Report of The Structural Biology Subcommittee of The Biological and Environmental Research Advisory Committee

REPORT
of the
STRUCTURAL BIOLOGY SUBCOMMITTEE
of the
BIOLOGICAL AND ENVIRONMENTAL RESEARCH ADVISORY COMMITTEE

In response to the charge letter of Dr. Martha Krebs, May 28, 1998

Executive Summary

Structural biology and especially macromolecular crystallography are playing an increasingly important role in biological discoveries. In order to achieve these breakthroughs, macromolecular crystallography has dramatically expanded its use of the nation's synchrotron sources. These developments have led to the reconvening of the Structural Biology Subcommittee of the Biological and Environmental Research Advisory Committee (BERAC) to review the needs and to make recommendations as to the resources and processes that are necessary for the proper support of the macromolecular crystallographic community at the synchrotrons. The Subcommittee met on July 13, 1998; this document is the product of those deliberations. The major issues addressed by the Subcommittee and its recommendations are as follows:

Improvements recommended for current beamlines

(for the purposes of this report, beamline is defined throughout as an independently operating station, i.e., an experimental station/hutch.)

Since the number of beamlines currently in use or in construction for macromolecular crystallography has increased significantly in the past six to seven years, it was the sense of the Subcommittee that, from an efficiency, productivity, and timeliness perspective, the first priority should be to upgrade these existing beamlines to maximize utilization. This involves the following investments:

  1. Staffing: increases are recommended in staffing at the rate of one FTE/beamline, on average, to improve round-the-clock user support, increase efficiency of operation, and better provide for less-experienced and non-specialist users. The Subcommittee regards this as a minimal increase relative to what is really needed for a fully efficient and functional operation.
    Estimated cost: annual incremental expenditure of $3M/yr (~$125K/FTE).

  2. Hardware upgrades: improved hardware such as newer CCD detectors, backup detectors to avoid time wasting catastrophic failures, and improved optics and instrumentation to increase beam intensity or broaden versatility of a beamline.
    Estimated cost: initial expenditure of ~$10M. Estimated annual capital requirement of $250K/yr/beamline to maintain effective state of hardware (some of which is already accounted for in existing budgets).

    Improved access process for synchrotron beamtime and user education

    Increased use and modern demands of research requires changing the way beamtime is allocated:

    1. Rapid access to beamtime within a month or two using rolling reviews.
    2. Regional or facility-wide access system for time slot applications.
    3. Increase number of longer term "program" allocations.
    4. Provide for "instant access" for hour-long time blocks for sample evaluation.
    5. Development of common cross-facility user-friendly graphics interfaces across beamlines.
    6. Institute programs, such as workshops, to educate less-experienced and non-specialist users.

    General Facility Operations and Upgrades

    1. The Subcommittee supports the recommendations of the Birgeneau/Shen Basic Energy Sciences Advisory Committee (BESAC) report urging continued funding of sustained basic operations for all four DOE synchrotrons. The Subcommittee also supports the beamline upgrades for APS and NSLS (priority 2) and the facility upgrades for NSLS and SSRL (priority 2.5), recommended by Birgeneau/Shen, as necessary for continued support of the macromolecular crystallographic community.

    2. Funding of the operating shortfall of CHESS by NSF at ~$0.5M/yr.

      Future research efforts: Detector, automation, and methodology development

      To foster continued expansion of facility access, efficiency, and capabilities, the Subcommittee recommends support of research in:

      1. Advanced detectors
      2. Develop beamline automation
      3. Methodology development, such as optimizing data collection, processing, and refinement strategies to enhance synchrotron use. Development of new web-based tools to permit active access to beamline experimentation by remote users.

    3. New beamlines

      A number of crystallography beamlines are currently in the commissioning, development, or proposal stage at six of the synchrotron facilities. The Subcommittee urges a case-by-case review of new beamline proposals by funding agencies with weight given to innovative applications.

