Design

D E S I G N

"Design is the intermediary between information and understanding"

- Hans Hoffman

Overview

Our project, siREN, consists of two pillars: the intracellular targeting of key genes in the progression of CLL (by using siRNAs) and the extracellular targeting of CLL cells in the circulatory system (by employing lipid nanoparticles). Both of these pillars required extensive design, in order to maximize the efficiency and specificity of our therapeutic approach. In this page, we present the pipeline we designed and followed for both siRNA and LNP structure design.

Further in silico analysis for both pillars of our project were conducted using various modeling approaches (see our Model Page). Throughout the design process for our project, we contacted multiple stakeholders to receive feedback and define our next steps (see our Integrated Human Practices Page).

Intracellular Targeting

RNA interference (RNAi) therapeutics has become a powerful method to address multiple diseases, by directly targeting their genetic roots 1. In the field of oncology specifically, by targeting and silencing disease-related genes at the mRNA level, multiple mechanisms that cancer cells rely on to survive and multiply are shut down, and apoptosis in cancer cells can be restored.

Our therapeutic approach is based on meticulously designed siRNAs that target genes related to CLL progression. siRNAs are short, double-stranded RNA molecules that silence specific genes by triggering their target messenger RNA (mRNA) to be degraded, through RNA interference. They incorporate into a protein complex called the RNA-induced silencing complex (RISC), which uses the siRNA to find and destroy complementary mRNA sequences, thereby preventing protein production. As of August 2025, FDA has approved a total of 7 siRNA therapies for different diseases 2 3 and many siRNA therapies are being examined at clinical trials today.

An siRNA-based therapy for Chronic Lymphocytic Leukemia can also be developed, by firstly identifying protein-coding genes, which are critical for CLL progression, and afterwards, extensively designing siRNAs that target the gene mRNA. By preventing the production of such proteins, their functions are blocked and the pathways in which these proteins are involved and lead to the accumulation of CLL cells are shut down. The diagram below showcases the siRNA design process we constructed and followed.

One-Hot Encoing
Diagram 1: siRNA Design Pipeline by iGEM Thessaloniki 2025 Team

Target Identification

In order to identify which proteins we should target using siRNAs, we conducted an extensive literature review to understand the underlying mechanisms that lead to CLL progression and find which genes are essential for these mechanisms. Our goal was to target and silence these proteins at the mRNA level using siRNAs, and, eventually, disrupt the CLL-progression mechanisms.

Throughout our literature review, we identified 5 potential target proteins: BCL2, BTK, STAT3, ROR1 and MCL1. By revising our literature review findings and consulting experts (RNAi researchers and doctors specialized in hematological cancers), we concluded that we should target the BCL2 and BTK genes. We chose two target genes instead of one, because we deem that a dual-targeting therapeutic approach would make it harder for CLL cells to evade immune response and apoptosis. Below, we present our rationale behind choosing these genes as potential targets for our siRNA therapeutic approach.

Role of BCL2 in CLL Progression

B cell leukemia/lymphoma 2 (BCL2) has been found to be overexpressed in various hematological malignancies, including CLL, and its main role is to help CLL cells evade immune system responses and inhibit apoptosis. Many current CLL treatment methods target BCL2, in an effort to restore CLL cell apoptosis. Overally, BCL2 is considered a hallmark of CLL 4.

In the majority of CLL cases, mir-15a and mir-16-1 are downregulated due to the frequent deletion of chromosome 13q14, resulting in the overexpression of BCL2 5. These two micro-RNAs normally suppress BCL2 expression by binding to the BCL2 mRNA and inhibiting its production, however in the case of CLL where mir-15a and mir-16-1 are downregulated, BCL2 tends to be overexpressed.

BCL2 also inhibits mitochondrial apoptosis in CLL mainly by sequestering pro-apoptotic proteins 6. It binds to activated monomeric BAX and BAK, preventing their homo-oligomerization. Without BAX/BAK oligomers, mitochondrial outer membrane permeabilization (MOMP) cannot occur, the cytochrome c and other pro-apoptotic molecules are not released, and the apoptosome/caspase cascade is not activated. Furthermore, BCL2 binds to and sequesters BH3-only “activators” (such as BIM, BID and PUMA), that normally bind to BAX/BAK and trigger their conformational activation. BCL-2 neutralizes them by binding their BH3 domains via its hydrophobic cleft, so they cannot engage BAX/BAK.

One-Hot Encoing
Diagram 2: Role of BCL2 in CLL Progression

Role of ΒΤΚ in CLL Progression

Bruton’s tyrosine kinase (BTK) is a non-receptor tyrosine kinase that plays a central role in B cell receptor (BCR) signaling, which is one of the main drivers of CLL progression. Its primary function is to transduce survival and proliferation signals from the BCR and the CLL microenvironment. By doing so, BTK enables CLL cells to escape apoptosis and remain dependent on continuous BCR signaling for their survival. Consequently, BTK is considered a critical therapeutic target in CLL, and several inhibitors have been developed to block its activity 7.

In normal mature B-cells, the B-cell receptor (BCR) is composed of a membrane-bound immunoglobulin (Ig) receptor associated with the signaling subunits Igα (CD79A) and Igβ (CD79B). Under normal conditions, the BCR delivers two types of signals: a constitutive, antigen-independent survival signal mediated through PI3K, and a second signal triggered by antigen engagement. The latter promotes the recruitment and activation of intracellular signaling molecules such as LYN, SYK, Bruton’s tyrosine kinase (BTK), and PI3K, leading to downstream signaling cascades that result in cell activation and proliferation. Upon antigen binding, LYN and SYK kinases phosphorylate BTK. In CLL, BTK is often overexpressed, which enhances pro-survival signaling pathways and contributes to disease progression 8.

BTK functions as a critical downstream mediator in BCR signal transduction. During B-cell development, its activation leads to the stimulation of four key tyrosine kinase pathways: PLCγ2, MAPK, NF-κB, and the serine/threonine kinase AKT, all of which are essential for promoting B-cell survival and proliferation 7. When BTK activity is pharmacologically inhibited in CLL, the signaling interactions between leukemic cells and their supportive microenvironment are disrupted, thereby preventing the transmission of survival cues and promoting apoptosis in malignant lymphocytes 7.

