Chemical selection by the Interagency Testing Committee: use of computerized substructure searching to identify chemical groups for health effects, chemical fate and ecological effects testing

The Science of the Total Environment, 109/110 (1991) 691-700 Elsevier Science Publishers B.V., Amsterdam

691

Chemical selection by the Interagency Testing Committee: use of computerized substructure searching to identify chemical groups for health effects, chemical fate and ecological effects testing John D. Walker Interagency Testing Committee (TS-792), 401 M Street, SW, Washington, DC 20460, USA

ABSTRACT Since 1977, the Interagency Testing Committee has convened over 300 meetings, issued 27 semi-annual reports to the Administrator of the US Environmental Protection Agency, screened about 26 000 chemicals, reviewed about 4500 chemicals and recommended between 5000 and 7600 health effects, chemical fate or ecological effects tests for 114 chemicals and 27 chemical groups. The Committee has used three chemical selection processes to screen and identify chemicals for priority testing consideration. From 1977 to 1980, the Committee's processes consisted of examining large lists of chemicals and designating chemical categories that satisfied generic definitions. From 1980 to 1989, the Committee used sequential exposure and biological scoring processes followed by in-depth review. Since 1989, the Committee has used computerized processes to identify chemical groups that are associated with potentials to cause adverse health or ecological effects or that are likely to involve occupational or environmental exposure and that have common substructures, uses, testing information deficiencies, risk assessment uncertainties, etc. This paper focuses on use of the computerized processes to identify chemical groups containing common substructures that are associated with potentials to cause adverse ecological effects. The purpose of this paper is to describe how: (i) chemical substructures are selected, (ii) substructures are used to identify chemical groups, (iii) groups are processed, and (iv) reliability of the processes that were used to select chemical substructures are being assessed.

INTRODUCTION

The Interagency Testing C o m m i t t e e

The Interagency Testing Committee (ITC) was created by the US Congress in 1976 to screen, select and recommend chemicals and chemical groups for priority testing consideration, to facilitate coordination of chemical testing sponsored or required by US Government organizations and to enhance information exchange to promote cost-effective use of US Government chemical testing resources. The ITC's statutory responsibilities are described 0048-9697/91/$03.50 © 1991 Elsevier Science Publishers B.V. All rights reserved

692

J.D.WALKER

in Section 4(e) of the Toxic Substances Control Act (TSCA;" Public Law 94-469, 90 Stat. 2003 et seq., 15 U.S.C. 2601 et seq.). The Committee prepares the Priority Testing List of chemicals or chemical groups recommended for testing (by the chemical's manufacturers), transmits the List to the Administrator of the US Environmental Protection Agency (EPA) and determines the order in which the EPA Administrator should implement the testing recommendations under TSCA Section 4(a). Congress directed the Committee to revise the Priority Testing List at least every 6 months and required the EPA Adminstrator to publish the Committee's Reports in the FEDERAL REGISTER. Since 1977, the Committee has screened about 26000 chemicals, reviewed about 4500 chemicals and recommended between 5000 and 7600 tests for 114 chemicals and 27 chemical groups. Recommended tests are provided as a range of the minimum and maximum number of tests because a test could be counted one or more times depending on, for example, the number of species tested. Chemical selection

Numerous approaches have been used to score, screen or select chemicals for testing. Recently, a number of these approaches have utilized qualitative or quantitative structure activity relationships or SAR (Hart, 1988; Klein et al., 1988; Jonsson et al., 1989; Tosato et al., 1990). The ITC's approaches are unique because they are driven by Congressionally-mandated directives that must focus on the statutory requirements of TSCA and withstand the criticism of public scrutiny. ITC's chemical selection processes From the ITC's first meeting in February 1977 to November 1980, the chemical selection process consisted primarily of examining large lists of chemicals and selecting chemical groups that satisfied generic definitions. Using this process the Committee designated between 3000 and 5500 tests for about 24 chemicals and 17 chemical groups. The chemical selection process used from December 1980 to May 1989 emphasized sequential exposure and biological scoring followed by in depth review (Walker and Brink, 1989). Using this process the Committee recommended or designated between 600 and 900 tests for 74 chemicals and three chemical groups. Since June 1989, the ITC has used chemical selection procsses that were developed in 1986 (Walker and Brink, 1989). These processes involve costeffective, computerized approaches to identify chemical groups that are likely to cause adverse health or ecological effects or that are likely to involve occupational or environmental exposure and that have c o m m o n substruc-

