BIOLOGICAL OVERVIEW

Drosophila latheo is involved in two unrelated aspects of cell biology: it was initially identified as a behavioral mutant and it also codes for a protein that functions at chromosomal sites known as origins of replication. latheo (lat) was initially identified in a behavioral screen for olfactory memory mutants (Boynton, 1992). Lat has been shown to be a presynaptic protein with a role in the Ca²⁺-dependent synaptic modulation mechanisms necessary for behavioral plasticity (Rohrbough, 1999). The original hypomorphic lat^P1 mutant shows structural defects in the adult brain. Homozygous lethal lat mutants lack imaginal discs, show little DNA replication of cell proliferation in the CNS of third instar larvae, and die as early pupae. lat has also been found to encode for ORC3, a protein with homology to a subunit of the origin recognition complex (ORC). Human and Drosophila Lat both associate with the known ORC protein ORC2 and both are related to yeast ORC3, suggesting that Lat functions in DNA replication during cell proliferation (Pinto, 1999).

Latheo is a conserved protein. The 80 kDa protein HsLat (for Homo sapiens Lat) coprecipitates with human ORC2. HsLat also associates with human ORC4 and ORC5 and thus appears to be an integral member of the ORC. Drosophila Lat has been shown to associate with Drosophila ORC2. The conservation of this Lat-ORC2 association between fly and human homologs, combined with the similar functional (cell proliferation) defects of lat and l(1)K43 (ORC2) mutant flies, further strengthens the notion that Lat is a bona fide ORC subunit. Sequence comparisons strongly suggest that Lat is the functional equivalent of yeast (S. cerevisiae) ORC3. Recently, a similar protein has been identified in Xenopus ORC (Pinto, 1999 and references).

If one assumes that the learning and memory phenotypes of mutant lat flies are due to morphological defects caused by cell proliferation defects (nuclear in origin) and related to the ORC function of Lat, then the expression of Latheo at synaptic (axonal) connections of the larval neuromuscular junction (NMJ) and its enriched expression in presynaptic boutons comes as a surprise. This expression at axonal terminals opens the possibility that it is this axonal expression that is responsible for learning and memory defects, rather than or in contrast to proliferation defects. Latheo, then, operates on both the nuclear level and in the axon terminal, far from the nucleus: experiments have shown that it is the axonal, rather than the nuclear, protein actions that explain the learning and memory defects.

The lat- mutations involved in the learning and memory defect exhibit only mild alterations in synaptic morphology. At the more complex muscle 12 NMJ, mutant terminal complexity appears to be slightly reduced on the basis of terminal branching and bouton number. In contrast, at the simpler muscle 4 NMJ, mutant terminals have normal or slightly increased bouton numbers and terminal complexity. These morphological differences are completely insufficient to account for the observed functional alterations in transmission at mutant synapses. Lethal lat mutant alleles exhibit strikingly altered synaptic transmission properties at the NMJ. Evoked transmission is not impaired; on the contrary, transmission is strongly elevated relative to normal. Moreover, mutants exhibit a reduced Ca²⁺ dependence with regard to transmission and strongly impaired activity-dependent synaptic plasticity, including short-term forms of facilitation, longer-term augmentation, and posttetanic potentiation. In these terms, lat mutants strongly resemble dunce mutants (Zhong, 1991). Lat's synaptic localization, coupled with the lat mutant synaptic phenotype, suggests that the protein may function to modulate synaptic transmission levels during the learning process, possibly through Ca²⁺- or cAMP-dependent mechanisms (Rohrbough, 1999).

This overview details the defects in neural transmission in lat mutants. The available information shows how synaptic function can be examined in Drosophila. The results of these analyses suggest that there is a great deal of similarity between the physiological effects of lat mutants and those of dunce and rutabaga mutants, known to effect the cyclic AMP learning and memory pathway. Lat regulates both evoked transmission amplitude and activity-dependent forms of synaptic facilitation and potentiation. Functional synaptic defects in lat mutants are strikingly similar to those in the cAMP mutant dunce. However, lat mutations produce far less distinctive alterations in NMJ synaptic morphology than those observed for dunce (Zhong, 1992) and thus lat mutantions specifically affect functional aspects of synaptic transmission and plasticity. Lat appears to have two roles in the presynaptic terminal, independent from a possible function in DNA replication: (1) regulation of Ca²⁺-dependent, basal-evoked transmission, including the Ca²⁺ sensitivity of release, and (2) regulation of both short- and longer-term aspects of Ca²⁺- and/or cAMP-dependent synaptic modulation. It is suggested that Lat may serve similar functions at central synapses involved in learning in the adult brain (Rohrbough, 1999).

