Limits...
Probing solid-state nanopores with light for the detection of unlabeled analytes.

Anderson BN, Assad ON, Gilboa T, Squires AH, Bar D, Meller A - ACS Nano (2014)

Bottom Line: Much progress has been made toward biotechnological applications; however, electrically probing the ion current introduces nonideal noise components.We show that by fine adjustment of the CaCl2 gradient, EGTA concentration, and voltage, the optical signals can be localized to the immediate vicinity of the pore.Consequently, the noise spectral density distribution in the optical signal exhibits a nearly flat distribution throughout the entire frequency range.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering Boston University Boston, Massachusetts 02215, United States.

ABSTRACT
Nanopore sensing has enabled label-free single-molecule measurements on a wide variety of analytes, including DNA, RNA, and protein complexes. Much progress has been made toward biotechnological applications; however, electrically probing the ion current introduces nonideal noise components. Here we further develop a method to couple an ionic current to a photon-by-photon counting of fluorescent signal from Ca(2+)-sensitive dyes and demonstrate label-free optical detection of biopolymer translocation through solid-state nanopores using TIRF and confocal microscopy. We show that by fine adjustment of the CaCl2 gradient, EGTA concentration, and voltage, the optical signals can be localized to the immediate vicinity of the pore. Consequently, the noise spectral density distribution in the optical signal exhibits a nearly flat distribution throughout the entire frequency range. With the use of high-speed photon counting devices in confocal microscopy and higher photon count rates using stronger light sources, we can improve the signal-to-noise ratio of signal acquisition, while the use of wide-field imaging in TIRF can allow for simultaneous quantitative imaging of large arrays of nanopores.

Show MeSH
Electrical and optical noise spectra. The electrical noise (a) exhibits low frequencies “flicker” noise characteristic to the nanopore system, as well as ∼f2 growing noise amplitude (“capacitance noise”), on top of the fundamental Johnson noise. (Inset) The electrical current histogram fit with a Gaussian distribution (mean = 6.57 ± 0.03 nA, width = 0.77 ± 0.07, χ2 = 2.25) shows deviations at the peak and tails. Optical noise (b) is spectrally flat. (Inset) Optical noise shows a near perfect fit to a Poisson distribution (mean = 5.55 ± 0.01 cnts/4 μs, χ2 = 0.96). The signal to noise ratios (SNR) for the optical signals (c) were evaluated as a function of the measurement bandwidth, as explained in the text, showing positive function of the total number of counts.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4334260&req=5

fig4: Electrical and optical noise spectra. The electrical noise (a) exhibits low frequencies “flicker” noise characteristic to the nanopore system, as well as ∼f2 growing noise amplitude (“capacitance noise”), on top of the fundamental Johnson noise. (Inset) The electrical current histogram fit with a Gaussian distribution (mean = 6.57 ± 0.03 nA, width = 0.77 ± 0.07, χ2 = 2.25) shows deviations at the peak and tails. Optical noise (b) is spectrally flat. (Inset) Optical noise shows a near perfect fit to a Poisson distribution (mean = 5.55 ± 0.01 cnts/4 μs, χ2 = 0.96). The signal to noise ratios (SNR) for the optical signals (c) were evaluated as a function of the measurement bandwidth, as explained in the text, showing positive function of the total number of counts.

Mentions: To compare the electrical and optical noise characteristics, synchronous photon counts and ionic current were acquired for 5 s at 300 mV using a 4.5 nm pore (Figure 4a,b insets). This data was used to calculate the power spectra (Figure 4a, blue and Figure 4b, red) for the electrical and optical signals, respectively. The electrical spectrum displays the characteristic 1/f noise contribution (“flicker noise”)25,27 at low frequency and a ∼f2 noise term above roughly 10 kHz, as well as some spurious electrical pickups and resonances (somewhat exaggerated due to openings in our Faraday box for the objective lens and stage controls) . This shape and the various contributions across the frequency domain have been considered in the literature by multiple groups.25−27,40 In contrast, the corresponding optical spectrum is virtually flat from 5 to 105 Hz, and specifically no spurious noise is observed. We further analyzed our signals by fitting the raw data as shown in Figure 4a,b, right insets. An ideal ion current signal would result in a perfectly normal distribution; however, careful analysis shows significant deviation from a single Gaussian (black curve) for the electrical signal at the peak of the function, where the Gaussian fit overestimates the data. We note that this deviation is the direct consequence of the shape of the electrical noise spectra shown in Figure 4a, specifically caused by the 1/f flicker noise. Since fluctuations in the signal mean are used to detect translocation events, these deviations directly reduce the quality of the electrical signal for translocation detection. In contrast, the optical signal, measured synchronously with the electrical current, is very well modeled by a Poisson distribution (black curve), a strong indication that the optical signal exhibits shot noise only.


