A Combined Aptamer Pulldown‐DNAzyme Cleavage Assay for Intact Methicillin Resistant Staphylococcus aureus Cells

Abstract

Accurate and convenient detection of methicillin resistant Staphylococcus aureus (MRSA) plays a vital role in determining appropriate antibiotic interventions. Herein, we report the first example combining a DNA aptamer and an RNA‐cleaving DNAzyme (RCD) that bind different protein markers for selective preconcentration and detection of MRSA. An aptamer for the penicillin binding protein 2a (PBP2a) was generated by in vitro selection and was coupled to agarose beads to allow rapid pull down of intact MRSA cells from complex samples. A previously reported RCD for Staphylococcus aureus (SA) was then used to detect a second proteinaceous marker within the lysed cells based on a protein‐activated cleavage reaction that was linked to both fluorescence and lateral flow assays, allowing for selective detection of MRSA over methicillin sensitive SA. An optimized LFD assay could detect ∼10³ cfu mL⁻¹ of MRSA in either nasal mucus or serum with a total assay time of 1 h using minimal sample processing.

Rapid fluorescence and lateral flow assays were developed for methicillin resistant Staphylococcus aureus (SA) by combining a DNA aptamer for selective isolation and an RNA‐cleaving DNAzyme for detection of MRSA.

Introduction

Bacterial infections represent both a major health concern and a significant socioeconomical burden.^{[ 1} , ² , ^{3 ]} Staphylococcus aureus (SA) is a pathogenic bacterium that is responsible for infections of the skin, respiratory tract and blood (SA bacteremia),^{[ 4} , ⁵ , ⁶ , ^{7 ]} with up to 50% of the population having SA present on the skin or in the respiratory tract,^{[ 8 ]} resulting in high morbidity and mortality.^{[ 9} , ¹⁰ , ^{11 ]} Exposure to SA can occur through surgery,^{[ 4} , ⁵ , ⁶ , ^{7 ]} consumption of SA contaminated food or beverages,^{[ 12} , ¹³ , ^{14 ]} or through inhalation of SA carrying aerosol particles,^{[ 15} , ^{16 ]} and hence SA can often be detected in blood, mucus or sputum.

SA can exist in both methicillin sensitive (MSSA) and methicillin resistant (MRSA) forms, with the latter having a much higher mortality rate and fewer treatment options.^{[ 17} , ¹⁸ , ^{19 ]} Detection of SA has typically been done using plate culture.^{[ 20} , ^{21 ]} However, this method requires up to 72 h to provide results and requires additional assays to assess methicillin resistance. MRSA strains overcome antibiotic sensitivity by producing an altered penicillin‐binding protein 2a (PBP2a),^{[ 17} , ^{18 ]} which has decreased affinity for most semisynthetic penicillins. The protein is encoded by an acquired gene, mecA,^{[ 17} , ^{18 ]} and thus specific detection of MRSA can be done either through amplification of the mecA gene using the polymerase chain reaction (PCR),^{[ 22} , ^{23 ]} or by using various post‐culture immunoassays, including lateral flow immunoassays, to detect the PBP2a protein.^{[ 14} , ²⁴ , ²⁵ , ^{26 ]} However, both methods require an initial culture step as well as extensive sample processing, making them unsuitable for rapid point‐of‐care testing.

In recent years, several groups have evaluated the use of DNA aptamers with affinity for PBP2a to develop rapid assays for MRSA.^{[ 27} , ²⁸ , ²⁹ , ^{30 ]} Recent examples include: detection of PBP2a in nasal swabs using a fluorescence aptamer‐based assay on lysed MRSA cells that were first isolated using immunomagnetic beads targetting Protein A;^{[ 31 ]} use of a PBP2a aptamer‐modified cotton swab for collection of MRSA cells on surfaces, followed by binding of aptamer‐bound colored nanoparticles for detection of MRSA;^{[ 32 ]} and an interferometric biosensor using PBP2a aptamers to capture intact MRSA.^{[ 33 ]} However, such assays either showed poor detection limits^{[ 31 ]} required expensive equipment,^{[ 33 ]} or were not tested using spiked biological media or patient samples,^{[ 32 ]} reducing their potential for point‐of‐care testing.

