Monkey׳S Short-Term Auditory Memory Nearly Abolished by Combined

BRES : 44603 Model7 pp: 2210ðcol:fig: : NILÞ

brain research ] ( ]]]]) ]]]– ]]]

Available online at www.sciencedirect.com 121 122 123 124 125 126 www.elsevier.com/locate/brainres 127 128 129 Review 130 131 132 Monkey's short-term auditory memory nearly 133 134 abolished by combined removal of the rostral 135 136 superior temporal gyrus and rhinal cortices 137 138 a b b 139 Jonathan B. Fritz , Megan Malloy , Mortimer Mishkin , Richard b,n 140 Q1 C. Saunders 141 Q2 aNeural Systems Laboratory, Center for Acoustic and Auditory Research, Institute for Systems Research, University of 142 Maryland, College Park, MD 20742, United States 143 bLaboratory of Neuropsychology, National Institute of Mental Health, NIH, Bethesda, MD 20892, United States 144 145 146 article info abstract 147 148 Article history: While monkeys easily acquire the rules for performing visual and tactile delayed 149 Accepted 7 December 2015 matching-to-sample, a method for testing recognition memory, they have extraordinary 150 difﬁculty acquiring a similar rule in audition. Another striking difference between the 151 Keywords: modalities is that whereas bilateral ablation of the rhinal cortex (RhC) leads to profound 152 Auditory impairment in visual and tactile recognition, the same lesion has no detectable effect on 153 Cortex auditory recognition memory (Fritz et al., 2005). In our previous study, a mild impairment 154 Monkey in auditory memory was obtained following bilateral ablation of the entire medial temporal 155 Recognition lobe (MTL), including the RhC, and an equally mild effect was observed after bilateral 156 Memory ablation of the auditory cortical areas in the rostral superior temporal gyrus (rSTG). In order 157 Echoic to test the hypothesis that each of these mild impairments was due to partial disconnec- 158 tion of acoustic input to a common target (e.g., the ventromedial prefrontal cortex), in the 159 current study we examined the effects of a more complete auditory disconnection of this 160 common target by combining the removals of both the rSTG and the MTL. We found that 161 the combined lesion led to forgetting thresholds (performance at 75% accuracy) that fell 162 precipitously from the normal retention duration of 30 to 40 s to a duration of 1to2s, 163 thus nearly abolishing auditory recognition memory, and leaving behind only a residual 164 echoic memory. 165 This article is part of a Special Issue entitled SI: Auditory working memory. 166 & 2015 Published by Elsevier B.V. 167 168 169 170 171 181 172 182 173 183 174 184 n 185 175 Correspondence to: Laboratory of Neuropsychology, NIMH, Bldg 49, Rm 1B80, Bethesda, MD 20892, United States. Fax: þ301 402 0046 176 E-mail addresses: [email protected] (J.B. Fritz), [email protected] (M. Malloy), [email protected] (M. Mishkin), 186 177 [email protected] (R.C. Saunders). 187 178 188 http://dx.doi.org/10.1016/j.brainres.2015.12.012 189 179 0006-8993/& 2015 Published by Elsevier B.V. 180 190

Please cite this article as: Fritz, J.B., et al., Monkey's short-term auditory memory nearly abolished by combined removal of the rostral superior temporal gyrus and rhinal cortices. Brain Research (2015), http://dx.doi.org/10.1016/j.brainres.2015.12.012 BRES : 44603

2 brain research ] ( ]]]]) ]]]– ]]]

191 Contents 251 192 252 193 1. Introduction...... 2 253 194 2. Results ...... 3 254 195 2.1. Histology – extent of actual lesions ...... 3 255 196 2.2. Behavioral effects ...... 4 256 197 2.2.1. Preoperative training and memory assessment ...... 4 257 198 2.2.2. Postoperative retraining and memory assessment ...... 4 258 199 2.2.3. Postoperative training and assessment ...... 4 259 200 2.2.4. Postoperative training and assessment ...... 4 260 201 2.2.5. Postoperative performance summary...... 5 261 202 3. Discussion ...... 5 262 203 4. Experimental Procedure...... 7 263 204 4.1. Subjects...... 7 264 205 4.2. Apparatus ...... 7 265 206 4.3. Stimuli ...... 7 266 207 4.4 Preoperative training and memory assessment ...... 7 267 208 4.4.1. Stage 1 (touch to sound location) ...... 7 268 209 4.4.2. Stage 2 – (spatial to non-spatial cue and differential touch to match or non-match) ...... 7 269 210 4.4.3. Stage 3 (criterion training on memory task)...... 8 270 211 4.4.4. Stage 4 ...... 8 271 212 4.5. Surgery ...... 8 272 213 4.5.1. rSTGþRhC lesion ...... 8 273 214 4.6. Histology ...... 8 274 215 Uncited references ...... 8 275 216 Acknowledgments ...... 8 276 217 References ...... 8 277 218 278 219 279 220 1. Introduction training, the delay intervals between sample and test stimuli 280 – 221 last no more than 1 2 s. Even this achievement, however, 281 222 Monkeys easily form long-term memories in vision and touch requires a training period of one to two years (Fritz et al., 282 223 (Mishkin, 1978; Murray and Mishkin, 1983) but not in audition 2005), in contrast to the few weeks they require to learn the 283 224 (Fritz et al., 2005). Monkeys also employ working memory same rule in vision (Delacour and Mishkin, 1972; Colombo 284 ’ 225 (system that enables the temporary maintenance and manip- and D Amato, 1986) or touch (Murray and Mishkin, 1998). 285 226 ulation of new or previously acquired information necessary Although the training period is much longer for auditory 286 227 to guide behavior) in vision and touch (Murray and Mishkin, memory tasks, we believe there is still great value in using 287 228 1984) but, again, apparently not in audition (Scott et al., 2012, this arduous training procedure and task learning of the 288 229 2013). The two sets of results are consistent with the hypoth- auditory recognition memory task for ablation studies. 289 230 esis that working memory depends on the short-term activa- Despite the extensive training required, our lesion studies 290 231 tion of a long-term memory trace (Cowan, 1984), and have shown differential behavioral effects of rhinal lesions, 291 232 therefore, lacking the long-term trace in audition, monkeys MTL lesions, and rSTG lesions (Fritz et al., 2005). These 292 233 necessarily lack auditory working memory as well. An intri- studies, together with the work of others (such as Wright 293 234 guing possible parallel is that human auditory recognition and Rivera, 1997; Wright, 1999; D’Amato and Salmon, 1984; 294 235 memory is also inferior to visual recognition memory (Cohen Colombo and D’Amato, 1986; Colombo et al. 1990, 1996; Scott 295 236 et al. 2009, 2011; Bigelow and Poremba, 2014), although et al. 2012, 2013, 2014) have provided us with insights into the 296 237 human verbal and musical memory (enriched by acoustic- neuropsychology of auditory memory that would not have 297 238 semantic associations) can clearly be prodigious. Despite the been discovered without the use of this or similar auditory 298 239 apparent absence of both forms of memory in pure audition recognition memory tasks. 299 240 (working memory or long-term memory), monkeys do have Following extensive training, once having finally acquired 300 241 excellent associative auditory memory (involving crossmodal the auditory DMS rule for delays of 1–2 s, monkeys can apply 301 242 links between auditory and other sensory stimuli or motor it with about 75% accuracy but only up to delays of 30–40 s – 302 243 actions (e.g. Colombo and Graziano, 1994)). Monkeys also do their “forgetting threshold” in audition (Fritz et al., 2005) – in 303 244 have some form of short-term auditory memory (Wright and contrast to their forgetting thresholds for trial-unique stimuli 304 245 Rivera, 1997). For example, after extensive training, they can of 10 min in touch (Buffalo et al., 1999) and 20 min in 305 246 perform a short-term, recency memory task with two tonal vision (Murray and Mishkin, 1998). We have therefore pro- 306 247 stimuli (Colombo et al., 1990, 1996) or recall and compare the posed that the monkey's extremely short retention period in 307 248 Q3 rates of two tonal flutter stimuli (Lemus et al., 2009a, 2009b). audition is due to a weak sensory trace that dissipates rapidly 308 249 They can also learn the rule for delayed matching-to-sample and inexorably, reflecting a failure to encode or retrieve an 309 250 (DMS) with trial-unique sounds, provided that, during initial acoustic trace in long-term memory (Scott et al., 2012, 2013). 310

