Main

The physiologic and neurologic bases of different infant states, the means that mothers use to cause state changes, and the central mechanisms that underlie these changes hold interest to pediatricians, psychologists, neuroscientists, and other students of development. In studying the bases of maternally induced change, we and others have discovered that intraorall infusions of milk(1), fat(2), and certain sugars(3–6) calm infant rats(7) and humans(8, 9) and improve their ability to cope with pain(2, 3, 10, 11). In agitated human newborns, 0.1 mL of sucrose causes a 10% reduction in metabolic rate(12).

The effective substances operate pregastrically through their taste and flavor; as little as 20 μL of milk(13) or 250 μL of sucrose(8) delivered to the mouths of rat and human newborns, respectively, are effective against pain and distress. Milk flavor or sweet taste appear to cause central opioid release, because the calming and pain-reducing effects of sugars, fat, and milk are blocked in rats by opioid antagonists such as naloxone or naltrexone(1–3). Moreover, human infants born to women who had used methadone(9) during pregnancy were not calmed by sucrose. Their deficit was specific, because sucking a pacifier, the effects of which are mediated by nonopioid mechanisms, calmed these infants and reduced their heart rate and gross motor activity(9).

Milk and fat also act postgastrically through release of the gut hormone cholecystokinin, which reduces ultrasonic vocalization in isolated rats when delivered i.p.(14). Moreover, the quieting effects of milk and fat, but not sucrose, are severely attenuated in rats by devazepide, the cholecystokinin type A receptor antagonist(15). Thus, milk, through orogustatory stimulation and postingestive change, mobilizes systems that conserve infant energy and, therefore, promote growth. Energy conservation is also achieved via orotactile stimulation provided by sucking breast or nipple. Energy conservation complements suckings better known energy-obtaining function that is realized through the nutrients, water, and minerals provided in milk.

The current study was undertaken because milk's calming and antinociceptive effects have been blocked by opioid antagonists such as naloxone. Moreover,β-CM have been found in the milk and intestines of a number of species(16, 17) including humans (seeRefs. 18 and 19 for review). Milk is of particular interest from our perspective because of the possibility that β-CM may be behaviorally active in infant rats. In adult rats, β-CM is antinociceptive; ICV β-CM increases pain threshold(20, 21). β-CM also quiets baby chicks when injected systemically, although it did not quiet baby rats(22). The mechanisms underlying vocalizations in isolated chick and rat infants may differ(23). β-CM is active pharmacologically as well: 1) β-CM binds preferentially to μ opioid receptors in infant rat brain(23);2) it can cross the blood-brain barrier(24, 25); 3) β-CM7 immunoreactive peptide levels increased markedly in plasma of 4-wk-old puppies after milk but not soy-based formula. The increased levels were maintained for at least 4 h after the meal(26, 27). Singh et al.(27) suggest that physiologically active β-CM may be cleaved from the immunoreactive peptide. β-CM injected systemically very quickly disappear from plasma(28).

The present studies determine whether β-CM are effective as pain-reducing agents in infant rats. They evaluate the potential of milk acting postabsorptively in the brain via β-CM. β-CM's ability to influence infant rat behavior was determined in heat escape tests. In experiment 1, dose-response functions were obtained for a number of exorphins. Experiment 2 evaluated central activity of β-CM5 against systemic injection of naloxone or vehicle and the effects of systemic injections ofβ-CM5 against central injections of naloxone. We now report thatβ-CM5, but not β-CM4 or β-CM7, causes behavioral change when injected systemically or ICV. Moreover, β-CM's behavioral effectiveness is mediated by central μ receptor systems because it is blocked by naloxone delivered ICV or i.p.

EXPERIMENT 1

Experiment 1 evaluated whether systemic administration of β-CM and two of its analogues increased pain threshold. Accordingly, 10-d-old rats received i.p. injections of β-CM4, β-CM5, or β-CM7 and their response to thermal pain was assessed by placing their paw on a 48°C surface(29).

