Uncovering Molecular Quencher Effects on FRET Phenomena in Microsphere-Immobilized Probe Systems

Double-stranded (ds) oligonucleotide probes composed of quencher-dye sequence pairs outperform analogous single-stranded (ss) probes due to their superior target sequence specificity without any prerequisite target labeling. Optimizing sequence combinations for dsprobe design requires promoting a fast, accurate response to a specific target sequence while minimizing spontaneous dsprobe dissociation events. Here, flow cytometry is used to rapidly interrogate the stability and selective responsiveness of 20 candidate LNA and DNA dsprobes to a 24 base-long segment of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) RNA and ∼243 degenerate RNA sequences serving as model variants. Importantly, in contrast to quantifying binding events of dye-labeled targets via flow cytometry, the current work employs the Förster resonance energy transfer (FRET)-based detection of unlabeled RNA targets. One DNA dsprobe with a 15-base-long hybridization partner containing a central abasic site emerged as very stable yet responsive only to the SARS-CoV-2 RNA segment. Separate displacement experiments, however, indicated that ∼12% of these quencher-capped hybridization partners remain bound, even in the presence of an excess SARS-CoV-2 RNA target. To examine their quenching range, additional titration studies varied the ratios and spatial placement of nonquencher and quencher-capped hybridization partners in the dsprobes. These titration studies indicate that these residual, bound quencher-capped partners, even at low percentages, act as nodes, enabling both static quenching effects within each residual dsprobe as well as longer-range quenching effects on neighboring FAM moieties. Overall, these studies provide insight into practical implications for rapid dsprobe screening and target detection by combining flow cytometry with FRET-based detection.


S3
Screening candidates to select optimal dsprobe(s) S3 Table S1.Nomenclature, sequence, and function of LNA, DNA, and RNA oligonucleotides employed in this work.

S4
Rationale for probe design S6 Screening candidates to select optimal dsprobe(s) The complete list of LNA and DNA microsphere-immobilized probes, their hybridization partners, and RNA targets are provided in Table S1.The immobilized sequence choice is identical to a particular DNA probe for COVID-19 from the Centers for Disease Control website as detailed in Table S1 footnotes.To find a stable, yet selectively responsive dsprobe system to toehold-mediated displacment by SARS-CoV-2 RNA segment in Table S1, 20 different candidate quencher-capped hybridization partners were tested.The choice to incoporate a locked nucleotide at every third base position is based on displacement studies with LNA/DNA hybrids 1 .

LNA 21mQ
Quencher-capped LNA hybridization partners with central abasic site

21Q
Quencher-capped DNA hybridization partners The superscript "L" indicates a locked nucleotide in select FAM-functionalized probe and quenchercapped hybridization sequences; X = abasic nucleotide in select quencher-capped hybridization partners; B = C, G, or U; D = A, G, or U; and H = A, C, or U in a mixture of model RNA sequence variants, var_RNA, to SARS-CoV-2 RNA.Each dsprobe is comprised of a 5′ FAM moiety on the immobilized sequence and a 3′ quencher (Q) on its hybridization partner.The choice of DNA probe sequence is identical to the probe sequence named 2019-nCoV_N1-P on the Centers for Disease Control website (https://www.cdc.gov/coronavirus/2019-ncov/lab/rt-pcr-panel-primer-probes.html)accessed on February 11, 2022.