brain research ] ( ]]]]) ]]]– ]]] 3

311 Studies examining one-trial memory in vision and touch uncinate fasciculus and is sectioned by the MTL removal at a 371 312 have identified the rhinal cortices (i.e. perirhinal and entorh- level just dorsal to the amygdala. One alternate auditory 372 313 inal areas) as essential for this mnemonic function in both of projection to the ventromedial PFC may arise in more caudal 373 314 those sensory modalities (e.g. Murray and Mishkin, 1984; portions of the STG (cSTG) and may be relayed through the 374 315 Meunier et al., 1993; Suzuki et al., 1993; Buffalo et al., 1999), parahippocampal cortex and RhC (Suzuki et al., 1993). In the 375 316 with increasing impairments as delays increase. In audition, absence of the more rostral rSTG projections, these more 376 317 by contrast, bilateral removal of monkeys' rhinal cortices was caudal indirect pathways may be sufficient to support some 377 318 found to have no statistically significant effect at any delay level of memory performance. 378 319 (Fritz et al., 2005). This lack of impairment is somewhat To test this possibility and, to examine the consequences 379 320 surprising as there are extensive projections from the rostral of removing both auditory cortical inputs to the ventromedial 380 321 superior temporal cortical areas to the medial temporal PFC, or to any other target that is common to both sources, 381 322 Q4 rhinal cortical region (Muñoz-Lopez et al., 2015). Furthermore, we examined the effects on auditory short-term memory of 382 323 whereas ablation of the entire medial temporal lobe (MTL) or combining the rSTG and RhC lesions. 383 324 of the rostral third of the superior temporal gyrus (rSTG) 384 325 produced significant and equally mild auditory memory 385 326 impairments at all delays (Fritz et al., 2005), removal of either 2. Results 386 327 387 area alone failed to eliminate auditory memory. Other studies 328 388 (Colombo et al., 1990, 1996) have reported more severe 2.1. Histology – extent of actual lesions 329 389 impairment of auditory memory following much larger 330 390 lesions of auditory cortical areas that spared only primary The actual lesions in Monkeys A and K, reconstructed from 331 391 auditory cortex and some adjacent lateral belt but removed serial sections of the histological slides, are shown in Fig. 1. 332 392 the entire medial belt areas and all of the STG and adjacent The lesions were largely as expected with the rostral 15 mm 333 393 sulcal areas. of the rSTG removed. The rhinal cortex removal was near 334 394 The likely explanation for the auditory memory impair- complete with minor sparing of the most caudal extent of the 335 395 ment after the MTL ablation is provided by a neuroanatomi- entorhinal cortex. In Monkey A the rSTG removal was some- 336 396 cal study (Muñoz et al., 2009), which demonstrated that what asymmetrical with the right hemisphere slightly larger 337 397 aspirating this tissue transects a direct auditory projection than in the left extending back an additional 2 mm. In 338 398 from the rSTG to the ventromedial prefrontal cortex (PFC). addition, Monkey A sustained unintended damage bilaterally 339 399 This projection courses through the medial portion of the to the lower bank of the STS and to the ventral portion of the 340 400 341 401 342 402 343 403 344 404 345 405 346 406 347 407 348 408 349 409 350 410 351 411 352 412 353 413 354 414 355 415 356 416 357 417 358 418 359 419 360 420 361 421 362 422 363 423 364 424 365 Fig. 1 – The column on the far left shows the extent of the intended combined rSTG plus RhC lesion on lateral surface views 425 366 and in coronal serial sections. The two columns on the right show the limits of the lesions from Monkeys A and K. Histological 426 367 analyses show the lesion to be complete as intended with the rhinal cortex removed bilaterally in both cases. In Monkey A 427 368 the rSTG lesion extended caudally in the right hemisphere approximately 2 mm more than the left. In Monkey A the most 428 369 rostral 3 mm of area TE was removed in the left hemisphere and the lower bank of the superior temporal sulcus was also 429 370 included bilaterally. 430

4 brain research ] ( ]]]]) ]]]– ]]]