      I. Introduction and charge of the Subcommittee

      The progress of molecular biological research depends increasingly on knowledge of macromolecular crystal structures. Advances in our understanding of many basic biological systems, as well as powerful new methodology for structure-based drug design, have come directly as a result of the empowering information provided by atomic resolution macromolecular structures derived from crystallography. The growing demand for new crystal structures has created an urgent need to improve access to synchrotron radiation for macromolecular crystallography. Consequently, the past decade has seen a rapid expansion of structural biology and, especially, macromolecular protein crystallography at the nation's synchrotrons. This expansion and the impact it is having can be measured in a number of ways. In the course of just five years (see Table 1), the number of structures deposited in the protein databank has almost tripled, the number of new structures published using synchrotron radiation has grown from 16% to 40% and the number of beamlines at the nation's six synchrotrons with capabilities for protein crystallography has increased five-fold. The

      Table 1 - Growth of Structural Biology1

      1990-1991

      1995-1996

      Structures deposited in Protein Databank (PDB)

      512

      1437

      Percent of new structures using synchrotrons

      16%

      40%

      Number of beamlines for crystallography

      5

      24

      1

      Data are taken from the BioSync 1991 and 1997 reports.


      increase in beamlines has in part been due to the introduction of two new synchrotron sources in the past three years. The current synchrotron sources currently having or developing x-ray structural biology programs in the United States are summarized in Table 2. Four of these synchrotron sources were built as high energy rings and two as lower energy ones. However, the lower energy rings (ALS and CAMD) can produce high brightness x-rays for macromolecular crystallography by using specialized high field magnets.

      Table 2 -- Current Synchrotron Sources in the United States Supporting Structural Biology

      Synchrotron Source

      Designation

      Site

      Agency Funding for Basic Operation

      Advanced Light Source

      ALS

      LBNL

      DOE

      Advanced Photon Source

      APS

      ANL

      DOE

      Center for Advanced Microstructures and Devices

      CAMD

      LSU

      State of Louisiana

      Cornell High Energy Storage Ring

      CHESS

      Cornell

      NSF

      National Synchrotron Light Source

      NSLS

      BNL

      DOE

      Stanford Synchrotron Radiation Laboratory

      SSRL

      SLAC

      DOE

      Over the past year, three different studies have been conducted that have reviewed the use of synchrotron radiation by the structural biology community. The Birgeneau/Shen report of the Basic Energy Sciences Advisory Committee (BESAC) of the Department of Energy, the 1997 BioSync report (an update of the original report issued in 1991), and the Hodgson/Lattman report all examined various facets of this issue. These reports document the recent expansion of the crystallographic use of synchrotrons and forecast a sustained, continued, dramatic growth in the use of synchrotron sources for macromolecular crystallography for the foreseeable future. As a consequence of this conclusion, the Structural Biology Subcommittee was requested to consider the opportunities facing the granting agencies of the United States government in providing for the anticipated needs of macromolecular crystallography at synchrotron facilities. While a number of techniques comprise the synchrotron usage of structural biology, the Subcommittee was requested to focus on macromolecular crystallography because this discipline currently forms the largest segment, by far, of structural biology users at synchrotrons and is placing the highest demands on capacity.

      The issues considered by the Subcommittee were devoted to two major areas of concern:

      1. Consideration of the expected user needs

      a. Expected rate of increase of usage
      b. What types of facilities will be needed: monochromatic, MAD, Laue, high brightness, etc.
      c. How should beamtime be scheduled: need for rapid access, immediate access, should scheduling be regional, by beamline, or national system
      d. User support vs. methods development
      e. "Non-specialist" user needs at beamlines
      f. Issues of training and education

      2. Assessment of the status of existing facilities and need for further facilities

      a. Beamline staff needed
      b. Beamline hardware needed
      c. Increase in number of beamlines
      d. Synchrotron upgrades

      After discussion of the above issues, the recommendations of the Subcommittee, in order of importance, were as follows:

      II. Improvements recommended for current beamlines

      The Subcommittee concluded that the optimization of existing beamline utilization for protein crystallography will provide the greatest benefit most rapidly and is, therefore, of highest, most immediate priority. This can be most readily achieved through staffing increases and hardware upgrades at existing or soon to be completed beamlines. Such investments are viewed as a very cost effective way to leverage the current considerable investment in beamlines and user support for protein crystallography research and significantly increase capacity to meet the existing and growing level of overdemand.

