Immunotherapy: Novel Immunomodulatory Targets
-
Upload
paul-d-rennert -
Category
Health & Medicine
-
view
1.418 -
download
4
description
Transcript of Immunotherapy: Novel Immunomodulatory Targets
Paul D Rennert, Founder & Principal SugarCone Biotech LLC
Exploring evolution of diverse immune checkpoint pathways
ImVacs, August 11-12, 2014
www.sugarconebiotech.com
How can we discover novel immune checkpoints?
• Focus on specific gene familes that we know contribute to immune checkpoint modulation
• Query these target collections using an evolutionary
perspective, to derive specific new therapeutic hypotheses
• Why? Ad hoc sampling of such families, especially the large ones, is time-consuming and expensive
2
www.sugarconebiotech.com
This is what we usually see ... and it’s useful
Drew M Pardoll, 2012. Nat Rev Cancer
• e.g. we know that CTLA4, an Ig-superfamily member, is a critical immune checkpoint target
• We’ve found out through the very hard
work of many labs that PD-1 is another critical immune checkpoint target
• But, can we use this level of information to guide our investigation of other Ig-superfamily members?
Using clustering to inform hypotheses • The prior figure is an example of functional clustering and can only inform us
about members whose function has been elucidated
• Highly refined distance methods are of interest, and by design use distance cutoff algorithms to unlink related families
Yap et al. 2013 JMB 4
Butyrophilins Siglecs
B7s PD-Ls
TIMs
www.sugarconebiotech.com
What is this showing us? • Relatedness is a term of convenience/relativity
• We “look” for relatedness to CD28 and B7s when we identify novel proteins ! PD-1 is cited as CD28/CTLA4 relative: PD-1 binds PD-L1 and 2 and
B7-1; the receptor has inhibitory activity ! In strict distance analyses, PD-1 is a singleton, with distant homology
to CD8 and the antigen receptor Ig domains ! ICOS can be aligned with CD28 and CTLA4 using pairwise
comparison methods
• ICOS-L, B7-H3 and B7-H4 align more closely with butyrophilins than B7s
• VISTA is a weak PD-L1 homologue, structurally unique
5
www.sugarconebiotech.com
Where do B7-1 & 2 and CD28 lead us?
ICOS! CTLA4! CD28!
PD-1!
VISTA!
????!
?"
?"
B7-H3 !
B7-H4 ! ICOS-L !
?"
????!
?"
6
www.sugarconebiotech.com
We have a big butyrophilin/B7 family B7-H3 !
B7-H4 ! ICOS-L !
We have a small CD28 receptor family
ICOS! CTLA4! CD28!
7
www.sugarconebiotech.com
Interesting New Directions
• Focus on new targets ! VISTA, B7-H3, B7-H4
• Novel families ! Butyrophilins are likely underexploited targets
• Different cell types ! NK cells, NKT cells, γ/δ T cells
• An insight: ! There is no a priori reason to anticipate high homology
when looking for the receptors/ligands of orphan proteins (e.g. IgSF BTLA binds TNFRSF HVEM)
8
www.sugarconebiotech.com
B7-H3 and B7-H4 B7-H3 • Biology poorly understood; low expression in normal tissues • Macrogenics (MGNX) and Servier are co-developing MGA271, a
humanized, Fc-optimized mAb that is in a Phase 1 trial, recruiting refractory prostate cancer and melanoma patients based on abundant expression of the target on these tumors
B7-H4 • Biology poorly understood, marks diverse normal and cancer cell types • Well-established biomarker of tumor progression • B7-H4 is expressed on infiltrating myeloid cells, vasculature, and tumor
cells in melanoma, NSCLC, prostate, ovarian, pancreatic, breast, renal and other cancers
• Amplimmune (now owned by the Medimmune division of Astra Zeneca; AZN) had listed B7-H4 in its’ portfolio
• Five Prime Therapeutics (FPRX) also has promoted this target
9
www.sugarconebiotech.com
VISTA • Normally expressed on hematopoietic cells including myeloid
cells and T cells • In multiple cancer models VISTA is found at particularly high
levels on tumor-infiltrating myeloid cells
• Appears to be a negative regulator of T cell responses, may suppress myeloid cell responses in the tumor microenvironment
• Anti-VISTA antibody treatment blunts tumor development even in the presence of high PD-1 expression
• Antibodies to VISTA are being developed by a private company, ImmuNext, in collaboration with Johnson&Johnson (JNJ)
10
www.sugarconebiotech.com
New “B7s”
• you can always force a phylogeny by limiting the input sequences, as shown here
• utility: can highlight relationships of suspected novel family members
• B7H6, tumor-specific in human?, binds NKp30, function poorly understood, an outlier
• B7H7/HHLA2 – likely related to butyrophilins
PD-L2
PD-L1 B7-H7 B7-H4
B7-H3
ICOS-L
B7-1
B7-2
VISTA
B7-H6
Flajnick et al. 2012. immunogenetics
The “true” Butyrophilins
• The failure to identify clear ligand/receptor interactions has limited drug development
• A reasonable prediction is that further study of the butyrophilin family will yield novel targets
Arnett & Viney. 2014. Nat Rev Immunol.
Underexploited Ig-families • Precedence here may come from the distinct but instructive LIR/KIR and NKG2
families of NK cell surface proteins
• Despite a huge complexity of family members and confusing biology, useful targets are emerging, notably from Innate Pharma (IPH)
• Bristol-Myers Squibb (BMY) is sponsoring trials of lirilumab in combination with ipilimumab (anti-CTLA4) and nivolumab (anti-PD-1) in patients with solid tumors. These trials will start to read out over the next 2 years.
