Metabolic control of acclimation to nutrient deprivation dependent on polyphosphate synthesis

Acclimation to nutrient limitation requires polyP synthesis to control ATP homeostasis and sustain respiration and photosynthesis.


INTRODUCTION
Polyphosphate (polyP), an ancient polymer conserved in organisms from bacteria to humans, is composed of phosphate groups linked by the highly energetic phosphoanhydride bond found in adenosine 5′-triphosphate (ATP). PolyP is abundant in cells, with concentrations that range from micromolar in mammals to millimolar in unicellular prokaryotes and eukaryotes (1). This polymer mainly accumulates in acidocalcisomes, acidic vacuoles that store calcium and other divalent cations in both eukaryotes and prokaryotes (2), although polyP can also localize to the cell wall, mitochondria, nucleus, and endoplasmic reticulum (3). In bacteria and amoebae such as Dictyostelium discoideum, polyP synthesis is mediated by polyP kinases (4)(5)(6), whereas in most unicellular eukaryotes, the synthesis is performed by the vacuolar transporter chaperone (VTC) complex (7). In both cases, the terminal phosphate of ATP is transferred to the growing polyP chain, resulting in the accumulation of energy in orthophosphate bonds and the release of adenosine 5′-diphosphate (ADP). In higher eukaryotes, although polyP and acidocalcisomes are present (2,8), the mechanism for polyP synthesis is still unknown.
In yeast, the VTC complex contains three polypeptide components, VTC1, VTC4, and VTC2/VTC3 (9). VTC4, the catalytic subunit of the complex, has three functional domains. The N-terminal region contains the SPX domain, which binds inositol pyrophosphate and activates polyP polymerase activity, the central region includes the catalytic polyP polymerase domain, and the C-terminal region contains the VTC domain, which consists of three  helices that anchor the protein to the membrane (7,(10)(11)(12).
Intracellular polyP aggregates were found more than a hundred years ago, and over the past three decades, many studies have shown that different organisms require polyP to cope with various stress conditions including high temperature, oxidative stress, osmotic shock, survival during stationary phase, heavy metal accumulation, and nutrient deprivation, among others (1,5,(13)(14)(15)(16)(17)(18). Moreover, polyP synthesis affects critical cellular processes, including DNA replica-tion, transcription, translation, quorum sensing, and protein folding, and is essential for normal infectivity of many parasites (1,5,(18)(19)(20)(21)(22)(23). However, how this simple polymer can be involved in so many different processes is an unanswered and intriguing question.
Researchers are just beginning to unravel mechanisms by which polyP regulates a broad spectrum of biological processes. A recently found posttranslational modification, polyphosphorylation, involves the covalent attachment of polyP to lysine residues. This modification is present on proteins involved in ribosome biogenesis and DNA unfolding (topoisomerases) (24,25). Another feature of polyP that could explain its impact on numerous cellular processes is its ability to serve as a stabilizing scaffold for protein-folding intermediates and for the protection and maintenance of protein activities by preventing their aggregation, as reported for bacterial cells exposed to high temperature or oxidative stress (13,26). Similarly, low levels of polyP have been associated with protein aggregation (amyloid plaques) in neurons and neurodegenerative diseases (13). PolyP can also interact with proteases and promote the degradation of specific proteins. For example, the Lon protease interacts with polyP, activating degradation of ribosomal polypeptides during amino acid starvation of Escherichia coli (21). Last, polyP can serve as a chelator of divalent cations (Ca 2+ , Fe 2+ , Cd 2+ , Mn 2+ , and Cu 2+ , among others) and control intracellular cation availability (15,27). Despite this growing number of discoveries concerning the impact of polyP on cellular processes, many with seemingly unrelated functions, we still know little about how this polymer and its dynamic synthesis and degradation help cells acclimate to stress.
To elucidate the link between polyP and stress acclimation, we focused on nutrient deprivation. Organisms starved for sulfur (S) trigger a suite of responses that include induction of genes encoding high-affinity transporters and assimilatory enzymes responsible for efficient scavenging and utilization of the limiting nutrients, the reorganization of cellular metabolism and energetics (28,29), and the accumulation of polyP (16,30). In the green alga Chlamydomonas reinhardtii (Chlamydomonas throughout), the VTC1 protein, which only contains the VTC domain and seems to have a structural function in the VTC complex, is essential for polyP accumulation. Strains lacking this protein were unable to normally acclimate to S deprivation (16). However, the role of polyP in the acclimation of cells to nutrient deprivation is still not clear.
