1. Picture of Long-Evans rat retrieving a pup.

2. Diagram of rat hovering over pups while licking them (top), nursing in the low (middle) or high (bottom) crouch posture. This upright crouching posture, in which the spine is arched, has recently been termed kyphosis by Stern ('96); it is the opposite spinal curvature from lordosis, the posture of female sexual receptivity. Both kyphosis and lordosis are medical terms that have been applied to describe the postures of reproductive behavior in animals. During kyphosis, the dam is quiescent, i.e., she will ignore a pup that is under her nose, whereas before becoming quiescent she will like the pup or move it to a position where it conveniently can gain access to her nipples. The suckling stimulus brings about the change in the dam's posture and state as well as well-known neuroendocrine changes, i.e., prolactin secretion needed for milk synthesis and oxytocin secretion needed for milk release (let-down).

3. MATERNAL BEHAVIOR STIMULATES C-FOS ACTIVITY IN THE BRAINS OF LACTATING RATS. This is the title of the Master's thesis of Dr. Joseph Lonstein working in Dr. Stern's lab; this work has now been published (J. Neuroscience) or is in press (Neuroscience). The following slides summarize some highlights of this work. C-Fos is an immediate-early gene (IEG), one of a family of such genes. Immediate refers to its rapid activation after a particular stimulus and early refers to the relatively short duration of its activity (about 2 hours) after the stimulus. In conjunction with other IEGs, it begins a cascade of neuronal events. Although it is not yet known just what role it plays, for example with respect to maternal behavior, it is used as a marker of neuronal excitation. Either the protein product, Fos, or the messenger RNA can be visualized with immunocytochemistry (ICC). Thus, IEG activation is widely used by neuroscientists to identify which parts of the brain are activated simultaneously by a sensory or chemical stimulus or by engaging in particular behaviors.

4. Rat mothers (dams) were separated from their pups on Day 5 postpartum, in order to down-regulate the c-fos gene, and then the experiment was performed on Day 7. There were 5 groups in our study. Dams were reunited with pups capable of suckling (SK) or not capable of suckling (NSK) due to local anesthesia of their whisker pads with lidocaine which renders them unable to attach to a nipple. SK dams showed the full array of both active maternal behaviors (retrieval and licking of pups) and nursing behavior (mostly in the kyphotic posture), while NSK dams showed the active behaviors but were not stimulated to show nursing. In the third group dams were presented with pups in a double wire-mesh box containing pups which they could see, smell, and hear but not touch (Pups in Box, P/B). The fourth group received the box alone as a control (BOX), and the fifth group did not receive any stimulus as another control (NoStim). Many brain sites were analyzed. In the medial preoptic area (MPOA), well-known to be important for active maternal behaviors, and in the part of the primary somatosensory cortex (S-I) which represents the face, the quantitative pattern of Fos was as follows: SK = NSK > P/B = BOX > NoStim.

Thus, the trigeminal (5^th cranial nerve) stimulation from snout contact with pups, so important for activation of retrieval and licking of pups (Stern & Kolunie, '89, '91), also activated c-fos in the MPOA and S-I during maternal behavior, and to a less extent during sniffing of the box, with or without pups inside.

5-6. Shown are sections through the MPOA of a in the NoStim group (5) with very little Fos staining and of a in the SK groups (6) with abundant Fos staining. In this ICC procedure, the nuclei of neurons which express the Fos protein are stained black.

7. A bar graph shows the actual mean + standard error of numbers of Fos nuclei in the various groups, in this case in face area of S-I, also called barrel cortex because each whisker is prominently represented by a barrel-shaped structure. There were about 600-900 Fos-stained nuclei in groups SK and NSK, 400-500 in groups P/B and BOX, and about 100 in group NoStim.

8-9. In many sites, the pattern of Fos activity in the various groups was as follows:

SK = NSK > P/B = BOX = NoStim.

