This paper describes the application of cryoanalytical electron microscopy to the quantitative measurement of total calcium content and distribution at subcellular resolution in physiologically defined biological specimens.
In this article the tools, techniques, and instruments appropriate for quantitative measurements of intracellular elemental content using the technique known as electron probe microanalysis (EPMA) are described. Intramitochondrial calcium is a particular focus because of the critical role that mitochondrial calcium overload plays in neurodegenerative diseases. The method is based on the analysis of X-rays generated in an electron microscope (EM) by interaction of an electron beam with the specimen. In order to maintain the native distribution of diffusible elements in electron microscopy specimens, EPMA requires "cryofixation" of tissue followed by the preparation of ultrathin cryosections. Rapid freezing of cultured cells or organotypic slice cultures is carried out by plunge freezing in liquid ethane or by slam freezing against a cold metal block, respectively. Cryosections nominally 80 nm thick are cut dry with a diamond knife at ca. -160 °C, mounted on carbon/pioloform-coated copper grids, and cryotransferred into a cryo-EM using a specialized cryospecimen holder. After visual survey and location mapping at ≤-160 °C and low electron dose, frozen-hydrated cryosections are freeze-dried at -100 °C for ~30 min. Organelle-level images of dried cryosections are recorded, also at low dose, by means of a slow-scan CCD camera and subcellular regions of interest selected for analysis. X-rays emitted from ROIs by a stationary, focused, high-intensity electron probe are collected by an energy-dispersive X-ray (EDX) spectrometer, processed by associated electronics, and presented as an X-ray spectrum, that is, a plot of X-ray intensity vs. energy. Additional software facilitates: 1) identification of elemental components by their "characteristic" peak energies and fingerprint; and 2) quantitative analysis by extraction of peak areas/background. This paper concludes with two examples that illustrate typical EPMA applications, one in which mitochondrial calcium analysis provided critical insight into mechanisms of excitotoxic injury and another that revealed the basis of ischemia resistance.
Calcium ions are arguably the most important and versatile cell signaling entity in biology, playing an essential role in normal processes as diverse as synaptic transmission and gene expression. On the other hand, calcium is equally important in cell death. In particular, calcium deregulation is a key factor in neuronal injury in stroke, Parkinson's, Alzheimer's and other neurodegenerative diseases3,5. Thus, it is critically important to understand quantitatively how calcium is distributed within cells, and how this changes following physiological or pathophysiological stimuli. This goal is complicated by the fact that calcium is dynamically distributed between two physical states – free in solution or bound to a substrate – and that cellular calcium concentrations change over several orders of magnitude as a consequence of stimulation.
While there are several advanced methodologies available for the analysis of free intracellular calcium, the determination of total calcium concentrations in defined intracellular compartments is realistically limited to one approach, namely, electron probe microanalysis (EPMA). EPMA is a technique that couples an X-ray spectrometer to a transmission electron microscope (TEM). The TEM electron gun focuses a stationary, submicron electron probe on a subcellular region of interest and the element-specific X-rays emitted as a result of electron bombardment are collected and analyzed (see references7,4 for detailed technical reviews). Advantages of EPMA include single organelle-level resolution and submillimolar sensitivity. In practice, however, EPMA requires specialized cryotechniques and instrumentation for specimen preparation and analysis. Here, the tools, techniques and instruments appropriate for measurements of intracellular calcium using EPMA are described. Intramitochondrial calcium is of special interest on account of the critical role that mitochondrial calcium overload plays in neurodegenerative diseases.
The approach described here was developed using specific instruments, tools and software. Because labs will not be using the same experimental setup the approach is generalized where possible.
1. Rapid Freezing
The analytical method to be described is absolutely dependent on cryogenic approaches for: 1) the "cryofixation" of cells or tissues in a manner that quantitatively preserves the distribution of diffusible tissue components and chemical elements as they were in live cells at the instant of freezing; and 2) the preparation of ultrathin cryosections suitable for imaging and analysis in a TEM. These techniques are briefly described here, but details necessary to reproduce these procedures in other laboratories are beyond the scope of this article. Interested readers are referred to excellent recent articles9,14.
Caution: This section describes the use of liquefied ethane, which is highly flammable and potentially explosive; appropriate precautions should be taken. Operator must also be familiar with liquid nitrogen (LN2) safety precautions, including protective lab coat, glasses, and cryoresistant gloves. Note: Here and throughout, all tools (forceps, etc.) used to handle frozen specimens must be precooled in LN2 before use to avoid accidental thawing.
