Team:NJU China/Project


<!DOCTYPE html> CSS-Only Responsive Layout with Smooth Transitions

Overview Chassis Targeting module Killing module Safety Achievement


Biomissile: a novel drug delivery system with exosome

Targeted drug delivery has always been a challenge in medicine. Current techniques using virus, bacterial or artificial liposome as delivery vehicles have toxicity and side effect on human bodies. It is urgently needed to develop a biocompatible and effective system to overcome the disadvantage of the currently available approaches. The aim of this project is, therefore, to design a novel targeted delivery system that can achieve both site-specific delivery and safety.

Exosomes are lipid bilayer vesicles naturally secreted by endogenous cells, playing crucial roles in transport of bioactive molecules between cells. Given their intrinsic role as natural transporter, exosomes potentially represent a novel and exciting drug carrier for therapeutic purpose. As therapeutic delivery agents, exosomes are biocompatible and will potentially be better tolerated by the immune system. Thus, exosomes may be capable of delivering drugs to local cellular environment without causing cytotoxicity and immune response.

Using the principles of synthetic biology, we aimed at building up a targeted drug delivery system using exosome, named ‘biomissile’. Our exosome have been reconstructed by two strategies, one is for outside modification and another is for inside modification.

Outside modification:
We engineered our chassis, 293T cells, to express a fusion protein comprising of the exosomal membrane protein Lamp2b and a receptor-binding short peptide. The short peptides designed for CD4+ T cell targeting is the HIV envelope glycoprotein gp120, for liver targeting is the pre-S1 of Hepatitis B Virus, and for neuron targeting is the rabies virus glycoprotein. Lamp2b can bring the peptide to the surface of exosomes, and the peptide can recognize specific host cells. Through outside modification, exosome will, in theory, has the ability to target specific cells, tissues and organs.

Inside modification:
We encapsulated our ‘kill device’ siRNA into exosomes by transfecting 293T cells with plasmids expressing siRNAs. siRNA is emerging as a promising therapeutic drug against a wide array of diseases and it functions to degrade mRNA through the RNA interference pathway. By designing siRNA against certain disease-related gene, we can use siRNA as drug for disease treatment.

Theoretically, by expressing targeting protein on the surface of exosomes, filling exosomes with siRNA drug and injecting the modified exosomes into the bloodstream, we will achieve specific siRNA delivery to the targeted cells, whereas non-specific uptake of siRNA in other tissues will be avoided. Indeed, the results showed that our delivery system successfully delivered siRNA to CD4+ T cells, liver and brain with high specificity and efficiency, and significantly blocked target gene expression in liver and brain. Thus, modification of exosomes derived from engineered cells may realize the goal of delivering siRNA to specific cellular environment.

On the other hand, we investigated the effect of our exosome delivery system on cell viability and immune response. CCK-8 assay and ELISA analysis showed that our exosomes did not influence cell viability and did not induce immune responses. Thus, our delivery system using exosomes is safe and non-toxic, and is compatible by our immune system.

Under the direction of the targeting peptide, exosome will fuse with specific cells, and then siRNA will destroy targeted mRNAs. Our exosome is just like the ‘biomissile’. The targeting peptide on the surface of exosome is the ‘GPS’ of the ‘biomissile’, and the siRNA within exosome is the ‘TNT’. The engineered 293T cell to continually produce modified exosomes is the ‘ordnance factory’. Our project will open up new avenues for therapeutic applications of exosomes in the future.


Exosomes are lipid bilayer vesicles that are secreted by all cell types, and its diameter ranges from 30nm to 100nm. Given its role as a natural transporter of bioactive molecules, we want to utilize exosomes as our drug carrier. The first problem we met is which chassis to choose to produce the exosomes we want. After screening through a large amount of different cell types, we choose to use HEK 293T cells as our chassis.

