BIOLOGICAL OVERVIEW

Adenomatous polyposis coli tumor suppressor homolog 2 (Apc2), a second Drosophila APC homolog, was identified nearly simultaneously in three laboratories (van Es, 1999; Yu, 1999a and McCartney, 1999). Whereas, the previously characterized APC-like regulates Arm signaling in the Drosophila larval photoreceptors (Ahmed, 1998), two features of APC-like were surprising, given the widespread expression and essential function of mouse APC. Embryonic expression of APC-like is largely confined to the central nervous system (Hayashi, 1997), and null mutations in APC-like are viable and fertile, with strong effects only in the larval photoreceptors (Ahmed, 1998). These observations suggested the existence of a second APC gene in flies, now identified as Apc2.

As well as having an intimate involvement in Wingless signaling, a function well documented for APC-like, APCs in general have additional cellular roles. When human APC (hAPC) is overexpressed in cultured cells, it decorates microtubules (MTs) and can bind and bundle MTs in vitro (Munemitsu, 1994; Smith, 1994). In cultured cells, APC localizes at the cell cortex in membrane puncta where bundles of MTs often terminate (Nathke, 1996). If one expresses a stabilized form of ßcatenin (Drosophila homolog: Armadillo) that cannot be phosphorylated by GSK (Drosophila homolog Shaggy), in cultured cells, mutant ßcatenin accumulates with APC in membrane puncta, and these cells display altered migratory behavior. These data have prompted the suggestion that APC may regulate cell migration via its interaction with MTs, and that this role is modulated by ßcatenin. APC may also influence cytoskeletal dynamics by binding to EB1 (Su, 1995), which associates with the MT cytoskeleton in mammalian cells (Berrueta, 1998; Morrison, 1998). Yeast EB1 homologs contribute to MT function and may form part of a cytokinesis checkpoint (Beinhauer, 1997; Schwartz, 1997; Muhua, 1998). In addition to connections with MTs, APC may associate with the actin cytoskeleton via ß- and alpha-catenin (McCartney, 1999).

Both actin and MT cytoskeletons are targets of the Wg/Wnt pathway. Signaling by Wnt family members or by their Frizzled (Fz) receptors (see Drosophila Frizzled) is required to orient certain cell divisions in both nematodes and flies. In C. elegans, Wnt signaling directs the orientation of mitotic spindles in specific early embryonic blastomeres and orients postembryonic asymmetric cell divisions (for review see Han, 1997), whereas in Drosophila, Fz is required for orientation of the mitotic spindles of bristle precursor cells (Gho, 1998). Fz also plays a key role in orienting the cytoskeleton during formation of hairs and bristles, polarized outgrowths of the cell membrane (for review see Shulman, 1998). Both the actin and MT cytoskeletons are required for hair positioning and growth (Wong, 1993; Turner, 1998).

Discovery of Drosophila Apc2, and investigation of its cytoskeletal connection (McCartney, 1999), raises the possibility that Apc2 acts as an effector molecule through which Wnt signaling influences the cytoskeleton. Because Apc2 associates with the actin cytoskeleton in contexts where no Wg signaling is thought to occur, such as in pre-blastoderm embryos, Apc2 may play more fundamental roles in cytoskeletal regulation. Such functions may now be revealed by further genetic analyses of APC-like and Apc2 (McCartney, 1999).

Anti-Apc2 antisera were used to characterize Apc2 expression and subcellular localization. During nuclear division cycles 10-13, which take place without cytokinesis in the peripheral cytoplasm of the embryo, Apc2 shows dynamic changes in subcellular localization, coincident with changes in actin. Sequential changes in MT organization as nuclei proceed through mitosis direct the reorganization of the cortical actin cytoskeleton. Before nuclei migrate to the periphery, actin is found at the cortex in a random reticulum. When nuclei reach the periphery, actin condensations appear in interphase and prophase above each nucleus, forming an actin bud that overlays a cytoplasmic bud. This separates the mitotic apparatus of one nucleus from that of its neighbor. As division proceeds to metaphase, actin redistributes from the crown of the bud to its lateral cortex, forming an oblong ring around each spindle. During anaphase, actin redistributes into discs above each newly formed nucleus. Centrosomes and their associated MTs direct the changes in actin distribution, although the mechanism responsible for this interaction is not known (McCartney, 1999).

In cycle 10-13 embryos, Apc2 colocalizes with actin at all stages of mitosis. The Apc2/actin colocalization is most prominent in the microvillar projections at the surface of the bud in interphase and prophase. At metaphase and anaphase, Apc2 and actin condensations are observed at the lateral cortex of the bud; Apc2 staining is somewhat less intense here, relative to actin. Toward the base of the bud, condensations of actin and Apc2 are also found in the region of the centrosome and asters. These Apc2 condensations occur within 0.3-0.5 µm of the surface of the embryo, and thus are most prominent above the spindle apparatus; kinetochore MTs are not in uniform focus until ~1.25 µm from the surface of the embryo. The location of these Apc2/actin condensations above the plane of the spindle places them in a position to interact with the astral MTs as they reach toward the cortex. During later nuclear cycles when pseudocleavage furrows are present, more defined dots of actin and Apc2 staining are sometimes observed in the region of the centrosomes. In a wild-type stock infected with the bacterial endosymbiont Wolbachia (visible as small propidium iodide-positive bodies), an additional Apc2 localization is seen. Wolbachia associate with astral MTs in Drosophila and thereby disperse into newly formed cells. In infected embryos, Apc2 localizes with the actin cytoskeleton as in uninfected stocks, and also associates with bacteria at the asters. Another astral MT-associated protein, the kinesin-like protein KLP67A is known to associate with bacteria. EM studies have shown that the bacteria are encapsulated within a cytoplasmic vacuole attached to astral MTs via an electron-dense bridge, possibly composed of cellular MT-associated proteins. Apc2's localization to the aster region of noninfected embryos and its association with bacteria suggest that Apc2 may contribute to the binding of the vacuole to the asters (McCartney, 1999).

