Report of the 2008 Summer Internship Program

Report of the 2008 Summer Internship Program

Mengke Xing

Host Employer: Johns Hopkins University, School of Medicine

Advisor: Dr. Alex Kolodkin

During the two-a-half-months internship program, I mainly devoted my time to answer some simple questions: what kind of interaction does Nrp2 has with PDZ domain containing protein in the Post Synaptic Density (PSD) region? And what part does Nrp2 has in the formation of dendritic spines during the nervous development? The background information that led me to ask those questions was the significant phenotypes observed on Nrp2 knock-out mice, where spines of the pyramidal neurons in layer 5 cortex as well as of DG neurons in hippocampus were larger and longer than those in wild type mice, which suggests that Nrp2 might have specific roles in dendrite and especially synapse formation. Still, at the very C terminus of Nrp2, which is a type 1 transmembrane receptor to the secreted neuron guidance cue Sema 3F, there is a PDZ binding motif composing of 3 amino acids, SEA, that is implied to binding to PDZ domain containing proteins, which are abundant in the PSD region. So the logic suggests that Nrp2 maybe localized to the post-synaptic area through its interaction with PSD scaffold proteins, and therefore participates in synapse formation.

At first, in a search to find potential proteins that bind to Nrp2, I collected cell

lysates from embryonic and adult cortex and hippocampus tissues respectively, and performed pull-down experiments using a polyclonal Nrp2 antibody. As control, proteins were also pulled down by a nonspecific IgG antibody from the same species. I intended to look for a difference in the bands after running the samples on a polyacrylamine gel and silver staining the gel. However, after several repeats, I didn’t find any band particularly interesting. Then I tried running the lysates and blotting it with Nrp2 antibodies, I realized that the endogenous protein concentration was not high enough for the silver staining experiment to be successful. And the low specificity of the antibody was another problem.

As a result, I took a different approach and tried in vitro pull-down experiments. Firstly, I designed two sets of primers to PCR up the entire sequence of Nrp2 and the one without the last three amino acids, i.e. the PDZ binding motif. Both sequences had an N terminal flag tag. Then I inserted the two fragments into pCAG-IRES-EGFP vector, and transfected 293 cells respectively. Vectors encoding known PDZ domain containing proteins were transfected along with Nrp2 constructs. This time, a flag antibody was used in the pull-down experiment, followed by Western blotting analysis to examine if Nrp2 had any interaction with the PDZ domain containing protein. As it turned out, some interesting preliminary results were collected. However, some repeats and more control experiments must be done before reaching any conclusions.

A quicker and easier way to examine the interaction of Nrp2 and PSD proteins was to look at the sub-cellular localization of Nrp2. In order to do that, I inserted the whole and minus SEA sequences of Nrp2 into a different vactor, pCAG-Cl, to

make N-terminally EGFP fusion proteins. After transfection of neuronal cells with the two constructs respectively, fluorescent microscopy was used to detect the localization of the fusion protein. However, only vague signal could be seen under the microscope. It is probably due to the low efficiency of transfection and/or expression, which was supported by Western blotting experiments on the cell lysates.

The Nrp2 I used in the previous experiments was the a17 isoform. There were 5 other isoforms of Nrp2 present due to splicing variance. Particularly, the b isoform of Nrp2 had different temporal expression forms as compared to a isoform. Moreover, another receptor in the Neuropilin family, Nrp1, seemed to have different roles in terms of dendrite formation. So it would be intriguing to inquire the interaction specificity of each receptor, Nrp2 a, Nrp2 b, and Nrp1, to PSD scaffold proteins. Given the reason above, I also made constructs of Nrp1 and planned to make constructs of Nrp2 b.

To sum up, different approaches were used in order to reveal the likely interaction between Nrp2 and PDZ domain containing scaffold proteins in the PSD region. However, a number of problems were encountered in these experiments.

Besides training on bench experiments, I also got a chance to give a group presentation in the lab meeting on a recently released Cell paper. As I prepared for the presentation, I held several informative discussions with Dr. Kolodkin, and gained great knowledge and perspective in the field of neuron guidance.

In a word, I have really learned a lot from this experience and I appreciate Dr. Kolodkin, and Dr. Rao for this program. I also want to thank Tracy Tran for her enormous help during the experiments; she has been a great mentor to me.