Plug-and-play protein biosensors using aptamer-regulated in vitro transcription

Lee, Heonjoon; Xie, Tian; Kang, Byunghwa; Yu, Xinjie; Schaffter, Samuel W.; Schulman, Rebecca

doi:10.1038/s41467-024-51907-4

Download PDF

Article
Open access
Published: 12 September 2024

Plug-and-play protein biosensors using aptamer-regulated in vitro transcription

Nature Communications volume 15, Article number: 7973 (2024) Cite this article

5460 Accesses
6 Altmetric
Metrics details

Subjects

Abstract

Molecular biosensors that accurately measure protein concentrations without external equipment are critical for solving numerous problems in diagnostics and therapeutics. Modularly transducing the binding of protein antibodies, protein switches or aptamers into a useful output remains challenging. Here, we develop a biosensing platform based on aptamer-regulated transcription in which aptamers integrated into transcription templates serve as inputs to molecular circuits that can be programmed to a produce a variety of responses. We modularly design molecular biosensors using this platform by swapping aptamer domains for specific proteins and downstream domains that encode different RNA transcripts. By coupling aptamer-regulated transcription with diverse transduction circuits, we rapidly construct analog protein biosensors and digital protein biosensors with detection ranges that can be tuned over two orders of magnitude and can exceed the binding affinity of the aptamer. Aptamer-regulated transcription is a straightforward and inexpensive approach for constructing programmable protein biosensors that could have diverse applications in research and biotechnology.

Functional analysis of single enzymes combining programmable molecular circuits with droplet-based microfluidics

Article 26 February 2024

Independent control of the thermodynamic and kinetic properties of aptamer switches

Article Open access 07 November 2019

Automated design of protein-binding riboswitches for sensing human biomarkers in a cell-free expression system

Article Open access 27 April 2023

Introduction

Biosensors that detect proteins are crucial for diagnostics¹, smart therapeutics², and biomedical research. Traditional ways to detect or measure the concentrations of proteins, such as enzyme-linked immunosorbent assay (ELISA)³ or Western blot⁴ often rely on specific and sensitive binding of antibodies to a target protein. However, these assays are time-consuming and labor-intensive due to multiple incubation and washing steps. Mass spectrometry⁵, single-molecule microscopy⁶, and electrochemical sensors⁷ can detect diverse proteins, but these analytics require expensive and bulky instrumentation, infrastructure, and trained personnel. Further, the output signals of such assays cannot be easily tailored to applications besides protein quantification such as portable diagnostics^8,9 or molecular machines that perform targeted delivery^10,11 in response to disease biomarker detection. Lateral flow assays¹² that rely on antibody binding permit low-cost and rapid detection of proteins, but typically only report the presence of a biomarker and cannot be integrated with downstream reactions that process the input signals^13,14.

Ideally, molecular biosensors could generate a response to a target protein without manual intervention, could be read without specialized equipment, and could easily be programmed to produce a variety of responses. Antibodies¹⁵, protein switches¹⁶, and aptamers^17,18 can be designed to bind to specific proteins, but modularly designing circuits that transduce the protein binding event into a measurable output is challenging. For example, allosteric protein switches^19,20 or aptamers^21,22,23 that undergo conformational changes upon protein binding to transduce signals have been integrated with molecular circuits. However, designing these structure-switching molecules often requires extensive reengineering for each target protein or desired output²⁴. This makes it difficult to rapidly develop protein biosensors against additional targets or to adopt existing biosensors for additional functionalities.

Here, we present aptamer-regulated transcription for in vitro sensing and transduction (ARTIST), which enables rapid construction of protein biosensors that can detect diverse targets and easily integrate with downstream circuits for programmable responses. ARTIST uses aptamer-protein binding to regulate transcription of a DNA template in analogy to the regulation of transcription by protein transcription factor-DNA binding^9,25 (Fig. 1a). Using published DNA aptamer sequences, we show how to simply swap the aptamer domain in these DNA templates for aptamer-regulated transcription (dARTs) to selectively sense a range of proteins, including multiple cytokines involved in autoimmune diseases. Conversely, we modularly swap dART output domains to integrate dARTs with different molecular circuits to create biosensors with different capabilities (Fig. 1b).

Using ARTIST, we build analog biosensors whose outputs indicate protein concentration and digital biosensors in which substantial output is only produced when the protein’s concentration exceeds a chosen threshold. The dynamic range of these biosensors can be rationally tuned across multiple orders of magnitude to detect cytokines at physiologically relevant concentrations; even below the aptamer K_d, which often limits biosensor sensitivity in practice. We also demonstrate how ARTIST can be used to create biosensors that are robust to environmental parameters that biosensors are often sensitive to²⁶, such as ion concentration and enzyme activity. ARTIST is a powerful systems chemistry approach for the facile development of protein biosensors capable of programmable and rapid (<60 min) responses to diverse proteins.

Results

dART design and characterization

The formation of noncanonical DNA structures such as G-quadruplexes have been shown to regulate transcription in cells^27,28,29 and have been adapted to inhibit in vitro transcription via binding to ions³⁰ or thrombin³¹. We hypothesized that using this approach³¹ we could develop a general protein biosensing platform by designing DNA transcription templates, termed dARTs, with G-quadruplex-forming aptamers downstream of a promoter that repress transcription via protein-aptamer binding. dARTs consist of a promoter domain that T7 RNA polymerase (T7 RNAP) can recognize to initiate transcription, a single-stranded aptamer domain that can bind to a specific protein, and an output domain that T7 RNAP transcribes to produce an RNA output sequence that can react downstream (Fig. 2a).

**Fig. 2: Design of aptamer-regulated transcription templates (dARTs).**

The aptamer sequence was inserted into the template strand of the dART ^30,31, i.e., the strand the polymerase reads during transcription. The output domain of the dART was designed to be double-stranded to prevent spurious interactions between the dART and other species such as the RNA transcribed from the dART, as well as promoter-independent transcription of T7 RNAP (Supplementary Fig. 1). The aptamer and the output sequences are both transcribed from dARTs, so the resulting transcripts could have undesired secondary structure across different input and output sequences (Supplementary Fig. 2, Supplementary Information Note 1.1). To reduce the possibility of such interactions, we added insulation domains, composed of i and i’, to the dART design (Fig. 2b). The i and i’ domains are complementary and are located on either side of the aptamer sequence so that in the RNA transcript produced from the dART, these domains will hybridize to form a hairpin, sequestering the sequence transcribed from the aptamer domain. (Fig. 2a, Supplementary Information Note 1.2)^32,33. The i domain located just downstream of the promoter was designed to be 6 bases to facilitate efficient T7 RNAP transcription initiation³⁴ .