After cellularization, Apc2 is still enriched in the region of MTs. Increased levels of cytoplasmic Apc2 are observed in mitotic domains (groups of cells undergoing synchronous mitosis). Here, cytoplasmic condensations of Apc2 are observed in the region of the spindle in metaphase and anaphase, but are absent in prophase or telophase; serial sections reveal that these cytoplasmic condensations are most prominent within 2-4 µm of the cell apex. In mitotic domains of a Wolbachia-infected strain, punctate condensations of Apc2 are observed near the spindle poles, presumably astrally associated bacteria, consistent with Apc2 localization to bacteria associated with preblastoderm asters (McCartney, 1999).

Apc2 is also expressed in dividing cells of the larval brain. The optic lobes contain two proliferative regions, the inner and outer proliferative zones. Apc2 is highly expressed in dividing cells of the proliferative zones and in their immediate progeny, but not in differentiated neurons. In contrast, Arm is not enriched in the proliferative zones but is enriched in axons. In the ventral nerve cord, Arm is found in axons, whereas Apc2 is found in midline glial cells. In contrast, Drosophila APC-like localizes to axons, at least in embryos (Hayashi, 1997). However, in larval neuroblasts (neural stem cells) Apc2 and Arm share a striking asymmetric distribution. Neuroblasts divide asymmetrically to produce a large neuroblast and a smaller ganglion mother cell, which will divide symmetrically to produce two neurons. The asymmetric division requires specific orientation of the mitotic spindle. Inscuteable (Insc), localized in a crescent opposite the future daughter cell during prophase and metaphase, is required for both spindle orientation and localization of the neural determinants Prospero and Numb. In larval neuroblasts, both Apc2 and Arm colocalize to a cortical crescent next to the future daughter cell; this crescent also includes the neural determinant Prospero. In contrast to other asymmetric neuroblast components, the Apc2 and Arm crescents are present even at interphase. In some neuroblasts, cortical actin also accumulates in a crescent with Apc2, whereas in others this association is less apparent. To examine the relationship between Apc2 and the spindle, neuroblasts were triple labelled with antibodies against phosphohistone, ß-tubulin, and Apc2. One pole of the spindle apparatus colocalizes with the Apc2 crescent; Apc2 is enriched at this point relative to the rest of the crescent. At this stage of the cell cycle, low levels of Apc2 were also observed at the opposite cortex; this position often coincided with the other spindle pole. Whereas cortical Apc2 associates with spindle poles, neuroblasts do not have cytoplasmic condensations of Apc2 around the central spindle as are observed in epidermal cells. Apc2 is also asymmetrically localized in embryonic neuroblasts (McCartney, 1999).

In nondividing cells, Apc2 also associates with the cell cortex, and colocalizes with actin. In the embryo, Apc2 is most strongly expressed in the epidermis and other epithelial cells. In the epidermis, Apc2 is enriched at the cell cortex and is also found throughout the cytoplasm in a punctate distribution. At the cortex, Apc2 appears as numerous punctate condensations of protein that are most prevalent at the apical end of the lateral cell surface but are also found more basally. The most intense staining of Apc2 appears at points of contact between multiple epidermal cells. Apc2 condensations often colocalize with condensations of actin and phosphotyrosine, although actin and phosphotyrosine associate with the cortex more continuously. In fully polarized epithelial cells, like the embryonic hindgut or the larval imaginal discs, Apc2 is enriched in adherens junctions, where it colocalizes with Arm; Apc2 also accumulates on the apical plasma membrane. The intracellular distribution of Apc2, in contrast to that of Arm, is not modulated in a segmental fashion. A strikingly different localization of Apc2 occurs in the epidermis after stage 15. Apc2 becomes organized into very large apical structures in segmentally repeated subsets of ventral epidermal cells, just before the stage at which these cells initiate denticle formation. The Apc2 structures occur specifically in the anterior epidermal cells of each segment and colocalize with similar actin structures, which likely represent larval denticle precursors (McCartney, 1999).

Although Apc2 colocalizes with actin in many tissues, it does not colocalize with actin in all contexts. For example, during cellularization, actin is prominent at the cellularization front, whereas Apc2 is enriched at the apical cortex. In addition at the cortex of epidermal cells actin is present at the membrane in a continuous fashion, whereas Apc2 is restricted to regions of most intense actin staining. Finally, Apc2 is not found with actin in cytokinesis furrows. Thus, although Apc2 associates with the actin cytoskeleton, the context-dependent nature of this association suggests that this association is regulated (McCartney, 1999).