One-Hot Encoing
Diagram 3: Role of BTK in CLL Progression

Designing siRNAs with siDirect

Before using the siDirect software tool to design our siRNAs, we retrieved both BCL2 and BTK mRNA sequences as FASTA files from GenBank. The BCL2 mRNA has 6 transcript variants: transcript variant alpha, transcript variant beta, transcript variant X1, transcript variant X2, transcript variant X3 and transcript variant X4, while the BTK mRNA has 3 transcript variants: transcript variant 1, transcript variant 2 and transcript variant 3.

Since both gene mRNAs have multiple transcript variants, we decided to design siRNAs using the most prevalent transcript variants of each gene. Therefore, we chose the BCL2 mRNA transcript alpha (GenBank Accession Number: NM_000633.3) and the BTK mRNA transcript variant 1 (GenBank Accession Number: NM_000061.3).

siDirect is an online software tool that can be employed to design highly-effective siRNAs, by inputting the target mRNA sequence in FASTA format and setting options based on the desired characteristics of the output siRNAs. The software provides as many siRNAs for the input mRNA sequence as possible and for each output siRNA, it provides: the guide strand (antisense), the passenger strand (sense), the target mRNA site and some sequence characteristics (such as seed-duplex stability Tm).

Before moving on to using siDirect, we conducted literature review to identify sequence features and rules that can enhance siRNA efficacy and minimize off-target effects. There are specific thermodynamic and sequence features via which we can predict whether a siRNA will be effective or not. Also, off-target siRNA effects primarily result from off-target mRNA sequence complementarity with the siRNA seed region 9. More information about the sequence features that affect siRNA efficacy is documented at our Model Page.

We set the siDirect options following rules that would produce siRNAs with better silencing activity. More specifically, we set that the produced siRNAs would follow rules by Ui-Tei et al. 10, Reynolds et al. 11 and Amarzguioui et al. 12. The following dropdown buttons contain the rules from each source.

  • A/U at 5′ end of antisense strand
  • G/C at 5′ end of antisense strand
  • At least five A/U bases within the 5′ one-third of antisense strand
  • No GC stretch of over nine nucleotides in length
  • 30–52 % G/C content
  • A at positions 3 and 19 of sense strand
  • U at position 10 of sense strand
  • Absence of G at position 13 of sense strand
  • At least three A/U at positions 15 to 19 of sense strand
  • Absence of G/C at position 19 of sense strand
  • No internal repeats
  • 32–58 % G/C content
  • Denser A/U content in 3′ end of sense strand than the 5′ end
  • G/C at position 1 of sense strand
  • Absence of U at position 1 of sense strand
  • A at position 6 of sense strand
  • A/U at position 19 of sense strand
One-Hot Encoing
Diagram 4: siRNA Design Pipeline using siDirect

The siDirect software tool produced a list of 21nt-long siRNAs targeting each mRNA sequence. Through our discussions with our team’s PI and Instructors, we decided to select siRNAs that target the mRNA coding regions (CDS). Therefore, we found the CDS regions for both BCL2 and BTK mRNAs on GenBank, and, finally, created a list of the produced CDS-targeting siRNAs for each gene. The tables below present the siRNAs produced by siDirect that target BCL2 and BTK mRNAs and follow the rules listed above.

Resulting BCL2-targeting siRNA
Sense (5′→3′): GGAUUGUGGCCUUCUUUGAGU Antisense (5′→3′): UCAAAGAAGGCCACAAUCCUC
Resulting BTK-targeting siRNAs
Sense (5′→3′): CUAUGAGUAUGACUUUGAACG Antisense (5′→3′): UUCAAAGUCAUACUCAUAGUA
Sense (5′→3′): CAGUAAGAAGGGUUCAAUAGA Antisense (5′→3′): UAUUGAACCCUUCUUACUGCC
Sense (5′→3′): GUAAGAAGGGUUCAAUAGAUG Antisense (5′→3′): UCUAUUGAACCCUUCUUACUG
Sense (5′→3′): CAAAUUUCAAUCAUUGAAAGG Antisense (5′→3′): UUUCAAUGAUUGAAAUUUGCU
Sense (5′→3′): CUACAUUCCUAGUAACUAUGU Antisense (5′→3′): AUAGUUACUAGGAAUGUAGCC
Sense (5′→3′): GUAUGAGUGGUAUUCCAAACA Antisense (5′→3′): UUUGGAAUACCACUCAUACAU
Sense (5′→3′): GAAAGAAGGAGGUUUCAUUGU Antisense (5′→3′): AAUGAAACCUCCUUCUUUCCC
Sense (5′→3′): CAUAUCCAGGCUCAAAUAUCC Antisense (5′→3′): AUAUUUGAGCCUGGAUAUGAG
Sense (5′→3′): CCAAAGUCAUGAUGAAUCUUU Antisense (5′→3′): AGAUUCAUCAUGACUUUGGCU
Sense (5′→3′): GCUCGAAACUGUUUGGUAAAC Antisense (5′→3′): UUACCAAACAGUUUCGAGCUG
Sense (5′→3′): CCAUAUGAGAGAUUUACUAAC Antisense (5′→3′): UAGUAAAUCUCUCAUAUGGCA

Checking for homology with other genes

The last step before the siRNA thermodynamic and sequence analysis was to corroborate that the produced siRNAs did not have significant homology with other genes, minimizing off-target effects and ensuring precise BCL2 and BTK mRNA targeting. For this purpose, we employed BLASTn (Standard Nucleotide Basic Local Alignment Search Tool), entered all siRNA antisense strands and searched for homology across transcript databases. We set the ceiling value for the e-value parameter to be 0.05. All siRNAs target all transcript variants of their target mRNA with an e-value of 0.003, except for two BTK-targeting siRNAs that did not target the BTK mRNA transcript variant 2.

