Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2015 Apr 3.
Published in final edited form as: J Microsc. 2007 Dec;228(0 3):384–389. doi: 10.1111/j.1365-2818.2007.01856.x

Modular software platform for low-dose electron microscopy and tomography

MICHAEL P MARSH †,‡,*, JUAN T CHANG †,‡,*, CHRISTOPHER R BOOTH †,‡,*, NATHAN L LIANG ‡, MICHAEL F SCHMID †,‡,#, WAH CHIU †,‡,#
PMCID: PMC4384816  NIHMSID: NIHMS36574  PMID: 18045333

Summary

Transmission electron microscopy imaging protocols required by structural scientists vary widely and can be laborious without tailor-made applications. We present here the jeol automated microscopy expert system (james) api integrator, a programming library for computer control of transmission electron microscopy operations and equipment. james has been implemented on JEOL microscopes with Gatan CCDs but is designed to be modular so it can be adapted to run on different microscopes and detectors. We have used the james api integrator to develop two applications for low-dose digital imaging: james imaging application and the mr t tomographic imaging application. Both applications have been widely used within our NCRR-supported Center for routine data collection and are now made available for public download.

Keywords: Automation, cryoEM, data acquisition, software

Introduction

A modern transmission electron microscope can be employed in a number of different ways. It can be used to observe specimens ranging in scale from complete cells down to individual atoms from biological or non-biological origins. The techniques used to visualize structures at these different scales require a diverse combination of microscope accessories and detectors, often from several different manufacturers. Frequently, many of these components require computer control to be used effectively. Unfortunately, the software to run these components is rarely shared between companies, resulting in partially integrated components, supporting only limited experimental tasks. In particular we found no available solution for low-dose imaging that integrated software control of our Gatan CCD cameras with the fastem software package (Fukushima et al., 2000) that controls our JEOL microscopes. To meet our needs but also establish a more general solution, we found it necessary to develop a package that handles communication between electron microscope components and provides researchers with a single interface to their equipment for software engineering.

We have chosen to develop a modular software platform called jeol automated microscopy expert system (james), which is a programming library and set of prototype applications to assist the skilled microscopist with routine microscopy tasks on JEOL microscopes running fastem. james does not attempt to oversimplify the technique, but instead allows the operator to stay in control of most of the decisions throughout the course of a microscopy session. This approach has allowed our operators to quickly prototype new image collection protocols and optimize data quality without having to spend time working with software packages from multiple manufacturers.

Materials and methods

Hardware

All development and testing were carried out on a JEM-2010F and a JEM-3000SFF (JEOL USA, Peabody, MA), 200- and 300-keV transmission electron microscopes. These microscopes are under computer control managed by JEOL’s fastem 3.00 application (Fukushima et al., 2000). fastem is a client/server application, enabling the microscopist to use a remote fastem client as a graphical microscope control interface that delegates directives across the network to the fastem server connected directly to the microscope.

Both microscopes are connected to Gatan Ultrascan 4000 slow-scan CCD cameras (US4000) (Mooney, 2007) (Gatan, Pleasanton, CA). The US4000 is a Peltier-cooled 4096 × 4096 camera with a four-port readout. Camera control is managed by Gatan’s digital micrograph (dm) application (gms 1.6) (Mitchell & Schaffer, 2005), which runs on a workstation proximal to the microscope.

Programming languages and libraries

All c++ programming was carried out with visual studio.net (Microsoft Corp., Redmond, WA). python code was written to be compatible with the python 2.3 specification and later validated with python 2.4 (http://www.python.org). Access to the c++ libraries from python was established with the boost tool kit for c++ (http://www.boost.org). Our james-enabled workstations run on either windows 2000 or windows xp (Microsoft Corp.).