      Staffing

      The two BioSync reports (1991 and 1997) and the recent Hodgson/Lattman report have documented the chronic understaffing of protein crystallography beamlines; the 1997 BioSync report concludes: "The most important need is for increased funding support for scientific and technical staffing at the synchrotron facilities." Appropriate staffing impacts the efficient operation of beamlines at every level: user support, proper hardware operation, implementation of hardware upgrades and new technologies, beamline administration and scheduling, safety issues, and user training. The user-support issue was noted by the Birgeneau-Shen BESAC subcommittee to be a substantial concern at many beamlines, and one can anticipate that this need will only increase with the rise in less-specialized and "non-specialist" users. The staffing requirement for hardware operation, maintenance, and upgrading is self-evident. Finally, the effective implementation of rapid access protocols to beamlines (see below), listed as a substantial concern by users in the BioSync survey, will require appropriate staff to implement and administer development of streamlined web-based tools for proposal review and rapid scheduling.

      Staffing levels must be adequate to permit round-the-clock usage to maximally utilize the available synchrotron radiation. The storage rings operate typically 24 hours/day, 7 days/week, 9 or more months/year, thus creating a significant burden on sustaining effective and reliable beamline operations. Current staffing levels at protein crystallography beamlines average ~2.5 FTE/beamline, as documented in the Hodgson/Lattman survey of synchrotron facilities; on average, at least one additional FTE/beamline was indicated as being required by the facility operators to have a reasonably effective level. The Subcommittee discussed appropriate staffing levels at length, and concluded that 3.5 to 4 dedicated FTE/beamline on average was a minimaltarget for the goal of optimizing efficient beamline utilization and maximizing throughput. It might be noted that this is still well below the optimal level of 5 to 7 recommended by the first BioSync report.

      These 3.5 to 4 FTEs represent (1) a staff scientist (or beamline scientist) responsible for overall beamline operation as well as participating in research and technological developments; (2) a beamline engineer (typically B.S. or M.S. level) who understands the major systems, optics, etc., as well as providing user support; (3) a technician, primarily for mechanical issues such as alignment, crystal changes, checkouts, maintenance, etc.; (4) a part-time administrator to schedule beamtime, arrange training, etc.; and (5) software and electronics technician support (assuming the beamline technician handles mechanical problems) are also required for some subset of beamlines. If these beamlines have common detectors, operating software, etc., then more efficient staff utilization can be achieved. Under these conditions, one software and one electronics technicians could service ~4 beamlines. Scheduling and administration of beamlines requires ~1 FTE/4 beamlines. Safety officers are also essential, and will likely require additional staffing with the development of new sample handling techniques for crystal cooling, gas pressurization, and heavy atom derivatization (particularly with increased sample throughput), in addition to monitoring of any biohazards. Thus, a total of 3.5-4 FTEs are needed per beamline (assuming sharing of the software, electronics, administrative and safety support among about 4 total beamlines).

      It is important to point out that the minimal staffing levels recommended above do not, however, provide for significant scientific support that would be needed by large numbers of less experienced and "non-specialist" investigators.

      Budgetary considerations:

      The cost of adding an incremental 1FTE/beamline, on average, to the ~24 protein crystallography beamlines comes to a total annual incremental expenditure of ~$3M, assuming $125K average cost including benefits and overhead per FTE (based on the Hodgson/Lattman report).

      Hardware Upgrades

      Many protein crystallography beamlines have proposed upgrading their present detector with either a CCD or a later generation, faster readout image plate system (Hodgson/Lattman Report). By reducing the detector readout time, the efficiency of synchrotron beam utilization can be increased by a factor between two to four. This would have an immediate and dramatic effect on sample, and hence user, throughput. However, it must be recognized that increased operations costs are associated with these more advanced detectors. A related and significant request is for backup detector capabilities, in the event of accidental damage or other catastrophic failure of critical detectors. The consequences of a detector failure can create loss of weeks of data collection time for scheduled users who may have waited months for access to the beamline. Ideally, each facility would have a backup detector to minimize loss of beamtime due to instrument failure. Where a commonality of detectors exists, such backup detectors could be "shared" by several beamlines.

      In addition to detector replacements/upgrades, other components of the beamlines, in particular the optical components like mirrors and monochromators, have finite lifetimes and need to be replaced periodically due to radiation damage or to accommodate ring upgrades. More recent developments in optics and instrumentation could also be profitably retrofitted into some existing beamlines to improve their performance and reliability, as well as confer new capabilities (e.g. rapid tunability for MAD or polychromatic capability for Laue data collection).

