Optimizing the Click

September 4, 2018

by Abhilasha "Abby" Dehankar

Goal: Optimization of the process used to attach single-stranded deoxyribonucleic acid (ssDNA) to quantum dots (QDs) surface. The QD-ssDNA conjugate can bind to a gold nanoparticle (AuNP) labeled with complementary ssDNA through DNA hybridization. The resulting QD-DNA-AuNP conjugate can undergo reversible quenching of QD fluorescence using a mechanism described in Figure 1. These photoswitchable QD-DNA-AuNP conjugates will allow us to employ QDs for STORM imaging.

 QD-DNA-AuNP conjugate reversible photoswitching mechanism
Figure 1: QD-DNA-AuNP conjugate reversible photoswitching mechanism

Review of progress-to-date:

1. We successfully transferred the organic phase QDs purchased from NN-labs to aqueous phase. We achieved the transfer by exchanging the non polar ligands on QD surface with polar phytochelatin-3 (PC3). The resulting QDs (QD-PC3) are water soluble and terminate with carboxylate (-COOH) and amine (-NH2) groups on their surface.

2. We failed to conjugate QD-PC3 to ssDNA terminated with -NH2 group using the “carbodiimide” chemistry. Carbodiimide chemistry forms an amide (-CONH2-) bond between the –COOH groups and -NH2 groups. We hypothesized that the chemistry failed for ssDNA because of its non-specific charge interactions with QD surface groups. As reported last month, we attempted to decrease these non-specific interactions by modification of ssDNA sequence (Mixed base-pair versus Poly-Thymine) and modification QD surface groups with uncharged, hydrophilic polyethylene glycol. Note: We could successfully conjugate QD-PC3 to small molecules with -NH2 terminations (ex: dye) using the same chemistry.

3. As an alternative approach we started exploring “click” chemistry for conjugation of QD-PC3 to ssDNA. Click chemistry differs from carbodiimide chemistry described above. To implement this approach, first we reacted the –COOH groups on QD-PC3 with small -NH2 terminated molecules called dibenzocyclooctyne (DBCO) using carbodiimide chemistry to form QD-DBCO conjugates. Next, we mixed the QD-DBCO conjugates with azide (-N3) terminated ssDNA to permit strain promoted alkyne - azide cycloaddition via a click reaction to generate QD-ssDNA conjugate. In the last blog, I reported that our first attempt at click chemistry was successful.

Monthly Update:

Increase in the number of ssDNA per particle, increases the number of sites available for complementary ssDNA hybridization per particle. From our prior experience with AuNP – DNA origami conjugates, we observed that increasing the number of hybridization sites per particle improves the thermal stability of the final conjugate (Fig 2A). Further, AuNP – DNA origami conjugates with higher thermal stability have shown higher yields. Therefore, we reasoned that increasing the number of ssDNA per QD-PC3 (DNA conjugation efficiency) could improve the thermal stability and yield of our final QD-DNA-AuNP conjugate (Fig 2B). In this blog, I report our attempts to increase the QD-ssDNA conjugation efficiency.

 ssDNA density per NP)
Figure 2: (A) AuNP-DNA origami conjugates  (Left: Low ssDNA density per NP; Right: ssDNA density per NP) (B) QD-DNA-AuNP composite (Left: Low ssDNA density per NP; Right: ssDNA density per NP)

As mentioned earlier, click chemistry is a two-step process (Fig 3):

Step 1: Conjugate -COOH groups on QD-PC3 to -NH2 terminated DBCO using carbodiimide chemistry to produce QD-DBCO conjugates. DBCO contains a strained alkyne group.

Step 2: Conjugate the strained alkyne group of QD-DBCO conjugate with azide terminated ssDNA using click reaction to form QD-ssDNA conjugates. The strain in alkyne group of DBCO promotes the cyclic addition reaction with -N3 groups on ssDNA.

Steps in click chemistry for QD-ssDNA conjugation
Figure 3:Steps in click chemistry for QD-ssDNA conjugation

The click reaction in step 2 is a bio-orthogonal reaction. This means the alkyne group of DBCO conjugated to QD-PC3 and the azide termination of ssDNA reacts exclusively with each other. As a result, all the alkyne-azide pairs will click over long incubation times. Therefore, in presence of excess azide terminated ssDNA, the number of ssDNAs conjugated per QD (QD-ssDNA conjugation efficiency) will be determined by the number of DBCO molecules conjugated per QD (QD – DBCO conjugation efficiency) and its yield. However, both, QD – DBCO conjugation efficiency and yield is dependent on efficiency of step 1. Consequently, we focused on optimizing the step 1.

Step 1 Optimization:

During step 1, –COOH groups on QD-PC3 react with –NH2 terminations of DBCO molecules through the carbodiimide chemistry. This involves activation of the –COOH groups on QD-PC3 with carbodiimide to form activated QD-PC3 (QD*), followed by reaction with –NH2 groups of target molecules. After completion of the reaction, the unconjugated target molecules are separated from the conjugates. We have previously optimized the activation step for conjugation of QD-PC3 to small molecules with –NH2 termination (Please refer to this blog post for more details). Therefore, today’s blog is focused on increasing the reactivity of QD* to target molecules.

