DNA embedding - One step closer

January 15, 2019

by Kil Ho Lee

Making a DNA-embedded Quantum Dot


So far, we have tested a number of variables associated with the DNA embedding technique, including heating, QD to DNA ratio, pH, and centrifugal filtration. After embedding Cy5 fluorescent dye terminated - DNA strands on the QD surface (Figure 1), we performed fluorescence measurements on the QDs (λemission = 540 nm) and on the Cy5 (λemission = 664 nm). (see the (Nov 17th & Dec 7th, 2018) blog post “Toward DNA-embedded Quantum Dots: Take VI and VII (make this a link). These previously reported results indicated success embedding DNA, and it was reasonable to conclude that the DNA embedding procedure was reproducible.

Figure 1. Schematic of Cy5 terminated - DNA embedded CdSe/ZnS (Core/Shell) QDs

However, using centrifugal filtration, we’ve had difficulty in the purification (removal) of free DNA strands .  Without effective purification, we will have difficulty distinguishing the Cy5 fluorescence signals from the embedded DNA and the free DNA.  And, the continued presence of free DNA strands would impede the further hybridization of DNA-embedded QDs and gold nanoparticles (AuNP). Abhilasha Dehankar faced similar purificiation challenges, and she investigated a number of purification techniques including centrifugal filtration, column purification, and gel electrophoresis.


In this study, we tested using gel electrophoresis rather than centrifugal purification to separate DNA embedded QDs from free DNA strands, in order to both confirm the success of DNA embedding, and to test this alternative purification route.

In addition, we investigated the effect of two different buffers on the  centrifugal filtration (100kDa) process. , comparing the fluorescence intensity of both QDs and Cy5-terminated DNA strands after filtration using water versus PBS buffer (pH 7.2).

Result and Discussion

Gel Electrophoresis

In this experiment, we prepared a set of samples as listed in Table 1.

Table 1. Samples analyzed using Gel Electrophoresis


QD (0.3 mg mL-1)

DNA (100 nM)


Control 1. DNA only


100 µL


Control 2. QD only

100 µL



DNA embedded QD w/ Heating

100 µL

100, 200, 300 µL


DNA embedded QD w/o Heating

100 µL

100 µL


The final gel was analyzed by collecting an image screening for Cy5 fluorescence signal (λemission = 664 nm), shown in Figure 2.

Figure 2. Cy5 fluorescence signals (dark areas) from DNA embedded QDs and from free DNA strands

As expected, DNA only [lane 6] showed a high Cy5 emission and left no trace of Cy5 signal along the gel lane. QD only [lane 5] did not show any Cy5 emission. QDs embedded with Cy5 terminated DNA showed clear bands in the middle lanes [2, 3, 4], indicating successfully embedded DNA on the QD surfaces; however, high Cy5 emissions of free DNA strands were detected as well [4] This result explained the low Cy5 fluorescence emission detected using a fluorometer in earlier blog posts because only a small number of Cy5 terminated DNA strands were embedded on the QD surfaces. Lastly, as shown in the first lane of Figure 2, heating is a required step for embedding DNA because this result was nearly identical to the result of DNA only.


Centrifugal Filtration: washing with water versus PBS buffer (pH 7.2)

In this experiment, we investigated the effectiveness of removing un-embedded DNA strands using centrifugal filtration. Previously, 100kDa centrifugal filter devices were used to wash the QDs after the DNA embedding steps. In a typical experiment, 3 washing cycles were performed using distilled water. However, in recent trials washing DNA-only controls, we noticed a substantial level of Cy5 fluorescence signal even after repeated washing steps. This result may be attributed to DNA strands in a “stretched” configuration; the “stretched” DNA strands may be too “bulky” to pass through the filter. Hence, we tested the effectiveness of centrifugal filtration using water versus PBS buffer (pH 7.2); we hypothesized that PBS buffer with higher ionic strength than distilled water might reduce the electrostatic repulsion among free DNA strands, which could result in DNA strands balling up and passing through the filter easily. 

First, we prepared two solutions of QDs mixed with Cy5-terminated DNA in distilled water. Then, we used 100kDa centrifugal filter devices to wash the two solutions, one with distilled water and the other with PBS buffer (pH 7.2). We measured the fluorescence intensity of QDs and Cy5-terminated DNA before and after the washing steps. (Results shown in Figure 3.)

Cy5 emission from free DNA strands was lower after filtration using water (Figure 3(A)). Still, the noticeable peak indicated that only a portion of DNA strands were removed. However, Cy5 emission from free DNA strands after the filtration using PBS buffer, showed negligible fluorescence signal (Figure 3(B)). This result supported the hypothesis that the ionic strength of the solvent affects the electrostatic repulsion of DNA strands, causing balled-up DNA strands to pass through the filter easily.

Figure 3. Fluorescence intensity of QDs and Cy5 terminated DNA after 3 cycles of centrifugal filtration/washing using (A) water, and (B) PBS buffer (pH 7.2)

QD emission was also reduced after the centrifugal filtration. Interestingly, the degree of reduced fluorescence intensity of QDs was higher when QDs were washed using PBS buffer (pH 7.2). Loss of QDs in between each filtration step is expected because it is difficult to retrieve the QDs  trapped on the filter wall. Still, we believed the substantially reduced QD fluorescence after washing using PBS buffer (pH 7.2) was attributable to other factors.

As mentioned above, this particular experiment was performed using QDs mixed with DNA without applying the DNA embedding protocols. Hence, the surface of QDs are coated with 3-MPA. We hypothesize that PBS buffer causes charge screening, causing the negatively charged 3-MPA surface to lose its electrostatic potential. This may cause aggregation, followed by a higher degree of entrapment on the filter wall. Also, the reduced surface charge may impede the redispersion in water during QD retrieval step.

Furthering this analysis, we also decided to repeat the same study and to analyze the supernatant after each filtration step (Figure 4).

Figure 4. Fluorescence intensity of QDs after each washing step using PBS buffer: (A) retenate, and (B) supernatant.

We found that QD fluorescence is reduced in the retenate between each washing step using PBS buffer (pH 7.2) (Figure 4(A)). Surprisingly, the supernatants also showed QD emissions after each washing step, showing QD emission. that QDs are passing through the filter when they were washed with PBS buffer (pH 7.2). This contradicts our initial hypothesis that the reduced surface charge, which causes the aggregation, would result in QD entrapment on the filter. In fact, QDs readily passed through the filter. More investigation is required to understand this finding in the future.

Figure 5. Fluorescence intensity of Cy5 terminated DNAs after each washing step using PBS buffer: (A) retenate, and (B) supernatant.

Lastly, we found reduced Cy5 emission after filtering free DNA strands using PBS buffer, while the fluorescence measurement performed for supernatants showed increased Cy5 signal after the 2nd washing using PBS buffer (Figure 5(B)). This result agrees with the reduced Cy5 signal from the retenate (Figure 5(A)).


Conclusion and Summary

Based on the gel electrophoresis study, we are now confident that the embedding procedure is working well. We still need process optimization to maximize the number of DNA embedded on QD surface.

Based on the centrifugal filtration study, we found that PBS buffer (pH 7.2) is a better option for removing excess free DNA from DNA embedded QDs. However, the PBS buffer also induces more QDs to pass through the filter.

Next steps toward effective filtration include (1) testing the effect of PBS buffer (pH 7.2) on purifying DNA embedded QDs, and (2) running the gel electrophoresis on the retenate and the supernatant obtained from washing DNA embedded QDs using PBS buffer (pH 7.2)