Amyloid beta protein may initiate a cascade leading to AD pathology.

1
Amyloid beta protein may initiate a cascade
leading to AD pathology.
2
secretases
APP
A?
nucleus
aggregation, fibrils and amyloid plaques
Clearance Microglia and blood vessels
Enzymes break down A?
neurotoxicity
3
CSF A? equilibrium depends on
4

• Ab42 is the initiator and main culprit in
  amyloid deposition

• A?42 is the initial amyloid species deposited
  in brain
• A?42 exceeds A?40 in amyloid deposits
• Toxicity and amyloid fibril formation A?42 gt
  40
• Selectively ? in presenilin mutations
• ? in most APP mutations
• High plasma A?42 is linked to a LOAD locus on
  chr 10

Downs syndrome study by C. Lemere
5
Possible changes in A? in sporadic AD
Production overall A? - BACE relative A?42 - ?-secretase ? BACE in brain in AD ? sAPP? in CSF ? plasma A?42 ? CSF A?42 ? cholesterol may affect lipid rafts Yes ? ? Yes No ?
Breakdown Neprilysin Insulin degrading enzyme ? Neprilysin in brain ? IDE or ? insulin associated with AD ? ? Chr 10 locus
Clearance vs aggregation APO-E e4, clusterin, microglia, RAGE ? CSF A?42 Yes
6
CSF A? in AD

• Total A? or A?1-40 do not differ in AD and
  controls
• A?42 levels are decreased in CSF in AD vs
  controls, by about 50.
• A?42 levels increase in the brain.
• ? deposits act as a sink, which binds more
  A?42
• Meta-analysis of CSF A?42, AD vs controls
• 18 studies, 980 AD, 499 controls
• Effect size 1.56 (Sunderland 2003)
• A?42 levels decrease in CJD, and in about 15-25
  of non-AD dementias
• ? due to ? production, or concomitant AD
  pathology

7
CSF A? and brain A? deposition
8
CSF A?42 meta-analysis (Sunderland, JAMA 2003)
9
CSF A-beta42 and APO-E
10
CSF A?42 in very mild AD/MCI
11
CSF biomarkers in MCI and early
ADRiemenschneider et al, 2002
12
Measuring A? subtypes
Peptide CSF Stds.
A? was immunoprecipitated from 2 ml of CSF from
an AD patient, and visualized on a bicine gel
that resolves A?38, 40 and 42
13
A? species in CSF
Wiltfang et al, J Neurochem 2002
14
Relative decrease in A?42 in CSF in AD
15
CSF A? as an index of drug treatment?

• Half-life of A? in CSF is about 30 minutes
• CSF and plasma A? are not correlated in humans
• May be easier to show effects in controls than in
  AD, because levels are not already decreased.
• Limited published data .
• - ?-secretase inhibitors CSF and plasma A?40
  and 42 ? in APP tg mice
• - Some NSAIDs may selectively decrease A?42 in
  tg mice and increase A?38
• - Rivastigmine x 1 year had no effect on CSF A?42

16
CSF A?42 remains stable in AD over 12 months
17
Summary

• CSF A?42 is decreased in AD, in 70-85 of
  patients, but less consistently so in MCI.
• A?40 levels are not altered.
• Diagnostic potential of CSF A?42 is limited, but
  may improve if it is part of a panel of
  biomarkers.
• CSF and possibly plasma A? may be used to monitor
  certain types of anti-amyloid therapy, e.g. for
  proof of principle, or dose finding
• Several forms of A? can be measured in CSF data
  on A? subtypes and on oligomers will be of
  interest.

18
Plasma Ab in inherited and sporadic AD