Aptamer-Functionalized Magnetic Nanoparticles for Rapid Isolation of Environmental Escherichia coli

Abstract

Access to safe water remains a vital public health challenge, especially in low- and middle-income countries like Colombia, where untreated sources lead to severe diarrheal diseases in children under five. Escherichia coli (E. coli), a key indicator of fecal contamination, is often detected using culture-based methods that are time-consuming and rely on specialized infrastructure. To overcome these limitations, we developed an aptamer-based isolation system targeting environmental E. coli. Aptamers were obtained using a Cell-SELEX protocol, and after six enrichment rounds, two candidates—APT-EC-1 and its truncated version APT-EC-MUT—were synthesized and attached to carboxyl-functionalized magnetic nanoparticles (MNP-COOH). Both complexes demonstrated a strong binding affinity and high specificity, successfully isolating E. coli from environmental and ATCC reference strains in the laboratory. Sensitivity tests detected E. coli at dilutions up to 1:10,000, showing reliable performance. In early in-field testing with environmental water samples, APT-EC-1 consistently identified E. coli colonies, while APT-EC-MUT struggled with low bacterial levels, illustrating performance differences. These findings demonstrate the promise of aptamer-functionalized MNPs as the basis for quick, affordable, and portable biosensors for water quality testing, especially in resource-scarce areas. Future efforts will add colorimetric or electrochemical readouts to allow real-time, on-site detection of fecal contamination.

1. Introduction

Water is a vital resource for life, essential for human health, daily activities, agriculture, and industry. Securing access to safe drinking water is a global priority, as highlighted in Sustainable Development Goal 6, which calls for universal access to clean water and adequate sanitation [1]. However, ensuring drinking water safety remains a major public health challenge. In 2022, approximately 2.2 billion people lacked access to safely managed water, and 115 million did not even have basic services; additionally, 3.5 billion depended on sanitation systems that did not adequately treat human waste [2]. Poor water management contributes to contamination by heavy metals, toxic chemicals, agricultural runoff, and pathogenic microorganisms, leading to gastrointestinal diseases, chronic illnesses, and certain types of cancer [3,4]. Among these pollutants, microorganisms are particularly worrisome because they can cause acute infections that spread rapidly. Human activities such as excretion, urination, and farming further intensify microbiological pollution, increasing disease risk through both ingestion and direct contact with contaminated water [5,6].

Monitoring microbiological water quality is therefore essential. Total coliforms and Escherichia coli (E. coli) are commonly used as indicators because they are closely linked to fecal contamination, are cost-effective to detect, and indicate the effectiveness of treatment [7]. E. coli, a rod-shaped bacterium from the Enterobacteriaceae family, is prevalent in human waste and serves as a reliable contamination marker [8]. Its pathogenic strains, especially diarrheagenic types, are leading causes of diarrhea, particularly in low- and middle-income countries (LMICs), where limited resources worsen the impact of diarrheal diseases [9,10,11,12]. In Colombia, access to safe drinking water remains limited, affecting both rural and urban communities. Many rely on wells, ponds, and streams, which contribute to high rates of acute diarrheal diseases (ADDs), a leading cause of infant death. In 2024, 152 probable child deaths under five were reported, with 106 confirmed [13]. To address these issues, Colombia employs traditional microbiological detection methods such as membrane filtration and enzyme-based techniques focusing on E. coli and total coliforms [14,15]. While these methods are affordable and reliable, they take 18–48 h for results and cannot detect viable but non-culturable cells [16,17].

