SUPPLEMENTARY INFORMATION

Transcription

1 DOI:.8/NCHEM. Conditionally Fluorescent Molecular Probes for Detecting Single Base Changes in Double-stranded DNA Sherry Xi Chen, David Yu Zhang, Georg Seelig. Analytic framework and probe design.. Design Principles for double-stranded toehold exchange (Text S).. Conditionally fluorescent probes for dsdna (Text S)...Equilibrium hybridization yield χ at G intended = (Text S.)... Discrimination factor Q at G intended = (Text S.)... Representative equilibrium hybridization yields and discrimination factors (Text S.)... Effects of Stoichiometry n by hybridization yield of intended and SNP targets (Table S).. Mathematical relation between R and Q (Text S)... Hybridization Probe (Text S.)...Double stranded toehold exchange probe (Text S.) 8. Probe performance 9.. Fluorescence normalization and discrimination factor Q determination (Fig. S) 9.. Expand View of reaction between probe and SNP targets (Fig. S) 9.. Robustness... Probe discriminates correct target from SNP targets regardless of probe concentration (Fig S)... Robustness Discrimination of pm intended target from 8mCG SNP variant (Fig S)... Poly-N single-stranded DNA has no significant effect thermodynamically or kinetically (Fig S)... Temperature and salinity effects on probe performance (Fig S).. Alternative designs for four-way toehold exchange (Fig S).. Unbalanced initiation and dissociation toeholds... Reduced specificity with toehold sequences cause G o to be significantly negative (Fig S8)... Double-stranded probes without dissociation toehold react with both correct and SNP targets (Fig S9)... Comparison of the toehold exchange probe with the probe lacking balancing toeholds (Fig S).. Comparison with Molecular beacon... Performance of two different molecular beacon on discriminating SNPs (Fig S)... DPerformance of molecular beacon targeting nt subsequence of the E. coli rpob gene (Fig S). Biologically derived targets.. Detection of Rif-resistance in E.coli; wild type and resistant colonies (Fig S - S).. Schematic of the targets and probes used in Figure of the main paper.. PCR primers for E. Coli-derived target (Table S). Electrospray Ionization (ESI) mass spectrometry spectrum of DNAs.. Representative Electrospray Ionization (ESI) mass spectrometry spectrum (Fig. S9).. ESI summary for oligonucleotides used in Figure, Figure and Figure of the main paper.. ESI summary for oligonucleotides used in Figure of the main paper NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

2 DOI:.8/NCHEM. Text S. Design Principles for double-stranded toehold exchange Text S. Conditionally fluorescent probes for dsdna S.. Equilibrium hybridization yield χ at G intended = S.. Discrimination factor Q at G intended = S.. Representative equilibrium hybridization yields and discrimination factors Text S. Mathematical relation between R and Q S. Hybridization probes S. Double stranded toehold exchange probes 8 Fig. S. Fluorescence normalization and discrimination factor Q determination 9 Fig. S. Expand View of reaction between probe and SNP targets 9 Fig. S. Probe discriminates correct target from SNP targets regardless of probe concentration Fig. S. Discrimination of pm intended target from 8mCG SNP variant Fig. S. Poly-N single-stranded DNA has no significant effect thermodynamically or kinetically Fig. S. Temperature and salinity effects on probe performance Fig. S. Alternative designs for four-way toehold exchange Fig. S8. Reduced specificity with toehold sequences cause G o to be significantly negative Fig. S9. Double-stranded probes without dissociation toehold react with both correct and SNP targets Fig. S. Comparison of the toehold exchange probe with the probe lacking balancing toeholds Fig. S. Performance of two different molecular beacon on discriminating SNPs Fig. S. Performance of molecular beacon targeting nt subsequence of the E. coli rpob gene Fig. S. Detection of Rif-resistance in E.coli; wild type and resistant colonies # and #. Fig. S. Detection of Rif-resistance in E.coli; wild type and resistant colonies # and #. 8 Fig. S. Detection of Rif-resistance in E.coli; wild type and resistant colonies # and #. 9 Fig. S. Detection of Rif-resistance in E.coli; wild type and resistant colonies # and #8. Fig. S. Detection of Rif-resistance in E.coli; wild type and resistant colonies #9 and #. Fig. S8. Schematic of the targets and probes used in Figure of the main paper Fig. S9. Representative Electrospray Ionization (ESI) mass spectrometry spectrum Table S. Effects of Stoichiometry n by hybridization yield of intended and SNP targets Table S. PCR primers for E. Coli-derived target Table S. ESI summary for oligonucleotides used in Figure, Figure and Figure of the main paper Table S. ESI summary for oligonucleotides used in Figure of the main paper NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