      Budgetary considerations:

      From the facilities surveyed in the Hodgson/Lattman report, initial hardware upgrades of ~$10M were requested for protein crystallography beamlines. For budgeting purposes, the average annual capital requirement for hardware upgrades is estimated to be ~$250K/beamline (this figure was provided by Dr. Roland Hirsch of DOE-BER based on average capital needs of currently operating DOE supported beamlines), for a total of ~$6M/year for 24 beamlines.

      General budgetary considerations:

      It should be noted that DOE-BES, DOE-BER, and NIH-NCRR are already investing some funding to address these issues and completely new funds in these amounts are not required. The details would need to be sorted out when the appropriate review and allocation mechanism for new funds was defined and initiated. Furthermore, it should be emphasized, that while some of the beamlines referred to as currently operational for macromolecular crystallography are privately resourced, the funding requested here will significantly enhance the general user access that is required of all beamlines at the current synchrotron beamlines.

      III. Improved access process for synchrotron beamtime and user education

      Increasing the capacity and number of beamlines by implementing the recommendations outlined above and below will produce the most significant improvement in beamtime for the user population. However, in order to optimize user access to these facilities, important procedural improvements are urgently needed in two areas. The first is to develop a suitable process to streamline the review of proposals and speed the assignment of beamtime. The second is to broaden access to synchrotron radiation within the biological community.

      Under the current system for review of proposals and assignment of beamtime, the time between beam-time request and experiment is generally very long (many months). This is poorly matched to today's rapid pace of research in structural molecular biology and to the highly competitive nature of many structural projects. We recommend that operators of synchrotron facilities and the user community, with encouragement from the funding agencies, develop a plan for more rapid and streamlined access for macromolecular crystallography. The committee discussed several specific ideas for improving access.

      • A rolling review of proposals for single experiments should be instituted, requiring neither specific proposal deadlines nor review committee meetings. Review times should ideally be about a month and, therefore, beamtime for such experiments would be scheduled only about two months in advance.
      • Since the overwhelming number of users preferred regional access to a synchrotron source, we encourage development of a single point of entry for general-user proposals at each synchrotron facility and perhaps regionally. It would also be useful to standardize the beamtime access application forms across facilities, since users often have to apply to more than one facility in order to obtain beamtime in a timely fashion.
      • The number of long-term "programs" of structural research, as implemented at some synchrotron facilities, should be expanded to include more specialist investigators with funded research programs in structural biology.
      • Where possible, a system for "instant access" to hour-long blocks of time for sample evaluation should be implemented.
      • Since users receive time at different beamlines on different visits, there is a significant benefit to be gained by developing a common cross-facility user-friendly graphics interface that would present the user with the same or similar user environment regardless of which beamline they are working out on this particular visit. This is especially important for the less expert users whose participation in synchrotron research is increasing.

      Macromolecular crystallography is rapidly becoming a standard tool of biological research. This trend speeds the advancement of biology by expanding the array of problems that are subjected to structural study. Thus, it is important to broaden access to synchrotron radiation for non-specialist users in the molecular biology community. Improved training for non-specialist users is important in addition to providing beamtime and support for experiment. We strongly encourage development of workshops and other venues for training non-specialist or less specialist users in the practice of macromolecular crystallography.

      Budgetary considerations:

      There will be some administrative overhead needed for a centralized facility or regional access process as well as for organizing and executing effective training programs and workshops. However, the budgetary costs to implement appropriate access mechanisms and educational programs are relatively small.

      IV. General Facility Operations and Upgrades

      The Subcommittee discussed the need for effectively sustaining basic operations and upgrades of the synchrotron facilities currently being used for, or with potential to soon contribute to, beamtime for structural molecular biology experiments.

      Highly reliable and stable operation of the storage ring sources is one of several key and essential components in the successful development and remarkable growth of synchrotron-based protein crystallography over the last decade. Sustainable running of the storage rings for typically 9 or more months per year with minimal down time and high beam stability enables users to rely upon synchrotron sources for their x-ray diffraction data collection needs. This will increasingly be the case if the access issues discussed earlier in this report can be mitigated. Hence it is important to recognize the importance of providing adequate funding for the ongoing operations of the facilities themselves and for facility upgrades that improve overall performance and reliability to the benefit of the whole user community which includes a growing fraction of structural biologists at all of the facilities.