Lat protein is expressed at all morphological subtypes of boutons at the NMJ, including those primarily utilized for fast glutamatergic transmission (type I synapse) as well as those that may have modulatory (type II synapse) or regulatory (type III synapse) functions (Keshishian, 1993). Lat is distinctively colocalized at synaptic boutons with the presynaptic vesicle protein Cysteine string protein. Colabeling experiments with markers for postsynaptic proteins, including DGluR2a, Discs large, and ßPS integrin, also suggest that the Lat protein is restricted to the presynaptic compartment. These results are consistent with findings that presynaptic plasticity mechanisms are defective in lat mutants. It is proposed that Lat is presynaptically located on the basis of these results but the possibility of its presence postsynaptically cannot be excluded (Rohrbough, 1999).

Mutant lat NMJs exhibit strongly elevated basal-evoked synaptic transmission over a wide range of external Ca²⁺ concentrations. Excitatory junctional current (EJC) amplitude in lethal lat- mutants is three to four times larger than normal in low Ca²⁺ at both the muscle 12 and 6 NMJs, suggesting that Lat likely has a similar function at all NMJs. However, these NMJs in lat- mutants exhibit no significant morphological alterations that can be correlated with increased transmission strength, such as greater numbers of terminal branches or synaptic boutons. The muscle 12 NMJ is, by contrast, slightly simplified in lat- larvae, when compared to normal, forming fewer higher-order branches and synaptic boutons. Removal of the protein in lat- mutants thus has a primarily functional consequence on the regulation of presynaptic transmitter release. Furthermore, normal mEJC amplitude and frequency is observed for lat- mutants, indicating that, respectively, neither postsynaptic glutamate receptor density nor the rate of spontaneous presynaptic vesicle exocytosis appear to be altered. The increased evoked mutant transmission is thus presynaptically mediated and specific to the Ca²⁺-dependent, evoked release pathway (Rohrbough, 1999).

Though lat mutants have enhanced, prefacilitated basal-evoked release, the ability of mutant synapses to undergo further activity-dependent increase in transmission appears to be severely defective. Under low-Ca²⁺ conditions (0.2 mM) that ordinarily favor Ca²⁺- and activity-dependent forms of synaptic facilitation, lat mutants have strongly depressed paired-pulse facilitation (PPF) and short-term frequency-dependent facilitation (STF). For measurements of PPF, responses to five to ten consecutive paired stimuli (separated by 4 s rest) at 20-100 ms interpulse intervals are averaged. At 20-30 ms intervals, wild-type NMJs exhibit 2.5-fold PPF, while lat- mutant NMJs exhibit only weak PPF or even depression. For measurement of STF, trains of 20 stimuli are delivered to the NMJ at 0.5-20 Hz frequencies. The amplitudes of the last ten responses in each train are averaged and normalized to the average amplitude at 0.5 Hz. Wild-type NMJs exhibit 2.5-fold facilitation at 10 Hz stimulation. In lat- mutants, initial transmission at 0.5 Hz is strongly elevated relative to normal and shows on average only weak facilitation (Rohrbough, 1999).

These forms of facilitation common to most synapses are generally believed to reflect enhanced transmitter release due to transient, residual increases in presynaptic Ca²⁺ during repetitive high-frequency stimulation (Zucker, 1993; Fischer, 1997). The short-term plasticity defects in lat- mutants do not appear to result simply from already saturated Ca²⁺-dependent plasticity mechanisms. When external Ca²⁺ is further reduced to enhance short-term activity-dependent increases in presynaptic Ca²⁺, lat- mutant PPF and STF are incompletely restored even though basal EJC amplitude is normal (Rohrbough, 1999 and references).

The regulation of transmitter release by intracellular Ca²⁺ has been examined at various synapses using a combination of techniques. Evoked, prefacilitated exocytosis is triggered by extremely rapid Ca²⁺ influx, and Ca²⁺ binding to fast, low-affinity molecular targets directly at the active site. Short-term facilitation (lasting 1 ms to 1 s) as well as augmentation and PTP (lasting seconds to minutes) are thought to be due to the continuing action of Ca²⁺ at distinct domains in the presynaptic terminal, each with different affinities for Ca²⁺. Rapid forms of facilitation, including PPF and STF, appear to result from residual Ca²⁺ binding to sites at or near the site of rapid exocytosis. More prolonged forms of facilitation, including augmentation and PTP, may be induced by longer-lived, lower levels of residual Ca²⁺ acting at different sites in the terminal. The results of this study suggest that one possible functional role for Lat may be in regulating Ca²⁺ levels or dynamics at one or more of these putative presynaptic domains. The Lat protein appears to be localized in the same proximity as synaptic vesicles. The elevated level of evoked transmitter release, shift in Ca²⁺ dependence, and reduced Ca²⁺ cooperativity of release at lat- NMJs suggest that the protein may regulate Ca²⁺ concentration or binding affinity at the rapid release site. However, the loss of short-term facilitation, augmentation, and PTP in lat- mutants suggests the release dependence is also altered at other presynaptic Ca²⁺ domains involved in mediating these forms of plasticity (Rohrbough, 1999 and references).