3 DOI:.8/NCHEM. Text S. Design principles for double-stranded toehold exchange Recall that the double-stranded toehold exchange process can be modeled as a series of states connected by transition rates (Fig. SA). We omit intermediate states in which the four strands are co-localized, but in which the toeholds are incompletely hybridized (Fig. SB), because these states suffer both entropic and enthalpic penalties, and consequently will not be prevalent in solution. Intermediate states in which internal base pairs are breathing (Fig. SC) are likewise omited. The reduced states of the reaction are thus: A + B C... C i... C n D + E () In the consideration of the energies of each of these states with respect to determination of occupancy (concentration) at equilibrium, we assume a standard DNA thermodynamic model in which only Watson- Crick base stacking energies are considered. Furthermore, for simplicity, we assume that the energies of all -helix multiloops in the various branch migration states C i are identical in strength. Under these assumptions, the (n + ) different branch migration states C i are identical in energy, and can thus be grouped into a conglomerate state C with energy: G C = G C i RT ln(n + ) () Doing so preserves the partition function of the original system in eq. : n Z = e G A+B /RT +( e G C /RT i )+e G D+E /RT i= = e G A+B /RT + e ln(n+) e G C /RT i + e G D+E /RT = e G A+B /RT + e ( G C RT ln(n+))/rt i + e G D+E /RT In this reduced reaction system A + B C D + E, the equilibrium concentration of C is low (e G C /RT Z) when either G C > G A+B or G C > G D+E. The difference between the values of G A+B and G C is due to the combined G values of the orange and purple initiating toeholds. Furthermore, to correctly account for the bimolecular A + B state in the Markov model, we need to adjust its G by the a concentration term RT ln(c), where c is the current concentration of the excess species of A and B (scaled to M for hybridization initiation entropy). Although c will change with time and progression of the reaction, it is bounded; assuming that [A] < [B],[B] c [B] [A]. G A+B = G C G toeholds + RT ln(c) = G C + RT ln(n + ) G toeholds + RT ln(c) NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

4 DOI:.8/NCHEM. To satisfy G C > G A+B, G toeholds > RT ln(c(n + )) The conditions for G C > G D+E is similar, except using the blue and green toehold energies instead. The equilibrium concentrations of C must be low for the dsdna probe to function properly. When the equilibrium concentration of the C states are not low, much of the probe could be trapped in intermediates, which not only results in low fluorescence for correct target binding (due to failure to dissociate), but also slow the kinetics of the overall reaction when the probe is trapped in intermediates with the SNP targets. For our experiments, n varied between and, c was typically nm, so RT ln(c(n + )) varied between -9.8 kcal/mol and -. kcal/mol. Our blue and green toeholds were typically weaker ( base pairs each), with total G toeholds = 8.8 kcal/mol, according to established thermodynamic parameters [? ]. The interaction between the ROX fluorophore and the Iowa Black RQ quencher likely also contributes to the thermodynamic stability of the toeholds. Consequently, our experiments in Fig. and Fig. S that use a nt branch migration region may suffer from intermediate state trapping. Because higher temperatures and lower salinity concentrations tends to weaken the thermodynamic contributions of base paring (making G toeholds less negative), it is expected that the probes for very long regions will show superior performance at elevated temperatures and reduced salinities. Text S. Conditionally fluorescent probes for dsdna The high specificity conditionally fluorescent dsdna probe involves a special case of double-stranded toehold exchange, in which the standard free energy G of the probe binding to the intended target is approximately. This is achieved by designing the initiation and dissociation toeholds to be approximately the same length and strength. We aim for G because this gives a good tradeoff between hybridization yield of the intended target and the discrimination factor between intended and SNP targets:. Equilibrium hybridization yield χ at G intended =. The equilibrium constant for the A+B D +E reaction is K eq = [D][E] [A][B] = [D] [A][B], where A is the target, B is the probe, D is the fluorescent product, and E is the dark product. In a detection reaction, there are no product [D] and [E] initially; therefore any production of D must necessarily be accompanied by production of E, and the concentrations of the two products are always equal. We define n to be the stoichiometric ratio between the initial concentrations of A and B (n = [A] [B] ). The hybridization yield χ is calculated based off the limiting species of target and probe (χ = [D] min{[a],[b] } ), and varies with time. Here we show derivation for the case where n, but similar results follow for n<. With some simplification, the equilibrium hybridization yield χ can be expressed as a function of NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

5 DOI:.8/NCHEM. stoichiometry n and equilibrium constant K eq : K eq = χ ( χ )(n χ ) χ = K eq + K eq n K eq Keq (n ) +n (K eq ) () () except in the case of K eq =, in which case χ = n n+.. Discrimination factor Q at G intended =. The discrimination factor Q is defined as: Q = χ (intended) χ (SNP) With G intended =, the numerator becomes n n+. The denominator can be calculated based on substituting G SNP = G ds into equation (). The value of G ds depends on the identity and local neighborhood of the SNP, and generally varies between + and + kcal/mol at room temperature. Here, the exact value of χ (SNP) does not matter; we merely wish to establish a lower bound on Q. At very high values of n, the system may lose specificity because the overwhelming amount of SNP target can produce equilibrium yield χ >.. We will discuss this scenario later in text S; here we limit ourselves to in which χ <. (e.g. n<),. In this case, we can bound the value of Q: χ K eq = ( χ )(n χ ) < χ.(n.) n e G ds /RT < χ n e G ds /RT < χ (SNP) Q = χ (intended) χ (SNP) > n n+ n e G ds /RT n = (n + ) n e G ds /RT e G ds /RT n NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