      Continued effective operation of the CHESS facility is an issue of immediate concern. The Subcommittee endorses the conclusion of the Birgeneau/Shen report that a single funding agency should maintain responsibility for the operating costs of a given facility. However, the shortfall of CHESS operating funds at the currently projected level of funding will likely have an immediate and negative impact on protein crystallography users through reduced beamline support. We urge that the NSF allocate additional funds to address the projected shortfall of about $0.5M/yr in the CHESS operations budget.

      The four DOE-funded synchrotron sources (APS, ALS, NSLS, and SSRL) have recently undergone an exhaustive and thorough review by the Birgeneau/Shen BESAC subcommittee. The report issued by this group in October, 1997 directly addresses and prioritizes the funding needs for ongoing operation and upgrades of these four facilities. In particular, the Birgeneau/Shen report recommends continued operation of all four of the DOE synchrotrons, funding for new beamline and front end upgrades at NSLS and APS, and major facility upgrades for NSLS and SSRL. Front ends and beamlines at APS and improvements to the NSLS PRT beamlines are relevant to enhanced access and future developments which are important to the structural biology community. The major facility upgrade at SSRL (bringing its performance to the level of a 3rd generation source) is important given its positive impact on the existing stations and user community and the clear benefits for many experiments for very high brightness beams. Likewise, components of the NSLS Phase III upgrade will also directly benefit the structural biology users through beamline and insertion device developments.

      This BERAC Subcommittee concludes that full implementation of the specific recommendations of the Birgeneau/Shen Report (i.e., funding for 1,2,2.5 and 3) is essential to both the immediate and longer term health and growth of the field. The Subcommittee did not discuss that part of recommendation 1 that deals with 4th generation R&D since it was outside the purview of the Subcommittee's charge. In light of the science being enabled by synchrotron-based structural molecular biology research, we urge that DOE give consideration to the above recommendations in setting their funding priorities for FY 1999 and beyond.

      Budgetary considerations:

      The budgetary aspects of this recommendation for DOE synchrotrons are covered in detail in the Birgeneau/Shen BESAC report. The request for addressing the shortfall in funding for CHESS is $0.5M/yr. Interagency cooperation on both of these issues is urged by the Subcommittee.

      V. Future research efforts: Detector, automation, and methodology development

      In order to foster continued expansion of facility access and improved capabilities at existing and new beamlines, the Subcommittee recommends continuing research and development of detector systems and automation. These two priority items are described below.

      Research in advanced detectors for x-radiation

      One of the most critical and often rate limiting steps for utilization of the beamtime is the detector used to collect the scattered x-radiation. While image plate detectors currently offer sufficient definition for large unit cells, read out time is relatively slow. Charge coupled devices (CCDs) have much faster read out times, but these times may still be limiting if the exposures are rapid enough and do not have the resolution necessary for very large unit cells to be collected efficiently. With the very high intensity sources from wiggler and undulator insertion devices, current technology does not provide detectors to take maximum advantage of the potential capabilities. Research is ongoing in a number of laboratories to develop more advanced detectors using different methodologies, including pixel array or amorphous silicon technologies, for example. These detectors offer improvement in the accuracy and speed of data collection, thus increasing the efficiency of use of bright synchrotron beamlines for biological crystallography. They offer the potential of improvements in dynamic range, data throughput, and point-spread function, and the capability to perform time-resolved experiments at the level of a few milliseconds. The Subcommittee felt strongly that research in this area must continue and be fostered to produce the necessary detectors to improve the use of current beamlines and create the ability to discover new advances in the field.

      Beamline automation

      The development and implementation of a robotics workstation for high-throughput data collection of protein crystals is a very high priority in order to maximize use of valuable synchrotron radiation beamtime. Several laboratories are developing robotics workstations for protein purification, crystallization, and crystal analysis. The logical advancement is to develop automated crystal mounting and alignment systems for x-ray diffraction data collection. This is particularly important for microcrystal work where the crystals are often extremely difficult to observe and align. One automation system currently under development is using technology developed for the Human Genome Project. The goal is to reduce the amount of time necessary for crystal mounting and data collection using thoroughly automated robotics workstations. Traditionally, these steps take a significant amount of synchrotron beamtime (approximately 10-15 minutes) to go in and out of the beam "hutch" and to align the samples. As data collection times decrease to less than an hour in several cases, this time loss becomes significant. Ideas for future development include use of optical methods to rapidly optimize crystal centering and development of new hutch designs to allow for simultaneous alignment/mounting of crystals in parallel with data collection. In addition, small synchrotron beams allow selection of different parts of the crystal for optimizing diffraction and managing crystal decay, both areas that would benefit from automation.