Thus Apc2 colocalizes with actin in many but not all cell types, suggesting a regulated interaction. The association between the actin cytoskeleton and Apc2 may occur via Arm and beta-catenin, although in some places where Apc2 and actin colocalize, there is little or no detectable Arm. The colocalization of Apc2 and actin is intriguing given the effects of Wnt/Fz signaling on planar polarity in Drosophila. In the wing, the best studied example, Frizzled signaling triggers asymmetric polymerization of actin, leading to development of an actin-based wing hair in the distal vertex of each hexagonal wing cell. The colocalization of actin and Apc2 during the onset of denticle formation is particularly striking in this context, because the process of denticle formation is very similar to that of wing hair formation. Similarities are found in the nature of the structure, its strict orientation in the plane of the tissue, and in its cell biological and genetic bases. This raises the possibility that Wg/Wnt signaling directly affects the actin cytoskeleton and thus tissue polarity, using Apc2 as an effector (McCartney, 1999 and references therein).

Although Apc2 does not contain the basic region thought to mediate MT association of hAPC, these data are consistent with the possibility that Apc2, like hAPC, may associate with MTs under certain circumstances. The data for a microtubule association of Apc2 are less robust than those suggesting association with actin. Whereas Apc2 does not prominently localize to most microtubule-based structures (nor does hAPC, unless overexpressed), Apc2 localizes to several places consistent with a role in anchoring microtubules. In preblastoderm embryos, when actin is essential for tethering the spindle to the membrane, Apc2 colocalizes with cortical actin and subcortical actin puncta. Subcortical Apc2 is concentrated just above the spindle, placing it in a position to interact with astral MTs as they reach toward the cortex. Both Apc2 and actin also localize to a dot-like structure, which may be the centrosome. In postblastoderm embryos, Apc2 is subtly enriched in the vicinity of the spindle (McCartney, 1999).

The asymmetric localization of Apc2 in dividing neuroblasts is also consistent with a possible role for Apc2 in linking the spindle to the cortex. During neuroblast mitosis, the spindle is specifically oriented. Insc, which localizes to a crescent opposite the future daughter cell from late interphase through metaphase, coordinates the neuroblast asymmetric cell division. Other proteins are likely to act in this process; e.g., Bazooka acts upstream of Insc (Kuchinke, 1998). In C. elegans and yeast, actin or actin-associated proteins localize asymmetrically in cells in which spindle orientations are specified, suggesting a role for actin in this process. The position of actin in a crescent next to the future daughter cell in a subset of the neuroblasts suggests that actin may affect spindle orientation in Drosophila neuroblasts as well. Apc2 may also play a role in this process. Within the crescent, Apc2 localization is strongest in the region of the spindle pole. During later stages of mitosis, although Apc2 remains enriched in a crescent next to the future daughter, Apc2 also localizes to the cortex on the opposite side of the cell, often in the region of the other spindle pole. In contrast to other asymmetrically localized components of the neuroblast, Apc2 localizes to a crescent during all stages of the cell cycle; in fact, the Apc2 crescent is most apparent during interphase and prophase. Apc2 and actin can also act as polarity markers for other proteins; actin is required for the asymmetric localization of Insc, Prospero, and Staufen. Nevertheless, the Apc2/MT connection remains speculative. In the future, tests will be made to see whether Apc2 associates with MTs, whether Apc2 can affect spindle orientation, and whether Apc2 and Arm function in the neuroblast asymmetric cell division (McCartney, 1999).

This issue is raised in light of the influence of Wnt signaling on mitotic spindle orientation in both C. elegans embryos (Thorpe, 1997) and in Drosophila sensory cells (Gho, 1998 ). In C. elegans, Wnt signaling controls spindle orientation independent of transcription (Schlesinger. 1999 ), suggesting that the Wnt pathway directly targets the cytoskeleton. Since Apc2 regulates Wg/Wnt signal transduction and appears to have connections to the cytoskeleton, it is a candidate for a direct effector of this process. RNA interference studies of C. elegans relatives of Arm (WRM-1) and APC (APR-1) do not reveal defects in spindle orientation (Schlesinger, 1999), suggesting the existence of a branch to the cytoskeleton upstream of APC and Arm. However, because RNA interference may not completely remove gene function, and because of the divergence between APR-1 and the APC family, the involvement of Arm and Apc2 in the pathway remains plausible. These data are also intriguing in light of studies of the hAPC binding protein EB1 (Su, 1995), which colocalizes with the spindle, centrosome, and asters (Berrueta, 1998 ; Morrison, 1998). Budding and fission yeast EB1 homologs are required for spindle assembly and stability (Beinhauer, 1997; Schwartz, 1997; Muhua, 1998). However, it is worth noting that Apc2 appears to lack the binding domain for EB1 identified in hAPC (McCartney, 1999).