431 temporal pole. In both cases the rSTG removal and the rhinal 300 491 432 removal were comparable to those reported in Fritz et al. 492 433 (2005). 493 434 250 494 435 2.2. Behavioral effects 495 436 496 437 497 2.2.1. Preoperative training and memory assessment 200 438 Both monkeys learned the auditory memory tasks through a 498 439 series of intermediate shaping tasks including: Stage 1 – 499 440 touch to sound location, Stage 2 – changing touch from 500 150 441 spatial to non-spatial cues-and differential touch for match 501 442 – 502

and non-match trials, Stage 3 Criterion training on the Sessions 443 Memory Task for fixed delay (2 s). Monkeys A and K took a 503 444 total of 139 and 218 sessions, respectively, to complete all the 100 504 445 Task Stages 1–3 (or 10 K trials on average) and thereby 505 446 demonstrated memory for auditory stimuli after a delay of 2 s 506 447 507 (see Section 4.4 for further details). For comparison, preo- 50 448 perative learning of Stages 1–3 by the nine monkeys in the 508 449 study by Fritz et al. (2005) required an average of 200 509 450 sessions (12 K trials). Consistent with that earlier study, 510 451 0 511 learning at Stage 2 took about twice the number of sessions Rh rSTG Rh + rSTG 452 that were required to learn at each of the two other training 512 453 513 stages (i.e., stages 1 and 3). Preoperative progression through Fig. 2 – Total number of sessions postoperatively for all 454 514 all three training stages did not differ significantly from the stages to attain criterion of 85% correct responses for two 455 515 rate of learning of the nine monkeys in the earlier auditory consecutive days. Monkeys with RhC lesions (n¼4) and with 456 516 memory study (Fritz et al., 2005). rSTG lesions (n¼2) relearned the basic task (14 sessions and 457 517 57 sessions on average, respectively) with post testing at 458 518 2.2.2. Postoperative retraining and memory assessment stage 2. Neither of the two monkeys with combined RhC and 459 519 As in the earlier study (Fritz et al., 2005), postoperative testing rSTG lesions (n¼2) were able to reach criterion performance 460 520 was begun with a 2-s intrapair interval (IPI), but it quickly on Stage 2 with the 2 s delays. In attempts to retrain the 461 521 became evident that neither animal with combined animals on the task they received hundreds of additional 462 522 rSTGþRhC lesions could reliably retain the auditory stimulus sessions at shorter delays. Despite this extensive training 463 523 even across this short delay. In contrast, monkeys with single they both failed to relearn the task to criterion with retention 464 524 lesions (Rhinal cortex or rSTG alone) relearned the task (Fritz delays as short as 125 ms. 465 525 et al., 2005). When it became clear the monkeys with the 466 526 combined lesions were not capable of relearning the task at 467 25 sessions. Further, when returned again to Stage 2, Monkey 527 the 2-s IPI, an attempt was made to retrain them on the basic 468 A was unable to reattain criterion in 81 sessions. 528 task. Hence, they were returned to training on earlier stages 469 We then presented blocks of trials at delays that decreased 529 470 of the task with much shorter IPIs, each monkey progressing gradually from 1000 to 125 ms in order to determine whether 530 471 on its own schedule according to its level of performance. Monkey A might succeed in attaining criterion at shorter IPIs. 531 472 Below is a description of each monkey's postoperative As shown in Table 1, Monkey A failed to reach criterion at 532 473 assessment. Fig. 2 compares the number of sessions of the every delay, but did achieve an average performance level of 533 474 monkeys with either rSTG or RhC lesions alone to success- 82% correct at delays of 125 and 250 ms, suggesting that the 534 475 fully relearn the task at Stage 2 with the total number of monkey retained knowledge of the basic DMS task, but was 535 476 training sessions of the monkeys with the combined rSTG unable to retain the memory of a specific auditory stimulus 536 477 and RhC lesion (at all post-operative stages, including stage 2) for an IPI of more than a few hundred milliseconds. 537 478 that were unsuccessful in relearning the task. 538 479 2.2.4. Postoperative training and assessment 539 480 2.2.3. Postoperative training and assessment Monkey K Monkey K was also started postoperatively at the 540 481 Monkey A On the first day of retraining, Monkey A failed to 2-s delay (Stage 3 – Criterion test of auditory memory). Unlike 541 482 emit any response to the auditory stimuli and so was Monkey A, this animal demonstrated retention of the operant 542 483 returned to preoperative Stage 1 (Touch to sound location). rule by responding appropriately to the stimulus presenta- 543 484 This was completed in 10 sessions. On Stage 2 (Spatial to non- tions from the beginning of testing. Nevertheless, his scores 544 485 spatial cues) (with R-speaker volume gradually attenuated across 71 sessions were no higher than those of Monkey A 545 486 from full volume to zero), criterion was reached in 33 sessions (mean, 64% correct responses; range, 48–80%). Because of the 546 487 (compared to 97 sessions preoperatively). However, on Stage experience of prolonged efforts to retrain Monkey A with a 547 488 3 (Criterion test of auditory memory task at 2-s delay) (F lack of success in task relearning at the 2 s IPI, we terminated 548 489 speaker only, match/nonmatch), Monkey A failed to reach further training of Monkey K when it became clear that there 549 490 criterion in 43 sessions having succeeded preoperatively after was no evidence of improvement in performance. Hence, the 550

