Thermodynamically programmed one-pot CRISPR platform for point-of-care SNP genotyping

Thermodynamically programmed one-pot CRISPR platform for point-of-care SNP genotyping | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Thermodynamically programmed one-pot CRISPR platform for point-of-care SNP genotyping I-Ming Hsing, Xiaolong Wu, Yanan Li, Yumeng Cao, Zibin Zhao, Hongyu Lu, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8802024/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract One-pot CRISPR diagnostics face a fundamental incompatibility: nucleic acid amplification requires rapid target accumulation, whereas CRISPR activation irreversibly consumes those substrates, destabilizing reaction kinetics. Existing strategies rely on empirical parameter balancing or external staging but lack an intrinsic mechanism to enforce reaction order within a single reactor. Here we introduce thermodynamic encoding as a molecular design principle that programs reaction order directly into DNA primers, enabling autonomous, threshold-gated activation of CRISPR only after sufficient amplicon accumulated. By embedding a defined free-energy differential between competing primers, the system evolves through two kinetically ordered amplification regimes, decoupling amplification from CRISPR signal transduction without physical separation or external triggers. This architecture relocates PAM dependence from native genomic targets to primer-encoded design, enabling detection of otherwise inaccessible loci while preserving single-nucleotide discrimination. An ordinary differential equation model captures the threshold behavior and establishes a predictable framework for primer design. Building on this principle, we develop Thermodynamically Encoded Molecular Programming for One-pot diagnostics (TEMPO), which achieves attomolar sensitivity within 30 min and enables sequencing-concordant SNP genotyping and pathogen detection in a single-step microfluidic format. Biological sciences/Biotechnology/Assay systems Biological sciences/Molecular biology/CRISPR-Cas systems Physical sciences/Chemistry/Chemical biology/Nucleic acids Physical sciences/Nanoscience and technology/Nanobiotechnology/Biosensors Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Main Fully integrated molecular diagnostics aim to combine nucleic acid amplification, molecular recognition and signal transduction within a single reaction environment 1,2 . Achieving such integration, however, requires multiple biochemical processes to occur in a defined temporal order despite sharing the same reactor and molecular intermediates 3 . When reaction order is externally imposed, through physical separation, staged operation or manual intervention, robust performance can be maintained 4–6 . In contrast, when competing reactions proceed simultaneously without intrinsic ordering, system-level instability emerges, undermining sensitivity, reproducibility and predictability 7,8 . Designing molecular systems that autonomously enforce reaction order therefore represents a central but unresolved challenge in integrated diagnostics. This challenge is particularly acute in one-pot CRISPR-based diagnostics 3,9,10 . CRISPR–Cas12a has enabled highly sensitive nucleic acid detection by coupling sequence-specific target recognition to collateral nuclease activity 11,12 . When combined with isothermal amplification, such systems promise rapid and portable diagnostics 13–15 . However, amplification and CRISPR activation form a tightly coupled reaction network that competes for the same target molecules 16 . Amplification requires sustained accumulation of target DNA, whereas CRISPR activation irreversibly consumes those targets through cis- and trans-cleavage. In a one-pot format, this intrinsic conflict destabilizes reaction kinetics, leading to premature target depletion, suppressed amplification and reduced sensitivity 10,16,17 . These limitations are amplified in applications requiring high specificity, such as single-nucleotide polymorphism (SNP) genotyping, where both stringent sequence discrimination and stable reaction kinetics are required 18–20 . Many clinically relevant SNPs underpin pathogen drug resistance, cancer heterogeneity and pharmacogenomic decision-making, yet their reliable detection remains difficult in one-pot CRISPR systems 18,19,21 . In this context, minor kinetic instability or premature target consumption can obscure single-nucleotide differences, rendering high-resolution genotyping unreliable. Existing strategies to address this interference have focused on mitigating, rather than resolving, the underlying lack of reaction order. Approaches such as empirical tuning of reagent concentrations 22,23 , weakening protospacer adjacent motif (PAM) recognition 24 , engineering crRNA 25,26 or primer 27,28 , or introducing physical or temporal separation 29,30 can partially alleviate competition between amplification and CRISPR activity. However, these solutions rely on external control or context-specific parameter balancing and therefore cannot guarantee robust performance across different targets, sequence contexts or operating conditions. Moreover, PAM dependence fundamentally constrains genomic addressability: fewer than a minority of clinically relevant SNP loci are proximal to canonical Cas12a PAM sequences 31,32 , rendering most variants inaccessible to conventional CRISPR diagnostics. As a result, current one-pot CRISPR systems face a persistent trade-off between workflow simplicity, target scope and single-nucleotide specificity. Here we propose that reaction order in one-pot CRISPR diagnostics can be molecularly programmed rather than externally imposed, enabling autonomous and predictable coordination of amplification and CRISPR transduction within a single reactor. We introduce thermodynamic encoding as a design principle in which reaction sequencing is embedded directly into DNA primer energetics. By engineering a defined free-energy asymmetry between competing amplification primers, the system autonomously evolves through kinetically ordered amplification regimes, prioritizing target accumulation before activating CRISPR-mediated signal transduction. This intrinsic, threshold-gated coordination decouples amplification from CRISPR activity within a single, unsegmented reactor, without the need for physical separation or external triggers. Importantly, this architecture relocates PAM dependence from native genomic targets to primer-encoded design, enabling access to loci that are otherwise undetectable while preserving single-nucleotide discrimination. Building on this principle, we develop Thermodynamically Encoded Molecular Programming for One-pot diagnostics (TEMPO), a programmable CRISPR platform that integrates isothermal amplification, PAM design and kinetic control into a single molecular framework. Using a quantitative ordinary differential equation model to guide primer design, TEMPO achieves predictable reaction ordering, attomolar sensitivity and robust SNP discrimination in a fully integrated one-pot format. We validate this approach using simulated HIV samples and clinically relevant human SNPs lacking proximal PAM sites and further demonstrate its applicability through integration into a single-step microfluidic diagnostic chip. Together, these results establish thermodynamic encoding as a generalizable design paradigm for coordinating competing biochemical processes within shared reactors, with implications extending beyond CRISPR diagnostics to integrated molecular systems that require autonomous reaction ordering. Result Thermodynamically encoded reaction ordering enables autonomous one-pot CRISPR coordination Previous efforts have attempted to temporally separate amplification and CRISPR activation using external gating elements such as photocaged crRNAs 25 or temperature switch 33 . While effective, these systems lack an intrinsic mechanism to prioritize amplification over CRISPR activation, resulting in uncontrolled competition for shared target molecules. Under such conditions, premature CRISPR engagement inevitably suppresses amplification efficiency and destabilizes signal generation. We reasoned that enforcing reaction order requires not parameter tuning or weakened CRISPR activity, but the introduction of an explicit energetic hierarchy between competing amplification pathways, such that reaction sequencing emerges as an inevitable consequence of molecular thermodynamics. To achieve this, we introduced two forward primers with a defined thermodynamic asymmetry (Fig. 1 A): a fully complementary primer (FP, ΔG FP−Template ≈-37.20 kcal/mol) that drives efficient amplification of PAM-free amplicons, and a deliberately mismatched primer (FPP, ΔG FPP−Template ≈-30.71 kcal/mol) that encodes a PAM but incurs a higher hybridization free-energy penalty. This energetic imbalance necessarily partitions amplification into two kinetically distinct regimes. At low template abundance, only the high-affinity FP pathway is thermodynamically accessible, leading to rapid accumulation of pristine amplicons that cannot activate Cas12a (Fig. 2 B). As amplification proceeds and template concentration increases, according to the relationship ΔG eff = ΔG° − RT ln([Template]) + C, the effective free-energy barrier for FPP binding is progressively reduced, enabling a second amplification regime that installs PAM-containing amplicons. CRISPR activation therefore emerges as a threshold behavior rather than an externally triggered event. Our results reveal that a defined free-energy difference between the two forward primers gives rise to two kinetically ordered amplification phases. Single-primer reactions (FP/BP or FPP/BP) each produced a single amplicon of the expected size, whereas reactions containing both forward primers generated two distinct products corresponding to the pristine and FPP-extended amplicons ( Fig. S2A ). Sequencing verified that the FPP primer introduces the engineered PAM without affecting the original product ( Fig. S2B ). Real-time kinetics further indicated that the higher-affinity FP primer drives the early phase of amplification, while the FPP primer, penalized by reduced binding strength, engages more slowly. At high template input, the rate difference narrows, consistent with mass-action effects overcoming the thermodynamic penalty ( Fig. S3) . Additional assays verified that only PAM-containing dsDNA efficiently activates Cas12a ( Fig. S4 ). Consistent with this mechanism, reactions containing only FP produced robust amplification without detectable CRISPR signal, whereas reactions driven solely by FPP exhibited weak and unstable fluorescence due to simultaneous amplification and target consumption. In contrast, reactions containing both primers reproducibly displayed delayed but strong CRISPR activation, reflecting autonomous reaction ordering encoded by primer thermodynamics rather than empirical parameter balancing (Fig. 1 C–D). Importantly, ongoing FP-driven amplification continuously replenished templates, preventing target depletion even after CRISPR activation and yielding stable one-pot kinetics. Based on this principle, we establish Thermodynamically Encoded Molecular Programming for One-pot diagnostics (TEMPO) and integrate it into a microfluidic platform (Fig. 1 E) for rapid pathogen detection and SNP analysis. (A) Thermodynamic design of two competing primer-binding pathways. A fully complementary forward primer (FP) readily hybridizes to the target and drives rapid RPA, producing pristine amplicons. In contrast, the PAM-containing primer carries mismatches (FPP), creating an unfavorable hybridization energy and preventing early engagement. Numbers of mismatches in FPP can be designed via ΔΔG penalty, FPP−Template = RT ln ([Template Pristine amplicons ] / [Template Threshold ]). (B) Conceptual free-energy landscape illustrating chemical potential-driven pathway activation in the TEMPO reactor. Pristine RPA initiates immediately along a low-barrier pathway due to favorable primer–template hybridization, producing PAM-free amplicons that accumulate without engaging CRISPR. The PAM-introducing amplification branch is initially inaccessible because an intentional mismatch imposes a higher free-energy barrier. As pristine amplicons accumulate, the increasing chemical potential of the template pool provides a concentration-dependent driving force that progressively lowers the effective barrier, enabling PAM-containing amplification. The resulting mixed amplicon population activates Cas12a via PAM-dependent cis -cleavage, followed by trans -cleavage of the reporter for signal generation. (C and D) TEMPO validation using DNA target 1 (derived from HIV genome, 1 fM) and DNA target 2 (derived from Human genome, 1 fM). (E) Schematic of on-site TEMPO detection using a single-step microfluidic chip. Following sample loading, the one-pot amplification–CRISPR reaction proceeds autonomously, and fluorescence signals are captured and interpreted by a smartphone-based readout. Quantitative modeling establishes reaction ordering as a predictable design outcome To formalize this behavior and distinguish intrinsic reaction ordering from empirical tuning, we constructed a quantitative reaction-network model describing primer competition, amplification and PAM-dependent Cas12a activation (Supplementary Note 1). The model comprises coupled ordinary differential equations that encode primer hybridization kinetics, strand extension and CRISPR-mediated cleavage as mass-action and Michaelis–Menten processes. 1. Primer competition and thermodynamic gating We modeled primer binding as bimolecular association processes. Central to the engineered thermodynamic switch is the mismatch-bearing primer FPP, whose reduced hybridization free energy is encoded directly into slower kinetic rate constants: $$\:\frac{d[\text{F}\text{P}\text{P}\cdot\:\text{D}]}{dt}={k}_{6}\left[\text{F}\text{P}\text{P}\right]\left[\text{D}\right]-{k}_{6}^{\text{ext}}[\text{F}\text{P}\text{P}\cdot\:\text{D}]$$ in contrast to fully complementary FP: $$\:\frac{d[\text{F}\text{P}\cdot\:\text{D}]}{dt}={k}_{1}\left[\text{F}\text{P}\right]\left[\text{D}\right]-{k}_{1}^{\text{ext}}[\text{F}\text{P}\cdot\:\text{D}]$$ with \(\:{k}_{6}<{k}_{1}\) capturing the thermodynamic penalty imposed by the designed mismatch. This asymmetry produces two kinetically distinct amplification channels: a high-affinity pristine pathway leading to the native amplicon \(\:D\) , and a delayed mismatch pathway that generates the PAM-containing amplicon \(\:{D}_{2}\) . 2. Sequential amplification and PAM emergence The reactor evolves according to mass-action coupling between strand-displacement extension and template regeneration: \(\:\frac{d\left[D\right]}{dt}={k}_{3}\left[\text{B}\text{P}\right]\left[\text{S}\text{T}\right]+{k}_{4}\left[\text{F}\text{P}\right]\left[\text{A}\text{T}\right]+{k}_{5}\left[\text{S}\text{T}\right]\left[\text{A}\text{T}\right]-{k}_{13}\left[\text{C}\right]\left[D\right]\) \(\:\frac{d\left[{D}_{2}\right]}{dt}={k}_{8}\left[\text{B}\text{P}\right]\left[{\text{S}\text{T}}_{2}\right]+{k}_{11}\left[\text{F}\text{P}\text{P}\right]\left[{\text{A}\text{T}}_{2}\right]+{k}_{12}\left[{\text{S}\text{T}}_{2}\right]\left[{\text{A}\text{T}}_{2}\right]-{k}_{14}\left[\text{C}\right]\left[{D}_{2}\right]\) where the emergence of \(\:{D}_{2}\) is controlled by the mismatch-driven delay in FPP-mediated extension. These expressions quantitatively reveal that PAM introduction is not an external trigger but a threshold phenomenon emergent from the kinetic imbalance between the two amplification tracks. 3. PAM-dependent CRISPR transduction as a catalytic nonlinear activation Cas12a activation is represented using two parallel Michaelis–Menten channels with dramatically different catalytic efficiencies: $$\:{v}_{1}=\frac{{k}_{\text{c}\text{a}\text{t}1}\left[\text{C}\text{D}\right]\left[R\right]}{{K}_{m1}+\left[R\right]},{v}_{2}=\frac{{k}_{\text{c}\text{a}\text{t}2}\left[{\text{C}\text{D}}_{2}\right]\left[R\right]}{{K}_{m2}+\left[R\right]}$$ with \(\:{k}_{\text{c}\text{a}\text{t},2}\gg\:{k}_{\text{c}\text{a}\text{t},1}\) . This structure creates an activation nonlinearity normally seen in engineered chemical reactors where a slow path primes the system for abrupt activation once a specific intermediate (here, \(\:{D}_{2}\) ) reaches a critical concentration. 