Rationale for dsprobe design and selection
The first dsprobe candidates consisted of LNA N1_FAM and quencher-capped LNA hybridization partners of three different lengths: 21 bases (i.e., LNA 21Q), 15 bases (i.e., LNA 15Q), and 9 bases (i.e., LNA 9Q).To first assess the maximum fluorescence signal from unquenched probes, Figure S1(a) shows single-stranded LNA N1_FAM probes alone and incubated with various unlabeled RNA sequences.As also shown in Figure S1(a) and in contrast to the relatively high fluorescence signal of unquenched probes, these three dsprobe systems all exhibited little to no fluorescence signal prior to SARS-CoV-2 RNA addition as well in the presence of any noncomplementary RNA.Following the addition of SARS-CoV-2 RNA, the longest 21 base-long LNA dsprobe showed no increase in fluorescence while the 15 base and 9 base-long LNA dsprobes showed only modest increases in fluorescence indicating little displacement by SARS-CoV-2 RNA occurs in these perfectly-matched LNA dsprobes.To promote additional selective displacement activity while maintaining stability, these studies were repeated using analogous quenchercapped LNA hybridization partners with a central base missing (i.e., abasic site).Probes initially hybridized to LNA 21mQ exhibited the same lack of fluorescence activity in the both the absence and presence of any RNA sequences.In contrast, a relatively high background fluorescence signal was found for LNA 9mQ indicating fewer dsprobes formed initially.While fluorescence further increased upon the addition of SARS-CoV-2 RNA, the increase in fluorescence also occurred in the presence of noncomplementary RNA.The most promising LNA dsprobe system employed was LNA 15mQ as it allowed for fluorescence signaling only in the presence of SARS-CoV-2 RNA; however, its observed fluorescence was significantly lower than that of single-stranded LNA N1_FAM in the absence of a quencher.Thus, exchange of quencher-capped LNA 15mQ with SARS-CoV-2 RNA appears incomplete.Overall, with the exception of LNA N1_FAM:LNA 9mQ exhibiting both higher background signal and less RNA target specificity, the affinity of LNA:LNA duplexes in the dsprobes explored here appears too high to be susceptible to extensive displacement by SARS-CoV-2 RNA.
Since DNA is a reportedly weaker hybridization partner compared to LNA [2][3][4][5][6] , the stability and responsiveness of LNA N1_FAM hybridized to quencher-capped 9, 11, 13, and 15 base-long DNA with a central abasic site were also examined.For all but the longest of these DNA hybridization partners (i.e., 15mQ), the background fluorescence signal and fluorescence response to negative RNA targets otherwise increased as the DNA hybridization partner base length decreased as shown in Figure S1(a) for 13mQ, 11mQ, and 9mQ, respectively.While the fluorescence signal for LNA N1_FAM:15mQ dsprobes is nearly double that of LNA N1_FAM:LNA 15mQ dsprobes in the presence of SARS-CoV-2 RNA, it is still well below the fluorescence signal of all the remaining dsprobes possessing shorter, mismatched DNA hybridization partners (i.e., 13mQ, 11mQ, and 9mQ) in the presence of SARS-CoV-2 RNA.Thus, while selective displacement activity appears enhanced by replacing the quencher-capped, 15 base-long, central abasic LNA hybridization partner with its equivalent DNA sequence, further improvements in detecting SARS-CoV-2 RNA appeared possible with stable, yet weaker affinity DNA:DNA sequence pairings in dsprobes.
Next, pure DNA dsprobes comprised of N1_FAM and various quencher-capped DNA hybridization partners were examined.As before with LNA dsprobes, first the fluorescence signal of unquenched N1_FAM before and after incubation with various RNA sequences is shown in Figure S1(b).Next, the background fluorescence signal is measured and determined to be negligible for all perfectly-matched DNA dsprobes in Figure S1(b).In comparison to the weaker fluorescence activity of perfectly-matched LNA dsprobe analogues in Figure S1(a) in the presence of SARS-CoV-2 RNA, a stronger dependence of fluorescence signaling on base length of 21Q, 15Q, and 9Q in the presence of SARS-CoV-2 RNA is evident in Figure S1(b).To further explore the ability to further improve SARS-CoV-2 RNA detection without compromising dsprobe stability, a series of quencher-capped DNA hybridization partners with a central abasic site was then explored.Unlike the perfectly-matched dsprobes, however, a trend of increasing background fluorescence in the absence of RNA is observable for 13mQ, 11mQ, and 9mQ, respectively indicating this series of imperfectly-matched dsprobes is increasingly susceptible to spontaneous dissociation.Thus, while additional fluorescence activity in the presence of SARS-CoV-2 RNA does occur in each of these three imperfectly-matched dsprobe system in Figure S1(b), their lower stability broadens their susceptibility to either dissociation (likely followed by RNA hybridization to unoccupied N1_FAM) or to displacement by var_RNA sequences that resemble, but do not entirely match SARS-CoV-2 RNA sequences.LNA 15mQ and LNA 9mQ as hybridization partners to this DNA probe in Figure S1