Results and discussion

JAMES API INTEGRATOR

Software architecture

To implement microscope control software in a general way, we have developed the james application programming interface (api) integrator. The james api integrator is a programming library that scientists can use to develop individualized microscopy applications. The james api integrator unifies the microscope and its disparate accessories in one library so that application developers do not need to learn vendor-specific APIs for each component they wish to control. By decoupling all of the microscope and camera control from the highest layer of the programming interface, we have isolated the application programmer from all of the details of a particular vendor’s microscope or camera. Yet the software has been designed to be modular and replaceable; cameras can be added or replaced, and, in principle, the microscope itself could be replaced by a newer microscope of a different brand or model. Fig. 1 illustrates the overall layout of software in the james api integrator.

Fig. 1.

Fig. 1

Open in a new tab

JAMES architecture. JAMES unifies the software from JEOL and Gatan into a single PYTHON API. This is achieved by writing C++ modules that communicate with FASTEM and DM and interface with PYTHON using BOOST.

fastem interface

JEOL offers a client/server system called fastem for controlling the microscope. The fastem server is connected directly to the microscope and can receive communication from multiple fastem clients. JEOL provides the fastem server and a basic fastem client API that was written in c (http://www.jeolusa.com). We wrapped this fastem client API in c++ to abstract some of the existing libraries and make them callable from python (Fig. 1). python is a popular scripting language, which has been widely adopted by the open source community and especially by the structural biology community (Hohn et al., 2007). Exporting the c++ API to a python API was carried out with the boost toolkit.

DIGITAL MICROGRAPH INTERFACE

dm provides a rich scripting language called dmscript. It is similar to c++ and can be used to control the operation of the CCD camera. Based on a template provided by Gatan, we implemented our dm client API in c++. It communicates directly to dm by sending dmscript commands and allows us to make dm functions callable from python. We again used the boost toolkit for this purpose (Fig. 1).

james api integrator layer

The james api integrator unifies the underlying layers described above and makes the interface to the microscope and its accessories visible to the scientists in one simplified API. For example, the scientist uses the james api integrator to interrogate the microscope at runtime to determine which detectors (i.e. photographic film camera or CCD camera) are available, and then selects which detector to use. The software will appropriately delegate function calls to the chosen detector. The usefulness of this library is made evident by the two imaging applications described below.

james: a single-particle application

Transmission electron microscopy of ice-embedded biological specimens (electron cryomicroscopy or cryoEM) is a technique by which large numbers of macromolecules in different orientations are imaged and averaged together to produce a 3D reconstruction. Current data sets require several thousands to hundreds of thousands of particle images. The requirement for so many images has prompted a number of groups to propose and implement automation of data collection to reduce the amount of time that an operator must spend in the highly repetitive task of image acquisition (Potter et al., 1999; Stagg et al., 2006; Zhang et al., 2003).

The james single-particle application (Fig. 2) developed with the james api integrator is our approach towards automation in the JEOL microscope. The application supports low-dose imaging by controlling the microscope and the CCD camera with attention to beam-blanking and incidental specimen exposure. Without software integration, the seemingly simple proposition of acquiring and saving a single image was surprisingly tedious. It required a ten-step procedure of navigating the interfaces of fastem, dm and the microscope console.

Fig. 2.

Fig. 2

Open in a new tab

GUI for JAMES single-particle application. The GUI is integrated with a view of the database. The panel on the left shows the available microscopy sessions in the database to which acquired micrographs should be attached. The main panel shows the status of the microscope and parameters that control the behaviour of the ‘Acquire’ button, such as choice of camera and choice of single exposure or focal pair. The tabs at the top of the interface correspond to unrelated operations that can be performed on database entries.

  1. Console: select an area to image

  2. fastem: engage the beam blank

  3. Console: set the desired objective defocus

  4. Console: raise the viewing screen

  5. fastem: disengage the beam blank

  6. dm: acquire the image

  7. dm: enter a filename and save the image

  8. dm: close the image

  9. fastem: engage the beam blank

  10. Console: lower the viewing screen

This procedure becomes more complicated if the user wants to take multiple exposures of the same area such as a focal pair. Because james manages all of the devices for the user, the procedure is simplified. The james interface allows the user to make a persistent setting for either one exposure or two, each at a prescribed defocus. Each time the user pushes the ‘Acquire’ button on the james interface, all of steps 2–10 described above are performed automatically, significantly reducing the repetitive tasks as follows.

