IfkA, a presumptive eIF2α kinase of Dictyostelium, is required for proper timing of aggregation and regulation of mound size
© Fang et al; licensee BioMed Central Ltd. 2003
Received: 3 February 2003
Accepted: 9 April 2003
Published: 9 April 2003
The transition from growth to development in Dictyostelium is initiated by amino acid starvation of growing amobae. In other eukaryotes, a key sensor of amino acid starvation and mediator of the resulting physiological responses is the GCN2 protein, an eIF2α kinase. GCN2 downregulates the initiation of translation of bulk mRNA and enhances translation of specific mRNAs by phosphorylating the translation initiation factor eIF2α. Two eIF2α kinases were identified in Dictyostelium and studied herein.
Neither of the eIF2α kinases appeared to be involved in sensing amino acid starvation to initiate development. However, one of the kinases, IfkA, was shown to phosphorylate eIF2α from 1 to 7 hours after the onset of development, resulting in a shift from polysomes to free ribosomes for bulk mRNA. In the absence of the eIF2α phosphorylation, ifkA null cells aggregated earlier than normal and formed mounds and ultimately fruiting bodies that were larger than normal. The early aggregation phenotype in ifkA null cells reflected an apparent, earlier than normal establishment of the cAMP pulsing system. The large mound phenotype resulted from a reduced extracellular level of Countin, a component of the counting factor that regulates mound size. In wild type cells, phosphorylation of eIF2α by IfkA resulted in a specific stabilization and enhanced translational efficiency of countin mRNA even though reduced translation resulted for bulk mRNA.
IfkA is an eIF2α kinase of Dictyostelium that normally phosphorylates eIF2α from 1 to 7 hours after the onset of development, or during the preaggregation phase. This results in an overall reduction in the initiation of protein synthesis during this time frame and a concomitant reduction in the number of ribosomes associated with most mRNAs. For some mRNAs, however, initiation of protein synthesis is enhanced or stabilized under the conditions of increased eIF2α phosphorylation. This includes countin mRNA.
Dictyostelium is one of the simplest studied eukaryotes that possesses true multicellularity . Dictyostelium amoebae grow and divide asexually while feeding on bacteria or within an enriched broth. When the food supply is depleted, Dictyostelium cells shut down growth and cell division and enter a developmental program designed to produce and disperse spores. Mounds of about 105 cells form as cells stream together though chemotaxis in response to cAMP pulses. During late aggregation, the initially identical cells differentiate into several prestalk and prespore cell types, sort in specified ways, and form a finger/slug that undergoes transient or prolonged migration depending on the environmental conditions. Culmination eventually ensues, resulting in a fruiting body with a sorus of spores held several millimeters above the substratum by a vacuolated cellular stalk, and thus situated for dispersal.
The initiating events of development of Dictyostelium include sensing starvation and cell density, which in turn result in the dispersed cells acquiring the ability to aggregate. The mechanism of sensing the density of starved cells insures that aggregation occurs only when there are sufficient numbers of starved cells to form aggregates and subsequent structures of appropriate size for optimized spore dispersal [2–4]. Hence, starvation and a threshold of cell density are the two known prerequisites for the transition from growth to development.
Two secreted proteins or protein complexes are involved in sensing cell density . Prestarvation factor (PSF) is a glycoprotein that is secreted when cells are growing and accumulates as an indicator of the ratio of cell density relative to the supply of food [6, 7]. When the ratio of Dictyostelium cells to nutrients is above a certain threshold, PSF induces the expression of several early developmental genes including discoidin I, lysosomal enzymes, and some components of the cAMP pulsing system [8, 9]. Once nutrients are depleted, PSF production declines and a second cell density-sensing pathway is activated. Conditioned medium factor (CMF) is a 80 kDa glycoprotein that is essential for early development [3, 4, 10]. CMF is sequestered in vegetative cells and is secreted upon starvation . A critical concentration of extracellular CMF is required for subsequent development as CMF is involved in regulating aggregation, cAMP pulsing, and early developmental events [2, 3, 11].
Recently several secreted protein factors were identified that control the size of aggregates and subsequent developmental structures. A large protein complex, counting factor, was purified from conditioned medium and shown to be involved in causing streaming cells to break up into groups of cells in order to generate mounds and subsequent developmental structures of the desired size [12, 13]. One of the subunits of counting factor was identified and characterized, and cells that are null for the Countin subunit lack counting factor activity . The result is massively large mounds and subsequent structures, with fruits that are too large to maintain their normal upright posture. Recently, another protein factor, countin2, was identified as a regulator of the minimum size of aggregates .
Although recent investigations have revealed several components involved in regulating the initiation of development [15–18], little or no information exists on how the cells sense starvation and in particular amino acid deprivation. Early studies indicated that depletion of amino acids and not other nutrients is essential for initiation of development [19–21]. Upon amino acid starvation, among other events there is a substantial decrease in the rate of initiation of protein synthesis, resulting in a reduction in size and amount of polysomes and accumulation of free ribosomes [19, 22, 23]. It is believed that the decrease in protein synthesis in response to starvation occurs post-transcriptionally, since adding back amino acids rapidly restores protein synthesis and does not require new mRNA synthesis .
In yeast and mammalian cells, GCN2 is a protein that senses amino acid starvation and brings about altered cell physiology in response to such starvation [24, 25]. GCN2 is a protein kinase that phosphorylates the α subunit of the translation initiation factor eIF2 . The GCN2 protein from yeasts, Drosophila, and mammals is composed of three domains: a pseudokinase domain that may function as a regulatory domain, a kinase domain, and a HisRS domain that shares significant similarity to histidyl-tRNA synthetases [27–30]. Studies of yeast GCN2 revealed that the HisRS domain binds uncharged tRNAs that accumulate when cells are starved for amino acids [31, 32]. As a result the kinase domain is activated and phosphorylates eIF2α at Ser51. Investigations with yeast GCN2 also suggest that that C-terminus is required for ribosome binding and facilitates GCN2 dimerization and activation [33–35].
