Identified? Yes
-
Feasibility Start 2 year before clinical validation (Takano san, 20200415 call)
-
Feasibility Start with narrowed down to 2-3 compounds
-
아래 그림으로는 PS 인데 이건 너무한듯
-
아래 로는 PE 때 시작
Biomarker Types and Stage Gates — Alignment with Asset Milestones
(Takeda slide — transcribed labels, slide image not embedded)
Two parallel timelines, each annotated with stage-gate dots:
- Asset: TE — PS — LGE — PE — CN — CS — EDE — IND — EPOC
- Biomarker: BPS (at PS) — BFA (at PE) — BTV (at CN) — BCS TE, PD, DR (at CS) — BCS PS (at EDE/IND)
Four icon callouts (Target Engagement TE / Target itself, Pharmacodynamic PD / Target Pathway, Disease Related DR / Disease Pathway, Patient Segmentation PS / Diagnostic) feed into four gate boxes:
- BPS – at Asset PS — Biomarker Platform Start: White paper rationale, Develop BM “one-pager”, BMax for novel PET tracer
- BFA – at Asset PE — Biomarker Feasibility Assessment: Compatibility of reagents, assay format; Determine assay sensitivity in samples; internal vs external execution
- BTV – at Asset CN — Biomarker Technical Validation: In vitro / in vivo evaluation of BM; Sensitivity, specificity; Reproducibility, linearity, etc.
- BCS – at Asset CS — Biomarker Candidate Selection: In vivo Calibration (IVC); Preclinical PK-PD; Longitudinal response of BM in preclinical or human EM study
(Takeda Pharmaceutical Company Limited)
Biomarker type-specific definition of milestones
(Takeda slide table — transcribed text, slide image not embedded)
| BPS – at Asset PS Biomarker Platform Start | BFA – at Asset PE Biomarker Feasibility Assessment | BTV – at Asset CN Biomarker Technical Validation | BCS – at Asset CS Biomarker Candidate Selection | |
|---|---|---|---|---|
| Common | • White paper rationale • Develop BM "one-pager" | • Compatibility of reagents, assay format • Determine assay sensitivity in samples • internal vs external execution | • In vitro / in vivo evaluation of BM • Sensitivity, specificity, • Reproducibility, linearity, etc. | • In vivo Calibration (IVC) • Longitudinal response of BM in preclinical or human EM study |
| Molecular Biomarker | Preclinical PK-PD | |||
| PET ligand | BMax for novel PET tracer | Specific binding in rodent | All the work finished for NHP study | |
| Neurocircuitry biomarker | Ability collect in rodent/NHP/human | Define sensitivity and selectivity in rodent/HP/human | Quantitative validation in rodent/HP/human Change is observed I patients/disease model |
(Takeda Pharmaceutical Company Limited)
Comparison: in vitro vs in vivo
| Possibilities | Consequence | ||
|---|---|---|---|
| In vitro | homogenates | the brain membranes are trapped on glass fiberfilters and subsequently washed several times | ↓ free radiotracer, ↓ NSB (보통 ↑ SB) |
| radiotracer may bind to the glass-fiber filters When homogenized, the cellular components wash away, | ↑ SB (이겠네?) ↓ SB (이겠네) | ||
| (no-wash) autoradiography | |||
| (living) cell binding assays | 장점: i) no Homogenation-induced wash away of intracellular component (target) ii) maintained cell membrane, so can demonstrate the ability of tracer to pass the memb and accumulate in the cell (→ ↑ in vivo translatability) | ||
| In vivo | Whatever signal one may obtain in vivo is solely determined by the affinity (In vivo imaging does not have a wash step!) | ||
Preclinical study
| Studies to establish | Method | Examples | |||
|---|---|---|---|---|---|