brain research ] ( ]]]]) ]]]– ]]] 5

551 difference in the number of trials/sessions each rSTGþRhC auditory DMS task that the rSTGþRh monkeys had in com- 611 552 monkey received at the 2 s delay was a result of individual parison with the single lesion groups (rSTG or Rh alone) 612 553 adjustments in training based on postoperative performance. studied earlier (Fritz et al., 2005) with the average perfor- 613 554 Monkey A received 157 sessions at the 2 s delay, as this mance of the Rh lesion group is shown for comparison. 614 555 monkey required an extended period of retraining through 615 556 the basic stages of the task. 616 557 Monkey K was then tested on the same set of short, 3. Discussion 617 558 gradually decreasing delays from 1000 to 125 ms, again with 618 559 overall results similar to those of Monkey A. Following combined bilateral removal of the rostral superior Q5 619 560 temporal gyrus (rSTG) and rhinal cortices (RhC), neither of the 620 561 2.2.5. Postoperative performance summary two monkeys that had been trained preoperatively on the 621 562 The left panel of Fig. 3 shows the average percent correct auditory DMS task were able to perform the task reliably even 622 563 responses for each monkeys' performance in the last six after a year of postoperative training (9000 trials). Further- 623 564 sessions on the variable short-delay set. The performance at more, in comparison with monkeys given rhinal lesions 624 565 each delay, for each monkey, was never higher than 78% and alone, post-operative forgetting thresholds (the duration of 625 566 the average performance for both monkeys across all delays the delay at which they scored 75% correct responses) 626 567 was fairly flat (69.4%). As described above, the monkeys with plummeted from a delay of 30 s post-operatively for the 627 568 rSTGþRhC lesions had great difficulty in achieving criterion rhinal monkeys (Fritz et al., 2005) to just a few hundred ms 628 569 (defined as 85% correct overall on two successive days, with a postoperatively for the animals with combined lesions, as 629 570 minimum of 80% correct on same and different trials) at shown in Fig. 3. 630 571 delays of 1 s, and neither achieved criterion at 2 s. Fig. 3 (right In a previous study of auditory memory in monkeys (Fritz 631 572 panel) documents the greater impairment in performing the et al., 2005), we observed only a mild deficit on the DMS task 632 573 after ablation of the rSTG and no deficit after the RhC alone. 633 574 Table 1 – The table shows the number of sessions and the The devastating deficit in auditory memory described in the 634 575 range and average performance level attained by the two current study is consistent with those of an earlier set of 635 576 monkeys with the combined rSTG and RhC lesion during experiments (Colombo et al., 1990, 1996), which found severe 636 criterion training with short IPIs (125–1000 ms). Even with 577 impairment in the auditory memory of monkeys with exten- 637 a short IPI of only 125 ms the monkeys were not able to 578 638 relearn the basic task at preoperative criterion levels. sive lesions of auditory association cortex. In that study, three 579 cebus monkeys (Cebus paella) were trained preoperatively on 639 580 IPI Inter- Sessions Trials Range Average an auditory recency memory (DMS) task in which the stimuli 640 581 pair performance were restricted to two widely separated pure tones, high vs 641 Interval percent correct 582 low. After achieving criterion, the monkeys received a bilateral 642 583 Monkey A 1000 10 562 52–80 73 lesion that included most of the superior temporal gyrus, 643 584 500 39 2493 66–88 78 sparing only the primary auditory cortex (area A1), the adja- 644 585 250 10 567 76–87 82 cent lateral belt, and the most rostral 5 mm of the temporal 645 – 586 125 17 370 72 92 82 pole. Only one of the three monkeys succeeded in re-attaining 646 Monkey K 1000 18 760 65–82 75 587 criterion on the baseline DMS task, which required bridging an 647 500 10 476 64–88 77 588 intrapair interval of only 500 ms. Comparison of those results 648 250 16 730 67–90 76 589 125 53 2452 63–83 70 with ours suggests that the mild behavioral effect of a rSTG 649 590 lesion can be increased dramatically by adding removal of 650 591 651 592 652 593 653 594 654 595 655 596 656 597 657 598 658 599 659 600 660 601 661 602 662 603 663 604 664 605 665 606 Fig. 3 – The performance of the two monkeys A and K with the combined rSTG and RhC lesions (left panel) shows the average 666 607 percent correct responses for each monkeys' performance for the last six sessions on the variable short-delay delays (125– 667 608 4000 ms). Performance across delays remains constant (average of 69.4 for both monkeys) despite the increasing delays. In 668 609 comparison (right panel), is a plot of average performance of the rhinal lesion monkeys (n¼4) with increasing delays that 669 610 shows a forgetting curve with performance of over 85% at 10 s delay and 75% at 30 s delay. 670

6 brain research ] ( ]]]]) ]]]– ]]]