Here, we show that, similar to the phenotype of the vtc1 mutant (16), the loss of VTC4 in Chlamydomonas leads to impaired acclimation to sulfur deprivation, but we also elucidate the connection between polyP and the acclimation process, demonstrating that polyP-deficient strains initiate the acclimation process but are unable to sustain it. Chlamydomonas cells appear to need polyP synthesis to maintain the ATP homeostasis and levels of adenylates in cells, which is critical to avoid inhibition of electron flow in both mitochondria and chloroplasts. Moreover, we demonstrate that active electron flow, independent of the cellular ATP level, is necessary to induce the responses associated with S and N deprivation. We suggest that the links between the synthesis of polyP, maintenance of intracellular ATP levels, sustained chloroplast and mitochondrial electron flow, and the specific responses of the cells to stress represent a set of interacting activities that is conserved in many organisms and that becomes critical under conditions in which the cells are able to synthesize high levels of ATP but are unable to grow. This work integrates key roles of polyP synthesis in regulating the energetics of the cell and provides insights into how polyP can affect numerous cellular processes, including the ways in which the cells cope with changing environmental conditions.

VTC4 is essential for "sustaining" the acclimation to S deprivation
The VTC4 protein is highly conserved, and as in other organisms, the Chlamydomonas protein contains three domains (SPX, polyP polymerase, and VTC) that were previously described in yeast and trypanosomatids (7,10,31) (Fig. 1A and fig. S1). To elucidate the function of polyP in acclimation to S deprivation, we characterized two mutants disrupted for the gene encoding VTC4, which has not been previously characterized in algae. These mutants were obtained from a genome-wide mutant library generated by insertional mutagenesis (32). In the vtc4-1 mutant, the insertion was in the middle of the polyP polymerase domain, without causing a deletion or rearrangement at the insertion site ( Fig. 1A and fig. S2). In the second allelic mutant (vtc4-2), the insertion occurred just downstream of the polyP polymerase domain and caused a deletion of 4120 base pairs (bp; fig. S3, A and B).
To assess the impact of the loss of VTC4 on S deprivation responses (SdR), we quantified the activity of arylsulfatase (ARS), a sulfur deprivation-induced periplasmic enzyme involved in scavenging sulfate from organic molecules in the environment. This enzyme has been extensively used as a reporter of the SdR (33). The two allelic vtc4 mutants showed very low ARS activity following transfer to medium depleted for S [tris acetate phosphate (TAP)-S], while the ectopic expression of a wild-type (WT) VTC4 gene in the vtc4-1 mutant (C2-and C8-complemented lines) rescued the ARS phenotype ( Fig. 1B and fig. S3C). Moreover, the vtc4-1 mutant, but neither WT cells nor the complemented lines, bleached after 10 days of S deprivation (Fig. 1B), which confirmed the essential role of VTC4 in the acclimation process.
In a previous study, it was shown that the vtc1 mutant exhibited low ARS mRNA accumulation after 6 hours of S deprivation (16). To acquire a better understanding of how the lack of the VTC complex affects the acclimation process, we studied the kinetics of ARS activity and mRNA accumulation in S-depleted medium. Unexpectedly, ARS activity and mRNA accumulation (Fig. 1, C and D) increased similarly in WT and vtc4-1 for the first ~3 hours of starvation. However, the ARS mRNA level in the mutant dropped markedly by 6 hours of S deprivation and reached preinduction levels after 24 hours (Fig. 1D). The ARS enzyme, on the other hand, is very stable, and its activity was identical after 6 and 24 hours of S deprivation (Fig. 1C). These results clearly indicate that VTC4 is not necessary to trigger acclimation but is required to sustain the SdR (at least with respect to ARS expression). The initial induction of ARS followed by a "repressor effect" (Fig. 1, C and D) in the vtc4-1 mutant was not exclusive for ARS mRNA but was also observed for other transcripts induced during S deprivation. The transcripts for two high-affinity sulfate transporters [sulfate transporter 2 (SULTR2) and SAC1-like transporter 1 (SLT1)] and a cell wall protein with almost no S-containing amino acids [extracellular protein ~76 kDa (ECP76)] (28,29) were also quantified ( fig. S4, A to C). The genes encoding these proteins belong to the two different tiers of regulation identified for SdR. Genes in the first tier encode mainly transporters (for S scavenging), which are induced immediately after deprivation. The second-tier genes are induced later in the acclimation process and are associated with sustained deprivation conditions (ARSs and ECPs, among others) (34). All of these transcripts were similarly induced after 1 to 3 hours of S limitation in WT, vtc4-1, and the C2and C8-complemented strains, while, at later times, only the mutant exhibited a reduction of more than 10-, 5-, and 56-fold for SULTR2, SLT1, and ECP76, respectively ( fig. S4, A to C). Our results indicate that VTC4 has a broad impact on sustaining the transcriptional activity of the SdR genes, but the final transcript levels may vary depending on which tier of control they are associated with and the stability of their transcripts under deprivation conditions.