| In vitro | -In vitro selectivity profiles, while useful for screening, are insufficient to prove in vivo selectivity | ||||
| Bmax |
-SB window (percentage specific binding) -Bmax (, fmol/mg tissue) -Brain bio-distribution: , (while not impacting PET doability), is required to inform subsequent PET imaging studies such as specific binding assessment. For example, if a target is only expressed or enriched in certain brain regions then a target-free brain region could be used as a "reference region" (e.g. cbll) to determine SB. On the other hand, if a target is expressed throughout the brain it would be necessary to carry out both baseline and blocking or equivalent (e.g. knock-out animal) studies to determine SB | In vitro autoradiography, [3H]-compound |
KO 사용해 (competition 해도 되지 않나?) SB window 확립 (1:1?), 이러고 나서야 Bmax 잴 수 있나? (nsb 것들은 target과 무관한 것들이니) Bmax assessment in rats (figure, evidence-only — same WT vs HDAC6 KO Cbound/Cunbound saturation curve as `20240722_184809`, with Specific binding / off-target binding / Non-specific binding legend; equations Cbound/Cunbound = Bmax_KO / (Kd_KO + Cunbound) + k_KO (HDAC6 KO)Cbound/Cunbound = Bmax_SB / (Kd_SB + Cunbound) + Bmax_KO / (Kd_KO + Cunbound) + k_KO (WT)Parameter table: Bmax_SB 3.1 nM (0.0–10.6); Kd_SB 1.2 nM (0.0–4.5); Bmax_KO 33.5 nM (0–133.8); Kd_KO 22.4 nM (0–84.5); k_KO — 2.7 (1.9–3.5)) | ||
|
Moreover, quantification allows for ① the pharmacological characterization of ligand affinity by means of dissociation constants (Kd), inhibition constants (Ki) ② the density of binding sites (Bmax) in selected tissues. Thus, the method provides information about both target localization and ligand selectivity (Au – Griem-Krey et al. 2019, PMID) - tissue sample: (Ashley Knight) I usually request a larger block (1g when possible). Something with one dimension the size of my thumbnail at least (very scientific). When tissues are rare I try to get 1 cm cubed in size. It provides enough tissue for sectioning and several studies. - Radioligand concentration (다양한 approach 존재): 1. a concentration of radioligand approximating the KD is used which would occupy 50% of the receptors present. 2. when the binding site is highly abundant in the examined tissue (& because radioligands can be costly) lower concentrations e.g. 10-20% of the KD concentration can be used. This has a potential advantage of increasing the signal-to-noise ratio by reducing NSB more than SB. 3. a concentration reflecting 5-6 times Kd is used to make sure that all binding sites are saturated. | |||||
| Mice | ARG 20210120 Paul: $150K for any rodent work | • A suitable radioligand | |||
| Generating 3H-labelled ligand | (low non-selective binding, sufficient selective binding commonly demonstrated using KO mice) | Example: Usually 3H or 11C (대개 18F안 쓴다) | cost for tritiation with CRO is Tomimatsu 200 mil yen (20억원, 이렇게 비쌀리 없음?) | ||
| Sample preparation | microtome cryostat for tissue sectioning | (12 µm recommended for [3H] labelled ligands for mice brains) | 1. Kunugi-san's team 2. 미야 라는 사람 Price: $150/slide (TNTecnos, Tomimatsu) | ||
| ARG experiment (per se) | Incubation with radioligand Exposure of the film to the matrix (containing the radiolabeled sample) | Tomimatsu: just using ImageJ | |||
| Image development | |||||
|
analysis - Cf) microARG: radiolabeled substrate taken up by individual prokaryotic cells → visualized by means of a radiation-sensitive silver halide emulsion covering the radiolabeled organisms → subsequently processed by standard photographic procedures. | |||||