We next sought to characterize how our dART design performed in different experimental conditions. For a protein to reduce the rate of dART transcription, the protein must be able to bind to the dART’s aptamer and repress transcription, while under the same conditions dARTs should be efficiently transcribed when no protein is present. We began by designing a dART, termed the IFN-O1-dART, that contained an IFN-γ aptamer that forms a G-quadruplex to facilitate IFN-γ binding (Fig. 2c)^35,36. For characterization, we used a fluorescently labeled DNA reporter complex that reacts with IFN-O1-RNA via toehold-mediated strand displacement (TMSD) to displace one strand of the reporter complex to produce a fluorescent signal (Fig. 2d, “Methods”, Supplementary Fig. 3, Supplementary Information Note 1.3).

Many G-quadruplex aptamers require potassium ions for effective ligand binding^37,38. Consistent with this requirement, in the absence of K+, the same rate of fluorescence decrease (i.e., the rate of transcript reaction with the reporter) was observed for IFN-O1-dART in either the presence of 100 nM of IFN-γ or its absence (Fig. 2e), suggesting IFN-γ did not bind to the IFN-O1-dART to reduce transcription rate. When 100 mM KCl was also added, 100 nM IFN-γ was able to substantially repress transcription, consistent with IFN-γ binding to the dART. Without IFN-γ present, the rate of reporter reaction of the IFN-O1-dART was slightly lower with 100 mM of potassium than without potassium, which is consistent with the ability of G-quadruplexes to interrupt transcription^27,30. In addition, the insulin-linked polymorphic region^39,40 of the human insulin gene was able to fully repress transcription in the presence of 100 mM K+, suggesting that a strong G-quadruplex is sufficient to inhibit transcription in the dART design (Supplementary Supplementary Fig. 4, Supplementary Information Note 1.4)^30,36. Conversely, a dART with a “dummy” aptamer domain (Dummy-O1-dART), i.e., no specific affinity for IFN-γ and no G-quadruplex forming domain, had similar reacted reporter kinetics for 0 nM or 100 nM of IFN-γ at both 0 mM and 100 mM of K+ (Fig. 2c, e).

As additional controls, we verified that exogenous addition of the same aptamer³⁶ sequence after 30 min of transcription reversed inhibition of the IFN-O1-dART with 100 nM of IFN-γ (Supplementary Fig. 5). Conversely, adding IFN-γ after 30 min of transcription was able to fully repress transcription with 1000 nM of IFN-γ (Supplementary Fig. 6). We also tested dART variants with i domain lengths ranging from 2 bases to 22 bases (Supplementary Figs. 7, 8, Supplementary Information Note 1.5) and found the 6 base design yielded the greatest difference in transcription rates with and without protein in our experimental conditions (Supplementary Fig. 9).

We next developed a simple model⁴¹ to predict the dose–response and kinetic behavior for 10 nM of IFN-O1-dART incubated with IFN-γ concentrations ranging between 0 and 1000 nM. Our model assumes that transcription cannot occur when a protein is bound to a dART’s aptamer, and thus the apparent dissociation constant of the protein and the dART’s aptamer, which we termed K_d,apparent, will dictate the concentration of dART available for transcription at a given protein concentration (Supplementary Information Note 2)^42,43. This model qualitatively agrees with our measured dose–response curve for a K_d,apparent of 8 nM (Fig. 2f and Supplementary Figs. 10–12), which is on the same order as the measured K_d of the dART’s aptamer sequence^35,36. Simulations of reacted reporter kinetics were also in good agreement with those measured in experiment (Fig. 2g). Using the model, we then compared the predicted dose–response curve using the K_d,apparent of 8 nM to the experimental dose–response curve of 1 nM of IFN-O1-dART in the presence of (0, 1, 3, 5, 10, or 100) nM of IFN-γ. Simulated and experimental results were consistent with each other for both dose–response curve (Fig. 2h) and dose-dependent kinetic profiles (Fig. 2i). Using this model, we estimated the limit of detection^44,45,46 (LOD) for the 10 nM and 1 nM dARTs to be 6.3 nM (95% CI 3.7–8.9 nM) and 3.7 nM (95% CI 2.0–5.4 nM), respectively (Supplementary Figs. 13 and 14, Supplementary Information Note 3).

We then fit the dose–response curves to a logistic function. Using this fit, the LOD of 10 nM and 1 nM IFN-O1-dART were determined as 4.6 nM (95% CI 2.7–6.5 nM) and 2.0 nM (95% CI 1.0–3.0 nM), respectively, in agreement with the predictions of our kinetic simulations. Based on the slopes of the inflection point of the fit dose–response curves^47,48, the sensitivities of 10 nM and 1 nM IFN-O1-dART were −1.3 (95% CI −0.9–1.7) and −1.2 (95% CI −0.6–1.8), respectively (Supplementary Figs. 13, 14, and 15 Supplementary Information Note 3).

dART input and output modularity

We next asked whether the design of the IFN-O1-dART could be generalized to respond to different protein inputs and produce different RNA outputs. We replaced the IFN-γ aptamer in the IFN-O1-dART with G-quadraplex forming aptamers for thrombin⁴⁹, IL-6⁵⁰, and TNF-α⁵¹ to create Thr-O1-dART, IL6-O1-dART, and TNF-O1-dART (Fig. 3a, Supplementary Fig. 16, Supplementary Information Note 4.1). NUPACK⁵² predicted that the insulation domains would prevent undesired secondary structures from forming in these output RNA sequences (Supplementary Fig. 17).

**Fig. 3: dART dose–response, selectivity, and modularity.**

We measured the reacted reporter kinetics for 10 nM of each of these dARTs when combined with 0 to 1000 nM of their target proteins (Fig. 3b). For all dARTs, the rates of reacted reporter decreased with increasing protein concentrations. We next asked whether these changes in dART reporter reacted were due to specific binding between a target protein and its aptamer. We combined 10 nM of each dART and Dummy-O1-dART with 100 nM of each protein ligand and BSA as a control. Only when dARTs were subjected to their target protein was there a large decrease in the reacted reporter concentration (Fig. 3c and Supplementary Fig. 18). As expected, the dummy template produced a similar concentration of reacted reporter in the presence of all input proteins.

We next asked whether the output domain could also be exchanged modularly, which would allow dARTs to be easily connected to different downstream processes (Fig. 3d, Supplementary Information Note 4.2). We measured transcription profiles of IFN-O1-dART, IFN-O2-dART, and IFN-O3-dART which encode the O1, O2, and O3 output domains, respectively, for 0 to 1000 nM of IFN-γ (Supplementary Figs. 19 and 20). As expected, reacted reporter kinetics decreased with increasing IFN-γ input concentrations for each output (Fig. 3e). This suggests that the outputs of the dARTs could be easily customized to couple them to downstream circuits to create biosensors with different functionalities.