RNA‐cleaving DNAzymes (RCDs) are well established for biosensor development^{[ 34 ]} and are becoming more prominent in clinical diagnostics.^{[ 35 ]} Our group recently reported on the selection and characterization of a new RCD (denoted as RFD‐SA6T1), which we have confirmed is activated by a thermostable protein with a molecular weight over 100 kDa that is not PBP2a or a ribonuclease (see Figure S1).^{[ 36 ]} The RCD was adapted for both rapid fluorescence and lateral flow assays, however it was not able to distinguish between MSSA and MRSA. To overcome this issue, we hypothesized that the use of agarose beads carrying a PBP2a binding aptamer could allow for selective enrichment of intact MRSA cells from complex samples via sedimentation,^{[ 31} , ^{37 ]} followed by detection of the recovered bacteria using the RFD‐SA6T1‐based assay. Herein, we report on the development of coupled aptamer pulldown‐RCD fluorescence and LFD assays for MRSA cells, which represents the first time that aptamers and RCDs for separate protein targets within the same bacterium have been combined in a single assay system to improve both the selectivity and sensitivity of bacterial detection.

Results and Discussion

As a starting point, we evaluated the use of commercially available PBP2a binding aptamers for selective enrichment of intact MRSA cells from complex samples. However, these existing aptamers did not bind to intact MRSA cells and thus could not isolate these cells for subsequent RCD assays (see Figure S2 in the supporting information). We therefore utilized SELEX (Systematic Evolution of Ligands by EXponential enrichment)^{[ 38} , ^{39 ]} for in vitro selection of new PBP2a‐binding DNA aptamers with the goal of identifying aptamers that could bind to the native PBP2a protein present in the outer membrane of intact MRSA cells. For this work we utilized a soluble recombinant form of PBP2a (rPBP2a) as the target for aptamer selection. The rationale for selecting rPBP2a was that the target should produce multiple aptamers that could bind to different epitopes of the target protein, and that at least one aptamer from this collection would bind to an epitope that was present at the surface of the cell membrane. This method avoids the potential for generating aptamers for unknown surface proteins, which can occur for whole cell selection methods.^{[ 40} , ^{41 ]}

Figure 1a shows the DNA library (DL) and primers used for SELEX (all sequences used in this work are listed in Table S1), while Figure 1b shows the selection protocol. Positive selection (PS) was carried out with rPBP2a immobilized on magnetic nickel beads (MNB) through a histidine tag. The DL contains 50 random nucleotides flanked by a fixed DNA sequence at each of the 3′ and 5′ termini to allow PCR amplification. A small fraction of the DL was amplified by PCR using fluorescein labelled FP to generate a label on the DL to follow the selection progress (see Electronic Supporting Information for details), and this was combined with 1 nmol of unmodified DL. The DL was incubated with the rPBP2a‐MNB suspension to allow aptamer binding. The beads were magnetically isolated and washed to remove unbound DNA sequences, followed by recovery of the bound DNA molecules using imidazole. These sequences were amplified by PCR using fluorescein labelled forward primer to label the DL for subsequent rounds and applied to the next round of selection. Counter selection (CS) was carried out in each round using His‐tag modified nickel beads to remove the sequences that had affinity for the His‐tag or bead only. The unbound sequences were collected and applied in the next PS round. To ensure selectivity, in every PS round, the beads were washed with a buffer containing human serum albumin (HSA) and in every second round they were washed with a mixture of the cell lysates of methicillin sensitive staphylococcus aureus (MSSA), E. coli, and B. subtilis.

Figure 1.

a) DNA library and related oligonucleotide sequences used in in vitro selection. b) Schematic illustration of in vitro selection. c) Images of the selection monitored through the fluorescence signal on the beads in each round of positive selection. d) Fluorescein labeled truncated shorter aptamer sequence. e) Binding sensitivity test using dot blot assay in buffer with pure PBP2a. f) Specificity test in buffer using dot blot assay.