Antisense - Transcript Transcript Variant Alpha Transcript Variant Beta Transcript Variant X1 Transcript Variant X2 Transcript Variant X3 Transcript Variant X4
UCAAAGAAGGCCACAAUCCUC 0.003 0.003 0.003 0.003 0.003 0.003
Table 1: Predicted e-values for BCL2 mRNA transcripts using BLASTn
Antisense - Transcript Transcript Variant 1 Transcript Variant 2 Transcript Variant 3
UUCAAAGUCAUACUCAUAGUA 0.003 0.003 0.003
UAUUGAACCCUUCUUACUGCC 0.003 0.003 0.003
UCUAUUGAACCCUUCUUACUG 0.003 0.003 0.003
UUUCAAUGAUUGAAAUUUGCU 0.003 0.003 0.003
AUAGUUACUAGGAAUGUAGCC 0.003 0.003 0.003
UUUGGAAUACCACUCAUACAU 0.003 0.003 0.003
AAUGAAACCUCCUUCUUUCCC 0.003 0.003 0.003
AUAUUUGAGCCUGGAUAUGAG 0.003 - 0.003
AGAUUCAUCAUGACUUUGGCU 0.003 - 0.003
UUACCAAACAGUUUCGAGCUG 0.003 0.003 0.003
UAGUAAAUCUCUCAUAUGGCA 0.003 0.003 0.003
Table 2: Predicted e-values for BTK mRNA transcripts using BLASTn

Further siRNA Modeling

We conducted further in silico analysis using computational tools to conduct thermodynamic and sequence analysis for each siRNA and we also developed different machine learning models to predict the siRNA efficacy. The goal of this further analysis was to rank siRNAs based on their characteristics and proceed to order the best candidate siRNAs for wet lab experiments. These results are included in our Model Page.

Extracellular Targeting

We next had to tackle the challenge of the specific delivery of our siRNAs into the target cells, without toxicity. Naked-siRNA delivery is limited by several barriers (membrane impermeability due to their negative charge and hydrophilic nature, enzymatic degradation, mononuclear phagocyte system entrapment, fast renal excretion, endosomal entrapment, off-target effects 22), leading to a short blood half-life of merely around 10 minutes 23 that hinders clinical application. To overcome these challenges, several modifications have been employed, and, in recent years, nanoparticles have shown promising results in siRNA delivery with minimum toxicity and off-target effects.

Lipid-based nanoparticles, polymer-based nanoparticles, PLGA nanoparticles, and inorganic nanocarriers have all been and continue to be explored for the delivery of siRNA to the target site, showing in vivo stability, target specificity, and internalization of the siRNA to the cytosol with less cellular toxicity and minimum immune response. To date, only Lipid Nanoparticles (LNPs) have progressed past clinical trials to enter the pharmaceutical market, with the notable example of FDA-approved ONPATTRO (Patisiran): an siRNA therapeutic for hereditary transthyretin-mediated amyloidosis, encapsulated in a novel LNP carrier.

However, even with a carrier, when the drug enters systemic circulation, there are various barriers that it needs to overcome to reach the target cell (intravascular degradation (by plasma nucleases and lysosomal nucleases), renal clearance, plasma protein binding, mononuclear phagocyte system entrapment, membrane impermeability), and more that need to be surpassed for the siRNA payload to reach the cytoplasm and exert its gene silencing function (endosome entrapment) 22. These are some of the challenges that need to be accounted for when designing a nanoparticulate carrier, as off-target effects are extremely susceptible to siRNA doses and may be the primary source of toxicity at the large dosages necessary due to delivery limitations 13.

Furthermore, it is widely known that excessive liver homing caused by the discontinuous nature of the hepatic vasculature or the lipid metabolism 24 is a notable disadvantage of intravenously injected ionizable LNPs; creating a reservoir limiting the bioavailability of the drug and potentially causing a hepatotoxicity issue. That is why the implementation of LNP-mediated RNAi in disseminated diseases such as CLL requires precision in the form of targeted LNPs 25.

Because the efficient delivery of the therapeutic agent is crucial for the success of a treatment, we conducted an extensive literature review and consulted field experts to propose a nanocarrier system employing dual targeting strategies: conjugated Fab’ targeting ROR1, a CLL-specific surface marker, and LNP structure or administration optimized to enhance secondary lymphatic tissue tropism.

ROR1 Expression in CLL

ROR1, an oncofetal receptor tyrosine kinase-like orphan receptor, is composed of a distinguished extracellular region with one immunoglobulin domain (Ig), one frizzled domain (Fz), and one kringle domain (both cysteine-rich domains), followed by a transmembrane region and an intracellular region that contains a tyrosine kinase domain 16.

ROR family proteins have different functions in cell aggregation, differentiation, angiogenesis, and migration, the processes involved in metabolism and cell signaling during embryogenesis, playing important roles in the early development of skeletal, cardio-respiratory and nervous systems. In humans, the expression of ROR1 is predominantly limited to embryo and fetus and is mostly absent in adult tissues devoid of a malignancy 17.

However, ROR1 overexpression has been reported in many human cancers 17 as a driving force of tumor progression, resistance to therapy (NF-κB pathway), tumor cell survival (MAPK/ERK and JAK/STAT pathways) and cell proliferation (AKT/PI3K and JAK/STAT pathways) via the WNT/ROR1 signaling pathway axis. In CLL, ROR1 overexpression results from the loss of regulatory miRNA-15a and miRNA-16-1: the same mechanism that deregulates our target, BCL2 19.

Figure 1: ROR1 Receptor Structure

Since 2001, ROR1 has been identified as a specific marker for CLL 19, expressed in over 95% of CLL cells but in negligible amounts of healthy donor B cells 18, with an increasing expression upon CLL progression 19. ROR1 is expressed in immature B cells and tonsillar B cells but is mostly absent in normal blood cells 24 (only 3.7% of mature B-cells, 0.3% of T-cells, 1.4% of monocytes, 1.6% of granulocytes, and 1.9% of NK-cells naturally express ROR1) 17. The restricted expression of ROR1 protein on B-CLL cells clearly distinguishes ROR1 from other hallmark molecules commonly used as targeting moieties, such as CD20 and CD52, whereas its stably low percentage in sera from CLL patients distinguishes it from the considerably elevated and variable soluble CD52 and CD23 16. This points to ROR1 as a promising agent for cell-specific targeting, with less side effects than existing formulations.