6 DOI:.8/NCHEM. Stoichiometry n χ (intended, G =. kcal/mol) χ (SNP, G o = +. kcal/mol) Q = χ (Intended) χ SNP TABLE S: Effects of stoichiometry n on hybridization yield of intended and SNP targets, and corresponding discrimination factor Q. The value of Q increases monotonically as G intended increases [? ]. At the limit of G intended +, K eq = G /RT e = χ ( χ )(n χ ) χ χ (intended) = e G intended /RT χ (SNP) = e G SNP / RT = e G ds /RT e G intended /RT Q max = χ (intended) χ (SNP) = e G ds /RT Thus, at G intended = and reasonably small n<, the discrimination factor is within a small constant factor n of optimal.. Representative equilibrium hybridization yields and discrimination factors The solid lines shown in Fig. of the main text plots the analytic dependence of χ on stoichiometry n shown in Equation (), assuming the following best-fit G values: Target G (kcal/mol) Intended (black) -. i8ta (red) +.9 d8 (green) +. m8cg (blue) +. Table S shows analytic equilibrium hybridization yields for representative values of n and G. Dividing the equilibrium hybridization yield for G =. by that for G = +. kcal/mol, we obtain the analytic discrimination factor Q. Although discrimination factor is highest at n =, the SNP can still be effectively distinguished at x excess of target over probe. Because analysis is symmetrical with respect to A and B, stoichiometries n< will yield the same discrimination factor as stoichiometry n = n. Text S. Mathematical relation between R and Q. The concentration equivalence R is an alternative measure of specificity, and denotes the ratio of the quantity of SNP versus intended target needed to yield the same level of fluorescence (% of maximum NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

7 DOI:.8/NCHEM. here): Conceptually, an R-fold excess of the SNP target yields a false positive, so R determines the specificity of a diagnostic assay based on this technology. Hybridization probes. For typical (A + B D) hybridization probes, R Q. First, we show the derivation for Q in the case of known n: K eq = [D] [A][B] = χ = n + + χ (n χ )( χ )[B] K eq [B] K eq[b] n + K eq[b] ++K eq[b] n +K eq[b] K eq [B] n K eq [B] The last expression of χ is rather complicated, but if we assume that χ. (and thus in the specific regime of the reaction), the solution for simplifies to: χ K eq [B] n = e G /RT [B] n The discrimination factor Q is calculated as: Q = χ (intended) χ (SNP) e G intended /RT [B] n e G SNP /RT [B] n = e G intended /RT + G SNP /RT We define G ss = G SNP G intended, and the above simplifies to: Q = e G ss /RT Next, we derive R based on χ =.: e G /RT = K eq = [D] [A][B] = [B] (n.) = e G /RT From this expression, we can solve for R: n = [B] +e G /RT [B] χ (n χ )( χ )[B] R = n SNP n intended = [B] +e G SNP /RT [B] +e G intended /RT NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

8 DOI:.8/NCHEM. We define G ss = G SNP G intended, and the above simplifies to: R = [B] +e G ss /RT e G intended /RT [B] +e G intended /RT When the intended hybridization is too favorable (e G intended /RT [B] ), the reaction is not specific, and the concentration equivalence approaches. In contrast, when (e G intended /RT > [B] ), the concentration equivalence can be approximated as: R e G ss /RT. Thus, R Q for the A + B D hybridization probe reaction in the specific regime of the probe. Double stranded toehold exchange probes. For the A + B D + E reaction with G ds = G SNP G intended, similar logic as the above leads to: χ ne G /RT Q = e G ds /RT Similarly for R: n = +e G /RT R e G ds /RT Thus, for the double-stranded probe, R Q in the specific regime where hybridization yields are below.. Experimentally, the median discrimination factor Q that we observe is ; consequently, we expect that up to an up to R = Q 8-fold excess of an SNP target will be tolerated before yielded the same fluorescence signal as x of the intended target. Note that G ds G ss, since the double-stranded probe results in mismatch bubbles per base pair change in the target, whereas the conventional hybridization probes results in mismatch bubble per base change. Thus, the observed discrimination factor Q will be roughly the same for the double-stranded probe as under the conventional hybridization probe under optimal conditions, but the concentration equivalence R will be roughly quadratically larger. NATURE CHEMISTRY 8 Macmillan Publishers Limited. All rights reserved.

9 DOI:.8/NCHEM. Direct Trigger Probe ROX ATACACACACCACCAACC TATGTGTGTGGTGGTTGG + RQ Trigger AGTGAATACACACACCACCAACC ATACACACACCACCAACC + AGTGAATACACACACCACCAACC TATGTGTGTGGTGGTTGG RQ Fluorescent Signal (a.u.). F nm, χ = max =.9.8 [Probe] = nm... nm Correct Target. F =.9. nm, χ =..8. nm SNP Target (m8cg). Probe only F =.8. b =.8. nm, χ =. nm, χ = + nm Target + nm Trigger FIG. S: Fluorescence normalization and discrimination factor Q determination. Shown to the right are sample fluorescence traces for the probe alone (orange), the probe with correct target (black), and the probe with the m8cg SNP target (blue). At time t =, the reaction is started by addition of target, if any. At time t =. hr, the reaction is terminated by addition of a large excess of a direct trigger that reacts with the probe (left panel); this fluorescence value after equilibration of this latter reaction corresponds to % reaction yield. The hybridization yields of the correct and SNP targets are calculated as: χ = F F b F max F b, where F denotes the fluorescence value of the reaction after. hours of reaction, F b denotes the time-varying background fluorescence observed from addition of pure buffer, and F max denotes the equilibrium after post-trigger. Fluoresence Signal (a.u.) correct 8 baseline icg i8cg i8ta i8gc mgc d mgc i8at d8 ig-c d m8cg m8at m8ta FIG. S: Expanded view of reaction between probe and SNP targets from Fig., plotted as non-normalized fluorescence. The reaction of the probe with the correct target is shown in black and rises very quickly with respect to SNP traces. Note that the arbitrary fluorescence units of different figures differ from each other. The baseline trace shows the average of separate reactions of probe with no target; buffer was added at t = to preserve similarity of reaction conditions and correct for dilution effects. Fluorescence value for the baseline and SNP traces showed an initial decline of fluorescence at t< hr; this decline is repeatable and likely due to the settling of air bubbles introduced from mixing by repeated pipetting. NATURE CHEMISTRY 9 Macmillan Publishers Limited. All rights reserved.