Cultured cells on coverslips
Rapid freezing of cultured brain slices
2. Cryosectioning
CAVEAT: Successfully producing dry-cut ribbons of ultrathin sections, while straightforward and logical, requires training, patience and considerable practice.
3. Cryotransfer of Specimens to the Electron Microscope
The core instrument for EPMA in this laboratory is an analytical electron microscope operated at 120 kV and equipped for cryomicroscopy, that is, designed with a clean vacuum, specimen-area anticontaminator, a 2k x 2k high-sensitivity digital camera and cryotransfer specimen holder. Check beforehand microscope alignment and operating conditions in both low- and high-magnification modes and confirm a satisfactory column vacuum, ideally ≤10-7 Torr. Tune up as necessary.
4. Visual Survey of Sections
5. Freeze-drying of Sections
6. Imaging of Cells and Organelles
Structural images of freeze-dried sections are obtained at ca. -160 °C as low-dose, zero-loss images recorded digitally using a 2k x 2k slow-scan CCD camera controlled by appropriate software.
7. Acquisition of X-ray Spectra
X-ray spectra can be recorded using any of several commercial or custom-designed X-ray analysis systems minimally consisting of an energy-dispersive X-ray (EDX) detector, associated pulse-processor electronics and compatible acquisition and display software. (The system used in this lab is described in Table 1.)
8. Analysis of X-ray Spectra
Qualitative analysis
An EDX spectrum (Figure 2, inset) is essentially an x-y plot of the X-ray intensity vs. energy. Spectra contain qualitative and quantitative information about the elemental composition of the analyzed volume, in that the "characteristic" energy of a peak identifies the element giving rise to that peak while the intensity reflects the amount of that element. The characteristic peaks ride on top of a slowly varying background, the "continuum". (The legend to Figure 2 further discusses salient details of EDX spectra.) The energy of the peak manifolds for the entire periodic table are defined by the well known electronic structure of elements, thus all EDX software links to a database that automatically identifies components of an analyte. In a physiological context, elements of general interest that are well suited to EDX analysis include Na (Kα at 1.04 keV), P (2.01 keV), K (3.31 keV), and Ca (3.69 keV).
Quantitative analysis
Quantitative analysis of EDX spectra consists of extracting the integrated area of identified peaks and converting this value to a concentration. For biological analysis, the established approach is the Hall peak/continuum method4,7,10, which takes advantage of the fact that the intensity of the continuum (defined above and in Figure 2) is proportional to the dry mass of the analyzed volume. Thus, the ratio of peak area/continuum area, when compared to the same ratio in spectra of standards of known composition, specifies the concentration in the targeted cellular compartment. Note that this approach provides concentrations in units of moles per weight, typically expressed as mmol/kg dry weight. This unit is unusual and for interpretation may require additional conversion to, for example, mmol/L wet weight or mmol/mg protein, as described elsewhere4,10.
Brain cells typically sustain excitotoxic injury as a result of the pathological neurotransmitter release that occurs under ischemic conditions. EPMA was critical to discovering how the ability of neuronal mitochondria to sequester massive amounts of calcium underlies the mechanism of injury. The electron micrograph in Figure 3 illustrates the appearance of mitochondria in freeze-dried cryosections of cultured hippocampal neurons rapidly frozen after 30 min exposure to an excitotoxic stimulus (100 μM NMDA). Most mitochondria appear to be structurally damaged, as they are highly swollen and contain small, dark inclusions (red arrows). EPMA, which has resolution sufficient to perform elemental analysis within and exclusive of (the "matrix") mitochondrial inclusions (Figure 3, red and blue spectra, respectively), revealed that the inclusions are largely composed of calcium and phosphorus. The extremely high calcium content of these inclusions – ~1,300 mmol/kg dry weight, whereas <1 mmol/kg is typical of resting mitochondria – explains their extraordinary calcium-buffering capacity, and why overwhelming this buffering mechanism leads to calcium overload-triggered cell death11.
The hippocampus, an area of the brain critical for learning and memory, is a major site of injury after ischemia. Interestingly, and therapeutically important, the functionally distinct region named CA3 is far less vulnerable to ischemic injury than is the adjacent, synaptically connected CA1 region. EMPA analysis – in conjunction with cryosections of specific regions of hippocampal slices that maintain in situ structure and function (Figure 4, micrograph) – was used to show that toxic stimuli induce much larger calcium elevations in vulnerable CA1 neurons than in resistant adjacent CA3 neurons (Figure 4, bar graph). Consequently, CA1 mitochondria exhibit extensive injury and dysfunction, indicating that Ca2+-overload-induced mitochondrial dysfunction is a determining factor in the selective vulnerability of CA1 neurons13.
