2.1. Principles of the Direct PCR- LFA system
To detect ACE I/D polymorphisms, we established the Direct PCR-LFA system. The principles of the Direct PCR-LFA system is schematically illustrated in Figure. 1.
GMNPs were synthesized based on the methods described previously [22, 23]. The GMNPs were functionalized with cetyltrimethylammonium bromide (CTAB) surfactant, followed by modification of the polyacrylic acid (PAA) and then conjugation of anti-digoxin antibody through 1-[3-(dimethylamino)propyl]-3-ethylcarbodiimide hydrochloride chemistry based on a previous method [24, 25](Fig. 1A). The LFA device consisted of a strip composed of five overlapping pads, i.e., sample pad, conjugate pad, nitrocellulose membrane, absorbent pad, and plastic cushion (Fig. 1B). In brief, goat anti-mouse IgG and streptavidin were pre-immobilized in a control line (C line) and a test line (T line) respectively on a porous nitrocellulose membrane using the BioJet HM3010 dispenser (BioDot Inc., California, U.S.A.). Then, the probe solution containing the poly-acrylic acid (PAA) modified gold magnetic nanoparticles (PGMNPs) conjugated with anti-digoxin antibody was dispensed onto the conjugate pad of the LFA strips. The strips were dried and stored in a sealed aluminum foil bag at room temperature until required.
Fresh whole blood samples were treated with NaOH solution, followed by PCR amplification using two tubes with the same prepared blood template (Fig. 1C). After PCR amplification, the PCR products in the two tubes were added onto the sample pad of the LFA strip and the results read visually within 5 min (Fig. 1D). For insertion homozygous samples (II), a distinct red band was observed on the T line of the strip from the PCR products of the M tube but not from the WT tube. For deletion homozygous samples (DD), only a red band was observed on the strip from PCR products of the WT tube but not the M tube. If red bands with similar intensities are observed on the T lines of both strips, this indicates an insertion/ deletion heterozygous sample (I/D).
Our Direct PCR-GoldMag LFA was faster compared to existing methods. The pretreatment of the samples with NaOH is crucial for this method and makes the whole testing procedure convenient and rapid. This is because, compared to other detection methods, the DNA purification step is eliminated in our assay with the help of NaOH treatment. This shortens the processing time from 1 to 2 hr to a mere 5 min. Second, by using NaOH-treated blood samples for PCR, the problem of cross-contamination, which exists in traditional blood DNA purification processes is eliminated. Furthermore, various PCR inhibitors present in the samples such as hemoglobin, IgG, lactoferrin, and proteases are inactivated by NaOH treatment[26, 27]. When the direct PCR product is loaded onto the sample pad of the LFA strip, certain impurities such as hemoglobin in the blood are filtered by the sample pad and conjugate pad, hence the chromatography is not affected.
By introducing mismatches at the penultimate or antepenultimate at the 3' end of the primers, primers specificity was enhanced. We designed several primers to determine the optimal primers to use. After ARMS-PCR, the LFA device was used to detect the PCR products. With the help of gold magnetic nanoparticles and LFA, accurate and rapid genotyping results could be generated by visual inspection of colors on the T and C lines. It takes only 5 min to obtain the results without the need for expensive or high-end instruments. We also demonstrated that our method was more convenient and superior to the classical agarose gel electrophoresis method. Hence, this method could be used in medical and hospital laboratories with limited resources for purchasing specialized equipment.
2.2. Performance of the Direct PCR- LFA system
To determine the optimal assay conditions, optimization of whole blood direct amplification was performed under different PCR cycling conditions, i.e., annealing temperature, concentration of the primers, etc were optimized using whole blood samples with the three different ACE genotypes. The magnetic signal at the T line showed the best amplification efficiency and specificity when the cycle number was 31 (Figure S1A) and the annealing temperature was 60 °C (Figure S1B). In addition, the best primer concentration was found to be 2.5 µM (Figure S1C).
The specificity of the ACE genotyping was an important consideration. For specificity analysis, three known ACE genotypes (II, I/D, DD) were analyzed using the Direct PCR- LFA system (Fig. 2A). We also performed a crossover experiment to validate the primers. In addition, ACE genotyping results detected using the lateral flow assay (Fig. 2B) was compared with the classical agarose gel electrophoresis method (Fig. 2C). ACE gene I/D polymorphism is caused by either an insertion or deletion of the 287 bp Alu repeat. A gene fragment containing a repeat sequence cannot be accurately verified using the Sanger method. A homozygous deletion could be accurately measured after several rounds of sequencing, however, homozygous insertions are more difficult to accurately sequence (Fig. 2D-E).
To enhance primer specificity, we designed multiple primers to determine the optimal primers pairs to use in our assay (Table 1). In addition, we introduced mismatches in the penultimate or antepenultimate of the 3' end of the primer for PCR. We observed that these modifications had an impact on the specificity of our method, as some allele-specific primers completely lost specificity upon removal of the mismatches [28]. This is important when analyzing clinical samples that require high specificity and accuracy. We then performed PCR and agarose gel electrophoresis using these primers and selected the most optimal primes to use in our assay. The deletion-specific primer sequence was 5'-AACCACATAAAAGTGACTGTATCGG-3', and had a mismatch at the third of the 3' end, while the insertion-specific primer sequence was 5'-TCGAGACCATCCCGGCTAAAAC-3'. Other primers that were synthesized were excluded due to poor specificity. The list of common primers is shown in Table 5. The common forward primer sequence was 5'-AAGGAGAGGAGAGAGACTCAAGCAC-3'.
