National Academies Press: OpenBook

First Symposium on Chemical-Biological Correlation, May 26-27, 1950 (1951)

Chapter: Influence of Isosteric Replacements Upon Biological Activity

« Previous: The Effect of Drugs on Physiologically Active Thiol Systems
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 295
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 296
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 297
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 298
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 299
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 300
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 301
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 302
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 303
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 304
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 305
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 306
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 307
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 308
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 309
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 310
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 311
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 312
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 313
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 314
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 315
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 316
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 317
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 318
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 319
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 320
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 321
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 322
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 323
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 324
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 325
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 326
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 327
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 328
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 329
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 330
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 331
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 332
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 333
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 334
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 335
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 336
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 337
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 338
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 339
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 340
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 341
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 342
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 343
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 344
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 345
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 346
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 347
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 348
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 349
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 350
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 351
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 352
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 353
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 354
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 355
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 356
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 357
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 358
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 359
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 360
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 361
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 362
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 363
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 364
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 365
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 366
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 367
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 368
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 369
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 370
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 371
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 372
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 373
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 374
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 375
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 376
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 377
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 378
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 379
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 380
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 381
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 382
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 383
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 384
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 385
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 386
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 387
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 388
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 389
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 390
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 391
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 392
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 393
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 394
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 395
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 396
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 397
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 398
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 399
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 400
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 401
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 402
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 403
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 404
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 405
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 406
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 407
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 408
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 409
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 410
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 411
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 412
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 413
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 414
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 415
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 416
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 417
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 418
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 419
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 420
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 421
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 422
Suggested Citation:"Influence of Isosteric Replacements Upon Biological Activity." National Research Council. 1951. First Symposium on Chemical-Biological Correlation, May 26-27, 1950. Washington, DC: The National Academies Press. doi: 10.17226/18474.
×
Page 423

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

INFLUENCE OF ISOSTERIC REPLACEMENTS UPON BIOLOGICAL ACTIVITY Harris L. Friedman Lakeside Laboratories, Inc. Milwaukee 1, Wisconsin

296 PART 1 INTRODUCTION To the synthetic organic chemist interested in medicinal chemistry, every physiologically active compound of known structure is a challenge - a challenge either to better it, or perhaps merely to equal it. For it must be remembered that even the most innocuous drug is not toler- ated by some people. There are numerous ways of attackmg such a problem, and this audience is certainly familiar with them. One of the methods which has been used frequently, very often with success, is that of isosteric replacement. The examples of this type of replacement in the literature are very numerous, and the fruitful results in the fields of sulfonamides, antimetabolites, and anti- histamines are well known. The concept of isosterism, first introduced by Langmuir, 13 has been changed over the years by the work of many others. It will be the object of this paper to survey the history of isosterism, to classify the varieties of isosteric replacements which are recorded in the litera- ture, and to note the influence of these replacements on the biological activity of compounds. We shall then be able to see if any general relationships apply, and what conclusions may be drawn from such data. PART 2 THEORETICAL Langmuir in 19191.* pointed out the remarkably similar physical properties of carbon dioxide and nitrous oxide. He deduced from the octet theory that the number and arrangement of electrons in these molecules are the same. Compounds showing such relationship to one another were termed isosteric compounds or isosteres. These terms were not restricted to compounds, but were extended to groups of atoms which hold pairs of electrons in common (termed by Langmuir comolecules). Comolecules were likewise considered isosteric if they contained the same number of atoms and possessed the same number and arrangement of electrons. Langmuir predicted twenty-one types of isosteric groups of which only a few will be mentioned here:

297 CHART 1 Langmuir Type 2 O=, F*, Ne, Na+, Mg++ , A1++ + 3 c= Ci * A K+ Ca+"*" 5 Br', Kr, Rb+, Sr+ + 8 N2, CO, CM' 9 CH4, NH4+ 10 CO2, N2O, N3*, CNO" I angmuir postulated that when isosteric comolecules are also isoelectric, that is when they have the same total charge, all their physical properties should be closely similar. Only three such pairs occurred in Langmuir's tables: N2 and CO; CO2 and N2O; N3* and NCO' However, no direct comparison can be made of the physical properties of isosteres having different electrical charges. Even though the classes are distinct from one another, it was demonstrated that compar- isons of comolecules in different groups could still be made. If any two substances are very much alike in physical properties, then any isoelectric isosteres of these substances should show close relationships with one another. Thus in types 3 and 8, argon and nitrogen resemble each other closely, therefore chloride ion (isosteric with argon) should resemble the cyanide ion {isosteric with nitrogen). Likewise the similarity between K+ and NH4+ can be derived from argon and methane. It is to be emphasized that in Langmuir's terminology K+ and NH4+ are not isosteres of one another. Whereas Langmuir compared only physical properties, Seifriz^ showed in 1948 that CO2 and N2O are both reversibly anesthetic to the slime mold Physarum polycephalum. In 1921 , W. Huckel" pointed out that the imino group (=NH) in homopolar compounds corresponds to the oxygen atom and that the -NH2 and -OH groups correspond to the F atom. Somewhat later (1 925) GrimmS very markedly extended the concept of isosterism. In place of Langmuir's term "isosteric comolecule", Grimm preferred the term pseudoatom. His hydride displacement law states: "Atoms anywhere up to four places in the periodic system before an inert gas change their properties by uniting with 1 to 4 hydrogen atoms, in such manner that the resulting combinations behave like pseudoatoms, which are similar to elements in the groups 1 to 4 places, respectively, to their right. " Grimm showed this relationship by the following chart:

298 CHART 2 Group 4 5 6 7 0 1 Number of 0 C N O F Ne Na Hydrogens 1 CH NH OH FH 2 CH2 NH2 OH2 FH2+ 3 CH3 NH3 OH3 + 4 CH4 NH4 + Beginning in 1932, Professor Erlenmeyer at the University of Basel in Switzerland has published a series of papers on isosteric compounds. He has given great impetus to the modern concept of isosterism in organic chemistry, particularly in relation to biological activity. Erlenmeyer accepted Grimm's classification and has broadened it even further. His definition of isostere is: "Atoms, ions, or molecules in which the peripheral layers of electrons can be considered to be identical are termed isosteres. " By Erlenmeyer's definition all elements in the same group of the periodic table are isosteric so long as they have the same number of electrons in their outermost shell. In a unique application of this concept, Erlenmeyer in 1933° considered S and CH = CH in an aromatic nucleus to be isosteric by counting only the "peripheral" electrons in the C=C pseudoatom (whether it be written -CH=CH- or=CH-CH=). In 1946, Mentzer^ demonstrated that, in certain circumstances, the group -CH2-CH2- could be replaced by -CO-O- with no change in biological activity of the parent compound. He did not term these isosteric pairs. Some authors, as Mentzer^ and Erlenmeyer3 use the term potential-cycle or pseudocycle to bring out the steric relationships between the ring and opened form of physiologically active molecules. Occasionally the term "isolog" is used by some authors (as Fieser? in the United States and Steinkopfl® in Europe) where "isostere" is usually designated. Isologous compounds, however, need not be isosteric. It is obvious from this brief survey that the term "isosteric" has varied in meaning with different writers - from a narrow to a very broad concept. In this discussion we shall accept the term in its broadest meaning and study the influence of isosteric replacements on the biological activity of molecules. We shall not consider physical properties, although it is not implied that such properties as mixed-crystal formation are not of great significance for isosteric relation- ships. We shall term compounds "bio-isosteric" if they fit the broadest definition for isosteres and have the same type of biological activity. The biological equivalence of isosteric groups receives support from immunological studies. l andsteiner'Z was able to prepare artificial antigen-antibody systems by coupling diazotized aromatic amines with proteins, and injecting these protein complexes into animals to form antibodies. He discovered that these antibodies have the specific power of combining with the group attached by the azo linkage. This group, which is of known structure, he termed the haptenic group. The specificity of the antibody in any serum could be tested, by means of the precipitin reaction, with related complex proteins. In general, the combining power is highly,

299 but not completely, specific. Erlenmeyer,4, 19 using Landsteiner's method, demonstrated the serological similarity of several isosteric atoms and pseudoatoms. The following are illustrations of several types of isosteric replacement, where the resulting compounds give cross-reactions: CHART 3 SEROLOGICAL SPECIFICITY C6H5-0-C6H4-N=N-Protein C6H5-NH-C6H4-Protein -Protein -C6H4-PO3H2 -C6H4-As03H2 not -C6H4-SbO3H2 -C6H4-S03H -C6H4-SeO3H not -C6H4-SO2H -CO-NH-C6H4- >-NH-C6H4- not -CO-NH-C6H4- H2N I and II ++ + I and III t II and III ++ III H2N CH, From Erlenmeyer4'

300 While this work is not necessarily foolproof, and indeed has been criticized by Heidelbergerl0 as requiring more and better support, it is in good agreement with Pauling's views. Paulingl5 has also studied serological reactions to provide mformation about the molecular structure and configuration of simple substances. He compared the process of anti- body formation to the production of a replica by pressing a plastic material against a mold and permitting it to harden. The polypeptide chain, with its power of assuming alternative configura- tions, is the plastic material, and the surface of the antigen serves as the die or moid. The complementariness of antigen and antibody includes not only surface configuration, but also juxta- position of special combining groups, such as a negatively charged group in the antibody with a positively charged group in the antigen, and a hydrogen-bond-forming group carrying the proton with a similar group presenting an electron pair. Thus, isosteric replacements in an antigen which do not affect the shape or polarity of the molecule should not interfere with its reaction with the antibody. Pauling has extended this concept of spatial surface configuration to include biological specificity in general. Deductions from serological reactions are limited because strongly polar groups have predominant effects. However, this type of study should show when parts of a molecule are simple "space fillers", that is occupy specified geometrical bulk. Since the discovery that the antagonism of the suHonamides to £-amino benzoic acid is an antimetabolite effect, due to the close similarity of structure, isosteric replacements in other essential nutrients have yielded many compounds of interest. This field has been thoroughly reviewed in the literature, '6 and is the subject of a separate part of the symposium. We shall merely mention some of the types of isosteric replacement in the numerous antimetabolites which have been made, and only a few specific examples will be given in this paper: CHART 4 Essential Nutrient Atom or Group Replaced Replacement Ri bo flavin' 2-CH3 2-C1 Thymine -CH3 -OH. -Br, or -NH2 Mesoinositol 6 -OH 6-C1 Thymine, lysine, folic acid -NH2 -OH Folic acid -OH -NH2 £ • Ammo benzoic acid, glutamic acid -COOH -CONH2 £-Amino benzoic acid, niacin -COOH -COCH3 Arginine -O- -CH2- Uracil, thymine, niacin amide -O- -S- Methionine -s- -O-, or -CH2- Purines -CH= -N= Phenyl alanine, £-aminobenzoic acid benzene pyridine

301 CHART 4 (Cont. ) Essential Nutrient Atom or Group Replaced Replacement Phenyl alanine benzene thiophene , furan, pyrrole Niacin pyridine thiazole Thiamine thiazole pyridine Methionine -S- -CH=CH- Valine, niacin, pantothenic acid, aspartic acid, oxybiotin, heteroauxin -COOH -S03H p-Amino benzoic acid -COOH -AsO3H2 D-Amino benzoic acid -COOH -P03H2 •- The isosteric compound formed may have either the same activity as the original, or more usually it may have an antagonistic effect. In either case, it is proof that isosteric replacement gives compounds acting by the same mechanism, that they are truly bio-isosteric. Ideally, to make comparisons between structure and biological activity, two criteria are necessary: (1) Substances compared must act by the same mechanism and (2) The structure involved in the test should be the structure of the compound under study. However, in practice, for many types of biological activity only in vivo tests can be used, and even when using in vitro tests, we cannot be sure that the above criteria apply. In so far as possible, examples have been chosen which are based on in vitro activity, and mainly those using an isolated tissue or micro- organism. It is not claimed that the in vitro tests will necessarily correlate with in vivo or clinical studies; nonetheless the data obtained may be a useful guide for further work and may be adaptable to other series of compounds. Activities found in one screening test need not parallel the relative activities of the compounds in another test. Since data in the literature are usually lacking for tests other than those in which the authors were most interested, it is seldom possible to make such alternate lists. Biological activities, as absorption, distribution, conjugation (detoxification), taste, odor, side effects of drugs, will not be discussed.

302 PART 3 TABLES OF DATA In order to classify the data for presentation, the following chart is pertinent for organic compounds. CHART 5 Class 4 Class 3 Class 2 Class 1 Sb Te I As Se Br P S Cl C N 0 F Ne N+ P+ S + -CH- -NH- -PH- -OH -SH -CH2- -NH2 -PH2 -CH3 Table modified from Grimm** We have designated these four types from the number of covalent bonds. Above are the elements of the same periodic group, below are the isosteric hydrides. The following tables will show the effect of isosteric replacement on biological activity within each type. The -S- and -C=C-, and other special cases do not fit into this chart and will be treated separately. Discussion of Tables Class 1 - Halogens and Hydrides (OH, NH2, CH.Q Tables 1 -4 list examples of comparisons of Class 1 of the chart previously shown. Table 1 contains examples of multiple comparisons, Table 2 compares halogens only, Table 3 compares halogens with hydrides, Table 4 compares hydride with hydride. An attempt has been made to select examples from fields of current interest. Table 1. If one were to judge results of isosteric replacements from Table 1, it would be difficult to arouse enthusiasm. The most unusual case is the replacement of the chain -OH of epinephrine by NH2; the activities of other members of this series should be of great theoretical interest.

303 Table 2. In general the results are what would be expected, activities usually showing a gradient with the molecular weights. There are exceptions, mostly with the extreme members F and I. Tables 3 and 4. These again show unpredictability of response. The a-hydroxy-p-phenyl- ethylamine examples have been amply discussed by Hartung. 9 General Conclusions to Class 1. It is not possible to predict when members of this class will be bio-isosteric ; in most instances they will not be. Very often activity is specific to one member which would be called, in Ehrlich's terminology, an anchoring group. Differences in activities may be attributable to differences in polarity of the groups, to solubility differences, or to chemical reactivity. The most likely pairs of bio-isosteres are: halogen and CH3 halogen and OH, the most unlikely pairs are: OH and NH2. Class 2 - (S, O. NH. CH2) Tables 5-7 inclusive list examples of Class 2 replacements. For convenience, the examples have been divided into structural types: ether, ketone, and ester. Table 5: Ether Type. These show a better probability of bio-isosterism than Class 1 types. The member which fits in the least is -NH. Sulfur is not always bio-isosteric with oxygen, in fact surprisingly less than might be anticipated. Probably polarity differences play a predominant role. Table 6: Ketone Type. The most interesting examples are probably the thio-barbiturates. In general, these types have restricted comparability. Table 7: Ester Type. Many amides and thioesters related to the local anesthetic and antispasmodic esters are known, but practically none have come into use. More thought should be given to the'replacement of ester oxygen by the CH2 group. General Conclusions to Class 2. Isosteric replacement in this group has better promise of usefulness than in Class 1. While methoxy and ethyl often do not show similarity, in other cases interchanging O and CH2 yields compounds of similar activity. Here, as in Class 1, polarities probably play a dominant role. Class 3 - Tertiary N and Tertiary C Tables 8 and 9 list examples of Class 3. Most of the known examples of this class occur in the aromatic ring systems (Table 8). No attempt has been made to list the numerous examples in well-known fields as the sulfonamides and antihistamines but a few are given to refresh memories. Commercially this has been the most valuable application of bio-isosterism. Table 9 contains non-aromatic examples. Because of polarity differences the aliphatic types can seldom be expected to display bio-isosterism, but it is a more likely assumption that diphenylamine and benzhydryl derivatives would show such similarity. More examples are desi rable. While tertiary P, As and Sb theoretically are electronically isosteric with N and CH, practically, except between As and Sb, they show little bio-isosterism. Class 4 - Quaternary C, N, S. Etc. (Table 10) The spatial tetrahedral geometry and the positive charge are of paramount importance for this class. In general, a quaternary carbon, because it lacks a charge, is not interchangeable with quaternary nitrogen.

304 The Special Classes Table 1 1 - Aromatic C-C and O, S, NH. Following Erlenmeyer, it is generally agreed that the pairs, benzene and thiophene, and pyridine and thiazole, are isosteric. Furan and pyrrole differ markedly from benzene in both physical and chemical properties. Indeed, Bradlow, Vanderwerf and KleinbergZ in a brief discussion of the concept of isosterism, state that "Proponents of the principle of isosterism do not point out the fact that by definition pyrrole is also isosteric with benzene, thiophene and furan. " There are sufficient examples, however, to indicate that these rings may be bio-isosteric although the furan and pyrrole compounds are usually weaker in activity than those containing benzene and thiophene. The activity of furan isosteres in the antihistamine field indicates that such replacements cannot be ignored. The latter part of the table contains comparisons of an ethylenic bridge between two aromatic rings with S, O and NH. Not enough examples are available to draw conclusions, but this should be an interesting replacement type. Table 12 - Carbonyl and Sulfone (or Sulfoxide). The structural relationship between £-aminobenzoic acid and sulfanilamide has been emphasized by Bell and Roblinl in explaining the mode of action of sulfonamides. The exchange of -COOH for -SO3H in many metabolites to produce antagonistic substances has shown this to be a general phenomenon. Therefore, a comparison has been made in this table of compounds with carbonyl and sulfone groups. The sulfoxide has been added since spatially it more nearly resembles the carbonyl than does the sulfone grouping. Electronically neither the sulfoxide nor the sulfone group is isosteric with the carbonyl group. The ionic bond of the sulfur groups further emphasizes the difference. The table does not indicate any striking resemblances except for the amidone type example. Table 13 - -CO-O- and -CH2-CH;-. These groups are not electronically isosteric; it is most likely that they owe their activity to similar spatial fixation (as discussed in the following type). Table 14 - Spatial Fixation by Ring or Double Bond. That three dimensional spatial characteristics play a highly important role in biological activity is known to everyone. The vastly different activities often noted in optical or cis-trans isomers must be consi'dered in any attempt to explain the mechanism of biological activity. When two molecules are almost identical spatially, that is, are superimposable in three dimensions, we may expect similar activity provided the polarities are situated in corresponding parts of the molecules. Table 14 lists examples of ring and open forms. In many cases there is striking agree- ment; often however, one form is completely inactive. Where agreement between activities is found, it is interesting to assume that the open and closed forms can be superimposed in three dimensions. Benadryl is an interesting example. When the rings are forced into the planar form of the fluorene ring, activity is lost; one might speculate therefore that in the "active" form of Benadryl the two benzene rings do not lie in the same plane. From the example of Trasentin and Pavatrine the opposite appears to be the case. Table 15.- Polarity Shift, Exo-Endo Cyclic. This and the following table are offered in an attempt to systematize data scattered throughout the literature. This table demonstrates the effect of moving a polarity from without to within a ring, the shift being to the adjacent position. The probability is that the polarity of the atom must remain quite similar in order to retain the same activity. Table 16 - Reversed Adjacent Polarity. This table illustrates the effect of reversing adjacent polar groups. Many instances are of great interest; this is a transformation to be kept in mind in seeking new compounds.

305 CONCLUSIONS We have seen that similarity in biological action need not result from isosteric replace- ment - isosteres need not be bioisosteres. This is not surprising in view of the complexity of the simplest living systems. As we all know, simple isosteric replacements often give compounds of interest and value. In addition, there are two important types of information to be learned from such replacements. One is that we discover which groups cannot be eliminated in order to retain the desired activity (i. e. , the anchoring groups); the other is that we learn which parts of the molecule are important because of their bulk space characteristics. These facts enable a more intelligent approach to the synthesis of new compounds. The similarity of biological activity in so many instances, and the successful results already achieved through isosteric replacements, show that this is a type of variation which the synthetic chemist must keep in mind. If chemical reactivity and polarities are considered, the predictability of bio-isosteric replacement is quite high.

306 REFERENCES TO TEXT 1. Bell, P.H. and Roblin, R. O. , Jr.. J. Am. Chem. Soc. , 64, 2905(1942). 2. Bradlow, H. L. , Vanderwerf, C. A. and Kleinberg, J. , J. Chem. Ed., 24, 433(1947). 3. Erlenmeyer, H. , Bull. Soc. Chim. Biol. , 3J7, 792(1948). 4. Erlenmeyer, H. , Berger, E. and Leo, M. , Helv. Chim. Acta, l_b. 733(1933). 5. Erlenmeyer. H. and Leo, M. , Helv. Chim. Acta. 15, 1171 (1932). 6. Erlenmeyer, H. and Leo, M. , Helv. Chim. Acta, 16, 1381 (1933). 7. Fieser, L. F. and Hershberg, E. B. , J. Am. Chem. Soc.. 62, 1640 (1940). 8. Grimm, H. G. , Z. Electrochem. , 3J, 474 (1925); Naturwissenschaften, 17, 557(1929). 9. Hartung, W. H. , Ind. Eng. Chem., 37, 126 (1945); Chem. Rev., 9, 389(1931). 10. Heidelberger, M. , Ann. Rev. Biochemistry, 4, 569(1935). 11. Huckel, W. , Z. Electrochem. , 27, 305 (1921); C. A. , 1£, 514(1922). 12. Landsteiner, K. . "The Specificity of Serological Reactions", Harvard Univ. Press, 1945. 13. Langmuir, I., J. Am. Chem. Soc., 41, 1543(1919). 14. Mentzer, C. , Gley, P. , -Molho. D. and Billet. D. , Bull. Soc. Chim.. 1946, 271. 15. Pauling, L. , Endeavor, J, 43 (1948). 16. Roblin, R.O. , Jr., Chem. Rev., 38, 255 (1946); Chem. Eng. News, 27, 3624(1949). 17. Seifriz, W. , Science, 107, 15 (1948). 18. Steinkopf, W. , "Die Chemie des Thiophens", 1941. 19. Tomcsik, J. , Schwarzweiss, H. , Trissler, M. and Erlenmeyer, H. , Helv. Chijn. Acta, 32, 31 (1949).

307 EXPLANATION OF TABLES In the first column the structural formulas of the compounds under consideration are written with the variable isosteric group represented as X. In the second column the biological list is stated (e.g. anti- tubercular, narcotic, anesthetic, etc. ) together with other perti- nent data necessary - as to whether in vitro or in vivo, what organ, organism, or animal was used, what challenging agent, if any, was used, and in what terms the activity data are expressed with a designation of the reference compound if such was used. Below, in parenthesis, the references to the literature are given. This list is found immediately following the tables. The remaining column headings usually designate the atom or group represented by X. In cases where this is not so, the heading is self evident. The data in these columns are the activities of the compounds in.the terms used by the original workers; this varies with different authors from quantitative figures, a system of plusses, to mere statements of activity or non-activity.