Not all aptamer sequences we pulled from the literature worked in dARTs. For example, we tested four aptamer sequences for VEGF⁵³ (Supplementary Figs. 21 and 22, Supplementary Information Note 4.3), and observed that one had the desired response (Supplementary Figs. 23 and 24). The other three sequences showed repressed transcription with potassium even without VEGF. Thus, these aptamer sequences may form stable enough G-quadruplexes with K+ alone to repress transcription. Alternatively, modeling of the secondary structure of the outputs of the poor performing dARTs suggested that some of the transcripts may adopt undesigned secondary structures that impedes their ability to react with their reporters (Supplementary Fig. 21). These results suggest a simple screening protocol for adopting additional aptamer sequences to ARTIST by measuring transcription of dARTs with and without potassium prior to testing with ligand.

Analog biosensors

We next asked whether we could use ARTIST to build biosensors with different functionalities, such as analog or digital response, or signal amplification by coupling dARTs to downstream reactions. In principle, the rate of increase of reacted reporter by a dART is a measure of the target protein’s concentration. However, rates are difficult to measure because they require observing a signal over time. We thus asked whether we might build analog biosensors that indicate protein concentration as the steady-state concentration of reacted reporter by adding RNase H. RNase H degrades RNA outputs of a dART bound to the DNA strand O1’_q^32,54, allowing O1’_q to rehybridize to O1_f to reform the quenched reporter complex after degradation (Fig. 4a). A balance of RNA production and degradation should therefore produce a steady-state response whose magnitude is dependent on the concentration of the target protein. The reacted reporter concentrations of IFN-O1-dART with 0 to 1000 nM of IFN-γ indeed reached different steady-state values for different input protein concentrations in the presence of RNase H (Fig. 4b, c). When Thr-O1-dART, IL6-O1-dART, and TNF-O1-dART were combined with reporter and 5 to 1000 nM of their corresponding input proteins, distinct steady-state reacted reporter concentrations were also observed after 240 min (Supplementary Information Note 5, Supplementary Fig. 25, Fig. 4c).

Digital biosensors

We next asked how we might use ARTIST to construct digital biosensors that have a high sensitivity across a threshold concentration i.e., they produce a high output signal when the input exceeds the threshold concentration and a low output signal otherwise. Digital biosensors are also important for measuring whether a sample satisfies a specific diagnostic criterion. We sought to design digital biosensors that invert the output signal of individual dARTs, so that a high protein concentration yields a high concentration of RNA output. We designed a digital biosensor for IFN-γ by integrating two dARTs to produce a comparator circuit. A reference dART with an aptamer domain that does not bind IFN-γ was designed to produce the O1 output (Ref-O1-dART). Another dART with the IFN-γ aptamer domain was designed to produce an output with partial complementarity to the O1 sequence (IFN-O1’-dART) (Fig. 5a, Supplementary Information Note 6.1, Supplementary Fig. 26). If the IFN-O1’-dART is added at a higher concentration than the Ref-O1-dART, then IFN-O1’-RNA will sequester Ref-O1-RNA, resulting in a low reporter signal in the absence of protein. As IFN-γ concentration is increased, the rate of IFN-O1’-RNA transcription will decrease, allowing the Ref-O1-RNA to react with the O1 reporter once a threshold IFN-γ concentration is reached (Fig. 5b). This comparator circuit also inverts.

**Fig. 5: Comparator circuits for digital biosensing.**

We first sought to build a comparator circuit that could threshold the concentration of IFN-γ so that [IFN-γ] < 20 nM would produce a low reacted reporter signal (OFF) and [IFN-γ] ≥ 50 nM would produce a high signal (ON). We used simulations with the previously measured K_d,apparent for the IFN-O1-dART (Supplementary Information Note 6.2, 6.3, and 6.4, Supplementary Fig. 27) to determine that 25 nM of Ref-O1-dART and 50 nM of IFN-O1’-dART should produce a digital response with the desired threshold (Supplementary Fig. 28 and Fig. 5c). These predictions were then confirmed in experiments (Fig. 5d, e). We termed this IFN-γ O1 comparator circuit involving 25 nM of Ref-O1-dART and 50 nM of IFN-O1’-dART, IFN-C-50-O1. Compared to 10 nM of IFN-O1-dART, the digital IFN-C-50-O1 biosensor exhibited a 5-fold increase in sensitivity across the threshold concentration (Supplementary Fig. 15, Supplementary Information Note 3).

We reasoned we could tune the IFN-γ threshold of IFN-C-50-O1 by changing the concentration of Ref-O1-dART, which would change the amount of IFN-O1’-dART that must be repressed by input protein to produce a reporter signal (Supplementary Information Note 6.4)¹³. In line with this intuition, we found that an IFN-γ O1 comparator circuit with 15 nM of Ref-O1-dART required ≥100 nM IFN-γ to turn on (IFN-C-100-O1), while a circuit with 40 nM Ref-O1-dART only required ≥30 nM IFN-γ to turn on (IFN-C-30-O1; Fig. 5f).

Robust digital responses

We hypothesized that the digital response of IFN-C-50-O1 would also be insensitive to different environmental factors that often affect biosensors. For example, T7 RNAP activity can vary with solution composition⁵⁵ or temperature⁵⁶, and such variations could lead to false positives or false negatives. Since a comparator computes the ratio of the transcription rates of a dART and a reference template, its output should not change substantially with T7 RNAP activity, because changes in T7 RNAP activity would alter the transcription rates of both templates in the same way. Indeed, the IFN-C-50-O1 produced low reacted reporter concentrations for [IFN-γ] <20 nM and high reacted reporter concentrations for [IFN-γ] ≥ 50 nM, when T7 RNAP activities were either 2, 4, or 8 U µl⁻¹ (Supplementary Figs. 29 and 30, Supplementary Information Note 6.5).

Different biological samples contain a wide range of potassium ion concentrations such as blood (3.5–5 mM)⁵⁷, saliva (20–22 mM)⁵⁸, and urine (40–100 mM)⁵⁹. The affinity of G-quadruplex-forming aptamers for their ligands is also sensitive to potassium^37,60,61, meaning that affinity assays involving these aptamers would have different performance in these different media. Consistent with this finding, dART transcription rate decreased with increasing potassium concentration (Supplementary Information Note 6.6, Supplementary Fig. 31). However, both dARTs comprising IFN-C-50 contain G-quadraplexes (Fig. 5g), so their transcription rates are affected in similar ways by different potassium concentrations. As a result, IFN-C-50-O1 had similar digital responses and response thresholds across different potassium concentrations (Fig. 5h). In contrast, a comparator with a reference dART with a dummy aptamer sequence (Fig. 2c, Supplementary Table 1) that does not form a G-quadruplex (Fig. 5i), becomes non-responsive at lower potassium ions concentrations (Fig. 5j, Supplementary Fig. 31) because the IFN-O1’-dART transcription rate changes with potassium concentration, but the Dummy-O1-dART transcription rate does not, changing the threshold. Another challenge to the use of ARTIST for biological samples is the presence of RNases in most samples. Compared to our conventional dART assay conditions, we found IFN-C-50-O1 had a similar sensitivity in 10 (v/v)% serum supplemented with RNase inhibitor (Supplementary Fig. 33)^23,62. We also confirmed that IFN-C-50-O1 triggered at the designed IFN-γ threshold concentration in 10 (v/v)% supernatant from cultures of Chinese hamster ovary (CHO) cells that were secreting VRC01 monoclonal antibodies (Supplementary Fig. 34). These results show how ARTIST could be used to construct digital biosensors that reliably detect proteins under different environmental conditions.