Examples
| Example | Method | Readout | |
|---|---|---|---|
| Ashley's AD postmortem: [11C]NCGG401 {Ogata, 2022 #2023} |
-tracer 농도는 변화 [3H]MCC950 10-100 nM, -blocker 농도는 고정 (MCC950 10 µM) Tracer 와 blocker (10 µM NCGG401) 모두 농도 고정 | ||
| {Au – Griem-Krey, 2019 #1506} | Jove protocol. mouse 포함. | Fig4) total binding: as fmol/mg tissue equivalents (TE) for the radioligands [3H]GHB (30 nM) Non-specific binding was not detected in the presence of 1 mM GHB Data is presented as mean ± SD of four biological replicates each performed in three technical replicates. | |
| AC Immune's aSyn PET | Human brain section (p14, 조각인듯), showing n of 2 | 한번은 radio-labelled 처리 & 다른 한번은 self-blocking 해놓고 (with ACI-3710 5 µM), Self-blocking | Ki |
| SPAL | Mouse (good의 p9 fig3-2중 맨 아래) & human Patient (good의 p9 fig3-2중 위로부터 세개, 조각인듯) (n=1?) | non-radioactive test compounds were used in excess (10 µM, 이렇게 완전히 막아놓고 나오는 signal 보는 것이니 'non-specific binding') and mixed with the 18F-labeled tracer (0.8–2.4 kBq/Ml, | Tracer 의 affinity (specific binding)를 보려는 목적이니 labeled tracer 를 여러 dose 씀. quantification 없는 듯 |
| (Kroth et al. 2019, PMID 31264169) | Human brain section (frozen 18 µm): fig4 (Brakk I, III, V 당 n=1 인듯). Entorhinal cortex hippocampus, frontal cortex 의 각 조각에서 본 듯, fig5: only amygdala (n=1 인듯) | ||
| Micro-ARG (Frozen tissue from entorhinal cortex brain region of an AD donor) | - | ||
| 위 Bmax 에 (Hostetler, 2016 #889) | NHP Brain (n=4), human (n=3) 은 region별 (cortex, thal, Hippo etc) | - | |
| (Hall, 1998 #1259) | Human brain section (whole hemisphere) | - | DA & HT receptor subtypes as well as of their transporters |
| (Pan, 2016 #1260) | Human postmortem AD brain hippocampus sections (7 µm), 확실히 AD 2명, Control 2명 | - Chunks of frozen tissue were dissected | |
| (Martín-Cora, 2004 #1488) | Human postmortem frontal cortex (a, b), hippocampus (c, d), mesencephalum (e, f), and thalamus: fig 4 (coronal sections) | 상세한 정량화 table 3!, By caudate nucleu, putamen, GP, SN 등 Region별로 다 했음 human) | (small saturation-binding figure shown — `Bmax (fM/M) – 0XX dpm` annotations on a single tracer concentration curve, evidence-only) |
Uncertain Spans
| location | transcription | uncertainty |
|---|---|---|
| Asset–Biomarker timeline, Asset stage labels | TE — PS — LGE — PE — CN — CS — EDE — IND — EPOC | The stage abbreviations on the Takeda slide are small; the label LGE between PS and PE in particular is tiny and could be IGE or similar. |
| Biomarker timeline, BCS gates | BCS TE, PD, DR (at CS) and BCS PS (at EDE/IND) | The two BCS gate labels share the same letters but are color-coded with different annotations; the exact category mapping is small print. |
| Preclinical study, Mice ARG row, Generating 3H-labelled ligand cost cell | Tomimatsu 200 mil yen (20억원, 이렇게 비쌀리 없음?) | The numeric 200 mil yen token and the parenthetical 이렇게 비쌀리 없음? Korean comment are at the small-print right edge of the slide-style table — preserved but the exact mil/billion conversion remark has minor character ambiguity. |
| Examples row for SPAL | (good의 p9 fig3-2중 맨 아래) and (n=1?) | The figure references are written in shorthand Korean+number tokens; transcribed verbatim but small print obscures the trailing punctuation. |
| Examples row for Martín-Cora 2004 | small inset binding curve labels | The inset y-axis label Bmax (fM/M) – 0XX dpm is partly obscured; transcribed as visible. |