Experimental Details
Sensitivity limit in select dsprobe system N1_FAM and 15mQ were combined in a 1:2 ratio and diluted in TE pH 8.0 to a final concentration of 5 µM N1_FAM.The mixture was annealed by heating to 94 ºC for 2 min, then slowly cooling to room temperature.A working solution of microspheres was prepared by diluting 5 µL bead stock from 1% w/v to 0.1% w/v in wash buffer.Microspheres were washed 3 times by centrifuging at 14,000g for 3 min, aspirating supernatant, then resuspending in fresh wash buffer.Pre-washed 3.00 µm microspheres were incubated with the pre-annealed dsprobe mixture at a final concentration of 1 µM N1_FAM and 0.1% w/v microspheres, and agitated for 15 min at 22ºC, 800 rpm on a thermomixer.The microspheres were then washed 3 times as described above.Following the final wash, the coupled microspheres were diluted 1000-fold to a final concentration of 0.0001% w/v and incubated with SARS-CoV-2 RNA at final concentrations ranging from 1 µM down to 100 pM, and agitated for 15 min at 22 ºC, 800 rpm on a thermomixer.For flow cytometry, 20 µL was taken from each sample and added to 400 µL PBS.

Separate quantification of primary duplex formation and extent of 15m displacement by RNA
Unlabeled N1 and 15m_FAM were combined in a 1:2 ratio and diluted in TE pH 8.0 to a final concentration of 5 µM N1.The mixture was annealed by heating to 94 ºC for 2 min, then slowly cooling to room temperature.Working solution of microspheres was prepared by diluting 5 µL bead stock from 1% w/v to 0.1% w/v in wash buffer.Microspheres were washed 3 times by centrifuging at 14,000g for 3 min, aspirating supernatant, then resuspending in fresh wash buffer.Pre-washed 3.00 µm microspheres were incubated with the pre-annealed dsprobe mixture at a final concentration of 1 µM N1 and 0.1% w/v microspheres, and agitated for 15 min at 22ºC, 800 rpm on a thermomixer.The microspheres were then washed 3 times as described above.5 µL was taken from each triplicate sample of coupled-microspheres to be used as the pre-RNA sample for background fluorescence.The rest of each triplicate was split to be incubated with one of three different RNA solutions: SARS-CoV-2 RNA, var_RNA, or scr_RNA, each at a final concentration of 1 µM RNA and 0.1% w/v microspheres.Mixtures were agitated for 15 min at 22 ºC, 800 rpm on a thermomixer, then 5 µL was removed from each and added to 400 µL PBS for flow cytometry.