During initiation of protein synthesis, the ternary complex, eIF2/GTP/methionyl-tRNAi, is an obligate intermediate in the binding of the initiator methionyl-tRNAi to the 40S ribosomal subunit. Phosphorylated eIF2α competitively inhibits the guanine nucleotide exchange factor eIF-2B, which is responsible for the recycling of eIF2 by catalyzing the conversion of eIF2-GDP to the active form eIF2-GTP . Thus the overall initiation of translation is inhibited when eIF2α becomes phosphorylated. Although the overall rate of initiation of translation is reduced by eIF2α phosphorylation, the translation of some specific mRNAs is greatly enhanced under these conditions [24, 25]. The characteristic feature of mRNAs whose translation is enhanced by GCN2 phosphorylating eIF2α are several upstream open reading frames (uORF) within the 5' untranslated region (5'UTR) of the mRNAs. The uORFs are involved mechanistically in the enhanced initiation of these mRNAs .
We reasoned that a GCN2-like protein might be involved in sensing amino acid starvation in Dictyostelium and in regulating the growth to development transition. Searches of the Dictyostelium sequence databases revealed two potential GCN2-like genes that we named ifkA and ifkB (initiation factor kinase). The two Ifk proteins possess all of the hallmarks of typical GCN2 proteins, including the characteristic kinase domain and the HisRS domain. IfkB is unique among GCN2-like proteins in that it does not possess the amino terminal pseudokinase domain. The findings presented herein indicate that neither of the ifk genes is involved in initiating development. However disruption of the ifkA gene resulted in several defects during the first few hours of development. ifkA null cells aggregated earlier than normal due to a more rapid establishment of the cAMP pulsing system. Cells lacking a functional IfkA protein also formed larger mounds and subsequent developmental structures. This defect was linked to a reduced extracellular level of Countin. Efficient translation of countin mRNA required increased phosphorylation of eIF2α by IfkA. The increased phosphorylation by IfkA normally occurs from 1 to 7 hours after the onset of development, resulting in a secondary shift from polysomes to free ribosomes for bulk mRNA.
Sequence of ifkA and ifkBgenes
The overall sequence identity between yeast GCN2 and the predicted Dictyostelium IfkA protein is 32%. Within the amino terminal sequences, yeast and other GCN2 proteins have a degenerate kinase domain of unknown function . There is a 63 amino acid long sequence in IfkA that has 38% identity to the V and VI subdomains of the IfkA/B kinase domain, indicating the presence of an amino terminal pseudokinase domain in the Dictyostelium protein (fig. 1B). The sequence identity within the HisRS domain of IfkA/B and yeast GCN2 is 28%, with strong conservation within the M1, M2, and M3 motifs that are characteristic of class II aminoacyl-tRNA synthetases . The similarity between the kinase domains is much higher, with IfkA/B possessing 44% identity with yeast GCN2, with high conservation of all of the kinase subdomains. A significant difference between the Ifk kinase domain and yeast GCN2 kinase domain is the length of the insert sequence between subdomain IV and V, which is around 290 amino acids in the Dictyostelium Ifk protein and 110 amino acids in yeast GCN2.
Expression of ifkA and ifkBmRNA during growth and development
Disruption of the ifkAgene
To examine the function of the ifk genes, we attempted to disrupt each gene individually. To date, several disruption attempts, using several different disruption constructs, have been made for ifkB. However, no disruptions have been obtained, indicating that perhaps the ifkB gene is essential. Attempts to disrupt ifkA were carried out using a plasmid that contained a blasticidin resistance cassette flanked by two fragments of the ifkA gene, each about 800 bp long and derived from the pseudokinase and kinase domains. The blasticidin resistance cassette replaced a 391 bp fragment that includes the first four subdomains (the ATP binding site) of the kinase domain. As only the 3' flanking sequence of the blasticidin resistance cassette is found in ifkB but not the 5' flanking sequence, the disruption construct should be specific for ifkA.
Disruption of the ifkAgene results in earlier aggregation and formation of larger than normal mounds
In addition to precocious aggregation, BS153 mounds were significantly larger than those of the parental strain. The size of BS153 mounds was more variable than Ax4, but the majority of the mounds and subsequent structures were much larger than those of Ax4 (fig. 4). Attempt to determine cell numbers in BS153 fingers were hampered by apparently stronger than normal cell-cell adhesion and thus incomplete dispersal of the cells for counting. Based on the number and volume of the mounds relative to those of Ax4, we estimated that most BS153 mounds were more than 50% larger than Ax4 mounds.
The timing of the conversion of mounds to fingers was normal for BS153 cells. Fingers that formed from the smaller mounds went through a transient slug stag of about 1 hour, comparable to what occurred with Ax4 fingers. However, most of the larger fingers fell to the substratum and underwent extended migration as slugs (4–6 hours). During migration, more cells than normal were sloughed off and left behind in the slime trail. Although this reduced their size, after culmination (which was completed with normal timing) many of the fruiting bodies of BS153 possessed thickened stalks and large sori (fig. 4). Several fruits were unable to remain upright. Germination of BS153 spores was examined and found to be similar to that of Ax4.
When developed at low cell density in monolayers, 1.6 × 105 cells/cm2 instead of 1.0 × 106 cells/cm2, the ability of BS153 cells to form large aggregates was more evident. As shown in fig. 4, BS153 cells formed a few large fruiting bodies with thick stalks and a few small fruits, while Ax4 formed over 20 typically proportioned fruiting bodies of small but uniform size. Development of the cells in monolayers in buffer or in conditioned medium from BS153 or Ax4 cells  indicated that secreted factors were involved in causing both the earlier aggregation of BS153 cells and the larger size of the resulting multicellular entities. (not shown). One secreted factor that is involved in regulating aggregation, cAMP pulsing, and early developmental events is CMF [2, 3]. CMF mRNA levels were normal in BS153 during growth and development (not shown). In addition, CMF protein levels, both cellular and secreted, also were normal in growing and developing BS153 cells (fig. 8B).