      This technology will be invaluable to the efforts in high-throughput structure analysis that will be absolutely required in structural genomics and very high-throughput structure-based drug design.

      Methodology development

      Improved methods to optimize data collection strategy (including high resolution, poorly diffracting crystals, very large unit cells, etc.), data reduction, structure phasing/solution and refinement (especially at very high resolution) are important components that need further development. The current and new generation detectors are requiring new data processing software. Seamless interface between data acquisition, reduction, and analysis utilizing local and remote high performance computational capabilities will become increasingly important to enhance throughput and effective use of beamtime. Interactions of the beamline groups with researchers in their home institutions can significantly improve productivity and new tools utilizing network-based collaboratories offer new approaches, especially as robotics and automation become more common.

      VI. New Beamlines

      The Subcommittee considered the question of whether additional beamlines for macromolecular crystallography are needed at this time.

      New beamlines for structural biology develop through differing mechanisms: they may be initiated by the facility in response to perceived demand or recognized opportunities, or they may be initiated by a consortium of users (Participating Research Team [PRT] or Collaborative Access Team [CAT]) to meet the particular needs of that group. The route taken depends largely on the character of the individual storage-ring facility (i.e., mostly facility-administered beamlines at CHESS and SSRL; mostly PRT/CAT-based beamlines at NSLS and APS). New beamlines have typically exploited completely new real estate. However, as the facilities mature, new beamlines increasingly will arise through the redeployment and refurbishment of existing beamlines built for another purpose. Each of the US synchrotron facilities has opportunities for expansion or redeployment, and several initiatives relevant to structural biology and to macromolecular crystallography, in particular, have advanced to various stages of planning and to proposals for funding.

      A new beamline initiative may address needs at various levels. It may focus on a particular technology (e.g., MAD phasing); it may relate to high-throughput crystallography (as for structural genomics); it may address state-of-the-art diffraction needs for a specific consortium of users; or it may be designed to provide expanded capacity for the community at large. Several new beamline initiatives in structural biology have been proposed, and are seeking funding. These include Structural Genomics - CAT, Southeast - CAT, and Northeast - CAT at APS; portions of wiggler stations on beamline G at CHESS; and X6B and portions of an in-vacuum undulator (X13/X9) at NSLS. In addition, the opportunity for superconducting bend magnets has been examined and approved for crystallography at ALS; 3 are proposed with 3-4 MAD stations. There are also possibilities for additional insertion devices at SSRL.

      We encourage a case-by-case evaluation of beamline proposals by the funding agencies. We advise that priority be given to innovative applications. Since there is a relatively long lead-time from conception to operation for beamlines (three years or more), it is important to anticipate future needs now even though several beamlines for structural biology are just now becoming operational and plans are being made to upgrade others.

      Members of the Structural Biology Subcommittee:

      Dr. Jonathan Greer, Chair
      Department of Structural Biology
      Abbott Laboratories
      Abbott Park, Illinois

      Dr. Stephen Burley
      The Rockefeller University
      New York, New York

      Dr. Wayne Hendrickson
      Biochemistry and Molecular Biophysics
      Columbia University
      New York, New York

      Dr. Keith Hodgson, Chair BERAC
      Professor, Department of Chemistry
      Stanford University
      Stanford, California

      Dr. John E. Johnson
      Scripps Research Institute
      La Jolla, California

      Dr. J. Keith Moffat (not able to be present at the meeting)
      University of Chicago
      Cummings Life Science Center
      Chicago, Illinois

      Dr. Douglas Rees
      Division of Chemistry
      California Institute of Technology
      Pasadena, California

      Dr. Janet Smith
      Department of Biological Sciences
      Purdue University
      West Lafayette, Indiana

      Dr. Ray Stevens
      Department of Chemistry
      University of California
      Berkeley, California

      Also Attending

      :

      Dr. Marvin Cassman
      Director, National Institute of General Medical Sciences
      National Insitutes of Health
      Bethesda, Maryland

      Dr. Roland Hirsch
      Office of Biological and Environmental Research
      U.S. Department of Energy
      Germantown, Maryland