13
www.sugarconebiotech.com
Lets try this again: TIM family proteins • TIM-1 identified as a critical T cell control protein (2001)
• Member of a small subfamily of receptors featuring IgV domains atop mucin domains and short stalks ! Human proteins are TIM-1, TIM-3, TIM-4
• We knew as early as 2002 that these were Siglec homologues ! TIM-1 cell binding was cation-sensitive suggesting a carbohydrate
component but a screen by he Consortium for Functional Glycomics failed
! Based on publications we chased semaphorins, CD300b, others
! Galectin-1,8 identified as binding proteins, unclear specificity?
14
www.sugarconebiotech.com
TIM family proteins: nearest neighbors • Extensive homology modeling places TIM-1 between
Siglecs and Butyrophilins
15
Rubenstein et al. 2013 Structure.
TIM-1 relationship to Siglecs
16
• TIM-1 IgV domain resembles a Siglec with a modified sialic acid binding motif
• Minimal energy modeling suggest carbohydrate docks as in a typical Siglec
16
17 *: Miyanishi et al. Nature 450: 435; Kobayashi et al. Immunity 27: 927; Santiago et al. Immunity 27: 941; Ichimura et al. 2008. J. Clin. Immunol. 118: 1657.
• TIM-1, TIM-4 and TIM-3 bind PS
• PS binding site: a cation buried in a charged cavity
• The sialic-acid binding residues lie atop the PS-binding site.
The real key to unlocking TIM-family function was phosphotydylserine (PS)
www.sugarconebiotech.com
• A reasonable hypothesis is that TIM-3 antagonism is efficacious in preclinical tumor models because it prevents T-cells from receiving inhibitory signals delivered by PS expressed by tumor cells and tumor associated macrophages (MDSC) ! Whether the putative TIM-3 ligand Galectin-9 also plays a role is
unclear
• Novartis (NVS)/CoStim, Bristol-Myers Squibb (BMS), Tesaro (TSRO)/Anaptsys all have active programs targeting TIM-3, with many others coming along ... why?
PS, tumors and TIM-3
18
www.sugarconebiotech.com
TIM-3 preclinical data are compelling
19
www.sugarconebiotech.com
Not so TIM-1: our evolving view of TIM-1 expression
• Focal, bright expression on mucosal endothelium
• Expressed on kidney epithelium (lumenal), and highly expressed in the injured kidney
• Highly (drastically) upregulated not only on kidney and ovarian carcinoma but also non-cancerous kidney and liver cells after chemo/radiation
• It’s reasonable to predict that the anti-TIM-1-ADC being developed by Celldex’ (CLDX) will have a difficult time finding a safe therapeutic window in the context of standard of care
20
www.sugarconebiotech.com
Other programs targeting cell death pathways
• Cell surface expression of PS is triggered by apoptosis; the appearance of sufficient PS on the cell surface triggers phagocytosis
• However, PS expression is quite widespread even on
normal cells
• SIRPα/CD47 is an explicitly anti-phagocytic signaling mechanism
• However, CD47 is widely expressed, and is abundantly
expressed on platelets
• Both targets seem problematic, however... 21
www.sugarconebiotech.com
Other programs targeting apoptotic signaling
These independent lines of evidence support the hypothesis that TIM-3 is acting in the context of PS-binding
• Peregrine Pharmaceutical’s (PPHM) anti-PS antibody bavituximab has advanced to Phase 3 with fast-track desination in r/r solid tumors including NSCLC (+ platinum-based chemo)
• In Phase 2, the ORR was only 17%, those responding had a median increase in OS of just over 4 months. It is predictable that this program will seek rational combination therapy trials.
• Celgene (CELG) handed privately held Inhibrx LLC a 50MM USD upfront with potential royalties running to 500MM for an anti-CD47 antibody specifically designed not to activate platelets
• Additional CD47 and SIRPα programs are under development
22
www.sugarconebiotech.com
• uh-oh... One more example...
• shown are two proteins that been investigated in the context of immune oncology:
• TIGIT
• CD200
• The fact that these cluster in the complex Nectin family is a little worrisome: these family members homo and hetero-dimerize fairly promiscuously, and this may complicate analysis and use of biologics targeting these proteins
23
www.sugarconebiotech.com
• CD200 was being targeted by an antibody ALXN6000/samalizumab (Alexion, ALXN) although there are no recent clinical updates - the company abandoned oncology for rare diseases
• TIGIT is a relatively new IgSF protein, with clear cell inhibitory function. There is defined binding to PVR and Nectin-2, potentially leading the way to influencing both T cells and NK cells through inhibition of this pathway
• However, proteins in the nectin family may have diverse binding partners, complicating development, e.g. the TIGIT ligands may also bind DNAM
Targeting “Nectins”?
24
Finally ... such analyses go only so far
LAG3 BTLA
FLT3
25
www.sugarconebiotech.com
other applications ...
• this type of analysis can be productively applied to the TNF and TNFR superfamilies ... which are much smaller
• allowing us to look for interesting and distinct pathways ...
• recent work has highlighted TNFRSF25 as a critical node in the immune response to tumors as you will hear more about from Taylor Schreiber
26
www.sugarconebiotech.com
Acknowledgements
Slides will be available on SLIDESHARE
Link through the sugarconebiotech.com blog
27
• TIM-1 program: BiogenIdec • TIM-1 oncology: CoStim • TIM-1 homology and binding specificity: Alexey Lugovskoy (Merrimack), Yen-
Ming Hsu (AbBio), Veronique Bailly (BIIB), Patricia McCoon (AZN), Gabriela Constantin (U Verona, IT)
• TIM-1 virology: ! Ebola, Marburg: Wendy Maury (U Iowa), Rob Davies (Tx Biomed) ! HIV: Shan-Lu Liu (U Missouri)