amplification range [6]. For DNA sequences < 200 nucleotides (nt) or variable GC content, these approaches are not usually amenable [20]. Additionally, multiple primers with complex structures readily form primer dimers and generate nonspecific amplification signals during single enzyme-based reactions [12]. To overcome these shortfalls, novel amplification principles with simpler primer design are appropriate.

Herein, we developed a novel linear displacement isothermal amplification (LDIA) method using two pairs of conventional PCR primers and Bst DNA polymerase. Addition of one extra primer could efficiently accelerate the LDIA reaction. Using Pseudorabies virus gE (high GC content) or Salmonella fimW (low GC content) genes, the LDIA method was explored and verified. The results showed that LDIA assays were simple to perform, with high sensitivity and specificity similar with LAMP. Conveniently, the LDIA signal could be accurately reflected by strand displacement (SD) probes. Furthermore, we incorporated a nucleic acid-free virus extraction step and a pocket-sized device to make the LDIA-SD approach suitable for POCT, in which LDIA-SD reactions could be directly performed and monitored. The results revealed advantages for LDIA-SD methods including facile primer design, and high sensitivity and specificity. Thus, LDIA-SD is a good choice for POCT, especially for amplification and detection of complex sequences.

2. Materials and methods

2.1. Virus and bacteria

Pseudorabies virus (PRV) GD-WH (GenBank: KT936468.1), porcine circovirus 2 isolate HN6 (GenBank no: KM035762.1), Streptococcus suis serotype 2, Salmonella typhimurium (ATCC14028), Escherichia coli (ATCC25922), Campylobacter jejuni (NCTC 11168) and Staphylococcus aureus (CMCC26003) were from our laboratory stocks.

2.2. Regents

Bst 2.0 WarmStart DNA polymerase, MgSO₄ (100 mM) and 10 × Thermopol Isothermal buffer containing 200 mM Tris-HCl, 500 mM KCl, 100 mM (NH4)₂SO₄, 20 mM MgSO₄ and 1 % Tween-20 were purchased from New England BioLab (NEB, USA). A 10 mM deoxynucleotide (dNTP) solution was purchased from Vazyme Biotech (China). Eva Green dye was purchased from Biotium Inc. (USA). SYBR Green dye was purchased from Solarbio Science & Technology (China). Tween-20, Triton X-100, dithiothreitol and guanidine hydrochloride were purchased from Sangon Biotech (China).

2.3. DNA extraction using commercial kits

Viral DNA was purified using a HiPure Viral RNA/DNA Kit (Magen, China) according to the instructions of the manufacturer. Finally, viral DNA was dissolved in 50 μL of nuclease-free water and stored at − 80 °C. Bacterial DNA was extracted using HiPure Bacterial DNA Kit (Magen, China) according to the manufacturer’s protocol.

2.4. Standard template preparation

PRV gE and Salmonella fimW genes were amplified by PCR using primers listed in Table S1 and S2. PCR products were purified using a Cycle Pure Kit (Omega, USA) according to the instructions of the manufacturer. Purified fragments were cloned into the pMDTM19-T vector and transformed into DH5α competent cells (TaKaRa Biotechnology, China). Plasmid DNA was obtained using a Plasmid Mini Kit I (Omega, USA) and used as template.

2.5. Primer design

Conventional PCR primers for the LDIA method were designed using Primer Premier 5.0 software. LAMP primers for the PRV gE gene (GenBank: KT936468.1) were designed using online PrimerExplorer V5 software (http://primerexplorer.jp/lampv5e/index.html). LAMP primers for the Salmonella fimW gene (GenBank: AP019375.1) were published in our previous report [24]. Primers for PRV and Salmonella are respectively listed in Table 1, Table 2. All primers were synthesised by Sangon Biotech.

Table 1.

Primers of the LDIA or LAMP method for PRV gE gene.

Primer name	Sequences 5′− 3′	Genome positiona
gE-LIF	ACGCGCTCGGCTTCCACT	683–700
gE-LIR1	AGACCACGCGCGGCATCAG	733–751
gE-LIR2	GCGCGAGTCGCCCATGTC	754–771
gE-LIR3	AGCGTGGCGGTAAAGTTCT	773–791
gE-LIR4	CGTAGTACAGCAGGCACCG	820–838
gE-LOF	ACGAGCCCCGCTTCCA	668–683
gE-LOR	AGATGCAGGGCTCGTACA	839–856
gE-LAR	TGTCCCCGGGCGAGAAGA	707–724
gE-LAR-probe	ATCAGGTCGAACGTGTCCCCGGGCGAGAAGA	707–737
gE-LAR-quencher	GGGGACACGTTCGACCTGAT	718–737
gE-FIP (F1c+F2)	AGACCACGCGCGGCATCAG-GCGCTCGGCTTCCACT	F1c: 733–751 F2: 685–700
gE-BIP (B1c+B2)	GAGAACTTTACCGCCACGCTGG-CGTAGTACAGCAGGCACCG	B1c: 772–793 B2: 820–838

^aThe genome of PRV gE gene (GenBank: KT936468.1)

Table 2.

Primers of the LDIA method for Salmonella fimW gene.