Cell surface expression of ROR1, ranging from 3 to 5 times above background, is homogenous, uniform, and constitutive: independent of anatomic niches, biological or clinical heterogeneity of B-CLL, B-cell activation, or the mutational status of key prognostic indicators (ZAP70, IGHV) 16. Interestingly, ROR1 expression remains consistently high during and after chemotherapeutic treatment 17. All of that evidence supports the notion that ROR1 may be a reliable and consistent agent to facilitate targeted therapeutic approaches, at any stage of the disease, and regardless of previous treatment administration. Furthermore, higher ROR1 expression and signaling in CLL is associated with a worse prognosis 19 and with drug resistance (against venetoclax 15 and ibrutinib 19), further justifying the necessity of a new drug targeting it for therapeutic agent delivery.

The lower density of ROR1 molecules on the cell surface provides the possibility to reduce immune responses, including cytotoxicity mediated by antibodies and complement. In 30-45% of cases, cell surface ROR1 can mediate at least partial internalization of bound antibody by endocytosis starting 30 minutes after incubation and reaching maximum level of internalization capacity after 2 hours 17, with the antibody binding capacity determined to be in the range of 1000 to 10000 molecules per cell 16.

For all the aforementioned reasons, ROR1 is the target of many newly developed therapeutic molecules 17 19, several of which are currently being investigated in clinical trials. Importantly, studies by Chiang et al. (2019) and Mao et al. (2015) have demonstrated that the use of ROR1-targeted nanoparticles facilitates strong internalization and selective accumulation in ROR1-expressing leukemic cells of encapsulated miR-29b or cytotoxic agent OSU-2S, while control formulations show limited specificity. The ROR1-targeted approaches spared healthy donor B-cells and did not accumulate in the liver and kidneys of model animals in contrast with controls. Moreover, the targeting moieties did not show cytotoxicity as a single reagent or in the presence of antibody or liposomal cross-linkers, further underscoring the potential of ROR1-mediated targeting for precise therapeutic delivery in CLL 18 20.

Fab’ Selection

Since ROR1 is highly expressed on the surface of CLL cells and possesses all the aforementioned qualities, we decided to conjugate on our lipid nanoparticles a ROR1-targeting antibody that could ensure selective uptake by malignant cells. For this purpose, we turned to Fab’ antibody fragments directed against the extracellular domain of ROR1, and more specifically, against the outermost domains of the extracellular domain of ROR1 (the Ig and Fz domains).

Because this pillar of our project is currently proposed on a theoretical and modeling basis, and we do not move forward with experimental validation at this stage, we did not design a completely new Fab’ sequence from scratch. Instead, we searched for ROR1-specific antibodies reported in the literature and patents. From these sequences, we retrieved the complementarity-determining regions (CDRs) and the framework regions (FRs) of the heavy and light chains. Using these elements, we reconstructed the Fv fragment, which represents the variable portion of an antibody responsible for antigen recognition.

By conducting extensive literature review, we came across a patent that contained a total of 13 ROR1 Fabs 14 21. For our design, we considered the three Fabs whose dissociation constants were in the single digit nanomolar range: ERR1-TOP43, XBR1-402, and ERR1-306. After retrieving their CDRs and FRs, we reconstructed the Fv fragments using ABodyBuilder2 and moved on to model their structures and simulate their interaction with the extracellular domain of ROR1. Following these steps, we identified the most promising candidate to advance in our design pipeline.

Once the Fv fragment was defined, the next step was to construct a Fab’ fragment. This fragment consists of the Fv region linked to the constant domains of the antibody heavy and light chains, while preserving the hinge region. The presence of free thiol groups in the hinge is crucial, as they allow site-specific conjugation of Fab’ fragments onto lipid nanoparticles.

Choosing Lipid Nanoparticle Carriers

Due to constraints in time and funding, we did not proceed with wet lab experiments to validate our nanocarriers. Given that the field of nanoparticles for therapeutic use is relatively new with significant research gaps in the literature, as well as because there is a vast variety of possible components and compositions that typically arise through experimental screening, we –along with the experts we consulted– have decided to propose two different possible nanocarriers for the purposes of our project, which will need to be validated experimentally for our project implementation. Additionally, we chose to adopt an FDA-approved and therefore well documented and studied base composition, which we could then optimize for both our delivery approaches with more certainty that what we propose will be able to be translated into a clinical application in the future.

Out of the various nanoparticle categories, lipid nanoparticles (LNPs) are the leading delivery vehicle for RNA therapeutics, with many therapeutics in clinical trials, approved anti-SARS-CoV-2 vaccines, and, the only FDA-approved LNP-siRNA nanomedicine (ONPATTRO® (Patisiran)). RNA-encapsulating LNPs confer an efficient biological effect, close to 100% payload encapsulation, are biologically compatible as well as biodegradable, and possess a desirable amphipathic nature. Most importantly, lipids and polymer-based approaches are the most effective delivery systems for siRNA targeting specific sites, with minimal toxicity. Of these, only lipid nanoparticles have been approved for clinical use to date, demonstrating safety, efficacy and scallability, which is why we and the stakeholders we consulted decided to pursue an LNP approach in our project 22 26 28 32.

LNP-RNA systems form via hydrophobic interactions in an aqueous environment combined with electrostatic interactions between negatively charged RNA and cationic or ionizable lipids. Though LNP-RNA formulations initially used cationic lipids to allow electrostatic interactions with RNA, their toxicity (via interactions with negatively charged plasma proteins) have instigated a gradual shift to ionizable lipids. Ionizable lipids are positively charged at low pH (which allows RNA binding) and become neutral at physiological pH, a change that helps reduce the toxicity of LNP-RNA complexes in vivo. Currently, the LNPs that are most widely used in systemic nucleic acid delivery typically contain ionizable cationic lipids, helper (structural) amphipathic phospholipids, cholesterol and polyethylene glycol (PEG) lipids. As explained previously, we chose to adopt the base components of ONPATTRO Patisiran, a formulation that has been dubbed the gold standard for RNA-LNP therapy 26.