308 fO r- | u Psl U i 3 V o Ps] V o X - o - z V i u o we i ^ OD V 3 1 u O X o 3 o V in O o V E — .o V t3 o o m M c t a V 00 « so r- o • n "a U< U ffl — — h u u 0 U u c 0O h a .2 -2 1 iU II 05 rt M ZT ^ • w ifl ' — . a "- 14 ^-* ft V *- iA V 3 5 - Z t 3 ~ "33? H u Z C. u « h y s § S 1 £ ki t1 *J &l > ? > « u .§ o 1 4 1 ! ! s .a _ u r i ^ ..- .-- ° r i " 0n 'Z £ '•S "> .s o 1 £ & 1 * '•5 > "o 5 £ 4 £ 5 a < £ £ .c § o a fM E o U § .. O Z Ki Ps] X u X X . x^ u o X z z X 0 z 0 U 1 k^ L> \ / \ / x\ /* z 'h: s PsJ •3r_ X

309 X 0 X O C 0 U U J n H V) in J U BiologiUal A CoeffiUieet Artyphosus A A I o a o U a U) Aeti vitro lar Coe ube ^-e M X O O U uberUu reg % (57) itr is Ae ^-n Sta

310 o o. X Z § I .o o o n o. o u o b >. 'S • c k S. 0, U - >.§ -• a o v a i. = a 0 w .o I

311 X u 1M 2 X o c 0 U W 03 H J U V Biolo o (M I u o u CM u Compo .2 1 a a .8 " 0 J EH D Q S H ° •? 5 •C! 7^ Q «" S Q U J H Q Q rt TJ C a o a M a V 00 0 h ,-* H 3 00 E ct) an ' u i

312 I U X o 05 1 l .o o "o £ .o c 3

313 u to h ! 1 u 1 u u u u u i —, o o '•§ '•§ V t1 u u B u u ^ 9 h I B 3 3 U U rt 4 .o 0 u u Jjl sj CL in it "o 13 " "« S 1 I 1 n U B a. B x C r^3 3 S 1 U V o f M u Pi u "u" u a B B 4l H H |ij 3 t1 IM u h « ,0 '£ 2 o y '"^ ,, H .2 _. 1 5 m r- .2 2. V 2 -o" M ,^, H « 0 B p. in ti ^"^ C V h U ^f ^3 o "-" u 3 ^ i4 u |* So nt SK a .0 "o Z X Z < > •o iM . c 3 O a x* u E o-u VI* o U X x* u z 0 o=u O 1 M u 1 fM sc T* x ,-* X X u X *7* U"X^' o K n X .£* *!. NX\r g u u u u u 2^/Z / \ u m m X X X /^^* /^*>* O u u •X; 'Xy «k 1S>. X

314 n § U w J CO H en en J U I « 0. o £ 1 o a o U x it >. 2 c V Is u a , u-V

315 X U z X o U ffl H 3 u £ • a - o o 0.1 Comprt I m U. 8 u • c « ^ 0 o J 00 .5 ~ U 2 X U fM X U 8 U .o i f! o S*E <fl -iJ •" fl 11 ™ r 2 • in nJ - i u iM U 8 /\ ' is. la.

316 I u £ x z § *& o U U H 13 o ffl m yl Ae O fM i? IS 1n X *l a oid Ae Artirt eseUtiUidal s. Ulotrtes mot .o c 3 O a O X u a>

317 X U X 2 I O r- "H n c 0 U U J n CO J U jtf U c < Q U .= 8 3 i B u § JS s "S 3 O , o U 0= an X U X <J X ° X X 0-2 Sa E ?

318 X u I 2 O 0 g| s:s i u u .o c 3 — l U o .5 o .5

319 a j n in w < 0 f » rt £ is i x * '£ a a* 0 X, + V X, - * " S ^1 *- ^ g U U **-" y "^ h t1 C 4 iD E K « 3 r. j 2 « S m ^ X Z iU l* «l ** 2 > « Jo. ^D « u *j (j ... ^3 .". + g — . •j < JD 0 .0 U V I J: 0 *• I a e reeUe) U E 0 V 13 S ^ r- E > . . £. o C it *.? w .o ** o .r" u V rt u in £ oo hi > iri^ s ^ o. E t - *° N *" u 3. ^ 0 - >s - w .*^- 2 *• .2 5 * -fc— U ^ u ri .^ > "« cm u SB 'is i <I .5 in < fl PI 3 o 0. > 10 < "o (Q TJ c 3 0 .r' CL U O E ifi* 6 u 0 *.^* X u o u r Z X* X U N V> rg . ^.' P^ 'i z OJ o o a Is X ui U u + Z o Z 1 m 3 fM V I co £ 2 X

320 U ! X u X 2 « \ ? sO V X O r- - H- u i H X l IJ u u c u h V | c o O O — .o 5» 00 in V ! 2 I *o o 1 u I- .o ^ o ^ O i, 0 f >. 22, iH 13 u H :s o * 1/1 0 J3 i 2! z c x 5 *J 1n < c J3 :, .5 la. a u < *- Q ^J "3 £ .o c s 0 a. E o U m - X JT X i x. V g 2=6 \ en / CO A .'^. O ^* '- r ' X X U— 2 z z c ^ J^ X O u — ;x . 2=U 0

321 = u rg I Z o § U w a H in o u .S o .g-s E>. Is u C - u N U rg U 6 O 'x' u ."' i fM X X u—u c .o o o § £ ^ 3 a.

322 X U CO u ffl J U o "5 s 3 O a o U I U S ... 0-X;

323 W H3 in in d O fsJ o o O V 0* X 0 -o u J a TJ a > O X N ?T B) o B i0 V h * a u O O o .^ 2 + h o O O i4 a O O c^ in JX e « £ 14 + i X at u t3 Q1 h V iM c 2 « •=" V a H 1 Aeeol CoeffiUi vs. Staprt. aure (14O) AetituberUular SaliUylate No. Aetihistamirt (17,97, 127 Blood pressure (13O) 5. •4-1 (16) .o ^^ w U iS V H o c "it u 14 a H H "3 x, A O Z u u X ' 1 1 r g rg .o c B X U U g Y rsJ i*J O X X U U a gr £ o u I 1 X Z Z O rg U O rS rg .w U u fM gl ^s. ^^ ^w ^0) K X 2=U ''*; cX' u / \ u Z fM IS Ta "Qi l.S, 13. X

324 B o U a H < J u u fM o X u u h S 0" ** V X t*i U .s z O + i u u 1 rg "o u S o N + h u s . iD u V V u u w m $ 1 2 U A u c u I O >. VI V i f 2 Si *- iA 3 x V . — - y ^ , . j: QO « V f- « ™ U i*i u 0 ~ « «}, = i« .^. m a — — i i ^ rM o 0l * i, ~ rt u O 7> 1 ? CO U o a M .O V 2 u O J "3 O Q, J n lSL fM s z Comprted O i*J 8 iM X U u u 1 fM rsJ X X § x u u-z X 6 rg u „, 0 U fM S u Q *"! X i X -i, X 2 = U -Jj ^ u •><. u H !*'' / \ I1 .'*'' co \==/ U +J X PS u

325 W 03 < J OS X! A >. H X u U -U-ia Z — U Z X O X y S o c £ 0. 1 T3 c < X u X u

.In z I 6 g- ° c >< o. ;i ^-\ u-z G C

327 X u X z W J ft M J U i< OS • o >. .2 o S 0 c5 Comprti o*u iM Z U X u U-Z U— fe u — z O X

32% U) y. 2 5 5 "J s X <. U b » « z s 2 ^ 1 C «*- Jl * VS c fa 5 S - u .5 c - fa £ r- | S js • ** • v. a 1 3 u 1 1 c v c < f .c < ffl £ *M MJ A 1 X X iM iM u u X X fM rM u u X X !M JL u u u u \/ X X 1 1 x1 x" 0 V v a 1 o X i z U u x1 u fM X X X U 0 u I — G*( \C /^^s\ ci X V .*" A O-u-< \ / \ / c o is Ha <

329 u J 03 U) 3 O fM £ X U U V U X 2 U 1 a c m O ~ 00 0 * o 1 ^ * O v I O u c V hi "0 E V c "1 o Il t1 ** « O rg C *"" — .2 2 -. Co" a^ u ~- I i 1- s a 1 S in . — . H U 1l *O ^ > u -$ S. ! I V ^2 V M 00 j? u V ^ i h *. K s 5 o X 5 < ^ E < S 3 .o § 0 rsJ a S 2 0 £ U t m X '"*' X 2 — U — 2 r^3 X X X 2 u-u rsJ * ? »" io 5? U U — u— 8 f 1 2 X X2 6— u— u 1 Z m X r-4 rs3 CC X U •n 2 H

330 N U I 2 J U X-U as v P. H X r g X X U-2-U l 1 fM fM x x U U \ / u

331 X u X z u J 03 lO en U « X 6 u a 9* H o ffl o E a a. £ .5 2 2 « ^* « tM s ^ u • N i u I3 u ;k. O u X /°\ J3 o 3 c u 'V O 00 A lS. IS.

332 TABLE 8 CLASS 3 Aromatic Rings Compound Biological Test (Reference) = CH- =N- . . Antibacterial moderate active (8, 120) O**e\ S02-NH^ J ?y(p=J Ditto moderate very active Q3 NCH2-CH2-NMe2 Ant ihis tam ine active active (148,78,80,59) Q'^ CH2^ ^,(OCH3) N \=/ CH2-CH2-NMe2 Ditto active active . NH-CH2-CH2-NEt2 Antimalarial inactive active Cl/'A-NH-^'^CH3 '-' CH3 (102) ,. Anti spasmodic L JL Jl J vs. acetylcholine 15 5 COO-CH2-CH2-NEt2 Atropine = 100 (92) f^1NH2 Antitubercular vs. 607 1/16 1/32 Bu"oSxr mg. % stasis (52)

333 TABLE 8 (Cont. ) CLASS 3 Aromatic Rings Compound Biological Test (Reference) = CH- =N- Antitubercular (X;jjNH2 vs. 607 1/16 1/16 Bu-oSx; mg. % stasis (52) Effect on blood pressure pressor a depressor P pressor CH2-CH2-NH2 (66,77,60) y pressor (weak) Local anesthesia active a inactive COO-CH2-CH2-NEt2 (30) P active (weak) Y inactive vs. T. pallidum in vitro 0AsO = 100 97 74 .=*, (41) oTT0 Syphilis active active HOV (56,123) ^N— CH2 Effect on blood pressure p ressor depressor ri^^ XNH-CH2 l%/*\^ (130) T3^ O-CO-Et Analgesic w ^ strong weak b1utyl (125)

334 TABLE 8 (Cont. ) CLASS 3 Aromatic Rings Compound Biological Test (Reference) = CH- * HN CO Antic on vuls ant 1 vs. electroshock ++++ ++++ OC f^^l HN — c ^2r rt *" (98) Antitubercular H^/^V/^-^ salicylate no. 1600 600 \ — f \-=f5/ (44)

335 TABLE 9 CLASS 3 Compound Biological Test (Reference) = CH- =N \ Antipyretic * .f CH3-N CO 1 1 ,VC"3 CH3 (50, 121) COOH OOH CH3 NH-CO-CH2.!X; " XCH3 Antitubercular stasis at mol. conC. 1/1200 1/160 (72) ,X-COO-CH2-CH7-NX NO - /. . ^ \ s 9 XH2-CH2' Antispasmodic intestinal strip 75 30 Papaverine = 100 (28) ^X!-CH2-CH2-NMe2 Antihistamine Benadryl = 1 1 1. 25 (126, 104) HO/-W/'). Estrogenic 10 mg. 10 mg. _ (131, 135) Syphilis Sb As OH unstable Salvarsan NH2 NH2 (149)

J .c .*- l0 •f' • - iM + o c 0, t O tn : O Z + o o .*. « V 1 (J w eeUe) h o o E £ h 11 V i 3 >. i4 «j u 0 25 l I £ ° 3 - TJ « — u hi « • u U1 a — — o >. -^ 5.8 a. 13 i4 V U ki 2 2- ~ h J3 I! 5 o. 9 IJ a u ** ** O Ii .5- 2 0 2 2. Q < j o ffl .o c u .3 O 0 o. u O o U iM 9 U 1M i iM U X 0 I t U 5? r g ife-. % U +5 « **i iXi IK, m N fO M V V 2 2 2

337 I I u o I W J eg u J o u 0. tO u ji u 00 o O u o a o U U Z X u X u u D* U—is. DO a o o £ Z I v B ** i a, « < 3 2 S Z i o u X z 6 u

338 X c o U U H J u U U a. X u i u V OS o a w x O U o o u-u-u ti I u la is ub mg 0 u C — - u ^J X u rM u

339 I z I u I u I/) u u u 0. _J 3 £ c < N 03 Q. .o O i4 S c 5 Z rg I 'J ^J X u c u .a i X Z-U-is. i ^J LJ 00 00 c ^ < - .o c

340 X CO « in en U J U W 0 , m X u X u E O h < .2 O .o o g a 8-i j: it .O C 3 SN Lf U O

341 X z X u o u Id 3 en en U U , a, tn 1 o 0, o u a. >, U M £ e & a o £ 2! « 0 U -a

342 TABLE 12 SPECIAL CLASS CO. SO, S02 Compound Biological Test (Reference CO SO S02 ci / \ ''x:- / \ ci Insecticidal \=/ \=/ vs. clothes moth inactive active active (88) H2N\ /^XV\ /NH2 Antibacterial inactive weak active _ (120) ^.XJ-NH-hete rocycle Antibacterial inactive active (83,120) 1 Analges.ic 100 120 Et*)C-C-CH?-CH-NMe7 '. 1 * 1 * Amidone = 100 0 CH3 (42) cH^*OAso vs. T. pallidum 42 49 0AsO =100 (39) H2N;x-^J*AsO vs. T. pallidum 0AsO = 100 45 29 HOs'XV 7 3 Me2N(X*. " (41) 8 112 Oj' \ .X-OH f *, Antitubercular in vitro active inactive (72) f^-OH Ditto IJ°H active inactive NH2 (94)

343 TABLE 13 SPECIAL CLASS -COO-, CH2-CH2 Compound Biological Test (Reference) -CO-O -CH2-CH2- HO OH Estrogenic (112) 1 00 gamma 100 gamma \.^ OH CH, Vitamin K activity (112) active Li. active

344 TABLE 14 SPECIAL CLASS Spatial Fixation by Ring or Double Bond Compound Biological Test (Reference Open Form Ring Form Antispasmodic vs. acetylcholine Atropine = 100 (22,92) 1 2. 3 2 0. 7 14 Adrenergic blocking (119) CH3 Analgesia strong (134) weak I I CH2—CH-CH2-NEt2 Antispasmodic vs. acetylcholine active (19) weak Antispasmodic vs. acetylcholine, Ba, histamine (84,17) active less active Antihistamine active inactive (126)

345 TABLE 14 (Cont. ) SPECIAL CLASS Spatial Fixation by Ring or Double Bond Compound Biological Test (Reference) Open Form Ring Form N-CH2-CH2-NMe2 Antihistamine (79) active inactive HO Estrogenic 1/40,000 of stilbestrol active (3D Estrogenic in gammas 0.4 (115) HO' OH Estrogenic 1 ++ 1 and 2 +++ (135) HO Estrogenic in gammas 0. 5 0.4 (115) CH.,— NH l : > CH2*-NH CS Antithyroid Thiouracil = 100 (26) 35 63

346 TABLE 14 (Cont. ) SPECIAL CLASSES Spatial Fixation by Ring or Double Bond Compound Biological Test (Reference Open Form Ring Form HN CO 1 Antithy roid 1 1 SC CH Thiouracil = 100 100 100 1 I /CH2 HN C-CHQ-f-- H2 (137) OCH/c=s Antithy roid Thiouracil = 100 14 116 (109) OC NH Ditto 1 1 14 10 CH2 C=S CH24-NH (109) /CH2-C-S^ CH2 , | C-NH2 Effect on blood pressure depressor pressor 2' (45) 0 Analgesic active (Amidone) inactive -..., N , CH3 1 '...-.-. 3.,CHj. (H) Antimalarial active inactive' i I ; v "iij" /CH3 (Paludrine) ^^.^NH-C-NH-C-NH-CH (1) r. y. NH N — f H3 C\/ ^NH-C-NH/ % Antimalarial + + + + + (34)

347 TABLE 14 (Cont. ) SPECIAL CLASS Spatial Fixation by Ring or Double Bond Compound Biological Test (Reference) Open Form Ring Form /-\ '. /*'X Antitubercular \ /~^-s / active inactive c 1 0 (57) Sympathomimetic "^jrl'?"2 * Vasoconstrictor + + ^/^CH2 Pressor weak weak (66) i^^N ' CHOH Ditto l^^ ^'L XCH-NH2 Vasoconstrictor + + 2 Pressor * weak (66) f^^i CHOH B rone hod ilator good good .' CH2 (69) 0 CH2-CH2 Antihistamine 9' CH2-CH2 Benadryl = 1 0. 1 2 (153)

348 TABLE 15 SPECIAL CLASS Polarity Shift TEST Exocyclic Endocyclic Analeptic activity f##Nsj|CO-NEt2 OcO-NEt2 N02 (46) Active Active f^, vs. Hem. strep, in mice H2N\ /So2-NH\ ^ 02N H2Nf^ /SO2-N"^^ (43) 15% greater activity than Sulfanilamide (Sulfapyridine) Histamine -like activity U^xJcH2-CH2-NH2 CH2-CH2-NH2 N02 N (90) None (pressor compound) Weak (0. 02 x histamine) Parasympathomimetic f#*#VNO-CO-NMe2 OG-CO-NMe2 R3N+ I (154,73,3) Active Active vs. Syphilis AsO AsO ^SsXNH2 OH o OH (123) Active Active

349 TABLE 15 (Cont. ) SPECIAL CLASS Polarity Shift TEST Exocyclic Endocyclic Antibacte rial vs. Staph. aureus Hexadecyl-N+-CH2-CH=CH2 Me2 + #~\ Hexadecyl- N V V-L_/ (146) 1/25,000 1/25,000 Antibacterial H2N\ ^SO2NH2 // \ N x> SO,NH7 \=/ 2 2 (145) Active Inactive ? Local anesthetic H2N/ \COO-CH2CH2-NEt2 \=/ N ^COO-CH7-CH?NEt7 \=/ (30) Active Inactive Analgesic 0S-COOR 0 COOR (^ (^ ^J NMe2 9 Me (129) Weak Strong O Vit. K. activity ^^Y^ ^CH3 ii i^V^ScHa 6 kAo (114) Potent Weak, but active

350 TABLE 16 SPECIAL CLASS Reversed Polarity TEST Analgesic (82,55) O g N i•. Demerol O ii Me 30 x Demerol Choline-Hke (Muscarinic) Acetylcholine = (33) Me3-N+-CH2-C-0-CH3 (Betaine ester) ++ K ? CH-C-O-CH2-CH2-NMe2 V very weak moderate O P\ u CH-C-O-CH:i-CH2-NEt2 (Trasentin) O 9-Fluorenyl-C:q-CH2-CH2-NEt2 (Pava.trine) O Me3-N+-CH2-O:C-CH3 (Acetyl - formocholine) A nti spasmodic vs. Histamine vs. Acetylcholine (5) O "CH-O.-C.-CH2-CH2-NMe2 very weak mode rate Antispasmodic (35) O 1l ,CH-p-C_-CH2-CH2-NEt2 Activity? Ditto (35) O I 9-Fluorenyl-O-C-CH2-CH2-NEi2 Activity ? Local anesthetic (61) :.O-NH-CH2-CH2-PiPeridine 0 NH2 NH:c.O-CH2-CH2-Piperidine NH2 weak? Activity?

351 TABLE 16 (Cont. ) SPECIAL CLASS Reversed Polarity TEST Local anesthetic COOMe COOMe OH NH2 NH2" (56) (Orthoform) Active (Orthoform new) Active Antitube rcular COOH COOH molar conc, for Q. stasis OH NH2 NH2 (72) 1/200 1/200 vs. T. pallidum AsO AsO in vitro {^i 0AsO = 100 NH/* NH2 OH"'* (40) 40 38 Pressor "1^i HoXN HO^s^XCH— CH-CH3 OH NH2 HOS^JcH— CH-CH, 1 1 3 NH2 OH (85,66) 1/12 epi Inactive

352 REFERENCES TO TABLES 1. Acheson, R.M. , King, F. E. , andSpensly, P. C. , Nature, 160. 53(1947). 2. Adams, D. F. , Biochim. Biophys. Acta, J, 1 (1949); CA. 43, 5062 (1949). 3. Aeschlimann, J. and Reinert, M. , J. Pharmacol. Exp. Therap. , 43, 413 (1931). 4. Alles, G.A. and Feigen, G. A. , J. Pharmacol. Exp. Therap., 72, 265(1941). 5. Alles, G.A. and Redeman, M. A. , J. Pharmacol. Exp. Therap.. 96, 338(1949). 6. Ashley, J. N. , Barber, H. J. , Ewins. A. J. , Newbery, G. and Self, A. D. H. . J. Chem. Soc. , 1942, 103. 7. Astwood, E. B. , J. Pharmacol. Exp. Therap., 78, 79(1943). 8. Bell, P.H. and P obi in, R. O. , Jr., J. Am. Chem. Soc., 64, 2905(1942). 9. Berger, F. M. , J. Pharmacol. Exp. Therap., 93, 470(1948). 10. Bernstein, S. , Oleson, J. J. , Ritter, H. B. and Sax, K. J. . J. Am. Chem. Soc.. 71, 2576 (1949). 11. Biel, J.H. , J. Am. Chem. Soc., 71, 1306(1949). 12. Blair, D. M. , Hawking, F. , Meeser, C. V. and Ross, W. F. , Brit. J. Pharmacol., 4_, 68 (1949). 13. Blicke, F. F. and Chanin, M. , Abstracts A. C. S. Meeting, Apr. 1946. 54K. 14. Blicke, F. F. and Krapcho, J. , Abstracts A. C. S. Meeting, Apr. 1948, 3K. 15. Blicke, F. F. and Tsao, M. U. , J. Am. Chem. Soc., 66, 1645(1944). 16. Bloch, H. , Brubach. G. , Erlenmeyer, H. and Suter, E. , Helv. Chim. Acta, 30, 539 (1947). 17. Bockmiihl, M. , Erhart, G. , Eisleb, O. and Stein, L. , U.S. 2,446,522(1948). 18. Bretschneider. H. , Monatsh., 76, 355(1947); CA. 42. 3359h(1948). 19. Brown, B. B. and Werner, H. W. , J. Pharmacol. Exp. Therap., 97, 157(1949). 20. Burgen, A. S. V. , Brit. J. Pharmacol. , 4, 219(1949). 21. Burtner, R.R. and Cusic, J. W. , J. Am. Chem. Soc., 65, 262 (1943). 22. Burtner, R.R. and Cusic, J. W. , J. Am. Chem. Soc., 65, 1582(1943). 23. Burtner, R.R. and Lehmann, G. L. , J. Am. Chem. Soc. . 62. 527 (1940). 24. Buth, W. , Kiilz, F. . and Rosenmund, K. W. , Ber. 72B, 19(1939). 25. Bywater, W. G. , Coleman, W. R. , Kamm, O. and Merritt, H. H. , J. Am. Chem. Soc., 67, 905 (1945).