Precocious aggregation of BS153 cells reflects earlier than normal establishment of the cAMP pulsing system
To examine at the molecular level the reason for the earlier aggregation of ifkA null cells, the mRNA levels were examined for two proteins involved in the cAMP pulsing mechanism of aggregation and chemotaxis. The cAMP cell surface receptor Car1 (carA gene) is responsible for sensing secreted cAMP and activating adenylyl cyclase (acaA gene), which in turn results in the production and secretion of cAMP in order to relay the cAMP pulses . Normally both of these genes are expressed at very low levels in growing cells, and their expression is induced during the first few hours of development, including a stimulation of expression by the cAMP pulses themselves .
Disruption of ifkAleads to reduced eIF2α phosphorylation and altered polysome profiles
Percent polysomes and monosomes for different RNA samples isolated from AX4 and BS153.
O hr. Ax4
O hr. BS153
10 min. Ax4
10 min. BS153
4 hr. Ax4
4 hr. Ax4
4 hr. BS153
4 hr. BS153
Disruption of ifkAresults in misregulation of countin
Counting factor is a secreted, large protein complex that regulates the number of cells within and thus the size of mounds . One of the polypeptides of the counting factor complex is Countin. Disruption of the countin gene results in extremely large mounds and fruiting bodies . As noted above, one of the phenotypes associated with disruption of the ifkA gene was large aggregates and fruiting bodies, and these aberrations appeared to result from a secreted factor(s). Thus, the expression of countin was examined in the mutant cells.
In the parental strain Ax4, the level of countin mRNA gradually decreased with time after the onset of development (fig. 8A), consistent with previous findings . In contrast, the decrease of countin mRNA was much more dramatic in BS153, with little detectable mRNA present by 8 hours. To study the synthesis and secretion of the Countin protein, BS153 and Ax4 cells were removed from the bacterial food source and shaken in normal starvation buffer for varying times. Samples of both cells and (conditioned) medium were processed for western analysis using antiserum specific for the Countin protein (fig. 8B). In both Ax4 and BS153 cells, Countin remained at a constant level within cells with time after starvation. At each time point, however, less countin was found in BS153 whole cell lysates relative to the levels in Ax4 lysates (Cell panels).
Analysis of the conditioned medium from Ax4 cells indicted that extracellular Countin increased with time as was expected fig. 8B, CM panels). However for BS153, levels of secreted Countin were reduced relative to those of Ax4, and by six hours only modest amounts of Countin had accumulated that were roughly equivalent to the accumulated level after 1.5 hours for Ax4 (CM panels). In contrast to the findings for extracellular Countin, levels of secreted CMF were the same for Ax4 and BS153 cells (CMF panels), thus ruling out a general decrease in sectreted proteins in the mutant strain. Although extracellular Countin did appear to accumulate with starved BS153 cells, its concetration was reduced and accumulation was slower compared to Ax4 cells. These findings suggest that the increased size of mounds and fruiting bodies characteristic of the ifkA null strain may be a result of decreased production and accumulation of extracellular Countin.
Disruption of ifkA results in altered translation efficiency and stability of countinmRNA
RNA was extracted from the fractions collected after sucrose gradient centrifugation of cell extracts isolated four hours after starvation. The distribution of countin mRNA was determined by RT-PCR. Fractions 3 through 11 represent polysomes with decreasing numbers of associated ribosomes going from fraction 3 to fraction 11, while fractions 12 through 16 represent free ribosomes (fig. 7). In both BS153 and Ax4 developing cells, countin mRNA was found mostly in the polysome fractions (fig. 9B). However in BS153 in comparison to Ax4, the relative abundance of countin mRNA was reproducibly less in the high polysome fractions and greater in the low polysome and free ribosome fractions (fig. 9B and 9C). The shift in relative abundance of countin mRNA within the polysome profiles of the two strains is opposite the secondary shift identified above for bulk mRNA and for the shift seen for H7 mRNA (fig. 9B). No difference in the distribution of countin mRNA was found between the two strains when polysome profiles from growing cells were examined (fig. 9C). The results indicate that in the absence of IfkA and its phosphorylation of eIF2α during early development, countin mRNA is retained less well on polysomes and has fewer associated ribosomes. This suggests that normally during early development in the wild type strain, enhanced translation of countin mRNA occurs as a result of IfkA phosphorylating eIF2α.
Recent findings indicate the presence of another Countin-like factor, Countin2, that is involved in regulating aggregate size . Countin2 mRNA levels were found to be the same in BS153 and Ax4 during growth and development (not shown).
The multicellular developmental program of Dictyostelium is a response to starvation with the goals being the survival of some cells, in the form of protected spores, and dispersal of the survivors in hopes of finding an environment that supports growth. Depletion of amino acids as opposed to other nutrients such as glucose and vitamins was shown to be responsible for shutting down growth and cell division and initiating the developmental program [19–21]. Although substantial progress on the regulation of the initiation of development in Dictyostelium has been made [15–18], how the cells sense starvation and in particular amino acid deprivation is unknown. GCN2 of yeast and mammals is a known sensor of amino acid starvation and regulates cellular responses to such starvation [24, 25]. Using degenerate primers and PCR and database searching identified two GCN2-like genes in Dictyostelium, ifkA and ifkB. Thinking that one or both of these genes may play a role in sensing amino acid starvation and in initiating development, we carried out the work described herein to examine their functions.