General methods . Subjects. Litters of primiparous and multiparous Sprague-Dawley female rats (Camm Laboratory, Wayne, NJ) mated in our colony served as subjects. Females were individually housed in plastic tubs (38 × 30 × 27 cm) with stainless steel wire lids. The floor of each cage was covered with a bedding of wood shavings. Tap water and Purina Lab Chow (Ralston-Purina, St. Louis, MO) were available ad libitum in the cage top. Lights in the colony were illuminated from 1100 to 2100 h eastern standard time. Colony room temperature ranged from 20 to 27°C, with humidity uncontrolled.

Newborn litters discovered at 1100 h were considered born on that day and designated 0 d old. Litters were adjusted to 12 pups on the day after birth by culling pups from litters of up to 16 pups and by adding pups, younger than 24 h, to litters of 8 or more pups. In all experiments, 10-d-old pups were used and were tested between 1400 and 2000 h eastern standard time. Each pup was studied once only. It received one substance at one specific dose.

Apparatus. The neonate's behavioral response to heat nociception was determined as described by Kehoe and Blass(29). A stainless steel hot plate (maintained at 48°C by circulating hot water) was connected in series to a variable DC power supply and an electric clock/counter (Lafayette Instruments Co.). The pup's right forelimb, two hind limbs, and trunk were gently yet firmly supported by the experimenter, allowing the left forelimb of the pup to rest lightly on the heated surface. The timing circuit was activated when the limb was brought into contact with the hot plate surface and stopped upon forelimb withdrawal. This defined the latency for heat withdrawal. The same apparatus was used in all experiments. Data were collected by an experimenter who was uninformed about the pups' experimental condition. A 30-s cut-off was used to prevent tissue damage. This only had to be used twice during the course of the experiments.

Procedures. One hour before injection, litters were separated from their dam and housed as a group in a plastic tub with clean bedding. Forty-five minutes later the pups were weighted, identified, and randomly assigned to one of the experimental conditions. Subsequently, each pup was injected i.p. Light pressure was applied to the injection site for 5 s to prevent leakage. Each pup was then returned to its siblings and remained housed in the group for a fixed period of time until testing. Testing was conducted at room temperature (22-25°C), humidity was uncontrolled.

Method. One hundred and twenty, 10-d-old pups from a total of 16 litters were studied in 8 experimental runs. Rats were injected i.p. with either β-CM7 (Tyr-Pro-Phe-Pro-Gly-Pro-Ile), β-CM5(Tyr-Pro-Phe-Pro-Gly), β-CM4 amide (Tyr-Pro-Phe-Pro-NH2, also called morphiceptin) at doses of 0, 0.1, 0.5, 1.0, or 2.5 mg/kg, for a total of 13 groups (n = 8 rats/group). Isotonic saline served as the control (0) injection. All compounds (except saline) were dissolved in distilled water. Each rat was tested for heat nociception 15 min after receiving its injection. Because 15 groups were used in each experimental run, rats from two different litters were used for each run. They were randomly assigned to a condition with the restriction that at least one rat from each litter had to be saline-injected. The nine rats that remained from each pair of litters were not studied.

Statistical evaluation. Analysis of variance was conducted on the raw scores by following a 3 (drug) × 5 (dose) design. Because of significant main and interaction effects, we then evaluated the effects of each substance separately to evaluate dose effects for a compound. Because there was considerable variability among the paw lift latencies of control rats, we will present mean difference scores in Figure 1 for the sake of clarity. Difference scores were obtained by subtracting the control rat's score from that of the experimental animal in question.

Figure 1
figure 1

The effects of β-CM4, -5, and -7 on the paw lift latency of 10-d-old rats whose paw was held in contact with a 48°C surface. Mean control value was 7.77 s (±1.05 SEM).