Recent technological advances have introduced faster, more sensitive options, such as PCR, genomic sequencing, mass spectrometry, and immunological tests [18,19]. However, these methods have limitations: PCR cannot distinguish between live and dead cells, requires specialized equipment, and trained personnel [20,21]. Immunological techniques also face challenges such as cross-reactivity among similar bacterial strains [19]. In this context, aptamers have emerged as promising tools for detection. These short DNA or RNA sequences form specific three-dimensional structures, allowing high-affinity and high-specificity binding to various molecular targets [22,23,24]. Created through SELEX, aptamers offer several advantages over antibodies, including lower production costs, greater stability, and the ability to bind to intracellular or hard-to-reach targets [25,26,27,28]. Coupling aptamers with magnetic nanoparticles (MNPs) further enhances analytical performance by enabling rapid and efficient separation of analytes, increasing sensitivity, and reducing interference [29,30]. This combination has expanded usage in microbiological detection, medical diagnostics, food safety, and environmental monitoring [31,32,33,34,35,36,37,38,39]. Although aptamer–MNP strategies for detecting E. coli have been explored, this study focuses on developing aptamers targeting a local environmental isolate from regional water sources. Environmental strains often differ genetically and phenotypically from reference strains, necessitating tailored detection tools. Therefore, the main goal of this research is to develop aptamers conjugated with MNPs that can detect an E. coli strain isolated from environmental water samples.

2. Materials and Methods

2.1. Bacterial Isolation

A water sample (~420 mL) was collected from a natural stream located at these coordinates: 6°09′29″ N 75°36′16″ W. The sample was obtained using a Schott glass bottle (500 mL capacity), ensuring a sampling depth of 7 cm and leaving an air chamber to maintain the viability of aerobic microorganisms. The sample was then filtered with a 0.45 µm filter and plated on nutrient media. The different colonies obtained were then isolated to obtain a pure culture.

2.2. Confirmation of Isolated Bacteria

Bacterial isolates were identified using the VITEK 2 COMPACT automated system (bioMérieux Industry, Marcy-l’Étoile, France), using Gram-negative cards as appropriate. Both the equipment and the required reagents were provided by the Institución Universitaria Colegio Mayor de Antioquia (IUCMA).

2.3. Bacterial Culture

For preservation, the bacteria were inoculated into BHI broth and subsequently subcultured on blood agar (Columbia base). For the SELEX strategy, E. coli was used for positive selection and A. baumannii (NS1) was used for counter-selection; both were inoculated into BHI broth and incubated at 37 °C with agitation at 150 rpm overnight.

2.4. Cell-SELEX

We performed the Cell-SELEX strategy using the protocol published by Ospina et al. [40]. An ssDNA library of 78 nucleotides, with a variable region (N) of 40 nucleotides and two conserved regions (5′-TGACACCGTACCTGCTCT-40N-AAGCACGCCAGGGACTAT-3′) at the ends, was used to facilitate detection and amplification. This library was flanked by a forward oligonucleotide complementary to the 5′ conserved region (5′-TGACACCGTACCTGCTCT-3′) and a reverse oligonucleotide complementary to the 3′ conserved region (5′-ATAGTCCCTGGCGTGCTT-3′). The oligonucleotides and the library were synthesized by the company Macrogen, Rockville, MD, USA. We performed a counter-selection round against 100 μL of an overnight culture of the NS1 strain. For this, 1000 ng of the library, diluted in 100 μL of binding SELEX buffer (PBS 1X, 5 mM MgCl₂, and BSA (1 mg/mL), pH 7.4 in milliQ water), was heated to 95 °C for 10 min and cooled at room temperature to promote 3D folding. After 30 min of interaction with NS1 bacteria, the unbound molecules (NBMs) were recovered, and 5 μL was used for PCR amplification according to the protocol described below amplification profile: 95 °C for 2 min, then 95 °C for 30 s, 61 °C for 30 s, 72 °C for 30 s and a final step of 72 °C for 2 min for 20 cycles. For the first positive selection round (PS1), we used 100 μL of E. coli cultured in BHI medium. This was incubated with 10 μL of the PCR product from the negative round in 90 μL of binding SELEX buffer for 30 min at 27 °C and 50 rpm to promote interaction between ssDNA and E. coli surface targets and E. coli surface targets. The unbound molecules were separated by centrifugation at 8000 rpm, and bound ssDNA molecules were eluted with 100 μL of Milli-Q water at 95 °C for 5 min. Then, bound molecules (BMs) were amplified by PCR according to the previously described PCR protocol. Six positive selection (PS) rounds were performed to isolate the molecules capable of recognizing the E. coli bacteria from the previously obtained environmental isolate. The final PCR products were sent for identification by next-generation sequencing (NGS) at Admera Health (South Plainfield, NJ, USA) using the Illumina 2 × 250 platform (San Diego, CA, USA). In the PS rounds, the incubation time and the amount of E. coli culture used were gradually reduced. This was done to increase the detection capacity of the aptamers in the shortest possible time.