We have observed that higher target molecule-to-QD* ratio result in improved QD – target conjugation efficiency because of the improved availability of the reactive –NH2 group. So, we decided to increase the DBCO (target molecule)-to-QD* ratio. However, DBCO is a charge neutral hydrophobic molecule. Consequently, increase in the QD – DBCO conjugation efficiency can make the QDs hydrophobic, leading to complete precipitation or lower yields. In contrast, as described earlier, high QD – DNA conjugation efficiency requires both, a high yield and a high QD – DBCO conjugation efficiency. Thus, use of DBCO establishes a fundamental limit on QD – DNA conjugation efficiency.

To determine optimal conditions for highest yield and QD-DBCO conjugation efficiency within the DBCO established fundamental limit, we evaluated three different DBCO-to-QD* ratios: 10, 100 and 1000. Successful QD-DBCO conjugation yields a peak in their absorbance spectrum at ~265 nm. As indicated by the absorption spectrum (Fig 4), successful conjugation was seen at DBCO-to-QD* ratios of 10:1 and 100:1. As expected, the DBCO-to-QD* ratio of 100:1 displayed higher conjugation efficiency as compared to 10:1, indicated by its higher absorbance. Unfortunately, QD-DBCO conjugates at the ratio of 1000:1 precipitated out of solution. Therefore, we selected the DBCO-to-QD* ratio of 100 as the optimal condition for all future experiments with DBCO as the intermediate molecule.

QD-DBCO conjugation absorption spectra

1 simply precipitated out of solution.
Figure 4:The DBCO peak shows successful conjugation of QD-to-DBCO. Higher absorbance values indicate higher efficiency of conjugation.  Conjugates at ratios of 1000:1 simply precipitated out of solution.

To overcome the fundamental limit introduced by the hydrophobicity of DBCO on QD – DNA conjugation efficiency, we decided to try “sulfoDBCO (sDBCO)” as the intermediate molecule. sDBCO is just a sulphate (SO4-) modified DBCO that is hydrophilic because of SO4- charge stabilization.  –NH2 terminated sDBCO can potentially conjugate to –COOH groups on the surface of QD-PC3 while preserving the hydrophilicity of the QD – sDBCO conjugates. Therefore, QD – sDBCO conjugation efficiency can be improved without QD precipitation concerns.

To test the QD-sDBCO conjugation, we performed step 1 of click chemistry with sDBCO-to-QD* ratio: 100 and 1000. Unlike DBCO, even the 1000:1 ratio of sDBCO-to-QD* formed stable QD-sDBCO conjugates and successfully surpassed the limitations of DBCO (Results not shown). Next, we conjugated both the QD-sDBCO conjugates to azide terminated ssDNA to test their conjugation efficiency. To test the conjugation results, a fluorescent dye modified azide-ssDNA was used for studying the efficiency of conjugation. After purification from excess, unconjugated ssDNA, a successful conjugation of ssDNA to QD displays fluorescence peaks from both, the QD and the dye, in the fluorescence spectra. Additionally, the intensity of fluorescence is directly proportional to the amount of QDs and dye present in the solution. Unfortunately, DNA did not conjugate at sDBCO-to-QD* ratio of 100 (Fluorescence data not shown). This indicates that under similar conditions, DBCO is more efficient than sDBCO. However, DNA conjugation at sDBCO-to-QD* ratio of 1000 was successful. As seen in Fig 5, for the same number of QDs (indicated by equivalent fluorescence peak of QDs), the number of DNA conjugated per QD was higher for QD-sDBCO (sDBCO: QD* = 1000) than QD-DBCO (DBCO: QD* = 100) (indicated by increase in the peak fluorescence intensity of dye in QD-ssDNA conjugates with sDBCO as the intermediate molecule). Therefore, although sDBCO is less efficient than DBCO, it can provide better DNA conjugation efficiency by allowing higher QD-sDBCO conjugation efficiency. Therefore, for future experiments, we will employ sDBCO as the activating molecule.

Comparison of conjugation by DBCO and sulfoDBCO

 100 (Equivalent QD peaks indicate the same concentration of QDs, while the increase in the dye peak for the sDBCO conjugates indicates an increase in the amount of DNA conjugated per particle)
Figure 5:QD-DNA conjugates with sDBCO: QD = 1: 1000 and DBCO: QD = 1: 100 (Equivalent QD peaks indicate the same concentration of QDs, while the increase in the dye peak for the sDBCO conjugates indicates an increase in the amount of DNA conjugated per particle)

Currently, we are testing the new QD-ssDNA conjugates with complementary ssDNA strands on specific locations of DNA origami hinges. Presence of QDs at desired locations on the DNA origami hinges will confirm that the ssDNA on QD is available for eventual binding to complementary AuNP-ssDNA conjugates. Future experiments will focus on formation and analysis of photo-switchable QD-DNA-AuNP conjugates.