DLin-MC3-DMA is the main component of the FDA-approved formulation of Patisiran, an ionizable cationic lipid which is thus proven to be safe for use in humans 34. Within the context of Dlin-MC3-DMA nanoparticles, the DMA head group grants to the formed particles a nitrogen to phosphate ratio (N/P) of 3 to 6 depending on the formulation parameters, values optimal for siRNA encapsulation and gene silencing in vivo. Design of ionizable cationic lipids is balanced twofold using their pKa value: the pKa should be high enough that at low endosomal pH, the ionizable lipids gain a positive charge that allows interactions with the endogenous anionic lipids: thus leading to the disruption of the endosomal structure and the cytoplasmic release of LNP cargo. Simultaneously, the pKa value of the ionizable lipids should be sufficiently low so at physiological pH the surface charge of LNPs will remain relatively neutral. This allows to modulate toxicity and immunogenicity of resulting LNPs and increases their circulation time 28. The pKa of Dlin-MC3-DMA is 6.2 to 6.5, meaning that the lipid molecules will be relatively neutral at physiological blood pH, but will be protonated inside the increasingly acidic endosome turning into a lysosome. That results in an increased proton sponge effect over time, leading to the quick rupture of the endosome and siRNA release 37.

Polyethylene glycol (PEG) provides a protective coating around the nanoparticle surface, stabilizing it, shielding it from the immune system, phagocytes, and plasma proteins, avoiding non-specific binding as well as slowing renal clearance 22. It also plays an important role in the particle’s morphology, as PEG content is inversely proportional to LNP size 28. However, the presence of PEG on LNP surfaces may induce anti-PEG IgM production, especially after repeated administrations. Because excessive immune response against PEGylated LNPs is detrimental for gene delivery, faster shedding from LNPs after administration is beneficial in this regard 28. The DMG-PEG used in Patisiran provides initial coverage before it detaches from the nanoparticle, reducing immunogenicity and improving transfection efficiency. Because of this fact, however, we chose to conjugate our Fab’ fragments on DSPE-PEG, which stays more stably associated on the LNP surface, providing a stable linkage between LNP and Fab’. DSPE-PEG is, for the same reason, more suitable for intramuscular administration of the LNP, as circulation time in the lymphatic fluid is much slower than in the blood and therefore more taxing on nanoparticle stability 13 30 34 35.

Cholesterol as a component enhances LNP stability and cell entry, promotes endosomal escape, and shields against degradation and renal filtration. Cholesterol additionally contributes to the speed at which siRNA release occurs, a factor crucial for efficient gene silencing. Liposomes consisting of 17% cholesterol can release over 90% of their encapsulated cargo within 5 minutes upon endocytosis. Seeing as this LNP/siRNA has 38.5% cholesterol, the siRNA release is expected to be even faster. 22 28 34 37.

Helper lipid DSPC, as the name suggests, aids in the formation of LNP structure, as well as in surpassing membrane impermeability 22.

Use of this formulation, and especially of Dlin-MC3-DMA, has been reported in various blood malignancies. A study validated DLin-MC3-DMA lipid-based LNP-siRNA in vitro in Chronic Myeloid Leukemia (CML) cells and in vivo in leukemic mice, targeting the bone marrow with high efficiency and lack of toxicity 29. In another study, antibody-functionalized LNP-siRNA comprising of DLin-MC3-DMA, DSPC, cholesterol, PEG-DMG, and PEG-DSPE-maleimide, targeted Mantle Cell Lymphoma (MCL) –a B-cell malignancy– in vitro and in vivo in SCID mice 31.

In general, primary B cells such as our target CLL cells are notoriously hard to transfect in vivo. However, in recent years, nanotechnology-based therapeutic strategies have piqued oncologists' interest due to their potential to provide a new paradigm to overcome challenging curative issues in CLL patients. At least theoretically, nanocarriers may have the potential to deliver an adequate amount of drug to malignant CLL cells and overcome some overactivated multidrug resistance mechanisms. To treat leukaemia, the diameter of nanoparticles should be below 100 nm, which is in alignment with Patisiran’s hydrodynamic diameter of 60-100nm. Other properties of this formulation are desirable for a delivery system, such as its PDI index of less than 0.2, a close to neutral z potential, and nearly 100% siRNA encapsulation efficiency 22 26 27 37. The in vitro proof of concept came from a study where CLL cells were transfected with siRNA-LNPs of the FDA-approved lipid formulation of Dlin-MC3-DMA (MC3), cholesterol, DSPC, and PEG-DMG, mixed at a molar ratio of 50:38:10.5:1.5, proving that transfection of B-CLL cells by these LNPs is possible 26.

However, in vivo applications need to surpass additional challenges. Excessive liver homing due to the organ’s morphology, due to cholesterol, and due to the opsonization by plasma proteins implicated in lipid metabolism such as ApoE, is a notable disadvantage of intravenously injected ionizable LNPs. Researchers have proposed that a selective organ targeting (SORT) strategy could overcome this issue. Along with standard LNP components including an ionizable cationic lipid, phospholipids, cholesterol and PEG, the authors proposed adding SORT molecules which allow lung-, spleen- or liver-specific gene delivery. The rationale was to tune efficacious LNP formulations without destroying the core 4-component ratios that are essential for mediating RNA encapsulation and endosomal escape. Researchers tested numerous molecules with different charges, including a negatively charged 1,2-dioleoyl-sn-glycero-3-phosphate (18PA) SORT molecule that, at 10-40% incorporation in an LNP formulation, resulted in exclusive, spleen-specific delivery. Incorporating SORT molecules into several classes of ionizable LNPs –including DLin-MC3-DMA– produced similar outcomes, indicating this system is compatible with commonly used ionizable lipid-LNPs and can be altered according to therapeutic goals. We decided to incorporate this discovery in our first group of LNPs, to enhance spleen tropism and target the B-CLL cells accumulating in that organ. The results of the study are promising, as at 30% implemented 18PA, the transfection efficiency in the spleen was 40%, and 12% of transfected cells were B-cells: a percentage that we can only assume will increase depending on the dose and the addition of the targeting Fab’ 28 32.

As for the transport of particles to the lymph nodes (LN), the intramuscular route is favourable in comparison with the iv route, and has the added benefit of increasing patient autonomy and decreasing discomfort. For transitivity, a diameter of less than 100nm is necessary, but, while 20–50 nm seems to be the optimal range for direct transport, larger diameters are able to facilitate longer retention of the particles in the lymph node. This way, there is sustained release of the drug thanks to the depot effect. For designing LN-targeting nanoparticles, their charge has a substantial influence on the transport to the LNs via the lymphatic route in addition to their size. Efficient absorption via the lymphatics has been reported in cases of neutral or, especially, anionic charged nanoparticles, which have shown efficient translocation to the cortex and paracortex, namely, B cell and T cell zones, in the LNs. Interestingly, large amounts of PEG simultaneously hinder transportation but enhance retention in the lymph nodes. Therefore, efficient delivery and retention in the lymph nodes is highly dependent on a fine balance of size, charge, and PEG content of the nanoparticle 32 33 34.