353 26. Bywater, W. G. , McGinty, D. A. and Jenesel, N. D. , J. Pharmacol. Exp. Therap. , 85, 14 (1945). 27. Campaigne, E. E. , Starke, A. C. , Jr., Fosdick, L. S. and Dragstedt, C. A. , J. Pharmacol. Exp. Therap.. 71, 59(1941). 28. Cheney, L. C. and Bywater, W. G. , J. Am. Chem. Soc. , 64, 970(1942). 29. Cheney, L. C. Smith, R. R. and Binkley, S. B. , J. Am. Chem. Soc., 7_^, 53 (1949). 30. Chiang, M. C. and Hartung, W. H. , J. Org. Chem., 1_0, 26(1945). 31. Clark, E.P. andLinnell, W. H. , J. Pharm. and Pharmacol. , 1949, 211. 32. Cope. A.C., Kovacic, P. and Burg, M. M. , J. Am. Chem. Soc., 71, 3658(1949). 33. Craig, L.E. , Chem. Rev. , 42, 285 (1948). 34. Curd, F. H. S. and Rose, F. L. , Chem. and Ind. , 1946. 75. 35. Cusic, J. W. and Robinson, R. A. , U.S. 2,480,224, CA. 44, 1140(1950). 36. Dawes, G. S. . J. Pharmacol. Exp. Therap., 84, 160(1945). 37. Denton, J. J. , Turner, R. J. , Neier, W. B. , Lawson, V.A. andSchedl, H. P. , J. Am. Chem. Soc., JJ, 2048, 2050 (1949). 38. Eagle, H. , Doak, G. O. , Hogan, R. B. and Steinman, H. G. , J. Pharmacol. Exp. Therap., 7J2, 21 1 (1940). 39. Eagle, H. , Hogan, R.B. , Doak, G. O. , and Steinman, H. G. , J. Pharmacol. Exp. Therap., Ifl, 221 (1940). 40. Eagle, H. , Hogan, R. B. , Doak, G. O. , and Steinman, H. G. . J. Pharmacol. Exp. Therap., 74, 210 (1942). 41. Eagle. H. , Hogan, R.B. , Doak, G. O. . and Steinman, H. G. , J. Pharmacol. Exp. Therap., 81, 142 (1944). 42. Eddy, N. B. , Touchberry, C. F. and Lieberman, J. E. , J. Pharmacol. Exp. Therap., 98, 121 (1950). 43. Erlenmeyer, H. , Aeberli, M. and Sorkm. E. , Helv. Chim. Acta, 30, 2066(1947). 44. Erlenmeyer, H. , Becker, C. , Sorkin, E. , Bloch, H. and Suter, E. , Helv. Chim. Acta, 30, 2058 (1947). 45. Erlenmeyer, H. and Jung, J. P. , Helv. Chim. Acta, 3j, 35(1949). 46. Erlenmeyer, H. and Jung, J. P. , and Sorkin, E. , Helv. Chim. Acta, 29, 1960(1946). 47. Erlenmeyer, H. and Kleiber, A., Helv. Chim. Acta, 21, 111 (1938). 48. Erlenmeyer, H. and Morel, C. J. , Helv. Chim. Acta, 28, 362(1945). *9. Erlenmeyer, H. and Miiller. M. , Helv. Chim. Acta, 28, 922 (1946). 50. Erlenmeyer, H. and Willi, E. , Helv. Chim. Acta, Iji, 740(1935). 51. Fastier, F. N. and Smirk, F. H. , J. Pharmacol. Exp. Therap., 89, 256(1947). 52. Feinstone. W. H. , Friedman, H. L. , Rothlauf, M. L. , Kelly, A.M., and Williams, R. J. , J. Pharmacol. Exp. Therap., 89, 153(1947).

354 53. Feldkamp, R. F. and Faust, J. A. , J. Am. Chem. Sac.. 7J , 4012(1949). 54. Fieser, L. F. and Richardson, A. P. , J. Am. Chem. Soc. , 70, 3156(1948). 55. Foster, R. H. K. and Carman, A. J. , J. Pharmacol. Exp. Therap. , 91, 195(1947). 56. Frankel, S. , Arzneimittel Synthese, Springer 1927. 57. Freedlander, B. L. , Am. Rev. Tuberc. , 49, 543(1944). 58. Friedman, H. L. , Braitberg, L. D. , Tolstoouhov, A. V. andTisza, E. T. , J. Am. Chem. Soc. , 69, 1795 (1947). 59. Friedman, H.I. and Tolstoouhov, A. V. , U.S. 2,465,865. CA, 43, 6244d(1949). 60. Fromherz, K. and Spiegelberg, H. , Helv. Physiol. Acta, 6, 42 (1948). 61. Gaind, K. N. and Vohra, P. N. , J. Indian Chem. Soc.. 23, 9(1946); CA. 40, 57092 (1946). 62. Gilman, H. andPickens, R. , J. Am. Chem. Soc. , 47, 245(1925). 63. Halpern, B. N. , Arch, intern, pharmacodynamie, 5.9, 149 (1938). 64. Hambourger, W. E. and Jamieson. R. B. , Jr., J. Pharmacol. Exp. Therap., 58, 53(1936). 65. Hansen, H. L. and Fosdick, L. S. , J Am. Chem. Soc., 55, 2872(1933). 66. Hartung, W. H. . Ind. Eng. Chem., 37, 126(1945); Chem. Rev., 9, 389(1931). 67. Hartung, W. H. , Munch, J. C. and Crossley, F. S. , J. Am. Chem. Soc., 57, 1091 (1935). 68. Hazard, R. , Cheymol, J. , Chabrier, P. , Smarzewska, K. , Compt. rendu, 226, 1850 (1948). 69. Heinzelmann, R. V. , Aspergren, B. D. and Hunter, J. H. , J. Org. Chem., 14, 907(1949). 70. Henne, A. L. , J. Am. Chem. Soc., 59, 1200(1937). 71. Hewitt, R. I. , Gumble, A., Kushner, S. , Safir, S. R. , Brancone, L. M. and Subbarow, Y. , J. Pharmacol. Exp. Therap. , 96, 305 (1949). 72. Hirt, R. and Hurni, H. , Helv. Chim. Acta, 32, 378 (1949). 73. Hoffman-La Roche, F. and Co. , Swiss 246,834. CA. 43, 5050d (1949). 74. Hoggarth, E. , Martin, A. R. . Storey, N. E. , Young, E. R. P. , Brit. J. Pharmacol. , 4, 248 (1949). 75. Halpieu. H. R. , Kitchel, J. H. and Weatherby, J. H. . J. Pharmacol. Exp. Therap., 68, 395(1940). 76. Hunt, R. andRenshaw, R. R. , J. Pharmacol. Exp. Therap. ,£5, 315(1925). 77. Hunt, W. H. and Fosbinder, R. J. , J. Pharmacol. Exp. Therap., 75, 299(1942). 78. Huttrer, C. P. , Enzymologia. 1_2, 277(1948). 79. Huttrer, C.P. , J. Am. Chem. Soc., 71, 4148(1949). 80. Huttrer, C. P. , Djerassi, C. , Beears, W. L. , Mayer, R. L. , and Scholz, C. R. , J. Am. Chem. Soc. , 68, 1999 (1946).

355 81. Jenkins, G. L . and Hartung, W. H. , The Chemistry of Organic Medicinal Products, Wiley, 1949. 82. Jensen, K. A. , Lindquist, F. , Rekling, E. and Wolffbrandt, C. G. , Dansk. Tids. Farm., 1J7, 173(1943); CA, 39, 25066(1945). 83. Johnson, O. H. , Green, D. E. andPauli, R. , J. Biol. Chem. , 153, 37(1944). 84. Kleiderer, E.G., Rice, J. B. , Conquest, V., Williams, J.H. , Department of Commerce, PB-981, July 1945. 85. Kovacs, J. and Horvath, T. , J. Org. Chem., 14, 306(1949). 86. Kushner, S. , Brancone. L. M. , rlewitt, R. I. , McEwens, W. L. , Subbarow, Y. , Stewart, H. W. , Turner, R. J. , and Denton, J. J. , J. Org. Chem., 1.3, 144(1948). 87. Kushner, S. / Dalalian, A., Cassell, R. T. , Sanjurjo, J. L. , McKenzie, D. and Subbarow, Y. , J. Org. Chem., 1_3, 834(1948). 88. Lauger, P., Martin, H. and Miiller. P., Helv. Chim. Acta, 27, 892 (1944). 89. Lee, H. M. , Dinwiddie, W. G. and Chen, K. K. , J. Pharmacol. Exp. Therap., 89, 83(1947). 90. I.ee, H. M. and Jones, R. G. , J. Pharmacol. Exp. Therap., 95, 71 (1949). 91. Lehmann, G. and Knoefel, P. K. , J. Pharmacol. Exp. Therap., 63, 453 (1938). 92. Lehmann, G. and Knoefel, P. K. , J. Pharmacol. Exp. Therap. 74. 274(1942). 93. Lehmann, G. and Randall, L. O. , J. Pharmacol. Exp. Therap., 93, 115(19.48). 94. Lehmann, J. , Lancet, 250, 15 (1946). 95. Lerman, J. and Harington, C. R. , J. Clin. Endocrinol. , 9, 1099(1949). 96. Lipschitz, W. L. and Hadidian, Z. , J. Pharmacol. Exp. Therap., 8_1 , 84(1944). 97. l.oew, E.R., Kaiser, M. E. and Moore, V., J. Pharmacol. Exp. Therap., 83, 120(1945). 98. Long, L. M. , First Natl. Med. Chem. Symp. , Am. Chem. Soc. , 1948, 114. 99. Lott, W. A. , Trans. N. Y. Acad. Sci. (II), 11, 2 (1948). 100. Loubatieres, A. , Compt. rend. soc. biol. , 143, 155 (1949); CA, 43, 9238e (1949). 101. Macht, D. I. , Bryan, H. F. and Grumbein, M. L. , J. Pharmacol. Exp. Therap., 63, 279 (1938). 102. Mann, F. G. and Porter, J. W. G. , J. Chem. Soc., 1947, 910. 103. Marsh, D. F. , J. Pharmacol. Exp. Therap., 94, 192(1948). 104. Marsh, D. F. , Relationship between chemical constitution and antihistaminic activity (mimeographed) Morgantown, Va. , 1949. 105. Marsh, P. B. , Butler, M. L. and Clark, B. S. , Ind. Eng. Chem., 4_1, 2176(1949). 106. Masson, G. , Rev. can. biol., 3, 491 (1944). 107. McElvain, S. M. and Carney, T. P. , J. Am. Chem. Soc., 68, 2592(1946). 108. McGinty, D.A. and Bywater, W.C., J. Pharmacol. Exp. Therap., 85, 129 (1945).

356 109. McGinty, D. A. and Bywater, W.C., J. Pharmacol. Exp. Therap. , 84, 342(1945). 110. Meier, R. and Hoffmann, K. , Helv. med. Acta, 2. 106(1941). 111. Melville, K.I. andStehle, R. L. , Can. J. Research E 22. 95(1944). 112. Mentzer, C. , Gley, P., Molho, D. and Billet, D. , Bull. Soc. Chim. , 1946, 271. 113. Metcalf, R. L. , Mode of Action of Organic Insecticides, Nat'l Res. Council, Wash. , D. C., 1948). 114. Meunier, P. and Mentzer, C. , Bull. soc. chim. biol. , 2_5, 80 (1943). 115. Miescher, K. , Chem. Rev., 43, 367(1948). 116. Miller, O. H. and Fischer, L. , J. Am. Pharm. Assoc. , Sci. Ed., 30, 45(1941). 117. Molitor. H. , J. Pharmacol. Exp. Therap., 58, 337(1936). 118. Moore, M. B. , J. Am. Pharm. Assoc., Sci. Ed., 33, 193(1944). 119. Nickerson, M. and Gump, W. , J. Pharmacol. Exp. Therap., 97, 25 (1949). 120. Northey, E. H. , The Sulfonamides and Allied Compounds, Reinhold, 1948. 121. Orestano, G. , Arch. ital. sci. farmacol. , 8, 353(1939); CA, 3J, 1926(1945). 122. Prelog, V., Juhasz, S. , Rezek, A. and Stern, P., Helv. Chim. Acta, 25, 907(1942); CA, 3J, 12263 (1943). 123. Raiziss, G. W. , Severac, M. andClemence, L. W. , U.S. 2,476,508; CA, 43, 8105c (1949). 124. Ramsey, H. and Richardson, A. G. , J. Pharmacol. Exp. Therap., 89, 131 (1947). 125. Randall, L. O. and Lehmann, G. L. , J. Pharmacol. Exp. Therap., 9_3. 314(1948). 126. Rieveschl, G. , First Nat'l Med. Chem. Symp. , Am. Chem. Soc., 1948, 125 (oral presentation). 127. Rieveschl, G. , U.S. 2,483,436(1949). 128. Rohmann, C. and Scheurle, B. , Arch. Pharm., 274. 110(1936). 129. Rubin, M. and Wishinsky, H. , Abs. A. C.S. Meeting, April 1946, 53K. 130. Scholz, C.R. , Ind. Eng. Chem., 37, 120(1945). 131. Schueler, F. W. , Science, 103, 221 (1946). 132. Seifter, J. and Ehrich, W. E. , J. Pharmacol. Exp. Therap.. 92, 303(1948). 133. Shaffer, C. B. and Critchfield, F. E. , J. Biol. Chem., 174, 489(1948). 134. Shapiro, D. , J. Org. Chem., 14, 839 (1949). 135. Solmssen, U. V. , Chem. Rev., 37, 481 (1945). 136. Spielman, M. A. , First Natl. Med. Symp., Am. Chem. Soc., 1948, 119. 137. Stanley, M. M. and Astwood, E. B. , Endocrinology, 41, 66(1947). 138. Stanley, M. M. and Astwood, E. B. , Endocrinology, 41, 66 (1947).

357 139. Stewart, H. W. , Turner, R. J. , Denton, J. J. , Kushner, S. , Brancone. L. M. , McEwen, W. L. , Hewitt, R.I. and Subbarow, Y. , J. Org. Chem. , 13, 134(1948). 140. Suter, C. M. , Chem. Rev., 28, 269(1941). 141. Swanson, E. E. and Fry, W. E. , J. Am. Pharm. Assoc. , Sci. Ed., 29, 509(1940). 142. Tabern, D. L. and Volwiler, E. H. , J. Am. Chem. Soc. , 57, 1961 (1935). 143. Tainter, M. L. , Arch, intern, pharmacodynamie, 42, 128 (1932). 144. Thompson, R. L. , Wilkin, M. L. , Hitchings, G. H. and Russell, R. B. , Proc. Soc. Exp. Biol. Med. . 72, 169(1949). 145. Tiesler, A. E. , U.S. 2,349,060, CA, 39, 820(1945). 146. Valko, E.I. and OuBois, A. S. , J. Bact. , 50, 481 (1945). 147. Vaughan, J. R. , Anderson, G. W. , Clapp, R. C. , Clark, J. H. . English, J. P. , Howard, K. L. , Marson, H. W. , Sutherland, L. H. , and Denton, J. J. , J. Org. Chem., 14, 228 (1949). 148. Viaud, P. , Produits pharm. 2, 53 (1947). 149. Wagner-Jauregg, T. , Therapeutische Chemie, Huber, 1949. 150. Welch, A. D. and Landau, R. L. , J. Biol. Chem., 144, 581 (1942). 151. Welch, A. D. and Roepke, M. H. , J. Pharmacol. Exp. Therap. , 55, 118(1935). 152. Wick, H. , Arch. exp. Path. Pharmakol. , 205. 490 (1948); CA, 43, 7137c (1949). 153. Wright, J. B. , Kolloff, H. G. and Hunter, J. H. , J. Am. Chem. Soc., 70, 3098 (1948). 154. Wuest, H. M. and Sakal, E. , Abs. A. C, S. Meeting, March 1949, 7k.

358. DISCUSSION DR. D. W. WOOLLEY (Rockefeller Institute for Medical Research, New York. New York): This question of isosterism is one I have given considerable thought to, and I would like to tell you how I look at the matter. As Dr. Friedman pointed out, Erlenmeyer gave great impetus to this study, with investigations in the early thirties. He came against a blank wall, though, when the isostere of thiamine was made; that is, a thiazole ring in the vitamin was replaced by a pyridine ring and, contrary to what had been expected (that is, that it would have thiamine activity), it was an exceedingly powerful antagonist; it had the exactly opposite effect from thiamine. This was one of the serious challenges to the idea of isosteric replacement giving rise to biological materials of the same potency, and it still is. The confusion which seems to reign in the field, and to which Dr. Friedman referred, resides partly in this fact. If in a drug, one makes isosteric replacements, such as the exchange of a pyridine ring for a thiazole, one may obtain a compound of similar qualitative action, although in all respects the two may not be identical. Actual experience showed that drugs of similar activity could be found by making such isosteric changes in existing drugs. This may have been because various members of a series of antimetabolites were being made. If, however, the limit is exceeded and the metabolite rather than the. analog is made, an entirely different situation arises. Among pharmacologists and others as well, hormones have been considered as drugs. Choline, for instance, for years, was thought of as just a synthetic drug, until it was isolated from tissues and gradually shown to be an essential metabolite. The case of adrenalin sometimes can be viewed in the same way. The confusion of drugs with metabolites was therefore profound with the hormones. With the vitamins, however, the matter can be seen more clearly. Because the vitamins are essentially nontoxic, enormous doses can be given to normal animals with no detectable effect. This is not true of many of the hormones. If in a vitamin various isosteric replacements are introduced, an antagonist, frequently a very powerful one, which has exactly the opposite effect to that of the vitamin, may result. The same is true with many amino acids. With this in mind, we can regard some of the groups of isosteric drugs as being mere families of antimetabolites. If the real metabolites to which they may be related are tested, then there should be no surprise if their activity is quite the opposite of the series to which they are isosteric. I think, in coming down to the practical matter of thinking about synthetizing useful agents, we must never lose sight of this fact that, if we start with the biologically active compound (that is, the metabolite), it is an entirely different problem than if we start with the product in the laboratory; to wit, an already existing analog. Isosteric replacement may have quite opposite effects in these two situations. DR. FRIEDMAN: This may be an erroneous way of looking at it, but I always consider that an antimetabolite really has the same mechanism of action as the metabolite. Thinking in a pictorial manner, both the metabolite and its isostere start with the same mode of action: the isostere is hooked into the system in the same place as the metabolite would have been. But the rest of the mechanism beyond this stage is blocked when the antimetabolite is present and thus growth, or whatever consequence is involved, is prevented.

PANEL DISCUSSION ON ANTIMETABOLITES, CARCINOGENESIS AND CANCER CHEMOTHERAPY M.J. Shear, Moderator National Cancer Institute National Institutes of Health Bethesda, Maryland

360 ANT1METABOLITES D. W. Woolley Rockefeller Institute for Medical Research New York. New York Mr. Chairman, I feel I have used up my five minutes already. To keep within the time limit, I think we can dispense with most of the slides. I believe it is needless to say to you all that antimctabolites exist, or to show you the structure of sulfanilamide, p-aminobenzoic acid, and so forth; I believe you all know those. The basic phenomenon, it should be recalled, is that, if one takes a vitamin - it has been done with all the water-soluble ones and some of the fat- soluble ones, too - or an amino acid, and in some cases a hormone or a purine, any one of a number of metabolically essential compounds, and if one modifies its structure in some small way, replacing an atom by some other atom or replacing a group by some other group, one frequently obtains a substance which produces, in various kinds of living things, the signs of deficiency of that metabolite. Frequently, one can overcome the toxicity attendant on such a demonstration, and nullify the effect by giving the structurally related metabolite. This has been done with bacteria and with viruses; it has been done in animals; it has even been done, in a few cases, in man. This idea arose some fifteeen or twenty years ago among enzymologists, and received great impetus by the discovery that a clinically useful agent, sulfanilamide, was of this type: that is, its action was overcome by a structurally related, metabolically important substance, p-aminobenzoic acid. Many people thought it would be easy to reverse the idea and make new bactericidal and therapeutic agents, by alteration of the structure of other metabolites. In general, this has not proven to be so. There have been far more failures than successes. One just cannot select at random some amino acid or some vitamin and alter its structure in such a way as to produce an anti-metabolite, and thus arrive at a fairly useful therapeutic agent. One of the problems before us this afternoon, is how to arrive at therapeutically active agents. It has therefore seemed to me that one of the ways we should approach this problem is to try to understand selective toxicity. As you know, sulfanilamide derivatives are essentially nontoxic to higher animals, yet are quite harmful to many pathogenic bacteria. I think we should inquire why this is so, because I believe any attempt, other than those based on chance, to make therapeutic agents will have to delve into the reasons why such selective effects are obtained. In the case of the sulfanilamide drugs, something is known as to why they have this selective effect. Why don't they kill animals as well as bacteria? From the work of Woods and from that of Miller, it is clear that p_-amino- benzoic acid is used as a substrate from which, by a'series of reactions, the living cell synthesizes folic acid. Practically all of the pathogenic bacteria carry out this reaction very well; that is, they produce their own j>-aminobenzoic acid, and they metabolize that further and produce folic a.cid. Higher animals, however, apparently do not do this, at least not at a rate sufficiently fast to meet their needs. Woods has also shown that sulfanilamide inhibits this enzymic synthesis of folic acid from p-aminobenzoic acid. One cannot inhibit the synthetic reaction in animals, because they do not possess it; hence, they go largely unharmed, but the bacteria do possess this reaction; they depend on it for multiplication and, when the reaction is inhibited, their growth (multiplication) is inhibited. So I think the basis of selectivity, one of the reasons why the sulfonamide drugs are selective, is the difference in their nutritional requirements for folic acid.