HEK 293T cell is a subtype of human embryonic kidney cells and we choose this as our chassis for three main reasons. The first reason is that HEK 293T cells can secrete large amounts of exosomes, so we can get enough exosomes by using HEK 293T cells as our chassis.

The second reason is that HEK 293T cells are derived from human, so the exosomes they secrete will be more human compatible and have little chance of inducing immune response compared to other non-human cells.

The last reason is that HEK 293T cells are immortalized cells, which means that after genetically engineering them to produce the exosomes we want, we can simply subculture the cell line and keep them as cell factory to produce our desired exosomes massively.

Targeting module

Specificity is one of the most vital goals that many different drug delivery system want to achieve, and that is also exactly what we are expecting from our biomissile.

To achieve specificity, we need to make our exosomes able to recognize distinctive sites within the body. In order to do that, the first step we took towards our exosome is outside modification. We wanted to add a target protein onto the surface of the exosomes to endow it with specificity, therefore we need to first find a protein naturally expressed on the surface of the exosomes and then fuse our target protein to the membrane protein of exosomes.

After a round of screen of the different membrane proteins on the exosomes’ surface, we chose to use lamp 2b to construct our fusion protein.

Lamp 2b (lysosomal associated membrane protein 2b) is a protein ubiquitously expressed on the surface of the exosomes. By genetically engineering a target protein to the outmembrane part of the lamp 2b, we can use the lamp 2b to bring our target protein to the surface of the exosomes. Thus we can endow the exosome with site-specificity by the addition of the fusion protein.

As the target protein in the lamp 2b is an exchangeable part, we can use different target protein to direct our exosomes to different parts of human body.

Liver,T cells and especially brain are poor targeting sites for previous drug delivery system. In order to prove that our target module does work, we choose to target these sites for testing.
To see more about Brain , Liver and T cell targeting.

Killing Module

Based on the utilization of natural exosome produced by HEK 293T cells and the modification of the surface protein, lamp-2b, now we have got a site-specific drug carrier, which can bring medicine to cells with a certain kind of receptors. In order to expand power of our system, we decide to pack the carrier with our disease-killing device, siRNA.

Recently, small interfering RNA (siRNA) is emerging as a promising therapeutic drug against a wide array of diseases. siRNA functions through the RNA interference pathway. Normal double-stranded RNA is first processed by Dicer and Argo to become short double-stranded RNA, which is about 21-25bp in length. Then it will recruit other proteins to form RISC( RNA induced silencing complex). One of the two RNA stands in the RISC will be degraded and the remaining strand can specifically recognize other mRNA by base pairing. Once the RISC bind to other complementary mRNA, it will destroy the mRNA through the RNAi pathway. By designing siRNA against certain virus genes, we can use siRNA as molecular medicine for diseases treatment. siRNAs are well tolerated and have suitable pharmacokinetic properties.

This discovery encouraged us to harness siRNA as specific targeted drugs producing a therapeutic benefit in our system. Now, not only can we carry drugs into a specific site, but also the drug itself specifically turn off the disease-causing gene expression rather than destroy the whole cell or disrupt normal protein production in a healthy cell.

We carried out this experiment in the HBV testing model to verify that our big idea can work in real human disease. The first step in designing a siRNA for viral gene silencing is to choose the siRNA target sites. Firstly, we should find 21 nt sequences in the target mRNA from Hepatitis b virus (HBV) genome that begin with an AA dinucleotide.
And then, choose target sites from among the ‘AA sequences’ based on guidelines like ‘Target sequence should have a GC content of around 50%’; ‘Avoid stretches of 4 or more bases such as AAAA, CCCC; ’‘Avoid regions with GC content <30% or > 60%.’
We completed the first two steps in the software, siRNA Designer.
Finally, we performed BLAST homology search to avoid off-target effects on other genes or sequences

After screening the HBV genome using the methods mentioned above, we ultimately find three candidates as our anti-HBV siRNA
siRNA 308 TATGCCTCAAGGTCGGTCGTT against HBx gene in HBV genome
siRNA 467 TCCCATAGGAATCTTGCGAAA against HBsAG gene in HBV genome
siRNA 516 ACAAATGGCACTAGTAAACTG against HBsAg gene in HBV genome
The number 308, 467 and 516 in the siRNA indicates their target sites within their target gene.