It is difficult to propose a Ca²⁺ regulatory scheme that can account for this combination of lat transmission defects. The increased mutant transmission phenotype could be produced by increased basal levels of presynaptic Ca²⁺ resulting from alterations in presynaptic Ca²⁺ sequestration or buffering. For example, presynaptic reticulum have been proposed to play important roles in presynaptic Ca²⁺ sequestration at other synapses and can regulate both basal Ca²⁺ level and transmitter release as well as Ca²⁺-dependent synaptic plasticity. However, other aspects of lat mutant transmission are at odds with a generalized role in Ca²⁺ regulation. Most significantly, no increase in lat- mutant mEJC frequency is observed, as would be expected from increased basal presynaptic Ca²⁺ levels resulting from a mitochondrial or Ca²⁺ buffering defect. Whereas the inhibition of presynaptic Ca²⁺ sequestration increases facilitation at Aplysia synapses (Fossier, 1993 and Fossier, 1998), synaptic facilitation in lat- mutants is absent or depressed even at low external Ca²⁺. Further work will be necessary to determine how Lat may be involved in presynaptic Ca²⁺ action at various domains following Ca²⁺ influx as well as during activity-dependent synaptic facilitation (Rohrbough, 1999 and references).

Lethal lat mutants display strongly impaired synaptic augmentation during prolonged tetanic stimulation and severely depressed PTP. The defects in these longer-term forms of synaptic modulation also remain significant even if initial lat- mutant synaptic transmission amplitude is reduced to the wild-type level by further reducing external Ca²⁺. These results support the idea that Lat may have a specific role in the synaptic modulation mechanisms that support long-term augmentation and PTP at the NMJ. At the Drosophila NMJ, these more prolonged (minutes or longer) forms of plasticity have been shown to depend, at least in part, on the activation of cAMP and the cAMP-dependent signaling pathway (Zhong, 1991). In the presence of cAMP analogs, basal low-frequency stimulation reversibly induces prolonged (>30 min) potentiation. Tetanic stimulation accelerates PTP formation in the presence of cAMP, indicating the cAMP modulation process is also activity dependent (Rohrbough, 1999 and references).

Learning and memory mutations known to alter cAMP synthesis and degradation also alter synaptic augmentation and potentiation properties (Zhong, 1991 and Storm, 1998). In particular, the Drosophila mutants of dunce, which encode a cAMP-specific phosphodiesterase, show increased cAMP levels and impaired olfactory learning and short-term memory. Both dunce and lat mutants display markedly similar functional phenotypes at the larval NMJ. In low external Ca²⁺, dunce basal-evoked transmission is characterized by a strongly (4-fold over normal) elevated EJC amplitude, similar to the 3- to 4-fold increase observed for lat; and like lat, dunce basal-evoked transmission also has a reduced Ca²⁺ dependency. Moreover, dunce mutant NMJs also show a nearly complete loss of STF, longer-term augmentation, and PTP. The only striking difference between the lat and dunce mutant phenotypes is that dunce also significantly increases NMJ morphological growth, including bouton number and terminal branching, by 30%-50%. The lat mutation is therefore of particular interest as a second example in which abnormally elevated synaptic transmission strength is correlated with a deleterious effect on learning or memory but that is unique among Drosophila learning/memory genes in that it appears to affect presynaptic function independent of morphology (Rohrbough, 1999 and references).

Parallel evidence that cAMP levels critically regulate long-term synaptic plasticity and memory formation comes from Drosophila rutabaga mutants, which have reduced cAMP levels due to the loss of Ca²⁺/CaM-dependent AC activity. Basal-evoked transmission amplitude at rutabaga mutant NMJs is normal, but STF and augmentation are weaker than normal, and mutants cannot sustain PTP. Though the effect on cAMP level and synaptic phenotype are distinctly different, rutabaga and dunce mutants have similar defects in adult olfactory learning and memory ability. In mammals, the impaired cerebellar LTP observed for AC mutant mice (Storm, 1998) further demonstrates the importance of cAMP activation in learning and memory processes. These studies strongly suggest that cAMP elevation, activated at least in part by sustained presynaptic elevations of Ca²⁺ occurring during high-frequency activity, initiates the functional long-term synaptic potentiation necessary for normal learning and memory to occur. However, habitually elevated cAMP levels, as in dunce mutants, can result in permanently potentiated synapses without the ability to dynamically upregulate Ca²⁺- and cAMP-dependent pathways, with equally negative behavioral consequences (Rohrbough, 1999 and references).