Aside from the innate targeting properties of our nanoparticles, they are going to be functionalized with anti-ROR1 Fab’, to improve cell-specificity, and that needs to be considered when designing the LNP composition. The researchers that have used the formulation we’ve chosen to target B cells in MCL, conjugated antibodies via maleimide-thiol chemistry. As is logical and common because of the greater stability of DSPE-PEG, the antibodies were conjugated on DSPE-PEG-maleimide chains. These comprised a small percentage of the total PEG chains, and a mere 0.05% of the total LNP content. Total PEG content was 2% of the total LNP content, with the non-antibody-conjugated chains being exclusively DMG-PEG. That is a larger percentage than the one used in Patisiran, which will have an impact on size (predictably smaller) and other properties of the nanoparticle, making it more suitable for hematologic malignancies. The percentage offset is made by decreasing Cholesterol content from 38.5 to 38%, which is not projected to have any negative consequences, as this percentage is already far larger than the minimum requirement for efficient transfection. This approach was the one we implemented in our first group of intravenous, spleen-targeting LNPs, with the percentage of each component adjusted to account for the addition of SORT molecule 18PA, which was incorporated at 30% as per the relevant publication. The final percentage of each component was also externally validated by experts Dr. Pippa and Dr. Demetzos, and is shown in Table 3 31 32.

Figure 2: Fab' Functionalized LNP Structure

The concentration of DSPE-PEG-maleimide used in the aforementioned paper on MCL was also kept in the second group of intramuscular, lymph-node targeting LNPs. However, in that case, we chose to replace DMG-PEG with the more stable DSPE-PEG, to account for the longer circulation time in the lymphatic fluid and thus to increase the particle’s half-life, as previously explained. Additionally, we chose to limit the total PEG content to 1.5%, as a slightly larger thermodynamic diameter can help increase retention in lymph nodes and the depot effect. The composition of the second group of LNPs was also corroborated with experts Dr. Pippa and Dr. Demetzos, and is shown in Table 3.

LNP GROUP DLin-MC3-DMA DSPC Chol DMG-PEG DSPE-PEG DSPE-PEG-MAL 18PA
SPLEEN-TARGETING 35% (50%) 7% (10%) 26.6% (38%) 1.365% (1.95%) – 0.035% (0.05%) 30%
LYMPH NODE-TARGETING 50% 10% 38.5% – 1.45% 0.05% –

Table 3: Final concentrations of each LNP component for each of our two LNP groups 31 32 34.

The components of these LNPs were visualized with modeling approaches, as well as the complete bilayer of the first group of LNPs. The next steps include experiments to determine their thermodynamic properties (z potential, thermodynamic diameter, polydispersity index, morphology, stability), and to proceed with either fine-tuning or in vitro experiments in our target cell lines, to validate these theoretically proposed carriers.

siRNA Modifications

Studies have shown that aside from the carrier composition, modifications of the siRNA itself play a role in the efficacy of the drug. For therapeutic applications, the most useful siRNA modifications are a limited number of terminal phosphorothioate linkages (which increase the half-life of the agents, enhance the target knockdown, and are necessary for the rest of the modifications to be functional, but have been implicated in hepatotoxicity and local injection site reactions), and several 2′-O-methyl (which enhance activity and stability, aid in immune system evasion, avoidance of intravascular nucleases and decelerated renal filtration) and/or 2′-F-nucleotides (which enhance activity and stability, and attenuate the immune response) scattered throughout the sequence. 2′-OMe substitution specifically at position 2 of the guide strand. Furthermore, it has been demonstrated that replacing uridine (U) with naturally occurring pseudouridine (Ψ) attenuates the innate immune response, aids in avoiding cleavage by RNase, and reducing RNA-dependent protein kinase (PKR) activity. In contrast to the antisense technology, a high degree of backbone modifications renders siRNA oligonucleotides unable to activate the RNAi machinery, so there is a fine balance that must be kept 13 22 24 28. The way to do this is usually through screening of a modified siRNA library, which we were not at liberty to accomplish in the context of the iGEM competition.

After our consultation with Dr. Stathopoulos (see our Integrated Human Practices Page), we were directed to consider a modification pattern similar to the FDA-approved siRNA therapeutic Inclisiran. Throughout the siRNA structure of this drug, there are 32 ribonucleotides chemically modified with 2-O-methyl (2′-O-methyl), 11 modified with 2′-fluoro-ribonucleotide (2′-fluoro) and one 2′-deoxy-ribonucleotide. In addition, two phosphorothioate groups are added to the 5′ end of the sense and both the 5′ and 3′ ends of the antisense strands 36. All these insights from the literature and from expert input will serve as the guideline which will direct the design of our modified siRNA sequence for in vivo applications, in the next experimental steps of our project.