4. Model-experiment agreement validates a predictable sequential reactor In this framework, the mismatch-bearing FPP primer is represented by a reduced effective association rate constant relative to FP, directly reflecting its higher hybridization free-energy. This asymmetry necessarily generates two amplification pathways with distinct kinetic accessibility. Model simulations predict three defining features of thermodynamically encoded reaction ordering: (i) suppression of CRISPR activation during early exponential amplification, (ii) delayed emergence of PAM-containing amplicons once a threshold template concentration is reached, and (iii) a sharp, nonlinear onset of trans-cleavage activity following PAM installation. Experimental kinetics closely matched these predictions across multiple targets and primer designs (Fig. 1 C to D; Fig. S5 ). The observed delay in CRISPR activation, its threshold-like onset, and the sustained amplification after activation all emerged without invoking external triggers or time-dependent parameter changes. Importantly, altering primer free-energy penalties or PAM identity shifted the predicted threshold and activation kinetics in a manner quantitatively consistent with experimental observations, confirming that primer thermodynamics act as a tunable control variable rather than an empirical adjustment. Programmable DNA engineering operates as a multi-axis kinetic control element Building on the model, we next examined whether primer engineering could systematically shape reaction dynamics. Conventional RPA–CRISPR-Cas12a coupling, primarily regulates reaction kinetics by modifying crRNA–target interactions 34–36 . RPA primers are seldom altered, as sequence changes frequently compromise amplification efficiency. In contrast, in TEMPO the amplification efficiency is primarily contributed by the pristine primer FP, while FPP serves as a molecular bridge linking RPA output to Cas12a activation. This architecture means that programming FPP allows simultaneous tuning of both RPA progression and CRISPR kinetics, offering a capability inaccessible to traditional approaches. We infer that FPP exposes three orthogonal axes of design flexibility: (i) Varying FPP length modulated hybridization free-energy and predictably shifted the timing of PAM-containing amplification, thereby tuning the delay between amplification and CRISPR activation. (ii) Independently, altering the encoded PAM sequence modulated Cas12a activation strength without substantially affecting amplification kinetics, decoupling signal transduction from target generation. (iii) Finally, modifying the FPP–crRNA overlap region effectively introduced programmable mismatches at the recognition interface, providing an additional layer of kinetic control without rescreening crRNAs. Firstly, we systematically adjusted FPP length while maintaining mismatch position constant (Fig. 2 A). Shorter FPP variants, with weaker binding affinity, slowed and reduced PAM-containing product formation, extending the temporal separation between amplification and Cas12a activation (Fig. 2 B to C ). Conversely, increasing FPP length lowered the thermodynamic penalty, advancing the onset of CRISPR activation. These results indicate that mismatch-encoded ΔG operates as a kinetic valve, determining when the secondary amplification branch begins contributing to reactor output. We next tuned Cas12a activation independently of amplification by varying the PAM triplet within FPP (Fig. 2 D to F ). PAM identity modulated Cas12a activation strength without substantially affecting RPA kinetics. Canonical PAMs (TTT) induced rapid Cas12a activation driven by efficient cis -cleavage, whereas sub-optimal PAMs (TTG, ATT) produced more gradual and sustained activation kinetics. This behavior likely arises because weaker PAMs mediate reduced Cas12a cleavage activity, thereby alleviating competition between CRISPR cleavage and PAM-containing amplification. In contrast, excessively weak PAMs resulted in minimal signal output, consistent with their intrinsically low Cas12a activation efficiency. Finally, inspired by prior studies showing that crRNA mismatches modulate Cas12a kinetics 36 , but usually require extensive crRNA screening, we hypothesized that mutating the FPP–crRNA overlap region would indirectly introduce mismatch into the crRNA recognition interface, providing an alternative regulatory handle. Indeed, kinetic profiles confirmed this effect (Fig. 2 G to I ). Across these perturbations, reaction trajectories followed the same underlying sequence predicted by the model, differing only in timing and amplitude. This consistency demonstrates that thermodynamic encoding transforms primer design into a compact, multi-dimensional control space for shaping one-pot reaction dynamics, rather than a trial-and-error optimization problem. (A) Schematic illustrating modulation of primer–template hybridization thermodynamics by varying the length of the FPP primer, thereby tuning the rate of PAM-containing secondary amplification. Primer length serves as a programmable handle to adjust − ΔG FPP−Template . All reactions were performed with a fixed template input of 1 fM. (B and C) Kinetic profiles of TEMPO reactions programmed by FPP primer length. “Full” denotes the full-length FPP primer, and “−n” indicates truncation by n nucleotides. Panel B and panel C correspond to DNA targets derived from the HIV and human genomes, respectively. (D) Schematic illustrating FPP primers encoding different PAM sequences modulate CRISPR signal output. (E and F) Kinetic profiles of TEMPO reactions programmed with different PAM variants. Panel E shows reactions evaluated using DNA target 1 derived from the HIV genome (PAMs: TTG, TTT, TAT, TAG, and AAG), whereas panel F shows reactions evaluated using DNA target 2 derived from the human genome (PAMs: TTT, ATT, TCT, ACT, and ATA). (G) Schematic of crRNA-overlap engineering, where FPP primer modification introduces an effective mismatch at the CRISPR recognition locus. ( H and I ) Kinetic profiles of TEMPO reactions programmed with effective mismatches at defined positions downstream of the PAM. MUT-n indicates a single-nucleotide mismatch introduced n bases after the PAM via FPP–crRNA overlap engineering. Panel H and panel I correspond to DNA targets derived from the HIV and human genomes, respectively. DNA engineering enables broad-spectrum and highly specific SNP detection CRISPR-based diagnostics have been widely applied to SNP genotyping 20,37 . However, practical implementation remains constrained by two longstanding challenges: (i) strict PAM requirements limit accessible loci 11,38 , and (ii) single-nucleotide discrimination is not consistently achieved, particularly in one-pot RPA–CRISPR formats 39,40 . As a result, many clinically important SNPs remain undetectable. Given the programmable kinetic control demonstrated previously in TEMPO, we reasoned that FPP primer engineering could convert arbitrary SNPs, regardless of PAM availability, into high-fidelity, one-pot detectable targets. We first validated that TEMPO could redirect Cas12a targeting solely through FPP design (Fig. 3 A to C) . Within a fixed RPA amplicon, repositioning the FPP primer selectively introduced PAMs at user-defined locations, enabling Cas12a activation at either of two sites. Additional sequence designs further support the generality of this strategy (Fig. S6 ). Importantly, this targeting flexibility did not compromise sensitivity, as amplification driven by FPP remained intact. This property is particularly critical for SNP genotyping, where PAMs are often absent near the variant. Previous work has shown that crRNA or PAM engineering can enhance SNP discrimination 34 ; here we demonstrate that DNA-level primer engineering achieves comparable, and in some cases superior, control. Building on earlier results showing that PAM identity modulates reaction kinetics, we evaluated whether PAM tuning could improve allelic contrast. Systematic substitution of the PAM triplet within FPP (Fig. 4 D) yielded tunable genotype separation: canonical PAMs (TTT) produced 2.8-4.5-fold discrimination between SNP and WT alleles (Fig. 4 E), whereas suboptimal PAM variants (ATT) further elevated contrast to 5.9-18.4-fold (Fig. 4 F). We attribute this enhancement to moderate reductions in Cas12a affinity that amplify thermodynamic discrimination between matched and mismatched dsDNA substrates. However, certain SNPs remained poorly resolved even with optimized PAM weakening (Fig. 4 H). We therefore introduced a second layer of control by programming the FPP–crRNA overlap region, effectively inserting an additional mismatch at the recognition interface (Fig. 4 G). This approach mirrors crRNA engineering but eliminates the need for extensive crRNA rescreening. A single additional mismatch increased wild-type/mutant separation up to 3.6-fold (Fig. 4 I), converting previously non-resolvable SNPs into clearly distinguishable genotypes. Together, these results show that FPP primer design enables direct control of both CRISPR target accessibility and allele-level discrimination within a single amplification reaction. In summary, TEMPO expands the CRISPR detection landscape from PAM-constrained to universally addressable SNP sites, while substantially simplifying assay design. DNA primer programming alone is sufficient to convert undetectable SNPs into high-fidelity readouts, positioning TEMPO as a generalizable molecular framework for precision genotyping and sequence-specific diagnostics. ( A ) Relocate targeting site via primer engineering. Diagram of the two forward-selective primers (FPP1 and FPP2) positioned at different sites on the target sequence. (B) Amplification curves for the rs199476104 target (1 fM input) with/without FPP1. The pristine forward primer initiates amplification earlier, whereas FPP1 exhibits a delayed onset due to reduced binding affinity. (C) Amplification curves for the rs199476104 target (1 fM input) with/without FPP2, showing a similarly delayed and sequential onset of amplification as observed with FPP1. ( D ) Design strategy for PAM switching via FPP mismatch engineering. Mismatches within the FPP-binding region reassign the PAM sequence, enabling allele-dependent Cas12a activation. ( E and F ) Cas12a cleavage results for canonical PAMs (TTT) and suboptimal PAM variants (ATT) using the rs671 target (1 fM input). ( G ) Design strategy using a single-base insertion to enhance allele discrimination. The insertion is introduced downstream of the suboptimal PAM-encoding position within FPP, deliberately shifting the crRNA–target register to create a defined single-base mismatch in the seed region. (H and I) Cas12a cleavage results for the rs1229984 target (1 fM input) comparing designs without and with an insertion mutation. Rapid pathogen detection via TEMPO To assess the translational potential of TEMPO for rapid pathogen detection, we evaluated its performance using HIV RNA spiked into human serum as simulated clinical specimens. As illustrated in Fig. 4 A, extracted RNA was directly introduced into a single-tube TEMPO reaction and reported by real-time fluorescence, requiring no workflow branching or post-amplification handling. Building on its programmable reaction design, TEMPO demonstrated high analytical sensitivity across a broad dynamic range. Fluorescence output scaled proportionally with input RNA from 1,000 aM down to 1 aM, enabling both quantitative readout and naked-eye endpoint discrimination, with a limit of detection of 1 aM (Fig. 4 B; Fig. S7 ). Kinetic trajectories showed strong linear correlation with RNA input (R² = 0.9684), closely matching RT–qPCR calibration ( Fig. S8 ). Additional specificity testing against four non-target species confirmed that off-target templates produced only background-level Cas12a activation ( Fig. S9 ), confirming high sequence specificity of TEMPO. We next tested diagnostic performance using simulated clinical samples spanning clinically relevant viral loads. Among the 11 samples tested, TEMPO classification was fully concordant with RT-qPCR results: all nine samples with Ct values below 35 were identified as positive, whereas the two samples with Ct values ≥ 35 were classified as negative. This binary classification was fully concordant with RT–qPCR (Fig. 4 C). Notably, positive–negative separation emerged within ~ 20 min under isothermal conditions, substantially faster than thermal cycling qPCR and without the need for multistep sample manipulation. Collectively, these results establish TEMPO as a single-tube, isothermal platform that achieves qPCR-comparable sensitivity, rapid turnaround, and robust sequence selectivity. The combination of attomolar detection, minimal workflow complexity, and compatibility with crude clinical matrices highlights TEMPO as a promising framework for point-of-care pathogen surveillance and decentralized molecular diagnostics. (A) Schematic workflow for HIV detection using TEMPO. HIV mock samples were prepared and subjected to RNA extraction, followed by isothermal one-pot CRISPR detection or RT-qPCR analysis as a reference method. (B) Sensitivity of TEMPO for detecting different concentrations of HIV DNA (10000 aM,1000 aM, 100 aM, 10 aM,1 aM,0.1 aM, and negative control). Strong fluorescence signals were observed down to 1 aM concentration, with clear visual fluorescence under blue light. NTC, non-template control reaction. Error bars represent the means ± standard deviation (s.d.) from replicates. The unpaired two-tailed t-test was used to analyze the statistical significance. (C) Fluorescence output of TEMPO across extracted HIV RNA mock samples with varying RT–qPCR Ct values. Signal intensity decreases progressively with increasing Ct, consistent with the reduced RNA in put in higher Ct samples. On-site human genotyping by using TEMPO Having established the flexibility, extensibility, and allele-level selectivity of TEMPO under controlled benchmark conditions, we next evaluated its performance in human SNP genotyping. We analyzed 20 clinical genomic samples spanning three medically relevant loci, rs1229984 41 , rs671 41 , and rs199476104 42 , each representing distinct challenges in thermodynamic discrimination. For rs671 and rs199476104, single-nucleotide resolution was achieved solely through PAM attenuation, whereas rs1229984 required a combined strategy involving suboptimal PAM insertion and a single engineered mismatch encoded within the FPP primer. This comparative result underscores the modularity of FPP design and demonstrates that TEMPO can be rationally programmed to resolve thermodynamically recalcitrant SNP sites. To translate TEMPO into a clinically viable POC workflow, we developed an integrated platform coupling sample collection, room-temperature lysis, lyophilized one-pot amplification–CRISPR activation, and fluorescence readout into a 35 min sample-to-answer pipeline (Fig. 5 A). Buccal-swab lysates were processed by 5 min extraction-free lysis and directly loaded into a palm-sized microfluidic chip (diameter 4 cm, Fig. S10 ) preloaded with freeze-dried TEMPO reagents. Individual reaction chambers can be customized with locus-specific RPA primers and crRNAs. Upon rehydration, reactions self-initiate and produce a quantifiable readout after a 30 min incubation at 37°C. Fluorescence interpretation is automated through a custom smartphone application, removing the need for manual thresholding or trained operators. This architecture generatively supports on-site genotype determination within 35 min. We first validated genotyping accuracy using allele-specific crRNAs. TEMPO clearly distinguished wild-type (WT), heterozygous (HET), and homozygous mutant (SNP) states via distinct kinetic trajectories ( Fig. 5 B ) , while maintaining attomolar analytical sensitivity (Fig. S11) . All 20 clinical samples showed full concordance relative to Sanger-validated genotypes across the three loci (Fig. 5 C to D; Table S4 ): rs1229984 (4 WT / 6 SNP / 10 HET), rs671 (16 WT / 4 HET), and rs199476104 (20 WT). The system remained robust under direct-lysis conditions ( Fig. S12 and S13 ), and a sucrose–mannitol lyophilization matrix retained near-native activity after storage and rehydration, enabling operation without cold-chain logistics ( Fig. S14 and S15 ). For on-site SNP genotyping, the validated TEMPO reaction was integrated into a portable microfluidic chip (Fig. S10, S16 and S17) . Distinct chip architectures were developed to support either dual-target or multi-target detection, allowing the number of loci interrogated to be adjusted according to application-specific requirements while preserving a unified one-pot workflow. Clinical sample testing demonstrated clear and reliable signal outputs (Fig. 5 E; Fig. S18 ), enabling both direct visual interpretation and automated genotype assignment via a custom smartphone application (Fig. 5 F), with minimal user intervention. Collectively, these results establish TEMPO as a rapid, lyophilization-compatible SNP genotyping platform with programmable resolution down to single-nucleotide variation. By relocating PAM control from the native genome to FPP-encoded design, TEMPO unlocks previously inaccessible loci and greatly simplifies assay re-targeting without extensive crRNA screening. This modularity, coupled with chip integration and mobile analytics, positions TEMPO as a practical foundation for point-of-care precision genotyping and population-scale genetic screening. (A) Workflow of the SNP genotyping using TEMPO. Swab samples are collected and lysed at room temperature for approximately 5 min. The crude lysate is applied directly to the SNP Chip, where a one pot RPA–Cas12a reaction is performed at 37°C for 30 min. Fluorescence signals are then recorded and interpreted using a smartphone interface. (B) Representative amplification trajectories illustrating allele-dependent kinetics for rs199476104. Distinct curves corresponding to the wild-type, heterozygous, and mutant genotypes are generated within the one-pot reaction using a 1 fM standard template input. (C) Heat map summarizing TEMPO genotyping outcomes for 20 clinical samples across three SNP loci (rs1229984, rs671, and rs199476104). The discrimination factor (DF) encoded in each cell quantitatively captures the relative fluorescence contrast between SNP and wild-type alleles, enabling unambiguous genotype assignment with consistent performance across all samples and loci. Genotype calls were assigned based on DF thresholds, with DF values of 0–0.5 classified as wild type, 0.5–2.0 as heterozygous, and 2.0–10 as SNP. (D) Cleavage rate measured from Sample 20 as a representative example. The observed cleavage patterns match the expected allelic state and show strong agreement with Sanger sequencing traces provided above each bar. (E) Images of the SNP chip shown with a Hong Kong two-dollar coin for size reference, including the assembled microfluidic chip and representative endpoint fluorescence patterns from sample 20 after a 30 min reaction, highlighting allele-dependent signal differences. (F) Smartphone interface used for real-time data acquisition, visualization and automated SNP reporting. Discussion Fully integrated one-pot CRISPR diagnostics promise simplified workflows and improved accessibility, yet their practical deployment has been persistently limited by the difficulty of coordinating nucleic acid amplification and CRISPR-mediated signal transduction within a single reaction environment 10 . In most existing systems, these processes compete for shared intermediates, resulting in unstable kinetics, premature target consumption, and inconsistent signal generation 43 . Previous solutions, such as empirical parameter tuning 17 , physical compartmentalization 44 , or externally imposed temporal control 45,46 , have alleviated these conflicts only partially, often at the cost of robustness, generalizability, or operational simplicity. From an engineering perspective, these constraints underscore a fundamental limitation: reaction order in one-pot CRISPR diagnostics is rarely designed, but instead indirectly enforced. In this work, we introduce thermodynamically encoded reaction ordering as a molecular engineering strategy that represents a