Output amplification

We next asked whether we might increase the amount of output produced by a comparator circuit in response to the same concentrations of protein input. We designed an amplified comparator (Fig. 6a, Supplementary Information Note 7) in which the RNA output of the comparator circuit activates transcription of a downstream genelet³² via strand displacement reactions that complete the promoter sequence of the genelet (Fig. 6b, Supplementary Figs. 35 and 36). To prevent the low level of comparator output that is produced even when [IFN-γ] is low from triggering the output, we included RNase H, so that only when the rate of Ref-C1-RNA production exceeds its degradation rate can it activate the genelet to produce an output.

**Fig. 6: Sensitive protein detection with amplified comparators.**

The amplified comparator (IFN-AC-50-O4) reacted with 2000 nM of O4 reporter (Fig. 6c) in response to 50 nM of IFN-γ in <45 min, more than 40 times the output produced by the non-amplified comparator to the same input (IFN-C-50-O4; Supplementary Fig. 37). Without IFN-γ, about 200 nM of O4 reporter reacted transiently, but this amount then decreased, presumably because of RNase H-catalyzed O4 RNA degradation. Interestingly, the ON/OFF ratio of IFN-C-50-O4 was only 2-fold to 3-fold (Supplementary Fig. 37), compared to IFN-AC-50-O4, which was nearly 10-fold (Fig. 6c). This increase in the fold-change between the ON and OFF signals may be due to two factors. First, Ref-O1-RNA reacts faster with the O1 reporter in IFN-C-50-O1 than Ref-C1-RNA does with B1 in IFN-AC-50-O4³². Thus, when IFN-C1’-RNA is in excess, a greater fraction of Ref-C1-RNA is sequestered in IFN-C-50-O4 than IFN-O4-RNA in IFN-C-50-O4. Second, in IFN-AC-50-O4, both Ref-C1-RNA and O4-RNA are degraded by RNase H, which results in a lower background signal without protein for IFN-AC-50-O4 than for IFN-C-50-O4.

We hypothesized that an amplified comparator such as IFN-AC-50-O4 could be tuned to produce a similar amount of high output to its corresponding non-amplified comparator (IFN-C-50-O4) beginning above a much lower threshold protein concentration. To build this more sensitive amplified comparator, we reduced the concentrations of the Ref-C1-dART and IFN-C1’-dART of IFN-AC-50-O4 50-fold to create a diluted amplified comparator, termed IFN-AC-1-O4. IFN-AC-1-O4 maximized reporter signal for IFN-γ concentrations of 1 nM or higher and otherwise produced <12% of the maximum reporter signal. Amplifying the output of the comparator thus makes it possible to reduce the threshold input concentration of the circuit 50-fold (Supplementary Fig. 38), decrease the limit of detection below 1 nM, and maintain high sensitivity (Fig. 6d, e, Supplementary Fig. 15, Supplementary Information Note 3). Furthermore, the concentration of reacted reporter, a measure of the RNA output, produced by IFN-AC-1-O4 at this threshold is >250-fold higher than the input protein concentration (Fig. 6d, e).

Discussion

ARTIST is a versatile platform for rapidly developing biosensors that report on protein concentrations. By coupling downstream reactions to a dART’s protein-controlled transcription processes, ARTIST can be easily tailored to produce biosensors with analog responses, tunable digital responses, and high sensitivities below the K_d of an aptamer.

ARTIST’s performance suggests it could be used for sensing proteins of interest in research and clinical settings. For example, IFN-γ is a ubiquitous cytokine involved in various malignancies (e.g., cancer, autoimmune diseases)⁶³, as well as a key signalling factor in immunotherapy. T-cells can secret up to 30 nM of IFN-γ during mitogenic stimulation⁶⁴ and activated lymphocytes can secrete up to 3.5 nM of IFN-γ after transfection of immunotherapies⁶⁵. The ARTIST biosensors presented here are already suited to detect proteins within these biologically relevant ranges, with the IFN-C-30-O1 and IFN-AC-1-O4 biosensors designed to respond at 8.7 nM and 0.4 nM of IFN-γ, respectively. Incorporating additional layers of genelet amplification⁶⁶ or other isothermal RNA amplification schemes that have been able to detect femtomolar concentrations of nucleic acids⁶⁷ could enable ARTIST biosensors that respond well below the current LOD of 0.4 nM^66,67.

Aptamers identified in previous studies were readily integrated into dARTs, suggesting the promise of developing a larger library of biosensors for a range of important protein targets. The selectivity of dARTs suggests that they may be combined to allow for multiplexed sensing or decision-making involving multiple protein inputs. They could also be integrated with methods such as proximity ligation⁶⁸ or multi-epitope targeting⁶⁹ to enhance selectivity for detection in more complex environments. G-quadruplex-forming aptamers, which have been developed for more than 2000 proteins⁷⁰, were investigated here, but non-G-quadruplex-forming sequences have also been reported to repress in vitro transcription³⁰, suggesting that ARTIST may apply to an even broader range of aptamers.

It should be possible to incorporate ARTIST into existing field-deployable workflows for rapid point-of-need biosensing. The constituents of ARTIST (T7 RNAP, RNase H, DNA complexes) can be easily freeze-dried for storage and the outputs are compatible with existing portable systems to measure fluorescence^8,9. Testing in matrices such as serum, blood, urine, and saliva can severely inhibit reaction conditions, and will need to be systematically studied. Our results in serum and supernatant from antibody secreting CHO cell cultures indicate RNase inhibitors can help with operating biosensors within biological samples by preventing RNA degradation^23,62.

The effects of environmental conditions such as ion concentration on aptamer affinity can limit their applicability in biosensors. We demonstrate how the ARTIST system can produce a measurement of IFN-γ concentration that is mostly insensitive to that concentration of potassium ions, despite the effects of potassium on aptamer affinity³⁸. Biological samples such as serum, blood, urine, and saliva, have different salt concentrations that can confound the readout of biosensors. ARTIST, through the ability to self-calibrate in response to its environment, may potentially ameliorate this issue. The incorporation of further self-calibration, background, or crosstalk subtraction⁷¹ or even feedback control methods such as adaptation⁷² might allow ARTIST to overcome many of these limitations, or further, even to achieve robustness levels exceeding those of many traditional affinity assays.

