Exploiting a natural conformational switch to engineer an interleukin-2 ‘superkine’ May 22, 2012 Joseph Argus, Pardeep Singh, Uland Lau.

Exploiting a natural conformational switch to engineer

an interleukin-2 ‘superkine’

May 22, 2012

Joseph Argus, Pardeep Singh, Uland Lau

IL-2• IL = interleukin = cytokine of immune system• 15.5 kD, variably glycosylated• Necessary for growth and function of T cells• Promotes differentiation and proliferation of

natural killer cells• Used in clinic to upregulate immune system

(chronic viral infection, adjuvant for vaccines, cancer therapy)

• Also adverse effects, at least partially due to upregulation of Treg cells

Goal: Create modified IL-2 that stimulates cytotoxic T cells and natural killer cells with

less Treg activation (fewer side effects)

IL-2 Receptor

• Treg and cytotoxic T both contain low levels of beta and gamma

• Only Treg contain high levels of alpha (in resting state)• Locking IL-2 in the active (purple) conformation will bypass

the need for alpha and increase the relative proportion of cytotoxic:regulatory T cell activation

Summary:

• Developed versions of IL-2 (“superkines”) that bypass the need for the alpha subunit of receptor using directed evolution

• Verified nature of mutations using physical biochemistry, crystallography

• Verified biological significance using:– in vitro assays (pSTAT5)– in vivo assays (splenic lymphocyte number, tumor

volume, and lung metastases)

Directed Evolution

-Five of the six mutations clustered on the B-C loop and within the C helix core.-V85, F80, andV86 substitutions appeared to collapse into a hydrophobic cluster to stabilize the loop by fixing helix C into the core of the molecule.

Crystallization of D10 IL-2 superkine

Low-resolution structure of D10 ternary complex

-Is this heterodimeric architecture the same when D10 binds as compared with wild type IL-2?Answer-Found to be essentially identical r.m.s.d.=0.43 angstoms

Conformation of unliganded IL-2/D10 and ligand bound CD25

-Unliganded D10 is conformationally similar to the IL-2Ralpha[CD25] as compared to the unliganded IL-2

Molecular Dynamics simulations of IL-2 and D10

-Analysis of anatomically detailed Markov state models showed that D10 was more stable than IL-2-B/B-C/and C all had lower visible deviations compared to wild type IL-2

Comparison of average IL-2 wt vs.D10

Conclusion from set of experiments

• The reduced flexibility of helix C in the IL-2 superkine is due to improved core packing with helix B.

• Structural and molecular dynamics results show that evolved mutations cause a conformational stabilization of the cytokine, reducing the energetic penalties for binding to IL-2Rβ.

Dose response curves using flow cytometry to assay STAT5

phosphorylation

Absence of CD25 Presence of CD25

-Do IL-2 superkines demonstrate signal potencies?-Do they depend on cell surface expression of CD25?

Probing CD25-independence with a mutation of IL-2

• F42A= Phe 42 replaced with Ala. Reduces binding to CD25 by 220-fold for H9 and 120-fold for IL-2.

Dose response curves on T cells from mice with absent CD25.

Flow cytometry fluorescence assaySuperkines=spread throughout/low density.

IL-2=concentrated/ lacks replication/ high density.

Antitumor activities of IL-2 superkine

• IL-2 superkine H9, wildtype IL-2, and IL-2-anti-IL-2 mAb effects on CD25low vs CD25high T cells

• IL-2-anti-IL-2 mAb– Shown to reduce pulmonary

edema and have potent antitumor responses in vivo

• Memory-phenotype (MP) CD8+ T cells– Low levels of CD25– High levels of IL-2Rβγ

• Regulatory T (Treg) CD4+ cells– High levels of CD25

Different tumor models

• Mice injected subcutaneously with B16F10 melanoma cells, murine colon carcinoma, and Lewis lung carcinoma

• Treatments:– PBS-control– High-dose IL-2– IL-2-anti-IL-2 mAb complexes– H9 IL-2 superkine

• PBS-control : tumor reached 1500 mm3 at day 18

• IL-2 treatment: delayed as much as 39% at day 18

• Similar effects between IL-2-anti-IL-2 mAb and H9 IL-2 superkine– Reduced tumor growth by more than 80%– Compared to IL-2, >70% reduction

Conclusions

• Engineered IL-2 superkine via in vitro directed evolution

• Eliminated CD25 dependency of IL-2• Increased binding infinity towards IL-2Rβ• IL-2 superkine elicited proliferation of T cells

irrespective of CD25 expression• Improved antitumor responses in vivo (reduced

pulmonary edema)• Showed activation of cytotoxic CD8+ T cells

and NK cells – antitumor immune response• Minimal toxicity