671 either the cSTG or alternatively, the RhC. In either case it is auditory core and its projections to the belt and parabelt 731 672 unlikely that the deficit observed after the combined lesions is regions in cSTG. If so, it can be predicted that bilateral 732 673 a result of a perceptual defect. First of all the monkeys with removal of the ventromedial PFC should lead to a form of 733 674 singular lesions of the rSTG performed relatively well at short ephemeral auditory memory similar to that produced by 734 675 delays and monkeys with the complete RhC lesions had no combined ablation of the rSTG and the RhC. Recently, it has 735 676 effect on performance. Furthermore, in visual recognition been shown (Plakke et al., 2015) that inactivation of the 736 677 memory in macaques, the contribution of the RhC to memory ventrolateral PFC impairs auditory working memory (and 737 678 is well established (Mishkin, 1978; Zola-Morgan et al., 1989; auditory-visual working memory but not visual working 738 679 Meunier et al., 1993; Malkova et al., 2001; Buffalo et al., 2000) memory). 739 680 and can be distinguished and dissociated from the perceptual One of several alternative possibilities, is that both the 740 681 deficits that occur after lesions of visual area TE (Buffalo et al., rSTG and the rhinal cortex form short-term auditory mem- 741 682 2000). Thus we think it is unlikely the defect observed here is a ories irrespectively of their projection to the ventromedial 742 683 result of a perceptual deficit. prefrontal cortex. According to this hypothesis, a lesion of 743 684 In the previous lesion study of auditory recognition mem- one of these sites would determine a mild disturbance of 744 685 ory in monkeys (Fritz et al., 2005), we found no impairment memory processes while a combined lesion of both struc- 745 686 after selective removal of the rhinal cortices (RhC) (Fig. 3 – tures could dramatically affect auditory recognition memory. 746 687 right panel). This negative result had not been expected given However, this explanation seems unlikely to be true for our 747 688 the critical role of the RhC in both visual and tactile recogni- studies in the rhesus monkey. Although we have previously 748 689 tion. However, a similar, negative, result in auditory memory described a mild memory deficit after rSTG lesion alone, we 749 690 after RhC lesions had also been observed in dogs, even when found no observable behavioral deficit in memory following 750 691 the RhC lesion was combined with removal of the entire combined lesions of the perirhinal cortex and parahippocam- 751 692 medial temporal lobe, including the parahippocampal cortex pus and hippocampus in the monkey (Fritz et al., 2005). 752 693 (Kowalska et al., 2001). These negative findings were particu- Recently, auditory afferent-driven synaptic activation has 753 694 larly puzzling given the evidence obtained by Suzuki and been demonstrated in rodent perirhinal cortex (Kotak et al., 754 695 Amaral (1994) that area TH of the parahippocampal cortex 2015) and previous studies have shown that auditory inputs 755 696 receives substantial auditory inputs from the lateral surface play a role in auditory fear conditioning memory in rodents 756 697 of the superior temporal gyrus as well as from the polymodal (Lindquist et al., 2004; Bang and Brown, 2009). However, while 757 698 auditory-visual areas on the ventral bank of this gyrus. In the it is certainly likely that there are also some auditory inputs 758 699 previous study on monkeys (Fritz et al., 2005) we found only to rhinal cortex in monkeys, our evidence is that they do not 759 700 mild impairment after lesions of either the medial temporal appear to play a significant role in auditory recognition 760 701 lobe (MTL) or rSTG lesions alone. Neuroanatomical studies memory. 761 702 suggest that the effect of the MTL ablation, which included Whatever residual auditory sensory memory was 762 703 both the amygdala and the hippocampus, may have been observed in the rSTGþRhC lesioned monkeys may be related 763 704 due, at least in part, to disconnection of rSTG from one of its to the form of brief auditory “sensory memory” termed 764 705 major target areas in the prefrontal cortex (Muñoz et al., “echoic” in humans (Neiser, 1967). There is a striking simi- 765 706 2009). Aspiration of the rostral amygdala transects rSTG fibers larity between this “echoic” auditory memory in humans and 766 707 that form the medial part of the uncinate fasciculus, which the residual auditory memory in the monkeys examined 767 708 project to the ventromedial prefrontal cortex. Second, aspira- postoperatively in the current study. Using short-duration 768 709 tion of the rostral (or Pes) hippocampus transects rSTG fibers individual stimuli (300–500 ms) psychoacoustic studies in 769 710 that cross the ventrocaudal neostriatum on their way to humans have shown the presence of a transient auditory 770 711 medial thalamus. Thus, the effects of removing either the memory store of about 2–4 s in duration (Darwin et al., 1972) 771 712 MTL or the rSTG may be explained, at least in part, by as well as a short-term auditory memory store for frequency, 772 713 disconnection of auditory inputs to the ventromedial PFC. timbre, and loudness with a temporal span of about 1 s 773 714 The RhC lesion, while having no apparent behavioral effect (Clement et al., 1999; Demany and Semal, 2008). Brain 774 715 on its own (Fritz et al., 2005), also disrupts some of the damage in humans (extensive unilateral damage of the left 775 716 indirect projections to the ventromedial PFC, from caudal suprasylvian areas) can also lead to severe auditory memory 776 717 auditory cortex. In non-human primates, a possible indirect impairments that spare only this residual echoic memory 777 718 projection was shown in Suzuki and Amaral's (1994) neuroa- (Kojima et al. 2012). 778 719 natomical study, which demonstrated the presence of a Recent neurophysiological studies in PFC and rSTG of 779 720 substantial projection from cSTG to the parahippocampal monkeys performing an auditory DMS task have found clear 780 721 cortex. evidence for auditory memory-encoding cells. For example, 781 722 Thus, the striking deficit observed in the present study match suppression has been observed during performance of 782 723 may be the result of having transected two or more different an auditory DMS task in monkey dorsal temporal pole (Ng 783 724 pathways to the ventromedial PFC (Hackett et al., 1999; et al., 2014, Scott et al., 2012, 2013, 2014) and memory-related 784 725 Romanski et al., 1999a, 1999b, Carmichael and Price, 1995; responses have been described in Area 46 of lateral PFC 785 726 Kondo et al., 2003; Saleem et al., 2008; Muñoz et al., 2009; (Plakke et al., 2013). Auditory memory responses have also 786 727 2010; Plakke and Romanski, 2014; Medalla and Barbas, 2014) been observed in ventrolateral PFC (Romanski, 2012) and we 787 728 that each may play key roles in auditory memory. It seems would predict will be found in ventromedial PFC. 788 729 likely that the monkeys' residual ephemeral auditory mem- In summary, our results demonstrate that the more 789 730 ory depended on the limited memory store provided by the complete disconnection of auditory cortex from frontal cortex 790