conceptual shift in one-pot CRISPR diagnostics. Rather than mitigating interference through weakened interactions or external staging, TEMPO encodes reaction sequencing directly into molecular free-energy landscapes, converting reaction order from an operational challenge into a programmable system property. This framework simultaneously addresses a major barrier to SNP genotyping: dependence on protospacer adjacent motifs. By encoding PAM sequences within primers rather than relying on native targets, TEMPO enables PAM-independent access to arbitrary loci while preserving single-nucleotide discrimination, expanding genomic addressability for clinically relevant variants. At the platform level, TEMPO achieves attomolar sensitivity, single-nucleotide resolution, and sequencing-concordant genotyping across multiple human SNP loci within a fully integrated workflow. Compatibility with extraction-free lysis, lyophilized reagents, and single-step microfluidic implementation further supports deployment in point-of-care and resource-limited settings, achieving high performance without additional reaction steps, specialized hardware, or user intervention. Together, these features establish TEMPO as a programmable, robust, and clinically adaptable framework for one-pot CRISPR diagnostics. Several limitations define the current scope of the approach. First, while TEMPO supports robust qualitative and semi-quantitative readouts, its quantitative precision does not yet match that of RT–qPCR. This reflects fundamental differences in reaction control: qPCR enforces deterministic amplification stages through thermal cycling 47 , whereas isothermal amplification exhibits greater variability in initiation and growth kinetics 48,49 . In one-pot CRISPR systems, this variability is compounded by dynamic sharing of amplicons between amplification and CRISPR consumption, rendering signal intensity dependent on coupled reaction dynamics rather than template abundance alone. Consequently, TEMPO is presently best suited for presence–absence testing and genotyping applications rather than absolute quantification. A second limitation concerns SNP discrimination, which remains inherently dependent on local sequence context. Identical single-nucleotide variants can exhibit divergent discrimination performance depending on neighboring bases and thermodynamic landscape 50–52 . Although multilayer energetic modulation, combining PAM down tuning with engineered mismatches, can enhance allelic resolution, predictive design rules for SNP discrimination remain incomplete, necessitating empirical validation for new targets. Looking forward, data-driven and physics-informed modeling approaches may help address these challenges 51,53 . Machine learning frameworks trained on large-scale kinetic and sequence datasets could enable predictive mapping between primer energetics, reaction dynamics, and allelic discrimination performance, improving both quantitative accuracy and target generalizability 54,55 . More broadly, the principle of thermodynamically encoded reaction ordering is not limited to RPA or Cas12a and may be extended to other amplification chemistries and CRISPR effectors, providing a generalizable strategy for coordinating competing biochemical processes within shared reactors. In summary, TEMPO establishes thermodynamic encoding as a practical engineering paradigm for designing predictable, integrated, and deployable one-pot CRISPR diagnostics. By unifying reaction sequencing, PAM independence, and SNP discrimination within a single molecular design framework, this work advances the feasibility of scalable point-of-care and at-home nucleic acid testing, bridging the gap between CRISPR diagnostic concepts and real-world biomedical application. Method Materials Lachnospiraceae bacterium Cas12a (LbCas12a) protein was from Tolo Biotech (Shanghai). RNA and DNA oligonucleotides, as well as synthetic DNA templates (Table S1 ), were provided by GENEWIZ (Suzhou, China). Acrylamide/bisacrylamide solution was purchased from Bio-Rad. SYBR Green Nucleic Acid Gel Stain, and SYTO 82 Orange Fluorescent Nucleic Acid Stain were purchased from Thermo Fisher Scientific (Hong Kong). DEPC Treated Water was from Sangon Biotech (Shanghai). DNA Isothermal Rapid Amplification Kit, RNA Isothermal Rapid Amplification Kit and Rapid Nucleic Acid Release Reagent were from Amp-future (Changzhou). RNase inhibitor, and NEBuffer 2.1 were from New England Biolabs. Proteinase K, Magnetic Bead Virus DNA/RNA Kit and Magnetic Universal Genomic DNA Kit were from Tiangen (Beijing). One Step RT-qPCR Master Mix and Universal Blue qPCR SYBR Green Master Mix was from YEASEN (Shanghai). Sucrose and mannitol were from Macklin (Shanghai). RPA and real-time RPA reactions, gel analysis, and sequencing RPA reactions were carried out in a 50 µL reaction mixture containing 29.5 µL primer-free rehydration buffer, 1 µL forward primer (10 µM), 1 µL FPP primer (10 µM), 2 µL reverse primer (10 µM), 4 µL template, 2.5 µL MgOAc (280 mM), and 1 µL DNase/RNase-free water. Reactions were incubated at 37°C for 60 min. To terminate amplification, Proteinase K was added to the completed reaction and incubated at room temperature for 30 min. The reaction was then heat-inactivated at 95°C for 10 min. The resulting products were either analyzed on 10% denaturing PAGE for gel visualization or submitted for Sanger sequencing (GENEWIZ). For real-time RPA, the reaction was assembled as described above with the addition of 1 µM SYTO 82 stain, and fluorescence was recorded continuously during incubation at 37°C. TEMPO system One-pot RPA–Cas12a reactions were assembled in a 20 µL final volume as summarized in Table S2. Briefly, 25 nM LbCas12a and 50 nM crRNA were premixed and incubated for 5 min at room temperature to allow formation of the ribonucleoprotein complex. The reaction mixture contained 400 nM reporters, 125 nM each of primer FP and primer FPP, 250 nM primer BP, NEB Buffer 2.1, the rehydration buffer supplied with the RPA kit, and nuclease-free water. 2 µL Template DNA was added to each reaction. MgOAc (280 mM) was added last to initiate amplification, yielding a total reaction volume of 20 µL. All reactions were mixed gently and incubated under standard one-pot amplification conditions. The fluorescent signal (485 nm excitation and 535 nm emission) was monitored with a qPCR system (Bio-Rad) with an interval time of 1 min. Sample preparation HIV mock samples. HIV RNA standards were prepared by spiking 0.5 µL of HIV RNA into 49.5 µL of human serum, followed by nucleic acid extraction using a magnetic bead–based purification kit. The purified RNA was used as the template for the TEMPO one-pot assay as well as RT- qPCR benchmarking. Human genomic samples. Twenty volunteers were recruited at the Hong Kong University of Science and Technology under institutional protocol SP-2024-0197 and HREP-2021-0285. Buccal swabs were collected from each participant and divided into two aliquots. One aliquot was processed using the Magnetic Universal Genomic DNA Kit to obtain purified genomic DNA, which served as the input for one-pot TEMPO genotyping and for Sanger sequencing after RPA amplification of target. The second aliquot was subjected to rapid lysis and used directly for microfluidic chip–based detection. Briefly, the swab sample was immersed in a centrifuge tube containing 1 mL of lysis buffer and vortexed thoroughly, followed by incubation at room temperature for 5 min. The tube was briefly centrifuged for 5 s, and 10 µL of the supernatant was collected as the template for direct amplification. RT-qPCR validation of HIV mock samples. HIV mock samples were validated using the Hifair Advanced One-Step RT-qPCR SYBR Green Kit (Yeasen). Each 25 µL reaction mixture contained 12.5 µL of 2× Hifair Advanced SYBR Green Buffer, 1 µL of forward primer (10 µM), 1 µL of reverse primer (10 µM), 1 µL of Hifair Advanced UH Enzyme Mix, 4 µL of RNA template, and nuclease-free DEPC-treated water to the final volume. RT-qPCR was performed with an initial reverse transcription step at 50°C for 6 min, followed by pre-denaturation at 95°C for 5 min. Amplification was then carried out for 40 cycles consisting of denaturation at 95°C for 15 s and annealing/extension at 60°C for 30 s. Design and fabrication of the microfluidic diagnostic chip The microfluidic diagnostic chip was designed in SolidWorks to integrate multiple reaction chambers and fluidic pathways for parallel SNP detection, with chamber volumes defined to accommodate 20–25 µL reactions. Passive Tesla-type valve structures were incorporated to prevent contamination arising from liquid backflow during the reaction process 56 . The chip employed a two-layer architecture, consisting of a CNC-machined polymethyl methacrylate (PMMA) layer containing the microfluidic channels and a non-fluorescent adhesive sealing film (Microseal B, Bio-Rad) as the bottom layer. Vent ports were sealed with a waterproof, breathable expanded polytetrafluoroethylene (ePTFE) membrane (diameter, 3.5 mm) to allow gas exchange while preventing liquid leakage and external contamination. The finalized designs were fabricated from PMMA using CNC micromachining by a commercial microfabrication service provider (Shenzhen Huilixing Precision Manufacturing Co., Ltd., China). Reagent loading and lyophilization on chip The lyophilizable one-pot RPA–Cas12a reaction mixture was prepared according to the formulation summarized in Table S3. A 20 µL lyophilizable one-pot RPA–Cas12a reaction mixture was prepared, containing RPA reagents were reconstituted in DEPC-treated water, 125 nM each forward primers FP and FPP, 250 nM reverse primer R, 50 nM Cas12a, 100 nM crRNA, and 800 nM ssDNA reporter, supplemented with 5% (w/w) sucrose and 2% (w/w) mannitol as lyoprotectants. The reaction mixture was dispensed into the reaction chambers (~ 25 µL per chamber). Chips were snap-frozen by positioning them approximately 1 cm above liquid nitrogen vapor for 2 min, followed by lyophilization for 12 h in the dark. After returning to room temperature, the bottom sealing film and top vent membrane were affixed to complete device assembly. Statistics and reproducibility Statistical analyses were performed using GraphPad Prism. Quantitative data (endpoint fluorescence and time-to-positive measurements) are reported as mean ± s.d. from at least two independent replicates unless stated otherwise. Fluorescence data were normalized using DF to quantify the relative signal contrast between SNP and wild-type reactions. The DF was calculated as 29 , DF = (F SNP − F NTC ) / (F WT − F NTC ), where F SNP , F WT , and F NTC represent the fluorescence intensities measured for SNP, wild-type, and no-target control reactions, respectively. Data availability All nucleic acid sequences used in this study are provided in manuscripts or supplementary information. The raw datasets generated and analyzed during the study are available from the corresponding authors on reasonable request. Code availability The TEMPO analyzer mobile phone application and TEMPO model developed in this work are available in https://github.com/zibin-zhao/TEMPO . Declarations Acknowledgments We thank the developers of OpenAI’s ChatGPT for assistance with language editing of the manuscript. The funding is in part provided by the Research Grants Council (GRF# 16304225, 16303522) of the Hong Kong SAR Government, China. Author contributions Conceptualization: IM.H., X.W., YL. Methodology: IM.H., X.W., YL., Y.C., Z.Z. Software: Z.Z. Validation: YL., X.W., Z.Z., S.L. Formal Analysis: YL., X.W., Y.C., Z.Z. Investigation: YL., X.W. Resources: IM.H. Data Curation: YL., X.W., Z.Z. Visualization: YL., X.W., Y.C., H.L. Funding acquisition: IM.H. Project administration: IM.H. Supervision: IM.H. Writing – original draft: YL., X.W., Y.C. Writing – review & editing: IM.H., X.W., YL., Y.C., Z.Z. Competing interests I-ming HSING, Xiaolong WU and Yanan LI are co-inventors of a patent application related to this work filed by the Hong Kong University of Science and Technology. Additional information Correspondence and requests for materials should be addressed to I-ming Hsing. References Rolando, J. C., Melkonian, A. V. & Walt, D. R. The Present and Future Landscapes of Molecular Diagnostics. Annu. Rev. Anal. Chem. 17 , 459–474 (2024). Rolando, J. C., Melkonian, A. V. & Walt, D. R. The Present and Future Landscapes of Molecular Diagnostics. Annu. Rev. Anal. Chem. 17 , 459–474 (2024). Chen, L., Hu, M. & Zhou, X. Trends in developing one-pot CRISPR diagnostics strategies. Trends Biotechnol. 43 , 98–110 (2025). Kadimisetty, K. et al. Fully 3D printed integrated reactor array for point-of-care molecular diagnostics. Biosens. Bioelectron. 109 , 156–163 (2018). Rodriguez, N. M., Wong, W. S., Liu, L., Dewar, R. & Klapperich, C. M. A fully integrated paperfluidic molecular diagnostic chip for the extraction, amplification, and detection of nucleic acids from clinical samples. Lab. Chip 16 , 753–763 (2016). Hu, S. et al. All-in-One Digital Microfluidics System for Molecular Diagnosis with Loop-Mediated Isothermal Amplification. Biosensors 12 , 324 (2022). Branca, M., Calvet, C., Limoges, B. & Mavré, F. Specific Versus Non‐specific Response in Exponential Molecular Amplification from Cross‐Catalysis: Modeling the Influence of Background Amplifications on the Analytical Performances. ChemPhysChem 22 , 1611–1621 (2021). Rolando, J. C., Melkonian, A. V. & Walt, D. R. The Present and Future Landscapes of Molecular Diagnostics. Annu. Rev. Anal. Chem. 17 , 459–474 (2024). Liu, H. et al. One‐Pot Isothermal Nucleic Acid Amplification Assisted CRISPR/Cas Detection Technology: Challenges, Strategies, and Perspectives. Adv. Sci. 12 , e06716 (2025). Li, X. et al. CRISPR-based one-pot detection: A game-changer in nucleic acid analysis. Biosens. Bioelectron. 288 , 117786 (2025). Chen, J. S. et al. CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity. Science 360 , 436–439 (2018). Li, S.-Y. et al. CRISPR-Cas12a-assisted nucleic acid detection. Cell Discov. 4 , 20 (2018). Kaminski, M. M., Abudayyeh, O. O., Gootenberg, J. S., Zhang, F. & Collins, J. J. CRISPR-based diagnostics. Nat. Biomed. Eng. 5 , 643–656 (2021). Feng, W. et al. CRISPR technology incorporating amplification strategies: molecular assays for nucleic acids, proteins, and small molecules. Chem. Sci. 12 , 4683–4698 (2021). Zhang, L., Jiang, H., Zhu, Z., Liu, J. & Li, B. Integrating CRISPR/Cas within isothermal amplification for point-of-Care Assay of nucleic acid. Talanta 243 , 123388 (2022). Yao, Z. et al. Tuning the Dynamic Reaction Balance of CRISPR/Cas12a and RPA in One Pot: A Key to Switch Nucleic Acid Quantification. ACS Sens. 9 , 3511–3519 (2024). Ding, X. et al. Ultrasensitive and visual detection of SARS-CoV-2 using all-in-one dual CRISPR-Cas12a assay. Nat. Commun. 11 , 4711 (2020). Zhang, H.-X. et al. Cas12a-based one-pot SNP detection with high accuracy. Cell Insight 2 , 100080 (2023). Zhuang, S., Bai, B. & Liu, Y. CRISPR-based SNP detection technologies advance from classical methods to cutting-edge innovations. Chem. Commun. 61 , 13345–13358 (2025). Chai, L. et al. CRISPR/Cas Toolbox: Unveiling the secret weapon for SNP detection. TrAC Trends Anal. Chem. 192 , 118321 (2025). Kohabir, K. A. V. et al. In Vitro CRISPR-Cas12a-Based Detection of Cancer-Associated TP53 Hotspot Mutations Beyond the crRNA Seed Region. CRISPR J. 6 , 127–139 (2023). Ooi, K. H. et al. An engineered CRISPR-Cas12a variant and DNA-RNA hybrid guides enable robust and rapid COVID-19 testing. Nat. Commun. 12 , 1739 (2021). Chen, Y., Mei, Y., Zhao, X. & Jiang, X. Reagents-Loaded, Automated Assay that Integrates Recombinase-Aided Amplification and Cas12a Nucleic Acid Detection for a Point-of-Care Test. Anal. Chem. 92 , 14846–14852 (2020). Lu, S. et al. Fast and sensitive detection of SARS-CoV-2 RNA using suboptimal protospacer adjacent motifs for Cas12a. Nat. Biomed. Eng. 6 , 286–297 (2022). Hu, M. et al. Photocontrolled crRNA activation enables robust CRISPR-Cas12a diagnostics. Proc. Natl. Acad. Sci. 119 , e2202034119 (2022). Chen, Y. et al. Photoactivatable CRISPR/Cas12a Strategy for One-Pot DETECTR Molecular Diagnosis. Anal. Chem. 94 , 9724–9731 (2022). Cao, Y. et al. CoLAMP: CRISPR-based one-pot loop-mediated isothermal amplification enables at-home diagnosis of SARS-CoV-2 RNA with nearly eliminated contamination utilizing amplicons depletion strategy. Biosens. Bioelectron. 236 , 115402 (2023). Cao, Y. et al. Benchtop to at-home test: Amplicon-depleted CRISPR-regulated loop mediated amplification at skin-temperature for viral load monitoring. Biosens. Bioelectron. 267 , 116866 (2025). Wu, X. et al. RatioCRISPR: A ratiometric biochip based on CRISPR/Cas12a for automated and multiplexed detection of heteroplasmic SNPs in mitochondrial DNA. Biosens. Bioelectron. 241 , 115676 (2023). Chen, Y. et al. Dual-CRISPR/Cas12a-Assisted RT-RAA for Ultrasensitive SARS-CoV-2 Detection on Automated Centrifugal Microfluidics. Anal. Chem. 94 , 9603–9609 (2022). Ai, J. et al. PAM-Independent CRISPR-Cas12a System for Specific Assays of Single Nucleotide Variants. JACS Au 5 , 1392–1401 (2025). Dai, J. et al. Exploiting the Potential of Spherical PAM Antenna for Enhanced CRISPR-Cas12a: A Paradigm Shift toward a Universal Amplification-Free Nucleic Acid Test Platform. Anal. Chem. 97 , 1236–1245 (2025). Huang, Z. et al. Thermally programmed one-pot CRISPR assay for on-site pandemic surveillance. Nat. Commun. 16 , 10286 (2025). Moon, J. & Liu, C. Asymmetric CRISPR enabling cascade signal amplification for nucleic acid detection by competitive crRNA. Nat. Commun. 