Titration experiments to determine if self-quenching of immobilized FAM occurs
For the experiments depicted in Figure 4(a), varying ratios of N1 to N1_FAM were diluted in TE pH 8.0 to a final concentration of 5 uM.For the experiment depicted in Figure 4(c), varying ratios of N1 to N1_FAM were combined in a 1:2 ratio with 15m_noQ and diluted in TE pH 8.0 to a final concentration of 5 µM probe.All mixtures were annealed by heating to 94 ºC for 2 min, then slowly cooling to room temperature.Working solution of microspheres was prepared by diluting 5 µL bead stock from 1% w/v to 0.1% w/v in wash buffer.Microspheres were washed 3 times by centrifuging at 14,000g for 3 min, aspirating supernatant, then resuspending in fresh wash buffer.Pre-washed 3.00 µm microspheres were incubated with the pre-annealed solutions at a final concentration of 1 µM probe and 0.1% w/v microspheres, and agitated for 15 min at 22ºC, 800 rpm on a thermomixer.The microspheres were then washed 3 times as described above.5 µL was taken from each triplicate sample of coupled-microspheres to be used as the pre-RNA sample for background fluorescence.The rest of each triplicate was split to be incubated with one of three different RNA solutions: SARS-CoV-2 RNA, var_RNA, or scr_RNA, each at a final concentration of 1 µM RNA and 0.1% w/v microspheres.Mixtures were agitated for 15 min at 22 ºC, 800 rpm on a thermomixer, then 5 µL was removed from each and added to 400 µL PBS for flow cytometry.

Titration experiments to determine if residual quencher species has longer range effects on neighboring FAM moieties
For the experiment depicted in Figure 5(a), varying ratios of 15m to 15mQ were diluted in TE pH 8.0, and mixed with N1_FAM at a ratio of 2:1, with a final N1_FAM concentration of 5 µM.These mixtures were annealed by heating to 94 ºC for 2 min, then slowly cooling to room temperature.Working solution of microspheres was prepared by diluting 5 µL bead stock from 1% w/v to 0.1% w/v in wash buffer.
Microspheres were washed 3 times by centrifuging at 14,000g for 3 min, aspirating supernatant, then resuspending in fresh wash buffer.Pre-washed 3.00 µm microspheres were incubated with the preannealed solutions at a final concentration of 1 µM probe and 0.1% w/v microspheres, and agitated for 15 min at 22ºC, 800 rpm on a thermomixer.The microspheres were then washed 3 times as described above.5 µL was taken from each triplicate sample of coupled-microspheres to be used as the pre-RNA sample for background fluorescence.The rest of each triplicate was split to be incubated with one of two different RNA solutions: SARS-CoV-2 RNA or var_RNA each at a final concentration of 1 µM RNA and 0.1% w/v microspheres.Mixtures were agitated for 15 min at 22 ºC, 800 rpm on a thermomixer, then 5 µL was removed from each and added to 400 µL PBS for flow cytometry.
For the experiments depicted in Figure 6(a), varying ratios of 15_noQ to 15Q were diluted in TE pH 8.0, and mixed with N1 at a ratio of 2:1, with a final N1 concentration of 5 µM.Separately, B15 (5' -ATC AGC CGC AAT CCA -3') and A24_FAM (5' -biotin -TTT TTT TTT TGG ATT GCG GCT GAT -FAM -3') were mixed at a ratio of 2:1, with a final A24_FAM concentration of 5 µM.All of these mixtures were annealed by heating to 94 ºC for 2 min, then slowly cooling to room temperature.Then, each N1:15_noQ/15Q sample was mixed with the A24_FAM:B15 mixture described above, in a 1:1 ratio of N1 to A24_FAM.Working solution of microspheres was prepared by diluting 5 µL bead stock from 1% w/v to 0.1% w/v in wash buffer.Microspheres were washed 3 times by centrifuging at 14,000g for 3 min, aspirating supernatant, then resuspending in fresh wash buffer.Pre-washed 3.00 µm microspheres were incubated with the pre-annealed solutions at a final concentration of 1 µM probe and 0.1% w/v microspheres, and agitated for 15 min at 22ºC, 800 rpm on a thermomixer.The microspheres were then washed 3 times as described above, then 5 µL was removed from each and added to 400 µL PBS for flow cytometry.