10 DOI:.8/NCHEM. Fluorescnece (a.u.).. nm Probe.nM Correct target nm SNP target (m8cg).nm SNP target (m8cg) H. yield (χ) H. yield (χ).8 Fluorescnece (a.u.)..nm Probe.nM Correct target nm Mutated target (m8cg).nm Mutated target (m8cg)..8. FIG. S: Probe effectively discriminates correct target from SNP targets regardless of probe concentration, as predicted by theory. Consistent with analysis and Fig. from the main text, increased concentrations of SNP targets have a sub-linear effect on the hybridization yield. CCAAGGTGGTGTGTGTAT GGTTCCACCACACACATA Target mutate to C-G Probe AGTGA AGTGA ROX ATACACACACCACCAACC TATGTGTGTGGTGGTTGG RQ Fluorescnece (pm) [Probe] = nm, [Target] = pm Correct m8cg FIG. S: Discrimination of pm intended target from m8cg SNP variant. Hybridization Yield. nm polyn nm polyn M Na +, o C nm Probe + nm Correct Target μm polyn μm polyn μm polyn μm polyn no polyn nm Probe + nm m8cg 8 Discrimination Factor Q at t = hr [polyn] nm nm nm μm μm μm μm FIG. S: Poly-N single-stranded DNA has no significant effect on the thermodynamic or kinetic behavior of the probe reaction, even at high concentration excesses. This result contrasts that of probe based on three-stranded toehold exchange for discriminating single-stranded DNA and RNA [? ]. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

11 DOI:.8/NCHEM. a b c d.. Salinity Temperature Hybridization Yield Hybridization Yield..... x TAE Mg + x TAE Mg + x PBS x PBS M Na + Time(hr) o C o C o C o C Time(hr) Equilibrium Yield.. e f g h.. Equilibrium Yield. x TAE Mg + x TAE Mg + x PBS x PBS M Na + o C o C o C o C Rate Constant (M/s) Rate Constant (M/s) x TAE Mg + x TAE Mg + x PBS x PBS M Na + o C o C o C o C Discrimination Factor Q x TAE Mg + Discrimination Factor Q x TAE Mg + x PBS x PBS M Na + o C o C o C o C FIG. S: Temperature and salinity effects on probe performance. (a) and (e) show the hybridization yield over the first three hours of reaction for varying salinities and temperatures, respectively. (b) and (f) summarizes the fluorescence after hours of reaction. (c) and (g) summarizes the best-fit rate constants of the forward reaction (A + B D + E) to the kinetic traces shown in panels (a) and (e). (d) and (h) summarizes the discrimination factors Q observed at t = hr. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

12 DOI:.8/NCHEM. a b A B C C C n D E A B C C C n D E c A B C C C n D E Hybridization Yield A B C C C n D E Toeholds: nt/nt Toeholds: nt/nt Toeholds: 8nt/nt Toeholds: 8nt/nt.... Intended Intended Intended Intended.... SNP (m8cg) SNP (m8cg) SNP (m8cg) SNP (m8cg) Discrimination Factor Q nt/nt t = hr Annealed nt/nt 8nt/nt 8nt/nt FIG. S: Alternate designs for four-way toehold exchange, in which only a subset of the toeholds are used. Here, we only experimentally tested the design shown in panel (C), which uses exactly initiation toehold and dissociation toehold, and may be easier to prepare than probes with initiation and dissociation toeholds. This probe effectively discriminates intended target from SNP target, although with significantly slower kinetics than the design presented in the main paper. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

13 DOI:.8/NCHEM. a nt initiation toehold nt balancing toehold b nt initiation toehold nt balancing toehold Target CCAAGGTGGTGTGTGTAT GGTTCCACCACACACATA AGTGAG Correct (G-C, no change) m8cg: mutate to G-C to C-G AGTGAG Probe ROX ATACACACACCACCAACC TATGTGTGTGGTGGTTGG RQ Target CCAAGGTGGTGTGTGTAT AGTGA GGTTCCACCACACACATA AGTGA Correct (G-C, no change) m8cg: mutate to G-C to C-G Probe ROX ATACACACACCACCAACC TATGTGTGTGGTGGTTGG RQ nm Probe nm Probe Hybridization Yield.8 nm / Correct target.. 8.nM. nm / m8cg target.9nm nm / m8cg target.nm 8 Hybridization Yield.8. nm / Correct target.. nm / m8cg target nm / m8cg target 8.8nM.8nM.nM FIG. S8: Probes exhibit reduced specificity with toehold sequences cause G to be significantly negative. Here, the probes with nt initiation toeholds and nt dissociation toeholds show more reaction with nm m8cg SNP target than the default probes with nt initiation toeholds. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