Probing solid-state nanopores with light for the detection of unlabeled analytes.

Anderson BN, Assad ON, Gilboa T, Squires AH, Bar D, Meller A - ACS Nano (2014)

Electrical and optical noise spectra. The electrical noise (a) exhibits low frequencies “flicker” noise characteristic to the nanopore system, as well as ∼f2 growing noise amplitude (“capacitance noise”), on top of the fundamental Johnson noise. (Inset) The electrical current histogram fit with a Gaussian distribution (mean = 6.57 ± 0.03 nA, width = 0.77 ± 0.07, χ2 = 2.25) shows deviations at the peak and tails. Optical noise (b) is spectrally flat. (Inset) Optical noise shows a near perfect fit to a Poisson distribution (mean = 5.55 ± 0.01 cnts/4 μs, χ2 = 0.96). The signal to noise ratios (SNR) for the optical signals (c) were evaluated as a function of the measurement bandwidth, as explained in the text, showing positive function of the total number of counts.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4334260&req=5

fig4: Electrical and optical noise spectra. The electrical noise (a) exhibits low frequencies “flicker” noise characteristic to the nanopore system, as well as ∼f2 growing noise amplitude (“capacitance noise”), on top of the fundamental Johnson noise. (Inset) The electrical current histogram fit with a Gaussian distribution (mean = 6.57 ± 0.03 nA, width = 0.77 ± 0.07, χ2 = 2.25) shows deviations at the peak and tails. Optical noise (b) is spectrally flat. (Inset) Optical noise shows a near perfect fit to a Poisson distribution (mean = 5.55 ± 0.01 cnts/4 μs, χ2 = 0.96). The signal to noise ratios (SNR) for the optical signals (c) were evaluated as a function of the measurement bandwidth, as explained in the text, showing positive function of the total number of counts.
Mentions: To compare the electrical and optical noise characteristics, synchronous photon counts and ionic current were acquired for 5 s at 300 mV using a 4.5 nm pore (Figure 4a,b insets). This data was used to calculate the power spectra (Figure 4a, blue and Figure 4b, red) for the electrical and optical signals, respectively. The electrical spectrum displays the characteristic 1/f noise contribution (“flicker noise”)25,27 at low frequency and a ∼f2 noise term above roughly 10 kHz, as well as some spurious electrical pickups and resonances (somewhat exaggerated due to openings in our Faraday box for the objective lens and stage controls) . This shape and the various contributions across the frequency domain have been considered in the literature by multiple groups.25−27,40 In contrast, the corresponding optical spectrum is virtually flat from 5 to 105 Hz, and specifically no spurious noise is observed. We further analyzed our signals by fitting the raw data as shown in Figure 4a,b, right insets. An ideal ion current signal would result in a perfectly normal distribution; however, careful analysis shows significant deviation from a single Gaussian (black curve) for the electrical signal at the peak of the function, where the Gaussian fit overestimates the data. We note that this deviation is the direct consequence of the shape of the electrical noise spectra shown in Figure 4a, specifically caused by the 1/f flicker noise. Since fluctuations in the signal mean are used to detect translocation events, these deviations directly reduce the quality of the electrical signal for translocation detection. In contrast, the optical signal, measured synchronously with the electrical current, is very well modeled by a Poisson distribution (black curve), a strong indication that the optical signal exhibits shot noise only.

Bottom Line: Much progress has been made toward biotechnological applications; however, electrically probing the ion current introduces nonideal noise components.We show that by fine adjustment of the CaCl2 gradient, EGTA concentration, and voltage, the optical signals can be localized to the immediate vicinity of the pore.Consequently, the noise spectral density distribution in the optical signal exhibits a nearly flat distribution throughout the entire frequency range.

View Article: PubMed Central - PubMed

Affiliation: Department of Biomedical Engineering Boston University Boston, Massachusetts 02215, United States.

ABSTRACT
Nanopore sensing has enabled label-free single-molecule measurements on a wide variety of analytes, including DNA, RNA, and protein complexes. Much progress has been made toward biotechnological applications; however, electrically probing the ion current introduces nonideal noise components. Here we further develop a method to couple an ionic current to a photon-by-photon counting of fluorescent signal from Ca(2+)-sensitive dyes and demonstrate label-free optical detection of biopolymer translocation through solid-state nanopores using TIRF and confocal microscopy. We show that by fine adjustment of the CaCl2 gradient, EGTA concentration, and voltage, the optical signals can be localized to the immediate vicinity of the pore. Consequently, the noise spectral density distribution in the optical signal exhibits a nearly flat distribution throughout the entire frequency range. With the use of high-speed photon counting devices in confocal microscopy and higher photon count rates using stronger light sources, we can improve the signal-to-noise ratio of signal acquisition, while the use of wide-field imaging in TIRF can allow for simultaneous quantitative imaging of large arrays of nanopores.

Show MeSH