14 , 7504 (2023). Mao, X. et al. Iterative crRNA design and a PAM-free strategy enabled an ultra-specific RPA-CRISPR/Cas12a detection platform. Commun. Biol. 7 , 1454 (2024). Kohabir, K. A. V. et al. Synthetic mismatches enable specific CRISPR-Cas12a-based detection of genome-wide SNVs tracked by ARTEMIS. Cell Rep. Methods 4 , 100912 (2024). Kohabir, K. A. V., Sistermans, E. A. & Wolthuis, R. M. F. Recent advances in CRISPR-based single-nucleotide fidelity diagnostics. Commun. Med. 5 , 252 (2025). Karvelis, T. et al. PAM recognition by miniature CRISPR–Cas12f nucleases triggers programmable double-stranded DNA target cleavage. Nucleic Acids Res. 48 , 5016–5023 (2020). Chen, Y., Mei, Y. & Jiang, X. Universal and high-fidelity DNA single nucleotide polymorphism detection based on a CRISPR/Cas12a biochip. Chem. Sci. 12 , 4455–4462 (2021). He, Y., Shao, S. & Chen, J. High-Fidelity Identification of Single Nucleotide Polymorphism by Type V CRISPR Systems. ACS Sens. 8 , 4478–4483 (2023). Chang, T., Yen, T., Wei, C., Hsiao, T. & Chen, I. Impacts of ADH1B rs1229984 and ALDH2 rs671 polymorphisms on risks of alcohol‐related disorder and cancer. Cancer Med. 12 , 747–759 (2023). Catarino, C. B. et al. Characterization of a Leber’s hereditary optic neuropathy (LHON) family harboring two primary LHON mutations m.11778G > A and m.14484T > C of the mitochondrial DNA. Mitochondrion 36 , 15–20 (2017). Tong, X. et al. Fast and sensitive CRISPR detection by minimized interference of target amplification. Nat. Chem. Biol. 20 , 885–893 (2024). Zhou, H. et al. Phase Transition of Wax Enabling CRISPR Diagnostics for Automatic At‐Home Testing of Multiple Sexually Transmitted Infection Pathogens. Small 21 , 2407931 (2025). Cheng, Z.-H. et al. Tunable control of Cas12 activity promotes universal and fast one-pot nucleic acid detection. Nat. Commun. 16 , 1166 (2025). Lin, M. et al. Glycerol Additive Boosts 100-fold Sensitivity Enhancement for One-Pot RPA-CRISPR/Cas12a Assay. Anal. Chem. 94 , 8277–8284 (2022). Bustin, S. A. et al. The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. Clin. Chem. 55 , 611–622 (2009). Crannell, Z. A., Rohrman, B. & Richards-Kortum, R. Development of a Quantitative Recombinase Polymerase Amplification Assay with an Internal Positive Control. J. Vis. Exp. 52620 (2015). Lobato, I. M. & O’Sullivan, C. K. Recombinase polymerase amplification: Basics, applications and recent advances. TrAC Trends Anal. Chem. 98 , 19–35 (2018). You, Y. Design of LNA probes that improve mismatch discrimination. Nucleic Acids Res. 34 , e60–e60 (2006). Kotlarz, K., Mielczarek, M., Biecek, P., Guldbrandtsen, B. & Szyda, J. Exploring the impact of sequence context on errors in SNP genotype calling with whole genome sequencing data using AI-based autoencoder approach. NAR Genomics Bioinforma. 6 , lqae131 (2024). De Oliveira Martins, E. & Weber, G. Nearest-neighbour parametrization of DNA single, double and triple mismatches at low sodium concentration. Biophys. Chem. 306 , 107156 (2024). Listgarten, J. et al. Prediction of off-target activities for the end-to-end design of CRISPR guide RNAs. Nat. Biomed. Eng. 2 , 38–47 (2018). Wessels, H.-H. et al. Prediction of on-target and off-target activity of CRISPR–Cas13d guide RNAs using deep learning. Nat. Biotechnol. 42 , 628–637 (2024). Camacho, D. M., Collins, K. M., Powers, R. K., Costello, J. C. & Collins, J. J. Next-Generation Machine Learning for Biological Networks. Cell 173 , 1581–1592 (2018). Liu, Z., Shao, W.-Q., Sun, Y. & Sun, B.-H. Scaling law of the one-direction flow characteristics of symmetric Tesla valve. Eng. Appl. Comput. Fluid Mech. 16 , 441–452 (2022). Additional Declarations There is NO Competing Interest. Supplementary Files SupplementaryMaterialsNCTEMPOHsing.docx Supplementary Materials Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8802024","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":590995567,"identity":"dccc3f71-fc8e-4fb4-8fdb-4110f960cc73","order_by":0,"name":"I-Ming Hsing","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAu0lEQVRIiWNgGAWjYBACCQYeBoYEBgY5EIeZ4QAJWoxBGkjQAgSJDURrkWzvPfzh4Y7a9A3nzx9gLjhDhBZpnnNpEolnjuduuJHMwDzjBhFa5CRyzBgS244BtQAdxvOBGC3yb4w/ALWkG5w/TKQWaQkeA4nEtpoEgwNAh/EQ4zDJnhwzoJYDhjNvJBsc5iHG+xLHzxh//NlWJ893/uDDxzzHiNACBYfB5AHiNTAw1JGieBSMglEwCkYaAAAcfTih7ORUFgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-3326-3021","institution":"The Hong Kong University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"I-Ming","middleName":"","lastName":"Hsing","suffix":""},{"id":590995568,"identity":"acc6857e-c860-4e8f-871a-17779452ca00","order_by":1,"name":"Xiaolong Wu","email":"","orcid":"","institution":"The Hong Kong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Xiaolong","middleName":"","lastName":"Wu","suffix":""},{"id":590995569,"identity":"f900ef4c-3382-4d79-9352-43c5b6d11bc6","order_by":2,"name":"Yanan Li","email":"","orcid":"","institution":"Hong Kong University of Science \u0026 Technology","correspondingAuthor":false,"prefix":"","firstName":"Yanan","middleName":"","lastName":"Li","suffix":""},{"id":590995570,"identity":"87f57bee-af96-405a-a629-49e46b0a7be6","order_by":3,"name":"Yumeng Cao","email":"","orcid":"https://orcid.org/0000-0002-1729-9614","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yumeng","middleName":"","lastName":"Cao","suffix":""},{"id":590995571,"identity":"f9ad5f99-1f03-4e0a-afd9-477098f018ec","order_by":4,"name":"Zibin Zhao","email":"","orcid":"https://orcid.org/0000-0002-3121-9131","institution":"The Hong Kong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Zibin","middleName":"","lastName":"Zhao","suffix":""},{"id":590995572,"identity":"b0645774-b599-40f6-98c4-336999a2e761","order_by":5,"name":"Hongyu Lu","email":"","orcid":"","institution":"The Hong Kong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Hongyu","middleName":"","lastName":"Lu","suffix":""},{"id":590995573,"identity":"39898a6a-2a2f-415e-8181-298a32b6c715","order_by":6,"name":"Shaochong Liang","email":"","orcid":"","institution":"The Hong Kong University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Shaochong","middleName":"","lastName":"Liang","suffix":""}],"badges":[],"createdAt":"2026-02-06 03:05:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8802024/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8802024/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102818562,"identity":"9feb594a-3ed7-4a20-a684-bde502595060","added_by":"auto","created_at":"2026-02-17 06:52:15","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":368232,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThermodynamically encoded reaction ordering enables autonomous one-pot CRISPR coordination.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Thermodynamic design of two competing primer-binding pathways. A fully complementary forward primer (FP) readily hybridizes to the target and drives rapid RPA, producing pristine amplicons. In contrast, the PAM-containing primer carries mismatches (FPP), creating an unfavorable hybridization energy and preventing early engagement. Numbers of mismatches in FPP can be designed via ΔΔG\u003csub\u003epenalty, FPP-Template\u003c/sub\u003e = RT ln ([Template \u003csub\u003ePristine amplicons\u003c/sub\u003e] / [Template \u003csub\u003eThreshold\u003c/sub\u003e]). \u003cstrong\u003e(B)\u003c/strong\u003e Conceptual free-energy landscape illustrating chemical potential-driven pathway activation in the TEMPO reactor. Pristine RPA initiates immediately along a low-barrier pathway due to favorable primer–template hybridization, producing PAM-free amplicons that accumulate without engaging CRISPR. The PAM-introducing amplification branch is initially inaccessible because an intentional mismatch imposes a higher free-energy barrier. As pristine amplicons accumulate, the increasing chemical potential of the template pool provides a concentration-dependent driving force that progressively lowers the effective barrier, enabling PAM-containing amplification. The resulting mixed amplicon population activates Cas12a via PAM-dependent \u003cem\u003ecis\u003c/em\u003e-cleavage, followed by \u003cem\u003etrans\u003c/em\u003e-cleavage of the reporter for signal generation. \u003cstrong\u003e(C and D)\u003c/strong\u003e TEMPO validation using DNA target 1 (derived from HIV genome, 1 fM) and DNA target 2 (derived from Human genome, 1 fM). \u003cstrong\u003e(E) \u003c/strong\u003eSchematic of on-site TEMPO detection using a single-step microfluidic chip. Following sample loading, the one-pot amplification–CRISPR reaction proceeds autonomously, and fluorescence signals are captured and interpreted by a smartphone-based readout.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/e6ca52f1d0b326dc24fd9197.png"},{"id":102818563,"identity":"185676c0-06d0-42b6-a4ba-b9af7abdfbc6","added_by":"auto","created_at":"2026-02-17 06:52:15","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":525146,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eProgrammability of FPP primers via hybridization strength and PAM engineering\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Schematic illustrating modulation of primer–template hybridization thermodynamics by varying the length of the FPP primer, thereby tuning the rate of PAM-containing secondary amplification. Primer length serves as a programmable handle to adjust −ΔG\u003csub\u003eFPP-Template\u003c/sub\u003e. All reactions were performed with a fixed template input of 1 fM. \u003cstrong\u003e(B and C)\u003c/strong\u003e Kinetic profiles of TEMPO reactions programmed by FPP primer length. “Full” denotes the full-length FPP primer, and “−n” indicates truncation by n nucleotides. Panel B and panel C correspond to DNA targets derived from the HIV and human genomes, respectively. \u003cstrong\u003e(D)\u003c/strong\u003e Schematic illustrating FPP primers encoding different PAM sequences modulate CRISPR signal output. \u003cstrong\u003e(E and F)\u003c/strong\u003e Kinetic profiles of TEMPO reactions programmed with different PAM variants. Panel E shows reactions evaluated using DNA target 1 derived from the HIV genome (PAMs: TTG, TTT, TAT, TAG, and AAG), whereas panel F shows reactions evaluated using DNA target 2 derived from the human genome (PAMs: TTT, ATT, TCT, ACT, and ATA). \u003cstrong\u003e(G)\u003c/strong\u003e Schematic of crRNA-overlap engineering, where FPP primer modification introduces an effective mismatch at the CRISPR recognition locus. (\u003cstrong\u003eH and I\u003c/strong\u003e) Kinetic profiles of TEMPO reactions programmed with effective mismatches at defined positions downstream of the PAM. MUT-n indicates a single-nucleotide mismatch introduced n bases after the PAM via FPP–crRNA overlap engineering. Panel H and panel I correspond to DNA targets derived from the HIV and human genomes, respectively.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/091634924aef0ab0cfd9f2c0.png"},{"id":102818564,"identity":"2976da60-a0ad-459a-b017-da8e1e217c97","added_by":"auto","created_at":"2026-02-17 06:52:15","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":359157,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDNA engineering enables broad-spectrum and highly specific SNP detection.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(\u003cstrong\u003eA\u003c/strong\u003e) Relocate targeting site via primer engineering. Diagram of the two forward-selective primers (FPP1 and FPP2) positioned at different sites on the target sequence. \u003cstrong\u003e(B)\u003c/strong\u003eAmplification curves for the rs199476104 target (1 fM input) with/without FPP1. The pristine forward primer initiates amplification earlier, whereas FPP1 exhibits a delayed onset due to reduced binding affinity. \u003cstrong\u003e(C)\u003c/strong\u003eAmplification curves for the rs199476104 target (1 fM input) with/without FPP2, showing a similarly delayed and sequential onset of amplification as observed with FPP1. (\u003cstrong\u003eD\u003c/strong\u003e) Design strategy for PAM switching via FPP mismatch engineering. Mismatches within the FPP-binding region reassign the PAM sequence, enabling allele-dependent Cas12a activation. (\u003cstrong\u003eE and F\u003c/strong\u003e) Cas12a cleavage results for canonical PAMs (TTT) and suboptimal PAM variants (ATT) using the rs671 target (1 fM input). (\u003cstrong\u003eG\u003c/strong\u003e) Design strategy using a single-base insertion to enhance allele discrimination. The insertion is introduced downstream of the suboptimal PAM-encoding position within FPP, deliberately shifting the crRNA–target register to create a defined single-base mismatch in the seed region. \u003cstrong\u003e(H and I) \u003c/strong\u003eCas12a cleavage results for the rs1229984 target (1 fM input) comparing designs without and with an insertion mutation.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/d9f8deb74367fddf242ef983.png"},{"id":102962952,"identity":"ba81b4b7-1ad8-46c0-a778-44a8b0094c46","added_by":"auto","created_at":"2026-02-19 04:12:19","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":8495769,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAnalytical performance of TEMPO for pathogen detection.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Schematic workflow for HIV detection using TEMPO. HIV mock samples were prepared and subjected to RNA extraction, followed by isothermal one-pot CRISPR detection or RT-qPCR analysis as a reference method. \u003cstrong\u003e(B)\u003c/strong\u003eSensitivity of TEMPO for detecting different concentrations of HIV DNA (10000 aM,1000 aM, 100 aM, 10 aM,1 aM,0.1 aM, and negative control). Strong fluorescence signals were observed down to 1 aM concentration, with clear visual fluorescence under blue light. NTC, non-template control reaction. Error bars represent the means ± standard deviation (s.d.) from replicates. The unpaired two-tailed t-test was used to analyze the statistical significance. \u003cstrong\u003e(C)\u003c/strong\u003eFluorescence output of TEMPO across extracted HIV RNA mock samples with varying RT–qPCR Ct values. Signal intensity decreases progressively with increasing Ct, consistent with the reduced RNA in put in higher Ct samples.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/0029999cc30bd3e674b8cfa2.png"},{"id":102818567,"identity":"6c7590a2-dc7f-47ef-8fdb-46d6a806c6d1","added_by":"auto","created_at":"2026-02-17 06:52:15","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":10655906,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRapid genotyping of clinically relevant SNPs using TEMPO.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e Workflow of the SNP genotyping using TEMPO. Swab samples are collected and lysed at room temperature for approximately 5 min. The crude lysate is applied directly to the SNP Chip, where a one pot RPA–Cas12a reaction is performed at 37 °C for 30 min. Fluorescence signals are then recorded and interpreted using a smartphone interface. \u003cstrong\u003e(B)\u003c/strong\u003eRepresentative amplification trajectories illustrating allele-dependent kinetics for rs199476104. Distinct curves corresponding to the wild-type, heterozygous, and mutant genotypes are generated within the one-pot reaction using a 1 fM standard template input. \u003cstrong\u003e(C)\u003c/strong\u003e Heat map summarizing TEMPO genotyping outcomes for 20 clinical samples across three SNP loci (rs1229984, rs671, and rs199476104). The discrimination factor (DF) encoded in each cell quantitatively captures the relative fluorescence contrast between SNP and wild-type alleles, enabling unambiguous genotype assignment with consistent performance across all samples and loci. Genotype calls were assigned based on DF thresholds, with DF values of 0–0.5 classified as wild type, 0.5–2.0 as heterozygous, and 2.0–10 as SNP. \u003cstrong\u003e(D)\u003c/strong\u003e Cleavage rate measured from Sample 20 as a representative example. The observed cleavage patterns match the expected allelic state and show strong agreement with Sanger sequencing traces provided above each bar. \u003cstrong\u003e(E)\u003c/strong\u003eImages of the SNP chip shown with a Hong Kong two-dollar coin for size reference, including the assembled microfluidic chip and representative endpoint fluorescence patterns from sample 20 after a 30 min reaction, highlighting allele-dependent signal differences. \u003cstrong\u003e(F)\u003c/strong\u003e Smartphone interface used for real-time data acquisition, visualization and automated SNP reporting.\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/6a5cadf61f8cda289295f2ed.png"},{"id":103056337,"identity":"3f21d93c-f775-482e-a1bc-219dfda946a0","added_by":"auto","created_at":"2026-02-20 09:07:03","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":21687498,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/e0bc8d11-22e4-484d-a171-7db62cf9127e.pdf"},{"id":102818566,"identity":"e04a8e9d-5938-4231-b0d3-497eaf1f7fea","added_by":"auto","created_at":"2026-02-17 06:52:15","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":7846823,"visible":true,"origin":"","legend":"Supplementary Materials","description":"","filename":"SupplementaryMaterialsNCTEMPOHsing.docx","url":"https://assets-eu.researchsquare.com/files/rs-8802024/v1/310093104f8b966f6bcb20f0.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Thermodynamically programmed one-pot CRISPR platform for point-of-care SNP genotyping","fulltext":[{"header":"Main","content":"\u003cp\u003eFully integrated molecular diagnostics aim to combine nucleic acid amplification, molecular recognition and signal transduction within a single reaction environment\u003csup\u003e1,2\u003c/sup\u003e. Achieving such integration, however, requires multiple biochemical processes to occur in a defined temporal order despite sharing the same reactor and molecular intermediates\u003csup\u003e3\u003c/sup\u003e. When reaction order is externally imposed, through physical separation, staged operation or manual intervention, robust performance can be maintained\u003csup\u003e4\u0026ndash;6\u003c/sup\u003e. In contrast, when competing reactions proceed simultaneously without intrinsic ordering, system-level instability emerges, undermining sensitivity, reproducibility and predictability\u003csup\u003e7,8\u003c/sup\u003e. Designing molecular systems that autonomously enforce reaction order therefore represents a central but unresolved challenge in integrated diagnostics.