Thus, only maternal behavior, with or without suckling, significantly increased Fos expression. These sites include several known from other work to help mediate maternal behavior or to respond to maternal stimulation: Lateral septum, lateral habenula, nucleus accumbens (NA), and trunk representation area of S-I. Other sites had this pattern as well, such as paraventricular thalamic nucleus (PVT), rostral periaqueductal gray (rPAG, see below), and dorsal raphe (DR, near the PAG; serotonin neurons). Except for rPAG and DR, these sites are in the forebrain. Since the dam becomes quiescent (immobile) during nursing, it is likely that forebrain sites are inactivated at that time. Bar graphs showing the quantitative pattern above in the NA and PVT.

10. The Master's thesis of Ms. Stephanie Keer in Dr. Stern's lab (Keer & Stern, '96 abstract, and in preparation for publication) shows one example of how the Nucleus Accumbens in the ventral striatum helps mediate maternal behavior in rats, especially retrieval. It was previously shown in Stern's lab that systemic injections (5 mg/kg) of the dopamine (DA) receptor blocker haloperidol (Haldol, used in people to treat symptoms of schizophrenia), prevents or greatly impairs the dam's active maternal behaviors, but not nursing behavior if she is placed over a litter of hungry pups. The line graphs show the highlights of Keer's study, in which tiny quantities of another DA receptor blocker, cis-flu-penthixol (FLU), were microinfused. FLU in the NA, but not in the caudate putamen (CP, in dorsal striatum) or lateral ventricles (LV), caused a dose-dependent inhibition of pup retrieval, whereas nursing behavior was normal. Pup licking was also reduced by FLU in the NA, but it was reduced to an extent in the other groups, suggesting that this behavior may be affected by DA in many sites.

11. Diagrams of sections of rat brain showing the location of estrogen-receptor (ER) sites.

12. A Table showing that many sites that express FOS in response to maternal behavior also have high densities of ERs, oxytocin fibers, and/or dopamine terminals. This suggests the possibility that the same neurons which express FOS in response to maternal behavior may also have one or more of these other characteristics. Pictures were shown of a pilot study carried out by Dr. Stern this summer in Dr. Jeff Blaustein's lab at University of Massachusetts with his student Tony Auger. Using different flourescent labels to identify FOS and ERs, it was shown that there is a very high degree of co-localization in the MPOA, especially the medial preoptic nucleus (> 50%), of FOS and ERs. Thus, most ER-containing cells in MPN also express FOS in response to maternal behavior. Since maternal behavior is maintained, once induced, by stimulation from pups and without estradiol, these results suggest that somatosensory stimulation from pups may maintain maternal responsiveness by activating estrogen receptors, even in the absence of estrogen. Perhaps a similar mechanism accounts for the hormone-independent maintenance of sexual behavior in women!

13. Bar graphs of FOS in three brainstem sites, caudal (c) PAG (PAG is also called midbrain central gray), rostral (r) PAG, and peripeduncular nucleus (PP). The only site over of 30 analyzed to show more FOS in response to suckling than nonsuckling maternal behavior (SK > NSK) is the cPAG.

14. NURSING BEHAVIOR IN RATS IS IMPAIRED IN A SMALL NESTBOX AND WITH HYPERTHERMIC PUPS. Stern & Lonstein, 1996. In the 2^nd week postpartum rat mothers begin to spend progressively less time nursing. According to Leon and coworkers (Nelson, text p. 321-322 and Fig. 17), this is because the dam gets too hot from contact with the litter mass. Their work was based on automatic determinations of time spent in a very tiny nest box (without direct observations by the investigators), coupled with increases or decreases of room or nest box floor temperature. They did this work before work from my lab showed that suckling induced nursing behavior and that chilled pups do not suckle because they are immobile. In our report, we showed that hyperthermic pups also do not suckle. Therefore, nursing behavior is not induced if the pups are too cold or too warm. Thus, it is important to look at your animals to assess your manipulation. We also found that hyperthermic (abnormally hot) rat mothers (due to combination of morphine and naloxone) nursed normally.