361 We can test this hypothesis by applying it to some other metabolic system. If we choose a vitamin for which we know a biological precursor, then we can make an ana- log of that precursor and should be able to show that it is poisonous to those species which synthe- size the vitamin, and not poisonous to those which depend on their diets and do not synthesize it. If I might have the last three slides, please. We chose biotin because it is a vitamin which is required by animals, and not required by certain pathogenic bacteria, such as the tubercle bacillus. COOM (CH2)5 COOH Pjmelic acid C- [2,4-Dichloro-sulfanilidq] - cappoic acid Slide 1 It fits our needs beautifully. We also chose it because it had previously been known that pimelic acid was probably a precursor from which it was formed in vivo. Therefore, the problem was to make analogs of pimelic acid. This could easily be done. Because of the experi- ence accumulated over the past decade, a rather good guess can be made as to how the structure of a'metabolite should be changed in order to form an anti-metabolite. The compound shown in the first slide was the first one tried; it is the dichloro-sulfanilide corresponding to pimelic acid, and it has the selective activity envisioned for it. Parenthetically, if one wishes to correlate activity and structure, this analog presents an interesting case. If this compound had been found accidentally and its activity had been observed, without knowledge of how it was brought about, the ring system would have attracted attention, because rings are rather attractive to manufacturers of drugs. The alkyl side chain might have been pictured as something attached to make the compound soluble, or otherwise to improve its properties. Actually, however, it is quite the other way. The ring exists as part of the structural alteration which makes this compound an antimetabolite. The structural analogy lies in the side chain. Among the antihistamines, this same consideration can sometimes be seen. For example, in phenergan, the phenothiazine ring is part of the structural alteration of histamine; the part which bears the structural analogy is the side chain. In the case of the pimelic acid analog, it was found that, by moving the halogens to different positions around the ring or by using bromine atoms instead of chlorines, activity could be markedly enhanced or diminished. The second slide will show some of the effects of this analog on Bacillus tenuis. The addition of the sulfanilide inhibited growth, and the effect was overcome by pimelic acid or biotin.

36Z Inhibition of Growth of B. tennis Cultured in Flnsks b?t c-(£,4-Dirhlorosulfonili'h\ caproic Acid and Its Reversal by Piinelic Acid or Biotin Antlofue Bloun Pinulic acid Turbidity 7 tf «. 1 tv cc. > P*r fc 0 0 0 70 20 0 0 60 50 0 0 82 100 0 0 98 100 0.0005 0 66 100 0 3 70 Slide 2 The antagonism with pimelic acid was competitive whereas that with biotin was noncompetitive. If an amount of biotin which would meet the needs of growth was present, then no amount of the analog was poisonous until a very high concentration was reached (2 mg. per cc.). As predicted, this compound was nonpoisonous to mice, a species which requires a dietary source of biotin; 10 mg. per day injected into a mouse caused no detectable effect. In the third slide a number of micro-organisms are shown along with their susceptibility to this analog. Except for E^. coli, which does not fit the prediction, the ones which require biotin are not affected by the analog; the ones which make their own are. Correlation of Toxicity of *-(t,4-Dichloro8ulfanilido)-caproic Acid with Nutritional Needs for Biotin Organism Analomie neeoM for Half ma.imal inhibition Nutritional requirement for biotin iKrec. Bacillia tenuit 58* Not required, but stim- ulatory under certain conditions A cetobacter suboxydant 350 Not required, but some- what stimulatory Mycobacterium tubercvlorii H37Rv 20t Not required Etcherichia coli No effect at 1000 i, *i " " biotinless 1000 Required Proteus strain 4 " " " 1000 " Staphylococcui aureut ii « « 300 " Lactobacillua catei 500 " " arobinotut ii ii ii 500 11 Lcuconoatoc mrsenteroiiles " " " 500 " Hemolytic streptococcus H6QD . . . . . 500 " Saccharomycet cereviriac i* ii ii 1000 " * When cultured in test.tubes, a value of 20 was uniformly found. t I0cubation time 4 days. Slide 3 This is a laboratory model to think about in trying to visualize one way in which selectively active agents can be made. This is undoubtedly not the only basis on which to proceed; it is merely one.

363 CHEMICAL-BIOLOGICAL COPRELATION IN THE FIELD OF CARCINOGENESIS I. Berenblum The Weizmann Institute of Science Rehovoth, Israel Chemical carcinogenesis became an established science when Kennaway, Cook and their associates, isolated and identified 3,4-benzpyrene as a potent carcinogenic constituent of coal-tar. 9 Previous tests of a wide range of compounds had already demonstrated that tumors could be artificially induced in mice by repeated applications to the skin of the synthetic 5-ring polycyclic hydrocarbon, 1 , 2, 5, 6-dibenzanthracene, though no such action was obtained with anthracene, phenanthrene, carbazole, acridine, chrysene, pyrene, picene, perylene, and many other hydrocarbons and related compounds. Here, then, was a promising field for the study of the relation between chemical structure and biological activity of a unique kind; and both Kennaway and Cook, in England, and Fieser and Shear, in the United States, were quick to realize the potentialities of this new approach. In less than a decade, prior to the outbreak of World War II, several hundred new compounds were synthesized and tested for carcinogenic activity, including interesting series of homologues, and related derivatives, of certain key compounds, carefully chosen with an eye to the elucidation of the biochemical mechanism of carcinogenesis. In the short time available to me here, I shall only be able to pick out one or two illustrative examples from this important branch of carcinogenesis, and to deal equally briefly with other important groups of carcinogens of different chemical structures and biological properties, in order that I may devote a little time to a consideration of postulated mechanisms of action and chemical-biological correlations, and to conclusions that may be drawn about the validity of such associations. Of the fifteen possible 5-ring polycyclic hydrocarbons, twelve were already known at the •ime these early investigations were in progress, and of these, only two - 1 , 2 . 5, 6-dibenz- anthracene and 3,4-benzpyrene - had been found to be carcinogenic. Z These two compounds could be considered structurally related, in the sense that the 1 , 2-benzanthracene configuration j was present in both. Since 1, 2-benzanthracene itself displayed little or no carcinogenic activity, it seemed reasonable to extend the study to substituted derivatives of this compound. The accumulated data from these studies have been tabulated by Hartwell, 17 and the extensive literature on the subject ably reviewed by Cook and Kennaway, 10,11,12 Fieser,14,15 and more recently by Badger. ' The more important conclusions derived from this work may briefly be summarized as follows: 1,2,5. 6-dibenzanthracene 3,4-benzpyrene

364 (1) Carcinogenic activity was conferred on the 1, 2-benzanthracene molecule by methyl substitu- tion in positions 5, 9 and 10. and somewhat less so in positions 6, 3 and 4, with no activity when substitution was in any other position. (21 Di- and trimethyl substitution conferred even greater carcinogenic activity, provided again that these occupied the above-mentioned 'active' positions. (Certain anomalous results were obtained, as. for instance, the absence of carcinogenic action by 3, 9-dimethyl-l. 2-benzanthra- cene. ) (3) Ethyl and propyl substitutions were also effective, but increasing the length of the side chain beyond a point, led to diminution, and finally to extinction, of carcinogenic activity. (4) Polar substituents were, on the whole, unfavorable for carcinogenic activity. (5) Hydrogenation of any part of the aromatic ring structure usually led to loss of activity, though there were a few exceptions to this. (6) Substitution of a thiophene or pyridine nucleus for one of the benzene rings of a carcinogenic hydrocarbon, did not seem to interfere with carcinogenic activity, though an oxygen-containing ring destroyed activity. The results, so far, were encouraging, in favor of a close correlation between chemical structure and carcinogenic activity. There were, however, already indications at an early stage, that the 1, 2-benzanthracene structure was not an essential prerequisite for carcinogenic activity, since 3, 4-benzphenanthrene was found to be carcinogenic. When Uacassagne24 demonstrated that mammary tumors could be induced in mice by injections of the ovarian hormone, estrone, both sterol chemistry and endocrinology came within the ambit of experimental carcinogenesis. On the basis of, what then seemed, a plausible hypothesis2l , 2 . tnat tl,e body might, under certain conditions, convert sterols into carcinogenic polycyclic hydrocarbons - attempts were made to carry out this transformation synthetically. By cyclization of the side chain, followed by dehydrogenation, bile acids were successfully converted into 20-methylcholanthrene, which was subsequently shown to be highly carcinogenic. CH3 3,4-benzphenanthrene desoxycholic acid 2O-methylcholanthrene The idea that estrone might act by preliminary conversion into a methylcholanthrene-type of carcinogen, became somewhat weakened when it was found that mammary tumors could also be induced by stilbestrol and other synthetic estrogens of simpler structure. It seems possible, moreover, that the function of estrone in mammary tumor production might be merely to prepare the tissue for subsequent action by a viral or other factor, rather than that it was itself a straight- forward carcinogen. At the same time, it is now known that there are sterols in the body that possess weak, though definite carcinogenic activity. ^ The concept of a correlation between chemical structure and carcinogenic activity was faced with a more serious difficulty when Yoshida3S demonstrated that liver tumors could be induced in rats and mice by feeding or injections of an azo dye - 2-amino-5-azotoluene.

365 KinositaZ2 extended this observation to other azo compounds, and showed that p-dimethylamino- azobenzene was even more potent in producing liver tumors in the rat. The work has been followed up by several investigators, using a large group of azo compounds, including isomers, analogues, and split products of those previously shown to be carcinogenic. (The early literature is well reviewed by Shear, 31 while reviews of subsequent work are presented by Kirby, 23 Orr,29 Cook, 8 Miller and Miller,25 and others. ) -NH2 CH3 2-amino-5-azotoluene p -dime thy laminoazobenzene Within the group of azo compounds, as in the case of polycyclic hydrocarbons, it is possible to discern a certain correlation between structure and biological activity. Starting with p-dimethylaminoazobenzene, carcinogenic activity is maintained when one of the two methyl substituents is replaced by hydrogen, or by an ethyl group, but is entirely lost when both methyl groups are replaced by ethyl or other groups. Additional methyl groups in other parts of the molecule also appear to influence carcinogenic activity according to a pattern, though one which is not easy to interpret. The azo group appears to be important for carcinogenesis, in so far as all the reduced products tested proved to be inactive. (Yet, it was recently found that £-dimethylaminostilbene - where the azo group is replaced by a -CH=CH- group - is also carcinogenic. 1°) Related to the azo group of carcinogens are certain amino compounds, such as p-naphthyl- amine, responsible for tumor production in the urinary bladder, 20 and amino- and acetylamino- fluorene, which induce a variety of tumor types, including those of the liver, breast, external auditory canal, etc. , 14 when these compounds are incorporated in the diet. Attempts have been made to find a structural link between these carcinogens of somewhat 'simple' structure and those belonging to polycyclic hydrocarbons, on similar lines to the suggested relation between sterols and polycyclic hydrocarbons. An interesting example of such a postulated conversion is from two molecules of p-naphthylamine to 3, 4, 5, 6-dibenzcarbazole, the latter being known to be carcinogenic. 6, 8 H-CO-CH3 NH2 Acetylaminofluorene - naphthy lam ine 3,4,5,6 -dibenzcarbazole There is, as yet, no experimental evidence in support of this hypothesis; the arguments in its favor are: (1) that some of these postulated conversions in vivo can be carried out synthet- ically, and (2) that the pattern of carcinogenic action by these relatively simple compounds differs from that of polycyclic hydrocarbons, in that in the case of these 'simple carcinogens', tumors do not usually arise at the site of application or injection, but appear in certain specific organs, the distribution being different for the different compounds. This might be expected to occur if the 'simple' carcinogens were really precursors of carcinogens, and that tumors only arose in those organs in which the chemical conversion, from precursor to carcinogen, took place.

366 The little that is known about the metabolism of carcinogens in the body, is insufficient to confirm or contradict the above hypothesis. (See review by Boyland and Weigert. ?) Polycyclic hydrocarbons are oxidized in the body to phenolic derivatives, presumably through an inter- mediate stage of dihydrodihydroxy- derivative, and there is more recent evidence^ that the oxidation can go further in the body, leading to the break-up of the molecule. The phenolic deriv- atives, so far isolated, possess little or no carcinogenic activity. More is known about the metabolism of azo carcinogens. (See Miller and Miller. ") Such compounds tend to be demethylated, and also undergo cleavage at the azo linkage. Except for the first stage (demethylation to the monomethyl compound), metabolic changes in the compounds lead to loss of carcinogenic activity. It will be noted that, so far, it is possible, by fanciful speculation, to correlate chemical structure and carcinogenic activity. But this breaks down completely when one passes on to other types of carcinogens, such as urethane (ethylcarbamate), which produces tumors of the lung, 27 and carbon tetrachloride, which produces tumors of the liver;13 to say nothing of the claims in the literature of skin tumor production by painting with HC1 or NaOH, 26 or sarcomas by injection of glucose, 28 or even by implantation of bakelite disks subcutaneously. 33 An attempt has been made, in recent years, to find a common feature among carcinogenic compounds, based on physio-chemical criteria. French scientists^ have been trying to correlate carcinogenic potency of hydrocarbons and related compounds with the electron density at the K region of the molecule. The comparisons are based on quantitative evaluations of the electron densities, and on a quantitative grading of the respective carcinogenic potencies. This, at once, raises the important issue, as to whether carcinogenic potency can be considered as an absolute value. It will be recalled that the early studies on carcinogenesis were restricted to skin paint- ing in mice, by the English group, and to subcutaneous injections in mice, by the American group. Though the results, in most cases, were remarkably similar by the two methods of assay, this was not so in every case. For instance, 10-methyl- 1, 2-benzanthracene was found to be highly potent by the subcutaneous route, but relatively weak when tested on the skin. 32 Several other such examples were noted, and in the case of some of the weaker carcinogens, some were active only on the skin, and others, only subcutaneously. Such differences became very much more accentuated when these studies were extended to other tissues, and to other species of animals. For instance, the mouse and the rabbit behave very differently in their response to carcinogenesis according to the carcinogen used. Tar readily induces skin tumors in both species; but benzpyrene, one of the constituents of tar, is strongly carcinogenic to the skin of the mouse, but only weakly so to the skin of the rabbit. A benzpyrene-free fraction of tar has recently been isolated from tar which behaves in the opposite manner to benzpyrene: it is very strongly carcinogenic to the skin of the rabbit, but not to the skin of the mouse. 4 Several other examples may be quoted, such as the fact that p-naphthylamme produces bladder tumors in the dogs but not in the rabbit; or the fact that 2-amino-5-azotoluene produces liver tumors in mice even more effectively than in rats, whereas £-dimethylaminoazo- benzene is very active in rats but hardly at all in mice. Perhaps the most striking example is the behavior of 9, 10-dimethyl-1, 2-benzanthracene, which is the most potent carcinogen for skin in rabbits and mice, and very active for sarcoma production when injected subcutaneously in mice, but entirely non-carcinogenic by that route in the rabbit. What, then, is meant by the carcinogenic potency of a compound? Should 9, 1 0-dimethyl- 1 , 2-benzanthracene be considered the most potent of all carcinogens, because it undoubtedly is so when tested on the rabbit's skin, or should it be deemed non-carcinogenic, which it is when tested in the same animal by the subcutaneous route ? Our present-day concept of carcinogenic potency is based on very limited, and arbitrarily chosen, conditions of assay, and there is no doubt that, by extending the tests to other tissues and other species of animals, a very different picture would be obtained. The problem is even more complex, since it is now known that carcinogenesis is not a single biological process, but comprises at least two independent phases - initiating action and promoting action. 5 Some agents can induce promoting action without being capable of

367 inducing the initiating process; and it is conceivable that there are other agents that produce the opposite effect. Carcinogens, as we know them, produce both effects, but we cannot say yet whether a relatively weak (over-all) carcinogenic effect is due to deficiency of either one or other, of both effects. ' In the light of all this, one cannot avoid the conclusion that the search for a close correla- tion between chemical structure and carcinogenic activity may be founded on an unjustifiable premise. . It is tempting for the scientist to look for correlations, since without them, mere accumulation of data lead to confusion and chaos; but he must always beware of seeking relation- ships that do not, in fact, exist. A whiff of chloroform and a blow on the head with a stick both produce unconsciousness. That does not mean that there is any relationship between the chemical constitution of chloroform and the wood from which the stick is composed. It is implicit in the argument of a chemical-biological correlation that the chemical agent produces the biological effect by a one-stage process. Where there are many stages, and particularly where some involve alternative routes, no absolute correlation can be expected. Perhaps the most important outcome of the work on chemical carcinogenesis, and of the failure to find a close over-all correlation between chemical structure and carcinogenic activity, is the indication it affords that carcinogenesis is a more complex phenomenon than has hitherto been imagined.

368 REFERENCES 1. Badger, G. M. , Brit. J. Cancer, 2, 309(1948). 2. Barry, G. , Cook, J. W. , Haslewood, G. A. D. , Hewett, C. L. , Hieger, I., and Kennaway, E.L. , Proc. Roy. Soc. , B. 117. 318(1935). 3. Berenblum, I., J. Nat. Cancer Inst. , 10, 167(1949). 4. Berenblum, I. and Schoental, R. , Brit. J. Cancer, 1., 157 (1947). 5. Berenblum, I. and Shubik, P., Brit. J. Cancer, 1, 383(1947). 6. Boyland, E. and Brues, A.M. , Proc. Roy. Soc. , B. 122. 429 (1937). 7. Boyland, E. and Weigert, F. . Brit. Med. Bull. , 4, 254 (1947). 8. Cook, J.W., Brit. J. Nutrition, \. 245(1948). 9. Cook, J. W. , Hewett, C. L. and Hieger, I. , J. Chem. Soc. , 395 (1933). 10. Cook, J. W. , Haslewood, G. A. D. , Hewett, C. L. Hieger, I., Kennaway, E. L. , and Mayneord, W.V. , Am. J. Cancer, 29., .219 (1937). 11. Cook, J. W. and Kennaway, E. L. , Am. J. Cancer. 33, 50(1938). 12. Cook, J. W. and Kennaway, E. L. , Am. J. Cancer, 39., 381 and 521 (1940). 13. Edward. J. E. , J. Nat. Cancer Inst. , 2, 197(1941). 14. Fieser, L. F. , Am. J. Cancer, 34, 37(1938). 15. Fieser, L. F. , Univ. Pennsylvania Bicent. Conf. (1940). 16. Haddow. A., Harris, R.J.C. , Kon, G. A. R. , and Roe, E. M. F. , Phil. Trans., Roy. Soc., A. 241. 147 (1948). 17. Hartwell, J. L. , Survey of Compounds Which Have Been Tested for Carcinogenic Activity. Fed. Security Agency, U.S. Publ. Health Service (1941). 18. Heidelberger, C. , Kirk, M. R. , and Perkins, M.S., Cancer, I, 261 (1948). 19. Hieger, I., Brit. Med. Bull. ,4, 360(1-947). 20. Hueper, W. C. , Wiley, F. H. , and Wolfe, H. D. , J. Indust. Hyg. and Toxicol. . 20, 46 (1948). 21. Kennaway, E.L. and Cook, J. W. , Chem. and Industry, 1_0, 521 (1932). 22. Kinosita, R. , Yale J. Biol. and Med. , \2. 287(1940). 23. Kirby, A. H. M. , Cancer Research, 5, 683 (1945). 24. l.acassagne, A., C. r. Acad. Sciences, 195, 630(1932). 25. Miller, J.A. and Miller, C. E. , J. Exp. Med., 87. 139(1948). 26. Narat, J. K. , J. Cancer Res. , .9, 135 (1925).

369 27. Nettleship, A. , and Henshaw, P. S. , J. Nat. Cancer Inst. , 4, 309 (1943). 28. Nishiyama, Y. , Gann. 3J, 223 (1937) (quoted by Hartwell). 29. Orr, J.W. , Brit. Med. Bull. , 4. 385(1945). 30. Pullman, A., Assoc. franc etude da cancer, 3j, 120(1946). 31. Shear, M. J. , Am. J. Cancer, 29. 269(1937). 32. Shear, M. J. , Leiter, J. , and Perrault, A. , J. Nat. Cancer Inst. , 1., 303 (1940). 33. Turner, F.C., J. Nat. Cancer Inst. , Z, 81 (1941). 34. Wilson, R.H. , DeEds, F. , and Cox, A. J. , Cancer Research, ±, 595(1941). 35. Yoshida, T. , Trans. Jap. Path. Soc. . 2_3_, 636 (1933).