693

CHEMICAL SELECTION BY THE INTER-AGENCY TESTING COMMITTEE

Ecological Effects Panel

Health Effects Panel

Substructure Questionnaire Developed

Substructures Likely to be Associated with Chemicals Causing Adverse

~ v

Ec°l°gicalE f f e c t ~ / " ~

Chemical Groups With Potential for Adverse Ecological Effects

Substructure Questionnaire Developed

[ -. \ , ] ~ k Gnemlcals / " J ~.~ ~./ ~'/ ~,,,,,,~L

Chemical Groups With Potential for Adverse Ecological and Health Effects

Substructures Likelyto be Associated with Chemicals Causing Adverse

HealthEffects

Chemical Groups With Potential for Adverse HealthEffects

Reliability Assessment and Application o! SARto Structurally-Related Groups

Fig. 1. lnteragency Testing Committee's processes for using substructures (likely to be associated w i t h chemicals causing adverse health or ecological effects) to identify chemical groups with potential to cause adverse health or ecological effects.

tures, uses, testing information deficiencies, risk assessment uncertainties, etc. Using these processes, the Committee has recommended or designated between 1100 and 1200 tests for 16 chemicals and seven chemical groups. SELECTING CHEMICAL SUBSTRUCTURES The process for selecting chemical substructures is outlined in Fig. 1. Organization of the ecological effects panel, development of the ecological effects substructure questionnaire and the use of expert opinion to select substructures likely to be associated with chemicals causing adverse ecological effects were previously described by Walker and Brink (1989). Over 120 internationally recognized experts provided opinions related to a substructure's potential to cause adverse ecological or health effects. Eighteen substructures likely to be associated with chemicals causing adverse health and ecological effects are listed in Table 1. Substructures currently being used by the ITC to identify potentially-hazardous chemical groups are listed in Table 2.

/

H

s \R, °

N ~- C = S

N= C = O

R m~0

R --

lsothiocyanates

Thioxiranes

R --

R , x ~ O

II

O

R--C-O-C~R'

II

O

R-O-P-O--R'

O II

Substructure

Isocyanates

~,fl-Unsaturated aldehydes

Anhydrides, carboxylic

Organophosphates

Group

Substructure-associated potential adverse health or ecological effects

TABLE 1

Oncogenicity mutagenicity reproductive effects teratogenicity membrane irritation

Membrane irritation acute toxicity

Membrane irritation acute toxicity

Oncogenicity mutagenicity membrane irritation

Membrane irritation

Neurotoxicity

Potential health effects

Mammals

Mammals, soil invertebrates

Mammals, microbes

Birds, fish, mammals

Mammals

Aquatic and soil invertebrates, birds, fish, mammals

Organisms potentially susceptible to adverse ecological effects

p-

C

Azo aromatics, heteroaromatics

I

N--N

\ \

I R'"

R ))

NH 2

Az- N-- N - Ar'

R'

R

Ar

Aromatic amines

Hydrazines

Ar

Nitro aromatics NO 2

X - - Cl[ 2 - O - - R

f

C

N

/\

Haloalkyl ethers

Quinones (o, p)

Aziridines

Oncogenicity mutagenicity

Oncogenicity mutagenicity membrane irritation acute toxicity

Oncogenicity mutagenicity chronic toxicity

Oncogenicity mutagenicity chronic toxicity

Oncogenicity mutagenicity membrane irritation

Oncogenicity mutagenicity membrane irritation

Oncogenicity mutagenicity reproductive effects teratogenicity membrane irritation

(continued)

Birds, algae, aquatic invertebrates, fish, mammals, microbes

Birds, fish mammals

Algae, aquatic invertebrates, birds, fish, mammals, microbes

Algae, fish, mammals, microbes

Fish

Algae

Algae, aquatic and soil invertebrates, birds, fish, mammals

NC/

R'

/N

R

R~C

O

OH

°

-R'

C~---N

~kH

//

0 U R~N~C--$

R, R', R" and R" = H, alkyl, or aryl. Ar = aromatic.