Primer name	Sequences 5′− 3′	Gene positiona
fimW-LIF	TTATCAGATACCTATGCATACCCA	183–206
fimW-LIR1	TGAACATGAGCTTTTCTTTATCGC	239–262
fimW-LIR2	AACATCATCTTCCCGATA	292–309
fimW-LIR3	TCCGGGTAATTTCTTCAACATC	304–325
fimW-LOF	CTGGATGATGATTGGTTCAG	154–173
fimW-LOR	GAAGGGACGCTATGTCGA	357–374
fimW-LAR	TACAAATAATCGCCCGTAGCTGAT	209–232

^aThe genome of Salmonella fimW gene (GenBank: AP019375.1)

2.6. LDIA and LAMP reactions

The mixtures for LDIA and LAMP reactions were consistent with typical isothermal methods [1], [23], [8], and contained 1 × Thermopol Isothermal Buffer, 1 × Eva Green dye, 1.6 mM dNTPs and 8 U Bst WarmStart DNA polymerase. The concentration of LDIA primers was optimized as 1.6 µM LIF/LIR, 0.8 µM LAR and 0.2 µM LOF/LOR. The optimal concentration of LAMP primers was 1.6 µM FIP/BIP, 0.8 µM LAR and 0.2 µM LOF/LOR. Reactions were performed on a Roche Light Cycler 96 real-time detection system (Roche, Switzerland). Reactions included 60 cycles at 63 °C for 1 min, and fluorescence signals were measured at the end of each cycle.

2.7. Identification of LDIA products

After real-time detection, LDIA products were subjected to 3 % agarose gel electrophoresis and stained with GoldView II Nuclear Staining Dyes (Solarbio, China). Gel images were visualized and captured using a ChemiDoc XRS Imaging System (Bio-Rad Laboratories, USA). After agarose gel electrophoresis analysis, a single band consistent with the theoretical size was cut and purified by using the Gel Extraction Kit (Omega, USA). The purified DNA was sequenced by Sangon Biotech.

2.8. Sensitivity analysis

Ten-fold serial dilutions of plasmid DNA harbouring the PRV gE gene or the Salmonella fimW gene were used to determine the detection limit of the LDIA method, and this was compared with the detection limit of LAMP.

2.9. LDIA-SD

To develop the LDIA-SD method, a pair of SD probes were designed using Primer Premier 5.0 software (Table 1). The length of the LAR primer was increased to 31 bp. It was labelled with fluorescein at the 5′ end and named LAR-probe. Meanwhile, a ∼20 bp complementary primer labelled with quencher at the 3′ end was used as the LAR-quencher. The reaction mixture for the LDIA-SD method contained 1 × Thermopol Isothermal Buffer, 1.6 mM dNTPs and 8 U Bst WarmStart DNA polymerase. The concentrations of LDIA-SD primer/probe were 1.6 µM for LIF/LIR, 0.3 µM for LAR-probe, 0.4 µM for LAR-quencher and 0.2 µM for LOF/LOR. For quantitative detection, reaction tubes were incubated in a Roche Light Cycler 96 real-time detection system (Roche). For direct analysis, reaction tubes were incubated in the pocket-sized instrument.

2.10. Nucleic acid-free virus extraction

A mixture of 100 mM Tris, 100 mM guanidine hydrochloride, 250 mM dithiothreitol, 5 % Triton-X100 and 5 % Tween-20 was prepared as nucleic acid-free extraction solution. Briefly, 100 μL of each sample was separately added to 400 μL of nucleic acid-free extraction solution to release DNA. Solutions were mixed three times and incubated at room temperature for 2 min, and then 1 μL of the mixture was used as the template for subsequent LDIA detection.

2.11. Incorporating a pocket-sized instrument

The pocket-sized device measured 80 mm (l) × 60 mm (w) × 64 mm (d). We designed it by using SolidWorks 2019 software (Concord, MA, USA) and 3D-printed in black resin. A temperature control module powered by a high-capacity lithium-ion battery was connected to two positive temperature coefficient ceramic heating plates attached to the back of the metallic pedestals using thermal grease. A compact heat transfer design between the heating plate and the metallic pedestals stablized the temperature of the pedestals by conducting excess heat through a miniature fan, thereby maintaining the temperature of the metallic pedestals at 63 °C. The lithium-ion battery was charged through a generic Type-C charging port, which increases the versatility of the device. On each metallic pedestal there are two mutually orthogonal microvoids; the bottom microvoid allows excited light to enter through a Band-Pass Filter (maximum passing wavelength of 498 nm, bandwidth of 7 nm). Reaction tubes in each metallic pedestal can be excited to generate fluorescence that is emitted from the microvoid on the side. The excitation light is provided by a set of light-emitting diodes (LEDs) with a peak wavelength of 494 nm. The emitted fluorescence passes through an optical lens and edge stray light is removed. Uniform, intense outgoing fluorescence then passes through a 495 nm narrow band filter positioned 4 cm away from reaction tubes and can be observed by the naked eye, while negative samples emit no fluorescence. The pocket-sized instrument has advantages of being small and easy to carry, providing adjustable temperature, observing the results by the naked eye, and being suitable for on-site detection based on fluorescence signal output. A national utility model patent has been applied for our pocket-sized instrument (Application Number: 202220951202.4).

