The UPR turns on early during neurodegenerative conditions, including amyotrophic lateral sclerosis, Parkinson’s, Alzheimer’s, and other tauopathies (Atkin et al., 2008; Hoozemans et al., 2012; Nijholt et al., 2011; Ferreiro and Pereira, 2011). “The lessons learned through these axonal damage studies might have implications beyond injury-related cell death and neural repair,” noted Francesco Roselli and Pico Caroni of the Friedrich Meischer Institute in Basel, Switzerland, in a commentary accompanying the Neuron paper.

Study author Yang Hu, now at Temple University in Philadelphia, set out to understand the consequences of ER stress pathways after injury. Hu started the work in the Harvard lab of Zhigang He, co-senior author with Dong Feng Chen. Hu, a former ophthalmologist, chose to work with retinal ganglion cells because of their clear, straightforward anatomy. Together with co-first authors Liu Yang, Kevin Park, and Xin Wei in the Chen lab, Hu crushed the axons and purified the cell bodies to examine the expression of ER stress markers.

Upon ER stress, three biochemical pathways awaken, leading to the activation of three transcription factors: CHOP (CCAAT/enhancer binding homologous protein), XBP-1, and ATF6 (activating transcription factor 6). The precise downstream effects of these factors are far from straightforward. XBP-1 acts on multiple gene targets, including those responsible for ER membrane synthesis, protein-folding chaperones, and destroying misfolded peptides (Sriburi et al., 2004; Lee et al., 2003). Yet XBP-1 is not always beneficial; for example, it contributes to ALS-like symptoms in mice (see ARF related news story on Hetz et al., 2009). CHOP’s effects include starting up apoptosis, but it can also protect neurons (Oyadomari and Mori, 2004; Halterman et al., 2010). The sum of the three UPR pathways “can be anti- or pro-apoptotic depending on the trigger, intensity, and cellular context of UPR activation,” Roselli and Caroni wrote (Han et al., 2009).

Hu and colleagues focused on the activities of XBP-1 and CHOP, which have been studied more than ATF6. Since the ER stress pathways are thought to work together (Ron and Walter, 2007), Hu was surprised to find that the axon crush strongly upregulated CHOP for at least a week, but XBP-1 only slightly and transiently.

To examine what the two proteins do following axon injury, the researchers repeated the axon crush experiments in CHOP knockout mice or conditional knockouts missing XBP-1 in the retinal ganglia, and then counted surviving retinal neurons in the optic nerve. Getting rid of CHOP was beneficial: Fifty-two percent of the neurons lived past axon crush, while only 24 percent survived in wild-type controls. However, XBP-1 knockout did not protect the neurons. Surprised again, Hu said, “We thought, this molecule may have the opposite effect of CHOP.” When the researchers overexpressed XBP-1, 64 percent of neurons survived axon damage versus 20 percent in the control animals. “Taken together, these results suggest that XBP-1 and CHOP play opposite roles in controlling neuronal survival after axonal injury,” wrote the authors.

Hu repeated the experiments in a mouse model of glaucoma. This eye disease results from increased pressure inside the eye; the pressure squeezes the axons that reach into the eye’s fluid-filled cavity. Hu and colleagues injected beads into the eye’s anterior chamber to block fluid draining, boosting the internal pressure. As with the acute crush injury, the pseudo-glaucoma turned on CHOP but only a bit of XBP-1, and neurons degenerated. Adding XBP-1 or limiting CHOP rescued them. The XBP-1 treatment was effective even a week after the bead injection. This last scenario is more relevant to people, who would be treated after damage had occurred, noted Lawrence Wrabetz of the State University of New York at Buffalo in an e-mail to ARF. “This paper provides further support for the idea that ER stress responses are a mixed bag—some beneficial and some detrimental,” he added (Gow and Wrabetz, 2009).

Despite the UPR pathway, there is no evidence that misfolded proteins fill the retinal neurons upon axon injury, Hu noted in an e-mail to ARF. Thus, he speculated, the UPR activated in Alzheimer’s and other conditions might not be the result of misfolded proteins, but of axonal injury. “The results of Hu et al. suggest that, in neurons, a UPR can be an intrinsic response to disturbances in axonal integrity and flow, possibly unrelated to the load of un/misfolded proteins,” concurred Roselli and Caroni. This supports the idea that the UPR is not wholly involved in managing un- or misfolded proteins, but responds to a variety of cell stresses and demands (see review by Rutkowski and Hegde, 2010). Therapeutics promoting XBP-1 activity might be applicable in a number of neurodegenerative diseases, Wrabetz suggested.

Plenty of questions remain about these pathways. “We still do not know how XBP-1 protects these injured neurons,” Hu wrote in an e-mail. Also by e-mail, He speculated it might upregulate chaperone proteins. Keeping the cell bodies alive is only half the battle; they will need axons to function. It is not yet clear, Hu said, if adding XBP-1 or removing CHOP assists axons as well. “Whether this also can preserve the axon is still the question we need to answer,” he said.—Amber Dance.

References:
Hu Y, Park KK, Yang L, Wei X, Yang Z, Cho KS, Thielen P, Lee AH, Cartoni R, Glimcher LH, Chen DF, He Z. Differential effects of unfolded protein response pathways on axon injury-induced death of retinal ganglion cells. Neuron. 2012 Feb 9;73(3):445-52. Abstract

Roselli F, Caroni P. Life-or-death decisions upon axonal damage. Neuron. 2012 Feb 9;73(3):405-7. Abstract