The predicted amino acid sequences of IfkA and IfkB are nearly identical within the kinase domain and the HisRS domain, and in turn these two domains share significant sequence identity with GCN2 proteins from yeast, Drosophila, and mammals. This includes the characteristic features peculiar to the kinase domain of GCN2 proteins, the m1, m2, and m3 motifs of the HisRS domain, and a cluster of positively charged residues at the C-terminus that are thought to function in ribosome binding and dimerization [31, 37, 41]. Within the kinase domain, the Dictyostelium proteins possess an inserted sequence between the 4th and 5th kinase subdomains that is also found in this position in all known GCN2-like proteins. The insert in IfkA and IfkB is significantly longer than that of other GCN2 proteins. The major difference between IfkA and IfkB is the amino terminal extension of about 900 amino acids that is found in IfkA but not IfkB. Similar extensions amino terminal to the kinase domain are found in GCN2 proteins from other species, and the function of this portion of the protein is unknown. This region has limited sequence identity to the kinase domain, and thus has been termed the pseudokinase domain . Similarly, the extension in IfkA shows limited sequence identity to the kinase domain. In contrast, IfkB begins at the kinase domain and is the only known GCN2-like protein that does not possess a pseudokinase domain.
Even though a number of attempts have been made, we have been unable to disrupt the ikfB gene. This suggests it is an essential gene, a property not found for eIF2α kinases of other organisms, and we currently have no insight into its function. Although it is possible that IfkB plays a role in the transition from growth to development, we do not think it does. A basal level of eIF2α phosphorylation was found in growing cells, and this basal level was maintained during development in the ifkA null strain. Thus we have not detected any changes in eIF2α phosphorylation that can be attributable to regulated IfkB activity. Instead, only a basal and unchanging level of phosphorylation during growth and development can be attributed to IfkB.
A very early event marking the initiation of development is a substantial reduction in overall translation initiation and a concomitant major shift of ribosomes from mRNA-associated ribosomes (polysomes) to free ribosomes [22, 23]. The percentage of mRNA associated with polysomes drops from about 80–90% in growing cells to 50–60% within several minutes of the removal of the food source. We detect no increase in eIF2α phosphorylation within this time frame, indicating that enhanced eIF2α phosphorylation is not the mechanism used to bring about the rapid, initial decrease in the initiation of protein synthesis.
Disruption of the ifkA gene was accomplished, and several defects were found in the null strain (BS153) that indicted a role for IfkA and eIF2α phosphorylation during the first several hours of development. Examining the morphology of the null strain revealed two major aberrations relative to the normal morphology attained during development. BS153 cells aggregated and formed mounds two to four hours earlier than the parental cells. Earlier aggregation was seen when cells were developed at both high and low cell densities.
The early aggregation phenotype associated with BS153 suggests that the cAMP pulsing system was established earlier than normal. This was confirmed, indirectly, by examining the expression of several of the components of the pulsing system. mRNA for both Carl, the cell surface cAMP receptor, and AcaA, the adenylyl cyclase responsible for cAMP synthesis, was found to be induced 2 to 4 hours earlier than in the parental cells. The levels of these mRNAs were also greater in developing BS153 cells. Thus, the components of the pulsing system are produced earlier and to higher levels when IfkA is not functioning. In addition, these genes are "pulse-induced" genes and thus their induction during early development normally is enhanced and promoted as the cAMP pulses are established . The early and more abundant induction of their mRNAs in BS153 indicate that the pulsing system is functioning earlier in the absence of IfkA. The earlier and more rapid loss of discoidin mRNA reinforces this conclusion as discoidin is a "pulse repressed" gene .
Examination of eIF2α phosphorylation in ifkA null cells indicated that the IfkA protein is responsible for increased phosphorylation eIF2α from 1 to 7 hours after the onset of development. Although an increase was seen in the parental strain during this time frame, eIF2α phosphorylation levels in developing BS153 cells remained at the lower basal level found in growing cells.
Concomitant with the increased eIF2α phosphorylation in the parental strain was a secondary shift of ribosomes from polysomes to free ribosomes. A decrease in polysomes of about 15 % and an equivalent increase in free ribosomes was observed when the polysome profile of Ax4 cells developed for 2 or 4 hours was compared to the profile from BS153 cells developed for the same time or to the profile from Ax4 cells developed for only 10 minutes. These results indicate that IfkA is normally active from about 1 hour to about 7 hours after starvation and normally phosphorylates eIF2α during this time period (however, we have no direct data that demonstrates IfkA phosphorylates eIF2α).. IfkA activation at these times results is a reduction in overall protein synthesis initiation as indicated by the secondary shift of polysomes to ribosomes. Based on the normal function of GCN2-like proteins , these findings suggest that uncharged tRNAs are more prevalent within the cells during this time relative to either growing cells or to cells immediately after removal of the food source.
In addition to the earlier aggregation phenotype of developing BS153 cells, the IFkA null cells also formed atypically large aggregates, mounds, and subsequent developmental structures. This phenotype is reminiscent of cells that lack Countin, a component of counting factor that regulates the size of aggregates [12, 13]. Countin mRNA was found to be lost significantly more rapidly in developing BS153 cells relative to the parental Ax4 cells. Examination of the stability of countin mRNA indicated that it was turned over more rapidly in BS153 cells during early development. This suggests that countin mRNA is destabilized in developing cells that lack IfkA function, and we assume that countin mRNA synthesis is not affected in these cells. We suggest the destabilization of countin mRNA may be a result of the loss of ribosomes, which otherwise would afford protection from degradation, from this mRNA in BS153 cells.