2.12. POCT of clinical specimens by the LDIA-SD method

Oral fluid specimens of pigs suspected to be infected by PRV were collected by resident farm veterinarians and sent to our laboratory for routine detection and surveillance. The detection method was the national standard real-time PCR based on TagMan probes targeted to PRV gE gene (National Standard Number: GB/T 35911–2018). Primers and probes for real-time PCR were listed in Table S3. Samples collecting treatment were conducted in accordance with national and local laws and guidelines. All animal experiments were reviewed and approved by the ethical and ethics commission (Institute of Animal Health, Guangdong Academy of Agricultural Sciences, China). The license number was SYXK (Yue) 2011–0116. According to results of real-time PCR, a total of 30 oral fluid specimens were selected and used to evaluate the ability of LDIA-SD in POCT, including 18 positive ones with variant CT values and 12 negative specimens. These specimens were processed by nucleic acid-free virus extraction and analyzed by the LDIA-SD method in the pocket-sized device.

2.13. Statistical analysis

Statistical analysis was conducted by using GraphPad Prism 5 software. P values were calculated by using an unpaired Student’s t-test (*P < 0.05; **P < 0.01; ***P < 0.001).

3. Results and discussion

3.1. Principle of LDIA

Fig. 1 shows the principle of the LDIA process. Four common primers targeting the template initiate the LDIA reaction, including outer primers LOF and LOR, and inner primers LIF and LIR. Due to the higher concentration of inner primers relative to outer primers, they readily bind the template. A single-stranded DNA (ssDNA) is formed upon extension of the outer primers and the strand displacement activity of BST DNA polymerase, and a short double-stranded DNA (dsDNA) is formed by the inner primers. At ∼ 60 °C, these short DNA fragments (60–140 bp) undergo dsDNA breathing, reaching a balance between melting and annealing, even melting into ssDNA. Subsequently, LIF and LIR respectively bind during dsDNA breathing and generate new amplicons. These dsDNA become new templates and the reaction cycles continue. According to this principle, an accelerated primer (LAR) is included in the reaction, which can form a much shorter product with LIF or LIR, generating a great number of amplicons.

Fig. 1.

Principle of LDIA method. Firstly, inner primers (LIF and LIR) bind the target template and are extended. A single-stranded DNA (ssDNA) is formed from LIF or LIR. Extension of outer primers (LOF and LOR) occurs when the ssDNA displaces LIF or LIR on the target template, resulting in the formation of a short double-stranded DNA (dsDNA) from the inner primers. At ∼ 60 °C, these short DNA fragments (60–140 bp) undergo dsDNA breathing, reaching a balance between melting and annealing, even melting into ssDNA. LIF and LIR bind during dsDNA breathing to generate new amplicons. The resulting dsDNAs become new templates and the reaction cycle continues.

3.2. Proof of concept assay

To prove the principle of LDIA, we used the gE gene of PRV, a type of Suid herpes virus 1, as a model. Firstly, we evaluated the necessity of the outer primers for LDIA with a product length of 109 bp. A typical amplification curve is shown using Eva Green staining to monitor the real-time fluorescence of LDIA ( Fig. 2 A). In the absence of outer primers signals were delayed. The real-time fluorescence products were further analyzed by 3 % agarose gel electrophoresis. In addition to the expected 109 bp band formed by inner primers, ladder-like bands were produced by the LDIA reaction (Fig. 2B). Furthermore, DNA sequence analysis revealed that this 109 bp band was consistent with the theoretical product (Fig. 2 C).

Fig. 2.

LDIA proof of concept. (A) Real-time fluorescence curve for the LDIA method with and without outer primers. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA method with outer primers in the presence of 10⁴ copies of template; 2, LDIA method without outer primers in the presence of 10⁴ copies of template. (B) Agarose (3 %) gel electrophoresis analysis of real-time LDIA products. M, 50 bp DNA ladder; 1, LDIA method with outer primers in the presence of 10⁴ copies of template; 2, LDIA method without outer primers in the presence of 10⁴ copies of template. Red stars indicate expected bands. (C) Sequence analysis of expected bands for the real-time LDIA method. Bands were sequenced using the gE-LIF primer. Black underline indicates sequences complementary to the gE-LIR3 primer. (D) Real-time fluorescence curve for the LDIA method (with outer primers) with different length amplification products. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–4, LDIA method (with outer primers) with gE-LIF and gE-LIR1, gE-LIR2, gE-LIR3 or gE-LIR4 primers in the presence of 10⁴ copies of template. (E) Real-time fluorescence curve for the LDIA method (without outer primers) with different length amplification products. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–4, LDIA method (without outer primers) with gE-LIF and gE-LIR1, gE-LIR2, gE-LIR3 or gE-LIR4 primers in the presence of 10⁴ copies of template. (F) Agarose gel electrophoresis analysis of the LDIA method with different length amplification products. M, 50 bp DNA ladder; 1, 3, 5 and 7, LDIA method (without outer primers) with gE-LIF and gE-LIR1, gE-LIR2, gE-LIR3 or gE-LIR4 primers in the presence of 10⁴ copies of template. 2, 4, 6 and 8, LDIA method (with outer primers) with gE-LIF and gE-LIR1, gE-LIR2, gE-LIR3 or gE-LIR4 primers in the presence of 10⁴ copies of template. Red stars indicate expected bands.