14 DOI:.8/NCHEM. Correct i8ta: d8: m8cg: Target GGTGGTGTGTGTAT CCACCACACACATA AGTGA AGTGA (G-C, no change) insert T-A before G-C delete G-C mutate to G-C to C-G + Probe ATACACACACCACC TATGTGTGTGGTGG ROX RQ ATACACACACCACC AGTGATATGTGTGTGGTGG ATACACA CACCACC AGTGATATGTGT CTGGTGG + Correct + m8cg AGTGAATACACACACCACC TATGTGTGTGGTGG AGTGAATACACA G ACCACC TATGTGT G TGGTGG ΔG o = -.8 kcal/mol ΔG o = -. kcal/mol Hybridization Yield x TAE Mg nm Probe + nm Target Correct m8cg i8ta d8 x TAE Mg + nm Probe + nm Target Correct m8cg i8ta d8 M Na + nm Probe + nm Target Correct m8cg i8ta d FIG. S9: Double-stranded probes without dissociation toeholds react with both correct and SNP targets with thermodynamic favorability ( G < ). Experimental results across a variety of salinities demonstrate that that SNP target react signficantly with the probe, leading to a false positive signal. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

15 DOI:.8/NCHEM. Target Target a b Probe ROX Probe ROX GGTGGTGTGTGTAT CCAAGGTGGTGTGTGTAT ATACACACACCACCAACC ATACACACACCACC CCACCACACACATA GGTTCCACCACACACATA TATGTGTGTGGTGGTTGG TATGTGTGTGGTGG RQ RQ mutate to G-C mutate to G-C Correct (G-C, no change) Correct insert C-G insert A-T i8ta: insert T-A before G-C delete delete d8: delete G-C Correct m8cg: mutate to G-C to C-G 9 Discrimination Factor Q 8 AGTGA AGTGA mutate to C-G, T-A, A-T insert A-T, T-A, G-C, C-G delete H. Yield (χ) Q > after mins m8ta m8at m8cg d igc d8 i8at mgc d mgc i8gc i8ta i8cg icg 9 Discrimination Factor Q 8 AGTGA AGTGA Insertion (i8ta) H. Yield (χ) Mutation (m8cg) Deletion(d8) FIG. S: Comparison of the (a) toehold exchange probe with the (b) probe lacking balancing toeholds. The observed discrimination factor Q are plotted versus time. The values of Q quickly approached their equilibrium values, and in all cases were higher than after minutes. In contrast, the probe lacking balancing toeholds can be used to kinetically discriminate SNPs in the early timepoints of the reaction, but observed discrimination factors Q decrease as the reactions approach equilibrium. Note that even the highest Q observed for the probes lacking balancing toeholds is lower than corresponding Q values for the toehold exchange probes. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

16 DOI:.8/NCHEM. A Target CCAAGGTGGTGTGTGTAT AGTGA Correct i8ta: d8: m8cg: GAGCG ATACACACACCACCT CGCTC ROX RQ Beacon (G, no change) insert T before G delete G mutate to C GAGCG CGCTC CTATACACACACCACCTT ROX RQ Beacon8 B Hybridization Yield (χ) C Hybridization Yield (χ). Beacon8, C. Time (min) Time (min) Beacon, C Beacon, C Beacon, C. Time (min) D Hybridization Yield (χ).. Double-Stranded Toehold exchange Probe Beacon nm nm.um um 8uM Concentration of m8cg Time (min) FIG. S: Performance of two different molecular beacons on discriminating SNPs. Compare with Fig. b for performance of double-stranded probe. At elevated temperatures ( C), molecular beacons show improved specificity, but at the cost of hybridization yield. We also tested molecular beacons targeting a nt sequence, but these did not show significant difference between correct and SNP targets (data not shown). Beacon E. Coli rpob -8 GAACAACCCGCTGTCTGAGATTACGCACAAACGTCGTATCTCCGCACTCGG GAGCG ACGTTTGTGCGTAAT CGCTC Tye FQ Correct Mutate to T Hybridization Yield (χ). C. C FIG. S: Performance of molecular beacon targeting nt subsequence of the E. Coli rpob gene, nucleotides - 8. Compare with Fig. for performance of double-stranded probe on subsequences ranging from nt to 98 nt. Molecular beacon SNP discrimination performance is significantly worse than in the case of the synthetic targets (Fig. S8), because of secondary structure inherent in the target sequence. In contrast, double-stranded probes are not affected by secondary structure. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

17 DOI:.8/NCHEM. E. coli rpob gene and Rif-resistance regions Codon Nucleotide nt tested by Probe (ROX) nt tested by Probe(Tye) Codon 98 nt tested by Probe (ROX) Nucleotide 8 Wild Type No Mutation Wild Type No Mutation Rif-Resist # (ATC CTC) Rif-Resist # (ATC CTC) Rif-Resist # (CAC TAC) Time (hours) Rif-Resist # (CAC TAC) 8 Time (hours) FIG. S: Detection of Rif-resistance in E. coli using double-stranded probes; wild type and resistant colonies # and #. (Summarized in Fig. of main paper). NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

18 DOI:.8/NCHEM. E. coli rpob gene and Rif-resistance regions Codon Nucleotide nt tested by Probe (ROX) nt tested by Probe(Tye) Codon 98 nt tested by Probe (ROX) Nucleotide 8 Wild Type No Mutation Wild Type No Mutation Rif-Resist # (GAC GGC) Rif-Resist # (GAC GGC) Rif-Resist # (ATC CTC) Time (hours) Rif-Resist # (ATC CTC) 8 Time (hours) FIG. S: Detection of Rif-resistance in E. coli using double-stranded probes; wild type and resistant colonies # and #. NATURE CHEMISTRY 8 Macmillan Publishers Limited. All rights reserved.