370 A CONSIDERATION OF CHEMICAL-BIOLOGICAL CORRELATION IN EXPERIMENTAL CANCER CHEMOTHERAPY C. Chester Stock Division of Experimental Chemotherapy Sloan-Kettering Institute for Cancer Research New York, New York In a brief introduction to a discussion on aspects of chemical-biological correlation in experimental cancer chemotherapy, there is time to do little more than highlight some of the observations in the hope that others will be stimulated to discuss more extensively the questions presented. When the symposium was being arranged, we expressed doubt that experimental cancer chemotherapy would at this time be a suitable part of a program on chemical-biological correlation; however, it may be worthwhile to review the present status of the field even though from published data only a few generalizations may be made with respect to correlations of chemical structure with antitumor activity. The introductory remarks have been collected under the title "A Consideration of Chemical-Biological Correlation in Experimental Cancer Chemotherapy", rather than a more imposing one such as "The Correlation of Anti-Tumor activity with Chemical Structure". This choice of title reflects the present status of experimental chemotherapy. While it has advanced considerably during the past fifteen years, there still is a paucity of active compounds and a complete lack of adequately effective substances. With only a few exceptions, 32, 6, 35, 43 there is an acute lack of collected data of a uniform nature on sufficiently large series of compounds to supply the basis for correlations. Many of the existing data have not been adequately covered in publications and may offer unknown possibilities for correlation. In the evaluation of data from various sources it must be recognized that the results from one search for anti-tumor activity cannot a priori be combined with data from another study in which different experimental tumors, different test conditions or differences in both have been employed. This would appear to be emphasizing the obvious but experience has indicated the need to point out the possible differences in effects observed with different tumors. Table 1 shows the differences in inhibition of the development of certain tumors by several compounds of various types. Table 2 illustrates differ- ences in the response of several strains of mouse leukemia to selected materials. Many of the general factors concerned with any chemotherapy study apply equally well to experimental cancer chemotherapy studies. 22,42,37 Thus, the choice of the tumor and test conditions in a chemotherapy screening study may be made in such a manner as to permit the discovery of activity with many materials or none at all. This is suggested by the data in Table 1 and by other observations. The rat tumors, for example, have appeared more susceptible to the inhibiting action of the nitrogen mustards than have the most susceptible mouse tumors in the spectrum of tumors examined. *9 if the ultimate goal of an experimental cancer chemotherapy program is the selection of materials with an adequate, differential adverse effect upon abnormal tissue under conditions of practical use, the ideal procedure would appear to be that which would reveal leads for selection of compounds for study in addition to pointing out the more effective compounds worthy of clinical trial. A program of this nature need not exclude the intelligent employment of m vitro techniques, such as tissue culture, which yield data within the limitations of the procedure. It is appropriate that the discussion on experimental cancer chemotherapy be held jointly with those on carcinogenesis and anti-metabolites. Inhibition of the growth of tumors by carcinogenic substances has been noted and proposed as a possible basis for the therapy of cancer. 18,2 The category of antimetabolites includes many compounds of diverse structures which are considered to act by blocking the utilization of the structurally related normal metabo- lites. 52 This represents one of the most interesting and fruitful approaches in cancer chemo- therapy studies, as illustrated by recent studies with desoxypyridoxine,44

371 h V as .*- •+. + ' + + 4 - «J o o 3 "-> C ' +• i .*- + •+. arUoma rt .*. + *. en c . HI il + .M rt ^* a Q, J OT u .ii « v S E a Ji o o0 o u < 2 *• i .*. i + ' j I * n U) in O 0 a 3 P. S H V i i +i i jn 3 0 2 5 r- 0 + U rt u « *.1 + +' t- ^^ sl en rg rg o in r^ 0 & 00 in o o in E J, .2 h C . — . u 0 S O i ^ X 2 C ^ .S Compoued tl — o ^ 1" E H 1 | 5 3.8 c .S i. >, -s 3h u u at i. - fl *J >. rt ! ?! | u w .B ^ ". i E s^ .5 « ^ 00 rg < 5 o

372 TABLE 2 CHEMOTHERAPEUTIC ACTIVITY OF COMPOUNDS AGAINST DIFFERENT STRAINS OF MOUSE LEUKEMIA Compound Representative per cent prolongation in survival time for leukemia strains Ak 1394 Ak 4 F T[3 line 15 F T8 line 291 Methyl bis(p-chlor- ethyljamine 100 40 - - Amethopterin 20 150 - - Urethane 100 20 100-200 25-50 Potassium arsenite 60 0 300-350 10-40 Benzene 100 0 40-60 0-25 Data derived from Burchenal4. 5, 6 anej from Kirschbaum. 24 anti-folic acids, 36, 10, 5,28, 31,48, 39,27 anti-riboflavins, 45 and the anti -purines, 2, 6-diamino- purine,4and 8-azaguanine. 23, 40, 14, 47, 26 As might have been anticipated, the usefulness of this approach has been limited in one or more instances by the metabolic requirements of the normal cells. 30,51,11,8 Analogs of folic acid, particularly the 4-amino-analogs, have been of interest for studies of their ability to act as antagonists of folic acid in bacteria, 12 and animals. 1 1 , I 3, 29, 51, 30 Some of the resulting information led to study of these compounds against leukemia, 10, 5,36 and solid tumor in animals. 28, 31, 48, 39 We have not observed, as yet, any published correlations of anti-bacterial with anti-tumor activity. It is to be expected that the anti-metabolite approach will be extended to include other anti-vitamins and anti-metabolites. This is being done in the case of anti-amino acids, with analogs of carbohydrate metabolism and with purines, pyrimidines, and pteridines for possible interference with nucleic acid metabolism. 38 jn the latter category, weak tumor inhibitions have been observed which suggest some correlations. 41 A few general conclusions on anti-tumor activity of other classes of compounds may be drawn from published data. The nitrogen mustards have been studied extensively in leukemia.' and solid tumors. 16, 19,38. it has become evident that most of the nitrogen mustards with two p-chlorethyl groups have been effective, whereas those with only a single group have been inactive. The presence of two reactive groups in compounds such as the nitrogen mustards and other alkylating agents has been postulated as a requirement for anti-tumor activity based upon the mechanism of chromosome damage;l 5, 19 however, there is evidence that this postulate may not be well founded. 1 In contrast to the general effectiveness of the bis -[3-chlorethyl amines against susceptible tumors are the data on carbamates. Against Sarcoma 180* and mouse leukemia^t 35 the activity of urethane has been unique among the carbamates and thiocarbamates thus far tested. With respect to inhibition of another tumor, Walker carcinosarcoma 256,Z0 and to other biolog- ical effects , such as hypnosis , 1? carcinogenesis,25 leukopenic action, 34 ancj rn it otic inhibition, ^ the carbamates are more generally active. Preliminary reports of the extent of activity have been made by Shear and his associates * 33 for various groups of compounds, including arsenicals. acridines, diphenyl ethyl amines, and podophyllin and colchicine derivatives. Further reports on correlations in these studies may be made in this symposium.

373 Recently we have had an opportunity to see relationships in the structure of certain steroids and their ability to inhibit the development of transplanted mouse lymphosarcomas. 50 Figure 1 shows the structure of cortisone acetate with the system of numbering indicated for »CH2OCCH3 1 1l = 0 0 COMPOUND E Figure 1 1 1-Dehydro-1 7-hydroxycorticosterone acetate reference ir» the use of Tables 3 and 4. Table 3 shows the structural relationships of the three steroids, compounds E, F, and A, thus far found active against the lymphosarcoma. Several inactive steroids have been included and 21-desoxycortisone, which remains questionable because limited results available indicate activity with ten times the dose level of cortisone acetate. 53,46 Some of the formulas of the compounds are presented in Figure 2. Table 4 shows a number of steroids which have been tested at five to ten times the dose level of cortisone with- out apparent effect upon the lymphosarcoma. In addition to the other variations in structure, it is important to note that all of the compounds lack the 11-oxygen function. Thus far the steroids which have inhibited the mouse lymphosarcoma have possessed in common an 11-oxygen group, a 2O-keto group and a 3-keto group with the ^4 unsaturation. The 21-hydroxy group is important from the quantitative aspect, if it is not essential.

374 H2-C-OH -OH DEHYDRO E Active 21-DESOXY E ? Active DEHYDRO E Inactive CH2-OH CH2-OH COMPOUND F Active COMPOUND A Active COMPOUND S Inactive Figure 2 Many of the points have been referred to briefly because of the limitations in time and with the understanding that some of the investigators, who have made contributions in this field, are here to present their viewpoints. It is to be hoped also that some of the panel members will be able to present from their unpublished data correlations in areas of interest which we have not discussed. Although the data at hand may permit only a few limited correlations of structure with anti-tumor activity, it can be anticipated that the increasing efforts devoted to experimental cancer chemotherapy studies will provide within the next few years chemical-biological correla- tions as important as those discussed in other meetings of this symposium.

375 TABLE 3 STRUCTURAL RELATIONSHIPS OF A FEW STEROIDS WITH RESPECT TO INHIBITION OF MOUSE LYMPHOSARCOMA 1 1 -Keto 11 -OH 20 Keto 3-OH 17-OH 21 OH Activity Keto Cortisone + Compound F + + * * + - * * Compound A (17-Desoxy- + cortisone) * - * + '- + - 21 -Desoxycortisone ? * - - + + + + Dihydrocortisone 1 1 -Keto pregnanolone Compound S (11 -Desoxy- cortisone) - - * * - - * Desoxycorticosterone (1 1 , 1 7-Desoxycortisone)

37h TABLE 4 ASPECTS OF STRUCTURE OF STEROIDS INACTIVE AGAINST MOUSE LYMPHOSARCOMA 1 1-Oxy 20-Keto 3-OH 17-OH 20-OH 21 -OH 3-Keto - + - - + - - Testoste rone - - i + - - - Progesterone A -Pregnenolone _ + + . - - + Desoxycorticosterone 21 -OH-Pregnenolone - •f - .f - - 17a OH-Progesterone (A4) 17a OH-Progesterone (A ) 17p OH-Progesterone 1 7a OH-Alloprcgnanolone (Cmpd. L .f •f - * * - * - * 17a, 21 -OH-Progesterone (Cmpd.! - 4. - - * * : I 7a-Triolone 17p-Triolone

377 BIBLIOGRAPHY 1. Biesele, J. J. , Philips, F. S. , Thiersch, J. B. , Burchenal. J. H. , Buckley, S. M. and Stock. C.C. , Nature, 166. 1112(1950). 2. Boyland, Eric, Biochimica et Biophysica Acta, 4, 293 (1950). 3. Buckley, S. M. , et al. , Unpublished data. 4. Burchenal, J. H. , Bendich, A., Brown, G. B. , Elion, G. B. , Hitchings, G. H. , Rhoads, C. P. , and Stock, C. C. , Cancer, 2, 119(1949). 5. Burchenal, J. H. , Burchenal, J. R. , Kushida, M. N. , Johnston, S. F. , and Williams, B. S. , Cancer. 2, 1 13 (1949). 6. Burchenal, J. H. , Lester, R. A. , Riley, J. B. . and Rhoads, C. P. , Cancer, .1, 399(1948). 7. Burchenal, J. H. , and Riley, J. B. , Cancer Research, 2, 553 (1949). 8. Cartwright, G. E. , Palmer, J. E. , Hitchings, G. H. , Elion, G. B. , Gunn, F. D. and Wintrobe, M. M. , J. of Lab. and Clin. Med. , 35, 518 (1950). 9. Cornman, I., J. Nat. Cancer Inst. , 1_0, 1123(1950). 10. Farber, S. , Diamond, L. K. , Mercer, R. D. , Sylvester, R. F. , Jr., and Wolf, J. A. , New England J. Med. , 238, 787 (1948). 11. Ferguson, F. C. , Jr., Thiersch, J. B. and Philips, F. S. , J. Pharm. and Exp. Ther. , 98, 293 (1950). 12. Franklin, A. L. , Belt, M. , Stokstad, E.L.R., and Jukes, T. H. , J. Biol. Chem. , 177, 621 (1949). 13. Franklin, A. L. , Stokstad, E. L. R. , and Jukes, T. H. , Proc. Soc. Exp. Biol. and Med. , 67, 398 (1948). 14. Gellhorn, A. , Engelman, M. , Shapiro, D. , Graff, S. , and Gillespie, H. , Cancer Research, 10, 170 (1950). 15. Goldacre, R.J. , Loveless, A. and Ross, W. C. J. , Nature, 163, 667(1949). 16. Goldin, A. , Noe, H. A. , Landing, B. H. , Goldberg, B. , and Fugmann, R. A. , Acta Unio Internal. Contra Cancrum, j6, 501 (1949). 17. Goodman, Louis and Gilman, Alfred, The Pharmacological Basis of Therapeutics, New York, 1941. 18. Haddow,. Alexander, Nature, 136, 868(1935). 19. Haddow. A., Kon, G. A. R. , and Ross, W. C. J. , Nature, 162. 824(1948). 20. Haddow, Alexande r and Saxton, W. A. , Nature, 157, 500(1946). 21. Hartwell, Jonathan, L. , and Shear, M. J. , Am. Assoc. Cancer Research Meeting, May 17, 1947. 22. Karnofsky, David A. , New England J. Med. , 239, 226, 260, 299 (1948).

378 23. Kidder, G.W., Dewey, V. C. , Parks, R. E. , Jr., and Woodside, G. L. , Science, 109. 511 (1949). 24. Kirschbaum, Arthur, Judd, Sister Theresita, Lu, Chen Shan, Engstrom, Ruby M. , and Mixer, Harry, Proc. Soc. Exp. Biol. and Med. , 68, 377 (1948). 25. Larsen, C. D. , J. Nat. Cancer Inst. , 5, 35(1948). 26. Law, L. W. , Cancer Research, 10, 186(1950). 27. Little, P. A. , Sampath, A. and SubbaRow, Y. , J. Lab. and Clin. Med. , 33, 1144 (1948). 28. Moore, A. E. , Stock, C. C. , Sugiura, K. , andRhoads, C.P., Proc. Soc. Exp. Biol. and Med. , 70, 396 (1949). 29. Oleson, J. J. , Hutchings, B. L. and SubbaRow, Y. , J. Biol. Chem. , 175. 359(1948). 30. Philips, Fredericks., and Thiersch, J. B. , Proc. Soc. Exp. Biol. and Med. , 72, 401 (1949). 31. Schoenbach, E. B. , Goldin, A. , Goldberg, G. , and Ortega, L. G. , Cancer, 2, 57 (1949). 32. Shear, M. J. , Downing, V., Hartwell, J. L. , Leiter, J. , MacCardle, R. C. , Perrault, A., and Vivian, D. L. , Cancer Research, 9, 625 (1949). 33. Shear, M. J. , Hartwell, Jonathan, L. , Peters, Virginia B. , Dalton, Albert J. , Dunn, Thelma B. , Hauschka, Theodore S. , Diller, Irene Corey, McConnell, Jane Royle, Oakey, R. , Reimann, Stanley P. , Rees, Charles W. , Beck, Lyle, V. and Holloman, A. Leon. , A. A. A. S. , 1947. 34. Skipper, Howard E. , Bryan, Carl E. , Riser, William H. , Jr. , Welty, Margaret and Stelzenmuller, Anne, J. Nat. Cancer Inst. , .2, 77 (1948). 35. Skipper, Howard E. , and Bryan, C. E. , J. Nat. Cancer Inst. , 3, 391 (1949). 36. Skipper, Howard E. , Edwards, P. C. , Bryan, C. E. , Chapman, J. B. , Bell, Martella and Hutchison, O. S. , Cancer, 1, 348 (1950). 37. Stock, C. Chester, Am. J. of Med. , 8, 658 (1950). 38. Stock, C. Chester, et al. , Unpublished data. 39. Stock, C. Chester, Biesele, John J. , Burchenal, Joseph H. , Karnofsky, David A. , Moore, Alice E. and Sugiura, Kanematsu, N. Y. Academy of Science Conference on Develop- ment and Use of Antimetabolites, Feb. 12, 1949: 40. Stock, C. C. , Cavalieri, L. F. , Hitchings, G. H. , and Buckley, S. M. , Proc. Soc. Exp. Biol. and Med. , 72, 565 (1949). 41. Stock, C. Chester, Hitchings, G. H. , et al. , Unpublished data. 42. Stock, C. Chester and Rhoads, C.P. , Chapter in Evaluation of Chemotherapeutic Agents, Columbia U. Press, 1949. 43. Stock, C. Chester, Sugiura, Kanematsu, Moore, Alice E. , and Rhoads, C.P. , Cancer Research, 9, 598 (1949). 44. Stoerk, H. C. , J. Biol. Chem., 171. 437(1947). 45. Stoerk, H. C. , and Emerson, G. A. , Proc. Soc. Exp. Biol. and Med. , 70, 703(1949). 46. Sugiura, K. , et al. , Unpublished data.

379 47. Sugiura. Kanematsu, Hitchings, G. H. , Cavalieri, L. F. , and Stock, C. Chester, Cancer Research, 10, 178 (1950). 48. Sugiura, K. , Moore, A. E. , and Stock, C. C. , Cancer, 2, 491 (1949). 49. Sugiura, Kanematsu, and Stock, C. Chester, Cancer Research, 10, 244(1950). 50. Sugiura, Kanematsu, Stock, C. Chester, Dobriner, Konrad and Rhoads, C.P., Cancer Research, 1_0, 244 (1950). 51. Thiersch, J. B. and Philips, F. S. , Proc. Soc. Exp. Biol. and Med. , 7_1. .484 (1949). 52. Wooley. D. W. , Physiol. Rev., 27, 308(1947). 53. Wooley, George W. , Unpublished data.

380 ANTIMETABOLITES Bernard D. Davis U. S. Public Health Service Tuberculosis Research Laboratory 41 1 East 69th Street New York 21, New York Mr. Chairman, Ladies and Gentlemen: I should like to speak on the application of bacterial mutants to several problems that concern antimetabolites. In the first place, certain antimetabolites have given rise to the concept of competitive inhibition which has played such an important role in pharmacology that it might be called the cornerstone of much of the work being considered here today. We propose to extend that concept to the normal physiology of the growing cell, using it as the basis of an integrative mechanism for which we believe adequate, though indirect, evidence is now available. After discussing this topic, a moment will be spent in comparing the results provided by inhibition analysis and by mutants in two types of studies: those designed to determine paths of biosynthesis and those designed to determine the site of action of an inhibitor. Finally, possible application of mutants to the development of new chemotherapeutics will be noted. I am sure you are all familiar with the work which was started in 1941 when Beadle and Tatum isolated various Neurospora mutants, each of which lacked the activity of the enzyme of one essential biosynthetic reaction. These mutants grow only if supplied with the product of the blocked reaction, or a derivative of that product. These organisms are elegant tools for metabolic study, as it is possible to eliminate, with complete specificity, any one of a variety of enzymes. Similar biochemical mutants can be obtained with bacteria. Our interest in this field arose rather accidentally through an interest in chemotherapy which led to the development of an efficient method of mutant isolation, based on a unique property of penicillin: namely, penicillin sterilizes bacteria only when they are growing. The colon bacillus grows on a minimal medium containing only glucose and salts. It, therefore, synthesizes all its amino acids and other components from these materials. The mutants, however, by definition, are cells which cannot grow on this medium. When a large population of bacteria containing a few mutants is exposed to penicillin in minimal medium, the predominant parent cells grow, and are promptly sterilized by the penicillin. The mutants, on the other hand, do not grow; hence they survive. With this simple trick a wide variety of interesting mutants were soon isolated. These have been employed in studies of certain paths of biosynthesis which will be briefly described in order to illustrate the techniques involved. It is known that in a variety of cells, from bacteria to mammalian liver, ornithine is converted to citrulline which in turn is converted to arginine. We have ^. coli mutants blocked before each of these three compounds. In this series, as in many others, a mutant, trying vainly to synthesize the required compound, accumulates in the medium the precursor of the blocked reaction, normally present only in traces. The technique of syntrophism (cross-feeding) can be used to permit very simple demonstration, without any elaborate biochemical techniques, of precursor accumulation. The arginine-requiring mutant spills out its precursor, citrulline; by diffusion through the agar, this gives rise to a gradient of stimulation of the neighboring streaks which can grow on this compound. In turn, the citrulline-requiring mutant feeds the ornithineless mutant. These observations are illustrated in the slides. This technique is now being used to analyze unknown paths of biosynthesis. What concerns us at present, however, is the broader problem of trying to find how the various single reactions in the growing cell are so beautifully integrated. E^. coli is exceedingly economical in its pattern of synthesis; the normal cell excretes no significant amount of any amino acid. This fact was easy to ascertain by the use of appropriate mutants. The cell therefore synthesizes these components in exact proportion to their relative requirements, and hence must possess efficient

381 regulatory mechanisms. A possible clue to their nature arose when we observed that a block in synthesis of one amino acid causes excretion not only of a precursor, as was illustrated above, but often also of another amino acid, usually one with a close structural relationship. It seemed likely that an analysis of this breakdown in the regulatory mechanism would throw light on the nature of the mechanism, for medical investigators well know that physiological mechanisms are often first recognized when distorted by pathological conditions. An analysis was consequently carried out - in a much less logical and sequential fashion, I fear, than will be conveyed in the following account. The key was furnished by a phenomenon discovered by Bonner in a mutant of Neurospora requiring a closely related pair of amino acids, isoleucine and valine. This double requirement results from a single genetic block. It could be shown that the block interrupts only isoleucine synthesis, but as a result a precursor accumulates which competitively inhibits an enzyme concerned with a similar precursor of valine. The o-keto analogue of isoleucine was originally thought to be the inhibitory precursor, but Adelberg and Tatum subsequently showed that the mutant with the double requirement accumulates primarily the a, p-dihydroxy acid rather than the keto acid. The principle of internal inhibition, however, remains unchanged by this later development. We have here an example of an internal or physiological block secondary to a genetic block. What is the significance of this internal inhibition for the cell? A rather teleological altitude toward general physiology gives rise to the following considerations. If an excess of a given metabolite can completely block a reaction, then the normal amount of that metabolite might govern that reaction, and complete absence might permit excessive synthesis by removal of the physiological brake. This hypothesis can be tested. We fortunately have not only E. coli mutants of the type described above, with a double requirement for isoleucine and valine; there is also one blocked later in isoleucine synthesis, which requires only this amino acid, and another blocked very early, which responds to either the 4-carbon compound a-aminobutyric acid (AAB) or the 6- carbon isoleucine. According to the notion outlined above, the mutant which is blocked very early and can respond to AAB should fail to make the normal governor; hence, it should excrete valine. Furthermore, there is a second, more critical test of this hypothesis. An excess of AAB should restore to the system the governing intermediate which is absent on account of the genetic block, and hence should abolish valine excretion. Similarly, if the biosynthetic reactions are reversible, an excess of isoleucine should have the same effect. The culture plates illustrated on the slides show that these expectations are fulfilled. The AAB-requiring mutant, growing to a limited extent on a small amount of isoleucine or of AAB, feeds a valineless mutant heavily. Furthermore, an excess of either isoleucine or AAB completely abolishes excretion of valine. The control tests required for these conclusions have been carried out: it has been shown that the compound excreted is valine, and it is abolition of excretion, and not of response, which is caused by the excess of isoleucine or AAB. This rather complicated explanation of amino acid excretion might seem to rest on a wobbly foundation, were it to rely only on this pair of mutants. Identical behavior, however, has been observed with another pair of related amino acids, tyrosine and phenylalanine. Several phenylalanine-requiring mutants excrete tyrosine; reciprocally, tyrosineless T9 excretes phenylalanine, but T8, another tyrosineless mutant with a different genetic block, fails to excrete phenylalanine. We would postulate that a metabolite in the sequence of synthesis of either compound is a governor of the other sequence. Excretion would be explained by an early block, which prevents formation of the governing compound. On the other hand, the tyrosine mutant which fails to excrete phenylalanine could be explained by a later block, which does not interfere with the production of this governing compound. Just as in the previous series, an excess of phenylalanine completely abolishes excretion of tyrosine, and an excess of phenyl- pyruvic acid, a known precursor, has the same effect. Similarly with the tyrosineless mutant, an excess of its growth requirement completely abolishes excretion of phenylalanine. In all these cases the compounds excreted have been identified by paper chromatography as well as microbiological response, and appropriate controls have shown that the excess of the growth requirement abolishes excretion rather than response.