The distribution of countin mRNA within the polysome profile was altered in developing BS153 cells. As described above, bulk mRNA undergoes a secondary shift from polysomes to free ribosomes in Ax4 cells, and this shift was not seen in BS153 cells as it is dependent upon eIF2α phosphorylation by IfkA. More ribosomes were found to be associated with countin mRNA in early developing Ax4 cells relative to BS153 cells, which is the converse ribosome association with bulk mRNA in the two strains. This finding indicates that countin mRNA is translated more efficiently when IfkA is functioning properly. This is similar to the enhanced translation of certain yeast and mammalian mRNAs (GCN4, ATF4, both transcription factors regulated by GCN2) when GCN2 in these organisms is activated by amino acid starvation, even though bulk mRNA undergoes a decrease in translation efficiency [24, 25]. The hallmark of mRNAs that are upregulated by GCN2 and eIF2α phosphorylation is the presence of several upstream open reading frames within their 5' UTRs. The uORFs are involved mechanistically in bringing about enhanced translation initiation under conditions that otherwise favor a general decrease in translation initiation . Within the 5' UTR of countin mRNA, we find uORFs similar in number and in arrangement to those of GCN4 mRNA, the major mRNA regulated by yeast GCN2 . Thus, we suggest that countin mRNA, and probably other as yet identified mRNAs, are translationally upregulated when IfkA phosphorylates eIF2α during the preaggregation period, or from 1 to 7 hours after the onset of development.
A result of the lack of increasing the translational efficiency of countin mRNA in developing ifkA null cells should be less Countin protein in these cells. Indeed we found less intracellular Countin throughout the first 8 hours after the onset of development of BS153 cells. In addition, there was significantly less Countin secreted during the first 8 hours of development. Countin, as an essential component of counting factor, regulates aggregate size after it is secreted. A high extracellular concentration of counting factor is interpreted by the cells of indicating a large number of cells within an aggregating stream. This causes the break-up of the aggregating stream into groups of the normal number of cells . Thus, the large size of the structures formed by developing BS153 cells can be accounted for by the lower extracellular Countin concentration and thus a lack of breaking up into smaller groups of cells. The heterogeneity of the size of the aggregates of developing BS153 might arise from the fact that extracellular countin levels eventually reach near normal levels, at about 10 to 12 hours after the onset of development. The reduced levels of Countin are not thought to account for the early aggregating phenotype of cells lacking IfkA since early aggregation was not found for countin null cells .
Why would the translation efficiency of countin mRNA be maintained and not reduced as is bulk mRNA during the secondary shift of polysomal RNA? Since the extracellular concentrations of Countin reflect cell density, perhaps it is important to maintain a constant level of Countin production and secretion once development has been initiated. Any drop in translation efficiency might alter the counting mechanism and result in aberrantly sized mounds.
In summary, the findings suggest that IfkA is an eIF2α kinase of Dictyostelium that normally phosphorylates eIF2α from 1 to 7 hours after the onset of development, or during the preaggregation phase. This results in an overall reduction in the initiation of protein synthesis during this time frame and a concomitant reduction in the number of ribosomes associated with most mRNAs (the secondary polysome shift). For some mRNAs, however, initiation of protein synthesis is enhanced or stabilized under the conditions of increased eIF2α phosphorylation. This includes countin mRNA. As a result, appropriate amounts of Countin are secreted and aggregate size is regulated properly. In the absence of IfkA, eIF2α phosphorylation does not increase, resulting in inefficient translation of countin and perhaps other mRNAs. Decreased Countin leads to larger than normal aggregates, slugs, and fruiting bodies. The larger size of many of these fruits results in their toppling over and thus in an inability to properly disperse the spores. Earlier than normal aggregation and cAMP pulsing also result from a lack of IfkA activity during preaggregation, and we postulate that this is due to the production and secretion of an as yet unidentified factor in early developing ifkA null cells.
Cell growth and development
The axenic strain, Ax4 , was used as the parental or "wild type" strain in all experiments. Cells were grown axenically in HL5 media or on SM plates with Klebsiella pneumoniae . Cells grown in the presence of bacteria were using for development as described [46, 47] unless otherwise stated.
Identification of the ifk genes and disruption of the ifkA gene
Initial studies identified a portion of the ifkB gene using degenerate PCR procedures based on the shared sequences of yeast, Drosophila, and mammalian GCN2 proteins. Sequences were eventually identified from the Dictyostelium cDNA project in Japan http://www.csm.biol.tsukuba.ac.jp/cDNAproject.html and the Dictyostelium genomic sequencing program (genome.imb-jena.de/dictyostelium (Genome Sequencing Center Jena); http://www.uni-koeln.de/dictyostelium (University of Cologne); dictygenome.bcm.tmc.edu (Baylor College of Medicine); and http://www.sanger.ac.uk/Projects/D_discoideum (Sanger Centre)). Two GCN2-like genes were identified and named ifkA (NCBI accession AAM43647) and ifkB (NCBI accession AAM33717). A plasmid was constructed that carried 825 bp from the pseudokinase domain ofifkA and 731 bp from the kinase domain, placed at either end of a 1.4 kb blasticidin resistance cassette . A 3.95 kb EcoR I/Hind III fragment, containing the ifkA sequences and the blasticidin gene, was released from the plasmid and it's ends were made blunt using the klenow fragment and dephosphorylated to prevent recircularization and reduce nonspecific integration of the fragment. This fragment was transformed into exponentially growing Ax4 cells using electroporation . Transformants were selected as described . Clonal isolates were obtained, and disruption of the ifkA gene was confirmed by PCR and Southern analysis. All disrupted strains had the same atypical morphology when developed under standard conditions. One strain was used for most of the experiments detailed herein, and is termed BS153.
For determination of the extent of eIF2α phosphorylation, Ax4 and BS153 cells were plated for development, and samples were taken at various times. Cells were lysed in a phosphate lysis buffer to prevent protein dephosphorylation (200 mM sodium phosphate, pH 7.5, 20 mM KCl, 5% sucrose, 0.5% NP-40, 1 mM PMSF). The lysate was cleared by spinning at 5000 × g for 10 minutes. The protein concentration of the cleared lysates was determined using the BCA protein assay kit (Pierce, Rockford IL). 40 mg of protein from each sample was separated using PAGE (NuPAGE, Invitrogen, Carlsbad CA). Phosphorylated eIF2α was detected with antiserum specific for the phosphorylated protein (Research Genetics, Carlsbad CA). The amino acid sequence of the phosphopeptide used to raise the antibodies (EGMILSEIS*RRRIRSI, * indicates phosphorylated serine) is highly conserved among the eIF2α proteins of all species, including that of Dictyostelium (I at position 8 is a L in Dictyostelium).