19 DOI:.8/NCHEM. E. coli rpob gene and Rif-resistance regions Codon Nucleotide nt tested by Probe (ROX) nt tested by Probe(Tye) Codon 98 nt tested by Probe (ROX) Nucleotide 8 Wild Type No Mutation Wild Type No Mutation Rif-Resist # (GAC AAC) Rif-Resist # (GAC AAC) Rif-Resist # (GAC TAC) Time (hours) Rif-Resist # (GAC TAC) 8 Time (hours) FIG. S: Detection of Rif-resistance in E. coli using double-stranded probes; wild type and resistant colonies # and #. NATURE CHEMISTRY 9 Macmillan Publishers Limited. All rights reserved.

20 DOI:.8/NCHEM. E. coli rpob gene and Rif-resistance regions Codon Nucleotide nt tested by Probe (ROX) nt tested by Probe(Tye) Codon 98 nt tested by Probe (ROX) Nucleotide 8 Wild Type No Mutation Wild Type No Mutation Rif-Resist # (CAC (ACC TAC) AAC) Rif-Resist # (CAC (ACC TAC) AAC) Rif-Resist #8 (GAC (ACC AAC) GCC) Time (hours) Rif-Resist #8 (GAC (ACC AAC) GCC) 8 Time (hours) FIG. S: Detection of Rif-resistance in E. coli using double-stranded probes; wild type and resistant colonies # and #8. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

21 DOI:.8/NCHEM. Wild Type No Mutation Rif-Resist #9 (CAC (ACC E. coli rpob gene and Rif-resistance regions Codon Nucleotide nt tested by Probe (ROX) nt tested by Probe(Tye) Codon 98 nt tested by Probe (ROX) Nucleotide 8 (CAC TAC) AAC) Rif-Resist # TAC) Time (hours) Wild Type No Mutation Rif-Resist #9 (CAC (ACC (CAC TAC) AAC) Rif-Resist # TAC) 8 Time (hours) FIG. S: Detection of Rif-resistance in E. coli using double-stranded probes; wild type and resistant colonies #9 and #. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