382 While other theories might explain the fact of excretion, it would be difficult to explain in any other way its abolition by excess of the growth factor. Though the evidence for this theory is indirect, its consistency with the varied data is impressive. The cell is not a bag of independent enzymes, but rather appears to control, in at least some cases, not only the relative number of molecules of each enzyme per cell, but also the activity per enzyme molecule. This idea, though hardly novel, is not frequently stressed. To it we have added a specific control mechanism which involves a dual role for certain metabolites: substrate for one enzyme and governor of another. This concept is an extension to physiology of the principle of competitive inhibition, so well established in pharmacology. Such integrative mechanisms are surely not restricted to bacteria but undoubtedly enter into normal and disturbed growth regulation in animal and plant cells, including differentiation, regeneration, and neoplasia. And now we might discuss briefly the use of mutants in connection with studies employing growth inhibitors. Following the development of the concept of competitive inhibition of essential metabolites by structural analogues, growth inhibitors have been rather widely used in recent years to analyze paths of biosynthesis. This methodology, called inhibition analysis, involves the assumption that a competitive antagonist is a precursor, and a non-competitive antagonist is a product, of the affected reaction. This assumption is not always safe; Hitchings has shown that this interpretation of certain data would lead to the absurd conclusion that bromouracil is a normal metabolite. We have several examples that further illustrate the dangers inherent in inhibition analysis. Beerstecher and Shive, using this approach with 1C. coli. concluded that tryptophan can be converted to phenylalanine, and the latter is a precursor of tyrosine; furthermore, phenyl- pyruvic acid did not act as a precursor of phenylalanine. All three of these conclusions are in conflict with those derived from studies on mutants. For in addition to the strains noted above, we have mutants blocked early in aromatic synthesis which require all three aromatic amino acids; others, blocked even earlier, require' in addition PABA and, in some cases, a previously unknown bacterial vitamin. These observations, supplemented by others, conclusively demon- strate that these three amino acids are derived from a common precursor; no one is a normal precursor of any other. Furthermore, the absolute requirement of these strains for all three amino acids implies that these compounds are not interconvertible in the mutants even by reversal of normal biosynthetic paths; one would, therefore, not expect them to be interconvertible in the parent wild type either, since the genetic blocks in these mutants do not lie in the path of the potential reversed syntheses. Finally, phenylpyruvic acid is used practically as well as phenylalanine by all of these strains. As another example, in our laboratory Dr. Werner Maas, using both inhibition analysis and mutants, has shown that D-serine inhibits the growth of a wild type strain of £. coli by interfering with the conversion of p-alanine to pantothenate ; the inhibition is overcome competi- tively by the former compound, and non-competitively by the latter. Glycine also overcomes the inhibition competitively, and would, therefore, be interpreted, according to the principles of inhibition analysis, as a precursor of p-alanine. However, glycine cannot satisfy the growth requirement of a mutant blocked in the synthesis of p-alanine, and yet it does reverse the inhibitory effect of D-serine on the growth of this mutant. It is clear that glycine does not act simply by promoting the synthesis of p-alanine; a more subtle mechanism must be present. These several conflicts suggest that in some cases inhibition antagonisms involve more complex mechanisms than those that have been considered in studying paths of biosynthesis by this approach. The above discussion has been concerned with the limitations of growth inhibitors, compared with mutants, in analyzing paths of biosynthesis. In addition, mutants can throw light on the mode of action of inhibitors themselves. As was stated above, and will be documented elsewhere, the use of mutants has considerably strengthened the evidence for the site of action of D-serine. Furthermore, the combined study of bacterial mutants and growth inhibitors leads naturally to an investigation of the biochemical mechanism underlying drug resistance. The D-serine-pantothenate system appears in some ways to be more suitable for an attack on this problem than the clinically effective chemotherapeutics. The compounds involved possess simple chemical structures ; the metabolic path involved is better understood than most others ; the affected reaction can be studied with non-growing cells; mutants with a high degree of resistance to D-serine can be obtained easily. Preliminary experiments have ruled out several mechanisms that have been widely assumed for drug resistance: it does not involve an increase in the production of either pantothenate or p-alanine, nor does it involve formation of pantothenate via

383 an alternate metabolic pathway bypassing p-alanine. Among the remaining possibilities, D-serine may be prevented from reaching its site of action, or the previously sensitive enzyme may have been changed, qualitatively or quantitatively, in such a manner that it can now function adequately in the presence of the inhibitor. This problem is under investigation. Finally, mutants can be useful in the development of new chemotherapeutics. The system- atic synthesis of metabolite analogues has provided a host of cytotoxic agents that are effective m vitro, but has been disappointing in its failure to yield chemotherapeutics that are useful in vivo. It is now well recognized that this failure can be accounted for by the requirement of the animal host for the very metabolites that have served as models. This situation arises from the fact that the known bacterial metabolites are nearly all mammalian metabolites as well. But this fact does not depend solely on the unity of biochemistry in nature; it depends also on the unity of purpose of biochemists. In general, microbial nutrition has served as a handmaiden (or investigators primarily interested in mammalian metabolism. But metabolites peculiar to microorganisms do exist. The success of sulfonamides depends on the fact that PABA as such is not used as a metabolite by animals, whereas the PABA-containing folic acid of the animal is not used, at least when supplied externally, by many bacteria. It follows that we should search further for specifically microbial metabolites. For some time it has been clear that mutants might be used in the search not only for precursors of known metabolites, but also for new growth factors. We have recently been studying a mutant blocked early in aromatic synthesis, which requires tyrosine, phenylalanine, tryptophan, and PABA; for rapid growth it requires further a fifth substance, present, like PABA, in the filtrate of the wild type E. coli. This substance has been tentatively identified as a simple aromatic compound, active in trace amounts. From the point of view of this audience, it is particularly significant that this substance, like PABA, is present in relatively large amounts in yeast extract but not in liver extract. It therefore offers real promise as a model for chemo- therapeutic synthesis.

384 CARCINOGENESIS L. F. Fieser Harvard University Cambridge, Massachusetts It is a pleasure to see here so many of the people with whom I collaborated in the work you mention, including Dr. Shear* Dr. John Wood. Dr. Arnold Seligman, Dr. Fred Novello; I guess Dr. Konrad Dobriner can be included as well. I would not attempt to try to find a common factor between a blow on the head and chloro- form, as suggested so aptly by Dr. Berenblum. As a matter of fact, I think it is asking a little bit too much to try to find a common factor between carcinogenetic hydrocarbons, the azo dyes, the estrogenic substances, and so on. It seems to me, if one could understand a little about the way in which any series of compounds of these different groups functions, we would be well ahead. At the time I stopped work on the problem of the carcinogenic hydrocarbons, it seemed to me that the very pronounced chemical activity exhibited by some of the most potent ones was probably concerned in the reaction of carcinogenesis, as some of the hydrocarbons exhibit chemical activity of a specific type which far surpasses that of related, less carcinogenic hydro- carbons. Then, in a summarizing lecture at the University of Pennsylvania celebration, 1 postulated that the carcinogen reacts with sulfhydryl groups of a protein, and that this is the reaction which is responsible for initiation of malignancy. That was a nice guess at the time, but I am sure it has been ruled out by now, because the hydrocarbons just do not react with proteins in a chemical way. Careful examination of the postulated correlation between structure and chemical reactivity or carcinogenic activity shows that there are rough parallelisms but that no exact relationship exists; any attempt at exact relationship seems to fall down. So chemical reactivity is not quite the whole story. As I see it now, a hydrocarbon - I will stick to hydrocarbons - to be a potent carcinogen, has to have various properties. One property is highly developed chemical reactivity of a certain type; probably another factor is the shape of the molecule. It looks as though the hydrocarbon has to have a fairly flat surface in order to be effective. A methyl group attached to the benz- anthracene nucleus may permit carcinogenic activity, whereas the homolog with an ethyl or propyl or butyl group begins to lose that property. That may be because only the methyl group can stick out flush with the flat surface of the rmg system. Then, perhaps, the over-all size is important. You could pick out a hydrocarbon such as aceanthrene, which has the same kind of chemical reactivity as methylcholanthrene but still lacks carcinogenic potency. It may be that the molecule is a little too small to give the right distribu- tion characteristics. I am impressed with the idea that, if we are ever going to understand the relationship between chemical structure and any kind of biological activity, we are going to have to take into account another factor, and that probably is the distribution characterisistic of the molecule, distribution particularly between lipophilic and hydrophilic phases. A compound which has all the other properties required but does not have the right distribution characteristics may be outside the field of biologically active compounds. What kind of interaction with a cell constituent involved in gene determination could occur with a hydrocarbon of the right chemical activity, shape, size and distribution characteristics? l do not think it is a chemical reaction, as I said before. The nearest guess at present is that it

385 is a process of adsorption, or of some form of complex formation, and the requirement for specific chemical reactivity is merely to give the molecule sufficient pep so that it can form suitable complexes. It seems to me possible that the phenomenon of carcinogenesis may have some relationship to the ability of the carcinogenic hydrocarbons to form complexes with poly- nitro compounds. The trinitrobenzene derivatives of methylcholanthrene, benzpyrene, and other particularly potent carcinogens are deeply colored, whereas complexes of less reactive hydro- carbons are less highly colored. The color probably is just one indication of stability of the complex and of the reactivity of the components. Possibly forces similar to those involved in the formation of these polynitro complexes are involved in carcinogenesis. So, as a way of moving ahead in the future, it seems to me - just a guess at the moment - that any way of setting up experimental methods of detecting or measuring complex formation or adsorbability of the agents might shed new light on the general problem.

386 CANCER CHEMOTHERAPY Howard E. Skipper Southern Research Institute Birmingham, Alabama I have been told that, after about two minutes. Dr. Shear will lower the boom, so I arfi certainly going to get through with my short discussion of carbamates in chemotherapy of leukemia within that time. Our work with carbamates has shown several things with regard to the relation between structure and biological activity (anti-leukemic activity). Certainly it is mostly negative in this respect since we have found that ethyl carbamate (urethan) is the only one of a rather large series of this type compound which is very effective in prolonging life span in mouse leukemia. However, when one compares the anti-leukemic activity and the carcinogenic activity of carbamates (work by Dr. Larson of the National Cancer Institute), we see a very nice correlation; ethyl carbamate is the most active by far in either of these respects. There seems to be little correlation between leukopenic activity (ability to depress the total white blood count) of the carbamates ill normal mice and anti-leukemic activity of this series. Many carbamates have been found to be leukopenic while urethan is the anti-leukemic carbamate. There does seem to be a very good correlation between leukopenic action in mice and mitotic poisoning activity of carbamates in such organisms as the sea urchin egg (work by Dr. Ivor Cornman, George Washington University). We have been most interested in trying to find out why urethan is so specific with regards to anti-leukemic activity. In certain mechanism studies we have labeled the carbonyl and the methylene carbon atoms of the urethan molecule: XOCH2CH3 -OC*H2CH3 O = C* H O = C'C ^H Carbonyl -labeled urethan Methylene -labeled urethan and have compared the distribution of the active atoms in normal and neoplastic mice as well as excretion and exhalation. It appears quite clear that urethan is broken down in the body to carbon dioxide, ethyl alcohol, and ammonia. We have found that cancerous mice do not degrade the molecule nearly so rapidly as do normal mice. In an isolated system, living sperm from the sea cucumber, which is largely living desoxyribose nucleoprotein, Dr. Cornman and I have observed that carbon 14 from carbonyl- labeled and methylene -labeled urethan are fixed to a higher degree than labeled CO2 or methylene- labeled ethyl alcohol, the respective labeled hydrolysis products of the two labeled urethans. This is not the case in sea urchin eggs where the cell is fairly low in nucleoprotein. We think that this result may be suggestive of the mechanism of the carcinogenic and anti-leukemic actions of urethan.

387 CARCINOGENESIS H. P. Rusch University of Wisconsin Madison, Wisconsin I have a few slides which illustrate some of the relationships between the chemical structures of certain compounds and their carcinogenic activities. These studies have been carried out by Drs. J. A. and E.G. Miller of our department, and the data are published in Cancer Research, J, 227; %, 504, 652 and J. Exp. Med. , 87, 139. First, I will mention some studies on certain aminoazo dyes. These compounds induce liver tumors when they are fed to rats for several months. As shown by Haddow and Kon, replacement of the azo linkage by an ethylene linkage to give 4-dimethylaminostilbene results in a carcinogen which is, as far as the liver is concerned, a slower acting carcinogen than the aminoazo dyes. relotive octivity -CH—CH- weak(Haddow and Kon) -CH=N- 0 -N=CH- 0 Slide 1 Substitution of -CH=N-, -N=CH-, or -C-NH- for the azo linkage results in inactive compounds. The second slide illustrates the effect of alterations in the substituents on the amino group on the potency of the aminoazo dyes. 4-Dimethylaminoazobenzene is assigned an activity of six, and 4-monomethylaminoazobenzene with only one methyl substituent and 4-ethylmethylaminoazo- benzene with one ethyl and one methyl substituent also have activities of six. Thus, in order to

388 relative activity -N/CH2CH3 Slide 2 be carcinogenic the compounds must have at least one N-methyl group. However, as illustrated by the non-carcinogenic dyes 4-p-hydroxyethylmethyl- and 4-benzylmethylaminoazobenzene, the second substituent on the ami no group can also affect the potency of the compound. Introduction of substituents into the rings also strongly affects the carcinogenic activities of the aminoazo dyes. Either a hydroxy or trifluoromethyl group in any of the prime positions abolishes the activity of 4-dimethylaminoazobenzene. With other substituents the activity of the compound depends on the position of the group. Thus the 3'-methyl derivative of 4-dimethyl- aminoazobenzene is about twice as active as the parent compound while the 2'-methyl derivative is only one-third to one-half as active and the 4'-methyl derivative is very weak. 2- and 3- methyl-4-dimethylaminoazobenzene are both inactive under our conditions. Similarly, intro- duction of a chloro or nitro group in the 3' position results in a more active compound than if the same group is inserted into the 2' or 4' position. On the other hand, compounds containing fluorine in any one of the prime positions are more active than the unsubstituted dye. More recent work has shown that the 2-fluoro and 2', 41 -difluoro derivatives of 4-dimethylaminoazo- benzene are also about twice as active as 4-dimethylaminoazobenzene. Since substituents in the 2, 2', or 4' positions should hinder a benzidine rearrangement of the corresponding hydrazo derivatives, the high activities of these fluorinated derivatives largely eliminate the possibility that such a rearrangement is involved in the mechanism by which these dyes induce the carcino- genic process; such a theory was proposed several years ago by certain British experimenters.

389 2 3 Relolive Activities (unsubst1tuted dye = 61 4' 3' 2' 2 3 H0- 0 0 0 0 0-12 2 3 0 0 NOg- 0 9 3 Cl- i-e 5-6 2 F- 10 10 7 Slide 3 Recent studies on the effects of various alterations in the structure of 2-acetylamino- fluorene on its carcinogenic activity are presented in the fourth slide. Replacing the -CH2-bridge in 2-acetylaminofluorene by -S- as in 3-acetylaminodibenzothiophene did not alter the carcino- TABLE I TX UnONOGEMC AC1MTES OF ANALOGS OF J.ACETYLAMNOFLUOnENE wiTH THE GC'CRAL FORMULA Of-f^lt^ - .HP* rj^i^:^wT" ™r / '.^ /"--*. H -vy: •.ra 4 a 7* 3 6 0 1 O 0 0 2 0 5 0 *'•ET 8 7 0 • 0 i 0 1 0 ' i 0 0 0 0 0 i 0 0 0 -." 0 1 0 i 0 0 0 0 0 0 0 7 0 0 0 0 0 0 il3l W0, 3»H W8l W6l 7t7l 6Ml 717l _T Ml 7l7} KD S49l 5l2l 7l71 4(4l 6l6l 0 0 not am al0l W^ Kit SOl MW rnn iM tut, '..•Ml ..itam vt . - .- - . . .' i*11 . ~~- * Slide 4

390 genicity of the molecule for either mammary gland or ear duct tissue. Substitution of -S- for the O -CHZ as in 3-acetylaminodibenzothiophene-5-oxide greatly lowered the activity towards these two tissues, while insertion of an -O- bridge as in 3-acetylaminodibenzofuran only partially dimin- ished the activity of the molecule in these respects. Unlike 2-acetylaminofluorene, however, none of these three compounds had any carcinogenic activity towards the rat liver. Further studies will be necessary to determine what features these various carcinogenic processes have in common and how these relatively small variations in structure interfere with or promote the reactions which lead to tumor formation.

391 CANCER CHEMOTHERAPY R. B. Angler Lederle Laboratories American Cyanamid Company New York, New York I feel, somehow, out of place up here, being an organic chemist, but I presume that Dr. Farber wanted to make it official, to get these formulas on the screen so that everybody would know what he is talking about. A rather large number of analogs of folic acid have been prepared. The basic structure we have up here, which includes the pteridine nucleus, the methylene linkage to the benzoic acid portion of the molecule and the PABA residue. We see there are four different substituents on here. These four different groups have been changed in a number of ways to produce the various analogs which are being described. 1 do not expect Dr. Farber will talk about all of these but, in order to cover the ones he will mention, I am putting these all on. Number one is merely a change of the glutamic acid to aspartic acid on the end of the molecule. Following, is the dj aspartic acid derivative. Of course, in all other cases, the other substituents are the ones you will find in folic acid itself. This is your 9-methyl pteroyl glutamic acid, in which you have a methyl group on the bridge carbon atom. The antagonist activities are shown here, also. We will pass them by. These involve some of the more potent antagonists which have been discussed. In amino- pterin we see merely the change of the R-group in the 4-position to an amino group. That comes under the class of isosteres, which were discussed earlier. This compound becomes, then, an extremely potent antagonist, microbiologically as well as in the animal. Amino anfol, has the 4-amine but, in the place of the glutamic acid out here, it has the aspartic acid. A-methopterin which is even more potent than is aminopterin, has an amino group in the 4-position, replacing the OH and a methyl group on the N^ position. A-ninopterin has two methyl groups, one on the 9-position and one on the NlO position. Adenopterin also has an amino group in the R-position, and two methyl groups, one in the 9- and one in the 10-position. Aminoteropterin has three glutamic acids out here, instead of just one. They are linked in the gamma position; each linkage is gamma. Also aminoteropterin has an amine in the 4-position. Amino alanfol has an amino group in the 4-position while the glutamic acid residue of folic acid has been replaced by .dj-alanine. And lastly, amino treofol has an amino group in the 4-position while the glutamic acid residue has been replaced by dl-threonine.

392 CANCER CHEMOTHERAPY Sidney Farber Children's Hospital Boston, Massachusetts Mr. Moderator, one comment, first, about the chemical structures which Dr. Angier just showed you. Our studies were made, first, with two folic acid antagonists, without the amino in the 4-position. They were made by the American Cyanamid Co. group of chemists; and both had anti-leukemic activity, but the activity was weak, and no prolonged or real remissions were attained. There is no close correlation whatsoever between the action on mouse leukemia and the action on man. All of the excellent results have been attained with compounds with the amino in the 4-position, and every compound with the amino in the 4-position has been toxic. I think we can state, first of all, that the initial impression we had concerning a correla- tion between toxicity and carcinolytic action does not obtain. The same mechanism may be working in both instances, but it is perfectly possible to obtain clinical results without bringing about any important toxic changes in the patient. To summarize clinical results very briefly, our group has had the opportunity to treat and care for some two hundred children with acute leukemia in the last three and one-half years, with the various folic acid antagonists. An over-all remission rate of approximately 50 per cent was obtained for this group. In one instance with the 4-aminopteropterin, to which Dr. Angier referred, the remission rate has reached 66 per cent for a period of time. Prolongation of life up to twenty-two months has been achieved. In all instances, after remissions have been attained, resistance to the treatment has occurred eventually and, in all instances, death has occurred. As the moderator has pointed out, no cure has, therefore, been attained in acute leukemia. We have seen, however, for the first time, a carcinolytic effect which has been produced by no other series of compounds. This holds true for aminopterin, for the 9-methyl, the 10-methyl, the 9, 10-dimethyl, the 4-aminoteropterin and for other compounds to which reference has been made. No one is really that much better than another in this series, so that no strong preference can be stated at this time. It is of considerable interest that, when empirical trial was made with the action of these compounds on other solid tumors in man, a carcinolytic or carcinostatis action was demonstrated in a number of totally unrelated forms of cancer, cancers unrelated to leukemia, such as neuroblastoma, carcinoma of the breast, or carcinoma of the prostate, with metastatic lesions elsewhere, and so on. In all instances, after a period of excellent tumor effects, varying from three months to, in some instances, twelve months, the "resistance" occurred and, in every instance, in every patient observed long enough to permit this statement, death has occurred. I think, certainly, this is the most effective group of compounds in acute leukemia. They are of great interest in other forms of cancer. The problems which the workers concerned with the further study of these compounds would like to have answered are, first, the nature of the toxic changes; second, the mechanism of their carcinolytic and carcinostatic action. Finally, there is the actual mechanism of resistance of the cancer cell which has once responded. There is considerable information coming from different laboratories which bears upon this subject. Eventual solution of that problem, however, will be necessary before further progress can be reported in the treatment of patients with incurable cancer with these antifolic compounds. The use of these compounds in conjunction with, or following, ACTH, cortisone, and related compounds has been of great interest. There is an effect which can be produced by the two in series or at the same time which, in some instances, cannot be produced by the antifolic compounds alone.

393 One last comment, if I may make it, Mr. Moderator. When one begins to work with patients with cancer incurable by surgery or radiation, the patient is included in the research team, as is the mouse. Since the purpose of cancer chemotherapy is the cure of cancer in man, it is essential that the splendid work going on with other agents, must be accompanied by actual observation on the living human being with cancer which is incurable by present methods of the rapy. A final point in this discussion: If that is done, then the finest medical and surgical care available must be given to the patient with incurable cancer before chemotherapy is instituted, and chemotherapy must be stopped if, at any time, other forms of treatment promise more for the patient than the form of trial therapy which has been instituted.

