Using Affymetrix gene chips, Yankner and colleagues compared the transcriptional profile of wild type and PS1/PS2 double knockout embryonic stem (ES) cells. Given the dramatic effect of the PS double knockout, it was perhaps not surprising that the gene expression profile of double KO ES cells was substantially different from wild-type ES cells. Indeed, one percent of the total genes detected were altered in the double knockout ES cells.

Mindful of the burgeoning list of putative γ-secretase substrates (see below) Yankner further refined his analysis to attribute the observed changes to the two best characterized γ-secretase substrates, APP and Notch. Cells were transfected with constructs encoding an AICD-like fragment (the C-terminal 59 amino acids of APP) or NICD, RNA extracted and transcriptional profiles determined as before. Interestingly only seven genes were rescued by AICD: the kinesin receptor, Follistatin, TGFβ induced early growth response, caspase 7, Calbindin, Fragile X, and a novel gene product PRT. Of these the kinesin receptor (see Kamal et al. Nature 2002) and the calcium-binding protein Calbindin are particularly interesting. At the Elan satellite symposium on 22 July, Lennart Mucke also reported that Calbindin was elevated in AβPP transgenic mice. PRT, a novel gene cloned from ESTs with homology to the tetraspanins, was markedly increased in PS double knockout ES cells and elevated in ES cells treated with inhibitors of γ-secretase. The tetraspanins are inhibitors of metasis, involved in cell migration and synapse formation, and tetraspanin Kal1 has been shown to be directly activated by a ternary complex containing AICD (Baek et al. 2002 Cell 110, 55-67).

As one would predict based on knockout models of AβPP and Notch, transfection of ES cells with NICD recovered a substantially larger number of genes than the handful of genes rescued by AICD, with many of the NICD-rescued genes acting to stimulate cell migration. In addition, a third set of genes (including APOE and PrP) were altered in the PS double knockout cells but were not recovered by either AICD or NICD.

Although gene array analysis is still fraught with difficulties, and notwithstanding caveats about certain technical aspects of the reported study, the innovative approach of identifying transcriptional targets by comparison of PS double KO cells with wild type or ICD-rescued cells should aid in the identification of the gene targets regulated by ICDs of the various γ-secretase substrates.-Dominic M. Walsh, Ph.D. (Harvard Institutes of Medicine)

Putative Substrates for PS/γ-secretase

1. Erb4 (Todd Golde #549)

2. LRP*

3. APLP1 and 2*

4. Syndecan 3 (Bart DeStrooper #1043)**

5. CD44 (Kwang-Mook Jung & Tae-Wan Kim #782; Sven Laminch & C. Haass, have detected p3-like peptides #1059)**

6. Notch (Okochi & C. Haass, have detected p3-like peptides, #1059)

7. E-Cadherin (Robakis et al. #2079)

8. Delta/Jagged (Sisodia # 1572; LaVoie & Selkoe abstract# 658)**

9. Nectin 1 (Dora Kovacs #532)**

* These proteins were mentioned as γ -secretase substrates by several speakers but to my knowledge have yet to be confirmed.
** Novel presentation at this meeting.