Secondly, the applicability of LDIA to amplification products of different lengths was analyzed. We found that LDIA could amplify a product of at least 140 bp in length within 60 min (Fig. 2D). However, the reaction only yielded a typical amplification curve with a 69 bp product within 60 min when outer primers were absent. After 60 min, the amplification signal with a product longer than 69 bp was very weak (Fig. 2E). Agarose gel electrophoresis analysis also proved that fewer amplification products was produced when outer primers were absent (Fig. 2 F). These results confirmed that the outer primers are necessary for LDIA to function as intended. A previous report showed that strand exchange amplification (SEA) can amplify short sequences (40–45 bp) using a pair of common primers based on denaturation bubbles [19], [4]. In our study, only LIF and LIR primers failed to generate amplification products with lengths > 69 bp within 60 min. This might be attributed to the strict limit of product length for SEA [4], resulting in failed amplification of long products.

3.3. Effects of accelerating primers and temperature on LDIA

To develop a more robust LDIA reaction system, we explored the effects of additives. According to the LDIA principle, we speculated that it might be accelerated by factors enhancing the generation of shorter products or the formation of dsDNA denaturation bubbles. When adding LAR paired with LIF, threshold time (TT) values of LDIA with a 109 bp product were markedly reduced ( Fig. 3 A). This proved that an extra common primer was effective at accelerating LDIA, consistent with our prediction that shorter products were necessary for LDIA. It also supported our idea that the existence of dsDNA denaturation bubbles of short dsDNAs facilitates the LDIA principle. To evaluate the stability of LDIA amplification by adding LAR primers, we performed a concentration-dependent experiment. The results displayed a gradient acceleration effect of LAR primers on the LDIA assay. With a concentration of LAR primers up to 1.6 μM, there are no non-specific results, indicating the stability of LDIA amplification with LAR primers (Fig. 3B). To exclude the effects of fluorescent stains on LDIA, we compared the results of Eva Green and SYBR Green used in LDIA detection. The results were shown in Fig. 3 C, both of them can be used for LDIA assay at a suitable concentration, except an inhibitory effect of SYBR Green at high concentration.

Fig. 3.

Effects of accelerating primers and temperature on LDIA. (A) Real-time fluorescence curve for the LDIA method with or without addition of LAR primers. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA method with gE-LAR primers in the presence of 10⁴ copies of template; 2, LDIA method without gE-LAR primers in the presence of 10⁴ copies of template. (B) The concentration dependent analysis of LAR primers for the LDIA method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–6, Real-time fluorescence curves for the LDIA method (Positive control) with 0, 0.1, 0.2, 0.4, 0.8 and 1.6 μM LAR primers. 7–12, Real-time fluorescence curves for the LDIA method (Negative control) with 0, 0.1, 0.2, 0.4, 0.8 and 1.6 μM LAR primers. (C) The effect of fluorescence stains on the LDIA method with LAR primers. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–3, Real-time fluorescence curves for the LDIA method with 0.5 × , 1 × , 2 ×Eva Green dye. 4–6, Real-time fluorescence curves for the LDIA method with 0.5 × , 1 × , 2 ×SYBR Green dye. (C) Real-time fluorescence curve for the LDIA method over a reaction temperature range from 60 °C to 69.5 °C. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, 60 °C; 2, 61 °C; 3, 63 °C; 4, 65.9 °C; 5, 69.5 °C; 6, 72.5 °C; 7, 74.2 °C; 8, 75 °C.

Next, we analysed the influence of temperature on LDIA, which has been reported to play a positive role in the formation of dsDNA denaturation bubbles [26]. Considering the inactivation temperature of Bst DNA polymerase is ∼80 °C, assays were performed from 60 °C to 75 °C to balance high temperature and enzyme activities. As shown in Fig. 3D, the LDIA reaction proceeded across a temperature range from 60 °C to 69.5 °C, and the lowest CT values were observed at 63 °C. Therefore, the optimal temperature of the LDIA method was 63 °C. It has the same optimal temperature with many LAMP methods [27].