APC2 and APC1 have overlapping roles in the larval brain despite their distinct intracellular localizations

The tumor suppressor APC and its homologs, first identified for a role in colon cancer, negatively regulate Wnt signaling in both oncogenesis and normal development, and play Wnt-independent roles in cytoskeletal regulation. Both Drosophila and mammals have two APC family members. The functions of the Drosophila APCs is further explored using the larval brain as a model. Both proteins are expressed in the brain. APC2 has a highly dynamic, asymmetric localization through the larval neuroblast cell cycle relative to known mediators of embryonic neuroblast asymmetric divisions. Adherens junction proteins also are asymmetrically localized in neuroblasts. In addition they accumulate with APC2 and APC1 in nerves formed by axons of the progeny of each neuroblast-ganglion mother cell cluster. APC2 and APC1 localize to very different places when expressed in the larval brain: APC2 localizes to the cell cortex and APC1 to centrosomes and microtubules. Despite this, they play redundant roles in the brain; while each single mutant is normal, the zygotic double mutant has severely reduced numbers of larval neuroblasts. These experiments suggest that this does not result from misregulation of Wg signaling, and thus may involve the cytoskeletal or adhesive roles of APC proteins (Akong, 2002).

APC2 localizes asymmetrically in larval neuroblasts. This study focused on central brain neuroblasts, which are located medially to the proliferation centers of the optic lobes. Neuroblast divisions are asymmetric in size and fate, with the larger daughter remaining a neuroblast and the smaller daughter becoming a ganglion mother cell (GMC). Each central brain neuroblast divides a series of times to produce a 'cap' of GMCs that remain joined to the mother, and divide mitotically themselves (Akong, 2002).

To characterize APC2 dynamic localization during the cell cycle, two parallel approaches were taken. Immunofluorescence and confocal microscopy were used to colocalize APC2, microtubules, and mitotic DNA (via the phosphohistone H3 epitope) in fixed tissue, and an APC2-GFP fusion under the control of the GAL4-UAS system was used; APC2 was driven in a subset of neuroblasts by prospero-GAL4 (pros-GAL4). The localization of APC2 revealed by these approaches is quite similar, though not identical. APC2-GFP accumulated somewhat more uniformly around the cortical circumference; accumulation was present at higher levels in GMCs; it is suspected that these differences reflect the elevated levels of APC2-GFP, although they could be due to localization differences between the GFP fusion and endogenous APC2. Attempts were made to acquire images from neuroblasts that were dividing in the apical/basal plane (Akong, 2002).

During interphase, APC2 forms a strong asymmetric crescent, with APC2 accumulating at the border between the neuroblast and its previous daughters; this region is where the next daughter will be born. The APC2 crescent remains strong through prophase, as the centrosomes separate and begin to set up the spindle. The crescent becomes less pronounced at metaphase. The orientation of the neuroblast spindle at metaphase determines where the GMC is born. Two different relationships were observed between the cortical APC2 crescent and the mitotic spindle, which were present in approximately equal numbers. In about half of the neuroblasts, the forming mitotic spindle was directed toward the center of the APC2 crescent. In the other neuroblasts, the forming mitotic spindle was directed toward one edge of the crescent. The difference was first apparent during late prophase, and continued through metaphase and anaphase. The reason for this difference in APC2 localization in different neuroblasts remains to be determined. One speculative possibility is that the relationship of GMC birth position to the APC2 crescent differs depending on how many GMCs have already been born. The new GMC daughter is usually born adjacent to one of the earlier daughters. At early stages, when there are relatively few GMC daughters, the spindle may be directed toward the middle of the APC2 crescent. In later divisions, when there are more GMC daughters, the new GMC may be born at the edge of the APC2 crescent, which is found at the interface between the neuroblast and the cap of previous GMC daughters (Akong, 2002).

As neuroblasts enter anaphase, APC2 remains cortical. In neuroblasts where the spindle is pointed toward the center of the APC2 crescent, APC2 surrounds the budding daughter. In cells in which the spindle is directed toward the edge of the crescent, APC2 localizes along one side of the GMC. As cytokinesis begins, APC2 localizes prominently to the cleavage furrow in all cases. This enrichment is even more prominent at the end of cytokinesis; at times, the cleavage furrow localization resolved into an apparent double ring. APC2 remains enriched at the division site after cytokinesis, marking the spot where the last daughter was born, and it only gradually reexpands to the interface between the neuroblast and all of its GMC daughters (Akong, 2002).

The live imaging of APC2-GFP also provided a glimpse of the timing of neuroblast cell cycles. Mitosis lasts ~15-20 min. In two cases, a single neuroblast that divided twice was observed, allowing an assessment of the minimum cell cycle length. In these cases, the two mitoses were completed in 106-120 min. This is comparable to the cell cycle length of embryonic neuroblasts (about 50 min), and fits previous estimates of neuroblast cell cycle length from BrdU-labeling experiments. Other neuroblasts within the same brain, however, divided only once, while still others did not divide at all during the ~2.5 h of the movie (Akong, 2002).

One striking feature of the asymmetric localization of APC2 is that it is present throughout the cell cycle and is particularly strong during interphase. During embryonic neuroblast divisions, most asymmetric markers are localized only during mitosis. However, less is known about their localization in larval neuroblasts. Several asymmetric markers in larval neuroblasts were examined, and their localization was compared with that of APC2. In embryonic neuroblasts, the transcription factor Prospero (Pros) and its mRNA are GMC determinants that are asymmetrically localized to the GMC daughter. Pros protein then becomes nuclear and helps direct cell fate. In larval neuroblasts, a similar localization is observed. Pros is not detectable in interphase neuroblasts, when the cortical APC2 crescent is strongest. A small amount of Pros transiently localizes to an asymmetric crescent during mitosis. Pros is present at low levels in GMC nuclei and at higher levels in the nuclei of ganglion cells (Akong, 2002).