2.5. Next-Generation Sequencing (NGS) and Analysis

DNA libraries were prepared using the KAPA Hyper Prep PCR-free kit (Roche, Basel, Switzerland) for amplicon sequencing. Sample quality was assessed by Qubit and capillary electrophoresis, and libraries were quantified by qPCR. The sequencing was performed on an Illumina platform with 2 × 250 bp paired-end reads. The resulting fastq.gz files were processed with FAST QC software v0.12.1 [41] to determine the quality of the selected sequences. Subsequently, we used a custom Python v3.13.2 script to identify: (i) the total number of sequences obtained; (ii) sequences containing Illumina adapters; and (iii) sequences with the 40-nucleotide random region (N40) flanked by the conserved library regions 5′-TGACACCGTACCTGCTCT…AAGCACGCCAGGGACTAT-3′. Sequences corresponding to the N40 region were extracted and analyzed using the AptaSuite software platform v0.9.8 [42,43,44]. We performed analyses including enrichment profiling, aptamer family composition, identification of representative secondary structures, clustering (via AptaCluster), and motif discovery [45,46].

2.6. Modular Aptamer Design

The top 10 motifs identified were aligned in Clustal W [47], and the common sequences were selected to design modular aptamers. These were incorporated into short sequences (<35 nucleotides) using artificial structural elements such as stems (e.g., 5′-GCAGC-[motif]-GCTGC-3′) or linker sequences. Designs were optimized to ensure (i) both motifs remain exposed and accessible for target recognition, (ii) the predicted secondary structure was stable (i.e., negative ΔG), and (iii) shorter sequence length.

Structural predictions were performed using the RNAfold web server [48] configured for DNA, with temperature set at 27 °C (matching Cell-SELEX conditions) and a salt concentration of 1 M. Special attention was given to the structural exposure of motifs within loops or unpaired regions to favor specific target binding.

2.7. Aptamer Synthesis

The most representative aptamer, APT-EC-1, and the most optimally designed modular aptamer, APT-EC-MUT, were synthesized by Macrogen, Rockville, MD, USA.

2.8. Magnetic Nanoparticle (MNP) Synthesis and Functionalization

We conducted the synthesis of magnetic nanoparticles (MNPs) using a co-precipitation method in an alkaline solution using the publicly available protocols from Bomb.bio [49,50,51]. For this, we used FeCl₂ (Sigma^®, Burlington, MA, USA) and FeCl₃ (Sigma^®), ensuring that the solutions were previously degassed and heated to 80–85 °C. The reaction was carried out under normal conditions, yielding magnetite (Fe₃O₄) nanoparticles ranging from 5 to 20 nm in size. Then, we washed and stored them in deionized water. For their functionalization, we applied a radical polymerization of methacrylic acid (MAA) in an aqueous medium. We used sodium dodecyl sulfate (SDS) (Sigma^®) as a stabilizing agent and potassium persulfate (K₂S₂O₈) (Sigma^®) as an initiator. The reaction proceeded at 70 °C for 2 h. Subsequently, we purified the functionalized MNPs by magnetic separation and successive washes. Finally, we stored the carboxyl-coated product in deionized water at room temperature [51,52]. The characterization of the magnetic nanoparticles obtained is found in Supplementary Materials S2.