3.4. Sensitivity and specificity of LDIA assays

The sensitivity of LDIA was evaluated using 10-fold dilutions of dsDNA. Meanwhile, a LAMP method targeting the same region of the gene was designed for comparison. The main difference between LDIA and LAMP is that ∼150 bp products are targeted by LAMP inner primers because a product smaller than this is not compatible [15]. The results in Fig. 4B show that LDIA had a detection limit as low as 100 copies, comparable with LAMP (Fig. 4 A). Next, the amplification signal of LDIA was explored in excluding cross-reactivity for isothermal amplification, which has been reported for some LAMP assays in 90 min reactions [12]. For the LDIA method, the blank control yielded a specific signal during this reaction time (Fig. 4 C). Notably, the baseline of the LDIA method was lower than that of the LAMP method, especially for the negative control group. This might be attributed to non-specific amplification between LAMP primers, because LAMP require multiple primers for target amplification, the likelihood of primer dimer-based nonspecific amplification is increased. Further, specificity assay of LDIA indicated that it has no cross-reactivity with pathogens inducing similar clinical syndromes with PRV, including porcine circovirus 2, Streptococcus suis serotype 2 and Salmonella typhimurium (Fig. 4D). According to sensitivity and specificity assays, LDIA can have advantages similar to those of LAMP. However, we found that LAMP primers targeted to the gE gene were difficult to design using the online PrimerExplorer software. Due to the high GC content of this gene, only two groups of suitable PRV LAMP primers were generated. Therefore, we believe that facile primer design is another advantage of LDIA compared with LAMP.

Fig. 4.

Sensitivity and specificity of the LDIA assay. (A) Sensitivity of the LAMP method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–8, 10⁷, 10⁶, 10⁵, 10⁴, 10³, 10², 10 and 1 copies of templates. (B) Detection limit of the LDIA method. 1–8, 10⁷, 10⁶, 10⁵, 10⁴, 10³, 10², 10 and 1 copies of templates. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). (C) Amplification signal of the LDIA method over a 90 min incubation. In the statistical graph, each error bar represents the standard deviations at 3 replicates (n = 3) of endpoint fluorescence intensity of the LAMP products and LDIA products after 90 min reactions. 1, LAMP method in the presence of 10⁴ copies of template; 2, LAMP method negative control. 3, LDIA method in the presence of 10⁴ copies of template; 4, LDIA method negative control. (D) Specificity assay of the LDIA method for PRV detection. 1, PRV genome DNA; 2, Porcine circovirus genome DNA; 3, Streptococcus suis serotype 2 genome DNA; 4, Salmonella typhimurium genome DNA; 5, Negative control.

3.5. Applicability of LDIA for Salmonella detection

Subsequently, we utilised LDIA to detect the Salmonella fimW gene to test its universality. For this gene, we observed that outer and inner primers were efficient for amplifying a target sequence ranging from 80 bp to 143 bp ( Fig. 5A), and an accelerating primer markedly enhanced amplification (Fig. 5B). Sensitivity analysis showed that the detection limit was the same as that of LAMP, as low as 100 copies (Fig. 5C and 5D). As expected, no non-specific signal was observed for the Salmonella LDIA method within 90 min (Fig. 5E). Interestingly, the Salmonella LDIA method also displayed a lower baseline compared with LAMP methods, implying a decreased primer dimer-based nonspecific amplification. The results of specificity assay showed it has no cross-reactivity with other pathogens (Fig. 5F). Thus, decreased primer dimer-based nonspecific amplification is another advantage of LDIA.

Fig. 5.

Applicability of the LDIA method for Salmonella detection. (A) Real-time detection of the Salmonella fimW gene by the LDIA method with different length amplification products. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–3, LDIA method with fimW-LIF and fimW-LIR1, fimW-LIR2 or fimW-LIR3 primers in the presence of 10⁴ copies of template. (B) Real-time fluorescence curve for the LDIA method with or without addition of LAR primers. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA method with fimW-LAR primers in the presence of 10⁴ copies of template; 2, LDIA method without fimW-LAR primers in the presence of 10⁴ copies of template. (C) Detection limit for the Salmonella fimW gene by the LAMP method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–7, 10⁶, 10⁵, 10⁴, 10³, 10², 10 and 1 copies of templates. (D) Detection limit for the Salmonella fimW gene by the LDIA method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–7, 10⁶, 10⁵, 10⁴, 10³, 10², 10 and 1 copies of templates. (E) Amplification signal of the LDIA method targeting the Salmonella fimW gene over a 90 min incubation. In the statistical graph, each error bar represents the standard deviations at 3 replicates (n = 3) of endpoint fluorescence intensity of the LAMP products and LDIA products after 90 min reactions. 1, LAMP method in the presence of 10⁴ copies of template; 2, LAMP method negative control. 3, LDIA method in the presence of 10⁴ copies of template; 4, LDIA method negative control. (F) Specificity assay of the LDIA method for Salmonella detection. 1, Salmonella typhimurium genome DNA; 2, Escherichia coli genome DNA; 3, Campylobacter jejuni genome DNA; 4, Staphylococcus aureus genome DNA; 5, Negative control.

3.6. Combining LDIA with SD probes

Considering that all primers in the LDIA method share a common linear structure, we predicted that strand displacement (SD) probes might be compatible, since they have been used in LAMP to achieve probe-specific detection [17]. The 31 bp gE-LAR-probe was labelled with fluorescein at the 5′ end and a 20 bp complementary primer labelled with quencher at the 3′ end, forming a pair of SD probes. These complementary probes were included in the LDIA reaction mixture. Due to the strong complementary between these two probes, only robust amplification can exchange the quencher probe and result in a signal output. Robust amplification can exchange the quencher probe, resulting in a signal output ( Fig. 6 A). As shown in Fig. 6B, increasing the length of the NAR probe to 31 bp did not dampen the reaction. When this 31 bp LAR probe was labelled with fluorescein at the 5′ end and a 20 bp complementary primer labelled with quencher at the 3′ end, a pair of SD probes was formed. Using the LDIA method, a typical amplification curve was detected when SD probes were included in the reaction system (Fig. 6 C). Moreover, its sensitivity was as low as 100 copies, the same as that of the Eva Green-based LDIA method (Fig. 6D).