Evidence from Drosophila, Aplysia, and mice strongly indicates that the downstream effector of lasting cAMP-mediated functional and behavioral modifications is gene expression regulated by the CRE and CRE-binding protein (CREB). In Drosophila, the induced transgenic expression of activator or repressor isoforms of the Drosophila form of CREB (dCREB2) specifically enhances or blocks long-term memory formation, respectively. Expression of the dCREB2 repressor isoform during development in dunce mutant larvae also partially rescues the dunce mutant synaptic transmission and functional plasticity defects at the NMJ (Davis, 1996). These results suggest that the permanently altered plasticity at dunce mutant NMJs is indeed a consequence of the overexpression of genes under the transcriptional regulation of dCREB2, brought about by abnormally high levels of cAMP (Rohrbough, 1999 and references).

The lat mutant olfactory learning defect, combined with the striking functional similarities between the lat and dunce mutants, leads to a speculation that Lat may also be involved in cAMP-dependent forms of synaptic modulation. The behavioral defects of the viable lat^P1 mutant appear to be specific to the initial acquisition or learning step rather than to subsequent forms of short- or long-term memory. Lat may therefore function in one or more short-term, transcription-independent forms of modulation, including cAMP-dependent aspects of synaptic vesicle mobilization, docking, and release as well as downstream protein kinase A-dependent modulation of synaptic efficacy, or in another protein kinase or signal transduction pathway also involved in synaptic modulation. Future work must focus on synaptic ultrastructural correlations to the lat mutant functional phenotype and the relationship of this pioneer synaptic protein to identified synaptic proteins and modulation pathways. Examination of mutant synaptic ultrastructure will be necessary to reveal alterations in the number or structure of presynaptic release sites and the availability of synaptic vesicles at these sites. Ca²⁺ imaging at both resting and stimulated synaptic terminals in lat mutants may provide insight into a possible function in presynaptic Ca²⁺ regulation. Involvement in cAMP-dependent synaptic modulation can be tested by examining functional synaptic transmission and plasticity in lat double mutant combinations with dunce and rutabaga. These studies, together with continued molecular analysis, promise to advance the understanding of Lat's molecular role (Rohrbough, 1999).

The involvement of a DNA replication protein in functional synaptic plasticity is, to put it mildly, unexpected, and leaves three interpretations of the synaptic phenotype of latheo mutants. (1) The synaptic defects may be downstream of LAT's role in the nucleus, and the presence of LAT at the synapse may be a fluke. (2) The ORC has two functions: one in the nucleus and one at the synapse. (3) LAT has two functions, one as a part of the ORC and a second at the synapse. Fortunately, the reagents exist to test these three models. LAT biochemically interacts with another subunit of the Drosophila ORC, DmORC2. Mutants for DmORC2 show a proliferation phenotype identical to lat: late larval lethality, missing imaginal discs, and defective cell proliferation (Gatti, 1989 and Landis, 1997). In addition, antibodies have been generated to DmORC2 (Gossen, 1995). Future experiments can address whether DmORC2 mutants have a synaptic phenotype and if the protein is synaptically localized. If DmORC2 mutants do show the latheo synaptic phenotype but DmORC2 is not present at the synapse, this would suggest that the synaptic defects are downstream of the ORC's role in replication. If DmORC2 is at the NMJ, this would lead to the exciting hypothesis that the ORC has a second function at the synapse. Finally, if DmORC2 mutants have no synaptic phenotype, this would argue that Lat has a second function at the synapse. Regardless of the final explanation for the unexpected results with Lat, genetics has served its purpose in calling attention to a molecule that is not a usual suspect at the synapse (Rohrbough, 1999).

GENE STRUCTURE

Northern blot analyses reveal two transcripts of 2.6 and 3.9 kb in wild-type flies (Pinto, 1999).

Exons - 8

PROTEIN STRUCTURE

Amino Acids - 671

Structural Domains

A basal local alignment search technique (BLAST) search of the GenBank database indicates that this lat transcript encodes a novel protein. One human expressed sequence tag (EST) clone (U50950), however, shows significant amino acid sequence homology to Lat (Pinto, 1999). Sequence comparisons (Pinto, 1999) strongly suggest that Lat is the functional equivalent of ScORC3. Recently, a similar protein has been identified from Xenopus ORC (Carpenter, 1998).

date revised: 1 July 99

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.