\u003c/p\u003e \u003cp\u003eThis challenge is particularly acute in one-pot CRISPR-based diagnostics\u003csup\u003e3,9,10\u003c/sup\u003e. CRISPR\u0026ndash;Cas12a has enabled highly sensitive nucleic acid detection by coupling sequence-specific target recognition to collateral nuclease activity\u003csup\u003e11,12\u003c/sup\u003e. When combined with isothermal amplification, such systems promise rapid and portable diagnostics\u003csup\u003e13\u0026ndash;15\u003c/sup\u003e. However, amplification and CRISPR activation form a tightly coupled reaction network that competes for the same target molecules\u003csup\u003e16\u003c/sup\u003e. Amplification requires sustained accumulation of target DNA, whereas CRISPR activation irreversibly consumes those targets through cis- and trans-cleavage. In a one-pot format, this intrinsic conflict destabilizes reaction kinetics, leading to premature target depletion, suppressed amplification and reduced sensitivity\u003csup\u003e10,16,17\u003c/sup\u003e. These limitations are amplified in applications requiring high specificity, such as single-nucleotide polymorphism (SNP) genotyping, where both stringent sequence discrimination and stable reaction kinetics are required\u003csup\u003e18\u0026ndash;20\u003c/sup\u003e. Many clinically relevant SNPs underpin pathogen drug resistance, cancer heterogeneity and pharmacogenomic decision-making, yet their reliable detection remains difficult in one-pot CRISPR systems\u003csup\u003e18,19,21\u003c/sup\u003e. In this context, minor kinetic instability or premature target consumption can obscure single-nucleotide differences, rendering high-resolution genotyping unreliable.\u003c/p\u003e \u003cp\u003eExisting strategies to address this interference have focused on mitigating, rather than resolving, the underlying lack of reaction order. Approaches such as empirical tuning of reagent concentrations\u003csup\u003e22,23\u003c/sup\u003e, weakening protospacer adjacent motif (PAM) recognition\u003csup\u003e24\u003c/sup\u003e, engineering crRNA\u003csup\u003e25,26\u003c/sup\u003e or primer\u003csup\u003e27,28\u003c/sup\u003e, or introducing physical or temporal separation\u003csup\u003e29,30\u003c/sup\u003e can partially alleviate competition between amplification and CRISPR activity. However, these solutions rely on external control or context-specific parameter balancing and therefore cannot guarantee robust performance across different targets, sequence contexts or operating conditions. Moreover, PAM dependence fundamentally constrains genomic addressability: fewer than a minority of clinically relevant SNP loci are proximal to canonical Cas12a PAM sequences\u003csup\u003e31,32\u003c/sup\u003e, rendering most variants inaccessible to conventional CRISPR diagnostics. As a result, current one-pot CRISPR systems face a persistent trade-off between workflow simplicity, target scope and single-nucleotide specificity.\u003c/p\u003e \u003cp\u003eHere we propose that reaction order in one-pot CRISPR diagnostics can be molecularly programmed rather than externally imposed, enabling autonomous and predictable coordination of amplification and CRISPR transduction within a single reactor. We introduce thermodynamic encoding as a design principle in which reaction sequencing is embedded directly into DNA primer energetics. By engineering a defined free-energy asymmetry between competing amplification primers, the system autonomously evolves through kinetically ordered amplification regimes, prioritizing target accumulation before activating CRISPR-mediated signal transduction. This intrinsic, threshold-gated coordination decouples amplification from CRISPR activity within a single, unsegmented reactor, without the need for physical separation or external triggers. Importantly, this architecture relocates PAM dependence from native genomic targets to primer-encoded design, enabling access to loci that are otherwise undetectable while preserving single-nucleotide discrimination.\u003c/p\u003e \u003cp\u003eBuilding on this principle, we develop Thermodynamically Encoded Molecular Programming for One-pot diagnostics (TEMPO), a programmable CRISPR platform that integrates isothermal amplification, PAM design and kinetic control into a single molecular framework. Using a quantitative ordinary differential equation model to guide primer design, TEMPO achieves predictable reaction ordering, attomolar sensitivity and robust SNP discrimination in a fully integrated one-pot format. We validate this approach using simulated HIV samples and clinically relevant human SNPs lacking proximal PAM sites and further demonstrate its applicability through integration into a single-step microfluidic diagnostic chip. Together, these results establish thermodynamic encoding as a generalizable design paradigm for coordinating competing biochemical processes within shared reactors, with implications extending beyond CRISPR diagnostics to integrated molecular systems that require autonomous reaction ordering.\u003c/p\u003e"},{"header":"Result","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eThermodynamically encoded reaction ordering enables autonomous one-pot CRISPR coordination\u003c/h2\u003e \u003cp\u003ePrevious efforts have attempted to temporally separate amplification and CRISPR activation using external gating elements such as photocaged crRNAs\u003csup\u003e25\u003c/sup\u003e or temperature switch\u003csup\u003e33\u003c/sup\u003e. While effective, these systems lack an intrinsic mechanism to prioritize amplification over CRISPR activation, resulting in uncontrolled competition for shared target molecules. Under such conditions, premature CRISPR engagement inevitably suppresses amplification efficiency and destabilizes signal generation. We reasoned that enforcing reaction order requires not parameter tuning or weakened CRISPR activity, but the introduction of an explicit energetic hierarchy between competing amplification pathways, such that reaction sequencing emerges as an inevitable consequence of molecular thermodynamics.\u003c/p\u003e \u003cp\u003eTo achieve this, we introduced two forward primers with a defined thermodynamic asymmetry (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA): a fully complementary primer (FP, ΔG\u003csub\u003eFP\u0026minus;Template\u003c/sub\u003e\u0026asymp;-37.20 kcal/mol) that drives efficient amplification of PAM-free amplicons, and a deliberately mismatched primer (FPP, ΔG\u003csub\u003eFPP\u0026minus;Template\u003c/sub\u003e\u0026asymp;-30.71 kcal/mol) that encodes a PAM but incurs a higher hybridization free-energy penalty. This energetic imbalance necessarily partitions amplification into two kinetically distinct regimes. At low template abundance, only the high-affinity FP pathway is thermodynamically accessible, leading to rapid accumulation of pristine amplicons that cannot activate Cas12a (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). As amplification proceeds and template concentration increases, according to the relationship ΔG\u003csub\u003eeff\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;ΔG\u0026deg; \u0026minus; RT ln([Template])\u0026thinsp;+\u0026thinsp;C, the effective free-energy barrier for FPP binding is progressively reduced, enabling a second amplification regime that installs PAM-containing amplicons. CRISPR activation therefore emerges as a threshold behavior rather than an externally triggered event.\u003c/p\u003e \u003cp\u003eOur results reveal that a defined free-energy difference between the two forward primers gives rise to two kinetically ordered amplification phases. Single-primer reactions (FP/BP or FPP/BP) each produced a single amplicon of the expected size, whereas reactions containing both forward primers generated two distinct products corresponding to the pristine and FPP-extended amplicons (\u003cb\u003eFig. S2A\u003c/b\u003e). Sequencing verified that the FPP primer introduces the engineered PAM without affecting the original product (\u003cb\u003eFig. S2B\u003c/b\u003e). Real-time kinetics further indicated that the higher-affinity FP primer drives the early phase of amplification, while the FPP primer, penalized by reduced binding strength, engages more slowly. At high template input, the rate difference narrows, consistent with mass-action effects overcoming the thermodynamic penalty (\u003cb\u003eFig. S3)\u003c/b\u003e. Additional assays verified that only PAM-containing dsDNA efficiently activates Cas12a (\u003cb\u003eFig. S4\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eConsistent with this mechanism, reactions containing only FP produced robust amplification without detectable CRISPR signal, whereas reactions driven solely by FPP exhibited weak and unstable fluorescence due to simultaneous amplification and target consumption. In contrast, reactions containing both primers reproducibly displayed delayed but strong CRISPR activation, reflecting autonomous reaction ordering encoded by primer thermodynamics rather than empirical parameter balancing (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC\u0026ndash;D). Importantly, ongoing FP-driven amplification continuously replenished templates, preventing target depletion even after CRISPR activation and yielding stable one-pot kinetics. Based on this principle, we establish Thermodynamically Encoded Molecular Programming for One-pot diagnostics (TEMPO) and integrate it into a microfluidic platform (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE) for rapid pathogen detection and SNP analysis.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e(A)\u003c/b\u003e Thermodynamic design of two competing primer-binding pathways. A fully complementary forward primer (FP) readily hybridizes to the target and drives rapid RPA, producing pristine amplicons. In contrast, the PAM-containing primer carries mismatches (FPP), creating an unfavorable hybridization energy and preventing early engagement. Numbers of mismatches in FPP can be designed via ΔΔG\u003csub\u003epenalty, FPP\u0026minus;Template\u003c/sub\u003e = RT ln ([Template \u003csub\u003ePristine amplicons\u003c/sub\u003e] / [Template \u003csub\u003eThreshold\u003c/sub\u003e]). \u003cb\u003e(B)\u003c/b\u003e Conceptual free-energy landscape illustrating chemical potential-driven pathway activation in the TEMPO reactor. Pristine RPA initiates immediately along a low-barrier pathway due to favorable primer\u0026ndash;template hybridization, producing PAM-free amplicons that accumulate without engaging CRISPR. The PAM-introducing amplification branch is initially inaccessible because an intentional mismatch imposes a higher free-energy barrier. As pristine amplicons accumulate, the increasing chemical potential of the template pool provides a concentration-dependent driving force that progressively lowers the effective barrier, enabling PAM-containing amplification. The resulting mixed amplicon population activates Cas12a via PAM-dependent \u003cem\u003ecis\u003c/em\u003e-cleavage, followed by \u003cem\u003etrans\u003c/em\u003e-cleavage of the reporter for signal generation. \u003cb\u003e(C and D)\u003c/b\u003e TEMPO validation using DNA target 1 (derived from HIV genome, 1 fM) and DNA target 2 (derived from Human genome, 1 fM). \u003cb\u003e(E)\u003c/b\u003e Schematic of on-site TEMPO detection using a single-step microfluidic chip. Following sample loading, the one-pot amplification\u0026ndash;CRISPR reaction proceeds autonomously, and fluorescence signals are captured and interpreted by a smartphone-based readout.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eQuantitative modeling establishes reaction ordering as a predictable design outcome\u003c/h3\u003e\n\u003cp\u003eTo formalize this behavior and distinguish intrinsic reaction ordering from empirical tuning, we constructed a quantitative reaction-network model describing primer competition, amplification and PAM-dependent Cas12a activation (Supplementary Note 1). The model comprises coupled ordinary differential equations that encode primer hybridization kinetics, strand extension and CRISPR-mediated cleavage as mass-action and Michaelis\u0026ndash;Menten processes.\u003c/p\u003e \u003cp\u003e1. Primer competition and thermodynamic gating\u003c/p\u003e \u003cp\u003eWe modeled primer binding as bimolecular association processes. Central to the engineered thermodynamic switch is the mismatch-bearing primer FPP, whose reduced hybridization free energy is encoded directly into slower kinetic rate constants:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\frac{d[\\text{F}\\text{P}\\text{P}\\cdot\\:\\text{D}]}{dt}={k}_{6}\\left[\\text{F}\\text{P}\\text{P}\\right]\\left[\\text{D}\\right]-{k}_{6}^{\\text{ext}}[\\text{F}\\text{P}\\text{P}\\cdot\\:\\text{D}]$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ein contrast to fully complementary FP:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:\\frac{d[\\text{F}\\text{P}\\cdot\\:\\text{D}]}{dt}={k}_{1}\\left[\\text{F}\\text{P}\\right]\\left[\\text{D}\\right]-{k}_{1}^{\\text{ext}}[\\text{F}\\text{P}\\cdot\\:\\text{D}]$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewith \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}_{6}\u0026lt;{k}_{1}\\)\u003c/span\u003e\u003c/span\u003ecapturing the thermodynamic penalty imposed by the designed mismatch. This asymmetry produces two kinetically distinct amplification channels: a high-affinity pristine pathway leading to the native amplicon \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:D\\)\u003c/span\u003e\u003c/span\u003e, and a delayed mismatch pathway that generates the PAM-containing amplicon \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{D}_{2}\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e2. Sequential amplification and PAM emergence\u003c/p\u003e \u003cp\u003eThe reactor evolves according to mass-action coupling between strand-displacement extension and template regeneration:\u003c/p\u003e \u003cp\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{d\\left[D\\right]}{dt}={k}_{3}\\left[\\text{B}\\text{P}\\right]\\left[\\text{S}\\text{T}\\right]+{k}_{4}\\left[\\text{F}\\text{P}\\right]\\left[\\text{A}\\text{T}\\right]+{k}_{5}\\left[\\text{S}\\text{T}\\right]\\left[\\text{A}\\text{T}\\right]-{k}_{13}\\left[\\text{C}\\right]\\left[D\\right]\\)\u003c/span\u003e \u003c/span\u003e \u003cspan class=\"InlineEquation\"\u003e \u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{d\\left[{D}_{2}\\right]}{dt}={k}_{8}\\left[\\text{B}\\text{P}\\right]\\left[{\\text{S}\\text{T}}_{2}\\right]+{k}_{11}\\left[\\text{F}\\text{P}\\text{P}\\right]\\left[{\\text{A}\\text{T}}_{2}\\right]+{k}_{12}\\left[{\\text{S}\\text{T}}_{2}\\right]\\left[{\\text{A}\\text{T}}_{2}\\right]-{k}_{14}\\left[\\text{C}\\right]\\left[{D}_{2}\\right]\\)\u003c/span\u003e \u003c/span\u003e \u003c/p\u003e \u003cp\u003ewhere the emergence of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{D}_{2}\\)\u003c/span\u003e\u003c/span\u003eis controlled by the mismatch-driven delay in FPP-mediated extension. These expressions quantitatively reveal that PAM introduction is not an external trigger but a threshold phenomenon emergent from the kinetic imbalance between the two amplification tracks.\u003c/p\u003e \u003cp\u003e3. PAM-dependent CRISPR transduction as a catalytic nonlinear activation\u003c/p\u003e \u003cp\u003eCas12a activation is represented using two parallel Michaelis\u0026ndash;Menten channels with dramatically different catalytic efficiencies:\u003cdiv id=\"Equc\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equc\" name=\"EquationSource\"\u003e\n$$\\:{v}_{1}=\\frac{{k}_{\\text{c}\\text{a}\\text{t}1}\\left[\\text{C}\\text{D}\\right]\\left[R\\right]}{{K}_{m1}+\\left[R\\right]},{v}_{2}=\\frac{{k}_{\\text{c}\\text{a}\\text{t}2}\\left[{\\text{C}\\text{D}}_{2}\\right]\\left[R\\right]}{{K}_{m2}+\\left[R\\right]}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewith \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{k}_{\\text{c}\\text{a}\\text{t},2}\\gg\\:{k}_{\\text{c}\\text{a}\\text{t},1}\\)\u003c/span\u003e\u003c/span\u003e. This structure creates an activation nonlinearity normally seen in engineered chemical reactors where a slow path primes the system for abrupt activation once a specific intermediate (here, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{D}_{2}\\)\u003c/span\u003e\u003c/span\u003e) reaches a critical concentration.\u003c/p\u003e \u003cp\u003e4. Model-experiment agreement validates a predictable sequential reactor\u003c/p\u003e \u003cp\u003eIn this framework, the mismatch-bearing FPP primer is represented by a reduced effective association rate constant relative to FP, directly reflecting its higher hybridization free-energy. This asymmetry necessarily generates two amplification pathways with distinct kinetic accessibility. Model simulations predict three defining features of thermodynamically encoded reaction ordering: (i) suppression of CRISPR activation during early exponential amplification, (ii) delayed emergence of PAM-containing amplicons once a threshold template concentration is reached, and (iii) a sharp, nonlinear onset of trans-cleavage activity following PAM installation.\u003c/p\u003e \u003cp\u003eExperimental kinetics closely matched these predictions across multiple targets and primer designs (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC \u003cb\u003eto D; Fig. S5\u003c/b\u003e). The observed delay in CRISPR activation, its threshold-like onset, and the sustained amplification after activation all emerged without invoking external triggers or time-dependent parameter changes. Importantly, altering primer free-energy penalties or PAM identity shifted the predicted threshold and activation kinetics in a manner quantitatively consistent with experimental observations, confirming that primer thermodynamics act as a tunable control variable rather than an empirical adjustment.\u003c/p\u003e\n\u003ch3\u003eProgrammable DNA engineering operates as a multi-axis kinetic control element\u003c/h3\u003e\n\u003cp\u003eBuilding on the model, we next examined whether primer engineering could systematically shape reaction dynamics. Conventional RPA\u0026ndash;CRISPR-Cas12a coupling, primarily regulates reaction kinetics by modifying crRNA\u0026ndash;target interactions\u003csup\u003e34\u0026ndash;36\u003c/sup\u003e. RPA primers are seldom altered, as sequence changes frequently compromise amplification efficiency. In contrast, in TEMPO the amplification efficiency is primarily contributed by the pristine primer FP, while FPP serves as a molecular bridge linking RPA output to Cas12a activation. This architecture means that programming FPP allows simultaneous tuning of both RPA progression and CRISPR kinetics, offering a capability inaccessible to traditional approaches.