Fig. 6.

LDIA combined with SD probes. (A) Principle of the LDIA-SD method. The 31 bp gE-LAR-probe was labelled with fluorescein at the 5′ end and a 20 bp complementary primer labelled with quencher at the 3′ end, forming a pair of SD probes. These complementary probes were included in the LDIA reaction mixture. Robust amplification can exchange the quencher probe, resulting in a signal output. (B) Effect of the LAR probe on the LDIA method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA method with the 31 bp LAR probe in the presence of 10⁴ copies of template; 2, LDIA method with common LAR primers in the presence of 10⁴ copies of template. (C) Real-time fluorescence curve for the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA-SD method in the presence of 10⁴ copies of template; 2, Negative control. (D) Sensitivity of the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–7, 10⁶, 10⁵, 10⁴, 10³, 10², 10 and 1 copies of templates.

3.7. Incorporating nucleic acid-free extraction in LDIA-SD

Furthermore, a nucleic acid-free extraction solution was applied in LDIA-SD for POCT. As a precondition, we evaluated the influence of nucleic acid-free extraction mixture components, including Tween-20, Triton-X100, dithiothreitol and guanidine hydrochloride. Tween-20, Triton-X100 and dithiothreitol stimulated the LDIA method ( Fig. 7A, 7B and 7 C), but a high concentration of guanidine hydrochloride has an inhibitory effect (Fig. 7D). This was not consistent with a previous report showing that guanidine hydrochloride had a positive effect on LAMP [28]. Surprisingly, the mixture stimulated LDIA-SD, and did not have an inhibitory effect (Fig. 7E). We speculate that the stimulatory effect might be related to the influence of Tween-20, Triton-X100 and dithiothreitol on the formation of dsDNA denaturation bubbles because they are aqueous surfactants. However, this needs to be further explored. Regardless, the nucleic acid-free extraction mixture performed well when used to directly detect PRV, compared with a commercial nucleic acid extraction kit (Fig. 7 F). Therefore, our LDIA-SD method based on nucleic acid-free extraction has great potential for POCT.

Fig. 7.

Incorporating a nucleic acid-free extraction step in LDIA-SD. (A) Influence of Tween-20 on the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–5, Real-time fluorescence curves for the LDIA-SD method with 0, 0.125 %, 0.25 %, 0.5 % and 1 % (v:v) Tween-20. (B) Influence of Triton-X100 on the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–5, Real-time fluorescence curves for the LDIA-SD method with 0, 0.125 %, 0.25 %, 0.5 % and 1 % (v:v) TritonX-100. (C) Influence of dithiothreitol on the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–5, Real-time fluorescence curves for the LDIA-SD method with 0, 25 mM, 50 mM, 100 mM and 200 mM dithiothreitol. (D) Influence of guanidine on the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1–4, Real-time fluorescence curves for the LDIA-SD method with 0, 25 mM, 50 mM and 100 mM guanidine. (E) Enhancement of LDIA-SD with a nucleic acid-free extraction step. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA-SD method with a nucleic acid-free extraction step in the presence of 10⁴copies of template. 2, LDIA-SD method without a nucleic acid-free extraction step in the presence of 10⁴ copies of template. (F) Nucleic acid-free extraction and direct detection of PRV using the LDIA-SD method. In the statistical graph, each error bar represents the standard deviations of TT values at 3 replicates (n = 3). 1, LDIA-SD method in the presence of sample template including a nucleic acid-free extraction step; 2, LDIA-SD method in the presence of sample template using a commercial nucleic acid extraction kit.

3.8. LDIA-SD-based POCT using a pocket-sized device

To establish a user-friendly LDIA-SD method for POCT, we designed a pocket-sized device for simultaneous isothermal incubation and fluorescence screening. The basic working principle of this device is explained in Fig. 8 A. Metallic pedestals were utilised to heat the reaction tube. In addition, a microvoid was designed in the front and back of the metallic pedestals. The front microvoid facilitated fluorescence observation, while the back microvoid was stimulated by light at 494 nm. To ensure a stable temperature for the metallic pedestals, a temperature control procedure was incorporated (Fig. 8B). Using this pocket-sized device (58 mm × 62 mm × 75 mm), the results of LDIA-SD can be judged through the observation window by a smartphone (Fig. 8 C). Positive results yield fluorescence emission, while negative results do not (Fig. 8D). Sensitivity analysis showed the detection limit of LDIA-SD using the pocket-sized device was 100 copies, the same with its results on real-time detection system (Fig. 8E). When quantification fluorescence values of LDIA-SD in detecting different concentration of DNA was analyzed, we found the only drawback LDIA-SD using the pocket-sized device is that there is no correlation between the fluorescence values and the DNA copies. However, the repeatability LDIA-SD using the pocket-sized device was well done (Fig. 8 F).