The progress of the selection was monitored after each PS round based on both the fluorescence signal of the beads arising from the bound aptamers and that of the supernatant obtained after displacement of the bound aptamers. A significant amount of fluorescence was obtained in each case in round 8 (Figure 1c). The population of round 8 was deep sequenced (top 10 sequences are shown in Figure S3A) and the top 6 sequences, which had fractional abundance values >0.1% of the total DNA sequences, were synthesized with fluorescent tags on the 5′ terminus and evaluated for binding with bead immobilized rPBP2a. The results showed that the sequences bound to rPBP2a beads to different degrees, as reflected by the varying fluorescence signals for each aptamer (Figure S3B). None of the sequences were found to bind to the control His‐tag beads, suggesting that the sequences bound selectively to rPBP2a (Figure S3B).

The fluorescence binding assay showed that Apt4 bound most strongly to the rPBP2a‐modified beads, followed by Apt3 and Apt1 (Figure S3B). To remove any artifacts resulting from non‐specific binding to beads, all sequences were also tested using an electrophoretic mobility shift assay (EMSA). In this case, Apt2 and Apt5 showed the highest binding, followed by Apt1 and Apt 3 (Figure S4). The lack of binding by Apt4 suggests that this sequence interacted non‐specifically with the beads.

Truncated versions of the top aptamers were produced (Figure S5A) and tested for binding to rPBP2a. Interestingly, only Apt1 retained binding activity after truncation (Figure S5B), while truncated versions of Apt2–Apt5 showed a complete loss of binding activity. The truncated Apt1 was also determined to bind to rPBP2a to a greater degree than the full length Apt1 (compare Figures S5C and S5D). The full length and truncated versions of Apt1 are shown in Figure S5E,F. Based on these results we selected the truncated Apt1 (denoted as PBP2a‐Apt1T1, Figure 1d) for further studies.

Figure 1e shows a binding isotherm for PBP2a‐AptT1 using a fluorescence‐based dot‐blot assay utilizing rPBP2a immobilized on the membrane. The dissociation constant (K _d) was determined to be 58 ± 11 nM, which is consistent with previously reported PBP2a‐binding aptamers, (K_d values ranged from 15 to 50 nM).^{[ 28 ]} Figure 1f shows the selectivity of PBP2a‐AptT1 against a range of other proteins that would be expected to be present in nasal mucus or blood, which are two relevant biological matrixes for MRSA testing.^{[ 18} , ²⁴ , ^{31 ]} These proteins included mucin, myeloperoxidase (MPO) and eosinophil peroxidase (EPX) as representative mucus proteins, and human serum albumin (HSA), immunoglobulin G (IgG) and thrombin as representative blood proteins. In each case, dot‐blot binding assays demonstrated little to no background binding to these interfering proteins, though there was slight binding from EPX, which is a highly cationic protein with a pI of 11.4,^{[ 42 ]} and thus may have nonspecific ionic interactions with the anionic DNA aptamer.

PBP‐2a is a membrane associated transpeptidase, and thus resides in the outer membrane of MRSA.^{[ 18} , ²⁴ , ^{43 ]} As such, it should be possible for the PBP2a‐Apt1T1 aptamer to bind to intact MRSA cells. To evaluate this possibility, we undertook studies to evaluate binding of fluorescently labelled versions of Apt1–Apt6, as well as PBP2a‐Apt1T1, to MRSA cells. In this case, 100 nM of each aptamer was incubated with 10⁸ cfu mL⁻¹ of MRSA for 1 h, followed by centrifugation, removal of the supernatant, and fluorescence imaging of the reaction tubes. As shown in Figure 2a, only PBP2a–Apt1T1 was able to bind to the intact MRSA cells, while none of the full‐length aptamers, including Apt1, were able to bind to the native membrane‐bound PBP‐2a in MRSA. This suggests that the full‐length aptamers likely bound to PBP2a epitopes in the recombinant protein that were not accessible on the cell surface or had insufficient affinity to remain bound after washing. Importantly, the ability to bind to intact MRSA cells should allow the aptamer to be used for isolation of MRSA from complex matrixes through aptamer‐directed affinity pull down assays, as described below.

Figure 2.

Whole cell binding test. a) MRSA33591 was used to test binding with all 6 aptamers. Only PBP2a‐Apt1T1 bound strongly and efficiently. b) Selectivity test with PBP2a‐Apt1T1 against various Stapholococcus species. SL = S. lentus; SC = S. chromogenes; SS = S. saprophyticus; SP = S. pasteuri; SE = S. epidermis. c) Binding test with other aptamers and non‐aptamer sequences. S. pasteuri was found to bind with any DNA sequence. d) LoD for aptamer binding to intact MRSA cells.