  1. Console: select an area

  2. james: acquire an image or focal pair

After the image is collected, the user can select the next area for imaging and repeat.

The example above omitted the step of recording the details of the acquired images (such as defocus, exposure time, grid position, etc.), which can obviously increase the time required for manual data collection. james is integrated with the electron microscope electronic notebook, a database for archival of images and electron microscope metadata (Ludtke et al., 2003). This integration permits james to query the microscope for relevant metadata and then upload this data into the electron microscope electronic notebook in an automated way when acquiring images. The result is a more detailed account of the acquired images than a user is likely to record by hand.

Several projects have now been completed using data collected with the james single-particle application (Booth et al., 2004; Chang et al., 2006; Jiang et al., 2006a,b; Lee et al., 2003; Ludtke et al., 2004, 2005; Mao et al., 2004; Marsh et al., 2006; Pope et al., 2007; Yu et al., 2005). Each of these projects consists of many single-particle images collected and used for reconstruction. In many cases the data collected with james are sufficient for structural determination to subnanometer resolution. Figure 3(A) shows an example of using james to record images of ice-embedded virus particles at 83100× effective magnification onto the Gatan US4000 CCD camera and its corresponding power spectrum of the boxed-out particle images (Fig. 3B). A contrast transfer function ring is generally detectable beyond 9 Å after the background is properly subtracted from the 1D averaged power spectrum. In this project, james software was used to acquire 428 focal pairs spanning five imaging sessions, resulting in 6465 unique particles that were used to compute an icosahedral reconstruction at 9 Å resolution (Fig. 3C), in which long alpha helices and large beta sheets can be identified (Booth et al., 2004). The other projects followed similar imaging parameters.

Fig. 3.

Fig. 3

Open in a new tab

Cytoplasmic polyhedrosis virus images collected by JAMES and corresponding reconstruction. (A) A 200-keV image of cytoplasmic polyhedrosis virus is acquired with this software. Following microscope alignment and setup of FASTEM minimum dose system, areas of interest were identified in search mode, and JAMES was used to record all high-resolution images in photo mode. Images were acquired as focal pairs. (B) Power spectrum of the first image as computed according to Saad et al. (2001). The signal-to-noise ratio extends beyond 9 Å. (C) A slice through the 9 Å resolution reconstruction. Scale bar represents 1000 Å.

mr t: a tomography application

Tomographic imaging is a process by which images of a single area are acquired while tilting the goniometer through a range of angles. Many groups have automated tomographic data collection (Daberkow et al., 1996; Mastronarde, 2005; Nickell et al., 2005; Zheng et al., 2007; Ziese et al., 2002). However, there is currently no computer software available for electron cryotomography of dose-sensitive specimens on JEOL microscopes running fastem software. Because the james api integrator offers a unified interface that can control stage tilt as well as image acquisition, we have used it to develop a software package called mr t to facilitate the acquisition of low-dose tilt series. The result, which includes a simple graphical user interface (GUI) (Fig. 4), permits routine acquisition of tilt series on a JEOL 2010F microscope, even by inexperienced users.

Fig. 4.

Fig. 4

Open in a new tab

GUI for MR T tomographic imaging application. The GUI allows the user to enter parameters for the tilt experiment, including the tracking method, tilt range, tilt step and exposure time. In addition, the tilt plan (rightmost panel) can be arbitrarily modified.

mr t offers many features offered in tomography solutions implemented by other investigators for their own instruments. To keep the area of interest in the field of view, mr t offers specimen tracking by cross-correlation or by precalibration (Ziese et al., 2002). mr t’s data collection routine permits flexibility in the angular sampling, allowing for fixed interval or Saxton interval tilting (Saxton et al., 1984). Users can also choose to have all exposures acquired with equal dose or have their dose weighted by the secant of the tilt angle to compensate for the weaker signal-to-noise ratio at higher tilt angles, as implemented in serial em (Mastronarde, 2005) and also described in methods of Medalia et al. (2002). mr t’s approach to dose management allows the user to specify a total allowable dose, and then the software plans the dose partitioning for each exposure to meet these requirements without exceeding this limit.