2.9. Pull-Down

100 μL of carboxyl-functionalized magnetic nanoparticles (MNP-COOH) (10 mg/mL) was aliquoted into a 1.5 mL sterile Eppendorf tube. The tube was then placed on a magnetic rack for 1 min to allow complete separation of the MNPs from the supernatant, which was carefully removed. Next, three washes were performed using the aptamer-nanoparticle binding buffer (10 mM Tris, 100 mM NaCl, 10 mM CaCl₂, pH 7.4). Subsequently, 100 µL of aptamers (20 ng/μL) was added; the aptamers had been previously heated to 90 °C for 5 min and then cooled at room temperature for 10 min. The mixture was incubated at 4 °C for 15 min. Then, aptamers and MNPs were incubated for 5 min at room temperature with constant agitation at 50 rpm to facilitate the formation of the MNP-APT complexes. Following incubation, the coupled MNPs were magnetically separated, and the supernatant was discarded to remove any unbound aptamer. Subsequently, 100 µL of the isolated E. coli bacterial sample (pure and serial dilutions 1:10, 1:100, 1:1000, and 1:10,000) and ATCC reference strains as controls, including E. coli (positive control) and E. faecalis, S. aureus, K. pneumoniae, and S. Typhimurium (negative controls), previously cultured in BHI medium, were added according to the experimental design. The mixtures were incubated for 5 min at room temperature with constant agitation at 50 rpm. After incubation, magnetic separation was performed to remove the unbound supernatant, followed by three washes (500 μL) with injectable water to eliminate non-specifically captured cells. Finally, the MNP–Aptamer complex retaining the bound bacteria was resuspended in 100 µL of sterile water; 10 µL was then plated on the selective medium for Gram-negative bacteria, MacConkey agar, and incubated for 24 h at 37 °C. The number of colonies on the agar was counted to calculate the total number of viable bacteria in the samples, and the specificity of the aptamer against E. coli was verified.

In addition, we performed an in situ assay with a new environmental water sample to verify the aptamers’ ability to detect E. coli isolates. For this assay, the same protocol described above was followed, but at the sampling site under sterile conditions. The isolated bacteria were subsequently cultured in the laboratory to verify their quantity and specificity on MacConkey agar.

3. Results

3.1. Bacterial Isolation

The initial microorganisms isolated from blood agar showed whitish, opaque, and shiny colonies with convex elevations measuring 2–4 mm in the case of strain 1. Strain 2 showed shiny, colorless, nonhemolytic, smooth, and small colonies, 1–2 mm in size. In Gram staining, both strains were Gram-negative with a bacillary form for strain 1 and a coccobacillary form for strain 2 (Figure 1).

3.2. Confirmation of Isolated Bacteria

Strain 1 was identified as E. coli with a high level of confidence, which was the target microorganism for this study. Additionally, strain 2 was identified as Acinetobacter baumannii with a high level of confidence, which was used as a target for counter-selection in Cell-SELEX (

Table 1. Results of the VITEK examination report.

Identified Organism	Probability	Confidence Level
Escherichia coli	96%	Excellent identification
Acinetobacter baumannii complex	99%	Excellent identification

). The original reports obtained by VITEK equipment can be found in Supplementary Materials S2.

3.3. Cell-SELEX

The Cell-SELEX strategy that allowed for the selection of ssDNA sequences with high affinity and specificity to E. coli is shown below in Figure 2.

DNA quantification in each round initially showed an increase in the recovery of bound molecules, followed by a progressive decrease in later rounds, indicating an improvement in the specificity of the selected aptamers (Figure 3A). As the process advanced, the reduction in incubation time and the amount of E. coli used facilitated the elimination of non-specific sequences, allowing for the selection of those with higher affinity (Figure 3B).

Electrophoresis analysis in agarose gel at 2% confirmed the amplification of PCR products in each round, as defined bands were observed in the positive selection rounds, while no amplification was detected in the negative controls, thereby ruling out contamination or non-specific amplification (Figure 3C).

In

Table 2. Summary of Cell-SELEX rounds’ design.