15. Pictures of (left) a lactating dam nursing in the supine posture, stretched out like a cat, with her two-week old pups lined up at her side while suckling, and (right) the same dam placed, while anesthetized) in a nestbox the size that Leon's lab used. It is obvious that there is not enough room in the small nestbox for the dam to show supine nursing, a posture used increasingly with time postpartum as the older, mobile pups initiate nursing from their mother while she is lying down. Therefore, the apparatus Leon used altered the very behavior he was studying!

Transparency: For her doctoral work, Keer is showing that the development of satiety cues by the rat pups after the first week of life accounts for their dam's decreased time spent nursing. Thus, at two weeks if the pups are pre-loaded with 4% or 8% of their body weight with half-and-half cream they suckle less and their dams show less nursing behavior in the subsequent 12 hours. Conversely, if the dam's mammary ducts are ligated (tied off), so that milk cannot be transferred and they remain hungry, the pups suckle more and the mothers nurse more. IT IS THE PUPS WHICH INDUCE THE MOTHER TO SHOW MATERNAL BEHAVIOR.

16. For his Doctoral Dissertation, Lonstein followed up on the previous observation on highest levels of Fos in the cPAG after suckling with a lesion study to determine the function of this region in lactating rats. The title of his recently completed dissertation is: ROLE OF THE MIDBRAIN PERIAQUEDUCTAL GRAY MATTER IN NURSING BEHAVIOR AND MATERNAL AGGRESSION IN LACTATING RATS.

17. Conclusion: Electrolytic Lesions in the Caudal Periaqueductal Gray (cPAG) Impair Kyphosis, But Enhance Maternal Aggression. Lesions were done both prepartum or postpartum.

18. Stacked bar graphs showing duration of time (in minutes) spent in various nursing postures during a 1-hour test. Whereas sham-lesioned (= electrode inserted, but no current passed) and control-lesioned (lesions in the region of the cPAG but missed it on one or both sides) rat mothers were the kyphotic nursing posture for about 30 minutes, the cPAG-lesioned dams were in this posture for less than 5 minutes. In contrast, the cPAG-lesioned dams spent more time than controls in supine (lying on side, like a cat) or prone (over pups, but without limb support, so that pups have little or no room to breathe and swallow milk while suckling; an abnormal posture) nursing postures.

19. Two line graphs show results from prepartum lesion study on days 2-7 postpartum. Top shows the duration of kyphosis per 60-minute test in sham- versus cPAG-lesioned dams; again, about 30 vs. 5 min. Bottom shows reduced litter weight gains in the litters nursing cPAG-lesioned dams. We think this is due to the impaired nursing behavior rather than a neuroendocrine impairment because older litters, which typically suckle while their dam is in the supine posture, gain weight normally.

20. Female rats are most aggressive when they are lactating, which functions to defend the litter and nest. Surprisingly, cPAG-lesioned dams are even more aggressive than intact lactating dams! In fact, they show twice as many attacks on a strange male intruder placed in their home cage than do control mothers (data not shown). Lactating rats have also been shown to be less fearful than virgins, which is probably related to their increased aggression. That is, in a response to a frightening stimulus, a fearless individual is more likely to fight than flee or freeze. Slide shows a rat in an elevated plus-maze, an apparatus used to measure fearfulness. There are two open arms (no walls) and two closed arms (two clear Plexiglas high walls); the more time spent in the open arms, the less fearful the rat is considered to be. In this test, Lonstein and Stern confirmed that lactating rats are less fearful than virgin controls, and further found that cPAG-lesioned lactating rats are even less fearful than intact lactating dams. Additional work with cPAG lesions found greatly decreased female sexual behavior, including both proceptivity and lordosis, during the rat's postpartum estrus (night after giving birth). Thus, the cPAG is a sensorimotor integration site for both kyphosis and lordosis; it also influences fearfulness and aggression.