Full size image

Results and discussion. Paw-lift latency differed according to substance, F (2, 105) = 9.58, p < 0.001. The effect of dose was also statistically reliable, F (4, 105) = 2.89, p< 0.05 as was the interaction between substance and dose. It is obvious from Figure 1 that the effective substance was β-CM5. The one-way analysis of variance that evaluated dose effects revealed a statistically reliable effect F (4, 35) = 6.87, p < 0.001. The inert precursor β-CM7 was not effective at any dose; neither was β-CM4, morphiceptin. This last finding was surprising given morphiceptin's potency in adult rats(20). Lack of effect in the current experiment was probably not due to the interval between morphiceptin administration and testing because in subsequent efforts the interval was reduced to 7 min but to no avail. Also, it was not due to faulty supply or handling because morphiceptin reduced ultrasonic vocalizations in isolated rats. Despite these additional efforts, the current studies do not preclude the possibility that morphiceptin is antinociceptive in infant rats. Antinociception threshold in infant rats may require more available morphiceptin than does the threshold for reducing ultrasonic vocalization. Accordingly, we may not have adequately explored either the dose range or the time factors that would allow morphiceptin, which is antinociceptive in adults(20), to manifest its potential in 10-d-old rats. In short, experiment 1 demonstrated that β-CM5 at a dose range of 0.1-2.5 mg/kg significantly increased paw lift latency in a dose-related fashion in the dose range used here, and that related substances were ineffective against noxious thermal stimulation, at least within the time and dose parameters of the present study.

EXPERIMENT 2

Experiment 1 established that β-CM5 elevated heat-withdrawal latencies in 10-d-old rats. In Experiment 2, we sought to determine whether:1) β-CM's action was central in origin and 2)β-CM's effects could be reversed by naloxone. If such a reversal obtained, it would imply opioid mediation. These aims were achieved in two complementary procedures. In one, β-CM5 was delivered directly into the cerebral ventricles, and either isotonic saline or naloxone was injected systemically. In the other procedure, β-CM5 was injected systemically, and either isotonic saline or naloxone, at a dose that was ineffective peripherally, was delivered into the ventricles. Littermates that received isotonic saline injections both systemically and into the ventricles served as controls. The first procedure evaluates central receptivity to β-CM5; the latter evaluates whether the effects of systemic injections can be accounted for by central mediation.

Procedure. A total of 60 10-d-old pups participated in this study. Each pup was randomly assigned to one of six groups. Each rat received two injections, one ICV, one i.p. The ICV treatment always followed systemic injections within 10 min. Testing occurred as above 20 min after ICV treatment to ensure that the rats had recovered from the ICV injections.

ICV procedure. The procedure was modeled after that created by Ellis et al.(30) to study central control of drinking during development. We have previously used it to assess central mediation of morphine-induced analgesia in d 10 rats(31). Solutions in the current study were delivered to the lateral ventricles in a 1-μL volume of which 0.1 μL was India ink to verify accuracy of fluid delivery. The centrally injected substances were 0.25μg of β-CM5. Naloxone (0.25 mg/kg), or isotonic saline were delivered i.p. In the complementary study, isotonic saline or naloxone (0.25 μg) were delivered intracranially, and β-CM, 0.5 mg/kg, or isotonic saline was injected i.p. This constituted six groups, namely: 1) ICV saline-i.p. saline; 2) ICV β-CM-i.p. saline; 3) ICVβ-CM-i.p. naloxone; 4) ICV naloxone-i.p. saline; 5) i.p. β-CM-ICV saline; and 6) i.p. β-CM-ICV naloxone.

Solutions were delivered to the ventricles as follows. The rat's head was securely grasped between thumb and forefinger. A 23-gauge hypodermic needle was oriented perpendicular to the skull and entered the brain approximately 1 mm lateral to the sagittal suture and 2 mm posterior to the bregma. A stop(silastic tubing) arrested passage at a depth of 2 mm into the brain. A 30-gauge injector that had been secured flush with the tip of the guide cannula was then lowered into the brain to a final depth of 4 mm. The injector had been connected via polyethylene 10 tubing to a Hamilton microliter syringe that contained the designated fluid. This was delivered to the cerebral ventricles over 10 s. Both guide and injector were removed, slight pressure was applied to the injection site to prevent leakage, and the pup was returned to the litter. The experimenter was not aware of the contents of either syringe.

Histologic verification. At the completion of testing, an animal was killed by carbon dioxide, and its head was placed in 10% formalin. A single cut was made through the brain along the orientation of the cannula track. The location of the dye was drawn on tracings from the atlas of Konig and Klippel(32). Ink was always found in the lateral ventricles, and in 80% of the animals studied it also invaded the third ventricle. In 14 cases, the injecting cannula extended beyond the base of the lateral ventricle, and ink was present in the caudate nucleus as well as in the ventricular space. The behavior of these rats did not differ in any obvious way from that of others in their group, and their data are included.