Cyanohydrins

Aliphatic aldehydes

Thiocarbamates

R--N--C--O--R II 0

o" I At-- 6"+=N-- At'

Azoxy aromatics, heteroaromatics

Carbamates

Substructure

Group

TABLE 1 (continued)

Fish, mammals, microbes

Oncogenicity neurotoxicity

Acute toxicity

Mammals

Birds, fish, mammals

Aquatic and soil invertebrates

Oncogenicity mutagenicity neurotoxicity

Oncogenicity mutagenicity reproductive effects teratogenicity membrane irritation

Mammals, microbes, soil invertebrates

Organisms potentially susceptible to adverse ecological effects

Oncogenicity mutagenicity

Potential health effects

C H E M I C A L S E L E C T I O N BY T H E I N T E R - A G E N C Y T E S T I N G C O M M I T T E E

697

TABLE 2 Substrates current used by the ITC to identify chemical groups Acetals Acetonitriles Acid halides Acrylamides Acrylates Acrylonitriles Aldehydes Aliphatic amines Alkenes Alkoxysilanes Alkylbenzenes Aldyldiaromatics Alkylnitros Alkylphenols Alkyltriaromatics Alkynes Allylic compounds Allylics Alpha-hydroxamines Aminoanthraquinones Anhydrides Anilides Anthracenes Anthraquinones Aromatic amines Aromatic nitriles Aromatic phosphorus Aromatic sulfhydryls Aromatic sulfonates Aryl ureas Aziridenes Azo compounds Azoxy compounds Benzidines Benzothiazoles Benzotrihalides Benzyl chlorides Beta-diketones Beta-haloethyl amines Biphenyls Bisphenols Butylhydroxy toluenes Carbamates Carbazoles Carboimides Chlorinated benzoic acids Chlorocycloparaffins Chromenes Coumarins

Cyanohalides Cyanohydrins Cycloparaffins Dialkylamines Dianhydrides Dianisidines Dioxanes, 1,4Diphenylamines Diphenylethers Dithiocarbomates Epoxides Ethanolamines Ethoxylates Ethylhexyls Fluorenes Glycidyl amines Glycidyls Glycol ethers Haloalkyl ethers Haloalkyl sulfides Halobenzenes Halodiaromatics Haloformic ethers Halophenols Halotriaromatics Heterocyclics Hexyls Hindered amines Hindered phenols Hydrazines Hydroxamates lmidazoles Indoles Isocyanates Isocyanides Isophthalic acids Isothiocyanates Ketenes Ketoximes Lactones Methylene dioxybenzenes Methylenediamines Mustards Naphthylamines Nitric acid esters Nitriles Nitroaliphatics Nitroaromatics Nitrobenzenes

Nitrofuryls, 5Nitrohalides Nitroquinolenes Nitroso compounds, CNitroso compounds, NOrganophosphates Perftuoros Peroxides Phenanthridenes Phenols Phosphines Phosphites Phosphoniums Phosphoramidates Phosphoric esters Phosphorodithioates Phthalates Phthalimides Polyamines Propargylic esters Pyrazines Pyrethroids Pyridazinones Pyridines Pyrrolidines Quaternary amines Quinolenes Quinones Siloxanes Styrenes Sulfanilamides Sulfonates Sulfonyl halides Sulfuric acid esters Terephthalates Thiocarbamates Thiocyanates Thiol esters Thiol ethers Thiophosphates Thioxiranes Thiram disulfides Tolidines Triazines Triazoles Trimellitates Uracils Urethanes