Figure 1. Low-magnification electron micrograph of two ribbons of frozen-hydrated cryosections supported on a pioloform film. Characteristics of a good preparation are illustrated, including flat, well attached, minimally fragmented sections. Grid squares with rips in the support film should be avoided, as the film tears further under the electron beam. Two squares fully appropriate for analysis are indicated by asterisks. Scale bar = 100 μm.
Figure 2. Quantitative analysis of organelle elemental composition by EPMA. Low-dose, digital scanning transmission electron micrograph of sympathetic neuron in a freeze-dried cryosection prepared from rapidly frozen control ganglion illustrates the detail achievable in such specimens. Note the sharp plasma membrane, well preserved stacks of endoplasmic reticulum and abundant mitochondria. Inset – EDX spectrum recorded from a typical area of mitochondrial matrix, as indicated by the red dot. Elements corresponding to the major K shell X-ray peaks are identified; in the case of potassium and calcium, two K-shell lines arising from alternative electronic transitions are resolved, indicated as Kα and Kβ according to standard spectroscopic notation. Spectrum illustrates typical distribution of major elements in living neurons. Note the slowly varying continuum radiation, e.g. between 1.5-1.8 or 2.8-3.1 keV, which reflects the mass density of this cellular compartment. Quantitative analysis of calcium in healthy cells is complicated by the presence of relatively high levels of potassium (ca. 500 mmol/kg dry weight, approximately equivalent to 150 mM), which gives rise to overlap of the potassium Kβ and calcium Kα peaks in the EDX spectrum. There are standard algorithms for deconvolving this overlap. Scale bar represents 1 μm.
Figure 3. Excitotoxic stimulation induces variable and localized calcium accumulation in individual mitochondria. Digital transmission electron micrograph of freeze-dried cryosection prepared from unfixed, rapidly frozen hippocampal cell culture after excitotoxic stimulation of the NMDA subtype of glutamate receptors (100 μM NMDA for 30 min) shows numerous mitochondria containing small, naturally electron dense, punctate inclusions (red arrows). Corresponding X-ray spectra demonstrate that toxic stimulation resulted in loss of ion homeostasis, as evidenced by increased Na peak and reduced K peak in the inclusion-free mitochondrial matrix (blue arrows and spectrum). The large Ca peak in the red spectrum reflects strong mitochondrial calcium accumulation localized to characteristic inclusions, which also contain large amounts of phosphorus and oxygen. Average Ca concentration in inclusions and matrices was ~1,300 and ~60 mmol/kg dry weight, respectively. Scale bar represents 500 nm.
Figure 4. Mitochondrial calcium overload is responsible for selective ischemic vulnerability of hippocampal CA1 neurons. left panel – Electron micrographs of cryosections of cell bodies in the CA3 and CA1 regions of an organotypic hippocampal slice culture, rapidly frozen under ischemia-mimicking conditions (100 μM NMDA). right panel – EPMA analyses of mitochondrial calcium content shows that chemical ischemia induces much larger calcium elevations in mitochondria of vulnerable CA1 neurons than in adjacent, ischemia-resistant CA3 neurons (*, p<0.05 relative to NMDA-exposed CA1). NMDA-induced calcium overload was abolished by NMDAR antagonist MK-801. Scale bar represents 2 μm.
The electron microscope-based analytical method presented here allows for the detection, identification, and quantitation of several elements of biological interest, including Na, K, P, and especially Ca. These analyses can be carried out at subcellular, i.e., intra-organelle, resolution owing to the ability to locate and identify structures of interest in high-quality images of cryosections prepared from rapidly frozen specimens. Note that no staining is required to record electron images comparable in structural quality to conventionally fixed, plastic-embedded preparations, even while the location of tissue elements is quantitatively preserved.