To examine the production of Countin or CMF protein inifkA null cells and Ax4 cells after starvation, mid-log phase cells growing axenically were collected, washed, and resuspended in the standard development buffer (PDF, 20 mM potassium phosphate, pH 6.1, 5 mM magnesium chloride, 0.5% streptomycin sulfate) at 5 × 106 cells/ml. The resuspended cells were shaken at 21°C for 1.5, 3, 4.5, 6, and 12 hours. The cells and buffer from 2 ml samples were collected. The cells were lysed with HMK-NP40 (50 mM Hepes, pH 7.5, 40 mM magnesium acetate, 20 mM potassium phosphate, 5% sucrose, 0.4% NP-40), clarified, and protein concentration determined as above. After removal of the cells, the collected buffer was used as conditioned medium. For detecting Countin at each time point, 20 μl of conditioned medium or 10 μg of protein from the cell lysates was fractionated by PAGE. Countin protein was detected by western analysis using anti- Countin antiserum  kindly provided by Dr. Richard Gomer. For detecting CMF, the 6 and 12 hour conditioned medium samples were concentrated prior to fractionation by PAGE. CMF antiserum was kindly provided by Dr. Richard Gomer.
mRNA and polysome analysis
RNA was isolated, processed, and used for RT-PCR as described [51, 52]. For polysome analysis, BS153 and Ax4 cells were harvest from bacterial growth plates before noticeable presence of Dictyostelium cell growth. After removal of bacteria, cells were plated for development as described above. After 2 and 4 hours of development, cells were collected into polysome buffer (HMK-NP40 plus cycloheximide to stabilize the polysomes). To prepare polysome samples from growing cells, cells were scraped from bacterial growth plates and suspended directly into polysome buffer. The cell lysates were spun at 8000 × g for 5 minutes to remove nuclei, cell debris, and bacteria (for growing cell samples). The concentration of RNA in the supernatants was determined spectrophotometrically. 0.5 mg of RNA were layered onto a 11 ml linear sucrose gradient (15–50% sucrose (w/v) in polysome buffer without NP-40), and the gradients were centrifuged in a Beckman SW40Ti rotor at 28000 rpm for 4 hours at 7°C. Gradient fractions were collected and the profile traces were generated by continuous monitoring at 254 nm. RNA from each fraction was isolated by extraction with phenol/chloroform followed by ethanol precipitation. RNA from an equivalent volume of each fraction was subjected to RT-PCR as just described, except the reactions were spiked with a small amount of radioactive dATP for quantitation. Quantitation was carried out using phosphoimaging analysis. For all RNA samples used in RT-PCR reactions, H7 specific oligonucleotides were used as a control; H7 mRNA is expressed at a constant level during growth and development  and thus monitors relative levels of RNA used in the reactions.
eukaryotic initiation factor 2 alpha
conditioned medium factor
This work was supported by NIH grant R01 GM59358. We thank Richard Gomer for generously supplying antiserum against Countin protein and Conditioned Medium Factor.
- Raper KB: The Dictyostelids. Princeton University Press, Princeton, NJ. 1984,Google Scholar
- Gomer RH, Yuen IS, Firtel RA: A secreted 80*10(3)Mr protein mediates sensing of cell density and the onset of development in Dictyostelium. Develop. 1991, 112: 269-Google Scholar
- Yuen IS, Jain R, Bishop JD, Lindsey DF, Derry WJ, Van Haastert PJM, Gomer R: A density-sensing factor regulates signal transduction in Dictyostelium. J Cell Biol. 1995, 129: 1251-1262. 10.1083/jcb.129.5.1251.View ArticlePubMedGoogle Scholar
- Yuen IS, Taphouse C, Halfant KA, Gomer RH: Regulation and processing of a secreted protein that mediates sensing of cell density in Dictyostelium. Develop. 1991, 113: 1375-Google Scholar
- Clarke M, Dominguez N, Yuen IS, Gomer RH: Growing and starving Dictyostelium cells produce distinct density-sensing factors. Dev Biol. 1992, 152: 403-10.1016/0012-1606(92)90147-9.View ArticlePubMedGoogle Scholar
- Clarke M, Kayman SC, Riley K: Density-dependent induction of discoidin-I synthesis in exponentially growing cells of Dictyostelium discoideum. Differ. 1987, 34: 79-87.View ArticleGoogle Scholar
- Clarke M, Yang J, Kayman S: Analysis of the prestarvation response in growing cells of Dictyostelium discoideum. Dev Genet. 1988, 9: 315-326. 10.1002/dvg.1020090413.View ArticlePubMedGoogle Scholar
- Rathi A, Clarke M: Expression of early developmental genes in Dictyostelium discoideum is initiated during exponential growth by an autocrine-dependent mechanism. Mech Develop. 1992, 36: 173-182. 10.1016/0925-4773(92)90068-U.View ArticleGoogle Scholar
- Rathi A, Kayman SC, Clarke M: Induction of gene expression in Dictyostelium by prestarvation factor, a factor secreted by growing cells. Dev Genet. 1991, 12: 82-87. 10.1002/dvg.1020120115.View ArticlePubMedGoogle Scholar
- Jain R, Yuen IS, Taphouse CR, Gomer RH: A density-sensing factor controls develpment in Dictyostelium. Genes Develop. 1992, 6: 390-400.View ArticlePubMedGoogle Scholar