Methods

Materials

O4_f, O1’_q, O2’_q, O3’_q, and O4’_q were purchased from Integrated DNA Technologies (IDT) with HPLC purification (Supplementary Table 1). All other oligonucleotides were purchased under standard desalting conditions. Triphosphates (NTPs) were purchased from ThermoFisher Scientific (R0481). T7 RNAP was purchased in bulk (300,000 U) from Cellscript (200 U µl⁻¹, C-T7300K) as well as from ThermoFisher Scientific (200 U µl⁻¹, EP0113). NEB RNAPol reaction buffer (M0251S; 10X) and yeast inorganic pyrophosphatase (YIPP; M2403S; 0.1 U µl⁻¹) were purchased from New England Biolabs (NEB). RNase H was purchased from ThermoFisher Scientific (EN0201; 5 U µl⁻¹). Fetal bovine serum (FBS; 26140079) and RNase inhibitor (N8080119; 0.5 U µl⁻¹) were purchased from ThermoFisher Scientific. CHOZN cell line (Sigma-Aldrich) expressing the VRC01 monoclonal antibody was cultured in imMEDIAte Advantage Medium (Catalog No. 87093C; Millipore-Sigma-Aldrich) in batch mode. The suspension cells were cultured in 125 ml shake flask with working volume of 30 ml at 125 RPM, 37 °C, and 5% carbon dioxide. Cells were inoculated at 5.0 × 10⁵ cells mL⁻¹. Culture media supernatants were collected after 4 days of culture.

Recombinant Human IFN-γ (285-IF), IL-6 (206-IL), TNF-α (210-TA), and VEGF 165 (293-VE) were all purchased from R&D Systems Inc. (USA) in lyophilized form. Recombinant Human IFN-γ was reconstituted in sterile, deionized water, whereas Recombinant Human IL-6, TNF-α, and VEGF were all reconstituted in sterile PBS containing 0.1% BSA. Human α-thrombin (HCT0020) was purchased from Haematologic Technologies, Inc. (Essex, VT) and dissolved in 50% H₂O/glycerol.

dART annealing and preparation

dARTs were assembled by annealing their three strands, a promoter non-template strand (Prom-dART-nt), a non-template strand encoding the output sequence of choice (i.e., O1-dART-nt, O2-dART-nt, O3-dART-nt) and a template strand that contains the aptamer domain of choice that is complementary to the two non-template strands (i.e., IFN-O1-dART-t, Thr-O1-dART-t, IL6-O1-dART-t, TNF-O1-dART-t) at equimolar concentrations. As an example, Prom-dART-nt, O1-dART-nt, and IFN-O1-dART-t were combined in a standard 200 µL PCR tube (VMR; 20170-010) at concentrations of 1 µM per strand in 1X NEB RNAPol reaction buffer supplemented with KCl to a final concentration of 100 mM. To anneal, mixtures were heated to 90 °C, incubated for 5 min, then cooled to 20 °C at a rate of 1 °C min⁻¹.

Reporter annealing and preparation

DNA reporters were prepared by diluting the fluorophore-modified DNA strand with its partially complementary quencher-modified DNA strand at a concentration of 10 µM per strand in 1X NEB RNAPol reaction buffer. The reporter mixture was heated to 90 °C, incubated for 5 min, then cooled to 20 °C at a rate of 1 °C min⁻¹.

Amplified comparator annealing and preparation

Genelet initially in a blocked state (G1O4:B1) was prepared by mixing G1O4-nt, O4-t, and B1 together in 1X NEB RNAPol reaction buffer at equimolar concentrations. The genelet mixture was heated to 90 °C, incubated for 5 min, then cooled to 20 °C at a rate of 1 °C min⁻¹.

Reaction conditions

Reactions were all conducted at 37 °C in 1X NEB RNAPol reaction buffer supplemented with KCl to a final concentration of 100 mM and NTPs (ATP, UTP, CTP, GTP) at a final concentration of 2 mM each unless otherwise stated. We included 100 mM KCl to promote the proper folding of G-quadruplex structures within aptamer domains²⁷. In addition to T7 RNA polymerase, YIPP was also included in reactions (1.35 × 10⁻³ U μl⁻¹) to extend the duration of the transcription reactions. The reaction conditions mentioned above are referred to as ARTIST reaction conditions. The concentrations of each of the molecules in each experiment are given in Supplementary Tables 10–31 under Supplementary Information Note 8. The total volume of the reaction was set at 25 µL for all assays.

To perform experiments, solutions containing dART templates under ARTIST reaction condition mentioned were first added to wells of a 384-well plate. Proteins were then added at the concentrations described set at a volume of 0.5 µL and in most experiments incubated for 30 min to 60 min at room temperature. In Supplementary Fig. 6, IFN-γ was added after 30 min of transcription to measure repression of IFN-O1-dART. In Fig. 3b, we also added 0.5 µL of sterile PBS containing 0.1% BSA into the assays with 10 nM of IL6-O1-dART or TNF-O1-dART reacting with 0 nM of IL-6 or TNF-α, respectively. This was done to ensure equal salt concentrations from the PBS buffer. Similarly in Supplementary Fig. 15, we added 0.5 µL of sterile PBS containing 0.1% BSA into all assays that did not have 100 nM of IL-6 or TNF-α. IFN-C-O1-50 was mixed in 10 (v/v)% FBS (Supplementary Fig. 33) or 10 (v/v)% CHO cell culture media (Supplementary Fig. 34), and 0 to 100 nM of IFN-γ. The reaction mix was incubated with 0.5 U µL⁻¹ of RNase inhibitor for 2 h before adding in T7 RNAP and YIPP.

Data acquisition

Fluorescence readings were then taken for 10 min to 25 mins to measure minimum fluorescence values before T7 RNAP, and YIPP were added to initiate the reactions (Supplementary Information Note 1.2, Supplementary Fig. 3a). At the end of the experiments, 0.5 µL of a DNA strand fully complementary to o1’_q or o4’_q was mixed into each assay at a final concentration of 2.5 µM to obtain a maximum O1 or O4 DNA reporter fluorescence intensity. Fluorescence data were then normalized using Eq. 1 as follows (Supplementary Fig. 3b):

$$\left[{{{\rm{Reacted\; Reporter}}}}\right]= ({{{\rm{Total}}}} \, \left[{{{\rm{DNA\; Reporter}}}}\right]) \\ *\frac{{{{\rm{Fluorescence}}}}-{{{\rm{Min}}}} \, ({{{\rm{Fluorescence}}}})}{{{{\rm{Max}}}} \, \left({{{\rm{Fluorescence}}}}\right)-{{{\rm{Min}}}} \, ({{{\rm{Fluorescence}}}})}$$

(1)

All kinetic data were obtained using either a BioTek Synergy H1 or Cytation 5 plate reader (Agilent Technologies). All fluorescence readings were measured using Gen5 3.11 (BioTek Synergy H1) and 3.12 (Cytation 5). HEX was measured with an excitation peak of 533 nm and an emission peak of 559 nm with a gain of 80 to 100 to ensure fluorescence values were within the linear range of detection. Cy3 was measured with excitation peak of 555 nm and emission peak of 569 nm with a gain 60 to 100 to ensure fluorescence values were within the linear range of detection. FAM was measured with excitation peak of 487 nm and emission peak of 527 nm with a gain of 60. Fluorescence measurements were taken every minute during reactions.