Statistical evaluation. A total of four t tests evaluated whether β-CM influenced behavior and whether naloxone administration, in turn, affected the behavior of rats that had receivedβ-CM. Accordingly, the following comparisons were conducted: 1 versus 2 and 1 versus 5 to evaluate the effect of β-CM against saline; 2 versus 3 and 5 versus 6 to determine whether naloxone prevented the effects of β-CM from becoming expressed.

Results. The results of experiment 2 may be summarized as follows. 1) Injections of β-CM (0.25 μg) into the lateral ventricles elevated paw-lift latencies, thereby implicating central responsivity. 2) This was blocked by systemic naloxone, indicating that the effects were opioid mediated. 3) The elevation in PLL caused by systemic β-CM was reversed by central naloxone, suggesting that the effects of the systemic injections were due to central mediation.

These findings are documented in Figure 2 and substantiated by the t test comparisons. The effectiveness of central β-CM at 0.25 μg as an analgesic agent is appreciated by comparing the filled histogram in the left panel with the clear one in the center. A comparison of the filled histogram in the right panel with the clear histogram reveals the potency of peripherally administered β-CM. Further inspection of the left panel shows that the central effects of β-CM were blocked by systemic naloxone. The right panel demonstrates that naloxone (0.25μg) delivered to the ventricular space blocked the elevated paw lift latency caused by systemic β-CM. In short, experiment 2 demonstrates that increased heat escape latencies caused by systemic injections of β-CM are centrally mediated. The comparisons between β-CM and saline controls were both statistically reliable (p between saline and β-CM comparisons <0.01 and 0.001, corrected) as were the comparisons between rats that received β-CM with either saline or naloxone (p < 0.002). Reversal by naloxone cannot be attributed to a general hyperalgesia. The PLLs of the naloxone-treated rats were comparable to saline control rats in experiment 1. Moreover, systemic naloxone did not lower paw lift latencies relative to saline controls in other reports in the literature(1–4, 29). Finally, Kehoe and Blass(31) demonstrated that i.p. naloxone did not reduce PLL in rats that had received saline ICV relative to rats that had received saline ICV and i.p.

Figure 2
figure 2

The effects of ICV and i.p. injections of β-CM5 on paw lift latency and their reversal by ICV or i.p. injections of naloxone.N, naloxone; S, saline.

Full size image

According to these findings, the brain of a 10-d-old rat can detectβ-CM that is delivered into the ventricular space, and this detection is transduced into a delayed paw-lift response to thermal stress. The effectiveness of 0.25 μg of naloxone into the ventricles demonstrates that the central target tissue contains elements that are mediated via μ opioid receptors. This further implies that the effectiveness of peripherally administered β-CM is due to the drug reaching sensitive receptors in the brain. The techniques of the current experiment do not reveal the locations of the behaviorally active systems because injections were into the ventricular space. There are ample β-CM sensitive units in the circumventricular region(33) to engage periventricular pain-combating systems.

GENERAL DISCUSSION

These studies demonstrate that the behavior of infant rats is sensitive to elevations in central β-CM and that β-CM's effectiveness against pain is mediated through central opioid pathways. In this regard, the question posed by this study of can suckling rats respond to increases in brainβ-CM has been answered affirmatively, thereby leaving open the consideration that β-CM obtained normally from a casein-derived milk product may have behavioral consequences. This is an important issue from the perspective of nutritional and behavioral development. The studies necessary to verify or reject definitively a behavioral contribution by β-CM under normal nest circumstances have not been conducted. The available literature provides the following for consideration, however. First, β-CM has been chemically isolated from human and bovine milk(16–18). In humans, blood levels of β-CM are elevated during pregnancy, especially as term nears(34). β-CM has been obtained from cord blood, raising the possibility of its providing analgesic effects during parturition(35). Thus, β-CM activity is elevated at parturition and beyond.