- Jain R, Gomer RH: A developmentally regulated cell surface receptor for a density-sensing factor in Dictyostelium. J Biol Chem. 1994, 269: 9128-9136.PubMedGoogle Scholar
- Brock DA, Buczynski G, Spann TP, Wood SA, Cardelli J, Gomer RH: A Dictyostelium mutant with defective aggregate size determination. Develop. 1996, 122: 2569-2578.Google Scholar
- Brock DA, Gomer RH: A cell-counting factor regulating structure size in Dictyostelium. Genes Develop. 1999, 13: 1960-1969.PubMed CentralView ArticlePubMedGoogle Scholar
- Okuwa T, t Katayama, Takano A, Kodaira K-i, Yasukawa H: Two cell-counting factors regulate the aggregate size of the cellular slime mold dictyostelium discoideum. Develop Growth Differ. 2001, 43: 735-744. 10.1046/j.1440-169X.2001.00615.x.View ArticleGoogle Scholar
- Kon T, Adachi H, Sutoh K: amiB, a novel gene required for the growth/differentiation transition in Dictyostelium. Genes to Cells. 2000, 5: 432-55. 10.1046/j.1365-2443.2000.00305.x.View ArticleGoogle Scholar
- Souza G, Lu S, Kuspa A: YakA, a protein kinase required for the transition from growth to development in Dictyostelium. Develop. 1998, 25: 2291-2302.Google Scholar
- Souza GM, da Silva AM, Kuspa A: Starvation promotes Dictyosteliumdevelopment by relieving PufA inhibition of PKA translation through the YakA kinase pathway. Develop. 1999, 126: 3263-3274.Google Scholar
- Zeng C, Anjard C, Riemann K, Konzok A, Nellen W: gdt1, a new signal transduction component for negative regulation of the growth-differentiation transition in Dictyostelium discoideum. Mol Biol Cell. 2000, 11: 1631-164.PubMed CentralView ArticlePubMedGoogle Scholar
- Margolskee JP, Froshauer S, Skrinska R, Lodish H: The effects of cell density and starvation on early developmental events in Dictyostelium discoideum. Dev Biol. 1980, 74: 409-421. 10.1016/0012-1606(80)90442-X.View ArticlePubMedGoogle Scholar
- Marin FT: Regulation of development in Dictyostelium discoideum. Dev Biol. 1976, 48: 110-117. 10.1016/0012-1606(76)90050-6.View ArticlePubMedGoogle Scholar
- Marin FT: Regulation of development in Dictyostelium discoideum. Dev Biol. 1977, 60: 389-395. 10.1016/0012-1606(77)90136-1.View ArticlePubMedGoogle Scholar
- Alton TH, Lodish HF: Translational control of protein synthesis during the early stages of differentiation of the slime mold Dictyostelium discoideum. Cell. 1977, 12: 301-310. 10.1016/0092-8674(77)90208-2.View ArticlePubMedGoogle Scholar
- Cardelli JA, Dimond RL: Regulation of protein synthesis in Dictyostelium discoideum : the effects of starvation and anoxia on initiation. Biochem. 1981, 20: 7391-7398. 10.1021/bi00529a011.View ArticleGoogle Scholar
- Harding H, Novoa I, Zhang Y, EZeng H, Wek R, Schapira AM, Ron D: Regulated translation initiation controls stress-induced gene expression in mammalian cells. Mol Cell. 2000, 6: 1099-1108. 10.1016/S1097-2765(00)00108-8.View ArticlePubMedGoogle Scholar
- Hinnenbusch A: Translational control of GCN4: gene-specific regulation by phosphorylation of eIF2. Translational Control. Edited by: Hershey J, Mathews M, Sonenberg N. 1996, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 199-244.Google Scholar
- Dever T, Feng T, Wek R, Cigan A, Donahue T, Hinnebusch A: Phosphorylation of initiation factor 2α by protein kinase GCN2 mediates gene-specific translational control of GCN4 in yeast. Cell. 1992, 68: 585-596. 10.1016/0092-8674(92)90193-G.View ArticlePubMedGoogle Scholar
- Berlanga JJ, Santoyo J, De Haro C: Characterization of a mammalian homolog of the GCN2 eukaryotic initiation factor 2alpha kinase. Eur J Biochem. 1999, 265: 754-62. 10.1046/j.1432-1327.1999.00780.x.View ArticlePubMedGoogle Scholar
- Olsen DS, Jordan B, Chen D, Wek RC, Cavener DR: Isolation of the gene encoding the Drosophila melanogaster homolog of the Saccharomyces cerevisiae GCN2 eIF-2alpha kinase. Genetics. 1998, 149: 1495-509.PubMed CentralPubMedGoogle Scholar
- Santoyo J, Alcalde J, Mendez R, Pulido D, de Haro C: Cloning and characterization of a cDNA encoding a protein synthesis initiation factor-2alpha (eIF-2alpha) kinase from Drosophila melanogaster. Homology To yeast GCN2 protein kinase. J Biol Chem. 1997, 272: 12544-50. 10.1074/jbc.272.19.12544.View ArticlePubMedGoogle Scholar
- Sattlegger E, Hinnebusch AG, Barthelmess IB: cpc-3, the Neurospora crassa homologue of yeast GCN2, encodes a polypeptide with juxtaposed eIF2alpha kinase and histidyl-tRNA synthetase-related domains required for general amino acid control. J Biol Chem. 1998, 273: 20404-16. 10.1074/jbc.273.32.20404.View ArticlePubMedGoogle Scholar
- Dong J, Qiu H, Garcia-Barrio M, Anderson J, Hinnebusch AG: Uncharged tRNA activates GCN2 by displacing the protein kinase moiety from a bipartite tRNA-binding domain. Mol Cell. 2000, 6: 269-279. 10.1016/S1097-2765(00)00028-9.View ArticlePubMedGoogle Scholar