698

J.o. WALKER

USING SUBSTRUCTURES TO SELECT CHEMICAL GROUPS

As illustrated in Fig. 1, substructures likely to be associated with chemicals causing adverse health or ecological effects are currently used to probe a universe of more than 36000 discrete organic chemicals in the TSCA inventory and to identify chemical groups containing these substructures. The Committee uses two computerized approaches to identify groups of discrete organic chemicals that contain a common substructure. The first approach involves using a combination of letters and numbers to code chemicals (in the universe of more than 36 000 chemicals) for a substructure and an adverse effect likely to be associated with a chemical containing that substructure; e.g., all isocyanates (coded with number 24) would contain the codes HA24, HM24, EM24, ET24 to indicate potential to cause adverse health (H) effects involving acute (A) toxicity and membrane (M) irritation and potential to cause adverse ecological (E) effects to mammals (M) and microorganisms (T) (see Table 1). This approach was described in detail by Walker and Brink (1989). It is being supplemented with an effort to develop computer files of individual structures for each discrete chemical in a chemical group. This approach should provide sufficient flexibility to modify substructures used to search chemical groups and to identify subgroups of chemicals; e.g., this approach was used for a group of almost 1000 organophosphates to identify alkyl-, amino-, aryl-, glyco- or thio-phosphates. This approach should also provide sufficient flexibility to develop subgroups for structure activity analysis; e.g., to analyze the straight-chain versus branched alkylphosphates that were recommended for testing (ITC, 1990). PROCESSING SUBSTRUCTURE-BASED CHEMICAL GROUPS

Before substructure-based chemical groups are recommended for minimum (screening) testing information deficiencies, chemicals are removed that are unlikely to satisfy TSCA statutory requirements or unlikely to have been recently produced in or imported into the United States. Chemicals unlikely to satisfy TSCA statutory requirements were described by Walker and Brink (1989). Chemicals that are removed from a chemical group are publicly deferred in the Federal Register to promote a public understanding of the number of chemicals that the Committee processes. Deferred chemicals are recycles through the computerized process to detect changes in production, importation and exposure trends etc. The Committee recently examined four chemical groups, aldehydes, brominated flame retardants, isocyanates and sulfones. Four hundred and twenty-eight aldehydes, 95 brominated flame retardants, 80 isocyanates and 83 sulfones were deferred (ITC, 1991). The remaining 89 aldehydes, 33 brominated flame retardants, 43 isocyanates and 26 sulfones were recommended for testing (ITC, 1989, 1990, 1991). It is

C H E M I C A L S E L E C T I O N BY T H E I N T E R - A G E N C Y T E S T I N G C O M M I T T E E

699

important to note that the substructure-based processes have been used to identify chemicals in groups that may be of regulatory concern to ITC Member Agencies or external organizations. For example, aldehydes and isocyanates are of concern to the US Department of Transportation, the National Institute for Occupational Safety and Health and EPA, respectively, because there are insufficient data to reasonably determine or predict toxicity to algae or aquatic invertebrates and physical-chemical properties, respectively. After a chemical group is recommended for testing, EPA publishes final rules that automatically require industry to submit TSCA Section 8 information in 3 months. Production, importation and exposure data that are submitted under TSCA Section 8(a) and any non-public chemical fate, health effects or ecological effects studies that are submitted under TSCA Section 8(d) are reviewed. Chemicals with 8(a) data that do not satisfy production, importation or exposure thresholds are publicly deferred from further consideration and shared with others (e.g., the National Toxicology Program and the Organization for Economic Cooperation and Development) for optional testing [because requiring tests under TSCA Section 4(a) would be difficult]. These chemicals are shared with others for optional testing, if testing them as part of a structure activity group [for which other tests are likely to be conducted under TSCA Section 4(a)] could improve the ability of the resulting structure-activity model to predict the toxicity or persistence of chemicals with similar structures. The Committee reviews and summarizes all 8(d) submissions, identifies those submissions that appear to match testing recommendations and assesses the adequacy of the data reported in those submissions. Testing recommendations for which 8(d) submissions provide adequate data are publicly withdrawn. Chemicals remaining in the group are reviewed by ITC Member Agencies and any testing information deficiencies related to Member Agency data needs are identified. These remaining chemicals may also be partitioned into subgroups for structure activity analysis. Chemicals remaining in the group or subgroups of chemicals for which Member Agency data needs have been identified or for which structure-activity relationships can be developed may be designated for testing or tested voluntarily with the knowledge that extensive efforts have been made to analyze and eliminate public and non-public studies from ITC Member Agencies and manufacturers, processors and distributors. The Committee's processes for obtaining recent exposure data and other non-public information appear to satisfy one of the Congressional intentions for creating the ITC, viz. to facilitate coordination chemical testing that would conserve chemical testing resources. Industry supports the need to coordinate chemical testing through the ITC as well as the need to review non-public studies submitted under TSCA Section 8 to avoid duplicative and unnecessary testing.