Data analysis

Determination of standard deviations for all data points shown in Supplementary Figs. 13, 14, and 15 were determined from the covariance estimates using SciPy’s curvefit function.

Statistics and reproducibility

For technical replicates, all experiments were conducted using the same instrument, reagents, and experimental conditions on two or three separate days. Error bars and shadings indicate ± one s.d. Randomization is not relevant to our study. Under Supplementary Note 3, the 95% confidence intervals of EC₅₀ and p parameters were determined from the covariance estimates supplied by SciPy’s curvefit function.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data generated in this study are provided in the Supplementary Information and Source data file. Source data are provided with this paper.

Code availability

All data analysis, simulations for K_d,apparent, reacted reporter kinetics of dARTs, reacted reporter kinetics, dose–response curves of the comparator, and LogFit analysis of sensitivity and LOD are available at: https://doi.org/10.7281/T1/IRL0IE⁴¹.

References

Soper, S. A. et al. Point-of-care biosensor systems for cancer diagnostics/prognostics. Biosens. Bioelectron. 21, 1932–1942 (2006).
Article CAS PubMed Google Scholar
Shen, L., Wang, P. & Ke, Y. DNA nanotechnology‐based biosensors and therapeutics. Adv. Health. Mater. 10, 2002205 (2021).
Article CAS Google Scholar
Rissin, D. M. et al. Single-molecule enzyme-linked immunosorbent assay detects serum proteins at subfemtomolar concentrations. Nat. Biotechnol. 28, 595–599 (2010).
Article CAS PubMed PubMed Central Google Scholar
Rivero-Gutiérrez, B., Anzola, A., Martínez-Augustin, O. & de Medina, F. S. Stain-free detection as loading control alternative to Ponceau and housekeeping protein immunodetection in Western blotting. Anal. Biochem. 467, 1–3 (2014).
Article PubMed Google Scholar
Jannetto, P. J. & Fitzgerald, R. L. Effective use of mass spectrometry in the clinical laboratory. Clin. Chem. 62, 92–98 (2016).
Article CAS PubMed Google Scholar
Orrit, M., Ha, T. & Sandoghdar, V. Single-molecule optical spectroscopy. Chem. Soc. Rev. 43, 973 (2014).
Article CAS PubMed Google Scholar
Wu, J., Liu, H., Chen, W., Ma, B. & Ju, H. Device integration of electrochemical biosensors. Nat. Rev. Bioeng. 1, 346–360 (2023).
Article PubMed PubMed Central Google Scholar
Pardee, K. et al. Paper-based synthetic gene networks. Cell 159, 940–954 (2014).
Article CAS PubMed PubMed Central Google Scholar
Jung, J. K. et al. Cell-free biosensors for rapid detection of water contaminants. Nat. Biotechnol. 38, 1451–1459 (2020).
Article CAS PubMed PubMed Central Google Scholar
Sedlmayer, F., Aubel, D. & Fussenegger, M. Synthetic gene circuits for the detection, elimination and prevention of disease. Nat. Biomed. Eng. 2, 399–415 (2018).
Article CAS PubMed Google Scholar
Douglas, S. M., Bachelet, I. & Church, G. M. A logic-gated nanorobot for targeted transport of molecular payloads. Science (1979) 335, 831–834 (2012).
CAS Google Scholar
Bahadır, E. B. & Sezgintürk, M. K. Lateral flow assays: principles, designs and labels. TrAC Trends Anal. Chem. 82, 286–306 (2016).
Article Google Scholar
Jung, J. K., Archuleta, C. M., Alam, K. K. & Lucks, J. B. Programming cell-free biosensors with DNA strand displacement circuits. Nat. Chem. Biol. 18, 385–393 (2022).
Article CAS PubMed PubMed Central Google Scholar
Ma, Q. et al. An automated DNA computing platform for rapid etiological diagnostics. Sci. Adv. 8, eade0453 (2022).
Ni, Y. et al. A plug-and-play platform of ratiometric bioluminescent sensors for homogeneous immunoassays. Nat. Commun. 12, 4586 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Choi, J. H., Laurent, A. H., Hilser, V. J. & Ostermeier, M. Design of protein switches based on an ensemble model of allostery. Nat. Commun. 6, 6968 (2015).
Article ADS CAS PubMed Google Scholar
Tuerk, C. & Gold, L. Systematic evolution of ligands by exponential enrichment: RNA ligands to bacteriophage T4 DNA polymerase. Science 249, 505–510 (1990).
Article ADS CAS PubMed Google Scholar
Ellington, A. D. & Szostak, J. W. In vitro selection of RNA molecules that bind specific ligands. Nature 346, 818–822 (1990).
Article ADS CAS PubMed Google Scholar
Ribeiro, L. F., Warren, T. D. & Ostermeier, M. Construction of protein switches by domain insertion and directed evolution. 43–55. https://doi.org/10.1007/978-1-4939-6940-1_3 (2017).
Quijano-Rubio, A. et al. De novo design of modular and tunable protein biosensors. Nature 591, 482–487 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Simmel, F. C., Yurke, B. & Singh, H. R. Principles and applications of nucleic acid strand displacement reactions. Chem. Rev. 119, 6326–6369 (2019).
Article CAS PubMed Google Scholar
Oh, S. S., Plakos, K., Lou, X., Xiao, Y. & Soh, H. T. In vitro selection of structure-switching, self-reporting aptamers. Proc. Natl Acad. Sci. 107, 14053–14058 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Vezeau, G. E., Gadila, L. R. & Salis, H. M. Automated design of protein-binding riboswitches for sensing human biomarkers in a cell-free expression system. Nat. Commun. 14, 2416 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Stein, V. & Alexandrov, K. Synthetic protein switches: design principles and applications. Trends Biotechnol. 33, 101–110 (2015).
Article CAS PubMed Google Scholar
van der Meer, J. R. & Belkin, S. Where microbiology meets microengineering: design and applications of reporter bacteria. Nat. Rev. Microbiol. 8, 511–522 (2010).
Article PubMed Google Scholar
Boyd, M. A., Thavarajah, W., Lucks, J. B. & Kamat, N. P. Robust and tunable performance of a cell-free biosensor encapsulated in lipid vesicles. Sci. Adv. 9, (2023).
Broxson, C., Beckett, J. & Tornaletti, S. Transcription arrest by a G quadruplex forming-trinucleotide repeat sequence from the human c-myb gene. Biochemistry 50, 4162–4172 (2011).
Article CAS PubMed Google Scholar
Bochman, M. L., Paeschke, K. & Zakian, V. A. DNA secondary structures: stability and function of G-quadruplex structures. Nat. Rev. Genet 13, 770–780 (2012).
Article CAS PubMed PubMed Central Google Scholar
Lee, C.-Y. et al. R-loop induced G-quadruplex in non-template promotes transcription by successive R-loop formation. Nat. Commun. 11, 3392 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Zhang, P. et al. Topologically switchable and gated transcription machinery. Chem. Sci. 13, 10555–10565 (2022).
Article CAS PubMed PubMed Central Google Scholar
Iyer, S. & Doktycz, M. J. Thrombin-mediated transcriptional regulation using DNA aptamers in DNA-based cell-free protein synthesis. ACS Synth. Biol. 3, 340–346 (2014).
Article CAS PubMed Google Scholar
Schaffter, S. W. et al. Standardized excitable elements for scalable engineering of far-from-equilibrium chemical networks. Nat. Chem. 14, 1224–1232 (2022).
Article CAS PubMed Google Scholar
Bhadra, S. & Ellington, A. D. Design and application of cotranscriptional non-enzymatic RNA circuits and signal transducers. Nucleic Acids Res. 42, e58–e58 (2014).
Article CAS PubMed PubMed Central Google Scholar
Conrad, T., Plumbom, I., Alcobendas, M., Vidal, R. & Sauer, S. Maximizing transcription of nucleic acids with efficient T7 promoters. Commun. Biol. 3, 439 (2020).