Mira is basally localized in embryonic neuroblasts, and required there for localization of Pros protein and mRNA. In central brain neuroblasts, Mira is diffusely cytoplasmic during interphase, when the APC2 crescent is the strongest. As cells enter mitosis, Mira first becomes cortical and then begins to accumulate asymmetrically on the side of the neuroblast where the daughter will be born. By metaphase, Mira asymmetry is very pronounced. The center of the Mira crescent is always precisely aligned with one spindle pole. As a result, in cells with the spindle pointing toward the center of the APC2 crescent, the Mira and APC2 crescents substantially overlap, while in cells in which the spindle points to the edge of the APC2 crescent, the two crescents are offset. Mira is partitioned into the GMC during anaphase, while APC2 relocalizes to the cleavage furrow. Mira could still be detected in some GMCs, which are thought to be those that were recently born (Akong, 2002).

In contrast to Mira and Pros, Inscuteable (Insc) and Bazooka (Baz) localize to the apical sides of embryonic neuroblasts, where they play essential roles in asymmetric divisions. Insc is asymmetrically localized in larval neuroblasts. Insc localizes to the side of the neuroblast opposite that of APC2 through much, if not all, of the cell cycle. Interestingly, there is a weak Insc crescent during interphase, that becomes stronger through prophase and metaphase. During anaphase, Insc localizes to the neuroblast cortex but not the GMC daughter. Baz localization was similar to that of Insc, though no cortical localization during interphase was detected. During prophase and metaphase, Baz localizes to a crescent opposite APC2, and as the chromosomes begin to separate, Baz localizes to a tight cap opposite the future GMC. Together, these data confirm that larval and embryonic neuroblasts asymmetrically localize many of the same proteins, and that APC2 localizes on the GMC side (basal) of the neuroblast, overlapping Mira and opposite Baz and Insc, which localize apically (Akong, 2002).

Arm also localizes asymmetrically in neuroblasts. Extending this, an examination was made of the localization of Arm's adherens junction partners DE-cadherin and ß-catenin. When central brain neuroblasts undergo a sequential series of asymmetric divisions, the GMCs remain associated with their neuroblast mother, resulting in a cap of GMCs in association with each neuroblast. APC2 localizes strongly to the boundary between the neuroblast and each GMC, and more weakly to the borders between the GMCs. APC2 is present at lower levels in ganglion cells and differentiating neurons (Akong, 2002).

The adherens junction proteins DE-cadherin, Arm, and ß-catenin all show a striking and asymmetric localization pattern in central brain neuroblasts. All precisely colocalize both at the boundary between neuroblasts and GMCs and at the boundaries between GMCs. DE-cadherin, Arm, and ß-catenin are also all expressed in epithelial cells of the outer proliferation center. The localization of DE-cadherin and the catenins is consistent with the idea that cadherin-catenin-based adhesion could help ensure that GMCs remain associated with each other, via association with their neuroblast mother (Akong, 2002).

To further explore this, how successive GMCs are positioned relative to their older GMC sisters was examined using two different approaches. First Mira was used to mark the newborn GMCs and DE-cadherin was used to mark the neuroblast and all of her GMC daughters. Mira localizes to a crescent on the side of the neuroblast where the daughter will be born (basal side), and then is segregated into the daughter. Mira persists for some time in newborn GMCs, and it remains detectable in the other GMCs as well, thus allowing the position of newborn GMCs to be examined relative to their older sisters. In many cases, new GMCs are clearly born at the edge of the cluster of older GMCs. This is particularly striking in neuroblasts with many progeny. It is worth noting that the cluster of daughters is three-dimensional, comprising a 'cap' of daughters in three dimensions rather than a two-dimensional line of daughters. It is thus suspected that new daughters are born near the edge of this cap (Akong, 2002).

These data suggest that neuroblasts and their GMC progeny remain closely associated. The GMCs then divide to form ganglion cells and ultimately neurons. The data further suggest that these latter cells may also remain associated and send their axons together toward targets in the central brain. When sections were made more deeply into the brain, below each cluster of neuroblasts and GMCs, structures that appear to be axons were detected projecting from these groups of cells. These axons label with Arm, DE-cadherin, and APC1. Arm also localizes to the axons of the neuropil, while DE-cadherin and APC2 are present at low levels or are absent from this structure (Akong, 2002).

Despite its striking localization, APC2 is not essential for brain development. One possible explanation is that APC1 is also expressed in larval neuroblasts and plays a redundant role there. The localization and function of APC1 in brain development was examined. In embryos, APC1 is expressed in primordial germ cells and in CNS axons. Anti-APC1 antibody was used to examine its expression in the larval brain, using the null allele APC1Q8 as a negative control. APC1 accumulates at apparently low levels throughout the brain. It is largely diffuse throughout the cell, with slight cortical enrichment. There was a small amount of residual staining in the APC1 mutant, that might represent slight crossreactivity with APC2 (cross-reactivity is detectable in immunoblotting). There is a strong accumulation of APC1 in the axons emerging from clusters of neuroblast, GMCs, and their progeny. Together, these data indicate that low levels of APC1 accumulate in neuroblasts, GMCs, and their progeny, with higher levels found in axons projecting from these cells. Thus, APC1 could potentially compensate for loss of APC2 in the larval brain (Akong, 2002).