We next evaluated the selectivity of PBP2a‐Apt1T1 against a range of non‐target bacteria that may be present in blood or nasal mucus. As shown in Figure 2b, the aptamer showed excellent selectivity for MRSA over MSSA, providing a potential method for detection of MRSA without interference from MSSA. However, both S. pasteuri (SP) and to a lesser degree S. epidermis (SE) showed substantial background binding, which is not unexpected as both these staphylococcal species are able to acquire the MecA gene and thus may express PBP2a.^{[ 44 ]}

To determine the origin of the PBP2a‐Apt1T1 binding to non‐target bacteria, a series of other DNA sequences were evaluated for binding to S. pasteuri and S. epidermis, including PBP2a‐Apt4, a second aptamer with selectivity for thyroid specific hormone, and a scrambled non‐aptamer DNA sequence (sequences are shown in Table S1). Each of these sequences showed strong binding to S. pasteuri as well as weak binding to S. epidermis (Figure 2c). Hence, the background binding is likely not generated by endogenous PBP2a but rather occurs due to membrane proteins or other components that bind DNA non‐specifically. As shown in Figure 2d, the aptamer‐based fluorescence assay produced a detection limit of ∼10⁵ cfu mL⁻¹ for intact MRSA cells in buffer, where the intensity of the air fraction in each tube was subtracted from the intensity of the liquid fraction to account for any variations in autofluorescence of the tubes.

To increase the selectivity of MRSA detection and improve the detection limit we next evaluated the combination of aptamer‐directed MRSA isolation followed by fluorimetric detection using the RFD‐SA6T1 RCD, as this RCD is not activated by either S. pasteuri or S. epidermis.^{[ 36 ]} As shown in Figure 3a, non‐fluorescent aptamer coated beads are first suspended in 1 mL of a sample to capture MRSA cells, followed by removal of the supernatant and resuspension of the beads in 0.05 mL of a solution containing lysostaphin for 5 min to lyse the cells and release the protein that activates the RCD.^{[ 36 ]} A further 0.05 mL of 200 nM RFD‐SA6T1 RCD in 2x selection buffer (SB, which contains Mg(II) as a co‐factor; see ESI for buffer composition) was then added, producing a total preconcentration of 10‐fold for MRSA cells. Samples containing MRSA resulted in RCD cleavage and increased fluorescence upon resuspension, while samples without MRSA produced a low fluorescence intensity. Using this combined assay, samples containing MSSA should not be preconcentrated in the first step, while non‐target bacteria such as SP or SE should not lead to cleavage of the RCD. Hence, the combined assay should be highly selective for MRSA cells.

Figure 3.

a) Schematic of the combined assay using aptamer‐based pulldown of MRSA followed by RCD cleavage and fluorescence signalling. b) Selectivity of MRSA detection relative to MSSA and various non‐target cells, including E. coli, B. subtilis, S. epidermis and S. pasteuri. c) Fluorescence response for varying concentrations of MRSA present in selection buffer. d) Concentration‐response plot for MRSA detection based on fluorescence response at 90 min.

To confirm that the first step resulted in selective preconcentration of MRSA over MSSA, we first used the RCD for detection of MRSA or MSSA cells pulled down using the aptamer coated beads. Samples containing 1 x 10⁶ cfu mL⁻¹ of either MRSA or MSSA (intact cells with no extracellular media (CEM)) were incubated with PBP2a‐Apt1T1‐modified beads for 15 min, followed by removal of the supernatant, a single bead washing step, and resuspension of the beads in lysostaphin followed by RFD‐SA6T1, as noted above. In each case, the cleavage‐based fluorescence enhancement of the RCD with time was measured: 1) for 0.1 mL of the initial solution before the pulldown step but following the lysis step (F _B); 2) for 0.1 mL the lysed supernatant remaining after the pulldown step (F _A); and 3) for the beads after resuspension in 0.1 mL of the lysostaphin/RCD solution described above to assess enrichment.