22 DOI:.8/NCHEM. AGTGA AGTGA GAGTTCTTCGGTTCCAGCCAGCTGTCTCAGTTTATGGACCAGAACAACCCGCTGTCTGAGATTACGCACAAACGTCGTATCTCCGCACTCGGCCCAGGCGGTCTG CACC CTCAAGAAGCCAAGGTCGGTCGACAGAGTCAAATACCTGGTCTTGTTGGGCGACAGACTCTAATGCGTGTTTGCAGCATAGAGGCGTGAGCCGGGTCCGCCAGAC GTGG CCAC CAGACCGCCTGGGCCGAGTGCGGAGATACGA CGTTTGTGCGTAATC TCAGACAGCGGGTTGTTCTGGTCCATAAACTGAGACAGCTGGCTGGAACCGAAGAACTC GGTG GTGTGGCGGACCCGGCTCACGCCTCTATGCT GCAAACACGCATTAG AGTGACACGCCCAACAAGACCAGGTATTTGACTCTCTCGACCGACCTTGGCTTCTTGAG TCACTC TCACTC ACCAT CGGTCGCGTATGTCCAATCGAAACCCCTGAAGGTCCGAACATCGGTCTGATCAACTCTCTGTCCGTGTACGCACAGACTAACGAATACGG TCGG AGCC ACCATGCCAGCGCATACAGGTTAGCTTTGGGGACTTCCAGGCTTGTAGCCAGACTAGTTGAGAGACAGGCACATGCGTGTCTGATTGCTTATGCC GGCT CCGTATTCGTTAGTCTGTGCGTACACGGACA GAGAGTTGATCAGAC CGATGTTCGGACCTTCAGGGGTTTCGATTGGACATACGCGACCG ATGGTG CCGA GGCATAAGCAATCAGACACGCATGTGCCTGT CTCTCAACTAGTCTG GCTACAAGCCTGGAAGTCCCCAAAGCTAACCTGTATGCGCTGGC ATGGTG ACCAT TTCGGTTCCAGCCAGCTGTCTCAGTTTATGGACCAGAACAACCCGCTGTCTGAGATTACGCACAAACGTCGTATCTCCGCACTCGGCCCAGGCGGTCTGACCCGTGAACGTGCAGGCTTCGAAGTTCGAGACGTACACCCGAACGGTCGCGTATGTCCAATCGAAACCCCTGAAGGTCCGAACATCGGTCTGATCAACTCTCTGTCCGTGTACGCACAGACTAACGA CACC GTGG ACCATAAGCCAAGGTCGGTCGACAGAGTCAAATACCTGGTCTTGTTGGGCGACAGACTCTAATGCGTGTTTGCAGCATAGAGGCGTGAGCCGGGTCCGCCAGACTGGGCACTTGCACGTCCGAAGCTTCAAGCTCTGCATGTGGGCTGAGTGATGCCAGCGCATACAGGTTAGCTTTGGGGACTTCCAGGCTTGTAGCCAGACTAGTTGAGAGACAGGCACATGCGTGTCTGATTGCT CCAC TCGTTAGTCTGTGCGTACACGGACAGAGAGTTGATCAGA CCGATGTTCGGACCTTCAGG GGTTTCGATTGGACATACGCGACCGTAGTGAGTCGGGTGTACGTCTCGAACTTCGAAGCCTGCACGTTCACGGGTCAGACCGCCTGGGCCGAGTGCGGAGATACGACGTTTGTGCGTAATCTCAGACAGCGGGTTGTTCTGGTCCATAAACTGAGACAGCTGGCTGGAACCGAA GGTG AGCAATCAGACACGCATGTGCCTGTCTCTCAACTAGTCT GGCTACAAGCCTGGAAGTCC CCAAAGCTAACCTGTATGCGCTGGCATCACTCAGCCCACATGCAGAGCTTGAAGCTTCGGACGTGCAAGTGCCCAGTCTGGCGGACCCGGCTCACGCCTCTATGCTGCAAACACGCATTAGAGTCTGTCGCCCAACAAGACCAGGTATTTGACTCTGTCGACCGACCTTGGCTTATGGTG ATGGTG FIG. S8: Schematic of the targets and probes used in Fig. of the main paper. Probes are composed of four strands, rather than. The non-overlapping nicks in the probe are not expected to affect performance of the probes, except when the SNP is directly adjacent to the nick. Colored bases show subsequences of the rpob that are probed; black letters show primers binding sites for PCR amplification. Yellow background highlights complementary nonoverlapping nick region. Name Ecoli-rpoB-Colony-F Ecoli-rpoB-Colony-R Ecoli-rpoB-Region-F Ecoli-rpoB-Region-F Ecoli-rpoB-Region-R Ecoli-rpoB-Region-F Ecoli-rpoB-Region-F Ecoli-rpoB-Region-R Ecoli-rpoB-full-F Ecoli-rpoB-full-F Ecoli-rpoB-full-R Primer Sequence ATGATATCGACCACCTCGGCA ACGTAGTTGCCTTCTTCGATAGCAGACA AGTGA GAGTTCTTCGGTTCCAGCCAG TCACT GAGTTCTTCGGTTCCAGCCAG GGTG CAGACCGCCTGGGCC ACCAT CGGTCGCGTATGTCCAATC TGGTA CGGTCGCGTATGTCCAATC CCGA CCGTATTCGTTAGTCTGTGCGTACAC CACCAT TTCGGTTCCAGCCAG GTGGTA TTCGGTTCCAGCCAG GTG TCGTTAGTCTGTGCGTACAC TABLE S: PCR primers for E. coli-derived target. Shown in bold are initiation and dissociation toehold domains. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

23 DOI:.8/NCHEM. Sales Order Number: 8 Reference ID: 998 Manufacturing ID: 998 IDT Oligo ESI Analysis Calculated Molecular Weight:,8. Measured Molecular Weight:,8. Sequence Name: Oligo Sequence: Readout-R-Rox '- CTC ACT ATA CAC ACA CCA CCA ACC /Rox_N/ -' Results provided by Integrated DNA Technologies, Inc. July, FIG. S9: Representative Electrospray Ionization (ESI) mass spectrometry spectrum. The oligonucleotide analyzed here is the fluorophore-labeled strand of the probe. NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

24 DOI:.8/NCHEM. Sequence Name Oligo Sequence Calculated Measured Purification Molecular Weight Molecular Weight Readout-R-Rox CTC ACT ATA CAC ACA CCA CCA ACC /Rox N/,8.,8. HPLC WS-PrB/ /IAbRQ/GGT TGG TGG TGT GTG TAT TCA CTC,8.,8. HPLC WS-TT/ CCA AGG TGG TGT GTG TAT AGT GA,.,.9 PAGE WS-TB/ AGT GAA TAC ACA CAC CAC CTT GG,.,. PAGE WS-ST/-mc CCA AGG TGG TCT GTG TAT AGT GA,.,. PAGE WS-SB/-mg AGT GAA TAC ACA GAC CAC CTT GG,.,. PAGE WS-ST/-ma CCA AGG TGG TAT GTG TAT AGT GA,8.,9. PAGE WS-SB/-mt AGT GAA TAC ACA TAC CAC CTT GG,.,. PAGE WS-ST/-mt CCA AGG TGG TTT GTG TAT AGT GA,9.,. PAGE WS-SB/-ma AGT GAA TAC ACA AAC CAC CTT GG,.,. PAGE WS-TT/-ia CCAAGGTGGTAGTGTGTATAGTGA,8.9,88. PAGE WS-TB/-it AGT GAA TAC ACA CTA CCA CCT TGG,.8,. PAGE WS-TT/-ig CCA AGG TGG TGG TGT GTA TAG TGA,.9,. PAGE WS-TB/-ic AGT GAA TAC ACA CCA CCA CCT TGG,9.8,9. PAGE WS-TT/-ic CCA AGG TGG TCG TGT GTA TAG TGA,.9,. PAGE WS-TB/-ig AGT GAA TAC ACA CGA CCA CCT TGG,.8,. PAGE WS-TT/-it CCA AGG TGG TTG TGT GTA TAG TGA,8.9,9. PAGE WS-TB/-ia AGT GAA TAC ACA CAA CCA CCT TGG,.8,. PAGE WS-TT/-d CCA AGG TGG TTG TGT ATA GTG A,8.,8. PAGE WS-TB/-d AGT GAA TAC ACA ACC ACC TTG G,.,. PAGE WS-TT/-mg CCA AGG GGG TGT GTG TAT AGT GA,99.,. PAGE WS-TB/-mc AGT GAA TAC ACA CAC CCC CTT GG,9.,98. PAGE WS-TT/-ic CCA AGG CTG GTG TGT GTA TAG TGA,.9,. PAGE WS-TB/-i8g AGT GAA TAC ACA CAC CAG CCT TGG,.8,.9 PAGE WS-TT/-d CCA AGG GGT GTG TGT ATA GTG A,8.,8. PAGE WS-TB/-d AGT GAA TAC ACA CAC CCC TTG G,88.,88.8 PAGE WS-TT/mg CCA AGG TGG TGT GTG TGT AGT GA,9.,9. PAGE WS-TB/mc AGT GAA CAC ACA CAC CAC CTT GG,98.,98. PAGE WS-TT/-ia CCA AGG TGG TGT GTG TAA TAG TGA,8.9,8.9 PAGE WS-TB/-i8t AGT GAA TTA CAC ACA CCA CCT TGG,.8,. PAGE WS-TT/-d CCA AGG TGG TGT GTG TTA GTG A,8.,8. PAGE WS-TB/-d AGT GAA ACA CAC ACC ACC TTG G,9.,98. PAGE TABLE S: ESI summary for oligonucleotides used in Figure, Figure and Figure NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.