2.3. Performance of the PCR-GoldMag LFA
To evaluate the performance of the PCR-GoldMag LFA system, the sensitivity of LFA was evaluated using various amounts of whole blood samples with known ACE genotypes. Gradient dilutions of whole blood with physiological saline at 1:1, 1:2, 1:4, 1:5, 1:10, 1:15, 1:20, 1:30, 1:40, 1:60, 1:120 ratio were evaluated. Whole blood samples are compared with purified nucleic acid samples (Fig. 3). The LFA typing results were analyzed to determine the sensitivity of the whole blood direct PCR method. Simultaneously, we performed blood routine tests obtained from the People's Hospital of Shaanxi Province (Table S1). The amount of nucleic acids in a white blood cell is approximately 5.6 × 10− 9µg[29]. The results demonstrated that the minimum detection sensitivity was 0.75 ng.
Using our optimized conditions, PCR-GoldMag LFAs showed high specificity with no false-positive results and had a higher sensitivity. The detection limit of PCR-LFA could reach 0.75 ng of DNA, which was comparable to a PCR-DNA microassay commercial kit and had obvious advantages. As shown in Fig. 3, whole blood samples were more sensitive compared to purified DNA. At a dilution ratio of 1:60, bands could still be observed using whole blood, while from purified DNA, bands were fainter at the same dilution factor. For homozygous deletion samples, no bands appeared at a ratio of 1:15. This may be due to the loss of white blood cells during DNA extraction, while whole blood eliminated the DNA extraction step, and hence had no loss of white blood cells. The sensitivity was higher using whole blood samples compared to purified DNA. The capacity of antibody conjugation to nanoparticles is critical for the success of our assay. In general, the conjugating capacity of nanoparticles to antibodies at best is only about 50µgmg− 1. Because of the novel GoldMag nanoparticle structure (nanoflowers)[22], our antibody conjugation reached greater than 100µgmg− 1. This ensured the sensitivity and stability of our established PCR-lateral flow assay[22, 25].
2.4. Clinical applications
To evaluate the reliability of our optimized PCR-GoldMag LFA method, the accuracy was further validated using an additional 300 clinical samples obtained from the Shaanxi Provincial People's Hospital (Xi’an, China), with informed consent waived. Each sample was tested using our whole blood Direct PCR-GoldMag LFAs and agarose gel electrophoresis was used to compare our results. As shown in Table 2, the genotyping results were 100% consistent with the results obtained from agarose gel electrophoresis. In addition, we performed a Hardy-Weinberg equilibrium test on selected samples (Table 2) and analyzed the genotype and allele frequencies of the 300 cases (Fig. 4). The results were consistent with Hardy Weinberg's law of equilibrium. It demonstrated that the selected groups had good group representation. The genotype frequencies of ACE II, ID, DD types were 42% (127 cases), 44% (132 cases), and 14% (41 cases) (Fig. 4A). The allele frequency for the I allele was 64% and the D allele was 36% (Fig. 4B). Representative genotyping results of the two methods are showed in Figure S2. In addition, we evaluated blood samples from cases with high bilirubin levels and low and high white blood cell counts (Figure S3). Our results demonstrated that the established method had good accuracy and reliability. We believe our method has wide application value in clinical practice.
We established a sensitive, low cost, and easy-to-use large fragment insertion/deletion polymorphism detection platform using whole blood Direct PCR-LFA. This enabled us to accurately detect insertion/deletion polymorphisms for specific genetic diseases. With the help of gold magnetic nanoparticles and LFA, genotyping results could be rapidly visualized based on colored T and C lines.
Compared to conventional detection methods for I/D polymorphism, our assay had numerous significant advantages. These include (a) Using the PCR-GoldMag LFA system, it takes only 5 min to obtain results without the need for expensive or high-end instruments. (b) After sample treatment with NaOH, the whole testing procedure was rapid and convenient. (c) Previous methods were complex and performed on expensive and sophisticated instruments that may not be available in many laboratories, however, the PCR-GoldMag LFA method is easy-to-operate and affordable for on-site genotyping with high efficiency. Hence, this method could be run in laboratories that are not fully equipped with sophisticated instruments.
2.5 Case-control study
Previous studies have reported that insertion/deletion (I/D) polymorphisms in the ACE gene were associated with cardiovascular and cerebrovascular diseases. However, the association between ACE gene I/D and cardiovascular and cerebrovascular diseases is controversial. Some studies have shown that ACE I/D polymorphisms are associated with coronary artery disease (CAD) and cerebral ischemic stroke (CIS)[12, 14], however, other studies have found no association[13, 15]. We performed a case-control study and a meta-analysis to evaluate the association between ACE I/D polymorphisms and coronary atherosclerotic heart disease, as well as stroke.
We analyzed a total of 633 subjects (199 CHD patients, 207 CIS patients, and 227 control group) for the case-control study. These samples satisfied the Hardy–Weinberg Law (CHD, c2 = 0.68, p = 0.41 > 0.05; CIS, c2 = 1.58, p = 0.21 > 0.05), which signified a reliable representative group. Using the chi-square test, no statistical differences between these two groups were observed (Table 3–4). The ACE genotype frequencies are shown in Fig. 5. Based on statistical analysis, we found no association between ACE I/D polymorphisms and coronary heart disease or stroke. However, after grouping based on age, we observed a significant difference between the genotype and age of the patients in the CHD group (p = 0.02 < 0.05), with no significant differences in the stroke group (p = 0.07 > 0.05) (Table S-2).
Association studies are influenced by selection bias, population stratification, confounding factors, and clinical criteria used to define patient groups. Future genetic association studies should include larger patient cohorts and strict study designs to determine potential associations between genetic susceptibility and cardiovascular and cerebrovascular diseases.