\u003c/p\u003e \u003cp\u003eWe infer that FPP exposes three orthogonal axes of design flexibility: (i) Varying FPP length modulated hybridization free-energy and predictably shifted the timing of PAM-containing amplification, thereby tuning the delay between amplification and CRISPR activation. (ii) Independently, altering the encoded PAM sequence modulated Cas12a activation strength without substantially affecting amplification kinetics, decoupling signal transduction from target generation. (iii) Finally, modifying the FPP\u0026ndash;crRNA overlap region effectively introduced programmable mismatches at the recognition interface, providing an additional layer of kinetic control without rescreening crRNAs.\u003c/p\u003e \u003cp\u003eFirstly, we systematically adjusted FPP length while maintaining mismatch position constant (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Shorter FPP variants, with weaker binding affinity, slowed and reduced PAM-containing product formation, extending the temporal separation between amplification and Cas12a activation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB \u003cb\u003eto C\u003c/b\u003e). Conversely, increasing FPP length lowered the thermodynamic penalty, advancing the onset of CRISPR activation. These results indicate that mismatch-encoded ΔG operates as a kinetic valve, determining when the secondary amplification branch begins contributing to reactor output.\u003c/p\u003e \u003cp\u003eWe next tuned Cas12a activation independently of amplification by varying the PAM triplet within FPP (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD \u003cb\u003eto F\u003c/b\u003e). PAM identity modulated Cas12a activation strength without substantially affecting RPA kinetics. Canonical PAMs (TTT) induced rapid Cas12a activation driven by efficient \u003cem\u003ecis\u003c/em\u003e-cleavage, whereas sub-optimal PAMs (TTG, ATT) produced more gradual and sustained activation kinetics. This behavior likely arises because weaker PAMs mediate reduced Cas12a cleavage activity, thereby alleviating competition between CRISPR cleavage and PAM-containing amplification. In contrast, excessively weak PAMs resulted in minimal signal output, consistent with their intrinsically low Cas12a activation efficiency.\u003c/p\u003e \u003cp\u003eFinally, inspired by prior studies showing that crRNA mismatches modulate Cas12a kinetics\u003csup\u003e36\u003c/sup\u003e, but usually require extensive crRNA screening, we hypothesized that mutating the FPP\u0026ndash;crRNA overlap region would indirectly introduce mismatch into the crRNA recognition interface, providing an alternative regulatory handle. Indeed, kinetic profiles confirmed this effect (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG \u003cb\u003eto I\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eAcross these perturbations, reaction trajectories followed the same underlying sequence predicted by the model, differing only in timing and amplitude. This consistency demonstrates that thermodynamic encoding transforms primer design into a compact, multi-dimensional control space for shaping one-pot reaction dynamics, rather than a trial-and-error optimization problem.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e(A)\u003c/b\u003e Schematic illustrating modulation of primer\u0026ndash;template hybridization thermodynamics by varying the length of the FPP primer, thereby tuning the rate of PAM-containing secondary amplification. Primer length serves as a programmable handle to adjust\u0026thinsp;\u0026minus;\u0026thinsp;ΔG\u003csub\u003eFPP\u0026minus;Template\u003c/sub\u003e. All reactions were performed with a fixed template input of 1 fM. \u003cb\u003e(B and C)\u003c/b\u003e Kinetic profiles of TEMPO reactions programmed by FPP primer length. \u0026ldquo;Full\u0026rdquo; denotes the full-length FPP primer, and \u0026ldquo;\u0026minus;n\u0026rdquo; indicates truncation by n nucleotides. Panel B and panel C correspond to DNA targets derived from the HIV and human genomes, respectively. \u003cb\u003e(D)\u003c/b\u003e Schematic illustrating FPP primers encoding different PAM sequences modulate CRISPR signal output. \u003cb\u003e(E and F)\u003c/b\u003e Kinetic profiles of TEMPO reactions programmed with different PAM variants. Panel E shows reactions evaluated using DNA target 1 derived from the HIV genome (PAMs: TTG, TTT, TAT, TAG, and AAG), whereas panel F shows reactions evaluated using DNA target 2 derived from the human genome (PAMs: TTT, ATT, TCT, ACT, and ATA). \u003cb\u003e(G)\u003c/b\u003e Schematic of crRNA-overlap engineering, where FPP primer modification introduces an effective mismatch at the CRISPR recognition locus. (\u003cb\u003eH and I\u003c/b\u003e) Kinetic profiles of TEMPO reactions programmed with effective mismatches at defined positions downstream of the PAM. MUT-n indicates a single-nucleotide mismatch introduced n bases after the PAM via FPP\u0026ndash;crRNA overlap engineering. Panel H and panel I correspond to DNA targets derived from the HIV and human genomes, respectively.\u003c/p\u003e\n\u003ch3\u003eDNA engineering enables broad-spectrum and highly specific SNP detection\u003c/h3\u003e\n\u003cp\u003eCRISPR-based diagnostics have been widely applied to SNP genotyping\u003csup\u003e20,37\u003c/sup\u003e. However, practical implementation remains constrained by two longstanding challenges: (i) strict PAM requirements limit accessible loci\u003csup\u003e11,38\u003c/sup\u003e, and (ii) single-nucleotide discrimination is not consistently achieved, particularly in one-pot RPA\u0026ndash;CRISPR formats\u003csup\u003e39,40\u003c/sup\u003e. As a result, many clinically important SNPs remain undetectable.\u003c/p\u003e \u003cp\u003eGiven the programmable kinetic control demonstrated previously in TEMPO, we reasoned that FPP primer engineering could convert arbitrary SNPs, regardless of PAM availability, into high-fidelity, one-pot detectable targets. We first validated that TEMPO could redirect Cas12a targeting solely through FPP design (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA \u003cb\u003eto C)\u003c/b\u003e. Within a fixed RPA amplicon, repositioning the FPP primer selectively introduced PAMs at user-defined locations, enabling Cas12a activation at either of two sites. Additional sequence designs further support the generality of this strategy \u003cb\u003e(Fig. S6\u003c/b\u003e). Importantly, this targeting flexibility did not compromise sensitivity, as amplification driven by FPP remained intact. This property is particularly critical for SNP genotyping, where PAMs are often absent near the variant.\u003c/p\u003e \u003cp\u003ePrevious work has shown that crRNA or PAM engineering can enhance SNP discrimination\u003csup\u003e34\u003c/sup\u003e; here we demonstrate that DNA-level primer engineering achieves comparable, and in some cases superior, control. Building on earlier results showing that PAM identity modulates reaction kinetics, we evaluated whether PAM tuning could improve allelic contrast. Systematic substitution of the PAM triplet within FPP (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD) yielded tunable genotype separation: canonical PAMs (TTT) produced 2.8-4.5-fold discrimination between SNP and WT alleles (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE), whereas suboptimal PAM variants (ATT) further elevated contrast to 5.9-18.4-fold (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF). We attribute this enhancement to moderate reductions in Cas12a affinity that amplify thermodynamic discrimination between matched and mismatched dsDNA substrates.\u003c/p\u003e \u003cp\u003eHowever, certain SNPs remained poorly resolved even with optimized PAM weakening (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eH). We therefore introduced a second layer of control by programming the FPP\u0026ndash;crRNA overlap region, effectively inserting an additional mismatch at the recognition interface (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eG). This approach mirrors crRNA engineering but eliminates the need for extensive crRNA rescreening. A single additional mismatch increased wild-type/mutant separation up to 3.6-fold (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eI), converting previously non-resolvable SNPs into clearly distinguishable genotypes. Together, these results show that FPP primer design enables direct control of both CRISPR target accessibility and allele-level discrimination within a single amplification reaction.\u003c/p\u003e \u003cp\u003eIn summary, TEMPO expands the CRISPR detection landscape from PAM-constrained to universally addressable SNP sites, while substantially simplifying assay design. DNA primer programming alone is sufficient to convert undetectable SNPs into high-fidelity readouts, positioning TEMPO as a generalizable molecular framework for precision genotyping and sequence-specific diagnostics.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(\u003cb\u003eA\u003c/b\u003e) Relocate targeting site via primer engineering. Diagram of the two forward-selective primers (FPP1 and FPP2) positioned at different sites on the target sequence. \u003cb\u003e(B)\u003c/b\u003e Amplification curves for the rs199476104 target (1 fM input) with/without FPP1. The pristine forward primer initiates amplification earlier, whereas FPP1 exhibits a delayed onset due to reduced binding affinity. \u003cb\u003e(C)\u003c/b\u003e Amplification curves for the rs199476104 target (1 fM input) with/without FPP2, showing a similarly delayed and sequential onset of amplification as observed with FPP1. (\u003cb\u003eD\u003c/b\u003e) Design strategy for PAM switching via FPP mismatch engineering. Mismatches within the FPP-binding region reassign the PAM sequence, enabling allele-dependent Cas12a activation. (\u003cb\u003eE and F\u003c/b\u003e) Cas12a cleavage results for canonical PAMs (TTT) and suboptimal PAM variants (ATT) using the rs671 target (1 fM input). (\u003cb\u003eG\u003c/b\u003e) Design strategy using a single-base insertion to enhance allele discrimination. The insertion is introduced downstream of the suboptimal PAM-encoding position within FPP, deliberately shifting the crRNA\u0026ndash;target register to create a defined single-base mismatch in the seed region. \u003cb\u003e(H and I)\u003c/b\u003e Cas12a cleavage results for the rs1229984 target (1 fM input) comparing designs without and with an insertion mutation.\u003c/p\u003e\n\u003ch3\u003eRapid pathogen detection via TEMPO\u003c/h3\u003e\n\u003cp\u003eTo assess the translational potential of TEMPO for rapid pathogen detection, we evaluated its performance using HIV RNA spiked into human serum as simulated clinical specimens. As illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA, extracted RNA was directly introduced into a single-tube TEMPO reaction and reported by real-time fluorescence, requiring no workflow branching or post-amplification handling.\u003c/p\u003e \u003cp\u003eBuilding on its programmable reaction design, TEMPO demonstrated high analytical sensitivity across a broad dynamic range. Fluorescence output scaled proportionally with input RNA from 1,000 aM down to 1 aM, enabling both quantitative readout and naked-eye endpoint discrimination, with a limit of detection of 1 aM (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB; \u003cb\u003eFig. S7\u003c/b\u003e). Kinetic trajectories showed strong linear correlation with RNA input (R\u0026sup2; = 0.9684), closely matching RT\u0026ndash;qPCR calibration (\u003cb\u003eFig. S8\u003c/b\u003e). Additional specificity testing against four non-target species confirmed that off-target templates produced only background-level Cas12a activation (\u003cb\u003eFig. S9\u003c/b\u003e), confirming high sequence specificity of TEMPO.\u003c/p\u003e \u003cp\u003eWe next tested diagnostic performance using simulated clinical samples spanning clinically relevant viral loads. Among the 11 samples tested, TEMPO classification was fully concordant with RT-qPCR results: all nine samples with Ct values below 35 were identified as positive, whereas the two samples with Ct values\u0026thinsp;\u0026ge;\u0026thinsp;35 were classified as negative. This binary classification was fully concordant with RT\u0026ndash;qPCR (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC). Notably, positive\u0026ndash;negative separation emerged within ~\u0026thinsp;20 min under isothermal conditions, substantially faster than thermal cycling qPCR and without the need for multistep sample manipulation.\u003c/p\u003e \u003cp\u003eCollectively, these results establish TEMPO as a single-tube, isothermal platform that achieves qPCR-comparable sensitivity, rapid turnaround, and robust sequence selectivity. The combination of attomolar detection, minimal workflow complexity, and compatibility with crude clinical matrices highlights TEMPO as a promising framework for point-of-care pathogen surveillance and decentralized molecular diagnostics.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e(A)\u003c/b\u003e Schematic workflow for HIV detection using TEMPO. HIV mock samples were prepared and subjected to RNA extraction, followed by isothermal one-pot CRISPR detection or RT-qPCR analysis as a reference method. \u003cb\u003e(B)\u003c/b\u003e Sensitivity of TEMPO for detecting different concentrations of HIV DNA (10000 aM,1000 aM, 100 aM, 10 aM,1 aM,0.1 aM, and negative control). Strong fluorescence signals were observed down to 1 aM concentration, with clear visual fluorescence under blue light. NTC, non-template control reaction. Error bars represent the means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (s.d.) from replicates. The unpaired two-tailed t-test was used to analyze the statistical significance. \u003cb\u003e(C)\u003c/b\u003e Fluorescence output of TEMPO across extracted HIV RNA mock samples with varying RT\u0026ndash;qPCR Ct values. Signal intensity decreases progressively with increasing Ct, consistent with the reduced RNA in put in higher Ct samples.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eOn-site human genotyping by using TEMPO\u003c/h2\u003e \u003cp\u003eHaving established the flexibility, extensibility, and allele-level selectivity of TEMPO under controlled benchmark conditions, we next evaluated its performance in human SNP genotyping. We analyzed 20 clinical genomic samples spanning three medically relevant loci, rs1229984\u003csup\u003e41\u003c/sup\u003e, rs671\u003csup\u003e41\u003c/sup\u003e, and rs199476104\u003csup\u003e42\u003c/sup\u003e, each representing distinct challenges in thermodynamic discrimination. For rs671 and rs199476104, single-nucleotide resolution was achieved solely through PAM attenuation, whereas rs1229984 required a combined strategy involving suboptimal PAM insertion and a single engineered mismatch encoded within the FPP primer. This comparative result underscores the modularity of FPP design and demonstrates that TEMPO can be rationally programmed to resolve thermodynamically recalcitrant SNP sites.\u003c/p\u003e \u003cp\u003eTo translate TEMPO into a clinically viable POC workflow, we developed an integrated platform coupling sample collection, room-temperature lysis, lyophilized one-pot amplification\u0026ndash;CRISPR activation, and fluorescence readout into a 35 min sample-to-answer pipeline (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Buccal-swab lysates were processed by 5 min extraction-free lysis and directly loaded into a palm-sized microfluidic chip (diameter 4 cm, \u003cb\u003eFig. S10\u003c/b\u003e) preloaded with freeze-dried TEMPO reagents. Individual reaction chambers can be customized with locus-specific RPA primers and crRNAs. Upon rehydration, reactions self-initiate and produce a quantifiable readout after a 30 min incubation at 37\u0026deg;C. Fluorescence interpretation is automated through a custom smartphone application, removing the need for manual thresholding or trained operators. This architecture generatively supports on-site genotype determination within 35 min.\u003c/p\u003e \u003cp\u003eWe first validated genotyping accuracy using allele-specific crRNAs. TEMPO clearly distinguished wild-type (WT), heterozygous (HET), and homozygous mutant (SNP) states via distinct kinetic trajectories \u003cb\u003e(\u003c/b\u003eFig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB\u003cb\u003e)\u003c/b\u003e, while maintaining attomolar analytical sensitivity \u003cb\u003e(Fig. S11)\u003c/b\u003e. All 20 clinical samples showed full concordance relative to Sanger-validated genotypes across the three loci (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC \u003cb\u003eto D; Table S4\u003c/b\u003e): rs1229984 (4 WT / 6 SNP / 10 HET), rs671 (16 WT / 4 HET), and rs199476104 (20 WT). The system remained robust under direct-lysis conditions (\u003cb\u003eFig. S12 and S13\u003c/b\u003e), and a sucrose\u0026ndash;mannitol lyophilization matrix retained near-native activity after storage and rehydration, enabling operation without cold-chain logistics (\u003cb\u003eFig. S14 and S15\u003c/b\u003e). For on-site SNP genotyping, the validated TEMPO reaction was integrated into a portable microfluidic chip \u003cb\u003e(Fig. S10, S16 and S17)\u003c/b\u003e. Distinct chip architectures were developed to support either dual-target or multi-target detection, allowing the number of loci interrogated to be adjusted according to application-specific requirements while preserving a unified one-pot workflow. Clinical sample testing demonstrated clear and reliable signal outputs (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eE; \u003cb\u003eFig. S18\u003c/b\u003e), enabling both direct visual interpretation and automated genotype assignment via a custom smartphone application (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eF), with minimal user intervention.\u003c/p\u003e \u003cp\u003eCollectively, these results establish TEMPO as a rapid, lyophilization-compatible SNP genotyping platform with programmable resolution down to single-nucleotide variation. By relocating PAM control from the native genome to FPP-encoded design, TEMPO unlocks previously inaccessible loci and greatly simplifies assay re-targeting without extensive crRNA screening. This modularity, coupled with chip integration and mobile analytics, positions TEMPO as a practical foundation for point-of-care precision genotyping and population-scale genetic screening.