To test whether the observed results reflect a general effect on transcription, we quantified the ACT1 (actin) and GPX5 (glutathione peroxidase) transcripts. The levels of these transcripts were not affected in the S-deprived vtc4-1 mutant ( fig. S4, D and E). These results indicate that the lack of or inability to synthesize polyP does not affect transcription in general and appears to be specific to mRNAs associated with the SdR.
We also observed that the inability of the vtc4-1 mutant to acclimate to S deprivation resulted in a slight effect on growth after S replenishment. The S-depleted mutant cells exhibited a ~50% reduction in growth over a period of 24 hours after being transferred to nutrient-replete medium. By 48 hours following the transfer, as the cells approached stationary phase, the cells showed approximately the same level of growth as the WT and complemented strain ( fig. S5). These results indicate that after being starved for S, the mutant initially may have a slower metabolism and/or reduced capacity to assimilate S compared to the WT and the complemented strain (the WT and complemented strain would have higher levels of transcripts for the high-affinity S transporters and potentially other acclimation-associated proteins following growth in S-depleted medium).

VTC4-dependent polyP synthesis in both S-replete and S-depleted media
The accumulation of polyP under nutrient stress conditions has been previously reported in different organisms, including Chlamydomonas, although we know little about the kinetics of accumulation. To examine the correlation between the development of the "ars" phenotype and polyP accumulation, we quantified polyP in cells maintained in nutrient-replete conditions and those exposed to 1, 3, 6, and 24 hours of S deprivation. Chlamydomonas WT cells accumulated low levels of polyP when grown in mixotrophic, nutrient-replete conditions (TAP). In contrast, no polyP was detected in either the vtc4-1 or vtc4-2 mutants; the phenotype was rescued in the complemented lines ( Fig. 2A and fig. S6). Quantification of polyP in WT cells and the complemented vtc4-1 strains demonstrated an ~8-fold increase after 24 hours of S deprivation, with no polyP detected in the vtc4-1 mutant under the same conditions ( Fig. 2B and fig. S6, B and C). PolyP accumulation in WT and the complemented lines was not observed immediately but only after ~3 hours of S deprivation, which correlated with the time at which we observed transcrip-tional repression of the SdR. These results clearly establish that the initial pool of polyP is not required for the induction of the SdR. This finding raises the possibility that an inability of the mutant to synthesize polyP following 3 hours of S deprivation might be the cause of its "ars − " phenotype.

Lack of polyP synthesis leads to high ATP accumulation and reduced electron flow in chloroplasts and mitochondria
PolyP synthesis involves ATP hydrolysis and the regeneration of ADP (7). Therefore, we hypothesized that this polymer could serve as an energy buffer to reduce the concentration of excess ATP during S deprivation when growth, the primary sink for energy, is arrested. We measured ATP levels in WT, vtc4-1, C2, and C8 strains grown in both replete medium and in medium lacking S. While a loss of VTC4 did not affect ATP levels in the cells during the first hour of S deprivation, after 3 hours, the cellular ATP concentration of vtc4-1 rose, with a ~2.5-fold increase after 24 hours (Fig. 2C). This accumulation in the mutant paralleled a rise in the ATP/ADP ratio (Fig. 2D). In contrast, the intracellular ATP concentrations for the WT and complemented strains were stable at all time points after exposing the cells to S deprivation. The kinetics of ATP accumulation in the mutant was similar to that of polyP accumulation in WT cells (compare Fig. 2, B and C), suggesting a connection between polyP synthesis and ATP homeostasis and a previously unrealized potential link between cellular energetics and maintenance of the SdR.
