Neurofilament phosphorylation and axon diameter in the squid giant fibre system.

Electron energy loss spectroscopic analysis of squid giant axons in a phosphorus energy window yielded bright signals, which were shown to originate from highly phosphorylated neurofilaments. The frequency and distribution of these signals were analysed at defined intervals in cross-sections of the giant axon, starting from its origin in the stellate ganglion and extending distally along the stellar nerve. The analysis revealed a proximodistal gradient of increasing neurofilament phosphorylation. Within the stellate ganglion and for some distance beyond, the increase in frequency of signals correlated with the widening of the neurofilament meshwork and the radial growth of the axon. This agrees with the hypothesis that neurofilament phosphorylation regulates axon calibre by affecting interfilament spacing. In distal axon domains where the axon diameter diminished, contrary to expectations, the spacing of signals increased and the signals were significantly larger. Hyperphosphorylation apparently compensated for a diminishing supply of neurofilament protein. Contrary to predictions, the presynaptic terminal of the giant synapse contained a distinct and highly phosphorylated neurofilament meshwork. We conclude that the growth of the axon diameter is a function of neurofilament phosphorylation, interfilament spacing and neurofilament density. A mature and highly phosphorylated neurofilament cytoskeleton completely filled the presynaptic terminal of the giant synapse.

Videodensitometric analysis of electron spectroscopic micrographs--a tool for detection of biologically relevant elements with high resolution.

Electron energy-loss spectroscopic imaging (ESI) yields high-resolution, element-sensitive images. However, ESI suffers from difficulties in distinguishing element-specific and background contributions. New methods have therefore been introduced which use grey-level measurements in micrographic images for a more accurate detection of element distributions. A videodensitometric method allowed the detection of low phosphorus levels in axoplasmic neurofilaments of squid giant axons. Here we further verify these results by investigating the relationship of videodensitometry and electron energy-loss spectroscopy (EELS), particularly considering the peculiarities of these methods in terms of automatic background correction and representation of the results. Six biological specimens and two nonbiological specimens were examined both by EELS and by videodensitometry. In all cases comparable results were obtained. The overlapping PL2,3 and SL2,3 ionization edges could clearly be recognized individually by both methods, and controls showed that mass density variations within the specimens did not impair elemental analysis. Additional evidence supporting the detection of phosphorylation sites in squid neurofilaments was obtained in both EELS and videodensitometric measurements of neurofilament-enriched pellets and of aggregated axoplasmic particles. Thus, video-densitometry appears to be a useful tool for an improved exploration of the full imaging capabilities of energy filtering electron microscopy.

Detection of low phosphorus contents in neurofilaments of squid axons by Image-EELS contrast spectroscopy.

We show that Image-EELS is suitable for detecting relatively low phosphorus concentrations in very small axoplasmic structures of squid axons. Imaging plates and a CCD camera were used as electron sensors. From image series spanning a certain energy-loss range EELS (electron energy-loss spectra) were derived by averaging read-outs from many axoplasmic particles (APs). The ratio of these spectra to spectra of the background was plotted, showing the contrast modulation as a function of the energy loss. This new approach is called EELC (electron energy-loss-dependent contrast spectroscopy). A distinct phosphorus signal was found in APs of presynaptic terminals of the squid giant synapse, in the peripheral giant axon and, as controls, in ribosomes. Biochemical experiments supported this result. In neurofilament-enriched pellets a phosphorus signal could be directly detected by serial EELS and in electron spectroscopic micrographs. After dephosphorylation of either the pellets or the extruded axoplasm with alkaline phosphatase, phosphorus signals in electron spectroscopic micrographs were absent or much reduced in size and intensity. With Image-EELS inherent limitations of traditional element detection modes in energy filtering transmission electron microscopy can be overcome. Compared with serial EELS, the selective analysis of small areas with irregular shape is possible with greatly improved signal-to-noise ratio. The identification of the element-peak in Image-EEL spectra directly proves the presence of the element within the region of interest. For small peaks, the visualization is facilitated by the contrast presentation (EELC). However, the background subtraction modes used for elemental mapping in electron spectroscopic imaging are subject to uncertainties when elemental ionization edges like the P1,2,3 edge are examined. Imaging plates are very sensitive electron sensors with a wide dynamic range. Unlike photographic emulsions, they allow acquisition of image series covering a large energy-loss range without normalization of exposure times, and direct extraction of EEL spectra. Thus, the combination of Image-EELS and imaging plates is proposed as an efficient new tool for analytical electron microscopy.

High-resolution imaging of protein phosphorylation in squid axons and synapse by electron energy loss spectroscopy.

We demonstrate that clusters of phosphorus atoms can be detected in energy loss spectroscopic images (ESI) of cytoskeletal proteins of squid axons. In series of images taken at four energy windows below and three windows above the phosphorus P-L2,3 ionization edge, signal-to-background intensity differences were analyzed by videodensitometry. A distinct increase of relative intensities was recorded above the phosphorus edge in neurofilaments of the peripheral giant axon and in those of the presynaptic terminal. A high level of neurofilament phosphorylation in the peripheral axon supports previous biochemical and immunochemical results, but our finding of phosphorylated neurofilaments in the presynaptic axon conflicts with these studies. Our method may be advantageous for analysis with high elemental and spatial resolution of the phosphorylation state of cytoskeletal protein molecules in situ.