Several biological samples have been imaged and reconstructed using mr t, including lipid vesicles (Chang et al., 2005), cells(Marsh et al., 2005) and Herpes Simplex Virus capsid (Chang et al., 2007). Figure 5 shows our results from a 2-h session using mr t in manual-tracking mode to record a cryotomographic series of lipid vesicles at 34600× effective magnification and total cumulative dose of 30 e− Å−2. The imod reconstruction package (Kremer et al., 1996) was used to generate a tomogram from 61 images spanning −60° to +60° in 2° increments. A single image from the tilt series (Fig. 5A) does not resolve the features of the vesicle organization. The tomogram reveals three nearly concentric vesicles that are annotated as cut-away spheres (Fig. 5C). A slice of the tomogram (Fig. 5B) shows the organization of these three vesicles, in which the leaflets of the bilayer are resolved.

Fig. 5.

Fig. 5

Open in a new tab

Electron cryotomography of lipid vesicles. (A) Single image of frozen-hydrated vesicles and colloidal gold acquired with MR T at 0° tilt, 200 keV. A reconstruction was performed with the IMOD package using the colloidal gold for fiducial alignment. (B) A slice through the tomogram, reconstructed from 61 images, shows three concentric vesicles. (C) The same slice with hollow spheres modelling the position of the vesicles. The yellow sphere shows the position of the colloidal gold. Scale bar represents 300 Å.

Conclusion

Experimental science continues to push beyond the limitations of the instruments and the software. The james platform was born out of the need to perform specific tasks with the transmission electron microscopy, and cryoTEM in particular, that were not well supported by commercial or publicly accessible software. By implementing a modular design for a core programming interface to control the microscope, we have achieved a multipurpose library with the james platform. The library and the applications can be downloaded from our website (http://ncmi.bcm.tmc.edu/software/james).

We have shown how the library has been used to develop prototype applications for single-particle and tomographic imaging of frozen-hydrated specimens of biological origin. These applications can be extended into feature-complete solutions, similar to automated applications reported for other instruments (Mastronarde, 2005; Nickell et al., 2005; Stagg et al., 2006; Zheng et al., 2007). The strength of the james platform is its modularity, which permits replacement of any microscope component without changing the application layer but instead by merely implementing the low-level code for the new component.

Acknowledgments

The research has been supported by National Institutes of Health National Center for Research Resources (P41RR02250), National Institutes of Health through the NIH Roadmap for Medical Research (PN2EY016525), National Institute of General Medical Sciences (P01GM064676) and the Robert Welch Foundation. We thank Jaap Brink and Bob O’Donnell (JEOL), Jacob Wilbrink, Robin Harmon and Doug Hauge (Gatan) for their technical assistance. The vesicles were kindly provided by Ka Yee C. Lee and Guohui Wu (U Chicago).