Fig. 8.

POCT by LDIA-SD using a pocket-sized device. (A) Schematic diagram of the pocket-sized device. (B) Internal structure of the device. (C) Actual photos of the device with dimensioning. (D) Rapid detection of PRV by the LDIA-SD method using a pocket-sized device. Nucleic acid-free extraction of virus samples is described in the materials and methods. 1 μL of the mixture was used as sample template for subsequent LDIA-SD detection. 1, LDIA-SD method in the presence of sample template; 2, LDIA-SD method in the presence of 10⁴ copies of template as a positive control; neg, Negative control. (E) Sensitivity of the LDIA-SD method using a pocket-sized device. In the statistical graph, each error bar represents the standard deviations at 3 replicates (n = 3) of endpoint fluorescence values analyzed by Image J software. The average value of the green (G) channel within a circle which has certain pixel radius is calculated by ImageJ software as endpoint fluorescence values. 1–6, 10⁵, 10⁴, 10³, 10², 10 and 1 copies of templates. neg, Negative control. (F) Repeatability of the LDIA-SD method using a pocket-sized device. In the statistical graph, each error bar represents the standard deviations at 3 replicates (n = 3) of endpoint fluorescence values analyzed by Image J software. The average value of the green (G) channel within a circle which has certain pixel radius is calculated by ImageJ software as endpoint fluorescence values. 1–2, 10³ and 10² copies of templates. neg, Negative control.

3.9. POCT of clinical specimens by the LDIA-SD method

To evaluate the using effect of LDIA-SD method in POCT, we employed it to detect oral fluid specimens of pigs suspected to be infected by PRV. To simulate POCT application, a total of 30 specimens were used for nucleic acid-free extraction and analyzed by LDIA-SD method in the pocket-sized device. Meanwhile, DNA of these clinical samples were extracted by using commercial kits and examined by using the national standard real-time PCR. As was shown in Table 3, POCT of clinical specimens by the LDIA-SD method displayed 18 positive results and 12 negative results, which was consistent with the real-time PCR method. Therefore, the LDIA-SD assay was manifested to be accurate in detecting clinical samples and potential for POCT application. According to previous reports, testing oral fluid specimens of pigs was a mean of improving surveillance efficiency and lowering costs for animal infectious diseases [18], [2]. By using the LDIA-SD method in POCT of the oral fluid specimens, we think it was effective in surveillance of infectious diseases.

Table 3.

The application of the LDIA-SD method in POCT of clinical samples.

NO.	Real-time PCR (Ct values)	LDIA-SD
1	30.53	+
2	30.94	+
3	29.95	+
4	25.54	+
5	34.49	+
6	26.99	+
7	28.36	+
8	29.6	+
9	31.23	+
10	32.85	+
11	25.32	+
12	31.06	+
13	27.3	+
14	31.69	+
15	26.47	+
16	27.68	+
17	28.31	+
18	33.24	+
19	-	-
20	-	-
21	-	-
22	-	-
23	-	-
24	-	-
25	-	-
26	-	-
27	-	-
28	-	-
29	-	-
30	-	-

Note: (+) Positive, (-) Negative.

4. Conclusion

In this study, we developed a novel LDIA-SD method combined with a pocket-sized device ( Fig. 9). Using Bst DNA polymerase, only two pairs of conventional PCR primers targeting the template were needed to initiate the LDIA reaction. Moreover, an extra PCR primer could markedly enhance the reaction. Therefore, all LDIA primers can be easily designed according to standard primer design guidelines. The results of our proof-of-concept assay indicated that outer primers are necessary for the LDIA method, especially for amplification of the products ≥ 140 bp. Although primers with simple structures were used in the LDIA method, it achieved the same sensitivity as LAMP, and there was no cross-reactivity during the 90 min incubation. Another advantage of the LDIA method is that it can be combined with SD probes due to flexibility of primer design. Furthermore, combining LDIA-SD with a nucleic acid-free extraction step and a pocket-sized device afforded a set-up suitable for POCT. The LDIA-SD method is flexible, easy to operate, benefits from facile primer design, and it can amplify complex sequences.

Fig. 9.

Workflow of POCT using the LDIA-SD method.

Funding sources

This work was supported by grants from the open competition program of top ten critical priorities of Agricultural Science and Technology Innovation for the 14th Five-Year Plan of Guangdong Province (2022SDZG02), the Special fund for scientific innovation strategy-construction of high level Academy of Agriculture Science-Distinguished Scholar (R2020PY-JC001), the Project of Collaborative Innovation Center of GDAAS (XTXM202202), the Scientific innovation strategy-construction of high level Academy of Agriculture Science (202110TD), the Special Fund for Scientific Innovation Strategy-Construction of High Level Academy of Agriculture Science (R2017YJ-YB2005 and R2021PY-QY006)，the Start-up Research Project of Maoming Laboratory (2021TDQD002), the Science and Technology plan Program of Guangdong, China (2019B020217002), the National Broiler Industry Technology System Project (CARS-41-G10) and the Walmart Foundation (Project SA1703162 and 61626817) supported by Walmart Food Safety Collaboration Center.