brain research ] ( ]]]]) ]]]– ]]] 7

791 by removing both direct (rSTG) and indirect (RhC) inputs, About every 10 days of training, all 964 stimuli were repeated 851 792 devastates auditory recognition memory. However, the com- in a re-randomized order of presentation. 852 793 bined disconnection does reveal the residual presence of a 853 794 very short-term “echoic” auditory memory, likely to be 4.4 Preoperative training and memory assessment 854 795 encoded in the core and caudal belt areas of auditory cortex, 855 796 which store sounds for only 1-2 s. This brief, local, auditory Monkeys were trained on the auditory delayed match-to- 856 797 store is normally extended by a short-term memory system sample (DMS) task described by Fritz et al. (2005). The 857 798 that relies both on direct auditory projections to the pre- monkeys learned the DMS rule with a sample-to-test delay 858 799 frontal cortex and on indirect ones relayed through the rhinal of 2 s across three stages of training and then the retention 859 800 cortices. interval was increased in 1 s steps from 2 s to 5 s. Monkeys 860 801 had to reach criterion before advancing to each increased 861 802 retention interval. 862 803 4. Experimental Procedure 863 804 4.4.1. Stage 1 (touch to sound location) 864 805 4.1. Subjects This stage had two steps: The first was classical conditioning, 865 806 in which each successive sound, presented pseudorandomly 866 807 The subjects were two experimentally naïve, male rhesus through either the F or the R speaker, was paired with a 867 808 monkeys (Macaca mulatta), monkey A weighing 4 kg and banana pellet reward; the second step was operant condi- 868 fi 809 monkey K, 8 kg. They were ve (A) and seven (K) years of tioning, in which the animal received the reward by tapping 869 810 age at the onset of the study. They were fed a diet of Purina the touch plate attached to the speaker emitting the sound. 870 811 monkey chow (No. 5038, PMI Feeds, St. Louis) supplemented An overhead light dimmed at sound onset and brightened at 871 812 with fruit. All procedures were carried out in accordance with sound offset. During initial training, the interstimulus inter- 872 813 the Institute of Medicine's Guide for the Care and Use of val (ISI) varied from 3–30 s, and responses during this interval 873 814 Laboratory Animals and under a protocol approved by the caused the ISI to reset. Training continued until the animal 874 815 Animal Care and Use Committee of the National Institute of responded to 90% of the sounds in one session. Monkey A 875 816 Mental Health (NIMH). completed this stage in 27 sessions, and monkey K, in 31 876 817 The monkeys in this study were motivated to work for the sessions. 877 fl 818 avored food-pellet treats by controlling the amount and 878 819 timing of food intake. We used preferred treats for reinforcers 4.4.2. Stage 2 – (spatial to non-spatial cue and differential 879 820 as it is our experience that monkeys will work for these while touch to match or non-match) 880 821 minimizing any food control. The monkeys were highly Use of the light cue was discontinued, and a pair of acoustic 881 822 motivated, eager to perform the task, and showed consistent stimuli (sample and test, separated by a 2-s interval, the 882 823 alertness, interest and motivation throughout the daily block intrapair interval, IPI) were now presented through one or the 883 824 of 60 trials. other speaker, the animal being required to respond only to 884 825 the second, or test, sound. Thus, the Stage 2 task was 885 826 4.2. Apparatus fundamentally different from Stage 1 since the monkeys 886 827 now had to learn to distinguish between the two sounds 887 828 Automated testing of an auditory version of trial-unique DMS (sample and test) and respond only to the second sound of 888 829 took place in a sound-attenuated chamber. The monkeys the pair. Matching sound pairs were always presented 889 830 were trained to sit in a primate chair and respond to acoustic through the F speaker, and the correct response was a tap 890 831 stimuli by touching one of two small touch-sensitive copper on the F touch plate; nonmatching sound pairs were always 891 2 832 plates (10 cm ), each mounted on the front of a speaker presented through the R speaker, and the correct response 892 833 (Audiotex 30-5121, Rockford, IL). Each speaker was located was a tap on the R touch plate. The two trial types were 893 834 40 cm from the monkey, one directly in front (F speaker) and presented pseudorandomly (Gellerman, 1933), with the 2-s 894 835 the other 20 cm to the right (R speaker). Below each speaker/ IPI, and a variable interpair delay of 10–20 s. Up to this point, 895 836 touch plate was a food well into which a high-preference the animal could simply continue to follow, and respond to, 896 837 reward (a 190-mg, banana-flavored pellet) was automatically the speaker that was the source of the sound. However, after 897 838 dispensed immediately after a correct response. 8–10 sessions of this kind, the nonmatching sound pairs, 898 839 which had been presented only through the R speaker, were 899 840 4.3. Stimuli now presented through both F and R speakers simulta- 900 841 neously, though, initially, at a very low volume through the 901 842 The sound library consisted of 964 distinct acoustic stimuli, F speaker. After another 8–10 sessions, the volume of the 902 843 each 2 s long and presented at a sound pressure level of 70– second sound in the nonmatching sound pairs presented 903 844 75 dB. These acoustic stimuli included a highly diverse set of through the F speaker was gradually increased until the 904 845 animal sounds, musical segments, and FM sweeps, as well as sounds were played at equal volume through both speakers. 905 846 both natural and synthetic environmental sounds. Because During this period, the monkey continued to be rewarded for 906 847 each monkey typically received 60 trials/day (5–6 days/week), tapping the F touch plate to the second stimulus of matching 907 848 90 different stimuli were required for each DMS session (60 sound pairs only, and for tapping the R touch plate only to 908 849 sample sounds, 30 of which were presented again as the the nonmatching sound pairs, with training continuing until 909 850 matching sounds, and 30 different or nonmatching sounds). the animal reached 90% correct responses in a single session. 910