The cell biological properties of APC1 and APC2 were compared. The four known APC family members, two in flies and two in mammals, share certain structural features but differ in others. All share a block of N-terminal Arm repeats, followed by a set of short repeated sequences that serve as binding sites for Arm/ßcat and Axin. These regions comprise most of fly APC2, which is the shortest family member. Fly APC1 is significantly longer at its N and C termini. Its extended C-terminal region contains a sequence similar to the microtubule-binding domain of human APC (Akong, 2002).

Given these structural differences between fly APC1 and APC2, it was asked whether they would exhibit similar or distinct cell biological properties when both were expressed at equivalent levels. The GAL4-UAS system to overexpress each protein in the same cellular environments, and where each localized under these conditions was examined. When APC2-GFP was overexpressed, it localized asymmetrically to the cortex of the larval neuroblasts, as well as to the junctions between the GMCs and thus paralleled the localization of endogenous APC2. In contrast, when APC1 was overexpressed, it had a strikingly different localization. APC1 localized very strongly to the centrosome and the microtubules emanating from it. It also localized to interphase microtubule arrays in cells of the inner proliferation center. However, it did not strongly colocalize with all microtubule-based structures; for example, it was only marginally enriched in the spindle at metaphase. Similar localization experiments were carried out after mis-expression in the embryonic epidermis, with similar results; APC2 localizes to the cell cortex, while APC1 localizes to centrosomes and microtubules. Together, the data from brains and embryos suggest that sequence differences between APC1 and APC2 direct them to quite different intracellular locations, despite their strong similarity in the core region of the protein (Akong, 2002).

Mammalian APC protein can oligomerize, and thus whether fly APC1 and APC2 proteins might interact was examined. To begin to examine this, the effect of APC1 overexpression on the localization of endogenous and overexpressed APC2 was examined. This study suggested that APC1 and APC2 may interact, either directly or indirectly, allowing APC1 to recruit APC2 to a new location. APC2 may recruit APC1 to the cell cortex; however, this is subject to the caveat that the APC1 antibody may weakly cross-react with APC2 (Akong, 2002).

Given the striking localization of APC2 in the brain, it was surprising to find that APC2 mutants have no apparent brain defects. The brain is only one of several tissues where neither single mutant had any apparent phenotype; this was surprising as Wg signaling plays an important role in several of these tissues. The realization that APC1 is also expressed in the brain raised the possibility that it might play a partially redundant role in this tissue. To test this, the phenotype of animals mutant for both APC1 and APC2 was examined (Akong, 2002).

APC1 and APC2 single mutants are both zygotically viable to adulthood. The only known problem in APC1 mutants are morphological defects and inappropriate apoptosis in the photoreceptors of the eye, while APC2 zygotic mutants are morphologically wild type, with a phenotype only emerging in embryos maternally and zygotically mutant. In contrast, APC2;APC1 double mutants are zygotically lethal. Two allelic combinations, APC2deltaS;APC1Q8 and APC2g10;APC1Q8, die as second instar larvae, while the third, APC2d40;APC1Q8, dies primarily during the first larval instar. None of the double mutants had any apparent defects in segment polarity as first instar larvae. Mitotic activity during larval stages is restricted to the imaginal tissues and the brain, and imaginal discs are dispensable for larval development. Given this and the observed expression pattern of APC2 and APC1 in the CNS, brain development was examined in double mutant larvae (Akong, 2002).

During normal brain development, most embryonic neuroblasts exit the cell cycle during late embryogenesis. Immediately after hatching, the only mitotically active neuroblasts are the so-called mushroom-body neuroblasts, which are at a ventrolateral position. Eight hours after hatching, other neuroblasts become mitotically active, such that by 20 h after hatching, 20-30 central brain neuroblasts per hemisphere are dividing. During the early first instar, the cells of the optic anlage become epithelially arranged and also begin dividing. The number of proliferating neuroblasts continues to increase, plateauing 20-50 h after egg laying (Akong, 2002).