We cloned these three siRNAs into the vector pENTR/U6, which is a plasmid backbone for high yield of siRNA.
By cloning the siRNAs into eukaryotic expression vectors, large amounts of corresponding siRNAs can be produced by the cell, instead of chemically synthesizing the siRNAs.
In these vectors, we use an RNA polymerase Ⅲpromoter U6 to direct the transcription of siRNA, and an enhancer which greatly increases the gene transcription

1.siRNA screen
Since we have designed three types of siRNA target to different HBV genes, we need to determine which one would be optimal for HBV treatment.

To achieve this goal, we first tested their relative expression level in cells and exosomes. Expression of them was confirmed by quantitative PCR (qPCR) analysis of transfected HEK 293t cells and exosomes collected from the culture medium. The result suggested that 467 siRNA had much higher level of expression in both cells (Fig.1) and exosomes (36h post-transfection)(Fig.2). Since we need high yield of siRNA in the exosome, we decided to choose 467 as our ‘kill device’.

Fig.1 :qPCR analysis of relative siRNA level in 293t cells. Result showed that 467 siRNA has a relatively higher level of expression than that of 308 siRNA and 516 siRNA.

Fig.2 :qPCR analysis of relative siRNA level in 293t cells. 24 hours after transfection, 516 has a relatively higher expression level in exosomes. However, after 36h’s transfection, 467 performed a significant increase in expression level, almost 100 times of that of 516 and 308. Therefore, together with the data from Fig.1, we eliminated 516 siRNA and 308 siRNA from our list.

2.Silencing effect of siRNA 4667 towards HBsAG
We investigated the silencing effectiveness of siRNA 467 towards HBsAG to guarantee that it would work as we want. By qPCR analysis of HepG2 cell co-transfected with both siRNA 467 plasmid and HBsAg plasmid, we observed the significant down-regulation of HBsAg gene (Fig.3). This demonstrated that our siRNA 467 does function to silence the target gene, yet it could not prove that 467 siRNA would also work when it was encapsulated into our exosomes.

Fig. 3 :Silencing of HBsAg by siRNA 467. All groups were transfected by Lipo 2000. Control group was transfected with empty plasmid. siRNA-free group was transfected with HBsAg overexpression plasmid, and experimental group with both HBsAg overexpression plasmid and 467 siRNA plasmid. The result suggested that 467 siRNA successfully down-regulatethe HBsAg gene.

3.Silencing effect of siRNA 467 towards HBsAg after encapsulated in exosome
After transfecting 293t cells with 467 siRNA plasmids, we collected exsomes and then co-cultured them with HepG2 cells which had been transfected with HBsAg over-expression plasmid. qPCR analysis confirmed that the down-regulation of HBsAg was significant(Fig.4), and thus proved that our 467 siRNA could work effectively in exosomes.

Fig.4: All groups were transfected by Lipo. The first group was transfected without any plasmid while other three groups were transfected with HBsAg over-expression plasmid. Second group cultured without exogenous exosomes. Third group co-cultured with empty exosomes, and fourth with exosomes containing siRNA 467. These two kinds of exosomes were collected from 293t cells transfected without and with 467 siRNA plasmid, respectively. Then every group was analyzed for their HBsAg gene expression level through qPCR. Since group 2 and group 3 showed a similar level of expression, either of which is significantly higher than group 1, the HBsAg gene did express in those cells. Yet group 4, which was co-cultured with 467 exosomes, showed a relatively lower HBsAg expression level. That may indicate that 467 contained in exosomes had down-regulated the target gene.