25 DOI:.8/NCHEM. Sequence Name Oligo Sequence Calculated Measured Purification Molecular Weight Molecular Weight wsr-probe-fl-ecoli-rpob /-ROXN/CCA CCA GAC CGC CTG GGC,.8,.9 HPLC CGA GTG CGG AGA TAC GA wsr-probe-fl-ecoli-rpob CGT TTG TGC GTA ATC TCA GAC AGC,.,. Ultramer GGG TTG TTC TGG TCC ATA AAC TGA GAC AGC TGG CTG GAA CCG AAG AAC TCT CAC TC wsr-probe-q-ecoli-rpob GAT TAC GCA CAA ACG TCG TAT CTC CGC,.,. HPLC ACT CGG CCC AGG CGG TCT GGT GG/IAbRQSp/ wsr-probe-q-ecoli-rpob CTC ACT GAG TTC TTC GGT TCC AGC 9,8.9 9,8.8 Ultramer CTG TCT CAG TTT ATG GAC CAG AAC CAG AAC CCG CTG TCT GA Ecoli-rpoB-probe-region-fl /TYE/GGC TCC GTA TTC GTT,.,. HPLC AGT CTG TGC GTA CAC GGA CA Ecoli-rpoB-probe-region-fl GAG AGT TGA TCA GAC CGA TGT TCG,9.,9. Ultramer GAC CTT CAG GGG TTT CGA TTG GAC ATA CGC GAC CGA TGG TG Ecoli-rpoB-probe-region-q TGA TCA ACT CTC TGT CCG TGT ACG,8.,8. HPLC CAC AGA CTA ACG AAT ACG GAG CC/IABkFQ/ Ecoli-rpoB-probe-region-q GTG GTA CGG TCG CGT ATG TCC AAT,8.,8. Ultramer CGA AAC CCC TGA AGG TCC GAA CAT CGG TC Ecoli-rpoB-Probe-full-FL /-ROXN/CCA CTC GTT AGT CTG,.,. HPLC TGC GTA CAC GGA CAG AGA GTT GAT CAG ACC GAT GTT CGG ACC TTC AGG Ecoli-rpoB-Probe-full-FL GGT TTC GAT TGG ACA TAC GCG ACC,88.,88. Ultramer GTA GTG AGT CGG GTG TAC GTC TCG AAC TTC GAA GCC TGC ACG TTC ACG GGT CAG ACC GCC TGG GCC GAG TGC GGA GAT ACG ACG TTT GTG CGT AAT CTC AGA CAG CGG GTT GTT CTG GTC CAT AAA CTG AGA CAG CTG GCT GGA ACC GAA ATG GTG Ecoli-rpoB-Probe-full-Q TCT GAT CAA CTC TCT GTC CGT GTA,8.,8.8 HPLC CGC ACA GAC TAA CGA GTG G/IAbRQSp/ Ecoli-rpoB-Probe-full-Q GTG GTA TTC GGT TCC AGC CAG CTG,.,. Ultramer TCT CAG TTT ATG GAC CAG AAC AAC CCG CTG TCT GAG ATT ACG CAC AAA CGT CGT ATC TCC GCA CTC GGC CCA GGC GGT CTG ACC CGT GAA CGT GCA GGC TTC GAA GTT CGA GAC GTA CAC CCG ACT CAC TAC GGT CGC GTA TGT CCA ATC GAA ACC CCT GAA GGT CCG AAC ATC GG TABLE S: ESI summary for oligonucleotides used in Figure NATURE CHEMISTRY Macmillan Publishers Limited. All rights reserved.