8 brain research ] ( ]]]]) ]]]– ]]]

911 When that performance criterion was achieved, the sound provided as required, in consultation with the NIMH 971 912 volume through the R-speaker for both matching and non- veterinarian. 972 913 matching sound pairs was gradually reduced across succes- 973 914 sive 10-trial blocks, according to the following ‘staircase’ rule: 4.5.1. rSTGþRhC lesion 974 915 R-speaker volume was reduced one step after a score of 9/10 The rSTG ablation covered the rostral 15–17 mm of the 975 916 or 10/10; it was increased one step after a score of 7/10 or superior temporal gyrus, including, within these limits, the 976 917 lower; and it remained unchanged after a score of 8/10. ventral bank of the lateral sulcus and the dorsal bank of the 977 918 Training proceeded in this fashion until the animal superior temporal sulcus. The removal thus encompassed 978 919 responded correctly with a score of 9/10 or 10/10, after the auditory recipient cytoarchitectonic areas TPO, PGa, IPa, TAa, 979 920 volume of the R speaker had been reduced to zero. At this Ts1, Ts2, and most of Ts3 of Seltzer and Pandya (1978), as well 980 921 final level, all sounds now emanated only from the F speaker. as the dorsolateral part of their area Pro in the temporal pole. 981 922 The ITI (the intertrial interval between successive pairs of The rSTG projects to the frontal cortex (Romanski et al., Q6 982 923 sounds) was fixed at 20 s for the remainder of training and 1999a, 1999b) and also provides the majority of the superior 983 924 testing. Monkey A completed stage 2 in 97 sessions, and temporal projections to the perirhinal and entorhinal cortices 984 925 monkey K, in 146 sessions. (Pandya and Sanides, 1973; Amaral et al., 1983; Suzuki and 985 926 Amaral, 1994; Pandya, 1995; Muñoz-Lopez et al., 2015). For the 986 927 4.4.3. Stage 3 (criterion training on memory task) combined lesion, the rSTG removal was then extended 987 928 All sounds continued to be presented through the F speaker ventromedially to include the rhinal cortices (RhC); (i.e., the 988 929 only, just as at the end of Stage 2, so that sound spatial source entorhinal and perirhinal areas, including the perirhinal 989 930 no longer cued responses. The correct response on match extension into the dorsomedial temporal pole area 36p-dm 990 931 trials remained a tap on the F touch plate, and on nonmatch of Insausti et al. (1987) together witth Brodmann's areas 36, 991 932 trials, a tap on the R touch plate, following the presentation of 35, and 28). An illustration of the intended lesion is shown in 992 933 the second sound in the pair. Training with the 2-s intrapair, Fig. 1. 993 934 or retention, interval continued until the animals met the 994 935 criterion of 85% correct responses, with a minimum of 80% 4.6. Histology 995 936 correct on each trial type (matched pairs and nonmatched 996 937 pairs), for two consecutive sessions. Monkey A met the On completion of postoperative testing, the monkeys were 997 938 criterion on this stage in 25 sessions, and monkey K, in 41 injected with a lethal dose of sodium pentobarbital and 998 939 sessions. perfused transcardially with saline followed by 10% formalin. 999 940 The brains were blocked in the coronal plane, removed, 1000 941 4.4.4. Stage 4 photographed, cryoprotected through a series of glycerol 1001 942 IPI, or retention intervals, were increased from 2 s to 5 s, in 1- solutions, frozen, and then cut in 50-mm slices on a freezing 1002 943 s steps, the animal being required to meet the same criterion microtome. Every fifth section was mounted on slides, 1003 944 as it had in Stage 3 at each increase in delay. Monkey A stained with thionin, coverslipped, and examined under the 1004 945 achieved the criterion at 5-s delays in 33 sessions, and microscope. The extent of the damage in each case was 1005 946 monkey K, in 23 sessions. plotted on drawings of a standard rhesus monkey brain at 1- 1006 947 mm intervals, and the surface views of the lesion were 1007 948 4.5. Surgery reconstructed. 1008 949 1009 950 About a week after completing preoperative training and 1010 951 testing, each monkey was pretreated for surgery with an Uncited references Q7 1011 952 antibiotic (Ditrim, 24% solution, 0.1 ml/kg i.m.), sedated with 1012 953 ketamine (10 mg/kg i.m.), anesthetized with isoflurane (1–4% Hackett, 2011; Hwang and Romanski, 2015; Kondo et al., 2005; 1013 954 to effect), and given Mannitol (30%, 30 cc i.v. over 20 min) to Muñoz-Lopez et al., 2010; Saleem et al., 2014. 1014 955 reduce brain volume. Throughout surgery, which was per- 1015 956 formed with sterile techniques, the animal was kept on a 1016 957 heating pad and hydrated (Ringer's solution i.v.). The scalp Acknowledgments 1017 958 and temporal muscle on one side were incised and reflected, 1018 959 the zygomatic arch was removed, and a large frontotemporal Funded by NIH Intramural Program. Q8 1019 960 bone flap was turned, extending from the orbit ventrally 1020 961 toward the temporal fossa and caudally to the auditory references 1021 962 meatus. The opening was enlarged ventrally to the infra- 1022 963 temporal crest, and the dura was then opened over the 1023 964 posterior frontal and anterior temporal lobes. The rSTG and Bang, S.J., Brown, T.H., 2009. Perirhinal cortex supports acquired 1024 965 the rhinal cortex were viewed through an operating micro- fear of auditory objects. Neurobiol. Learn. Mem. 92, 53–62. 1025 966 scope and removed by aspiration through a 20-gauge stain- Bigelow, J., Poremba, A., 2014. Achilles’ ear? Inferior human short- 1026 term and recognition memory in the auditory modality. PLoS 967 less-steel tube. On completion of the tissue removal, the dura 1027 One 9, e89914. fl 968 was sewn, the bone ap reattached, and the scalp wound Buffalo, E.A., et al., 1999. Dissociation between the effects of 1028 969 closed in anatomical layers. The same lesion was then damage to the perirhinal cortex and area TE. Learn. Mem. 6, 1029 970 repeated on the second side. Postsurgical analgesics were 572–599. 1030