References

  1. Booth CR, Jiang W, Baker ML, Zhou ZH, Ludtke SJ, Chiu W. A 9 Å single particle reconstruction from CCD captured images on a 200 kV electron cryomicroscope. J Struct Biol. 2004;147:116–127. doi: 10.1016/j.jsb.2004.02.004. [DOI] [PubMed] [Google Scholar]
  2. Chang JT, Marsh M, Rixon F, Chiu W. Mr T. Automated electron cryotomography for JEOL 2010F TEM. Microsc Microanal. 2005;11:308–309. [Google Scholar]
  3. Chang J, Weigele P, King J, Chiu W, Jiang W. Cryo-EM asymmetric reconstruction of bacteriophage P22 reveals organization of its DNA packaging and infecting machinery. Structure. 2006;14:1073–1082. doi: 10.1016/j.str.2006.05.007. [DOI] [PubMed] [Google Scholar]
  4. Chang JT, Schmid MF, Rixon FJ, Chiu W. Electron cryotomography reveals the portal in the herpesvirus capsid. J Virol. 2007;81:2065–2068. doi: 10.1128/JVI.02053-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Daberkow I, Koster AJ, Tietz HR, Typke D, Walz JP. A system for automated electron tomography using Philips CM series transmission electron microscopes. Philips Electron Opt Bull. 1996;134:27–31. [Google Scholar]
  6. Fukushima K, ODonnell R, Fujiwara K, Kai H, Okunishi E, Kawasaki M, Kersker M, Naruse M. Computer controlled high-throughput integration system: FASTEM. Microsc Microanal. 2000;6:1144–1145. [Google Scholar]
  7. Hohn M, Tang G, Goodyear G, et al. SPARX, a new environment for Cryo-EM image processing. J Struct Biol. 2007;157:47–55. doi: 10.1016/j.jsb.2006.07.003. [DOI] [PubMed] [Google Scholar]
  8. Jiang W, Chang J, Jakana J, Weigele P, King J, Chiu W. Structure of epsilon15 bacteriophage reveals genome organization and DNA packaging/injection apparatus. Nature. 2006a;439:612–616. doi: 10.1038/nature04487. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jiang X, Jayaram H, Kumar M, Ludtke SJ, Estes MK, Prasad BV. Cryoelectron microscopy structures of rotavirus NSP2-NSP5 and NSP2-RNA complexes: implications for genome replication. J Virol. 2006b;80:10829–10835. doi: 10.1128/JVI.01347-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Kremer JR, Mastronarde DN, McIntosh JR. Computer visualization of three-dimensional image data using IMOD. J Struct Biol. 1996;116:71–76. doi: 10.1006/jsbi.1996.0013. [DOI] [PubMed] [Google Scholar]
  11. Lee S, Sowa ME, Watanabe YH, Sigler PB, Chiu W, Yoshida M, Tsai FT. The structure of ClpB: a molecular chaperone that rescues proteins from an aggregated state. Cell. 2003;115:229–240. doi: 10.1016/s0092-8674(03)00807-9. [DOI] [PubMed] [Google Scholar]
  12. Ludtke SJ, Nason L, Tu H, Peng L, Chiu W. Object oriented database and electronic notebook for transmission electron microscopy. Microsc Microanal. 2003;9:556–565. doi: 10.1017/S1431927603030575. [DOI] [PubMed] [Google Scholar]
  13. Ludtke SJ, Chen DH, Song JL, Chuang DT, Chiu W. Seeing GroEL at 6 Å resolution by single particle electron cryomicroscopy. Structure. 2004;12:1129–1136. doi: 10.1016/j.str.2004.05.006. [DOI] [PubMed] [Google Scholar]
  14. Ludtke SJ, Serysheva II, Hamilton SL, Chiu W. The pore structure of the closed RyR1 channel. Structure (Camb) 2005;13:1203–1211. doi: 10.1016/j.str.2005.06.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Mao Y, Vyas NK, Vyas MN, Chen DH, Ludtke SJ, Chiu W, Quiocho FA. Structure of the bifunctional and Golgi-associated formiminotransferase cyclodeaminase octamer. EMBO J. 2004;23:2963–2971. doi: 10.1038/sj.emboj.7600327. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Marsh MP, Chang J, Chen J, Lopez J, Chiu W. The 3-dimensional architecture of platelets in the native state by electron cryotomography. Blood. 2005;106:473A. [Google Scholar]