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cb\u003e(A)\u003c/b\u003e Workflow of the SNP genotyping using TEMPO. Swab samples are collected and lysed at room temperature for approximately 5 min. The crude lysate is applied directly to the SNP Chip, where a one pot RPA\u0026ndash;Cas12a reaction is performed at 37\u0026deg;C for 30 min. Fluorescence signals are then recorded and interpreted using a smartphone interface. \u003cb\u003e(B)\u003c/b\u003e Representative amplification trajectories illustrating allele-dependent kinetics for rs199476104. Distinct curves corresponding to the wild-type, heterozygous, and mutant genotypes are generated within the one-pot reaction using a 1 fM standard template input. \u003cb\u003e(C)\u003c/b\u003e Heat map summarizing TEMPO genotyping outcomes for 20 clinical samples across three SNP loci (rs1229984, rs671, and rs199476104). The discrimination factor (DF) encoded in each cell quantitatively captures the relative fluorescence contrast between SNP and wild-type alleles, enabling unambiguous genotype assignment with consistent performance across all samples and loci. Genotype calls were assigned based on DF thresholds, with DF values of 0\u0026ndash;0.5 classified as wild type, 0.5\u0026ndash;2.0 as heterozygous, and 2.0\u0026ndash;10 as SNP. \u003cb\u003e(D)\u003c/b\u003e Cleavage rate measured from Sample 20 as a representative example. The observed cleavage patterns match the expected allelic state and show strong agreement with Sanger sequencing traces provided above each bar. \u003cb\u003e(E)\u003c/b\u003e Images of the SNP chip shown with a Hong Kong two-dollar coin for size reference, including the assembled microfluidic chip and representative endpoint fluorescence patterns from sample 20 after a 30 min reaction, highlighting allele-dependent signal differences. \u003cb\u003e(F)\u003c/b\u003e Smartphone interface used for real-time data acquisition, visualization and automated SNP reporting.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eFully integrated one-pot CRISPR diagnostics promise simplified workflows and improved accessibility, yet their practical deployment has been persistently limited by the difficulty of coordinating nucleic acid amplification and CRISPR-mediated signal transduction within a single reaction environment\u003csup\u003e10\u003c/sup\u003e. In most existing systems, these processes compete for shared intermediates, resulting in unstable kinetics, premature target consumption, and inconsistent signal generation\u003csup\u003e43\u003c/sup\u003e. Previous solutions, such as empirical parameter tuning\u003csup\u003e17\u003c/sup\u003e, physical compartmentalization\u003csup\u003e44\u003c/sup\u003e, or externally imposed temporal control\u003csup\u003e45,46\u003c/sup\u003e, have alleviated these conflicts only partially, often at the cost of robustness, generalizability, or operational simplicity. From an engineering perspective, these constraints underscore a fundamental limitation: reaction order in one-pot CRISPR diagnostics is rarely designed, but instead indirectly enforced.\u003c/p\u003e \u003cp\u003eIn this work, we introduce thermodynamically encoded reaction ordering as a molecular engineering strategy that represents a conceptual shift in one-pot CRISPR diagnostics. Rather than mitigating interference through weakened interactions or external staging, TEMPO encodes reaction sequencing directly into molecular free-energy landscapes, converting reaction order from an operational challenge into a programmable system property. This framework simultaneously addresses a major barrier to SNP genotyping: dependence on protospacer adjacent motifs. By encoding PAM sequences within primers rather than relying on native targets, TEMPO enables PAM-independent access to arbitrary loci while preserving single-nucleotide discrimination, expanding genomic addressability for clinically relevant variants. At the platform level, TEMPO achieves attomolar sensitivity, single-nucleotide resolution, and sequencing-concordant genotyping across multiple human SNP loci within a fully integrated workflow. Compatibility with extraction-free lysis, lyophilized reagents, and single-step microfluidic implementation further supports deployment in point-of-care and resource-limited settings, achieving high performance without additional reaction steps, specialized hardware, or user intervention. Together, these features establish TEMPO as a programmable, robust, and clinically adaptable framework for one-pot CRISPR diagnostics.\u003c/p\u003e \u003cp\u003eSeveral limitations define the current scope of the approach. First, while TEMPO supports robust qualitative and semi-quantitative readouts, its quantitative precision does not yet match that of RT\u0026ndash;qPCR. This reflects fundamental differences in reaction control: qPCR enforces deterministic amplification stages through thermal cycling\u003csup\u003e47\u003c/sup\u003e, whereas isothermal amplification exhibits greater variability in initiation and growth kinetics\u003csup\u003e48,49\u003c/sup\u003e. In one-pot CRISPR systems, this variability is compounded by dynamic sharing of amplicons between amplification and CRISPR consumption, rendering signal intensity dependent on coupled reaction dynamics rather than template abundance alone. Consequently, TEMPO is presently best suited for presence\u0026ndash;absence testing and genotyping applications rather than absolute quantification. A second limitation concerns SNP discrimination, which remains inherently dependent on local sequence context. Identical single-nucleotide variants can exhibit divergent discrimination performance depending on neighboring bases and thermodynamic landscape\u003csup\u003e50\u0026ndash;52\u003c/sup\u003e. Although multilayer energetic modulation, combining PAM down tuning with engineered mismatches, can enhance allelic resolution, predictive design rules for SNP discrimination remain incomplete, necessitating empirical validation for new targets.\u003c/p\u003e \u003cp\u003eLooking forward, data-driven and physics-informed modeling approaches may help address these challenges\u003csup\u003e51,53\u003c/sup\u003e. Machine learning frameworks trained on large-scale kinetic and sequence datasets could enable predictive mapping between primer energetics, reaction dynamics, and allelic discrimination performance, improving both quantitative accuracy and target generalizability\u003csup\u003e54,55\u003c/sup\u003e. More broadly, the principle of thermodynamically encoded reaction ordering is not limited to RPA or Cas12a and may be extended to other amplification chemistries and CRISPR effectors, providing a generalizable strategy for coordinating competing biochemical processes within shared reactors.\u003c/p\u003e \u003cp\u003eIn summary, TEMPO establishes thermodynamic encoding as a practical engineering paradigm for designing predictable, integrated, and deployable one-pot CRISPR diagnostics. By unifying reaction sequencing, PAM independence, and SNP discrimination within a single molecular design framework, this work advances the feasibility of scalable point-of-care and at-home nucleic acid testing, bridging the gap between CRISPR diagnostic concepts and real-world biomedical application.\u003c/p\u003e"},{"header":"Method","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eMaterials\u003c/h2\u003e \u003cp\u003e \u003cem\u003eLachnospiraceae bacterium\u003c/em\u003e Cas12a (LbCas12a) protein was from Tolo Biotech (Shanghai). RNA and DNA oligonucleotides, as well as synthetic DNA templates (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e), were provided by GENEWIZ (Suzhou, China). Acrylamide/bisacrylamide solution was purchased from Bio-Rad. SYBR Green Nucleic Acid Gel Stain, and SYTO 82 Orange Fluorescent Nucleic Acid Stain were purchased from Thermo Fisher Scientific (Hong Kong). DEPC Treated Water was from Sangon Biotech (Shanghai). DNA Isothermal Rapid Amplification Kit, RNA Isothermal Rapid Amplification Kit and Rapid Nucleic Acid Release Reagent were from Amp-future (Changzhou). RNase inhibitor, and NEBuffer 2.1 were from New England Biolabs. Proteinase K, Magnetic Bead Virus DNA/RNA Kit and Magnetic Universal Genomic DNA Kit were from Tiangen (Beijing). One Step RT-qPCR Master Mix and Universal Blue qPCR SYBR Green Master Mix was from YEASEN (Shanghai). Sucrose and mannitol were from Macklin (Shanghai).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eRPA and real-time RPA reactions, gel analysis, and sequencing\u003c/h2\u003e \u003cp\u003eRPA reactions were carried out in a 50 \u0026micro;L reaction mixture containing 29.5 \u0026micro;L primer-free rehydration buffer, 1 \u0026micro;L forward primer (10 \u0026micro;M), 1 \u0026micro;L FPP primer (10 \u0026micro;M), 2 \u0026micro;L reverse primer (10 \u0026micro;M), 4 \u0026micro;L template, 2.5 \u0026micro;L MgOAc (280 mM), and 1 \u0026micro;L DNase/RNase-free water. Reactions were incubated at 37\u0026deg;C for 60 min.\u003c/p\u003e \u003cp\u003eTo terminate amplification, Proteinase K was added to the completed reaction and incubated at room temperature for 30 min. The reaction was then heat-inactivated at 95\u0026deg;C for 10 min. The resulting products were either analyzed on 10% denaturing PAGE for gel visualization or submitted for Sanger sequencing (GENEWIZ).\u003c/p\u003e \u003cp\u003eFor real-time RPA, the reaction was assembled as described above with the addition of 1 \u0026micro;M SYTO 82 stain, and fluorescence was recorded continuously during incubation at 37\u0026deg;C.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eTEMPO system\u003c/h2\u003e \u003cp\u003eOne-pot RPA\u0026ndash;Cas12a reactions were assembled in a 20 \u0026micro;L final volume as summarized in Table S2. Briefly, 25 nM LbCas12a and 50 nM crRNA were premixed and incubated for 5 min at room temperature to allow formation of the ribonucleoprotein complex. The reaction mixture contained 400 nM reporters, 125 nM each of primer FP and primer FPP, 250 nM primer BP, NEB Buffer 2.1, the rehydration buffer supplied with the RPA kit, and nuclease-free water. 2 \u0026micro;L Template DNA was added to each reaction. MgOAc (280 mM) was added last to initiate amplification, yielding a total reaction volume of 20 \u0026micro;L. All reactions were mixed gently and incubated under standard one-pot amplification conditions. The fluorescent signal (485 nm excitation and 535 nm emission) was monitored with a qPCR system (Bio-Rad) with an interval time of 1 min.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSample preparation\u003c/h2\u003e \u003cp\u003eHIV mock samples. HIV RNA standards were prepared by spiking 0.5 \u0026micro;L of HIV RNA into 49.5 \u0026micro;L of human serum, followed by nucleic acid extraction using a magnetic bead\u0026ndash;based purification kit. The purified RNA was used as the template for the TEMPO one-pot assay as well as RT- qPCR benchmarking.\u003c/p\u003e \u003cp\u003eHuman genomic samples. Twenty volunteers were recruited at the Hong Kong University of Science and Technology under institutional protocol SP-2024-0197 and HREP-2021-0285. Buccal swabs were collected from each participant and divided into two aliquots. One aliquot was processed using the Magnetic Universal Genomic DNA Kit to obtain purified genomic DNA, which served as the input for one-pot TEMPO genotyping and for Sanger sequencing after RPA amplification of target. The second aliquot was subjected to rapid lysis and used directly for microfluidic chip\u0026ndash;based detection. Briefly, the swab sample was immersed in a centrifuge tube containing 1 mL of lysis buffer and vortexed thoroughly, followed by incubation at room temperature for 5 min. The tube was briefly centrifuged for 5 s, and 10 \u0026micro;L of the supernatant was collected as the template for direct amplification.\u003c/p\u003e \u003cp\u003e \u003cb\u003eRT-qPCR validation of HIV mock samples.\u003c/b\u003e \u003c/p\u003e \u003cp\u003eHIV mock samples were validated using the Hifair Advanced One-Step RT-qPCR SYBR Green Kit (Yeasen). Each 25 \u0026micro;L reaction mixture contained 12.5 \u0026micro;L of 2\u0026times; Hifair Advanced SYBR Green Buffer, 1 \u0026micro;L of forward primer (10 \u0026micro;M), 1 \u0026micro;L of reverse primer (10 \u0026micro;M), 1 \u0026micro;L of Hifair Advanced UH Enzyme Mix, 4 \u0026micro;L of RNA template, and nuclease-free DEPC-treated water to the final volume. RT-qPCR was performed with an initial reverse transcription step at 50\u0026deg;C for 6 min, followed by pre-denaturation at 95\u0026deg;C for 5 min. Amplification was then carried out for 40 cycles consisting of denaturation at 95\u0026deg;C for 15 s and annealing/extension at 60\u0026deg;C for 30 s.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eDesign and fabrication of the microfluidic diagnostic chip\u003c/h2\u003e \u003cp\u003eThe microfluidic diagnostic chip was designed in SolidWorks to integrate multiple reaction chambers and fluidic pathways for parallel SNP detection, with chamber volumes defined to accommodate 20\u0026ndash;25 \u0026micro;L reactions. Passive Tesla-type valve structures were incorporated to prevent contamination arising from liquid backflow during the reaction process\u003csup\u003e56\u003c/sup\u003e. The chip employed a two-layer architecture, consisting of a CNC-machined polymethyl methacrylate (PMMA) layer containing the microfluidic channels and a non-fluorescent adhesive sealing film (Microseal B, Bio-Rad) as the bottom layer. Vent ports were sealed with a waterproof, breathable expanded polytetrafluoroethylene (ePTFE) membrane (diameter, 3.5 mm) to allow gas exchange while preventing liquid leakage and external contamination. The finalized designs were fabricated from PMMA using CNC micromachining by a commercial microfabrication service provider (Shenzhen Huilixing Precision Manufacturing Co., Ltd., China).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eReagent loading and lyophilization on chip\u003c/h2\u003e \u003cp\u003eThe lyophilizable one-pot RPA\u0026ndash;Cas12a reaction mixture was prepared according to the formulation summarized in Table S3. A 20 \u0026micro;L lyophilizable one-pot RPA\u0026ndash;Cas12a reaction mixture was prepared, containing RPA reagents were reconstituted in DEPC-treated water, 125 nM each forward primers FP and FPP, 250 nM reverse primer R, 50 nM Cas12a, 100 nM crRNA, and 800 nM ssDNA reporter, supplemented with 5% (w/w) sucrose and 2% (w/w) mannitol as lyoprotectants. The reaction mixture was dispensed into the reaction chambers (~\u0026thinsp;25 \u0026micro;L per chamber). Chips were snap-frozen by positioning them approximately 1 cm above liquid nitrogen vapor for 2 min, followed by lyophilization for 12 h in the dark. After returning to room temperature, the bottom sealing film and top vent membrane were affixed to complete device assembly.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eStatistics and reproducibility\u003c/h2\u003e \u003cp\u003eStatistical analyses were performed using GraphPad Prism. Quantitative data (endpoint fluorescence and time-to-positive measurements) are reported as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;s.d. from at least two independent replicates unless stated otherwise.\u003c/p\u003e \u003cp\u003eFluorescence data were normalized using DF to quantify the relative signal contrast between SNP and wild-type reactions. The DF was calculated as\u003csup\u003e29\u003c/sup\u003e,\u003c/p\u003e \u003cp\u003eDF = (F\u003csub\u003eSNP\u003c/sub\u003e\u0026minus; F\u003csub\u003eNTC\u003c/sub\u003e) / (F\u003csub\u003eWT\u003c/sub\u003e\u0026minus; F\u003csub\u003eNTC\u003c/sub\u003e),\u003c/p\u003e \u003cp\u003ewhere F\u003csub\u003eSNP\u003c/sub\u003e, F\u003csub\u003eWT\u003c/sub\u003e, and F\u003csub\u003eNTC\u003c/sub\u003e represent the fluorescence intensities measured for SNP, wild-type, and no-target control reactions, respectively.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eAll nucleic acid sequences used in this study are provided in manuscripts or supplementary information. The raw datasets generated and analyzed during the study are available from the corresponding authors on reasonable request.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003eThe TEMPO analyzer mobile phone application and TEMPO model developed in this work are available in \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/zibin-zhao/TEMPO\u003c/span\u003e\u003cspan address=\"https://github.com/zibin-zhao/TEMPO\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thank the developers of OpenAI’s ChatGPT for assistance with language editing of the manuscript. The funding is in part provided by\u0026nbsp;the Research Grants Council (GRF# 16304225, 16303522) of the Hong Kong SAR Government, China.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceptualization: IM.H., X.W., YL.\u003c/p\u003e\n\u003cp\u003eMethodology: IM.H., X.W., YL., Y.C., Z.Z.\u003c/p\u003e\n\u003cp\u003eSoftware: Z.Z.\u003c/p\u003e\n\u003cp\u003eValidation: YL., X.W., Z.Z., S.L.\u003c/p\u003e\n\u003cp\u003eFormal Analysis: YL., X.W., Y.C., Z.Z.\u003c/p\u003e\n\u003cp\u003eInvestigation: YL., X.W.\u003c/p\u003e\n\u003cp\u003eResources: IM.H.\u003c/p\u003e\n\u003cp\u003eData Curation: YL., X.W., Z.Z.\u003c/p\u003e\n\u003cp\u003eVisualization: YL., X.W., Y.C., H.L.\u003c/p\u003e\n\u003cp\u003eFunding acquisition: IM.H.\u003c/p\u003e\n\u003cp\u003eProject administration: IM.H.\u003c/p\u003e\n\u003cp\u003eSupervision: IM.H.\u003c/p\u003e\n\u003cp\u003eWriting – original draft: YL., X.W., Y.C.\u003c/p\u003e\n\u003cp\u003eWriting – review \u0026amp; editing: IM.H., X.W., YL., Y.C., Z.Z.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eI-ming HSING,\u0026nbsp;Xiaolong WU\u0026nbsp;and Yanan LI\u0026nbsp;are co-inventors of a patent application related to this work filed by the Hong Kong University of Science and Technology.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eAdditional information\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrespondence and requests for materials\u003c/strong\u003e should be addressed to I-ming Hsing.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eRolando, J. C., Melkonian, A. V. \u0026amp; Walt, D. R. The Present and Future Landscapes of Molecular Diagnostics. \u003cem\u003eAnnu. Rev. Anal. Chem.