Second, after milk ingestion, a β-CM-like material is hydrolyzed from casein and absorbed in a gradient fashion from specialized microvilli in the gastrointestinal tract as a 12-13-amino acid sequence that has been identified by Singh et al.(27) as a β-CM7 immunoreactive peptide. The peptide is not readily degraded because it was found in plasma some 4 h after milk ingestion. This is significant becauseβ-CM 7, 5, and 4 are not found in plasma; β-CM are rapidly degraded with their half-life estimated to be shorter than 5 min(27). The immunoreactive fraction, being extremely hydrophilic, can probably cross the blood-brain barrier. This is more likely in very young animals for whom the barrier is considered to “leak”(24).

According to Volterra et al.(23), naturalβ-CM show a preferential affinity to the μ receptors. The present study complements this line of analysis because it demonstrates that elements in the brain can recognize the β-CM peptide configuration, in a dose-related manner, and are intrinsic to or can engage central mechanisms that are concerned with pain mediation and escape from noxious thermal stimulation. In this regard, we find it extremely unlikely that the delayed paw-lift latency reported here is due to motor incapacity because doses of morphine up to 0.5 mg/kg that doubled paw-lift latency in newborn rats did not compromise motor behavior(13).

These data fit into two larger, behavioral-physiologic perspectives on mammalian development. One is concerned with the mechanisms through which milk affects infant state; the other, with the classes of energy-conserving events that are triggered by mother's milk and their short- and long-term behavioral consequences.

Previous studies on the effects of milk and other substances on infant state suggest that milk utilizes at least three pathways to calm infants and to reduce pain. One pathway is orosensory and includes olfactory and gustatory information. Triggered by milk flavor, this pathway is mediated in rat and human infants by central opioid systems. A second pathway, orotactile, is activated by the texture of the nipple or breast and is not opioid-mediated. The third pathway is postingestive. It refers to the ability of cholecystokinin to quiet vocalizing infant rats.

Evidence for orosensory mediation has come from a number of studies. Blasset al.(13) have demonstrated that 20 μL of milk injected into the mouths of rats, that had just been delivered by cesarean section, markedly increased heat-escape latency, an effect that was naloxone-reversible. This is consistent with findings in older (d 10) rats whose heat-withdrawal responses were also considerably elevated by intraoral infusions of milk(1) and fat(2), but not lactose, the milk sugar(4). Moreover, in isolated rats, ultrasonic vocalizations, which have been broadly interpreted as a sign of stress(36), are reduced by infusions that increase pain threshold(1, 2, 4). The quieting caused by these infusions was also blocked by pretreatment with naloxone or naltrexone. As indicated, naloxone does not appear to be a hyperalgesic agent because it does not cause behavioral change in otherwise untreated or vehicle-treated rats.

Parallel findings have been obtained for human newborn infants. A single 250-μL bolus injection of sucrose quiets them(8). Sucrose (12%, wt/vol), delivered orally in two 100-μL infusions arrests spontaneous crying, and this calm endures for at least 5 min. The calm is accompanied by decreased gross motor activity and heart rate(5, 6, 9) and a 10% reduction in energy metabolism(12). Sucrose's calming and cardiovascular effects were not obtained in human newborn infants whose mothers had used methadone(9) during the course of pregnancy. Thus, in rats and humans the taste and flavor of milk, fat, and certain sugars may cause the release of central opioids that gain access to cardiovascular and behavioral systems that are involved in energy conservation. The data of the present study raise the possibility that central opioid systems, μ receptor in particular, are also engaged via a postabsorptive pathway, throughβ-CM, that enables milk's calming and energy-conserving influences to extend beyond the nursing-suckling bout.

The present data suggest that mothers, through the postabsorptive qualities of their milk, cause protracted change in the infant state. They demonstrate, in infant rats at least, that the brain can detect changes in circulating levels of β-CM dose dependently. Whether such changes can occur under physiologic circumstances surrounding milk delivery provides an interesting challenge to help guide future research. Resolving this question has implications beyond the moment of the particular nursing episode. Because these changes may help infants learn about the particular features of their mother, they may also contribute to the formation of the mother-infant relationship(37, 38).