Expression of microRNAs and their precursors in synaptic fractions of adult mouse forebrain

Abstract

We have characterized the expression of microRNAs and selected microRNA precursors within several synaptic fractions of adult mouse forebrain, including synaptoneurosomes, synaptosomes and isolated post-synaptic densities (PSDs), using methods of microRNA microarray, real time qRT-PCR, Northern blotting and immunopurification using anti-PSD95 antibody. The majority of brain microRNAs (especially microRNAs known to be expressed in pyramidal neurons) are detectably expressed in synaptic fractions, and a subset of microRNAs is significantly enriched in synaptic fractions relative to total forebrain homogenate. MicroRNA precursors are also detectable in synaptic fractions at levels that are comparable to whole tissue. Whereas mature microRNAs are predominantly associated with soluble components of the synaptic fractions, microRNA precursors are predominantly associated with PSDs. For seven microRNAs examined, there was a significant correlation between the relative synaptic enrichment of the precursor and the relative synaptic enrichment of the corresponding mature microRNA. These findings support the proposal that microRNAs are formed, at least in part, via processing of microRNA precursors locally within dendritic spines. Dicer is expressed in PSDs but is enzymatically inactive until conditions that activate calpain cause its liberation; thus, we propose that synaptic stimulation may lead to local processing of microRNA precursors in proximity to the synapse.

Figure 1. Restriction enzyme digestion of qRT-PCR products for 4 microRNA precursors

PCR products were run (left lanes of each pre-miR as indicated) and gave the expected size appropriate to each microRNA precursor (pre-miR-99a = 62 bp; pre-miR-124a= 57–59 bp; pre-miR-125b1= 53 bp; pre-miR-134 = 58 bp) relative to lanes loaded with 10 bp DNA ladder (Invitrogen). Products were also incubated with restriction enzymes (middle lane) or incubated without enzymes (right lane) as indicated. Enzymes (New England Biolabs) were added to 10 ul PCR product + 1.3 ul enzyme buffer +2 ul water and incubated at 37C for 90 min, then resolved on a 15% TBE criterion gel (Bio-Rad) and stained with Syber Gold. Each restriction enzyme produced fragments of the expected sizes (as shown below the gel) appropriate to each microRNA precursor sequence.

Figure 2. microRNA intensity values in synaptoneurosomes vs. total forebrain homogenate as measured by microarray

Data were plotted before vs. after normalizing values to the mean value of the overall microRNA population as described in Results.

Figure 3

Enrichment ratio ± s.e.m. of forebrain microRNA intensity values (synaptoneurosomes/total homogenate) as measured by microarray

Figure 4. microRNA enrichment in synaptoneurosomes vs. total homogenate

Data were computed as the enrichment ratio (X-axis) vs. computed as quantile rank differences (Y-axis). See text for details.

Figure 5

Relationship between the microRNA intensity values observed in synaptoneurosomes and their enrichment ratios as measured by microarray

Fig. 6. Enrichment ratio (synaptoneurosomes/total homogenate) of various RNAs as measured by qRT-PCR

Data represent the average of 3 independent preps.

Figure 7. Distribution of synaptosomal proteins and RNAs in soluble vs. PSD fractions after extraction with Triton X-100

A. Immunoblotting. Total forebrain homogenate (T ) was processed to obtain a soluble cytoplasmic fraction (S2) and a synaptosomal fraction (Sy), that was then lysed with 1% TritonX-100 to yield (Ss) and insoluble fractions (Sp) as described in Methods. Equal amount of protein were loaded and blotted for different antibodies as indicated. The dicer antibody was chicken polyclonal anti-dicer (Lugli et al, 2005). B. qRT-PCR measurements of RNAs. Total RNA was prepared from Ss and Sp fractions and measured as described in Methods. The Sp/Sy ratio was calculated for 3 independent preps and is plotted as shown.

Figure 8. Within synaptoneurosomes, BC1 RNA and mature miR-124a are predominantly soluble components

Total forebrain homogenate (T), synaptoneurosomes (Syn), and Triton-soluble (Ss) and insoluble (Sp) fractions of synaptoneurosomes were prepared. Equal amounts of total RNA were separated on agarose gels, transferred to Nylon membranes and blotted with biotinylated LNA oligonucleotide probes for BC1 and mir-124a, respectively (see Methods). miR-124a was chosen as a representative microRNA because it is abundant, neuronal specific, and well characterized.

Figure 9. Co-immunoprecipitation of microRNAs and their precursors with PSD95 or synapsin I

Synaptosomes were diluted 1:1 with RIPA buffer and incubated with anti-PSD95 or anti-Synapsin I (see Methods). Negative controls include omitting antibody, or omitting synaptosomes. A. Equal fractions were loaded and immnostained for PSD-95 and Synapsin I. B. Total RNA was isolated from immunoprecipitates and measured for mature microRNAs (99a, 124a, 125b, 134, 143, 339) and their precursors (99a, 124a1–3, 125b1, 125b2, 134, 339) by qRT-PCR. Results were computed as the ratio between the amount of each RNA brought down by each antibody and the corresponding amount brought down in the baseline group lacking specific antibody. Because results were similar for all microRNAs and for all precursors, the data were pooled and displayed as red bars (precursors) vs. blue bars (mature microRNAs).

Figure 10. Co-immunoprecipitation of mature microRNAs and microRNA precursors with different proteins

Total forebrain homogenate was spun at low speed to remove nuclei and then the S1 low speed cytoplasmic supernatant was prepared (1500xg for 10 min; final concentration 1.8 mg/ml). This was diluted 1:1 with RIPA buffer; each tube had 800 ul final volume. Immunoprecipitations were performed with a variety of affinity purified antibodies (commercially available or made and characterized in Lugli et al. 2005): a) mouse monoclonal anti-PSD95 (clone 28/86); b) rabbit polyclonal anti-eIF2c that recognizes multiple isoforms; c) chicken polyclonal anti-dicer recognizing a region (1389–1404) located between the first and second RNAse III domains; d) rabbit polyclonal anti-dicer recognizing the C-terminus; e) mouse monoclonal anti-FMRP (Brown et al., 2001; clone 7G1-1, Developmental Studies Hybridoma Bank, Iowa City, IA); f) rabbit polyclonal anti-PACT (ProteinTech Group Inc., Chicago, IL); g) rabbit polyclonal anti-MeCP2 (Upstate Biotech, Lake Placid, NY). As a negative control, rabbit polyclonal anti-Synapsin I was used (Chemicon, AB1543P). All antibodies brought down their respective proteins as expected (fig. 7, 9 and data not shown). However, no detectable GAPDH or CAMK2a mRNA were detected in any groups (data not shown). Total RNA was extracted and each immunoprecipitate was measured for mature microRNAs (99a, 124a, 125b, 134, 143, 339) and their precursors (99a, 124a1–3, 125b1, 125b2, 134, 339) by qRT-PCR. Results were computed as the ratio between the amount brought down by each antibody and the corresponding amount brought down by anti-synapsin I. Because results were similar for all microRNAs and for all precursors, the data were pooled and displayed as red bars (precursors) vs. blue bars (mature microRNAs). Note the log scale.