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 459\u0026ndash;474 (2024).\u003c/li\u003e\n\u003cli\u003eRolando, J. C., Melkonian, A. V. \u0026amp; Walt, D. R. The Present and Future Landscapes of Molecular Diagnostics. \u003cem\u003eAnnu. Rev. Anal. Chem.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 459\u0026ndash;474 (2024).\u003c/li\u003e\n\u003cli\u003eChen, L., Hu, M. \u0026amp; Zhou, X. Trends in developing one-pot CRISPR diagnostics strategies. \u003cem\u003eTrends Biotechnol.\u003c/em\u003e \u003cstrong\u003e43\u003c/strong\u003e, 98\u0026ndash;110 (2025).\u003c/li\u003e\n\u003cli\u003eKadimisetty, K. \u003cem\u003eet al.\u003c/em\u003e Fully 3D printed integrated reactor array for point-of-care molecular diagnostics. \u003cem\u003eBiosens. Bioelectron.\u003c/em\u003e \u003cstrong\u003e109\u003c/strong\u003e, 156\u0026ndash;163 (2018).\u003c/li\u003e\n\u003cli\u003eRodriguez, N. M., Wong, W. S., Liu, L., Dewar, R. \u0026amp; Klapperich, C. M. A fully integrated paperfluidic molecular diagnostic chip for the extraction, amplification, and detection of nucleic acids from clinical samples. \u003cem\u003eLab. Chip\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 753\u0026ndash;763 (2016).\u003c/li\u003e\n\u003cli\u003eHu, S. \u003cem\u003eet al.\u003c/em\u003e All-in-One Digital Microfluidics System for Molecular Diagnosis with Loop-Mediated Isothermal Amplification. \u003cem\u003eBiosensors\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 324 (2022).\u003c/li\u003e\n\u003cli\u003eBranca, M., Calvet, C., Limoges, B. \u0026amp; Mavr\u0026eacute;, F. Specific Versus Non‐specific Response in Exponential Molecular Amplification from Cross‐Catalysis: Modeling the Influence of Background Amplifications on the Analytical Performances. \u003cem\u003eChemPhysChem\u003c/em\u003e \u003cstrong\u003e22\u003c/strong\u003e, 1611\u0026ndash;1621 (2021).\u003c/li\u003e\n\u003cli\u003eRolando, J. C., Melkonian, A. V. \u0026amp; Walt, D. R. The Present and Future Landscapes of Molecular Diagnostics. \u003cem\u003eAnnu. Rev. Anal. Chem.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 459\u0026ndash;474 (2024).\u003c/li\u003e\n\u003cli\u003eLiu, H. \u003cem\u003eet al.\u003c/em\u003e One‐Pot Isothermal Nucleic Acid Amplification Assisted CRISPR/Cas Detection Technology: Challenges, Strategies, and Perspectives. \u003cem\u003eAdv. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, e06716 (2025).\u003c/li\u003e\n\u003cli\u003eLi, X. \u003cem\u003eet al.\u003c/em\u003e CRISPR-based one-pot detection: A game-changer in nucleic acid analysis. \u003cem\u003eBiosens. Bioelectron.\u003c/em\u003e \u003cstrong\u003e288\u003c/strong\u003e, 117786 (2025).\u003c/li\u003e\n\u003cli\u003eChen, J. S. \u003cem\u003eet al.\u003c/em\u003e CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e360\u003c/strong\u003e, 436\u0026ndash;439 (2018).\u003c/li\u003e\n\u003cli\u003eLi, S.-Y. \u003cem\u003eet al.\u003c/em\u003e CRISPR-Cas12a-assisted nucleic acid detection. \u003cem\u003eCell Discov.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 20 (2018).\u003c/li\u003e\n\u003cli\u003eKaminski, M. M., Abudayyeh, O. O., Gootenberg, J. S., Zhang, F. \u0026amp; Collins, J. J. CRISPR-based diagnostics. \u003cem\u003eNat. Biomed. Eng.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 643\u0026ndash;656 (2021).\u003c/li\u003e\n\u003cli\u003eFeng, W. \u003cem\u003eet al.\u003c/em\u003e CRISPR technology incorporating amplification strategies: molecular assays for nucleic acids, proteins, and small molecules. \u003cem\u003eChem. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 4683\u0026ndash;4698 (2021).\u003c/li\u003e\n\u003cli\u003eZhang, L., Jiang, H., Zhu, Z., Liu, J. \u0026amp; Li, B. Integrating CRISPR/Cas within isothermal amplification for point-of-Care Assay of nucleic acid. \u003cem\u003eTalanta\u003c/em\u003e \u003cstrong\u003e243\u003c/strong\u003e, 123388 (2022).\u003c/li\u003e\n\u003cli\u003eYao, Z. \u003cem\u003eet al.\u003c/em\u003e Tuning the Dynamic Reaction Balance of CRISPR/Cas12a and RPA in One Pot: A Key to Switch Nucleic Acid Quantification. \u003cem\u003eACS Sens.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 3511\u0026ndash;3519 (2024).\u003c/li\u003e\n\u003cli\u003eDing, X. \u003cem\u003eet al.\u003c/em\u003e Ultrasensitive and visual detection of SARS-CoV-2 using all-in-one dual CRISPR-Cas12a assay. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 4711 (2020).\u003c/li\u003e\n\u003cli\u003eZhang, H.-X. \u003cem\u003eet al.\u003c/em\u003e Cas12a-based one-pot SNP detection with high accuracy. \u003cem\u003eCell Insight\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 100080 (2023).\u003c/li\u003e\n\u003cli\u003eZhuang, S., Bai, B. \u0026amp; Liu, Y. CRISPR-based SNP detection technologies advance from classical methods to cutting-edge innovations. \u003cem\u003eChem. Commun.\u003c/em\u003e \u003cstrong\u003e61\u003c/strong\u003e, 13345\u0026ndash;13358 (2025).\u003c/li\u003e\n\u003cli\u003eChai, L. \u003cem\u003eet al.\u003c/em\u003e CRISPR/Cas Toolbox: Unveiling the secret weapon for SNP detection. \u003cem\u003eTrAC Trends Anal. Chem.\u003c/em\u003e \u003cstrong\u003e192\u003c/strong\u003e, 118321 (2025).\u003c/li\u003e\n\u003cli\u003eKohabir, K. A. V. \u003cem\u003eet al.\u003c/em\u003e \u003cem\u003eIn Vitro\u003c/em\u003e CRISPR-Cas12a-Based Detection of Cancer-Associated \u003cem\u003eTP53\u003c/em\u003e Hotspot Mutations Beyond the crRNA Seed Region. \u003cem\u003eCRISPR J.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 127\u0026ndash;139 (2023).\u003c/li\u003e\n\u003cli\u003eOoi, K. H. \u003cem\u003eet al.\u003c/em\u003e An engineered CRISPR-Cas12a variant and DNA-RNA hybrid guides enable robust and rapid COVID-19 testing. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 1739 (2021).\u003c/li\u003e\n\u003cli\u003eChen, Y., Mei, Y., Zhao, X. \u0026amp; Jiang, X. Reagents-Loaded, Automated Assay that Integrates Recombinase-Aided Amplification and Cas12a Nucleic Acid Detection for a Point-of-Care Test. \u003cem\u003eAnal. Chem.\u003c/em\u003e \u003cstrong\u003e92\u003c/strong\u003e, 14846\u0026ndash;14852 (2020).\u003c/li\u003e\n\u003cli\u003eLu, S. \u003cem\u003eet al.\u003c/em\u003e Fast and sensitive detection of SARS-CoV-2 RNA using suboptimal protospacer adjacent motifs for Cas12a. \u003cem\u003eNat. Biomed. Eng.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 286\u0026ndash;297 (2022).\u003c/li\u003e\n\u003cli\u003eHu, M. \u003cem\u003eet al.\u003c/em\u003e Photocontrolled crRNA activation enables robust CRISPR-Cas12a diagnostics. \u003cem\u003eProc. Natl. Acad. Sci.\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, e2202034119 (2022).\u003c/li\u003e\n\u003cli\u003eChen, Y. \u003cem\u003eet al.\u003c/em\u003e Photoactivatable CRISPR/Cas12a Strategy for One-Pot DETECTR Molecular Diagnosis. \u003cem\u003eAnal. Chem.\u003c/em\u003e \u003cstrong\u003e94\u003c/strong\u003e, 9724\u0026ndash;9731 (2022).\u003c/li\u003e\n\u003cli\u003eCao, Y. \u003cem\u003eet al.\u003c/em\u003e CoLAMP: CRISPR-based one-pot loop-mediated isothermal amplification enables at-home diagnosis of SARS-CoV-2 RNA with nearly eliminated contamination utilizing amplicons depletion strategy. \u003cem\u003eBiosens. Bioelectron.\u003c/em\u003e \u003cstrong\u003e236\u003c/strong\u003e, 115402 (2023).\u003c/li\u003e\n\u003cli\u003eCao, Y. \u003cem\u003eet al.\u003c/em\u003e Benchtop to at-home test: Amplicon-depleted CRISPR-regulated loop mediated amplification at skin-temperature for viral load monitoring. \u003cem\u003eBiosens. Bioelectron.\u003c/em\u003e \u003cstrong\u003e267\u003c/strong\u003e, 116866 (2025).\u003c/li\u003e\n\u003cli\u003eWu, X. \u003cem\u003eet al.\u003c/em\u003e RatioCRISPR: A ratiometric biochip based on CRISPR/Cas12a for automated and multiplexed detection of heteroplasmic SNPs in mitochondrial DNA. \u003cem\u003eBiosens. Bioelectron.\u003c/em\u003e \u003cstrong\u003e241\u003c/strong\u003e, 115676 (2023).\u003c/li\u003e\n\u003cli\u003eChen, Y. \u003cem\u003eet al.\u003c/em\u003e Dual-CRISPR/Cas12a-Assisted RT-RAA for Ultrasensitive SARS-CoV-2 Detection on Automated Centrifugal Microfluidics. \u003cem\u003eAnal. Chem.\u003c/em\u003e \u003cstrong\u003e94\u003c/strong\u003e, 9603\u0026ndash;9609 (2022).\u003c/li\u003e\n\u003cli\u003eAi, J. \u003cem\u003eet al.\u003c/em\u003e PAM-Independent CRISPR-Cas12a System for Specific Assays of Single Nucleotide Variants. \u003cem\u003eJACS Au\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1392\u0026ndash;1401 (2025).\u003c/li\u003e\n\u003cli\u003eDai, J. \u003cem\u003eet al.\u003c/em\u003e Exploiting the Potential of Spherical PAM Antenna for Enhanced CRISPR-Cas12a: A Paradigm Shift toward a Universal Amplification-Free Nucleic Acid Test Platform. \u003cem\u003eAnal. Chem.\u003c/em\u003e \u003cstrong\u003e97\u003c/strong\u003e, 1236\u0026ndash;1245 (2025).\u003c/li\u003e\n\u003cli\u003eHuang, Z. \u003cem\u003eet al.\u003c/em\u003e Thermally programmed one-pot CRISPR assay for on-site pandemic surveillance. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 10286 (2025).\u003c/li\u003e\n\u003cli\u003eMoon, J. \u0026amp; Liu, C. Asymmetric CRISPR enabling cascade signal amplification for nucleic acid detection by competitive crRNA. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e14\u003c/strong\u003e, 7504 (2023).\u003c/li\u003e\n\u003cli\u003eMao, X. \u003cem\u003eet al.\u003c/em\u003e Iterative crRNA design and a PAM-free strategy enabled an ultra-specific RPA-CRISPR/Cas12a detection platform. \u003cem\u003eCommun. Biol.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1454 (2024).\u003c/li\u003e\n\u003cli\u003eKohabir, K. A. V. \u003cem\u003eet al.\u003c/em\u003e Synthetic mismatches enable specific CRISPR-Cas12a-based detection of genome-wide SNVs tracked by ARTEMIS. \u003cem\u003eCell Rep. Methods\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 100912 (2024).\u003c/li\u003e\n\u003cli\u003eKohabir, K. A. V., Sistermans, E. A. \u0026amp; Wolthuis, R. M. F. Recent advances in CRISPR-based single-nucleotide fidelity diagnostics. \u003cem\u003eCommun. Med.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 252 (2025).\u003c/li\u003e\n\u003cli\u003eKarvelis, T. \u003cem\u003eet al.\u003c/em\u003e PAM recognition by miniature CRISPR\u0026ndash;Cas12f nucleases triggers programmable double-stranded DNA target cleavage. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cstrong\u003e48\u003c/strong\u003e, 5016\u0026ndash;5023 (2020).\u003c/li\u003e\n\u003cli\u003eChen, Y., Mei, Y. \u0026amp; Jiang, X. Universal and high-fidelity DNA single nucleotide polymorphism detection based on a CRISPR/Cas12a biochip. \u003cem\u003eChem. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 4455\u0026ndash;4462 (2021).\u003c/li\u003e\n\u003cli\u003eHe, Y., Shao, S. \u0026amp; Chen, J. High-Fidelity Identification of Single Nucleotide Polymorphism by Type V CRISPR Systems. \u003cem\u003eACS Sens.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 4478\u0026ndash;4483 (2023).\u003c/li\u003e\n\u003cli\u003eChang, T., Yen, T., Wei, C., Hsiao, T. \u0026amp; Chen, I. Impacts of \u003cem\u003eADH1B\u003c/em\u003e \u0026nbsp;rs1229984 and \u0026nbsp;\u003cem\u003eALDH2\u003c/em\u003e \u0026nbsp;rs671 polymorphisms on risks of alcohol‐related disorder and cancer. \u003cem\u003eCancer Med.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 747\u0026ndash;759 (2023).\u003c/li\u003e\n\u003cli\u003eCatarino, C. B. \u003cem\u003eet al.\u003c/em\u003e Characterization of a Leber\u0026rsquo;s hereditary optic neuropathy (LHON) family harboring two primary LHON mutations m.11778G \u0026gt; A and m.14484T \u0026gt; C of the mitochondrial DNA. \u003cem\u003eMitochondrion\u003c/em\u003e \u003cstrong\u003e36\u003c/strong\u003e, 15\u0026ndash;20 (2017).\u003c/li\u003e\n\u003cli\u003eTong, X. \u003cem\u003eet al.\u003c/em\u003e Fast and sensitive CRISPR detection by minimized interference of target amplification. \u003cem\u003eNat. Chem. Biol.\u003c/em\u003e \u003cstrong\u003e20\u003c/strong\u003e, 885\u0026ndash;893 (2024).\u003c/li\u003e\n\u003cli\u003eZhou, H. \u003cem\u003eet al.\u003c/em\u003e Phase Transition of Wax Enabling CRISPR Diagnostics for Automatic At‐Home Testing of Multiple Sexually Transmitted Infection Pathogens. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e21\u003c/strong\u003e, 2407931 (2025).\u003c/li\u003e\n\u003cli\u003eCheng, Z.-H. \u003cem\u003eet al.\u003c/em\u003e Tunable control of Cas12 activity promotes universal and fast one-pot nucleic acid detection. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 1166 (2025).\u003c/li\u003e\n\u003cli\u003eLin, M. \u003cem\u003eet al.\u003c/em\u003e Glycerol Additive Boosts 100-fold Sensitivity Enhancement for One-Pot RPA-CRISPR/Cas12a Assay. \u003cem\u003eAnal. Chem.\u003c/em\u003e \u003cstrong\u003e94\u003c/strong\u003e, 8277\u0026ndash;8284 (2022).\u003c/li\u003e\n\u003cli\u003eBustin, S. A. \u003cem\u003eet al.\u003c/em\u003e The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. \u003cem\u003eClin. Chem.\u003c/em\u003e \u003cstrong\u003e55\u003c/strong\u003e, 611\u0026ndash;622 (2009).\u003c/li\u003e\n\u003cli\u003eCrannell, Z. A., Rohrman, B. \u0026amp; Richards-Kortum, R. Development of a Quantitative Recombinase Polymerase Amplification Assay with an Internal Positive Control. \u003cem\u003eJ. Vis. Exp.\u003c/em\u003e 52620 (2015).\u003c/li\u003e\n\u003cli\u003eLobato, I. M. \u0026amp; O\u0026rsquo;Sullivan, C. K. Recombinase polymerase amplification: Basics, applications and recent advances. \u003cem\u003eTrAC Trends Anal. Chem.\u003c/em\u003e \u003cstrong\u003e98\u003c/strong\u003e, 19\u0026ndash;35 (2018).\u003c/li\u003e\n\u003cli\u003eYou, Y. Design of LNA probes that improve mismatch discrimination. \u003cem\u003eNucleic Acids Res.\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, e60\u0026ndash;e60 (2006).\u003c/li\u003e\n\u003cli\u003eKotlarz, K., Mielczarek, M., Biecek, P., Guldbrandtsen, B. \u0026amp; Szyda, J. Exploring the impact of sequence context on errors in SNP genotype calling with whole genome sequencing data using AI-based autoencoder approach. \u003cem\u003eNAR Genomics Bioinforma.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, lqae131 (2024).\u003c/li\u003e\n\u003cli\u003eDe Oliveira Martins, E. \u0026amp; Weber, G. Nearest-neighbour parametrization of DNA single, double and triple mismatches at low sodium concentration. \u003cem\u003eBiophys. Chem.\u003c/em\u003e \u003cstrong\u003e306\u003c/strong\u003e, 107156 (2024).\u003c/li\u003e\n\u003cli\u003eListgarten, J. \u003cem\u003eet al.\u003c/em\u003e Prediction of off-target activities for the end-to-end design of CRISPR guide RNAs. \u003cem\u003eNat. Biomed. Eng.\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 38\u0026ndash;47 (2018).\u003c/li\u003e\n\u003cli\u003eWessels, H.-H. \u003cem\u003eet al.\u003c/em\u003e Prediction of on-target and off-target activity of CRISPR\u0026ndash;Cas13d guide RNAs using deep learning. \u003cem\u003eNat. Biotechnol.\u003c/em\u003e \u003cstrong\u003e42\u003c/strong\u003e, 628\u0026ndash;637 (2024).\u003c/li\u003e\n\u003cli\u003eCamacho, D. M., Collins, K. M., Powers, R. K., Costello, J. C. \u0026amp; Collins, J. J. Next-Generation Machine Learning for Biological Networks. \u003cem\u003eCell\u003c/em\u003e \u003cstrong\u003e173\u003c/strong\u003e, 1581\u0026ndash;1592 (2018).\u003c/li\u003e\n\u003cli\u003eLiu, Z., Shao, W.-Q., Sun, Y. \u0026amp; Sun, B.-H. Scaling law of the one-direction flow characteristics of symmetric Tesla valve. \u003cem\u003eEng. Appl. Comput. Fluid Mech.\u003c/em\u003e \u003cstrong\u003e16\u003c/strong\u003e, 441\u0026ndash;452 (2022).\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8802024/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8802024/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eOne-pot CRISPR diagnostics face a fundamental incompatibility: nucleic acid amplification requires rapid target accumulation, whereas CRISPR activation irreversibly consumes those substrates, destabilizing reaction kinetics. Existing strategies rely on empirical parameter balancing or external staging but lack an intrinsic mechanism to enforce reaction order within a single reactor. Here we introduce thermodynamic encoding as a molecular design principle that programs reaction order directly into DNA primers, enabling autonomous, threshold-gated activation of CRISPR only after sufficient amplicon accumulated. By embedding a defined free-energy differential between competing primers, the system evolves through two kinetically ordered amplification regimes, decoupling amplification from CRISPR signal transduction without physical separation or external triggers. This architecture relocates PAM dependence from native genomic targets to primer-encoded design, enabling detection of otherwise inaccessible loci while preserving single-nucleotide discrimination. An ordinary differential equation model captures the threshold behavior and establishes a predictable framework for primer design. Building on this principle, we develop Thermodynamically Encoded Molecular Programming for One-pot diagnostics (TEMPO), which achieves attomolar sensitivity within 30 min and enables sequencing-concordant SNP genotyping and pathogen detection in a single-step microfluidic format.\u003c/p\u003e","manuscriptTitle":"Thermodynamically programmed one-pot CRISPR platform for point-of-care SNP genotyping","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-17 06:52:10","doi":"10.21203/rs.3.rs-8802024/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"83bb12c2-bcc2-4c4d-828c-61166336f53b","owner":[],"postedDate":"February 17th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":62895380,"name":"Biological sciences/Biotechnology/Assay systems"},{"id":62895381,"name":"Biological sciences/Molecular biology/CRISPR-Cas systems"},{"id":62895382,"name":"Physical sciences/Chemistry/Chemical biology/Nucleic acids"},{"id":62895383,"name":"Physical sciences/Nanoscience and technology/Nanobiotechnology/Biosensors"}],"tags":[],"updatedAt":"2026-05-11T09:19:02+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-17 06:52:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8802024","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8802024","identity":"rs-8802024","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}