For MSSA, the F _A/F _B ratio was 0.0064/0.0069 = 0.93 (based on slopes of the response curves), and thus 93% of cells remained after the pulldown step, and 7% of cells were removed from the solution (Figure S6A). However, after washing, the beads produced no fluorescence response (slope of 0.0) and thus the MSSA removed from the solution (7%) is due to non‐specific adsorption and is not carried forward to the RCD step after the beads are washed. However, in the case of MRSA, the F _A value was reduced to the level of background (slope of 0.0) and thus no cells remained in solution after bead isolation (100% removal of cells) (Figure S6B). On the other hand, the fluorescence signal of the resuspended beads, based on the initial slope over the first 10 min, was 0.042, which was over 10‐fold higher than the slope of the F_B response (0.0038), consistent with the expected 10‐fold preconcentration. Hence, essentially all MRSA cells were removed from the original solution and remained bound after washing to allow resuspension in the lysostaphin/RCD solution, and thus the use of aptamer‐coated beads provides a robust method for selective preconcentration of MRSA cells.

To further assess the selectivity of the combined assay, MRSA, MSSA, E. coli (EC), B. subtilis (BS), S. chromogenes (SC), S. lentus (SL), SP or SE were individually spiked into 1 mL of buffer at concentrations of 1 x 10⁸ cfu mL⁻¹, followed by aptamer‐based preconcentration and resuspension of the beads in the RCD/lysostaphin solution (0.1 mL final volume). In addition, a mixture of the 7 non‐target bacteria (10⁸ cfu mL⁻¹ each) were evaluated with no MRSA, 10⁴ or 10⁸ cfu mL⁻¹ of MRSA present. Figure 3b shows the resulting fluorescence intensity of the reaction tubes due to RCD cleavage over a 30 min period, and clearly shows that MRSA is selectively detected over MSSA, which produced only a slight signal above background (likely owing to a small amount of non‐specific binding as the MSSA concentration was 100x higher than that used in Figure S6A), while none of the non‐target bacteria produced a signal above background, either individually or as part of a mixture, as none of these bacteria can cleave the RCD.^{[ 36 ]}

Figure 3c shows the fluorescence versus time graphs for samples containing MRSA at concentrations ranging from 0 to 10⁸ cfu mL⁻¹ following the combined aptamer‐RCD assay. These data demonstrate both the selectivity of the assay, with similar signals being produced for 10³ cfu mL⁻¹ of MRSA and 10⁸ cfu mL⁻¹ of MSSA (compare signal levels at 90 min), corresponding to a selectivity co‐efficient of 10⁵. Figure 3d shows the concentration‐response curves for MRSA when using buffer as the matrix, indicating that the detection limit was −10³ cfu mL⁻¹ at 30 min, which is 100‐fold better than the aptamer‐based fluorescence assay described above (Figure 2d), and an order of magnitude better than previously reported aptamer‐based MRSA assays.^{[ 31} , ³² , ^{33 ]} The improved LOD value is likely due to a combination of MRSA enrichment and an improved K_d for the RCD^{[ 36 ]} relative to the aptamer.

The data in Figure 3c demonstrate that the rate of signal development, and thus the cleavage rate, depends on MRSA concentration. To further evaluate the cleavage rate, dPAGE gels were obtained after 30 min (see Figure S7) and the maximum cleavage rate (at 10⁸ cfu mL⁻¹ of MRSA) was 2% per min, or 1.22 x 10¹⁰ molecules per min. Thus, each MRSA bacterium can result in 122 cleavage events per min. We note however, that each bacterium may produce hundreds or even thousands of copies of the unknown target protein, and thus it is not possible to calculate a rate of cleavage per target protein, though it is likely to be far lower than the rate per bacterium. We note that the use of a cis‐acting RCD means that the RCD operates in single turnover mode, and thus it is not possible to generate a k _obs from our data.