A high intracellular ATP/ADP ratio is known to inhibit ATP synthesis, which is coupled to electron transport and the dissipation of a proton gradient across the thylakoid and respiratory membranes (35,36). To assess the impact of high ATP on respiration in the mutant, we quantify mitochondrial O 2 uptake (respiration) under nutrient-replete and S-deprivation conditions. The WT and vtc4-1 strains exhibited almost the same respiration rate under nutrientreplete conditions, but vtc4-1 showed a ~50% reduction relative to WT when the cells were deprived of S for 6 hours (Fig. 2E). The mutant cells were similarly inhibited for photosynthetic electron transport (PET; Fig. 2F). Dark-acclimated WT and vtc4-1 mutant cells showed a similar maximum photosystem II (PSII) yield (Fv/Fm) after 6 hours of S deprivation ( fig. S7A), which indicates that PSII is not damaged in the mutant. However, after illumination, the PSII yield in the mutant displayed a more pronounced drop than in WT cells (fig. S7B); this reduction is congruent with a restriction in photosynthetic electron transport (PET) caused by a high intracellular ATP concentration that is generated in strains with no polyP polymerase activity. These data also suggest the importance of polyP synthesis in maintaining ATP homeostasis, which, in turn, is required to sustain electron flow in both mitochondria and chloroplasts during acclimation to S deprivation.

Reduction of ATP levels allows rescue of the vtc4-1 mutant phenotype
To determine whether the phenotypes of vtc4-1 (abnormal acclimation and reduced respiration/photosynthesis) were a consequence of the high ATP levels or impaired polyP synthesis, we deprived mutant cells of S under conditions in which, in addition to not making polyP, the cells were unable to accumulate ATP. This situation was created by placing cells in medium depleted of both S and phosphorus (P; TA-S medium); the S-deprived mutant in the absence of Pi still did not accumulate polyP but was now also unable to accumulate ATP (Fig. 3, A and B). Under these conditions, the vtc4-1 mutant exhibited rates of chloroplast and mitochondrial electron transport that were comparable to those of WT cells (Fig. 4A). Furthermore, the TA-S-grown mutant was also able to induce ARS activity to a level similar to that of WT cells (Fig. 4, B and C). These results not only prove the role of VTC4 in controlling ATP homeostasis when the cells are deprived of S but also suggest that the ele-vated ATP in the mutant was responsible for the reduced rates of respiratory electron flow and PET and that the accumulation of polyP is not directly responsible for allowing cells to acclimate to S deprivation. In sum, the high level of intracellular ATP that accumulates in the vtc4-1 mutant likely causes a slowing of respiratory electron flow and PET and suppression of the SdR.

Electron flow in mitochondria or chloroplasts independent of ATP synthesis triggers expression of SdR genes
Inhibition of SdR in the vtc4-1 mutant might be a consequence of diminished mitochondrial or chloroplast electron flow. During mixotrophic growth (TAP medium in the light), Chlamydomonas generates energy through both respiration and photosynthesis. To study the impact of electron flow in the organelles on the SdR, we deprived WT cells of S in the dark and added the respiration inhibitor Myxothiazol (Myx). Myx is a mitochondrial complex III inhibitor that blocks most mitochondrial electron flow and ATP synthesis. In the absence of mitochondrial and chloroplastic electron flow (Myx in the dark), there was no ARS induction, while under the same conditions without the addition of Myx, the cells exhibited normal acclimation (Fig. 5,  A and D). These results suggest that eliminating electron flow in both mitochondria and chloroplasts prevents SdR induction. However, the ATP levels in dark maintained cells treated with Myx dropped by more than 50% (Fig. 5A), which could be the cause of the repressed SdR. To distinguish between the effect of reduced electron flow and the lower  concentration of ATP on ARS expression, we used the uncoupler carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP). This protonophore, which has been previously used on Chlamydom onas (37,38), dissipates the proton gradient across the membranes, which prevents ATP synthesis in both mitochondria and chloroplasts without reducing the rate of electron transport. Myx in the dark and FCCP in the light both similarly lowered the ATP levels in WT cells (to ~50% ; Fig. 5A). These two treatments had an opposite effect on ARS mRNA accumulation. In the presence of FCCP, which promotes electron transport, ARS was strongly induced, similar to WT control conditions. However, under the condition in which all electron transport was strongly reduced (Myx in the dark), ARS expression was almost completely repressed (Fig. 5A). These results indicate that the elevated level of ATP observed in the S-deprived vtc4-1 mutant was not directly responsible for suppression of SdR, but SdR suppression was the consequence of the impact of high levels of ATP on electron transport. Under high (vtc4-1 mutant in TAP-S) and low (vtc4-1 mutant in TA-S and FCCP/Myx treatments) ATP conditions, the transcriptional activation of the SdR depended exclusively on active electron flow.