  1. Scientific American Custom Media. (2024, December 12). The power and promise of RNAi therapeutics. Scientific American. https://www.scientificamerican.com/custom-media/alnylam/the-power-and-promise-of-rnai-therapeutics/
  2. Padda, I. S., Mahtani, A. U., Patel, P., & Parmar, M. (2024, March 20). Small interfering RNA (siRNA) therapy. In StatPearls [Internet].StatPearls Publishing. https://www.ncbi.nlm.nih.gov/books/NBK580472/
  3. Lu, D., Dou, F., & Gao, J. (2025). Fitusiran: The first approved siRNA therapy for hemophilia via reducing plasma antithrombin levels. Drug Discoveries & Therapeutics. https://doi.org/10.5582/ddt.2025.01031
  4. Buggins, A. G., & Pepper, C. J. (2010). The role of Bcl-2 family proteins in chronic lymphocytic leukaemia. Leukemia Research, 34(7), 837–842. https://doi.org/10.1016/j.leukres.2010.03.011
  5. Cimmino, A., Calin, G. A., Fabbri, M., Iorio, M. V., Ferracin, M., Shimizu, M., Wojcik, S. E., Aqeilan, R. I., Zupo, S., Dono, M., Rassenti, L., Alder, H., Volinia, S., Liu, C., Kipps, T. J., Negrini, M., & Croce, C. M. (2005). miR-15 and miR-16 induce apoptosis by targeting BCL2. Proceedings of the National Academy of Sciences, 102(39), 13944–13949. https://doi.org/10.1073/pnas.0506654102
  6. Del Gaizo Moore, V., & Letai, A. (2012). BH3 profiling – Measuring integrated function of the mitochondrial apoptotic pathway to predict cell fate decisions. Cancer Letters, 332(2), 202–205. https://doi.org/10.1016/j.canlet.2011.12.021
  7. Singh, S. P., Dammeijer, F., & Hendriks, R. W. (2018). Role of Bruton’s tyrosine kinase in B cells and malignancies. Molecular Cancer, 17(1), Article 57. https://doi.org/10.1186/s12943-018-0779-z
  8. Tambaro, F. P., De Novellis, D., & Wierda, W. G. (2021). The role of BTK inhibition in the treatment of chronic lymphocytic leukemia: A clinical view. Journal of Experimental Pharmacology, 13, 923–935. https://doi.org/10.2147/JEP.S265284
  9. Kobayashi, Y., Tian, S., & Ui-Tei, K. (2022). The siRNA off-target effect is determined by base-pairing stabilities of two different regions with opposite effects. Genes, 13(2), 319. https://doi.org/10.3390/genes13020319
  10. Ui-Tei, K., Naito, Y., Takahashi, F., Haraguchi, T., Ohki-Hamazaki, H., Juni, A., Ueda, R., & Saigo, K. (2004). Guidelines for the selection of highly effective siRNA sequences for mammalian and chick RNA interference. Nucleic Acids Research, 32(3), 936–948. https://doi.org/10.1093/nar/gkh247
  11. Reynolds, A., Leake, D., Boese, Q., Scaringe, S., Marshall, W. S., & Khvorova, A. (2004). Rational siRNA design for RNA interference. Nature Biotechnology, 22(3), 326–330. https://doi.org/10.1038/nbt936
  12. Amarzguioui, M., & Prydz, H. (2004). An algorithm for selection of functional siRNA sequences. Biochemical and Biophysical Research Communications, 316(4), 1050–1058. https://doi.org/10.1016/j.bbrc.2004.02.157
  13. Abbasi Dezfouli, Saba, Marcel E. Michailides, and Hasan Uludag. "Delivery Aspects for Implementing siRNA Therapeutics for Blood Diseases." Biochemistry 63.23 (2024): 3059-3077.
  14. Haiyong Peng, Thomas Nerreter, Jing Chang, Junpeng Qi, Xiuling Li, Pabalu Karunadharma, Gustavo J. Martinez, Mohammad Fallahi, Jo Soden, Jim Freeth, Roger R. Beerli, Ulf Grawunder, Michael Hudecek, Christoph Rader, Mining Naïve Rabbit Antibody Repertoires by Phage Display for Monoclonal Antibodies of Therapeutic Utility, Journal of Molecular Biology, Volume 429, Issue 19, 2017, Pages 2954-2973, ISSN 0022-2836, https://doi.org/10.1016/j.jmb.2017.08.003.
  15. Hallek, M. (2025). Chronic lymphocytic leukemia: 2025 update on the epidemiology, pathogenesis, diagnosis, and therapy. American Journal of Hematology, 100(3), 450-480.
  16. Baskar S, Kwong KY, Hofer T, Levy JM, Kennedy MG, Lee E, Staudt LM, Wilson WH, Wiestner A, Rader C. Unique cell surface expression of receptor tyrosine kinase ROR1 in human B-cell chronic lymphocytic leukemia. Clin Cancer Res. 2008 Jan 15;14(2):396-404. doi: 10.1158/1078-0432.CCR-07-1823. PMID: 18223214.
  17. Aghebati-Maleki L, Shabani M, Baradaran B, Motallebnezhad M, Majidi J, Yousefi M. Receptor tyrosine kinase-like orphan receptor 1 (ROR-1): An emerging target for diagnosis and therapy of chronic lymphocytic leukemia. Biomed Pharmacother. 2017 Apr;88:814-822. doi: 10.1016/j.biopha.2017.01.070. Epub 2017 Feb 1. PMID: 28160756.
  18. Chiang CL, Goswami S, Frissora FW, Xie Z, Yan PS, Bundschuh R, Walker LA, Huang X, Mani R, Mo XM, Baskar S, Rader C, Phelps MA, Marcucci G, Byrd JC, Lee LJ, Muthusamy N. ROR1-targeted delivery of miR-29b induces cell cycle arrest and therapeutic benefit in vivo in a CLL mouse model. Blood. 2019 Aug 1;134(5):432-444. doi: 10.1182/blood.2018882290. Epub 2019 May 31. PMID: 31151986; PMCID: PMC6676131.
  19. Tigu AB, Munteanu R, Moldovan C, Rares D, Kegyes D, Tomai R, Moisoiu V, Ghiaur G, Tomuleasa C, Einsele H, Gulei D, Croce CM. Therapeutic advances in the targeting of ROR1 in hematological cancers. Cell Death Discov. 2024 Nov 17;10(1):471. doi: 10.1038/s41420-024-02239-1. PMID: 39551787; PMCID: PMC11570672.
  20. Mani R, Mao Y, Frissora FW, Chiang CL, Wang J, Zhao Y, Wu Y, Yu B, Yan R, Mo X, Yu L, Flynn J, Jones J, Andritsos L, Baskar S, Rader C, Phelps MA, Chen CS, Lee RJ, Byrd JC, Lee LJ, Muthusamy N. Tumor antigen ROR1 targeted drug delivery mediated selective leukemic but not normal B-cell cytotoxicity in chronic lymphocytic leukemia. Leukemia. 2015 Feb;29(2):346-55. doi: 10.1038/leu.2014.199. Epub 2014 Jun 20. PMID: 24947019; PMCID: PMC4272672.