  17. Marsh MP, Campos SK, Baker ML, Chen CY, Chiu W, Barry MA. Cryoelectron microscopy of protein IX-modified adenoviruses suggests a new position for the C terminus of protein IX. J Virol. 2006;80:11881–11886. doi: 10.1128/JVI.01471-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Mastronarde DN. Automated electron microscope tomography using robust prediction of specimen movements. J Struct Biol. 2005;152:36–51. doi: 10.1016/j.jsb.2005.07.007. [DOI] [PubMed] [Google Scholar]
  19. Medalia O, Weber I, Frangakis AS, Nicastro D, Gerisch G, Baumeister W. Macromolecular architecture in eukaryotic cells visualized by cryoelectron tomography. Science. 2002;298:1209–1213. doi: 10.1126/science.1076184. [DOI] [PubMed] [Google Scholar]
  20. Mitchell DR, Schaffer B. Scripting-customized microscopy tools for digital micrograph. Ultramicroscopy. 2005;103:319–332. doi: 10.1016/j.ultramic.2005.02.003. [DOI] [PubMed] [Google Scholar]
  21. Mooney P. Optimization of image collection for cellular electron microscopy. Methods Cell Biol. 2007;79:661–719. doi: 10.1016/S0091-679X(06)79027-6. [DOI] [PubMed] [Google Scholar]
  22. Nickell S, Forster F, Linaroudis A, Net WD, Beck F, Hegerl R, Baumeister W, Plitzko JM. TOM software toolbox: acquisition and analysis for electron tomography. J Struct Biol. 2005;149:227–234. doi: 10.1016/j.jsb.2004.10.006. [DOI] [PubMed] [Google Scholar]
  23. Pope WH, Weigele PR, Chang J, et al. Genome sequence, structural proteins, and capsid organization of the cyanophage Syn5: a ‘horned’ bacteriophage of marine synechococcus. J Mol Biol. 2007;368:966–981. doi: 10.1016/j.jmb.2007.02.046. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Potter CS, Chu H, Frey B, et al. Leginon: a system for fully automated acquisition of 1000 electron micrographs a day. Ultramicroscopy. 1999;77:153–161. doi: 10.1016/s0304-3991(99)00043-1. [DOI] [PubMed] [Google Scholar]
  25. Saad A, Ludtke SJ, Jakana J, Rixon FJ, Tsuruta H, Chiu W. Fourier amplitude decay of electron cryomicroscopic images of single particles and effects on structure determination. J Struct Biol. 2001;133:32–42. doi: 10.1006/jsbi.2001.4330. [DOI] [PubMed] [Google Scholar]
  26. Saxton WO, Baumeister W, Hahn M. Three-dimensional reconstruction of imperfect two-dimensional crystals. Ultramicroscopy. 1984;13:57–70. doi: 10.1016/0304-3991(84)90057-3. [DOI] [PubMed] [Google Scholar]
  27. Stagg SM, Lander GC, Pulokas J, et al. Automated cryoEM data acquisition and analysis of 284742 particles of GroEL. J Struct Biol. 2006;155:470–481. doi: 10.1016/j.jsb.2006.04.005. [DOI] [PubMed] [Google Scholar]
  28. Yu X, Acehan D, Menetret JF, et al. A structure of the human apoptosome at 12.8 Å resolution provides insights into this cell death platform. Structure (Camb) 2005;13:1725–1735. doi: 10.1016/j.str.2005.09.006. [DOI] [PubMed] [Google Scholar]
  29. Zhang P, Borgnia MJ, Mooney P, et al. Automated image acquisition and processing using a new generation of 4K × 4K CCD cameras for cryo-electron microscopic studies of macromolecular assemblies. J Struct Biol. 2003;143:135–144. doi: 10.1016/s1047-8477(03)00124-2. [DOI] [PubMed] [Google Scholar]
  30. Zheng SQ, Keszthelyi B, Branlund E, Lyle JM, Braunfeld MB, Sedat JW, Agard DA. UCSF tomography: an integrated software suite for real-time electron microscopic tomographic data collection, alignment, and reconstruction. J Struct Biol. 2007;157:138–147. doi: 10.1016/j.jsb.2006.06.005. [DOI] [PubMed] [Google Scholar]
  31. Ziese U, Janssen AH, Murk JL, Geerts WJ, VanderKrift T, Verkleij AJ, Koster AJ. Automated high-throughput electron tomography by pre-calibration of image shifts. J Microsc. 2002;205:187–200. doi: 10.1046/j.0022-2720.2001.00987.x. [DOI] [PubMed] [Google Scholar]

RESOURCES