394 ANTIMETABOLITES AS CHEMOTHERAPEUTIC AGENTS George H. Hitchings The Wellcome Research Laboratories Tuckahoe, New York There are two contrasting ways in which one may go about the preparation of a new chemotherapeutic agent. One is the purely empirical approach - the testing of a wide variety of chemical substances, the selection of a few which have a selective effect on the undesirable organism or tissue, followed by further testing of variants of the selected molecule. The other approach, the deductive or logical way, stems from previous knowledge of a difference in metabolism between the host and the parasitic tissues upon which one might capitalize in a variety of ways. Of the two, the empirical, inductive way, has been much the more productive and has yielded nearly all the chemotherapeutic agents which are available. This does not mean that the deductive method may not yet yield an abundance of new drugs; it means simply that insufficient knowledge of intermediary metabolism has been available or that insufficient thought has been applied to the problem of selection of metabolic differences worthy of exploration. Antimetabolite studies, which have as their aim the production of chemotherapeutic agents, might be thought of as one way in which one might capitalize on a metabolic difference between host and parasite. Thus, had one known of the bacterial requirements for £-aminobenzoic acid, and had the absence of a human requirement for this growth factor been clearly demonstrated, sulfanilamide might have followed as a logical consequence of this knowledge, rather than as an empirical discovery. However, much the greater part of the work in the antimetabolite field has been concerned with the demonstration of the phenomenon itself. For this purpose, the most suitable metabolites are those which are used by a wide variety of living organisms. Pyrithi- amine, for example, has the characteristics of a competitive inhibitor of thiamine in a bacterium, and produces symptoms in the mouse of thiamine deficiency which can be reversed with thiamine. One is convinced, therefore, that such a substance acts by competing with the metabolite for some cell surface, probably an enzyme system. Information of this sort does not, in general, lead to new chemotherapeutic agents, for thiamine appears to be required by all living cells and the requirements probably do not differ quantitatively by a large ratio. What is needed for the production of new chemotherapeutic agents in a logical way is knowledge either of a large quantitative difference in the requirement for the metabolite, or, much better, the existence of a qualitative difference between host and parasite in the performance of some metabolic task. This kind of knowledge is available only rarely. To some extent it may be gained through the antimetabolite study itself. Our approach to new chemotherapeutic agents has been via the biochemistry of nucleic acids. It was felt that quite generally there must be a differential requirement between host and parasitic tissues for the precursors of nucleic acids, since the rapid growth of the parasitic tissue must involve a rapid synthesis of nucleic acids. So the possibility exists that chemo- therapeutic agents for a variety of disease processes might arise on this purely differential rate basis. There were also reasons for believing that qualitative differences in the biosynthesis of nucleic acids do exist. When our work was begun it was known only that the white rat does not use guanine or uracil as precursors of its nucleic acid; however, it was known that certain bacteria and molds do require one or the other or both of these substances. Such observations offered the promise that antimetabolites of the natural purines and pyrimidines might selectively block pathways which are essential to certain organisms and not to others. In the cancer field, this approach has not yet produced a cure but it has given a few leads which may be regarded as hopeful. The activities of the antifolics and of diaminopurine probably are to be regarded as the result of quantitative differences between normal and neoplastic tissues in the need for certain precursors of nucleic acids, but the mode of action of both these substances requires clarification. Despite inferences to the contrary, it appears improbable that any real

395 qualitative differences in nucleic acid synthesis between normal and neoplastic tissues have yet been found. The activity of 8-azaguanine on certain tumors is still somewhat of a mystery. The utilization of guanine is certainly not the common denominator of neoplastic growth which it was hoped it might be, for many tumors fail to respond to the analogue. Moreover, studies by Dr. George B. Brown with isotopically labeled guanine have shown no clear-cut difference in the use of guanine by tissues of the host and the neoplasm. On the other hand, this substance, the antifolics, diaminopurine and the other purines and pyrimidines which have a selective inhibitory effect on neoplastic tissues should be regarded as leads which may be elaborated by further studies and modifications of the molecules; these may eventually result in chemotherapeutic agents of real value. One of the axioms of antimetabolite-metabolite studies is that maximal activity of the structural analogue is produced with a minimum number of changes of the metabolite molecule - the change of one grouping usually giving a stronger antimetabolite than an alteration of two groups or atoms and so on. This has a considerable validity with respect to antimetabolite studies per se but could be very misleading if applied to the problems of chemotherapy. For purposes of illustration, one may return to the example of the sulfonamides. The molecule representing a minimal change from £-aminobenzoic acid would be sulfanilic acid. If one stopped there he would not have a very effective chemotherapeutic agent. If he continued to work he might soon recognize the importance of the pKa's of the metabolite and antimetabolite and thus come to sulfanilamide. This substance is a useable chemotherapeutic agent but has several undesirable side effects. In other words, it becomes involved, unexpectedly, in the host's metabolism in a number of ways. In actual practice, of course, given the lead provided by the empirical discovery of the action of sulfanilamide, the chemists proceeded to modify that molecule in a variety of ways. A s work progressed some of the patterns of chemical-biological correlation became apparent, and in turn could be related to certain physical properties of the molecules. Eventually drugs were produced which were not only more active than the original, but in which many of the undesirable features had been modified or eliminated. I think it is not too far-fetched to suggest that we have found in the field of antagonists of nucleic acid derivatives as related to cancer, substances of potential chemotherapeutic value - substances which stand to the chemotherapy of cancer about as sulfanilic acid might stand to the chemotherapy of bacterial diseases. A substance like 2, 6-diaminopurine, for example, has only a small and rather questionable differential effect on sarcoma 180 in the whole mouse, but, as has been found by Dr. J. J. Biesele, it is very considerably more toxic to sarcoma 180 than to mouse embryonic tissue, when studied in tissue culture. The problem here may be essentially to learn how to deliver the purine to the cancer cell without allowing it to be dissipated in side reactions. The antifolics inhibit sarcoma 180 when tested in the whole mouse but have little or no effect on either the sarcoma or embryonic tissue in tissue culture. Does this mean that the antifolic itself is modified in the animal metabolism, or does it act by depriving the cell of some essential metabolite which is produced in a specific locus and then transported to the growing cell ? Perhaps some other explanation will be found, but answers to such questions as these may lead to modified and better approaches to the problem of cancer chemotherapy. The activities of a number of the pyrimidine derivatives are not readily interpretable in terms of metabolite: anti- metabolite relationship but nevertheless point the way to work which should be done. Undoubtedly many of the apparent openings to chemotherapeutic agents will be found, on exploration, to be blind alleys. However, the importance of the problem suggests that each should be explored with considerable diligence. The working out of these leads will be mainly if not completely empirical. Antimetabolite studies may produce leads but are no substitute for hard work - the kind of thoughtful work which has been termed "enlightened empiricism".

390 CARCINOGENESIS Arnold M. Seligman Beth Israel Hospital and Department of Surgery Harvard Medical School Boston, Massachusetts Since I have been called upon, 1 might make one brief comment. Dr. Berenblum referred to Dr. Haddow's discovery that many of the carcinogenic hydrocarbons inhibit tumor growth. In looking over the data which Dr. Haddow published in that regard, we were impressed with the few exceptions, i.e. , that some non-carcinogenic compounds also inhibited tumor growth. He was interested in showing a correlation between the ability to inhibit tumor growth, and the property of carcinogenicity, and he showed a fairly interesting correlation in that regard., but there were a few exceptions which interested us. A few tumors were inhibited by compounds which were not carcinogenic, but resembled superficially the carcinogenic hydrocarbons in structure and did inhibit tumor growth. It was at that point that we became interested and prepared a series of compounds similar to these compounds, to see if this property of inhibition of tumor growth was extensive. We prepared a number of azo dyes which simulated the ring structure of the hydrocarbons (J. Am. Chem. Soc. , 71 , 3010 (1949)) Some of these azo compounds showed inhibition of tumor growth (sarcoma 37 and Walker carcinoma 256, to be published in Cancer). Later, he published some data in which he showed that the inhibition of tumor growth with the carcinogenic hydrocarbons was very much related to the protein content of the diet, and he pointed out that the observations he had made were that the carcinogenic hydrocarbons inhibited tumor growth when a fairly low protein diet was used, 5 per cent protein. When this was increased to 25 per cent, the effect was less apparent. Our own animals had received the higher protein diet. It was at about this point, unfortunately, that we lost interest in the phenomenon, and I suspect it is at precisely that point that many of you would become very much interested in it.

397 CANCER CHEMOTHERAPY Alfred Gellhorn Department of Medicine and Cancer Research Institute College of Physicians and Surgeons Columbia University New York, N--w York Dr. Shear has asked me to make some comments on the possible relationship of pyridoxine to the metabolism of neoplastic cells in man. We explored this possibility several years ago in clinical cancer chemotherapy and although the results were negative we learned several lessons. The basis for the clinical experiment rested on a series of observations made by Dr. H. C. Stoerk and others. It had been noted that lymphoid tissue was dependent upon the vitamin, pyridoxine, for its maintenance and growth. In experimental animals placed on a pyridoxine deficient diet there was rapid involution of the thymus gland ami other structures rich in lymphocytes. Stoerk further reported that in mice bearing a transplantable lymphosarcoma there was striking inhibition of the tumor growth when the animals were placed on a pyridoxine deficient diet and in addition were fed an analog of the vitamin, desoxypyridoxine. Our clinical observations were carried out on a number of patients with disseminated lymphosarcoma and also patients with acute lymphatic leukemia. The patients were placed on a rigid pyridoxine deficient diet together with the oral administration of large doses of desoxy- pyridoxine. During a period of three weeks, no detectable changes in evident tumor masses or peripheral blood picture were noted and there was no evidence that the course of the disease had been modified. Biochemical studies in these patients failed to reveal an increased urinary xanthurenic acid excretion, a sign of pyridoxine deficiency in most lower species. The lessons that we re-learned from this experience were three: (1) The correlation between tumors in experimental animals and neoplastic disease in man is not known. Precise information on the biochemical similarities or dissimilarities between the animal and human disease would be of tremendous value for the evaluation of observations made in cancer chemotherapy studies in experimental animals. (2) In many instances there are species differences in the metabolism of drugs and therefore the translation of animal observations to man can not be anticipated a priori. (3) A very important consideration, it seems to me, is that if a chemical compound merely shows tumor inhibition in the experimental animal it is probably not wise to try this agent in man. This is so, I think, because a carcinostatic effect in man is too difficult to demonstrate due to the relative chronicity of malignant disease in man as compared with cancer in experi- mental animals. Pyridoxine deficiency inhibits the growth of lymphosarcoma in mice, urethane inhibits the Walter rat carcinoma, 8-azaguanine inhibits a number of experimental cancers but clinical observations have failed to duplicate the results. This does not mean that therapeutic regimens with these agents might not have a carcinostatic effect in man, it merely indicates that it is impractical to conduct an experiment on humans which will require many months or even years before a conclusion can be reached. For this reason I believe it is permissible to conclude that clinical trial should be reserved for chemical compounds which have been shown to be carcinolytic against experimental tumors.

398 PURINES AND PYRIMIDINES IN EXPERIMENTAL CANCER CHEMOTHERAPY Alfred Gellhorn Department of Medicine and Cancer Research Institute College of Physicians and Surgeons Columbia University New York, New York The observation by Kidder in 1949 that a guanine analog, 2-amino-4-hydroxy-triazolo- pyrimidine, inhibits the growth of certain tumors has been confirmed and extended in our labora- tory. In an attempt to gain some insight into structure-activity relationships of this carcinostatic chemical agent, a series of pyrimidines and purines have been synthesized by Dr. Morris Engel- man at the Columbia University Cancer Research Institute and tested against a spectrum of neoplasms in experimental animals. The technique of evaluation consists in transplanting the tumor fragments into the axillae of the appropriate hosts and instituting therapy 24 to 48 hours later. In every experiment a minimum of 10 animals was employed for the study of a compound, and untreated controls and 2-amino-4-hydroxy-triazolopyrimidine (azaguanine) treated series were always included. In previous experiments it had been determined that a daily dose of 50 mgm. per kilogram intra- peritoneally of the azaguanine was uniformly effective and without apparent toxicity. The unknown chemical Compounds were administered at do'ses equimolar to the azaguanine and also at maximally tolerated doses. The duration of an experiment was from 14 to 21 days. The single exception to the method just described was for the evaluation of chemotherapeutic effect against the Brown-Pearce tumor. In this case the tumor fragments were transplanted to the anterior chambers of rabbits' eyes, therapy was not initiated until 3 days after transplantation and the duration of drug administration was limited to 7 days. In all instances the animals were sacrificed at the termination of the experiment and the subcutaneous or intraocular tumors were dissected out and weighed. Table 1 presents a key to the tumors used and their hosts, results of all of the compounds thus far studied. Tables 2 and 3 present the TABLE 1 Tumor Designation* Tumor Type Host 755 Carcinoma. of the breast C57 black mice RC Carcinoma of the breast dba mice EO771 Carcinoma of the breast C57 black mice Brown-Pearce Undifferentiated squamous cell carcinoma AH rabbits Sarcoma 180 Undifferentiated sarcoma Paris Kill and Longacre mice 6C3HED Lymphosarcoma C3H mice 9417 Acute leukemia AK mice C1498 Acute myelocytic leukemia C57 black mice * All tumors are transplantable

399 TABLE 2 EXPERIMENTAL NEOPLASM Compound 755 RC EO771 Brown-Pearce Sa. 180 6C3HED 9417 C1498 OH

400 TABLE 3 EXPERIMENTAL NEOPLASM Compound 755 RC EO771 Brown-Pearce Sa. 180 6C3HED 9417 C1498 OH =N- H=N-OH 2 N OH C"3 ^>JH2

401 As can be seen, no purine analog other than the 2-amino-4-hydroxy-triazolopyrimidine had any demonstrable effect against the tumors tested nor did any of the pyrimidines. Obviously at this time it is impossible to make any statements about structure-activity relationships in this series. This rather discouraging state of affairs is in part ameliorated by a series of interesting observations made by Dr. Jacob Kream and Dr. Ada Graff. Dr. Kream, who has been studying the deamination of guanine by liver guanase, demonstrated that both liver and tumor extracts are able to deaminate azaguanine. The product of such enzymatic action was found to be 2,4- dihydroxytriazolopyrimidine. It is to be noted in Table 2 that this compound was tested and found to be inactive. In the hope of slowing the rate of inactivation, the diacetyl and 2-amino sulfoxylate derivatives of azaguanine were synthesized. These also proved inactive because, as Dr. Graff found, the acetyl and sulfoxylate radicles are not removed, at least in vitro, and the 2-amino group remains masked. Although our results to date have been negative, the pursuit of structure-activity relationships in this series would appear to be potentially useful because (1) azaguanine probably interferes with nucleic acid metabolism and further understanding in this area could not fail to be of interest, (2) azaguanine appears to have a more selective action on neoplastic tissue than on normal tissue and, therefore, more potent carcinostatic or carcinoclastic agents might have clinical application and (3) azaguanine, in our experience, inhibits epithelial neoplasms but does not alter the growth of sarcoma or leukemia and, therefore, insight into its mechanism of action may lead to greater understanding of differences in metabolism in various types of malignant tumors.

402 CARCINOGENESIS W. C. Hueper, M. D. National Cancer Institute National Institutes of Health Bethesda, Maryland It becomes increasingly apparent that the attempts at establishing consistent and generally applicable correlations between the structure and carcinogenicity of chemicals may suffer the same fate that has overtaken similar efforts aimed at correlating chemical structure and pharma- cologic and toxic action of chemicals or the various types of microorganisms and their patho- genicity. There does not appear at the present time any possibility of linking in a common pattern the numerous aromatic carcinogens with the aliphatic ones or with the scarcely studied cancer- igenic metals, such as arsenic, chromates, nickel, asbestos, selenium and beryllium, not to mention in this connection the non-ionizing and ionizing cancerigenic radiations. Since the bulk of the available information on this subject is based on observations made with synthetic aromatic chemicals most of which have no practical importance to man, and because in most of the assay work only mice were used, it is impossible to state how much of the little we know of such correlations has any direct or indirect application to man. Doubtlessly, the terms "carcinogenicity and carcinogenic potency" are in need of re interpretation. They should be defined in specific terms, such as species susceptibility, target organ, and route of contact, dose and vehicle used, before they acquire any real meaning. The demonstration of correlations between the structure and cancerigenic potency of chemicals would be not only of distinct scientific importance, but also of eminent practical significance. The rapidly increasing number of environmental human cancerigens has created problems in connection with appropriate screening methods of chemicals that are of great urgency and complexity. Their solution would be simplified and hastened by the availability of information of such correlations, even if they should apply only to restricted groups of compounds. However, the procurement of such data, obviously, depends not only on a much deeper insight into the carcinogen-host relation and their inter-reactions, but also on a widening of our present methods of bioassays, if we expect that this information may become of some value to the human cancer problem. For illustration, reference may be made to the extensive use of skin application of cancerigenic chemicals in experimental tests. Since about 75 to 95 per cent of human skin cancers can be cured with existing methods, any information as to correlation between structure and carcinogenicity of chemicals has only limited value. Likewise, the numerous studies on the hepatocarcinogenic action of certain azo compounds, while of definite value to the human problem, provide data restricted in their importance by the fact that primary liver cancer in man is uncommon. On the other hand, very little experimental work has been done to investigate the cancerigenic action of chemicals on the organs of the respiratory tract, although cancer of the lung has become a highly important tumor because of its rapidly increasing frequency and high rate of mortality. It is apparent from these considerations that a great deal more information is needed before the human problem can benefit from the present studies on the correlation of structure and carcinogenic properties of chemicals.

403 CANCER CHEMOTHERAPY Konrad Dobriner Sloan-Kettering Institute for Cancer Research New York. New York Drs. Stock and Shear have already discussed some phases of the usefulness of steroid hormones in the therapy of neoplastic disease in humans. Testosterone and the estrogens have been applied in cancer of the breast in females. Estrogens are also used in the therapy of cancer of the prostate, and recently the effects of large amounts of progesterone in cancer of the cervix have been reported. The influence of large amounts of adrenal cortical hormones (cortisone) and effects obtained after stimulation of the adrenal by ACTH on certain tumors of the lymphoid tissues are well recognized. The correlation of chemical structure and antihormonogenic activity is still very little understood. If one considers that a slight change in the structure of a steroid hormone, often affords a striking change of biological action in most instances, a wide field for therapeutic trial is open as long as the chemist is willing to produce variations in the steroid structure and make compounds available in large enough amounts for testing. There seems to be now more hope that this will be possible than it was ten years ago. It seems that with many new steroids being made available, more thorough information will be obtained on the relation between structure, biological, and therapeutic activity on these compounds for the benefit of the suffering patients.

404 CANCER CHEMOTHERAPY J. H. Burchenal Sloan-Kettering Institute for Cancer Research New York, New York Since I have a foot in both camps, I would like to put a plug in for the mouse. It has been somewhat maligned today. I feel that clinical research on new compounds is a very expensive thing. Granted, it is the final proof of the pudding, but I think you have to have pretty good reason for going to the patient before you go there. We have screened some three hundred compounds related to the folic acid antagonists against mouse leukemia. We must qualify this statement however. We have used a system in which the mice were on a normal diet, which I think arbitrarily rules out most of the weak antagonists. We have found that only the compounds which were full-dress folic acid antagonists - by that, I mean they contained the pteridine, the j)-aminobenzoic acid, and the amino acid moieties and had an amino group in the 4-position of the pteridine ring, were active against this mouse leukemia. There was one other compound which was active, 2, 6-diaminopurine, which is somewhat analogous to the 2, 4-diaminopteridine. We thought, originally, that possibly the two worked the same way, but I think we have fairly definite evidence that 2, 6-diaminopurine and A-methopterin (4-amino-N10-methyl-pteroylglutamic. acid) work in different ways on their effects on the mouse. We have a strain of mouse leukemia which has developed complete resistance to therapeutic doses of A-methopterin so that there is no prolongation of survival time. This same strain of resistant leukemia, however, is affected by 2, 6-diaminopurine, and there is a significant prolongation of survival time. One thing more I would like to "mention; that is that when Dr. Shear introduced the folic acid antagonists, he called them the so-called folic acid antagonists. There is a great deal of controversy as to whether they are really folic acid antagonists, but we have definitely shown that at least the anti-leukemic effect of A-methopterin can be prevented by folic acid. It also might be appropriate to mention that this anti-leukemic effect can be prevented much more readily by the citrovorum factor. We feel that, on a dry weight basis, the citrovorum factor is probably twelve to twenty-five times as potent as folic acid in preventing the anti-leukemic effect I hope Dr. Welch will tell us more about his very interesting studies on citrovorum factor shortly.

405 CANCER CHEMOTHERAPY E. B. Schoenbach Johns Hopkins University Baltimore, Maryland Dr. Shear, I do not think I understand your question too clearly. Therefore, I shall take the liberty of selecting the topic. I think one could summarize a good deal of what has been said today, and our .own attitude, in the words of Lewis Carroll that because we breathe when we sleep does not necessarily imply that we sleep when we breathe. When we first found that the antifolic acid compound, 4-aminopteroylglutamic acid, exerted a definite inhibitory effect on the growth of sarcoma 180 and changed the cytologic appearance of the cells, we felt we should investigate to see whether these alterations were induced by the antifolic properties of this compound or whether the structural alteration in the configuration of the molecule had resulted in some new chemical compound which induced inhibition of tumor growth. We, therefore, treated the animals with folic acid and the antimetabolite, and the tumor inhibition was reversed. However, we could not demonstrate any change in the toxicity to the host of this antifolic acid compound in the presence of folic acid. Time does not permit a detailed presentation of the various experiments which were performed, but one may summarize them by noting that, in acute toxicity, tests with a single injection, if the folic acid were given at least an hour before the antifolic acid compound, one could demonstrate protection to life and prevention of weight loss. When less than an hour intervened, or the compound was given simultaneously, no protection whatsoever could be shown with the folic acid. We also were able to show that, if a suitable interval intervened between the dosage of folic acid and the antifolic acid compound, suitable ratios of these two compounds could be used, and, still, protection would be observed; that is, we could use many times the lethal dose of antifolic acid compound and protect the mouse provided the ratio of folic acid to the antifolic acid was constant. If one then gave repeated dosages of folic acid and antifolic acid, still trying to maintain the early injection of the folic acid, one could not protect the animal. That was quite a stumbling block to us, and we have been unable to explain it. Dr. Welch's and Dr. Burchenal's observations, would indicate that the folic acid was converted to some other compound which protected the mouse. The intervening period of time between the injection of folic acid and the exhibition of the antagonist was probably the necessary interval for the in vivo conversion of the folic acid. Our studies on mechanism of action have taken us far afield into nucleo-protein and steroid metabolism. We cannot discuss these at present, but the results are quite interesting and, I hope, will be available to everyone in published form soon.

406 CANCER CHEMOTHERAPY A.D. Welch Department of Pharmacology School of Medicine Western Reserve University Cleveland, Ohio That is quite a task you have set me. The subject of my remarks relates to a topic which the previous speakers have alluded to, namely, our interest in the citrovorum factor and the mechanism of action of aminopterin. 1 cannot tell you about what we think may be the action of aminopterin without telling you a little bit about the citrovorum factor. Our interest in it developed as an outgrowth of our long concern with studies of the action of folic acid. It was reported in 1948, by Sauberlich and Baumann of Wisconsin, that a new factor was required by a microorganism called Leuconostoc citrovorum ; this substance is referred to as the citrovorum factor. Later it was shown by Sauberlich that administration of folic acid to rats or human subjects increased the urinary output of the new factor. I must skip over a great deal of evidence which has been offered by a number of contrib- utors, particularly the ones mentioned, to indicate that this substance is a derivative of folic acid and not something which is synthesized as a result of a catalytic function of folic acid. It has prematurely been referred to as folinic acid, but, until the substance has been characterized, I think we will continue to refer to it as the citrovorum factor or factors. When you study the urinary excretion of this material after giving folic acid, you find that there is a rather constant relationship between dosage and output of the citrovorum factor. We found, for example, in a study carried out in normal human subjects, that the output in units - and I must refer to it in units, because we have no absolutely pure material - following 5 mgm. oral doses of folic acid, is in the order of 50,000 units per day, instead of the two to four thousand units which normally are excreted daily. If ascorbic acid is given simultaneously with folic acid, there is approximately a three-fold additional increase in the output of the citrovorum factor. This intriguing observation I also cannot go into in great detail at this time. It seemed logical, however, in trying to explain the urinary excretion of the new substance and the effect of ascorbic acid on it, to study the liver in vitro. My colleague. Dr. C. A. Nichol and I turned, therefore, to rat liver homogenates (which do not work) and to liver slices, and found that the latter are able to convert folic acid into the citrovorum factor or, at least, into materials which permit growth of the particular microorganism used in assaying for the new factor. Using a normal rat liver, a considerable formation of the citrovorum factor is observed during incubation of liver slices for two hours; the amount of activity can be doubled, approx- imately, by the addition, per gram of liver, of 100 micrograms of folic acid, for example, raising the amount from 1700 units to about 3400 units. Use of liver obtained from rats deficient in folic acid gives even more striking results. The liver of the rat fed, during a period of several weeks, a purified diet free of folic acid and containing succinylsulfathiazole, contains negligible amounts of both folic acid and citrovorum factor - a further indication of the relation of the two substances. Incubation of liver slices from such an animal fqr a 2-hour period leads to the appearance of less than 100 units of activity, instead of nearly 2000 units, as is the case with the liver of normal rats. Addition of 100 micro- grams of folic acid, under those conditions, leads to a striking increase in the amount of citro- vorum factor. Thus, in one experiment, 60 units were found in the absence of folic acid, and 2500 units in its presence.