brain research ] ( ]]]]) ]]]– ]]] 9

1031 Carmichael, S.T., Price, J.L., 1995. Sensory and premotor connec- Lindquist, D.H., Jarrard, L.E., Brown, T.H., 2004. Perirhinal cortex 1091 1032 tions of orbital and medial prefrontal cortex of macaque supports delay fear conditioning to rat ultrasonic social 1092 1033 monkeys. J. Comp. Neurol. 363, 642–664. signals. J. Neurosci. 24, 3610–3617. 1093 1034 Clement, S., Demany, L., Semal, C., 1999. Memory for pitch versus Medalla, M., Barbas, H., 2014. Specialized prefrontal “auditory 1094 memory for loudness. J. Acoust. Soc. Am. 106, 2805–2811. fields”: organization of primate prefrontal-temporal path- 1035 1095 Cohen, M.A., Horowitz, T.S., Wolfe, J., 2009. Auditory recognition ways. Front. Neurosci. 8 (77), 1–15. 1036 1096 memory is inferior to visual recognition memory. Proc. Natl. Meunier, M., Bachevalier, J., Mishkin, M., Murray, E.A., 1993. 1037 Acad. Sci. 106, 60008–60010. Effects on visual recognition of combined and separate abla- 1097 1038 Cohen, M.A., Evans, K.K., Horowitz, T.S., Wolfe, J., 2011. Auditory tions of the entorhinal and perirhinal cortex in rhesus 1098 1039 and visual memory in musicians and non-musicians. Psy- monkeys. J. Neurosci. 13, 5418–5432. 1099 1040 chon. Bull. Rev. 18, 586–591. Mishkin, M., 1978. Memory in monkeys severely impaired by 1100 1041 Colombo, M., D’Amato, M.R., 1986. A comparison of visual and combined removal of amygdala and hippocampus. Nature 1101 1042 auditory short-term memory in monkeys (Cebus apella). Q. J. 273, 297–298. 1102 Exp. Psychol. B 38, 425–448. Munoz, M., Mishkin, M., Saunders, R.C., 2009. Resection of the 1043 ˜ 1103 Colombo, M., D’Amato, M.R., Rodman, H.R., Gross, C.G., 1990. medial temporal lobe disconnects the rostral superior tem- 1044 1104 Auditory association cortex lesions impair auditory short- poral gyrus from some of its projection targets in the frontal 1045 term memory in monkeys. Science 247, 336–338. lobe and thalamus. Cereb. Cortex 19, 2114–2130. 1105 1046 Colombo, M., Graziano, M., 1994. Effect of auditory and visual Mun˜ oz-Lopez, M.M., Mohedano-Moriano, A., Insausti, R., 2010. 1106 1047 interference on auditory-visual delayed match to sample in Anatomical pathways for auditory memory. Front. Neuroanat. 1107 1048 monkeys (Macaca fascicularis). Behav. Neurosci. 108, 636–639. 4, 129. 1108 1049 Colombo, M., Rodman, H.R., Gross, C.G., 1996. The effects of Mun˜ oz-Lopez, M., Insausti, R., Mohedano-Moriano, A., Mishkin, 1109 1050 superior temporal cortex lesions on the processing and M., Saunders, R.C., 2015. Anatomical pathways for auditory 1110 1051 retention of auditory information in monkeys (Cebus apella). J. memory II: information from rostral superior temporal gyrus 1111 Neurosci. 16, 4501–4517. to dorsolateral temporal pole and medial temporal cortex. 1052 1112 Cowan, N., 1984. On short and long auditory stores. Psychol. Bull. Front. Neurosci. 9, 1–21. 1053 1113 96, 341–370. Murray, E.A., Mishkin, M., 1983. Severe tactual memory deficits in 1054 Darwin, C.J., Turvey, M.T., Crowder, R.G., 1972. An auditory monkeys after combined removal of the amygdala and hip- 1114 1055 analogue of the Sperling partial report procedure: evidence for pocampus. Brain Res. 270, 340–344. 1115 1056 brief auditory storage. Cogn. Psychol. 3, 255–267. Murray, E.A., Mishkin, M., 1984. Severe tactual as well as visual 1116 1057 Q9 Demany, L., Semal, C., 2008. The role of memory in auditory memory deficits follow combined removal of the amygdala 1117 1058 perception (Springer Handbook of Auditory Research Series). and hippocampus in monkeys. J. Neurosci. 4, 2565–2580. 1118 1059 In: In: Yost, W.A., Fay, R.R. (Eds.), Auditory Perception of Sound Murray, E.A., Mishkin, M., 1998. Object recognition and location 1119 Sources, 29. Springer, New York. memory in monkeys with excitotoxic lesions of the amygdala 1060 1120 Fritz, J., Mishkin, M., Saunders, R.C., 2005. In search of an auditory and hippocampus. J. Neurosci. 18, 6568–6582. 1061 1121 engram. Proc. Natl. Acad. Sci. 102, 9359–9364. Neiser, U., 1967. Cognitive Psychology. Prentice-Hall, Inc., Upper 1062 Gellerman, L.W., 1933. Chance orders of alternating stimuli in Saddle River, NJ. 1122 1063 visual discrimination experiments. J. Genet. Psychol. 42, Ng, C.W., Plakke, B., Poremba, A., 2014. Neural correlates of 1123 1064 206–208. auditory recognition memory in the primate dorsal temporal 1124 1065 Hackett, T.A., et al., 1999. Prefrontal connections of the parabelt pole. J. Neurophysiol. 111, 455–469. 1125 1066 auditory cortex in macaque monkeys. J. Comp. Neurol. 400, Plakke, B., Ng, C.-W., Poremba, A., 2013. Neural correlates of 1126 1067 271–286. auditory recognition memory in primate lateral prefrontal 1127 Hackett, T.A., 2011. Information flow in the auditory cortical 1068 cortex. Neuroscience 244, 62–76. 1128 network. Hear. Res. 271, 133–146. Plakke, B., Romanski, L.M., 2014. Auditory connections and 1069 1129 Hwang, J., Romanski, L.M., 2015. Prefrontal neuronal responses functions of prefrontal cortex. Front. Neurosci. 199, 1–13. 1070 during audiovisual mnemonic processing. J. Neurosci. 35, Plakke, B., Hwang, J., Romanski, L.M., 2015. Inactivation of 1130 1071 960–971. primate prefrontal cortex impairs auditory and audiovisual 1131 1072 Insausti, R., Amaral, D.G., Cowan, W.M., 1987. Entorhinal cortex of working memory. J. Neurosci. 35, 9666–9675. 1132 1073 the monkey II: cortical afferents. J. Comp. Neurol. 264, Romanski, L.M., et al., 1999a. Dual streams of auditory afferents 1133 1074 356–395. target multiple domains in the primate prefrontal cortex. Nat. 1134 1075 Q10 Kojima, T., Karino, S., Yumoto, M., Funayama, M., 2012. A stroke Neurosci. 2, 1131–1136. 1135 1076 patient with impairment of auditory sensory (echoic) memory. Romanski, L.M., et al., 1999b. Auditory belt and parabelt projec- Q11 1136 Neurocase: Neural Basis Cogn.20133–143 http://dxdoi.org/ tions to the prefrontal cortex in the rhesus monkey. J. Comp. 1077 1137 10.1080/13554794.2012.732091. Neurol. 403, 141–157. 1078 1138 Kondo, H., Saleem, K.S., Price, J.L., 2003. Differential connections Romanski, L.M., 2012. Integration of faces and vocalizations in 1079 of the temporal pole with the orbital and medial prefrontal ventral prefrontal cortex: implications for the evolution of 1139 1080 networks in macaque monkeys. J. Comp. Neurol. 465, 499–523. audiovisual speech. Proc. Natl. Acad. Sci. 109 (Suppl. 1), 1140 1081 Kondo, H., Saleem, K.S., Price, J.L., 2005. Differential connections S10717–S10724. 1141 1082 of the perirhinal and parahippocampal cortex with the orbital Saleem, K.S., Kondo, H., Price, J.L., 2008. Complementary circuits 1142 1083 and medial prefrontal networks in macaque monkeys. J. connecting the orbital and medial prefrontal networks with 1143 1084 Comp. Neurol. 493, 479–509. the temporal, insular, and opercular cortex in the macaque 1144 Kotak, V.C., Mowery, T.M., Sanes, D.H., 2015. Characterization of 1085 monkey. J. Comp. Neurol. 506, 659–693. 1145 auditory synaptic inputs to gerbil perirhinal cortex. Front. Saleem, K.S., Miller, B., Price, J.L., 2014. Subdivisions and con- 1086 1146 Neural Circuits 9, 40. nectional networks of the lateral prefrontal cortex in the 1087 Kowalska, D.M., Kusmierek, P., Kosmal, A., Mishkin, M., 2001. macaque monkey. J. Comp. Neurol. 522, 1641–16090. 1147 1088 Neither perirhinal/entorhinal nor hippocampal lesions impair Scott, B.H., Mishkin, M., Yin, P., 2012. Monkeys have a limited 1148 1089 short-term auditory recognition in dogs. Neuroscience 104, form of short-term memory in audition. Proc. Natl. Acad. Sci. 1149 1090 965–978. 109, 12237–12241. 1150

10 brain research ] ( ]]]]) ]]]– ]]]

1151 Scott, B.H., Mishkin, M., Yin, P., 2013. Effect of acoustic similarity Suzuki, W.A., Amaral, D.G., 1994. Perirhinal and parahippocampal 1161 1152 on short-term auditory memory in the monkey. Hear. Res. 298, cortices of the macaque monkey: cortical afferents. J. Comp. 1162 1153 36–48. Neurol. 350, 497–533. 1163 Scott, B.H., Mishkin, M., Yin, P., 2014. Neural correlates of auditory Wright, A.A., Rivera, J.J., 1997. Memory of auditory lists by rhesus 1154 1164 short-term memory in rostral superior temporal cortex. Curr. monkeys (Macaca mulatta). J. Exp. Psychol. Anim. Behav. 1155 1165 Biol. 24, 1–9. Process. 23, 441–449. 1156 Suzuki, W.A., Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1993. 1166 1157 Lesions of the perirhinal and parahippocampal cortices in the 1167 1158 monkey produce long-lasting memory impairment in the 1168 1159 visual and tactual modalities. J. Neurosci. 13, 2430–2451. 1169 1160