Electronic Machine Intelligence Lab (EMIL)

Priority: July 07, 2008

• [0093] The pores use a molecular adaptor to facilitate interaction with nucleotides.

•[0141] …molecular adaptor…facilitates the interaction between the pore and the nucleotides. The adaptor typically alters the charge of the barrel or channel of the pore or specifically interacts with or binds to nucleotides
•[0142] The adaptor preferably constricts the barrel or channel so that it may interact with the nucleotides.
•[0143] The adaptor is typically cyclic. The adaptor preferably has the same symmetry as the pore.

• [0149] The adaptor is covalently attached to the pore.
• [0150] The site of covalent attachment is selected such that the adaptor is positioned at or near residue 139 of SEQ ID NO:2. (This denotes the vertical position of the adaptor)

• [0179] Measure the current when the nucleotide interacts with the pore.

• [0187] An individual nucleotide is one not connected to any other nucleotide.
• [0190] A nucleotide may be derived from digestion of DNA or RNA.

• [0197] An example apparatus consists of two chambers separated by the pore carrying membrane.
• [0199] Voltage clamp is preferred to patch clamp. What is the reason for this preference?

• [0200] The nucleotide is present in a sample.
• [0201] The sample can be biological.
• [0202] The sample can be non-biological.
• [0203] The sample is typically assayed (e.g. by centrifugation or membrane filtering) to remove unwanted molecules.

• [0204] Applied voltages range from -400 mV to + 400 mV. More preferably in the range 120 mV to 170 mV. Increasing the voltage increases the discrimination between different nucleotides.
• [0205] Use an alkali metal chloride salt. KCl is preferred. A variety of salt concentrations are mentioned (0.15 - 2.5 M). High salt concentrations boost SNR.
• [0206] A buffer is also present in the aqueous solution. A suitable buffer is Tris-HCl. A preferred pH is 7.5.
• [0207] Preferable to carry out at 37 C (supports enzyme function). But can range from 0 to 100 C. Good nucleotide discrimination can be achieved at low salt concentrations if the temperature is increased.

• [0208] In one embodiment …

1. Digest nucleotide from one end of target sequence using exonuclease.
   
2. Contact nucleotide to pore.
   
3. Measure current passing through pore.
   
4. Repeat steps 1-3.
   

• [0210] Lengths of up to 500 nucleotides are noted.

• [0212] Attach an exonuclease to the pore. An exonuclease is a sequence snipping enzyme. It processes DNA like a ribbon sequentially chopping off (digesting) nucleotides one after another.
• [0215] Preferable to covalently attach the exonuclease to the pore.
• [0216] Preferable exonuclease rates are 1, 10, 100, 500, 1000 nucleotides per second.
• [0217] Exonuclease activity falls off as pH is reduced. Preferable pH is 7.5.
• [0218] Exonuclease activity dependent on certain metal ions such as magnesium.
• [0219] Exonuclease activity falls as salt concentration increases. Typical to use 0.15 to 0.8 M.

• [0235] Bilayers formed in 60-150-um diameter, 25-um thick, Teflon films.

• [0236] Axopatch 200B patch clamp amplifier used. 4-pole 10-kHz Bessel filter. Sampled at 20 kHz by PC equipped with Digidata 1440A A/D (Axon instruments) running ClampEx 10 software (Molecular Devices).
• [0237] Event histograms are constructed as follows…
• [0238] Run statistics on a moving window consisting of WT points.
• [0239] Calculate the T-statistc between adjacent WT windows (i.e. windows differing by only one sample).
• [0240] If the T-statistic difference between the windows, PT, is greater than some threshold TT a step is interpreted.
• [0241] Average the data between interpreted steps to determine its average current. Also keep track of its duration.
• [0242] Plot histograms of the mean event current from this data (over a long sequence I presume). An even is defined as “in limits” if the mean of the previous event was between set values LCD and UCD and if the duration was greater than N data points. • [0243] Typical values for event detection, producing a 4-nucleoside monophosphate histogram were WT=8, PT=3, TT=20, N=8, with LCD and UCD corresponding to the limits of the cyclodextrin level. So with a 10-kHz filtered and 20-kHz sampled signal the speed must be…
• [0244] Multiple Gaussian fitting is performed on the event histograms.
• [0245] Gaussian overlaps are calculated.

• [0261] In the 115 position in the <m>\beta</m>-barrel the baseline was too noisy and showed large current fluctuations. Maybe due to movement of cyclodextrin in barrel.
• [0263] Further down in 119 position baseline got cleaner, but still showed some spikes.

• [0275] Good baseline (Fig. 15) and nucleotide differentiation achieved with attachment in the L135C position (Fig. 16).
• [0276] Baseline is “clearly” better than any other position and “shows almost complete separation of all four nucleotides”.

• [0278] Baseline experiments to identify the four peaks undertaken (only put in one at at time). At 160 mV in 800 mM KCl

From Fig. 15 baseline current is around 67 pA.
• [0280] As you increase the applied voltage you get increase in passing currents and not necessarily equal shifts in the histograms (Fig. 18).
• [0281] Histogram overlab (at different voltages) looks pretty small (Fig. 19).
• [0282] The dTMP histogram looks the sharpest. Why?
• [0283] Because its dwell time is the largest. The longer (in time) the signal it generates the more samples we accrue and hence the more accurate our sampling of it (therefore the sharper the histogram). A summary of avg. dwell times in Fig. 20 at 130 mV and 800 mM KCl (just eyeballed the max. dwell times from Fig. 20).

Recall, sampling time is 20-kHz so every 0.05 ms. Avg. dwell time is 9.4 ms, so on average they are oversampling by 188 times.