To further test whether the inability of the vtc4-1 mutant to sustain SdR was the consequence of reduced electron transport, we treated WT and the vtc4-1 mutant that had already experienced 6 hours of S deprivation with FCCP. After 6 hours in TAP-S, both respiration and photosynthesis were partially blocked in the mutant, and the level of ARS mRNA was declining (Figs. 1D and 4C). For WT cells, FCCP treatment barely affected ARS mRNA accumulation. However, FCCP-treated vtc4-1 cells exhibited a notable increase (>20-fold) in ARS expression (Fig. 5B), which paralleled an elevated rate of respiration, indicated by a 50% increase in O 2 uptake (Fig. 5C). Together, these data confirmed that the transcription of the SdR genes requires respiratory electron flow or PET independent of the ATP concentration, although other signals elicited by nutrient deprivation and probably sensed by plasma membrane receptors would also be required (28,39). Therefore, polyP synthesis has a vital role in maintaining active electron flow, which is necessary for the cells to acclimate to changing nutrient levels.
Last, we tested whether electron flow in both chloroplasts and mitochondria is needed to induce acclimation. We performed experiments in which we blocked electron transport in WT cells in only one organelle (Myx in the light to eliminate respiratory electron flow or only dark to eliminate PET). Under both of these conditions, we only observed a slight delay in ARS mRNA accumulation (Fig. 5D). Hence, electron flow in either organelle can generate the signal that induces and sustains ARS expression. Similar results were obtained when we used a complex III-deficient mutant instead of Myx ( fig. S8), which suggests that the Myx effect is specifically due to its impact on mitochondrial electron transport. The delay in ARS induction observed when electron flow in only one of the organelles was blocked and the almost total loss of induction when both photosynthetic and respiratory electron flow was inhibited were also observed for SULTR2, SLT1, and ECP76 ( fig. S9), demonstrating that the SdR requires the integration of signals coming from either PET or respiratory electron flow.

PolyP synthesis also affects cellular energetics and acclimation during nitrogen (N) deprivation
One of the highly induced genes in Chlamydomonas cells experiencing N deprivation is LAO1, which encodes l-amino oxidase (40). LAO1 is a periplasmic protein, similar to ARS, that scavenges amine groups from extracellular organic molecules (41). Aksoy et al. (16) previously reported that a Chlamydomonas vtc1 mutant showed reduced induction of the LAO1 protein when cells were deprived of N. To study whether polyP synthesis is also required to maintain ATP homeostasis and sustain LAO1 gene expression, we transferred the vtc4 mutation to a genetic background able to assimilate all inorganic N (the original parental strain in this study, CMJ030, uses ammonium, but not nitrate, as a sole N source) by crossing the vtc4-1 mutant and the WT strain 21gr (CC-1690), which can grow with either nitrate or ammonium as a sole N source; this is a more appropriate genetic background when examining N deprivation. We verified that the resulting mutant (vtc4-1*) in the new parental background showed the ars − phenotype when deprived of S and that we  As previously shown for S deprivation, polyP accumulated in the WT (21gr genetic background) and the complemented line, but not in the mutant lacking VTC4 (Fig. 6A). As expected, this inability to synthesize polyP impaired the capacity of the cells to buffer ATP levels, with the consequent increase in the intracellular ATP concentration (>2.5-fold higher) in vtc4-1* (Fig. 6B). This lack of polyP synthesis also led to substantial repression in LAO1 mRNA (~40-fold) and protein accumulation after 24 hours of N deprivation (Fig. 6,  C and D). Last, similar to the results for S deprivation, treatment with FCCP and Myx in the dark showed that LAO1 expression appears to be linked to electron transport independently of the ATP concentration (Fig. 6E). These results demonstrate a role for polyP synthesis in maintaining ATP homeostasis and sustaining the acclimation of cells to N deprivation.