4.Absolute quantification of siRNA 467 encapsulated into exosome
In order to better apply our exosomes in further experiment (mice experiment, for example), we needed to absolutely quantify the siRNA productivity in each μg of exosomes (whose quantity is measured base on protein concentration). By qPCR analysis of a series of siRNA with given concentration, we drew a standard curve (Fig.5). Then using this curve, we were able to determine the siRNA quantity within exosomes. Also, we tested the quantity of siRNA from both 24h post-transfection and 48h post-transfection in each experiment. The results showed that 24 hours after the transfection, the quantity of siRNA per 1μg exosome was 8*10-2 fmol in exosomes(Fig.6). However, after 48 hours, the number nearly doubled. It was 2*10-1 fmol(siRNA)/μg(exosome) in exosomes. The results helped us determine the dose of injection for mouse and establish the mathematical model.

Fig.5 siRNA 467 Standard Curve drawn through a series siRNA with given concentration.

Fig.6 Absolute quantification of siRNA 467 in exosomes, 24h and 48h post transfection, respectively.


Besides site-specific delivery of therapeutic drugs, safety is another essential point in clinical application. Current techniques using viruses or artificial liposomes as delivery vehicles are somewhat toxic and unsafe. The ability of exosome to transfer molecules within human body raises very exciting possibilities for therapeutic uses. As therapeutic delivery agents, exosomes are biocompatible and will potentially be better tolerated by the immune system because they are natural transporters derived from endogenous cells. Thus, exosomes engineered to deliver siRNAs may be capable of delivering siRNAs to local cellular environment without causing cytotoxicity and immune response.

To prove this concept, we conducted a series of experiments. First, we investigated whether our delivery system using exosomes will have an impact on cell viability. We evaluated the effects of exosomes on the viability of HepG2 cells using the CCK-8 assay. As shown in Figure 2, HepG2 cells treated with exosomes showed no reduction of cell proliferation. The result suggests that our delivery system using exosomes is safe and non-toxic.

Figure 1. MTT cell viability assay at 0, 12, 24, 36 and 48 h in HepG2 cells that were untreated or incubated with empty exosomes or modified exosomes.

Next, we investigated whether our delivery system using exosomes will induce immune responses. We assessed the levels of TNF-α, IL-6 and IL-8 in the cell culture medium when HepG2 cells were untreated or treated with empty exosomes or our modified exosomes. As shown in Figure 2, non-significant changes in the levels of these cytokines were observed, confirming that exosomes are immunologically inert and will not cause immune responses.

Figure 2. ELISA analysis of the levels of TNF-α, IL-6 and IL-8 in the supernatants of the HepG2 cells that were untreated or incubated with empty exosomes or modified exosomes.


Starting from March, our lab work continues for 7 months. During these 7 months, we first came up with the idea of ‘biomissile’ via brainstorming, then carefully designed every component of our biomissile. And the most difficult part was to experimentally prove that every part of the system can work properly. From the hard work of the past 7 months, we have successfully done the following:

We designed and created a new anti-HBV siRNA biobrick(BBa_K1180000), and experimentally validated that it can significantly suppress the HBV viral gene. Apart from that, after transfection of HEK 293T cells with this part, the siRNA can be successfully encapsulated into the exosmes produced by the cells.

We designed and constructed a liver-targeting fusion protein (BBa_K1180003) . After transfecting the HEK 293T cells with this plasmid, we monitored the exosomes with this fusion protein on its surface successfully get into the Hep G2 cell.

We designed and constructed a brain-targeting fusion protein(BBa_K1180002). After transfecting the HEK 293T cells with this plasmid, the exosomes produced by the HEK 293T cells can successfully bring the siRNA contained into the mouse brain.

Combining the targeting module and kill module together, firstly the modified exosomes can target specific sites, apart from that, the siRNA contained within the exosome can specifically destroy the disease related gene. Thus we realize the idea of boimissile by using our engineered exosome for target-destruction of disease.