700

J.D. WALKER

ASSESSING RELIABILITY OF SUBSTRUCTURES TO IDENTIFY POTENTIALLY HAZARDOUS CHEMICAL GROUPS

Preliminary reliability assessments of substructures to identify potentially hazardous chemical groups involves retrieving and analyzing toxicity data on discrete chemicals within the chemical groups. Readily-available data are retrieved and tabulated. These data are supplemented with data extracted from studies submitted under TSCA 8(d) and the total data base is used to assess the reliability of substructures to identify potentially hazardous chemical groups, to identify substituents on the substructures (for future structure-activity analysis) that appear to increase or decrease toxicity and to designate potential SAR-based subgroups for testing. When an SAR-based subgroup is designated for testing, the group of deferred chemicals is reexamined to identify chemicals with similar structures that might be tested by others to strengthen the SAR. Acute LCs0 values for aquatic organisms that were retrieved for over 30 aldehydes appear to support the expert's opinions that chemicals containing aldehyde substructures could cause adverse effects to fish (Table 1). These data also suggest that alpha-beta unsaturated aldehydes may be toxic to algae and aquatic intebrates and support the ITC's recommendation for algal and aquatic invertebrate testing (ITC, 1991). REFERENCES Hart, J.W., 1988. Some Danish aproaches to selection of existing chemicals. Chemosphere, 17: 1411-1418. ITC, 1989. Twenty-Fifth Report of the Interagency Testing Committee (November 1, 1989) to the EPA Administrator; Receipt of Report and Request for Comments Regarding Priority List of Chemicals. Fed. Reg. 54:51114-51130. ITC, 1990. Twenty-Sixth Report of the Interagency Testing Committee (May 8, 1990) to the EPA Administrator; Receipt of Report and Request for Comments Regarding Priority List of Chemicals. Fed. Reg., 55: 23050-23062. ITC, 1991. Twenty-Seventh Reort of the Intragency Testing Committee (November 19, 1990) to the EPA Administrator; Receipt of Report and Request for Comments Regarding Priority List of Chemicals. Fed. Reg., 56: 9534-9572. Jonsson, J., L. Eriksson, M. Sjostrom, S. Wold and M.L. Tosato, 1989. A strategy for ranking environmentally occurring chemicals. Chem. Intell. Lab. Syst., 5: 169-186. Klein, A.W., W. Klein, W. Kordel and M. Weiss, 1988. Structure-activity relationships for selecting and setting priorities for existing chemicals - a computer-assisted approach. Environ. Toxicol. Che., 7: 455-467. Tosato, M.L., S. Marchini, L. Passerini, A. Pino, L. Eriksson, F. Lindgren, S. Hellberg, J. Jonsson, M. Sjostrom, B. Skagerberg and S. Wold, 1990. QSARs based on statistical design and their use for identifying chemicals for further biological testing. Environ. Toxicol. Chem., 9: 265-277. Walker, J.D. and R.H. Brink, 1989. New cost-effective, computerized approaches to selecting chemicals for priority testing consideration. In: G.W. Suter II and M.A. Lewis (Eds), Aquatic Toxicology and Environmental Fate: Vol. 11, Materials, Philadephia, PA, pp. 507-536.