DISCUSSION
Acclimation responses are energy consuming, and before being initiated, the cells must integrate not only external cues, including nutrient availability through plasma membrane sensors (34,39), but also they must respond to other signals that report the cell's energetic status and potential for rapid acclimation. Thus, the cells would adjust their acclimation responses to energy availability. In this work, we demonstrate that the responses of Chlamydomonas to S and N deprivation are linked to polyP synthesis, ATP levels, and mitochondrial and chloroplast electron flow (Fig. 7). These connections provide a mechanistic/holistic view of the pleiotropic nature of the varied phenotypes associated with the inability of cells to synthesize polyP. Our studies indicate that the key to the SdR is not the polyP molecule itself but rather the synthesis of polyP in regulating ATP levels and maintaining electron transport activities that are critical for acclimation. Additional studies will be required to elucidate the Student's t test was performed. **P values <0.01 and ***P < 0.001. "ns" means nonstatistically significant difference.
impacts of these links on the cell's capacity to cope with other stressful situations and to extrapolate the results presented here to other organisms. For example, our data raise the possibility that polyPdependent virulence and survival of microorganisms (5, 13, 42) could be a consequence of the stabilization of ATP levels, which would sustain respiratory electron flow in the microbe during the first stages of infection. PolyP synthesis may also be essential to maintain active electron flow in the host as it acclimates to the stress associated with invasion. The virulence of trypanosomes, which affects millions of people in the world, is substantially affected by disruption of polyP synthesis (42). The VTC complex, the enzymatic machine in these protozoa that uses ATP to synthesize polyP, has been considered a potential target for the treatment of trypanosomacaused diseases. Therefore, a detailed understanding of the links between polyP synthesis and cellular metabolism could allow for the development of new treatments to fend off parasites (23).
A major finding of our work is that polyP synthesis is not essential to induce acclimation of Chlamydomonas to S deprivation but that it is vital for sustaining this acclimation. The vtc4-1 mutant, which is unable to synthesize polyP, exhibited acclimation only during the first 3 hours following the imposition of S deprivation, a period in which polyP was barely synthesized by WT cells. The role of polyP synthesis in sustaining the SdR appears to be the consequence of the capacity of the VTC complex to use ATP (storing the energy in the molecule as polyP), preventing high levels of its accumulation. The characterization of the vtc4-1 mutant, which is unable to synthesize polyP, has allowed us to identify links among polyP accumulation, the regulation of intracellular ATP levels, and respiratory electron flow and PET. During S deprivation, cells are unable to grow, and the cellular energy demand drops markedly. However, as long as Pi and ADP are abundant, cells continue to synthesize excess ATP that is used to accumulate polyP. PolyP accumulation not only serves in the storage of energy and Pi but also enables the regeneration of ADP, which allows for continued respiratory electron flow and PET. Here, we clearly demonstrate a role  of these electron transport activities in triggering and sustaining SdR; these findings suggest that the cells use electron flow as a proxy for energy availability. The inability to recycle ATP would lead to elevated ATP levels, which, in turn, retards both respiratory electron flow and PET and signals a compromised metabolic status. The growth of Chlamydomonas cells declines upon nutrient deprivation, leading to a reduced demand for energy, and yet, during nutrient deprivation, the transcripts and proteins associated with mitochondrial complexes are stable during mixotrophic growth (28,40), while the transcripts and proteins associated with photosynthetic complexes are stable during photoautotrophic growth (43)(44)(45). These observations are in accord with our conclusion that effective acclimation to S deprivation (and likely other deprivation responses) relies on active electron transport in the energy-related organelles. Moreover, polyP also appears to play a similar role in N deprivation; its synthesis also is required for normal acclimation to this deficiency. Together, our data indicate that the synthesis of the polyP polymer is crucial for the cells to cope with the energetic dynamics that occur under stress conditions.
A link between polyP and mitochondrial function has been previously shown in mammalian cells. Heterologous overexpression of a yeast exopolyphosphatase in mammalian cell lines, which degrades polyP, inhibits normal mitochondrial function, alters mitochondrial Ca 2+ metabolism, and may affect the development of the mitochondria permeability transition pore (46). On the basis of our data, the polyP molecule itself is not directly required for mitochondrial function in S-deprived Chlamydomonas cells. The vtc4-1 strain exhibited near full recovery of mitochondrial and chloroplastic electron flow when S and P were simultaneously removed from the medium (Fig. 4A), conditions that prevent accumulation of high ATP levels in the absence of polyP. On the basis of these physiological responses and our pharmacological results, we conclude that it is the synthesis of polyP that is critical for maintaining mitochondrial and chloroplastic electron transport by controlling the cellular ATP concentration.