CRediT authorship contribution statement

Hongchao Gou: Conceptualization, Methodology, Data curation. Writing. Qijie Lin: Methodology, Investigation. Haiyan Shen: Methodology, Investigation. Kaiyuan Jia: Methodology, Visualization. Yucen Liang: Investigation. Junhao Peng: Investigation. Chunhong Zhang: Validation. Xiaoyun Qu: Validation. Yanbin Li: Revision. Jianhan Lin: Revision. Jianmin Zhang: Conceptualization, Methodology, Supervision. Ming Liao: Supervision, Project administration. All the authors read and approved the final manuscript.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Biographies

Hongchao Gou received his Ph.D. degree of Veterinary Medicine in South China Agricultural University in 2016; He now works as an associate research fellow at the Institute of Animal Health, Guangdong Academy of Agricultural Sciences. His current research focused on developing POCT methods and exploring metaboliic immune mechanisms of infectious diseases.

Qijie Lin is currently a postgraduate student in the College of Veterinary Medicine in South China Agricultural University. He received his B.S. degree of Veterinary Medicine in Shandong Agricultural University. His research interest focuses on biosensors and detection equipment development. He works in the areas of rapid detection of foodborne pathogen.

Haiyan Shen is majoring in Veterinary Medicine. She got her Ph.D. degree in 2011. Now, she works as an associate research fellow at the Institute of Animal Health, Guangdong Academy of Agricultural Sciences. Her current research focused on molecular epidemiology and immune mechanism of swine infectious disease, and rapid detection technology of animal disease.

Kaiyuan Jia is currently a postgraduate student in the College of Electronic Engineering in South China Agricultural University. He received his B.S. degree of Electronic Information Engineering in China Agricultural University. His research interest focuses on biosensors and detection equipment development.

Yucen Liang is currently a postgraduate student in the College of Veterinary Medicine in South China Agricultural University. She received her B.S. degree of Veterinary Medicine in Guangxi University. Her research interest focuses on rapid detection method of foodborne pathogen such as Campylobacter and Salmonella.

Junhao Peng is currently a postgraduate student in the College of Veterinary Medicine in South China Agricultural University. He received her B.S. degree of Veterinary Medicine in Guangxi University. His research interest focuses on epidemiology and drug resistance of Salmonella.

Chunhong Zhang is majoring in Veterinary Medicine. Now, she works as an associate research fellow at the Institute of Animal Health, Guangdong Academy of Agricultural Sciences. Her current research focused on molecular epidemiology, immune mechanism and rapid detection technology of animal disease.

Xiaoyun Qu is currently an intermediate experimenter at the College of Veterinary Medicine, South China Agricultural University. Her research interests include the prevalence, transmission mode, pathogenic mechanism and rapid detection methods of poultry diseases.

Yanbin Li is a Distinguished Professor of Biological Engineering and Tyson Endowed Chair in Biosensing Engineering at the University of Arkansas. He received his PhD degree in Agricultural Engineering from Penn State University and he is a registered professional engineer in the state of Arkansas. Dr. Li’s expertise is in the areas of biosensing and biomodeling technologies. He develops innovative biosensors for detection of biological and chemical agents and predictive and risk assessment models for microbial hazard analysis. In particular, he is interested in nanomaterials-based biosensors for rapid detection of pathogenic bacteria and viruses in agricultural systems and food supply chains. Dr. Li is a fellow of ASABE, AIMBE and IBE, a senior member of IEEE and NAI, and an active member of AAAS, ASEE, IAFP, IFT, PSA, and SRA.

Jianhan Lin is currently a professor in the School of Information and Electronics. He received his Bachelor's degree in Electronic engineering, Master's degree in Agricultural Engineering and Doctor's degree in Agricultural Engineering from China Agricultural University in 2001, 2004 and 2007, respectively. He has long been engaged in the research of rapid detection technology and equipment for pathogenic microorganisms, and his research direction mainly includes biosensors, microfluidic chips, immunomagnetic separation, etc.

Jianmin Zhang is now an associate professor in College of Veterinary Medicine, South China Agriculture University. He obtained his Ph.D. degree in Veterinary Medicine in 2011 at the South China Agriculture University. He focuses on epidemiology, drug resistance, pathogenic mechanism and diagnostic methods of important zoonotic bacterial diseases.

Ming Liao is now the vice president of Guangdong Academy of Agricultural Sciences. He obtained his Bachelor's degree, Ph.D. degree in Veterinary Medicine from South China Agriculture University in 1992 and 1997 respectively and was a senior visiting scholar at University of Arkansas in US in 2015. He is also the vice president of the Chinese Animal Husbandry and Veterinary Association, the president of the Guangdong Animal Husbandry and Veterinary Association, and the head of the Animal Influenza Expert Group of the National Animal Epidemic Prevention Expert Committee. Topic of his research include the interaction between avian influenza virus and poultry immune system and development of related epidemic prevention products.

Appendix A. Supplementary material

Supplementary material.

.

Supplementary material.

.

Data Availability

Data will be made available on request.