- Wek SA, Zhu S, Wek RC: The histidyl-tRNA synthetase-related sequence in eIF-2 alpha protein kinase GCN2 interacts with tRNA and is required for activation in response to starvation for different amino acids. Mol Cell Biol. 1995, 15: 4497-4506.PubMed CentralView ArticlePubMedGoogle Scholar
- Qiu H, Garcia-Barrio MT, Hinnebusch AG: Dimerization by translation initiation factor 2 kinase GCN2 is mediated by interactions in the C-terminal ribosome-binding region and the protein kinase domain. Mol Cell Biol. 1998, 18: 2697-711.PubMed CentralView ArticlePubMedGoogle Scholar
- Ramirez M, Wek RC, Hinnebusch AG: Ribosome association of GCN2 protein kinase, a translational activator of the GCN4 gene of Saccharomyces cerevisiae. Molecular and Cell Biology. 1991, 11: 3027-36.View ArticleGoogle Scholar
- Wek RC, Ramirez M, Jackson BM, Hinnebusch AG: Identification of positive-acting domains in GCN2 protein kinase required for translational activation of GCN4 expression. Mol Cell Biol. 1990, 10: 2820-31.PubMed CentralView ArticlePubMedGoogle Scholar
- Price N, Proud C: The guanine nucleotide-exchange factor, eIF-2B. Biochemie. 1994, 76: 748-760. 10.1016/0300-9084(94)90079-5.View ArticleGoogle Scholar
- Ramirez M, Wek RC, Vazquez De Aldana CR, Jackson BM, Freeman B, Hinnebusch AG: Mutations activating the yeast eIF-2α kinase GCN2: isolation of alleles altering the domain related to histidyl-tRNA synthetases. Molecular and Cell Biology. 1992, 12: 5801-5815.View ArticleGoogle Scholar
- Devreotes PN: G protein-linked signalling pathways control the developmental program of Dictyostelium. Neuron. 1994, 12: 235-241. 10.1016/0896-6273(94)90267-4.View ArticlePubMedGoogle Scholar
- Firtel RA: Integration of signaling information in controlling cell-fate decisions in Dictyostelium. Genes Develop. 1995, 9: 1427-1444.View ArticlePubMedGoogle Scholar
- Mangiarotti G, Bulfone S, Gioda R, Morandini P, Ceccarelli A, Hames BD: Analysis of specific mRNA destabilization during Dictyostelium development. Develop. 1989, 106: 473-481.Google Scholar
- Wek R, Jackson B, Hinnebusch A: Juxtaposition of domains homologous to protein kinases and histidyl-tRNA synthetases in GCN2 protein suggests a mechanism for coupling GCN4 expression to amino acid availablity. Proc Natl Acad Sci, USA. 1989, 86: 4579-4583.PubMed CentralView ArticlePubMedGoogle Scholar
- Vauti F, Morandini P, Blusch J, Sachse A, Nellen W: Regulation of the discoidin I gene in Dictyostelium discoideum: identification of individual promoter elements mediating induction of transcription and repression by cyclic AMP. Molecular and Cell Biology. 1990, 10: 4080-4088.View ArticleGoogle Scholar
- Tang L, Gao T, McCollum C, Jang W, Vicker M, Ammann R, Gomer RH: A cell number-counting factor regulates the cytoskeleton and cell motility in Dictyostelium. Proc Natl Acad Sci, USA. 2000, 99: 1371-1376. 10.1073/pnas.022516099.View ArticleGoogle Scholar
- Knecht DA, Cohen SA, Loomis WF, Lodish HF: Developmental regulation of Dictyostelium discoideum actin gene fusions carried on low copy and high copy transformation vectors. Molecular and Cell Biology. 1986, 6: 3973-3983.View ArticleGoogle Scholar
- Singleton CK, Delude RL, McPherson CE: Characterization of genes which are deactivated upon the onset of development in Dictyostelium discoideum. Dev Biol. 1987, 119: 433-441. 10.1016/0012-1606(87)90047-9.View ArticlePubMedGoogle Scholar
- Singleton CK: Nucleotide sequence of V1, a ribosomal protein gene from Dictyostelium discoideum. Nucleic Acids Res. 1989, 17: 1989-View ArticleGoogle Scholar
- Singleton CK, Zinda MJ, Mykytka B, Yang P: The histidine kinase dhkC regulates the choice between migrating slugs and terminal differentiation in Dictyostelium discoideum. Dev Biol. 1998, 203: 345-357. 10.1006/dbio.1998.9049.View ArticlePubMedGoogle Scholar
- Sutoh KA: A transformation vector for Dictyostelium discoideum with a new selectable marker Bsr. Plasmid. 1993, 30: 150-154. 10.1006/plas.1993.1042.View ArticlePubMedGoogle Scholar
- Howard PK, Ahern KG, Firtel RA: Establishment of a transient expression system for Dictyostelium discoideum. Nucleic Acids Res. 1988, 16: 2613-2623.PubMed CentralView ArticlePubMedGoogle Scholar
- Zinda MJ, Singleton CK: The hybrid histidine kinase dhkB regulates spore germination in Dictyostelium discoideum. Dev Biol. 1998, 196: 171-183. 10.1006/dbio.1998.8854.View ArticlePubMedGoogle Scholar
- Pekovich SR, Martin PR, Singleton CK: Thiamine deficiency decreases steady-state mRNA levels for transketolase and pyruvate dehydrogenase but not for α-keotglutarate dehydrogenase in three human cell types. J Nutr. 1998, 128: 683-687.PubMedGoogle Scholar
- Zinda MJ: Two component sensor kinases of Dictyostelium discoideum. Ph D. Thesis; Vanderbilt University. 1997Google Scholar
- Singleton CK, Manning SS, Feng Y: Effect of protein synthesis inhibition on gene expression during early development of Dictyostelium discoideum. Molecular and Cell Biology. 1988, 8: 10-16.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all media for any purpose, provided this notice is preserved along with the article's original URL.