Article CAS PubMed PubMed Central Google Scholar
Kimoto, M., Yamashige, R., Matsunaga, K., Yokoyama, S. & Hirao, I. Generation of high-affinity DNA aptamers using an expanded genetic alphabet. Nat. Biotechnol. 31, 453–457 (2013).
Article CAS PubMed Google Scholar
Hashim, S. N., Tsuchiya, A., Kamiya, N. & Sando, S. A single fluorophore-labeled aptamer sensor for the detection of interferon gamma. Chem. Lett. 44, 1670–1672 (2015).
Article CAS Google Scholar
Bhattacharyya, D., Mirihana Arachchilage, G. & Basu, S. Metal cations in G-quadruplex folding and stability. Front. Chem. 4, 38 (2016).
Dong, J., O’Hagan, M. P. & Willner, I. Switchable and dynamic G-quadruplexes and their applications. Chem. Soc. Rev. 51, 7631–7661 (2022).
Article CAS PubMed Google Scholar
Lew, A., Rutter, W. J. & Kennedy, G. C. Unusual DNA structure of the diabetes susceptibility locus IDDM2 and its effect on transcription by the insulin promoter factor Pur-1/MAZ. Proc. Natl Acad. Sci. USA 97, 12508–12512 (2000).
Article ADS CAS PubMed PubMed Central Google Scholar
Connor, A. C., Frederick, K. A., Morgan, E. J. & McGown, L. B. Insulin capture by an insulin-linked polymorphic region G-quadruplex DNA oligonucleotide. J. Am. Chem. Soc. 128, 4986–4991 (2006).
Article CAS PubMed PubMed Central Google Scholar
Lee, H. et al. Data associated with the publication: Plug-and-play protein biosensors using aptamer-regulated in vitro transcription. https://doi.org/10.7281/T1/IRL0IE, Johns Hopkins Research Data Repository, V1 (2024).
Maganzini, N., Thompson, I., Wilson, B. & Soh, H. T. Pre-equilibrium biosensors as an approach towards rapid and continuous molecular measurements. Nat. Commun. 13, 7072 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Jarmoskaite, I., AlSadhan, I., Vaidyanathan, P. P. & Herschlag, D. How to measure and evaluate binding affinities. Elife 9, e57264 (2020).
De Acha, N., Socorro-Leránoz, A. B., Elosúa, C. & Matías, I. R. Trends in the design of intensity-based optical fiber biosensors (2010–2020). Biosensors 11, 197 (2021).
Article PubMed PubMed Central Google Scholar
Shrivastava, A. & Gupta, V. Methods for the determination of limit of detection and limit of quantitation of the analytical methods. Chron. Young. Sci. 2, 21 (2011).
Article Google Scholar
Mann, M. M. & Berger, B. W. A genetically-encoded biosensor for direct detection of perfluorooctanoic acid. Sci. Rep. 13, 15186 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Yu, J., Zhang, Y., Zhao, Y., Zhang, X. & Ren, H. Highly sensitive and selective detection of inorganic phosphates in the water environment by biosensors based on bioluminescence resonance energy transfer. Anal. Chem. 95, 4904–4913 (2023).
Article CAS PubMed Google Scholar
Swank, Z., Laohakunakorn, N. & Maerkl, S. J. Cell-free gene-regulatory network engineering with synthetic transcription factors. Proc. Natl Acad. Sci. 116, 5892–5901 (2019).
Article ADS CAS PubMed PubMed Central Google Scholar
Tasset, D. M., Kubik, M. F. & Steiner, W. Oligonucleotide inhibitors of human thrombin that bind distinct epitopes. J. Mol. Biol. 272, 688–698 (1997).
Article CAS PubMed Google Scholar
Spiridonova, V. A., Novikova, T. M. & Snigirev, O. V. Obtaining DNA aptamers to human interleukin-6 for biomagnetic immunoassay nanosensors. Mosc. Univ. Phys. Bull. 71, 135–138 (2016).
Article ADS Google Scholar
Orava, E. W., Jarvik, N., Shek, Y. L., Sidhu, S. S. & Gariépy, J. A short DNA aptamer that recognizes TNFα and blocks its activity in vitro. ACS Chem. Biol. 8, 170–178 (2013).
Article CAS PubMed Google Scholar
Zadeh, J. N. et al. NUPACK: analysis and design of nucleic acid systems. J. Comput. Chem. 32, 170–173 (2011).
Article CAS PubMed Google Scholar
Nonaka, Y. et al. Affinity Improvement of a VEGF Aptamer by in Silico Maturation for a Sensitive VEGF-Detection System. Anal. Chem. 85, 1132–1137 (2013).
Article CAS PubMed Google Scholar
Del Grosso, E., Franco, E., Prins, L. J. & Ricci, F. Dissipative DNA nanotechnology. Nat. Chem. 14, 600–613 (2022).
Article PubMed Google Scholar
Borkotoky, S. & Murali, A. The highly efficient T7 RNA polymerase: a wonder macromolecule in biological realm. Int J. Biol. Macromol. 118, 49–56 (2018).
Article CAS PubMed Google Scholar
Chamberlin, M. & Ring, J. Characterization of T7-specific ribonucleic acid polymerase. J. Biol. Chem. 248, 2235–2244 (1973).
Article CAS PubMed Google Scholar
Nakhoul, G. N. et al. Serum potassium, end-stage renal disease and mortality in chronic kidney disease. Am. J. Nephrol. 41, 456–463 (2015).
Article CAS PubMed Google Scholar
Aps, J. K. M. & Martens, L. C. Review: The physiology of saliva and transfer of drugs into saliva. Forensic Sci. Int. 150, 119–131 (2005).
Article CAS PubMed Google Scholar
Maciel, A. T., Vitorio, D. & Osawa, E. A. Urine biochemistry assessment in the sequential evaluation of renal function: Time to think outside the box. Front. Med. 9, 912877 (2022).
Ogasawara, S. Transcription driven by reversible photocontrol of hyperstable G-quadruplexes. ACS Synth. Biol. 7, 2507–2513 (2018).
Article CAS PubMed Google Scholar
Russo Krauss, I. et al. High-resolution structures of two complexes between thrombin and thrombin-binding aptamer shed light on the role of cations in the aptamer inhibitory activity. Nucleic Acids Res. 40, 8119–8128 (2012).
Article CAS PubMed PubMed Central Google Scholar
Soltani, M. & Bundy, B. C. Streamlining cell-free protein synthesis biosensors for use in human fluids: In situ RNase inhibitor production during extract preparation. Biochem Eng. J. 177, 108158 (2022).
Article CAS Google Scholar
Gocher, A. M., Workman, C. J. & Vignali, D. A. A. Interferon-γ: teammate or opponent in the tumour microenvironment? Nat. Rev. Immunol. 22, 158–172 (2022).
Article CAS PubMed Google Scholar
Zhu, H. et al. Detecting cytokine release from single T-cells. Anal. Chem. 81, 8150–8156 (2009).
Article CAS PubMed PubMed Central Google Scholar
Tzeng, S. Y. et al. In situ genetic engineering of tumors for long-lasting and systemic immunotherapy. Proc. Natl Acad. Sci. USA 117, 4043–4052 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Wan, X. et al. Cascaded amplifying circuits enable ultrasensitive cellular sensors for toxic metals. Nat. Chem. Biol. 15, 540–548 (2019).
Article CAS PubMed Google Scholar
Pardee, K. et al. Rapid, low-cost detection of Zika virus using programmable biomolecular components. Cell 165, 1255–1266 (2016).
Article CAS PubMed Google Scholar
Fredriksson, S. et al. Protein detection using proximity-dependent DNA ligation assays. Nat. Biotechnol. 20, 473–477 (2002).
Article CAS PubMed Google Scholar
Chang, X. et al. Construction of a multiple-aptamer-based DNA logic device on live cell membranes via associative toehold activation for accurate cancer cell identification. J. Am. Chem. Soc. 141, 12738–12743 (2019).
Article CAS PubMed Google Scholar
Shahsavar, K., Hosseini, M., Shokri, E. & Xu, G. New insight into G-quadruplexes; diagnosis application in cancer. Anal. Biochem. 620, 114149 (2021).
Article CAS PubMed Google Scholar
Müller, I. E. et al. Gene networks that compensate for crosstalk with crosstalk. Nat. Commun. 10, 4028 (2019).
Article ADS PubMed PubMed Central Google Scholar
Frei, T. & Khammash, M. Adaptive circuits in synthetic biology. Curr. Opin. Syst. Biol. 28, 100399 (2021).
Article CAS Google Scholar

