22 October 2010. Could the field be one step closer to a new potential therapy for Alzheimer disease? Possibly, if the target is striatal-enriched tyrosine phosphatase (STEP). Deletion of this gene restored cognition and repaired synapses in AD transgenic mice, according to a report published online October 18 in PNAS. Researchers led by Paul Lombroso of Yale University School of Medicine in New Haven, Connecticut, bred triple transgenic (3xTg-AD) mice onto a STEP knockout background, and found that their offspring performed as well as wild-type littermates on several tests of learning and memory. Given that these improvements came without reducing Aβ or tau pathology, and that STEP-deficient mice appear to have normal cognition, scientists say hitting STEP could provide a welcome alternative to the more typical Aβ- and tau-targeting therapeutic approaches for AD.
STEP is a brain-specific tyrosine phosphatase expressed primarily at post-synaptic terminals of neurons in the striatum, hippocampus, cortex, and related brain areas (Boulanger et al., 1995; Xu et al., 2009). Excess STEP has been found in J20 APP-overexpressing mice (Chin et al., 2005). More recently, Lombroso and colleagues reported the same in another AD mouse strain (Tg2576) and in human prefrontal cortical tissue (Kurup et al., 2010). Lombroso’s work also revealed connections between Aβ and STEP, one being that STEP dephosphorylates glutamate receptor subunits, thereby promoting Aβ-induced internalization of the receptors from synaptic sites. Secondly, in AD, excess Aβ inhibits the proteasome, leading to uncontrolled build-up of STEP protein. “We hypothesized that if STEP levels are elevated in AD, and if that's contributing to loss of glutamate receptors, then maybe we could reverse cognitive deficits in AD mice by reducing STEP levels,” Lombroso told ARF.
To test this idea, first author Yongfang Zhang, who is now at Shanghai Jiaotong University Medical School in China, and colleagues crossed triple transgenics with STEP knockout mice previously generated in the Lombroso lab (Venkitaramani et al., 2009). At six months of age, Aβ and tau pathology starts to develop in 3xTg-AD mice and they already have trouble finding the platform in the Morris water maze, a spatial memory task. However, when STEP-deficient triple transgenics were put through the maze at this age, they were “indistinguishable from wild-type mice,” Lombroso said. Without STEP, 3xTg-AD mice also had no problems negotiating the Y maze, which tests spatial working memory, and recognizing objects, a task that tests non-spatial hippocampal-dependent memory. Consistent with their cognitive prowess, 10-month-old STEP-deficient triple transgenics had enhanced long-term potentiation, as measured by field potential recordings from CA1 hippocampal slices. Surprisingly, these benefits came without any effects on underlying amyloid or tau pathology.
In six-month-old mice, biochemical changes at the level of NMDA receptors supported the behavioral data. As seen with other AD strains, hippocampal samples from 3xTg-AD mice had higher STEP expression than wild-type samples. Furthermore, in line with STEP’s role in promoting NMDA receptor internalization, triple transgenic mice had reduced phosphorylation at tyrosine 1472 of the NR2B subunit, as well as lower total amounts of NR2B and NR1 subunits—but no change in levels of NR2A, which is not thought to be a STEP target. In STEP-deficient triple transgenics, phospho-NR2B and NR1/NR2B levels shot up even higher than normal, also consistent with STEP’s known effects on glutamate receptors. The researchers reported similar biochemical data in Tg2576/STEP knockout mice, suggesting that the NMDAR effects are not strain-specific. (Because the cognitive phenotype in Tg2576 mice is subtler and generally not detectable until at least 12 months of age, STEP-deficient Tg2576 mice were not analyzed for behavior, Lombroso told ARF.)
The biochemical studies included analysis of Fyn and ERK, as both are STEP substrates. As expected, the researchers found phosphorylated (i.e. active) forms of each at higher levels in 3xTg mice lacking STEP, compared to wild-type and regular triple transgenic mice. In this scenario, increased fyn activation seems beneficial because it correlates with improved cognition in STEP-deficient 3xTg-AD mice. The data sets up an apparent conundrum, however, when considered alongside a recent report by Jürgen Götz and Lars Ittner at the University of Sydney in Australia. That study found that tau binds fyn and targets it to NMDA receptors, ultimately leading to increased Aβ-induced excitotoxicity (Ittner et al., 2010 and ARF related news story).
Scientists say it is hard to reconcile these incompatibilities because of differences in the mouse strains analyzed in each study. Triple transgenics have less Aβ deposition than other APP-overexpressing AD stains (e.g., APP23 used in Ittner et al., 2010), wrote Greg Cole of the University of California, Los Angeles, in an email to ARF. Given that Aβ itself induces downregulation of glutamate receptors (Snyder et al., 2005 and ARF related news story), strains with varying degrees of amyloid pathology could respond differently when crossed onto a STEP knockout background, Cole noted. Furthermore, Gloria Lee of the University of Iowa, Iowa City, pointed out that the mice in the Götz paper and in the current study also differ in terms of tau, which recruits fyn to NMDA receptors. Götz’s mice had no tau or just a truncated tau, whereas the triple transgenics overexpress mutated human full-length tau.
Another factor making it difficult to compare the studies on the fyn issue is that STEP has many additional substrates besides fyn, including downstream components of the NMDA receptor signaling cascade, Götz noted via email.
Nevertheless, STEP looks promising as a therapeutic target for AD. “The fact that STEP knockout mice appear to have normal LTP and no obvious impairment in other aspects of cognition would seem to make STEP a good target for therapy,” suggested Peter Davies of Albert Einstein College of Medicine in the Bronx, New York (see full comment below).
Timing may be an issue, though. The mice in this study were evaluated at an age when Aβ and tau pathology was just starting to develop, leaving some question as to how well this resembles AD patients who may already have lots of plaques and tangles, Cole noted. In subsequent studies of older (nine-month-old) STEP-/- 3xTg-AD mice that were not included in the paper, “you start to not see as robust a change (in preserving cognition),” Lombroso told ARF. “This suggests that at some point, the neurodegenerative effects may kick in.”
In addition, it is uncertain whether genetic manipulations present from birth will translate to interventions begun later in life. “These types of genetic studies do not demonstrate what might happen if STEP is inhibited only after AD pathological changes begin, as would be expected for any therapeutic approach,” Rebeck wrote (see full comment below).
Lombroso also acknowledged the shortcomings of genetic analyses and said his lab is trying to identify small molecules that inhibit STEP activity. In the meantime, he remains hopeful about STEP as a therapeutic target—not only for AD, but potentially other neurological disorders as well. Lombroso has unpublished data showing that STEP is elevated in Fragile X syndrome and schizophrenia, and that reducing STEP can alleviate symptoms in a Fragile X mouse model.—Esther Landhuis.
Zhang Y, Kurup P, Xu J, Carty N, Fernandez SM, Nygaard HB, Pittenger C, Greengard P, Strittmatter SM, Nairn AC, Lombroso PJ. Genetic reduction of striatal-enriched tyrosine phosphatase (STEP) reverses cognitive and cellular deficits in an Alzheimer's disease mouse model. Proc Natl Acad Sci U S A. 2010 Oct 18. Abstract