MICROTUBULE PROTEIN Identification in and Transport to Nerve Endings

The subunit protein of microtubules, tubulin, has been demonstrated to be present in isolated nerve endings by gel electrophoresis, amino acid composition, and peptide mapping. The tubulin constitutes approximately 28% of the soluble protein of the nerve endings . The transport of tubulin to the nerve endings has been demonstrated and its relationship to slow transport is discussed . ponent of transport, but this is based on indirect evidence. A second problem is that morphological studies (28) reveal few, if any, microtubules at nerve endings. Also, if tubulin is transported with slow flow, certain evidence suggests that slow flow may not enter the synaptic expansion itself (29) . To resolve these problems we have attempted to apply the axoplasmic flow technique of Barondes (11) in which flow is assessed by comparing the levels of labeled protein in isolated synaptosomes to levels of the same material in whole brain at varied time intervals after the injection of labeled precursor. The results demonstrate the transport of tubulin to the nerve ending and its presence in large quantity in the ending and lend support to the conclusion that this transport is in the slow component . THE JOURNAL OF CELL BIOLOGY . VOLUME 51, 1971 . pages 138-147 on Jauary 4, 2018 jcb.rress.org D ow nladed fom MATERIALS AND METHODS Synaptosomal Isolation CD 1 or BALB/c mice between the ages of 3 and 7 days were used for these experiments . Subcellular fractionation and isolation of the synaptosomal fraction were done by a slight modification of the procedure of Gray and Whittaker (10) as described previously (11) . The major modification involves the washing of the crude mitochondrial fraction three times with 0 .32 M sucrose to ensure removal of adsorbed soluble protein. The nerve ending (synaptosomal) fraction was lysed with 5 ml of water at 0 °C and the lysed preparation was centrifuged at 100,000 g for 1 hr to yield a soluble component of the nerve ending fraction (NES) and a particulate component (NEP). In the experiments in which radioactive amino acid incorporation into brain protein was studied, mice received a single intracerebral injection of 10 sCi of L-leucine 4,53H (4 mCi/mmole, New England Nuclear Corp ., Boston, Mass.) in 10 μl of isotonic saline in the temporal region of the brain 15 min, 90 min, or 24 hr before sacrifice . The whole soluble fraction (WS) was prepared by taking the supernatant from the crude mitochondrial preparation above and centrifuging it at 100,000g for 1 hr and utilizing the supernatant for further study while discarding the pellet . The fraction provides an index of neuronal cell body levels of labeled materials for comparison with the levels in isolated synaptosomes . A major problem here is that the WS contains not only neural but also glial cytoplasm and quite possibly cytoplasm from nerve endings which may have ruptured during initial homogenization . The former problem is eased slightly by the fact that the immature mouse brain is relatively poor in glial cells . Vinblastine Precipitation The protein concentrations of the 100,000 g supernatant from whole brain homogenate (WS) and from the soluble component of the nerve ending fraction (NES) were adjusted to 1-2 mg/ml, and MgC12, sodium phosphate buffer (pH 7.0), and vinblastine sulfate were added to final concentrations of 0 .01 M, 0.01 M, and 5 X 10-° M, respectively. The vinblastine sulfate was obtained from Eli Lilly & Co ., Indianapolis, Ind . The mixture was incubated at 37 °C for % hr (5) and the precipitate was collected by centrifugation at 35,000 g for 15 min . Control experiments revealed 1007 sedimentation of colchicine binding activity under these conditions, and polyacrylamide gels of supernatant and pellet revealed all protein which co-migrated with a tubulin standard to be precipitated . The purity of the precipitates was estimated at 90-95% by densitometric scans of gels stained in amido-Schwarz . Similar estimates of purity were obtained by fractionation and counting of the gels after radioactive labeling of the protein . Since the results presented in this paper are based on liquid scintillation counting and protein determination on the entire precipitate, they will overestimate the actual amount of tubulin by 5-10% of the experimental figures. Therefore, all determinations of quantities of tubulin in the following sections are diminished by 10% from their experimental values to compensate for this error. Scans on such precipitates are shown in Fig . 2 a. Since higher purity was desired for peptide mapping and amino acid composition studies, the vinblastine precipitates were resolubilized by dialysis against 0 .01 M phosphate buffer, pH 7 .0, and the final solution was clarified by centrifugation at 6000 g for 20 min. The final supernatant was then taken for study . The insoluble pellet amounted to 10-20% of the total protein of the original precipitate and contained up to 40% tubulin as determined by gel electrophoresis. The supernatant was 98-997 pure tubulin as determined by densitometry and counting of labeled protein . Typical gels of such a preparation are shown in Fig . 2 b. Polyacrylamide Gel Electrophoresis Protein samples were reduced in ß-mercaptoethanol and alkylated with iodoacetamide in the presence of 8 M urea (12), except in the case of samples to be run on neutral sodium dodecyl sulfate (SDS) gels in which case the urea was omitted . Three types of gel systems were used : (a) the neutral pH gel system incorporating 0.1 % sodium dodecyl sulfate (SDS) in which the proteins migrate according to molecular weight (13) . The bands in this system tend to be quite broad ; (b) the discontinuous gel system of Davis (14) modified by the inclusion of 8 M urea in the gel ; (c) the Davis system as modified by Albert et al . (15) with the inclusion of both 8 M urea and 0 .1% SDS in the gel and 0.1 % SDS in the buffer ; separation in this system is by molecular weight over the range of 30,000-135,000 daltons . The acrylamide concentration in each case was 7 .5%. Gels were fixed overnight in a 50/50 mixture of 7% acetic acid in water and 7% acetic acid in methanol . Staining was done for 8 hr in amido-Schwarz, followed by electrophoretic destaining in a transverse destaining apparatus. Amido-Schwarz was chosen for staining because it represented a reasonable compromise between sensitivity and quantitation . The more sensitive Coomasie blue is nonlinear for tubulin, while amido-Schwarz gave better agreement with actual protein concentrations. Samples of WS and NES stained in amido-Schwarz and in the more quantitative fast green (30) differed by only a small percentage for any single peak . Gels used for radioactive experiments were all run FEIT, DUTTON, BARONDES, AND SHELANSKI Microtubule Protein 1.39 on Jauary 4, 2018 jcb.rress.org D ow nladed fom with identical amounts of protein . The protein profiles within the NES and WS fractions remained constant with time as determined by densitometry . Differences exist, of course, between the profiles from NES and WS . Proteins were labeled with tritiated leucine, and a 14 C-labeled tubulin standard was used for calibration . Gel slices were solubilized for counting in NCS reagent and counted in Bray's solution (31) .