Although structural survey images are recorded at low applied electron dose, much higher doses are required to elicit X-ray emission at count rates sufficient to obtain good statistics. Therefore, a microscope useful for EMPA must be able to form a high current focused probe; 2-5 nA into 25-50 nm, suitable for encompassing organelles like ER cisternae or synaptic vesicles, is acceptable. This minimally requires a high-brightness lanthanum hexaboride electron gun. Under these instrumental conditions, calcium at typical tissue concentrations of 1-10 mmol/kg dry weight can be quantitatively analyzed to a standard error of ±0.1 mmol/kg in ~100 sec live time. The sensitivity of calcium analysis would be greatly improved if it were not for the unfortunate overlap of the major calcium x-ray line with a minor potassium line, deconvolution of which adds significant uncertainty to the integration of the calcium signal (see Figure 1 legend for details). Note that the boundary conditions described are greatly relaxed when local calcium concentrations are unusually high, as occurs during physiological, and especially pathological, mitochondrial calcium accumulation (Figures 2-3).
Despite numerous successful applications, it is clear that EPMA is not a particularly efficient technique, so there is much incentive to improve data throughput. Three promising avenues are mentioned here: 1) electron energy loss spectroscopy (EELS) instead of EDX; 2) 2D element maps instead of point probes; and 3) newer, state-of-the-art instrumentation. Rather than collecting emitted X-rays, EELS depends on recording incident electrons that have lost characteristic, element-specific amounts of energy after electron/atom collisions. Statistically, EELS is inherently ~4x better than EDX, and EELS hardware and software are now practically mature for biologists. (See references6,2 for reviews of EELS applications in biology.)
The improved sensitivity offered by EELS – and also by the newer high-throughput Si-drift EDX detectors, see below – allows shorter dwell times per analysis, e.g. milliseconds rather than hundreds of seconds, and therefore the ability to generate digital maps of selected areas in reasonable times. As an example, a 128 x 128 calcium map can be recorded in ~1 hr1. Note that this approach, known as "spectrum imaging"6,2, provides a complete EELS spectrum at each pixel, so information on several elements is embedded in the recorded "data cube" (x vs. y vs. E).
Lastly, there have been substantive advances in instrument technology and performance since the present system, as described in the Protocol section, was developed over 20 years ago. State-of-the-art technology that would be de rigueur in an EMPA lab being setup today would include: 1) a high-brightness field-emission gun that can pump several nanoamps (nA) of current into subnanometer-sized spots; 2) a large area silicon-drift detector for maximized X-ray collection efficiency and throughput8; and 3) modern software offering superior flexibility and performance for spectral acquisition, analysis and quantitation. Such software packages are commercially available from most manufacturers; alternatively, an updated and improved version of DTSA software, DTSA II, is available at no cost from NIST (see Table 1, footnote 4). The features just described are on top of any automation and/or robotics available on the base electron microscope.
The EMPA protocol as described has proven itself highly useful for attacking several issues of interest to neuroscientists, helping, for example, to reveal cellular mechanisms of neurodegeneration. Considering the improvements just recommended, EMPA should continue to be a solid contributor to future research.
The authors have nothing to disclose.
We would like to thank Ms. Christine A. Winters for excellent technical assistance. This work was supported by the Basic Neuroscience Program of the NINDS Intramural Research Program, NIH (Z01 NS002610).
REAGENTS/MATERIALS | |||
Thermanox plastic coverslips | Thermo Fischer Scientific | 72280 | |
Culture inserts | BD Falcon | 353090 | For 6-well plates |
Cryopins | Leica Microsystems | 16701952 | Grooved |
Wood applicators | EM Sciences | 72300 | |
Folding EM grids | Ted Pella | 4GC100/100 | 100 mesh |
Indium foil | Alfa Aesar | 13982 | 0.25 mm thick |
EQUIPMENT | |||
Plunge freezing device | Leica Microsystems | KF-80 | |
Slam freezing device | LifeCell | CF-100 | |
Ultramicrotome | Leica Microsystems | UC6 | |
Cryoattachment for microtome | Leica Microsystems | FC6 | |
Diamond cryotrimming tool | Diatome | Cryotrim 45 | |
Diamond cryoknife | Diatome | Cryo 35 | |
Antistatic device | Diatome | Hauf Static Line | |
Cryo electron microscope | Carl Zeiss Microscopy | EM912 Omega | |
EM cryo specimen holder | Gatan | CT3500 | |
Slow-scan CCD camera, 2k x 2k | Troendle (TRS) | Sharpeye | |
Image acquisition software | Olympus SIS | iTEM suite | |
ED x-ray detector | Oxford Instruments | Linksystem Pentafet | |
Pulse Processor | Oxford Instruments | XP-3 | |
PCI backplane card | 4pi Systems | Spectral Engine II | |
Desktop computer | Apple | Any OS9-compatible model | |
X-ray analysis software | NIST | DTSA, DTSA II | |
Spreadsheet software | Microsoft | Excel | |
|