While electron transport is one key regulatory component of acclimation, the integration of this component with signals that directly control SdR-associated genes requires further investigation. However, previous research has established the involvement of SNRK protein kinases (SNF1-related protein kinase) in the regulation of the SdR (47). These kinases in plants serve to integrate metabolic signals elicited by abiotic stress with acclimation responses (48). Chlamydomonas SNRK2.1 is essential for inducing transcription of most genes required for acclimation of the cells to S deprivation (28). Whether signals generated by electron flow in chloroplasts and mitochondria (e.g., reactive oxygen species or nitric oxide) modulate the activity of this kinase is not known. In plants, both phosphorylation and nitrosylation can alter the activity of some SNRK proteins (49). We raise the possibility that the SNRK proteins associated with S deprivation in Chlamydomonas (47) integrate signals coming from the environment (e.g., through phosphorylation) with intracellular indicators of energetic and metabolic conditions (e.g., through nitrosylation). Mitochondria and chloroplasts are hubs for generating retrograde stress signals that affect gene expression through redox signaling and the production of reactive oxygen species. Recently, it was reported that Chlamydomonas cells accumulate NO in response to S deprivation (43), but in-depth studies are required to determine whether or not this signaling molecule participates in the acclimation process.
As established in this study, polyP synthesis is required to maintain SdR after 2 to 3 hours in starvation, and this synthesis is likely activated by the binding of inositol pyrophosphate to the SPX do-main of VTC4. We speculate that once the cells experience nutrient limitation (before polyP synthesis), inositol phosphate metabolism could serve as the first "buffer" to control ATP levels. As cellular ATP increases, the most highly phosphorylated forms of inositol, the inositol pyrophosphates Ins7P and Ins8P, are synthesized. These two molecules bind to the VTC4 SPX domain, activating polyP synthesis (10), although with different efficiencies. Ins8P has been shown to be the most potent VTC4 activator in yeast (11). Therefore, we hypothesize that an initial increase in intracellular ATP would elicit Ins7P accumulation and partial activation of VTC4. With a further increase in the ATP concentration, Ins8P begins to accumulate, and VTC4 becomes fully active, providing cells a higher capacity to buffer ATP. The connection between inositol metabolism and polyP synthesis has been reported in trypanosomes and yeast (10,50), and a link between inositol pyrophosphates and the control of ATP homeostasis has also been reported in yeast. The kcs1 yeast mutant, which is unable to synthesize inositol pyrophosphate, accumulates ATP and exhibits a reduction in mitochondrial activity (51). Our data raise the possibility that this increase in ATP could reflect the inability of this mutant to activate the SPX domain of VTC4 and the initiation of polyP synthesis. This mechanism of regulation would allow for a tiered ATP buffering capacity based on the sequential activation of inositol pyrophosphate and a subsequent elevation in polyP synthesis. Starch and lipids synthesis may also function in ATP buffering and maintenance of mitochondrial and chloroplast electron flow when the capacity to store polyP in acidocalcisomes is exceeded.
How an increase in the ATP concentration in S-deprived vtc4-1 mutant cells inhibits electron flow is not absolutely clear, although part of the response could reflect limited availability of substrate (ADP and Pi) for sustaining high levels of electron transport in both mitochondria and chloroplasts. In addition, in higher eukaryotes, high ATP levels allosterically regulate complex IV in mitochondria (35), while in bacteria, the ATP synthase is regulated by the ATP/ADP ratio (36). Our studies demonstrate that vtc4-1 mutant cells were affected in their total ATP concentration, the ATP/ADP ratio, and the levels of respiratory electron flow and PET. Furthermore, the ATP level and ATP/ADP ratio in the vtc4-1 strain exhibited a similar increase, with no change in the ADP concentration. This result suggests that Chlamydomonas cells could either induce purine synthesis (elevated intracellular adenosine phosphate pool) or promote the phosphorylation of AMP to compensate for reduced regeneration of ADP through the polyP kinase. Over the short term, this strategy would allow cells to maintain basal electron flow rates (50% of respiratory electron flow and PET), which could be essential for basic organelle functions and survival.