APC2deltaS;APC1Q8 zygotic double mutants die during the second larval instar. Wild-type and double mutant animals were compared immediately after they completed the second instar larval molt. Double mutant brains are essentially normal in size, and the optic anlage becomes epithelial. However, double mutants have a strikingly different pattern of neuroblast proliferation than wild type. While presumptive mushroom body neuroblasts continued to proliferate, the number of other mitotic neuroblasts is drastically reduced relative to wild type, as assessed both by phosphohistone H3-labeling and by Mira and Pros staining. A similar, though less drastic, block in mitotic activity was seen in APC2g10;APC1Q8. To determine whether DNA synthesis was initiated but mitosis blocked, larvae were labelled with BrdU throughout the first instar, identifying cells that replicated their DNA during this period. The results were similar to those seen with mitotic markers. APC2deltaS;APC1Q8 double mutants have drastically reduced numbers of BrdU-labeled cells; most labeled cells remaining appeared to be mushroom-body neuroblasts. It was next examined whether embryonic CNS development was altered, by examining the axonal scaffold produced by progeny of the embryonic neuroblasts, using the marker BP102 that labels all axons. Zygotic double mutants were distinguished from wild-type siblings by the presence or absence of APC1 in the CNS. The axonal scaffold was unaltered in APC2deltaS;APC1Q8 double mutants, suggesting that embryonic neuroblast proliferation is at least roughly normal. Finally, whether defects could be detected in asymmetric divisions in double mutants was examined, as assessed by examining asymmetric localization of Mira. In the wild-type first and second instar brain, the localization of Miranda is less strikingly asymmetric, with reasonably high levels in the cytoplasm. However, crescents of Mira could be detected and they localize to smaller daughters. In double mutants, many fewer Mira-positive neuroblasts were seen, but neuroblasts with Mira crescents or Mira segregated to smaller daughters were found. In addition, Prospero-positive daughters are found in the double mutants, though in reduced numbers (Akong, 2002).

These data thus suggest that asymmetric divisions still can occur, though it remains possible that they are not entirely normal or occasionally fail. Together, these data suggest that most likely explanation for the mutant phenotype is that double mutant embryonic neuroblasts do not reenter the cell cycle during the first larval instar (Akong, 2002).

APC1 and APC2 regulate Wg signaling, and thus one mechanism that could underlie the zygotic double mutant phenotype is the failure to properly regulate Arm levels. To test this, Arm accumulation was examined in zygotic double mutant embryos. Stage 16 double mutants were identified by the absence of APC1 accumulation in the CNS, and the level of Arm accumulation was examined relative to wild-type embryos. Arm accumulation appeared completely normal in the epidermis of APC2g10;APC1Q8 double mutant embryos, suggesting that maternally contributed protein from the two loci is sufficient for normal Arm regulation in the embryo (Akong, 2002).

Arm accumulation was examined in the brains of midfirst instar larvae. In the wild-type brain, Arm accumulates heavily in axons of the ventral nerve cord and neuropil. Arm accumulation in the cellular portion of the first instar brain is much lower. In the brain lobes, weak cortical staining of most cells was found. The only significant Arm accumulation outside the neuropil is in cells that are believed to be epithelial cells of the developing optic lobe. The accumulation of Arm in zygotic double mutant brains is not substantially elevated from that of wild-type. Arm levels in the neuroblasts and cell bodies are unchanged, while Arm levels in axons are similar or only slightly elevated. These data thus do not support the hypothesis that elevated Arm levels cause the phenotype (Akong, 2002).

These data demonstrate that APC1 and APC2 play redundant roles in larval brain development. Further, the data suggest that this role is Wg-independent, since it is not mimicked by elevating Arm levels. This contrasts with the situation in embryos and imaginal discs, where the two proteins also have overlapping functions but where these clearly involve Wg signaling. APC1 and APC2 have also been found to play redundant roles in cell adhesion during oogenesis. Maternal contribution of the two proteins appears sufficient for many if not all aspects of embryogenesis, since double mutant embryos hatch with a normal cuticle pattern and an apparently normal brain. However, striking differences between the larval brains of wild-type and double mutant animals were seen. During wild-type development, most neuroblasts become quiescent during late embryogenesis, with only the mushroom body neuroblasts mitotically active upon hatching. In the middle of the first instar, however, other embryonic neuroblasts reenter the cell cycle and proliferate. In the APC2;APC1 double mutant, this appears not to occur, as far fewer total neuroblasts are seen, and the number of brain cells in mitosis and the number that have gone through S-phase is substantially reduced (Akong, 2002).

Adenomatous polyposis coli is present near the minimal level required for accurate graded responses to the Wingless morphogen

The mechanisms by which the Wingless (Wg) morphogen modulates the activity of the transcriptional activator Armadillo (Arm) to elicit precise, concentration-dependent cellular responses remain uncertain. Arm is targeted for proteolysis by the Axin/Adenomatous polyposis coli (Apc1 and Apc2)/Zeste-white 3 destruction complex, and Wg-dependent inactivation of destruction complex activity is crucial to trigger Arm signaling. In the prevailing model for Wg transduction, only Axin levels limit destruction complex activity, whereas Apc is present in vast excess. To test this model, Apc activity was reduced to different degrees, and the effects were analyzed on three concentration-dependent responses to Arm signaling that specify distinct retinal photoreceptor fates. It was found that both Apc1 and Apc2 negatively regulate Arm activity in photoreceptors, but that the relative contribution of Apc1 is much greater than that of Apc2. Unexpectedly, a less than twofold reduction in total Apc activity, achieved by loss of Apc2, decreases the effective threshold at which Wg elicits a cellular response, thereby resulting in ectopic responses that are spatially restricted to regions with low Wg concentration. It is concluded that Apc activity is not present in vast excess, but instead is near the minimal level required for accurate graded responses to the Wg morphogen (Benchabane, 2008).