Download references

Acknowledgements

The authors thank Michael Betenbaugh and Junneng Wen for providing the CHO cell culture media, as well as Marc Ostermeier, Elizabeth Strychalski, Simon d’Oelsnitz, Moshe Rubanov, Everett Kengmana, Colin Yancey, and Lei Zhang for insightful conversations and comments on the manuscript. H.L. was supported by the Asan Foundation Biomedical Science Scholarship. S.W.S. was supported by a National Research Council Postdoctoral Fellowship. R.S. acknowledges support from NIH R21CA251027-01A1, NSF CIF Medium 2107246, and ARO award W911NF2010057 and ARL award W911NF2220246. The National Institute of Standards and Technology notes that certain commercial equipment, instruments, and materials are identified in this paper to specify an experimental procedure as completely as possible. In no case does the identification of particular equipment or materials imply a recommendation or endorsement by NIST, nor does it imply that the materials, instruments, or equipment are necessarily the best available for the purpose.

Author information

Authors and Affiliations

Biomedical Engineering, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Heonjoon Lee
Biochemistry and Molecular Biology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA
Tian Xie
Institute for Nanobiotechnology, Johns Hopkins University, Baltimore, MD, USA
Byunghwa Kang
Chemical and Biomolecular Engineering, Johns Hopkins University, Baltimore, MD, USA
Xinjie Yu & Rebecca Schulman
National Institute of Standards and Technology, Gaithersburg, MD, USA
Samuel W. Schaffter
Computer Science, Johns Hopkins University, Baltimore, MD, USA
Rebecca Schulman
Chemistry, Johns Hopkins University, Baltimore, MD, USA
Rebecca Schulman

Authors

Heonjoon Lee
View author publications
You can also search for this author in PubMed Google Scholar
Tian Xie
View author publications
You can also search for this author in PubMed Google Scholar
Byunghwa Kang
View author publications
You can also search for this author in PubMed Google Scholar
Xinjie Yu
View author publications
You can also search for this author in PubMed Google Scholar
Samuel W. Schaffter
View author publications
You can also search for this author in PubMed Google Scholar
Rebecca Schulman
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

H.L., S.W.S., and R.S. designed the research. H.L. conducted most of the experiments and simulations. T.X. performed some experiments in Figs. 2f, h, and 3b of the main text and Supplementary Fig. 9 under Supplementary Information. X.Y. performed some experiments for Fig. 3e. H.L., B.K., and R.S. conducted simulations and analysis to estimate the LODs of dARTs in Supplementary Information Note 3. H.L., B.K., S.W.S., and R.S. wrote the paper with feedback from the other authors.

Corresponding authors

Correspondence to Samuel W. Schaffter or Rebecca Schulman.

Ethics declarations

Competing interests

H.L., B.K., S.W.S., and R.S. are co-inventors of a pending patent application (WO/2024/118806). The remaining authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Peer Review File

Reporting Summary

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Lee, H., Xie, T., Kang, B. et al. Plug-and-play protein biosensors using aptamer-regulated in vitro transcription. Nat Commun 15, 7973 (2024). https://doi.org/10.1038/s41467-024-51907-4

Download citation

Received: 18 March 2024
Accepted: 19 August 2024
Published: 12 September 2024
DOI: https://doi.org/10.1038/s41467-024-51907-4