While the fluorescence‐based assay provides a simple method for signal generation, it requires a fluorescence reader. To move the assay to an equipment‐free format, we also evaluated our previously reported RCD‐based lateral flow device^{[ 36 ]} for MRSA detection (see Figure S8 for a schematic of the LFD and Table S1 for the relevant sequences used for the LFD assay). Figure 4a shows data for the same samples described above in Figures 3b,c, using the RFD‐SA6T1‐based LFD for detection. In this case, following the aptamer pulldown and a 20 min RCD reaction the samples were further diluted by 2‐fold with a quenching buffer containing SDS, Tween 20 and pullulan^{[ 36} , ^{45 ]} and 90 µL of this solution was added to the LFD and eluted for 20 min before imaging and quantification of test line color intensity. In agreement with the fluorescence data, only samples containing MRSA produced a visible test line on the LFD, although 10⁸ cfu mL⁻¹ of MSSA produced a slight color intensity above background (Figure 4a). We note that our previous report showed that without the isolation step, MSSA samples produced strong test lines that were similar in intensity to those obtained with MRSA.^{[ 36 ]} The dynamic range of the assay was 10³–10⁷ cfu mL⁻¹ with a detection limit slightly above 10³ cfu mL⁻¹ (Figure 4b), which was close to a 10‐fold improvement over the LOD of 10⁴ cfu mL⁻¹ for MRSA obtained using the LFD without the isolation step,^{[ 36 ]} and is consistent with the 10‐fold preconcentration of MRSA using the aptamer isolation step.

Figure 4.

Combined aptamer pull‐down / RCD lateral flow assays of bacteria. a) Selectivity of LFD for MRSA detection relative to MSSA and various non‐target cells, including E. coli, B. subtilis, S. epidermis and S. pasteuri. LFD test line intensity for varying concentrations of MRSA present in: b) selection buffer; c) simulated nasal mucus; d) human serum. Error bars are 1 standard deviation for experiments done in triplicate.

As a preliminary step toward clinical validation of the LFD assay, we evaluated the ability to detect MRSA that was spiked into either simulated nasal mucus (Figure 4c) or human serum (Figure 4d) at concentrations of 10–10⁸ cfu mL⁻¹. Following spiking, both sample types were diluted to 10% (v/v) using selection buffer, followed by testing in the same manner as described above for MRSA in buffer, except that elution was done for 30 min prior to imaging. In each case, the LFD images showed no background for samples without MRSA, demonstrating that the combined aptamer‐RCD LFD assay was compatible with both nasal mucus and serum samples, and showing that the isolation step using the aptamer removed any potential background interferences from such samples. In addition, the response curves demonstrated that MRSA could be detected at 10⁴ cfu mL⁻¹ visually or 10³ cfu mL⁻¹ when using a scanner in either matrix, further highlighting the advantages of MRSA isolation and pre‐enrichment. We note that cut‐off values to diagnose MRSA infections can range from as low as 300 cfu mL⁻¹ for wound swabs using PCR,^{[ 46 ]} 10⁴ cfu mL⁻¹ for LFD assays of wound swab cultures,^{[ 47 ]} and up to 10⁵ cfu mL⁻¹ of MRSA for PCR of urine samples.^{[ 48 ]} Hence the new assay should be able to detect MRSA at clinically relevant concentrations in a short time (about an hour) and without the need for complex equipment.

Conclusion

In summary, we have generated a new DNA aptamer for PBP2a that can bind to the native membrane‐associated protein on the surface of intact MRSA cells. This allowed the aptamer to be used as a capture element for affinity‐based isolation of MRSA from complex samples without interference from MSSA or other background bacteria. The isolation step was coupled with a signaling assay utilizing an RCD that was activated by a different protein in SA, which allowed for selective detection of MRSA based on either fluorescence or lateral flow methods, with no interference from non‐target bacteria such as SE or SP, which can bind to the PBP2a aptamer. The entire assay for MRSA can be completed in under an hour and does not require complicated instrumentation or advanced technical skills. Additional studies using patient samples are still required to fully validate the combined aptamer‐RCD assay. However, this work provides a foundation for developing a low‐cost in vitro diagnostic test that should be amenable rapid detection of MRSA in both clinical settings and resource limited areas.

Conflict of Interests

The authors declare no conflict of interest.

Supporting information

Supporting Information

Ali M. M., Almeida A. F., White D., Capretta A., Brennan J. D., Angew. Chem. Int. Ed.. 2025, 64, e202509598. 10.1002/anie.202509598

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.