Accelerate Your Antibody Discovery with High throughput Array SPR

Transcription

1 ccelerate Your tibod Discover with High throughput rra SPR December 10, 2018 Daiel ediger, PhD Seior Field pplicatios Scietist, Carterra

2 Carterra Fouded i 2005, Salt Lake Cit, UT, as Wasatch Microfluidics - Licesed Cotiuous Flow Microspotter (CFM) techolog from the Uiversit of Utah Research Foudatio - Combied the CFM with the IIS MX96 SPR detector Received sigificat fudig i 2016 rebraded as Carterra Expaded maagemet, R&D, ad commercial teams Commercial lauch i Ja 2018; - Now with sales i North merica, Europe, ad sia-pacific 2

3 Hubble extreme deep field ifrared 3

4 disruptio i rra SPR LS Lodestar referece poit used for avigatio rra 4

5 Therapeutic mbs Comprise bout Half Pharma Sales 5 Therapeutic mbs are a lucrative ad growig segmet of the pharma idustr s portfolio iophsics is agostic of therapeutic area because it studies molecular-level bidig iteractios to better uderstad a drug s mechaism of actio to guide the desig of extgeeratio medicies

6 Which Oe? Large ivestmets made i mb geeratio altical tools lag orders of magitude behid Do t throw awa a blockbuster cadidate! 6

7 High Throughput SPR a Disruptive Techolog rra SPR shifts the paradigm i atibod screeig eablig detailed kietic, affiit, ad epitope characterizatio of our librar at the earliest stages of research Icreases the chaces ou ll fid leads withi our librar that require miimal egieerig ccelerates librar-to-lead selectios 7

8 LS Core pplicatios Kietics/ffiit Epitope iig Mappig 8

9 Carterra s LS TM Disruptive Techolog Sigle-chael mode (large flow cell) Multi-chael mode (96-chael prithead) 9 Itegrates rra SPR with Flow Pritig utomated switchig betwee sigle- ad multi-chael Flow cells 384 reactio spots + iterspots per arra I-lie reloadig of arra, up to 1152 ligads per ru (3x 384) Supports stadard amie couplig ad capture formats

10 10 LS Prithead-96 spots

11 Flow Prit a 384-Ligad rra o the LS Seriall prit 4 ested 96-spot arras to geerate a 384-spot arra 11

12 Miimal alte Cosumptio Oe-o-Ma ssa Format alte flows over etire arra (1 o 96) alte flows over etire arra (1 o 384) 12

13 What Does High Throughput Kietics Look Like? Ligad Kietics, 1 Parallel Ru, 7 µg tige, 8 Cocetratios Software automaticall flags the Good, ad, ad Ugl

14 Scree Ma or Icrease Your N Reproducibilit across the arra allows ou to scree with cofidece If <384 uique mbs, wh ot icrease our llows statistical aalsis of the reported kietic parameters mb Mea ± StDev of 12 spots k a = (2.4 ± 0.2) x 10 5 (1/Ms) k d = (2.1 ± 0.1) x 10-3 (1/s) K D = 8.8 ± 0.9 M 14

15 oth Slow ad Rapid Kietics are Well Described mb Mea ± StDev of 12 spots k a = (1.48 ± 0.06) x 10 5 (1/Ms) k d < 1 x 10-5 (1/s) K D < 67 (± 3) pm mb Mea ± StDev of 8 spots k a = (3.6 ± 0.2) x 10 5 (1/Ms) k d = (5.2 ± 0.3) x 10-2 (1/s) K D = 145 ± 13 M 15

16 Iso-ffiit Plot Iso-affiit plots facilitate assessmet of kietic diversit withi a mb pael ppl 5% rule to the k d value The blue cloes idicate three cloes with differet affiities Weak, medium, high ppl the 1 x 10-5 k d limit automaticall i the fittig routie to prevet iappropriate affiit estimates whe miimal dissociatio is observed Kietics are well described coverig a 10,000 fold K D rage

17 Get Stats o Small Mb Paels rraig Cloes o Replicate Spots k a (1/Ms) uique mbs, each o 8 or 12 spots

18 Get Stats o Small Mb Paels rraig Cloes o Replicate Spots k d (1/s) uique mbs, each o 8 or 12 spots

19 Get Stats o Small Mb Paels rraig Cloes o Replicate Spots K D (M) uique mbs, each o 8 or 12 spots

20 LS-HC30M K D (M) echmark LS vs iacore 8K K D (M) (36 cloes) = x R² = iacore8k-c1 K D (M) Excellet agreemet i kietic rate costats (withi 2fold, whe match chip tpes) Excellet agreemet across wide affiit rage <100pM to >100M LS is 50x faster tha iacore (384 vs 8 iteractios at oce) LS cosumes 1% sample of iacore LS aalzes 384 bidig iteractios i a sigle work da s ru vs multiple rus/das to achieve similar throughput o iacore LS has powerful batch-mode fittig software to facilitate aalsis 20

21 8_D12, K D = 6M 7_F10, K D = 92M 8_F02, K D = 106M 7_F04, K D = 248M 7_G01, K D = 2.1uM 7_D11, K D = 2.7uM 21

22 Wh It s ll bout The Epitope Uderstadig a mb s mechaism of actio (MO) is fudametal to the discover of efficacious ad safe therapeutics mb s epitope largel dictates its biological fuctio, so sortig mbs b their epitope is more relevat tha affiit rakig them mb s epitope is a iate propert that ca be either shifted ratioall b egieerig or predicted/desiged i silico, so relies o empiric selectio Idetifig mbs targetig uique epitopes is highl desirable because the ma offer differetiated MO s ad associated IP opportuities Surveig the epitope ladscape of our mb campaig at earl-stage research where large umbers of cloes are available i low quatities provides fuctioall-relevat iformatio for triagig hits ad prioritizig resources rra SPR ca rapidl sort mbs ito epitope families usig mappig ad biig methods 22

23 High Throughput Epitope iig 384x384 mb competitio matrix reveals exquisite epitope differetiatio Triage large mb paels, idetif uique/uaced biders Low sample cosumptio - facile setup - streamlied aalsis Ol 5 µg/mb required regardless of arra size Oe sample for arra pritig, oe sample as aalte µg atige for 384 x 96 arra, µg for 384 x 384 biig (147,000 Iteractios) 13 x 13 (LI) 384 x 384 (rra SPR) Commuit plot 23 Mousa et al, PNS 2016 Sivasubramaia et al, MS 2016

24 Trasformig the iig Paradigm LS 8K HTX tibod cosumptio per mb 5 µg >1000 µg >1000 µg well plates required 2 plates / 4 das 10 to 100+ / 1 moth 10 to 100+ / 1 moth

25 mb iig: ti-pgrn librar Mappig P P P P P P F F F F G G G G G E E E Mouse Cross is librar MappigMouse Cross is ID P P P F F F F F G G G G E E E Chig et al, MS 2017

26 Epitope iig Software User Iterface Data Liked cross Three Visualizatio Paels sadwich smmetric block (dotted lie) block 26 Sesorgrams Heat Map Networks

27 Dedrograms ad Commuit Plots Networks (most graular bis) Dedrogram Commuities (user-defied cutheight) 27

28 Color Orthogoal Data Librar Subdomai Chig et al, MS Mouse Cross-Reactivit