Round	Bacteria Specie	Bacteria Culture Volume (µL)	Incubation Time (min)	Type of Round
1	A. baumannii	100	30	Negative
2	E. coli	100	30	Positive
3	E. coli	10	20	Positive
4	E. coli	1	15	Positive
5	E. coli	1	10	Positive
6	E. coli	1	5	Positive
7	E. coli	1	5	Positive

, we describe the detailed experimental design of the Cell-SELEX process employed in this study for the selection of specific aptamers against isolated E. coli. It presents the selection rounds, bacterial culture volumes used at each stage, the target bacterial species, and the incubation times applied. Throughout the process, these conditions were progressively adjusted to enhance selective pressure and promote the enrichment of sequences with greater specificity. The reduction in culture volume and incubation time facilitated the elimination of low-affinity sequences, thereby optimizing the efficiency of aptamer selection for E. coli.

3.4. Next-Generation Sequencing (NGS) and Analysis

A total of 508,256 sequences were obtained; 438,980 contained Illumina adapters, and 191,707 contained N40 regions within the conserved regions of the initial library. Approximately 12% of sequences showed enrichment. The percentage of base distribution was as follows: A: 17.6%, C: 22.3%, G: 31.1%, and T: 28.9%, compared to the initial library’s A: 26%, C: 26.7%, G: 24.5%, and T: 22.6%. The aptamer candidates’ ID, sequences, count, and frequencies are shown in Figure 4A. The 2D structure of the most representative aptamer is shown in Figure 4B; the alignment of the 10 most frequent motifs is observed in Figure 4C; and the 2D structure of the modular aptamer designed based on the motifs found is observed in Figure 4D.

3.5. Pull-Down

The summary of the pull-down strategy carried out is described in Figure 5. The specificity of both aptamers was evaluated by monitoring bacterial growth in the presence of Escherichia coli. The bacterial strains and isolates used for this laboratory-scale validation are detailed in Figure 6A. The results obtained using the EC aptamer are shown in the upper and lower panels of Figure 6B, while those corresponding to the modular aptamer EC_MUT are shown in the upper and lower panels of Figure 6C.

In both cases, bacterial growth was observed with the pure E. coli environmental isolate and across all tested dilutions (1:10, 1:100, 1:1000, and 1:10,000), as well as with the pure E. coli ATCC reference strain. No growth was detected when magnetic nanoparticles without aptamers were used, nor with any of the other tested ATCC Enterobacteriaceae strains. These results demonstrate that both aptamers selectively bind to E. coli and do not exhibit non-specific interactions with non-target bacterial species under the experimental conditions applied.

4. Discussion

The detection of microorganisms has long been essential in fields such as public health, biotechnology, and environmental microbiology. In recent years, aptamers have emerged as a promising alternative for microbial detection due to their high specificity and adaptability. This study employed the Cell-SELEX technique to isolate aptamers targeting Escherichia coli, a key fecal contamination indicator. By using an environmental isolate, the selection strategy was designed to closely reflect real-world application scenarios.

Selective pressure progressively increased throughout the Cell-SELEX rounds to enhance specificity, which enabled the identification of high-affinity candidates in fewer cycles than traditionally reported [53,54]. A single counter-selection round was incorporated early to reduce non-specific binding, balancing sensitivity and specificity [55,56].

Next-generation sequencing (NGS) of the sixth positive round generated a complex dataset, which was refined using a custom Python pipeline to remove adapters, conserved regions, and short sequences. Structural predictions via RNAfold revealed common aptamer motifs, such as stem-loops, which facilitate specific target binding [54,57,58]. These conformations are critical for molecular recognition and are consistent with previous structural models describing aptamer–target interaction as a “molecular embrace” [57,59,60,61,62,63,64,65].

Candidate selection considered thermodynamic stability (ΔG), with top sequences showing values from 0.0 to −14.7 kcal/mol, aligning with commonly accepted functional ranges [66,67,68,69]. A representative aptamer and a modular version (under 40 nt) were synthesized, the latter designed to retain key binding motifs while minimizing synthesis costs, in line with recent trends toward truncated aptamer development [26,70,71,72,73,74].