Previous genetic studies have provided conclusive evidence that the two Drosophila Apc proteins are crucial negative regulators of Arm signaling. Simultaneous inactivation of both Apc proteins results in ectopic Arm signaling in nearly all, if not all, cells, indicating that Apc is required to prevent Arm signaling in the absence of Wg stimulation. In contrast with the prevailing model for Wg transduction, which proposes that Apc is present in vast excess, the work presented in this study reveals that a less than twofold reduction in Apc activity can shift the threshold for the response to Wg. It is concluded that by negatively regulating Arm, Apc prevents ectopic Arm activity not only where Wg is absent, but also within the range of the Wg gradient (Benchabane, 2008).

Translation of a gradient of Wg morphogen activity to quantitatively distinct levels of Arm signaling is required to induce concentration-dependent cellular responses, although the mechanisms by which this occurs remain uncertain. The current results reveal that in regions of low Wg concentration, reducing total Apc activity by less than twofold results in aberrant cell fate specification. A morphogen model predicts that the low Wg concentration present in this region of the gradient is below the threshold necessary to trigger a detectable cellular response. This is the only region within the Wg gradient where a relatively small reduction in total Apc activity elicits an ectopic cellular response, and this response is characteristic of intermediate-level Arm signaling. Thus, these results reveal that Apc activity is in excess in regions where Wg is absent, but is not in vast excess within the range of the Wg gradient. Together, these data indicate that Apc activity is present near the minimal level required to prevent ectopic Arm signaling and thereby ensure accurate graded responses (Benchabane, 2008).

In Xenopus egg extracts, the levels of Axin are several magnitudes lower than the levels of other proteins in the destruction complex, suggesting that only Axin is a limiting component in Arm proteolysis, whereas Apc is present in vast excess. How can these biochemical data be reconciled with the current in vivo data, which indicate that Apc is not present in excess within the range of the Wg gradient? One possibility is that the levels of Apc in Xenopus eggs are much greater than those present in Drosophila photoreceptors. Alternatively, total Apc levels could be present in excess regardless of cell type or organism, but the relevant pool contributing to destruction complex activity, distinguished by either post-translational modification and/or intracellular localization, might be present near threshold levels. A correlation between the degree of reduction in the activity of the fly and mammalian Apc proteins with the level of β-catenin/Arm signaling has been demonstrated in several other developmental contexts and in tumorigenesis. Thus data from diverse experimental models indicate that the level of Apc contributes to the level of β-catenin/Arm signaling (Benchabane, 2008).

How is a gradient of Wg concentration translated into quantitatively distinct levels of Arm activity? Upon Wg stimulation, inactivation of the Axin/Zw3/Apc destruction complex is the primary event that triggers Arm signaling. Inactivation of Axin is important for downstream signal transduction in response to Wg stimulation, and is likely to be mediated by the translocation of Axin to the plasma membrane, and/or the degradation of Axin. Thus the local Axin concentration is likely to have a significant role in determining whether the destruction complex is assembled, and consequently is important in regulating Arm stability. The current findings provide in vivo evidence that the level of destruction complex activity is crucial for accurate patterning in response to Wg, and is dependent not only on Axin, but also on the maintenance of Apc activity above a minimal level. It is concluded that within the range of the Wg gradient, both Axin and Apc are present near threshold levels, and that, together, they achieve the precise levels of destruction complex activity required for accurate graded responses (Benchabane, 2008).

cDNA clone length - 4044 bases

Bases in 3' UTR - 645

Amino Acids - 1067

Structural Domains

All APC family members share an NH2-terminal conserved domain, 6 Arm repeats, and a series of ßcat binding (15 and 20 amino acid repeats) and Axin binding (SAMP repeats) motifs. Drosophila Apc2 is shorter at its NH2 and COOH termini than other APCs. Apc2 lacks the COOH-terminal basic region (the putative MT binding site) found in hAPC and Drosophila Apc-like (Hayashi, 1997), as well as the hAPC region containing binding sites for Discs-large (DLG) and EB1. Substantial alternative splicing is unlikely, as there are only two small introns in coding sequences (63 and 197 nucleotides). Drosophila Apc2 is most similar to other APC family members in the Arm repeats, where it most closely resembles Apc-like; hAPC2 is more similar to hAPC (Apc2 is 81% identical to Drosophila Apc-like and 57% identical to hAPC). Thus, there is no correspondence between individual human and fly proteins, even though both phyla show neural-enriched isoforms (Drosophila Apc-like and hAPC2), suggesting independent gene duplications. All APCs have six Arm repeats; a putative seventh Arm repeat is much more divergent and is not identifiable in Drosophila Apc2. The NH2-terminal conserved region (61% identity to Drosophila Apc-like vs. 44% identity to hAPC) distantly resembles the Arm repeat consensus and may form one or two degenerate Arm repeats. APC family members also share similarity on the C-terminal side of the Arm repeats. hAPC has two sets of repeated ßcatenin binding sites, the 15 and the 20 amino acid repeats; hAPC2 lacks the 15 amino acid repeats). Apc2 shares two of the three 15 amino acid repeats of Apc-like. Drosophila Apc-like and Apc2 each have five 20 amino acid repeats, among which are interspersed SAMP repeats. Apc-like has four SAMP repeats, whereas Apc2 has two. Apc2 ends 40 amino acids after the last SAMP repeat (McCartney, 1999).

For information about APC homologs see Apc-like.

date revised: 12 February 2000

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