  21. EP3405496B1 ROR1 ANTIBODY COMPOSITIONS AND RELATED METHODS, Espacenet, https://worldwide.espacenet.com/patent/search/family/057966152/publication/EP3405496B1?q=EP3405496B1
  22. Moazzam, M., Zhang, M., Hussain, A., Yu, X., Huang, J., & Huang, Y. (2024). The landscape of nanoparticle-based siRNA delivery and therapeutic development. Molecular therapy : the journal of the American Society of Gene Therapy, 32(2), 284–312. https://doi.org/10.1016/j.ymthe.2024.01.005
  23. Yang, C., Lin, Z. I., Zhang, X., Xu, Z., Xu, G., Wang, Y. M., Tsai, T. H., Cheng, P. W., Law, W. C., Yong, K. T., & Chen, C. K. (2024). Recent Advances in Engineering Carriers for siRNA Delivery. Macromolecular bioscience, 24(4), e2300362. https://doi.org/10.1002/mabi.202300362
  24. Lorenzer, C., Dirin, M., Winkler, A. M., Baumann, V., & Winkler, J. (2015). Going beyond the liver: progress and challenges of targeted delivery of siRNA therapeutics. Journal of controlled release : official journal of the Controlled Release Society, 203, 1–15. https://doi.org/10.1016/j.jconrel.2015.02.003
  25. Mizrahy, S., Hazan-Halevy, I., Dammes, N., Landesman-Milo, D., & Peer, D. (2017). Current Progress in Non-viral RNAi-Based Delivery Strategies to Lymphocytes. Molecular therapy : the journal of the American Society of Gene Therapy, 25(7), 1491–1500. https://doi.org/10.1016/j.ymthe.2017.03.001
  26. Kon, E., Hazan-Halevy, I., Rosenblum, D., Cohen, N., Chatterjee, S., Veiga, N., Raanani, P., Bairey, O., Benjamini, O., Nagler, A., & Peer, D. (2020). Resveratrol Enhances mRNA and siRNA Lipid Nanoparticles Primary CLL Cell Transfection. Pharmaceutics, 12(6), 520. https://doi.org/10.3390/pharmaceutics12060520
  27. Khodakarami, A., Kashani, M. A., Nazer, A., Kheshti, A. M., Rashidi, B., Karpisheh, V., Masjedi, A., Abolhasani, S., Izadi, S., Bagherifar, R., Hejazian, S. S., Mohammadi, H., Movassaghpour, A., Feizi, A. A. H., Hojjat-Farsangi, M., & Jadidi-Niaragh, F. (2023). Targeted Silencing of NRF2 by rituximab-conjugated nanoparticles increases the sensitivity of chronic lymphoblastic leukemia cells to Cyclophosphamide. Cell communication and signaling: CCS, 21(1), 188. https://doi.org/10.1186/s12964-023-01213-1
  28. Żak, M. M., & Zangi, L. (2021). Lipid Nanoparticles for Organ-Specific mRNA Therapeutic Delivery. Pharmaceutics, 13(10), 1675. https://doi.org/10.3390/pharmaceutics13101675
  29. Jyotsana, N., Sharma, A., Chaturvedi, A., Budida, R., Scherr, M., Kuchenbauer, F., Lindner, R., Noyan, F., Sühs, K. W., Stangel, M., Grote-Koska, D., Brand, K., Vornlocher, H. P., Eder, M., Thol, F., Ganser, A., Humphries, R. K., Ramsay, E., Cullis, P., & Heuser, M. (2019). Lipid nanoparticle-mediated siRNA delivery for safe targeting of human CML in vivo. Annals of hematology, 98(8), 1905–1918. https://doi.org/10.1007/s00277-019-03713-y
  30. Estapé Senti, M., de Jongh, C. A., Dijkxhoorn, K., Verhoef, J. J. F., Szebeni, J., Storm, G., Hack, C. E., Schiffelers, R. M., Fens, M. H., & Boross, P. (2022). Anti-PEG antibodies compromise the integrity of PEGylated lipid-based nanoparticles via complement. Journal of controlled release : official journal of the Controlled Release Society, 341, 475–486. https://doi.org/10.1016/j.jconrel.2021.11.042
  31. Weinstein, S., Toker, I. A., Emmanuel, R., Ramishetti, S., Hazan-Halevy, I., Rosenblum, D., Goldsmith, M., Abraham, A., Benjamini, O., Bairey, O., Raanani, P., Nagler, A., Lieberman, J., & Peer, D. (2016). Harnessing RNAi-based nanomedicines for therapeutic gene silencing in B-cell malignancies. Proceedings of the National Academy of Sciences of the United States of America, 113(1), E16–E22. https://doi.org/10.1073/pnas.1519273113
  32. Nakamura, T., & Harashima, H. (2020). Dawn of lipid nanoparticles in lymph node targeting: Potential in cancer immunotherapy. Advanced drug delivery reviews, 167, 78–88. https://doi.org/10.1016/j.addr.2020.06.003
  33. Nakamura, T., Kawai, M., Sato, Y., Maeki, M., Tokeshi, M., & Harashima, H. (2020). The Effect of Size and Charge of Lipid Nanoparticles Prepared by Microfluidic Mixing on Their Lymph Node Transitivity and Distribution. Molecular pharmaceutics, 17(3), 944–953. https://doi.org/10.1021/acs.molpharmaceut.9b01182
  34. Zukancic, D., Suys, E. J., Pilkington, E. H., Algarni, A., Al-Wassiti, H., & Truong, N. P. (2020). The importance of poly (ethylene glycol) and lipid structure in targeted gene delivery to lymph nodes by lipid nanoparticles. Pharmaceutics, 12(11), 1068.
  35. Takasaki, J., Raney, S. G., Chikh, G., Sekirov, L., Brodsky, I., Tam, Y., & Ansell, S. M. (2006). Methods for the preparation of protein-oligonucleotide-lipid constructs. Bioconjugate chemistry, 17(2), 451–458. https://doi.org/10.1021/bc050052j
  36. Migliorati, J. M., Jin, J., & Zhong, X. B. (2022). siRNA drug Leqvio (inclisiran) to lower cholesterol. Trends in pharmacological sciences, 43(5), 455–456. https://doi.org/10.1016/j.tips.2022.02.003
  37. Hassan, I. (2019). Therapeutic targeting of the leukaemic fusion genes. Newcastle University. http://theses.ncl.ac.uk/jspui/handle/10443/4806

Accesibility Bar