Although chemical modifications were not implemented, the potential for enhanced stability through backbone alterations, end-capping, or locked nucleic acids remains relevant for future integration [28,75,76,77,78,79]. Aptamers were conjugated onto magnetic nanoparticles (MNPs) functionalized with carboxyl groups (-COOH), chosen for their favorable handling, stability, and cost-effectiveness (Table S1). MNPs allowed for rapid magnetic separation, eliminating the need for biotinylation while maintaining performance [80,81].

The carboxyl-functionalized nanoparticles were obtained following the BOMB.bio protocol. The surface of these MNPs allows aptamer binding through electrostatic interactions, which were optimized by carefully controlling the pH, buffer composition, and short incubation times. These parameters were deliberately designed during the SELEX strategy development to ensure rapid and stable formation of aptamer–MNP complexes, enabling efficient pull-down of the E. coli target.

In this study, the incubation time between the aptamer-functionalized magnetic nanoparticles (MNPs) and E. coli was set to 5 min. This parameter was also optimized during the SELEX process, where interaction times were progressively reduced from 30 min in the initial rounds to 5 min in rounds six and seven, favoring the selection of aptamers capable of rapid binding. While the literature on aptamer–bacteria binding kinetics remains limited, several studies have demonstrated that aptamer–target interactions can occur within short timescales. For instance, fluorescence-based kinetic studies have reported equilibrium binding within 5–10 min for thrombin-specific aptamers [82]. Similarly, recent work on aptamer–protein recognition confirmed association times under 5 min when using optimized conditions [83]. These findings support the feasibility of short incubation times in aptamer-based systems, which is particularly advantageous for the development of rapid, field-deployable detection platforms.

In vitro, assays demonstrated that both the full-length EC aptamer and the modular variant were able to successfully recognize the environmental E. coli isolate as well as the E. coli ATCC reference strain, detecting both in their pure form and across serial dilutions down to 1:10,000. For the initial in situ validation, a second water sample was collected. Unlike the first (turbid) sample, this one appeared clear and nearly transparent, likely due to environmental factors at the time of collection. Under these conditions, the EC aptamer successfully detected three E. coli colonies, while the modular aptamer failed to detect any. This discrepancy may be attributed to a low bacterial load in the sample and the still-uncertain mechanisms governing aptamer–bacteria interactions. These findings highlight the need for additional validation studies under diverse environmental conditions to better understand aptamer performance and binding dynamics in real-world applications.

Overall, this detection strategy represents a promising step toward aptamer-based biosensing for E. coli. Future work should focus on integrating colorimetric or electrochemical detection systems to enable rapid, accessible readouts. Approaches such as sandwich-type assays with gold nanoparticles [36] or phenol red-based indicators leveraging urease activity [84,85] offer practical pathways toward real-world deployment.

5. Conclusions

The results of this study highlight the potential of aptamers as highly specific molecular recognition elements for the detection of Escherichia coli, supporting their application in water quality monitoring. The aptamer–magnetic nanoparticle (MNP) system developed here represents a significant step toward the creation of a rapid, cost-effective, and portable biosensor, particularly suited for use in low-resource settings. Although the functionality of the aptamer–MNP complex was validated in vitro through culture-based assays, further optimization is required to enable direct, faster detection without the need for culturing. It should be noted that in this study, we did not assess the ability of the selected aptamers to distinguish between viable and non-viable Escherichia coli cells, since culture-based confirmation was used. Future work will include additional experiments using UV-inactivated or otherwise treated bacteria to evaluate aptamer performance with live versus dead cells, which will be essential for the development of a fully reliable detection platform. Preliminary in situ testing also revealed performance differences between the full-length and modular aptamers, emphasizing the need for robust validation under diverse environmental conditions. Future efforts should focus on integrating efficient signal transduction mechanisms, such as colorimetric or electrochemical outputs, to advance toward a fully field-deployable diagnostic platform for real-time microbial monitoring.