407 The addition of vitamin C will double the formation of citrovorum factor by liver slices from the normal rat; also, ascorbic acid will more than double the amount formed from added synthetic folic acid. However, ascorbate has little effect on the formation of the factor by the liver of the folic-deficient rat, unless folic acid is added. In other words, the 60 units which such a liver can form intrinsically, m vitro, is only increased by ascorbic acid to about 130.units. However, the addition of folic acid as well as ascorbic acid causes the formation of almost 13,000 units. There seems to be little doubt that this factor is produced from folic acid, and that the conversion is facilitated by ascorbic acid. It is suspected that the citrovorum factor is more closely related than is folic acid to a prosthetic group of which folic acid serves as a precursor. It seemed to us possible, then, that aminopterin, the 4-amino analogue of folic acid, might block specifically this "activation" of folic acid. As Dr. Schoenbach has pointed out, many have wondered whether aminopterin functions exclusively by interfering with folic acid metabolism, or whether it has other important effects as well. If it affects only folic acid, and is truly an anti- metabolite of this vitamin, why, the question is asked, is its action in animals not effectively reversed by the metabolite it resembles ? Accordingly, aminopterin was tested in the liver slice system, and we were delighted to find that it blocks completely, or nearly completely, depending on the concentration, the forma- tion of citrovorum factor from added folic acid. If the rat is treated with aminopterin before the liver is removed for slicing, the ability of the liver to form citrovorum factor from folic acid is found to be lacking. Even more convincing proof that aminopterin prevents the conversion of folic acid to a physiologically utilizable substance has been obtained by the use of concentrates of this factor. Or. Schoenbach and others have commented on the fact that the toxic effect of aminopterin cannot be reversed effectively by folic acid, if the latter is given simultaneously, and only slightly more efficiently if folic acid is given before the aminopterin. We had only a very small amount of a potent concentrate of the citrovorum factor which had been supplied to us by the Lederle Labora- tories, through the courtesy of Dr. T. H. Jukes. Under conditions where daily injections of 5 mgm. of folic acid had absolutely no effect on the toxicity of 25 micrograms of aminopterin per day, injected simultaneously, and all animals died within a few days, administration of the concentrate in a dose of only 1. 5 mgm. daily prevented completely the lethal effects of the. analogue. This still crude material, containing approximately 250,000 units in the daily dose of 1. 5 mg. , not only prevented death, but, in the folic-deficient rats, growth, which had previously ceased, was initiated. Further rapid growth was maintained during the period of treatment despite daily dosage with the analogue. I think it is safe to say that at least a major part of the action of aminoptezin is due to a profound interference with the conversion of folic acid into a metabolically more active form of the vitamin. No doubt, in addition, the analogue interferes also with the utilization of the citro- vorum factor. In other words, I think it may have a high affinity for at least one of the enzymes concerned with the conversion of folic acid to what we have termed citrovorum factor. In this manner, aminopterin prevents the effective utilization of folic acid, since its metabolic altera- tion appears to be essential to its biological efficacy as a vitamin. In addition, the utilization of the product of the conversion may be less efficient in the presence of aminopterin. The role of ascorbic acid in the transformation is another story. However, it appears reasonable to suggest that it may play a role in one of at least two steps in the conversion of folic acid to citrovorum factor, perhaps in a reductive stage, since other reducing agents, such as gluco-ascorbic acid, also have the same effect as ascorbic acid. The new factor may prove to be useful not only in the study of the enzyme systems involved in the metabolic alteration and utilization of folic acid for its hemopoietic and cell-growth function, but it may have a practical application in alleviating the manifestations of overdosage or of frank poisoning caused by aminopterin. However, it remains to be proved that the toxicity of the analogue, once it is fully developed, can be alleviated as effectively by the citrovorum factor, as the development of toxicity by aminopterin can be prevented by the simultaneous administration of a concentrate of this new derivative of folic acid. How the abnormal cells of leukemic mice become resistant to aminopterin is most intriguing and deserves careful study. How the formation of the substance is disturbed, perhaps, in the megaloblastic anemias is the subject of investigative effort in progress in our laboratories,

408 in those of Dr. Burchenal, and in others. The next few months should add materially to our knowledge of the subject.

409 CANCER CHEMOTHERAPY F.S. Philips Sloan-Kettering Institute for Cancer Research New York, New York I do not think there is need to add further comment to the story of the antifolics. It seems clear that we are in position to understand their mechanism of action from Dr. Welch's lucid presentation. I believe that the mechanism by which antifolics act against some tumor cells is similar to the mechanism involved in their actions against bone marrow cells and the epithelium of the intestinal mucosa. Presumably, certain neoplastic and normal types of proliferating cells have more specific and possibly greater requirements for folic acid than other cells of the mammalian organism. Unfortunately, such a mechanism is not useful in the ultimate therapeutic sense, since agents having a common effect in both normal and tumor cells have limited clinical value. A similar limitation of therapeutic usefulness is inherent in the application of "total body" irradia- tion and the nitrogen mustards to the treatment of neoplastic diseases. These agents have highly specific actions against all types of proliferating cells; and in some instances, they have proved temporarily, though significantly, effective in controlling neoplastic proliferations. The nitrogen mustards are known to act through, transformation in vivo into intermediates containing positively charged ethyleneimmonium rings. Sulfur mustards are also believed to act through in vivo transformation into ethylenesulfonium derivatives. The fact that nitrogen and sulfur mustards exhibit important biological properties in common is readily understood on the basis of their conversion in vivo into derivatives containing reactive ethylenonium moieties. This generalization was well established by investigations carried out during the recent war. More current research has provided additional information concerning the nature of the chemical configurations essential to mustard-like'activity. Thus, compounds containing tertiary ethylen- imine groups have been shown to inhibit and damage tumors and to elicit the profound pathological changes in normal hematopoietic tissues of mammals which characterize the actions of nitrogen and sulfur mustards. In addition, investigators in England have been finding that derivatives containing ethylenepoxy groups exert typical mustard-like activity against tumors. It would appear, then, that mustard-like properties are not only to be related to the ethylenonium config- uration, but more generally to three-membered heterocyclic structures. Presumably, deriva- tives of three-membered heterocycles, which have as yet not been investigated, could also be expected to have similar biological properties. The three-membered heterocycles mentioned above are characterized by high chemical reactivity and, when investigated, have been found to alkylate a large variety of important bio- chemical radicals. Alkylation takes place with groups in both simple organic molecules and in complex biochemical entities such as proteins or nucleic acids. The alkylating properties of the mustards have featured in explanations of their mechanism of action. In addition, certain hypotheses have attempted to account for the well-known fact that among nitrogen and sulfur mustard analogs only those containing at least two reactive groups per molecule elicit character- istic mustard-like actions. Accordingly, it has been suggested that mustard-like properties are to be related more to the capacity of single molecules to engage in multiple alkylations rather than to the possession of the three-membered heterocyclic configuration. Nevertheless, it remains unchallenged, at present, that only derivatives containing three-membered heterocycles have the capacity to elicit all of the characteristic actions of the mustards against normal and neoplastic proliferating cells both ui vivo and in vitro. Furthermore, recent investigations with new compounds containing only one tertiary ethylenimine group reveal certain of these agents to inhibit Sarcoma 180 in vivo and to damage normal mammalian hematopoietic tissue in the characteristic manner of the mustards.

410 Dr. John Biesele in our laboratory has compared the cytological effects of mono-ethylen- imines and mono-ethylenepoxy derivatives in dividing cells of the onion root tip with the effects of roentgen radiation, of a nitrogen mustard, and of a bis(ethylenimine) derivative. It is well-known that the characteristic chromosomal abnormalities caused by roentgen radiation in such cells can be duplicated by their exposure to HN2. Dr. Biesele has found the bis-(ethylenimine) to elici- similar chromosomal changes. However, his most interesting findings concern the observation that ethylenimine and mono-ethylenimme derivatives, as well as a simple ethylenepoxy derivative such as glycidol (2, 3-epoxy-propanol-1), cause chromosomal alterations which cannot be distinguished from those following treatment with roentgen radiation or HN2. In view of the properties of certain mono-ethylenimine derivatives, it becomes difficult to accept the hypothesis that polyfunctional configurations are essential to mustard-like activity. Nevertheless, it remains true that compounds in which there are at least two ethylenimine moieties are 50- to 100-fold more toxic and more readily capable of producing damage to pro- liferating cells in vivo than are analogous mono-ethylenimine derivatives. For example, the LD^Q of 2, 4-bis -(ethylenimino) -6-amino-s -triazine in the rats is less than 1 mgm. kgm.; on the other hand, the LD50 of 2-ethylemmino-4. 6-diamino-s-triazine is about SO mgm. /kgm. More- over, in effective doses the mono-ethylenimines may cause lesions in tissues other than those containing proliferating cells. Thus, the LD5Q of ethylenimine causes in the rat a severe necrotizing lesion of the renal papilla - an observation reported about 1890 by Ehrlich; when 4 x LD5Q doses of ethylenimine are administered typical mustard-like lesions are observed in bone marrow and lympoid organs. On the other hand, lesions caused by bis-(ethylenimines), like those caused by HN2, are restricted only to hematopoietic organs, intestinal epithelium and testis. It might be suggested that duplication of three-membered heterocyclic functions in molecules enhances the in vivo specificity of action against proliferating tissues.

411 DISCUSSION DR. W.R KIRNER (Director, Chemical-Biological Coordination Center): The compound CH2*CH-CH2OH is fairly readily available if Dr. Philips is interested in testing it. XS' It was obtained during the war when attempts were made to prepare BAL by a catalytic process. The method was not successful because dehydrogenation occurred producing the above compound. I believe a sample could be procured if Dr. Philips was interested in testing it in connection with his hypothesis. CHAIRMAN SHEAR: I am sure that Dr. Philips, in his stimulating discussion, did not for a moment mean to imply that this postulated mode of action would be responsible for the modus operandi of other types of agents. As a matter of fact, we hope such will not prove to be the case, because the little progress that haa already been made in the field of the chemical treatment of tumors has shown that even after a tumor responds, for a number of months, to a given agent, it may then, and usually does, become resistant. If we have a variety of types of compounds which operate by means of different mechanisms, that would offer much more hope than if they all operated by the same mechanism. Does that provoke comment from anyone ? DR. HOWARD E. SKIPPER (Southern Research Institute): Dr. Shear asked me to talk about carbamates which carried me away from things a little closer to my heart at the present moment. That is, some work on the mechanism of the anti-leukemic action of certain folic acid antagonists which fits in very well with Dr. Welch's observations; and secondly, some work on tracing 8-azoguanine (Kidder's guanazolo) which this group might find of interest. Taking up the work on labeled 8-azoguanine first, we have, working in cooperation with Dr. Stock and Dr. Sugiura of Sloan-Kettering Institute, observed the distribution of the active 2-carbon atom of this compound in C57 black mice with Eo771 tumors. The tumors did not selectively pick up the radioactive atom. Tumor and normal tissue throughout the mouse was found to contain active carbon from 8-azoguanine-2-C14 at 1, 6, and 24 hours after injection. Kidder had suggested that the tumor-inhibiting action of this compound might be due to differences in purine metabolism which might result in preferential 8-azoguanine incorporation in the nucleic acids of tumor tissue. We have injected active 8-azoguanine into normal mice and after 6 and 24 hours isolated quite active nucleic acids from the viscera of the animals. The hydrolysis products of these viscera nucleic acids were then chromatographed (with a small amount of carrier 8-azoguanine) and we were pleased to find that much of the activity which resided in these nucleic acids was there as 8-azoguanine. Similar isolations with added "free" 8-azoguanine showed that the isolation procedure did not concentrate 8-azoguanine. One cannot be too bold in interpretation of these results; however, we feel that the presence of this non-metabolite in nucleic acids might be expected to affect nucleic acid meta- bolism of the 8-azoguanine injected mice. The second point I wanted to bring up has to do with some observations on the effect of aminopterin and A-methopterin on nucleic acid metabolism. Perhaps most of you know of the interesting observation made by Dr. Woolley showing ' that E. coli inhibited by aminopterin build up 4-amino-5-carboxaminoimidazole.

412 H2N-C=O C-N I H H2N-C — ^ 4-amino-5-carboxamidoimidazole You can see that this compound looks something like the purine skeleton minus the 2-carbon atom. It has been shown by Buchanan and his co-workers at the University of Pennsylvania that formate is the precursor of the 2-carbon in uric acid and presumably in nucleic acid pu'rines. We have used the rate of incorporation of Cl4-formate into nucleic acid purines as a means of estimating the rate of uj vivo synthesis of nucleic acid purines and in turn, nucleic acids. Once one has established a base line for incorporation of formate, it is then possible to study the effects of various means of therapy on nucleic acid metabolism. When mice were treated with A-methopterin or aminopterin, it is interesting to note that incorporation of formate into nucleic acid purines (over a six-hour period) is depressed to a small fraction of that seen in normal animals. If, as has been suggested by Shive and co-workers, formyl folic acid (or more recently folinic acid or acids) is the prosthetic group of a co-enzyme system necessary for introduction of single carbon moieties into the purine skeleton, then these observations may help to explain the mechanism of the anti-leukemic action of folic acid antagonists. CHAIRMAN SHEAR: Dr. Skipper, d'id you say that there was not a preferential concentra- tion of the 8-azaguanine in the tumors? Did the tumors respond in any way which could be detected? DR. HOWARD E. SKIPPER: These studies on distribution of labeled 8-azaguanine were only 6 or 24-hour experiments so no observations on tumor growth were possible. I think it is well established by the work of Dr. Gellhorn, Dr. Sugiura, and, of course, Dr. Kidder, that certain tumors (the Eo771 included) will respond to this compound. With regard to the radioactive formate incorporation studies, we have observed that at 6 hours the specific activities of control viscera nucleic acid purines from mice were about seventeen times that of folic acid antagonist treated animals. These results have been confirmed by repeated experiments. We feel that these results fit in very well with the observations of Woolley, who found that in the presence of aminopterin, E. coli apparently cannot incorporate the single carbon unit, the precursor of which is formate, into the 2-position of the purine skeleton. DR. A.D. WELCH (Western Reserve University): I might add something to that. The effect of aminopterin on the incorporation of formic acid indicates that the reaction may be catalyzed by folic acid or a derivative of it (perhaps the citrovorum factor) or a product of its metabolic activity, and it appears possible that a general function of folic acid is the transfer of a single carbon unit. That idea is not original with me, of course; a number of people familiar with the field are in this room and will be cognizant of speculation along these lines. Sakami and I reported in Atlantic City in April (1950) at the meeting of the Federated Societies that formic acid also is converted by the rat, both in vivo and in vitro . to the methyl group of methionine and we described findings indicating that folic acid, or a derivative of it, is involved in the synthesis of the methyl group, a synthesis hitherto considered not to be performed by mammalian species. DR. HOWARD E. SKIPPER: One point 1 did not mention was that mice with leukemia incorporate formate into nucleic acids much more rapidly than do normal mice.

413 CHAIRMAN SHEAR: But, in your experiments with labeled formates, together with the experiments on the labeled 8-azaguanine, you have illuminated some questions of metabolism. So far as the treatment of tumors is concerned, just about a year ago there was tremendous publicity given to guanazolo as a possible chemotherapeutic agent. 1 wonder if you would care to comment. Dr. Gellhorn, on what results you got in patients with this substance. DR. A. GELLHORN (Columbia University): I am enthusiastic about 8-azaguanine, but 1 have not considered its use in patients. The basis for this decision rests on the fact that guanazola, or 8-azaguanine, has had a purely carcinostatic action in all observations thus far recorded. If a chemical compound inhibits the growth of an experimental tumor but fails to cause regression of the tumor it appears to me highly unlikely that it will show a qualitatively different effect on a malignant tumor in man and be carcinolytic. I do not wish to suggest that a carcino- static agent for human neoplastic disease would not be a great step forward; however, the chronicity of cancer in man as compared with the mouse is such that the clinical trial of a purely growth inhibiting drug is not practical. Although we have not given 8-azaguanine clinical trial, the drug has been used in other institutions. It is my understanding that no beneficial effects have been observed, but Karnofsky has recorded toxic manifestations in the form of skin rashes. The available clinical evidence is insufficient to determine whether the drug inhibits any human tumor. I would like to pursue the discussion of 8-azaguanine started by Dr. Skipper. From the tracer studies with this chemical compound which he mentioned, I understand that there is no specific localization in the nucleic acids of tumor when compared to normal tissues. Does this imply that there is no selectivity in the action of 8-azaguanine on tumors? I do not believe that this necessarily follows. Barbiturates do not selectively localize in the central nervous system nor does digitalis concentrate in the myocardium; yet it cannot be denied that these drugs do have specificity of action. In our laboratory, Shapiro, Weiss and I have studied the specificity of 8-azaguanine by counting the mitoses in normal and neoplastic tissues of treated and control animals. We have found that the mitotic activity in tumors from treated animals is markedly reduced, whereas the rate of cell division in the intestinal crypts and in testicular epithelium is the same in treated and control animals. We feel that this indicates some specificity of action on tumor cells. I would also like to ask Dr. Skipper a question. We have obtained evidence that the enzyme of normal tissues which deaminates guanine to xanthine is unable to distinguish 8-aza- guanine and also deaminates this compound. The resulting xanthazola is completely inactive against tumors. In your isolation of the guanine analog from nucleic acids could you distinguish between 8-azaguanine and 8-oxaxanthine ? DR. HOWARD E. SKIPPER: I cannot answer that question positively, but I believe that our method of isolation (paper chromatography) would distinguish between guanazolo and deaminated guanazolo. DR. A. D. WELCH: Did you find radioactivity in any of the other constituents of nucleic acids ? DR. HOWARD E. SKIPPER: Yes, we did, but I think we had something of a 10, 000 or 20,000 safety factor in making this estimate, so there was not much doubt in our minds. CHAIRMAN SHEAR: Among the novel ideas presented this afternoon was one which Dr. Gellhorn has mentioned. It has been an^ priori assumption that one ought to look for a compound which collects preferentially in the tumor, and here was discussion of a compound for which the concentration in the tumor was lower than elsewhere. There is another agent which is capable, at least on the first injection, of producing dramatic changes in some tumors, both in animals and in patients, viz. a polysaccharide fraction from certain bacteria. It had been expected, on the basis of the prevailing concept, that this material would be present in the tumor in much larger amounts than in the normal organs. However, when Seligman labeled it with radio-iodine, and tracer work was done it was found, somewhat to our astonishment, that the tumor contained much less of the labeled material than

414 did normal organs, such as liver and lung. Sot perhaps, we ought not to be unduly impressed with the requirement that a substance, to be of use in treatment of tumors, must necessarily collect preferentially in the tumor. DR. W.R. KIRNER: In doing the analysis on nucleic acids, were they distinguished as to type? DR. HOWARD E. SKIPPER: No. In this work we isolated combined nucleic acids and combined nucleic acid purines. Work is in progress in which desoxyribose nucleic acid and ribose nucleic acid as well as the individual purine and pyrimidine moieties of these polymers are being isolated for activity assays following injection of radioactive formate and known anti-cancer agents. DR. W. R. KIRNER: Histologically and histochemically, there is evidence that, when aminopterin is given to animals bearing sarcoma 180, the nucleic acid continues to be formed in those cells. You might be able to show the differential incorporation of the formate carbon by separation of the two components. DR. HOWARD E. SKIPPER: It is known that x-radiation will preferentially inhibit nucleic acid purine synthesis. Hevesy made such observations and we have confirmed them. CHAIRMAN SHEAR: Your moderator no longer feels it his duty to needle those in attendance to contribute further to the discussion. I feel the moderator's duty has been accomplished, and his conscience is free; nevertheless, I do not want to give any implication of steamrollering or calling to an abrupt clpse this brief afternoon's discussion. Is there any final comment, question, or suggestion? DR. C. CHESTER STOCK (Sloan-Kettering Institute for Cancer Research): I would like to talk briefly on a point Dr. Gellhorn raised. I believe he suggested that in tumor chemotherapy studies in which inhibition is observed as the criterion of value, the selected compounds might not have much value in the clinic. I think that is quite true for some tumors. The first slide I showed indicated that certain tumors are more sensitive to inhibitory actions than others, and that is particularly true with the rat tumors within the small group of effective compounds we have studied. One of the great misfortunes right now, I tnink, is that we are lacking in really adequate, effective compounds we can use as yardsticks to judge our chemotherapy screening programs. However, just as an indication of tumor inhibition studies is the following. We have been through quite a large number of compounds; we have found nine which meet our maximum grade, positive grading. We do not.mean that is the maximum we can get, but it is the maximum grading in the screening procedure. All nine of those have gone to clinical trial, I must say, not so much based on the sarcoma 180 results as the leukemia studies. All nine of those have shown some benefit clinically; seven of them are antifolics, and one is triethylenemelamine, which Dr. Gellhorn has had an opportunity to try. The ninth one is a bis-(P-chloroethyl)amme with a methoxyl pyridoxine as the third radical. Certainly, all of those compounds are inadequate clinically, but they have shown some benefit. I am not just sure how much you want to have achieved before you consider they are worth while testing clinically. Those have only shown inhibition in the growth of mouse tumors. 1 think their clinical limitations have been shown by the mouse tumors, in that the effects were temporary and there was not a large therapeutic index. Also, initiation of treatment could not be delayed long before activity against the tumor would be lost. However, somewhat consistent with your idea of needing definite regressions, one of the antifolics and the other two compounds have caused regression in at least one rat tumor. That meets your criterion for selection of compounds, but still another tumor has shown a marked inhibition with these compounds but no regression. This is a rather small number of compounds to argue about or to base our arguments on, but I did want to bring out this point again of differences in the responses of different tumors. Thus with certain tumors you certainly might

415 want to take regression as the criterion, I think there might be certain tumors or certain test conditions in which you would not need to get regression before considering the compounds for clinical trials. If you want one which is really going to be a cure, then 1 would agree with you that remarkable anti-tumor effects must be observed in the experimental tumors.

,-c-< Cl Cl H-C-C-d Cl Cl

First Symposium on Chemical-Biological Correlation, May 26-27, 1950 Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

First Symposium on Chemical-Biological Correlation is a summary of a symposium held on May 26-27, 1950 by the Chemical-Biological Coordination Center. The purpose of the symposium was to bring together scientists trained in chemistry and biology for discussion of problems concerned with the effect of structure of chemicals on their biological activity and the mechanism of such actions.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!