21-24. Nursing behavior changes the brain! 21) Ratunculus, or map of rat's body in the primary somatosensory cortex; based on electrophysiology. 22) Similar map, but actual cortex of neonatal rat, flattened and stained (for cytochrome oxidase), shows the different parts of the body. The size of the cortical representation is related to the sensitivity of the body part, not to the size of the body part. Thus, in humans this map shows very large lips and fingers. In rats, the head is enormous, taken up mostly by the whisker barrels, then the forepaw, and then the hindpaw representations. In all organisms studied, including humans, the trunk representation is very small. 23-24) Slides from recent work of Xerri, Stern, & Merzenich (1994) showing that the size of the representation of the ventral trunk, where the nipples are, doubles in size during lactation, and the receptive fields (area of skin surface which when touched excite cortical neurons) around the nipples shrink by two-thirds (i.e., less stimulation is needed to activate the neurons). These findings suggest that there are many other changes in the brain of lactating rats in response to stimulation from their young and, more generally, any change in stimulus pattern that is persistent results in changes in the neocortex.

25-37. Effects of breast feeding on women. Throughout the evolutionary history of our species, women breast fed their young at least until they became pregnant again. This practice still occurs and contributes greatly to birth-spacing because of lactational amenorrhea. With sufficient nursing, there is an inhibition of ovulation. Slides 25-29 show pictures of women of the Kung-San Bushmen, traditionally hunter-gatherers. These people have been studied extensively by anthropologists as a model of our evolutionary past, though Western influences have become increasingly prominent in recent years. 25) Woman bending down to nurse her child of about 4. 26) Woman nursing a baby in a sling, in which babies can nurse at will, with an older child at her feet. 27) The ultimate shlep! Woman about 8 months pregnant on a gathering trip with a toddler on her shoulders. 28) Proud and happy-looking traditional Kung woman with her three well-spaced children. 29) Unhappy-looking, Westernized Kung woman (wearing a dress, standing in front of a hut) holding two closely-spaced young children. Westernization includes the introduction of artificial milk and palatable weaning foods, which reduces the baby's nursing time, which reduces the contraceptive benefits of lactation. Ovulation returns and the woman gets pregnant again, sooner than she would with traditional nursing.

30. Line graph shows cumulative % of women menstruating postpartum with no breast feeding (50% by 6 weeks), with Western-style breast feeding (early supplementation) (50% by 5 months), and with "natural lactation" (ad libitum for 1 year) (50% at 1 year).

31-35. In a sample of well-fed American breast feeding women from Boston between 3 and 46 months postpartum, Stern, Konner, Herman& Reichlin ('86) measured prolactin every hour for 24 hours, along with nursing times. We found that although mean prolactin (PRL) levels drop with time postpartum, they remain higher than those of nonlactating women for 2 or more years postpartum, longer than thought previously. Nursing duration is highly correlated with levels of PRL; the women with the longest durations of nursing per day remained amenorrheic (without menses) the longest. Several 24-hour patterns of PRL secretion show the fluctuations of this hormones with nursing and time of day. When the 6 women with the lowest durations of nursing per day (5 of whom were menstruating) were compared with the 6 women with the highest durations of nursing per day, the latter had much higher PRL, remained amenorrheic for close to 18 months, and were comparable in these ways to Kung women studied previously by Konner.

36-37. In a sample of bottle-feeding and breast-feeding women in the Rutgers-New Brunswick vicinity, Stern & Leiblum ('86) confirmed the big difference in onset of menstruation and further found that libido is depressed in the breast-feeding women. For example, it takes them longer to resume intercourse (coitus) postpartum. Also, among the breast-feeding women, the longer the nursing duration per day, the lower the coital frequency per week! This may be related to high levels of PRL and reduced levels of estradiol, but it may also be related to fatigue and other factors relevant to frequent nursing. In brief, nature (both behavior and endocrine function) means the mother to be more focused on the current baby, and less on producing another one!