End-to-end length calculations to estimate separation distance range of various probe combinations
Scheme S1(a) shows the various possible combinations of immobilized probe and hybridization partner and the calculated end-to-end length of each.The ssDNA probe was modeled as flexible 24 base-long ssDNA; the dsDNA probe was modeled as a rigid 15 base-long dsDNA B-helix segment + a flexible 9 base-long ssDNA segment; the dsDNA:RNA was modeled as a rigid 24 base-long dsDNA:RNA A-helix; and the imperfect dsDNA:RNA was modeled as a rigid 11 base-long dsDNA:RNA A-helix segment + a semi-rigid 13 base-long dsDNA:RNA A-helix per Sim et al. 7 Scheme S1(b) shows estimated separation distances between upright pairs of heterogeneous nearest neighbor oligonucleotides.The estimated separation distance is based on the reported 8 2 nm separation distance between neighboring biotin-binding pockets on a given face of tetrameric streptavidin (modeled as a rectangle with two pairs of biotin-binding pockets on two opposing faces 8 ) and theoretical segment lengths 7,9 of unhybridized and hybridized DNA and RNA as illustrated in Scheme S1(a).

Effect of hybridization partner on N1_FAM fluorescence
In order to study the effects of different hybridization partners on the fluorescence of N1_FAM, microspheres coupled with N1_FAM were incubated with several different unlabeled RNA sequences.As shown in Figure S2, there was a slight decrease in fluorescence signal with the addition of SARS-CoV-2 RNA, but there was an increase in fluorescence signal with the addition of a truncated version of the sequence, SARS-CoV-2(15).The addition of several variant RNAs (var, var_0, and var_3) also induced a higher fluorescence signal, but not uniformly.These results indicate that the differences in fluorescence signal of these various N1_FAM:RNA duplexes were induced by the base-length and fidelity of the unlabeled hybridization partner.

Figure S1 .
Figure S1.Bar graphs of normalized molecules of equivalent soluble fluorochrome (MESF) of (a) LNA N1_FAM and (b) N1_FAM probe systems prior to RNA addition (gray series) and following the addition of SARS-CoV-2 RNA (green series), var_RNA (blue series), or scr_RNA sequence (red series) for 1.05 µm microspheres (note: the studies described in the main article used 3.00 µm microspheres).Each sample was performed in triplicate and error bars indicate standard deviation values.Based on their selective response to SARS-CoV-2 RNA, the three dsprobe systems yielding highest fluorescence measurements are marked as #1, #2, and #3, respectively, below each probe's nomenclature in the bar graphs above.
Scheme S1.(a) Various combinations of an individual N1_FAM DNA probe sequence unhybridized (left) and hybridized to a 15 base-long quencher-capped DNA sequence (middle left); a perfectlymatched RNA sequence (middle right); and an imperfectly-matched RNA variant sequence (right), each labeled with their respective, calculated end-to-end lengths.(b) Examples of pairwise heterogeneous combinations of a dsDNA (i.e., N1_FAM:15mQ) with a neigboring unhybridized N1_FAM (left); hybridized to perfectly-matched RNA sequence (middle); hybridized to imperfectly-matched RNA sequence (right).Each pair of neighboring oligonucleotides is labeled below with the 2.0 nm spacing between biotin binding sites on the same face of streptavidin and above with the estimated separation distance between two upright neighboring oligonucleotides.Though not illustrated in (b) the estimated separation distance between nearest neighbor ssDNA probe and imperfect dsDNA:RNA is 5.5 nm.

Figure S2 .
Figure S2.Bar graph of normalized molecules of equivalent soluble fluorochrome (MESF) of ssDNA N1_FAM probe systems prior to unlabeled RNA addition (gray) and following the addition of SARS-CoV-2 (green), 15-base SARS-CoV-2 (light green), model variant sequence mixture, var, and specific variant sequences, var_0 and var_3, (blue series), or scr sequence (red series).Each sample was performed in triplicate and error bars indicate standard deviation values.For simplicity, the term "RNA" is excluded in the sequence nomenclature in both the bar graph x-axis as well as the figure caption.