The formation of brain circuits requires molecular recognition between functionally related neurons. We report the cloning of a molecule that participates in these interactions. The limbic system-associated membrane protein (LAMP) is an immunoglobulin (Ig) superfamily member with 3 Ig domains and a glycosyl-phosphatidylinositol anchor. In the developing forebrain, lamp is expressed mostly by neurons comprising limbic-associated cortical and subcortical regions that function in cognition, emotion, memory, and learning. The unique distribution of LAMP reflects its functional specificity. LAMP-transfected cells selectively facilitate neurite outgrowth of primary limbic neurons. Most striking, administration of anti-LAMP in vivo results in abnormal growth of the mossy fiber projection from developing granule neurons in the dentate gyrus of the hippocampal formation, suggesting that LAMP is essential for proper targeting of this pathway. Rather than being a general guidance cue, LAMP likely serves as a recognition molecule for the formation of limbic connections.

Monoamines modulate neuronal differentiation, and alteration of monoamine neurotransmission during development produces specific changes in neuronal structure, function, and pattern formation. We have previously observed that prenatal exposure to cocaine in a clinically relevant animal model produces increased length of pyramidal neuron dendrites in the anterior cingulate cortex (ACC) postnatally. We now report that cocaine administered intravenously to pregnant rabbits at gestational stages preceding and during cortical histogenesis results in the early onset of hypertrophic dendritic outgrowth in the embryonic ACC. Confocal microscopy of DiI-labeled neurons revealed that the atypical, tortuous dendritic profiles seen postnatally in ACC-cocaine neurons already are apparent in utero. No defects in neuronal growth were observed in visual cortex (VC), a region lacking prominent dopamine innervation. In striking correlation with our in vivo results, in vitro experiments revealed a significant enhancement of spontaneous process outgrowth of ACC neurons isolated from cocaine-exposed fetuses but no changes in neurons derived from visual cortex. The onset of modified growth in vivo is paralleled by reduced D(1A) receptor coupling to its G-protein. These data suggest that the dynamic growth of neurons can be regulated by early neurotransmitter signaling in a selective fashion. Prenatal onset of defects in dopamine receptor signaling contributes to abnormal circuit formation and may underlie specific cognitive and behavioral dysfunction.

A morphogenic role of neurotransmitters during cellular differentiation in vitro has been demonstrated in recent years. Using in situ hybridization, we confirm the presence of the D1 receptor at E16 and show additionally that the transcript is relatively widespread and present in both proliferative and differentiating areas of the cerebral wall. Because DA receptor expression precedes the arrival of presynaptic terminals during forebrain development, we examined the role of DA in cerebral cortical neuron differentiation in vitro, using immunohistochemical markers of dendrites, microtubule‐associated‐membrane protein 2 (MAP2) and axons, neurofilament protein (NF‐H). Neurite length, cell size, and cell viability in response to D1 and D2 receptor agonists SKF38393 and quinpirole, respectively, and to DA were analyzed in neurons obtained from embryonic (E) day 16 rats. We have shown that 1) paradoxically, DA at different concentrations can either stimulate or inhibit neurite outgrowth; 2) there is a bimodal pattern of DA‐induced axonal outgrowth, i.e., at low and high doses; 3) D2 receptor activation induces neurite outgrowth while D1 receptor activation is inhibitory; 4) D2‐mediated neurite elongation is preferentially axonal while D1 receptor activation reduces both axonal and dendritic outgrowth; 5) low doses of DA promote the expression of cytoskeletal components of axonal maturation; and 6) D1 receptor activation decreases neuronal size. We suggest that DA may influence cellular differentiation and circuitry formation early in development of the cerebral cortex through receptor‐mediated effects on process outgrowth, which could lead to effects on circuit formation. © 1996 Wiley‐Liss, Inc.

The limbic system-associated membrane protein (LAMP) is a 64-68 kDa neuronal surface glycoprotein expressed in cortical and subcortical regions of the limbic system of the adult and developing rat central nervous system (CNS). LAMP is a member of the immunoglobulin superfamily of cell adhesion molecules with three Ig domains and is highly conserved between rat and human. In this study, the temporal and spatial pattern of lamp gene expression during fetal rat development was analyzed by using Northern blot analysis and in situ hybridization. In Northern blot analysis, two lamp mRNA transcripts, 1.6 kb and 8.0 kb, identical in size to those present in the adult rat nervous system, were detected in developing neural tissue. In situ hybridization analysis showed close correlation, though not identity, between the expression of lamp mRNAs and the distribution of LAMP in limbic regions of the developing rat CNS, indicative of a more complex regulation of gene expression than was previously thought to be the case. The expression of lamp mRNAs is first detected on about embryonic day (E) 13. The hybridization signal is not seen in the proliferative ventricular zone at any level of the neuraxis, indicating that lamp is expressed in postmitotic neurons. In the cerebral cortex, lamp mRNAs are expressed in limbic cortical regions, such as the perirhinal cortex, prefrontal cortex, and cingulate cortex. In the hippocampus, the hybridization signal is observed in Ammon's horn by E18. The neostriatum, amygdaloid complex, and most hypothalamic areas express lamp mRNAs from early stages (E13-E14) in a pattern consistent with the onset of neurogenesis. The emerging patterns of lamp expression at the outset are similar to those seen in adult hypothalamus and dorsal thalamus. Although the hybridization signal is observed in some nonlimbic areas, including midbrain and hindbrain structures, intense labeling is evident in more classic limbic regions. The high levels of expression of lamp in limbic regions, beginning in early developmental stages, combined with the results of previous functional in vitro and in vivo studies, support a role for LAMP as a recognition molecule involved in the formation of limbic connections.

The search for molecular markers common to neural structures that are functionally related has become an attractive strategy for neurobiologists interested in identifying mechanisms involved in the formation of patterned connections. One such molecule is the limbic system-associated membrane protein (LAMP), a 64-68 kDa glycoprotein that is expressed in the soma and dendrites of subpopulations of adult neurons in the brain that are functionally associated with classic limbic structures. Such patterned molecular specificity is established prenatally; LAMP is detected during development on the surface of neurons, axonal membranes and pathfinding growth cones. This molecule has now been cloned (lamp) and has been shown to be highly conserved in rat and human. It is a new immunoglobulin superfamily member that has three Ig domains and a glycosyl-phosphatidylinositol (GPI) anchor to the cell membrane. In this study, the distribution of the lamp transcript in the adult rat brain was determined by using in situ hybridization. Generally, the distribution of lamp corresponds well with that of the LAMP protein. Within the cerebral cortex, the transcript is more abundant in areas that are associated with learning/memory and viscerosensory tasks. It is less abundant in somatic sensory and motor areas. The lamp transcript is also ubiquitous in the basal forebrain, amygdala, and preopticohypothalamic areas. In short, the lamp transcript is expressed heavily in areas of the forebrain and diencephalon that have been classically considered limbic and sparsely or moderately in nonlimbic midbrain and hindbrain regions. Correlative analysis of the connectivity patterns of the regions that express greater amounts of the transcript is consistent with a stronger limbic-associated function relative to the regions expressing less lamp. These quantitative differences may be significant in determining the function of LAMP in the adult brain.