Overall, the aspects of polyP regulation described in this manuscript could extend well beyond S deprivation and nutrient deprivation responses and into modulation of cellular processes in response to fluctuations in environmental conditions that alter growth and elicit changes in the levels of the different adenylate species. PolyP, a polymer that had been experimentally neglected, appears to be an integral factor in controlling a cascade of energetic processes that enable dynamic acclimation of organisms to fluctuations of the natural environment.

MATERIALS AND METHODS
Chlamydomonas and yeast strains CMJ030 (CC-4533) was used as our WT strain and is the parental strain of the vtc4-1 (LMJ.RY0402.086990) and vtc4-2 (LMJ.RY0402.195991) mutants, which were generated by insertional mutagenesis and are strains in the CLiP (Chlamydomonas Library Project) mutant library (32). The dum11 mutant (CC-4098, affected in complex III of the respiratory chain) was obtained from the Chlamydomonas Resource Center (www.chlamycollection.org/), and 137c (CC-124) was used as the parental strain. 21gr (CC-1690) was the WT strain for the experiments in which cells were exposed to N deprivation.
Media and culture conditions TAP medium was prepared according to standard recipes available at www.chlamycollection.org/methods. S-depleted TAP (TAP-S) medium was prepared by substitution of the sulfate salts with chloride salts. For P-and S-free (TA-S) medium, potassium phosphate was replaced with potassium chloride (52). Ammonium-free TAP (TAP-N) medium was used for N deprivation experiments. Hutner's trace elements for TAP and TAP-S media were purchased from the Chlamydomonas Resource Center. The solid medium contained agar (15 g liter −1 ). Chlamydomonas cells were grown in Erlenmeyer flasks at 23°C with agitation under fluorescent tubes at 100 mol photons m −2 s −1 . For nutrient deprivation experiments, cells were grown in TAP medium, pelleted by centrifugation for 2 min (2300g), and washed once in TAP-S medium before resuspending the cells at 10 g ml −1 of chlorophyll in the appropriate medium.

Reagents
Myx and FCCP from Sigma-Aldrich were dissolved in 100% ethanol and used at 7.5 and 5 M, respectively.

ARS activity and chlorophyll determination
Cells growing on solid TAP medium were streaked and maintained on TAP-S agar for 5 days. ARS activity was assayed by spraying colonies with 10 mM 5-bromo-4-chloro-3-indolyl sulfate (Sigma-Aldrich) in 0.1 M tris-HCl (pH 7.5) as previously described (47); the blue signal appeared after 1 to 2 hours. In liquid medium, ARS activity was assayed by mixing 500 l of culture (10 g ml −1 chlorophyll) with 500 l of the reaction mixture [4.5 mM p-nitrophenyl sulfate as substrate, 10 mM imidazole, and 0.1 M glycine (pH 9.0)]. Incubation of the reaction was for 30 min at 30°C. The reaction was stopped with 0.2 N NaOH, the debris pelleted by centrifugation for 2 min (14,000g), and the absorbance of the supernatant at 405 nm was determined. Chlorophyll measurements were made as previously described (53) after extraction of the pigments with methanol.

Genotyping of vtc4
Characterization of the insertion sites in vtc4-1 and vtc4-2 mutants was performed as detailed on the website www.chlamylibrary.org/ help. DNA was extracted from fresh colonies grown in TAP medium and resuspended and vortexed in 50 l of 5% Chelex-100 (Bio-Rad). The resuspended cells were boiled for 8 to 10 min, cooled on ice for 1 min, and made homogeneous by vortexing the suspension for 10 s. Cells were then pelleted by centrifugation for 1 min (20,000g), and 2 l of the supernatant were used as the template for PCR, which was performed using Taq Polymerase from QIAGEN according to the following protocol: 95°C for 3 min, 40 cycles of 95°C for 20 s, 60°C for 20 s, and 72°C for 2 min. Primers used, indicated in figs. S2 and S3, were oMJ913 (GCACCAATCATGTCAAGCCT), oMJ944 (GACGTTACAGCACACCCTTG), 1 (CCGCTCTCCAGATC-CCCTTTGAC), 2 (GCTAGAGCCTGGGCGGGTAGA), 3 (GAA GGGAAACGGAGGAAATGAGC), and 4 (TACGCCATCCTC-GTCAGCACAG).

Complementation of vtc4-1 mutant
The bleomycin cassette from the pSP124S plasmid was replaced by a hygromycin resistance cassette (from pRAM118 vector) using the restriction sites Kpn I-Hind III and generating the vector pSP124SH. The full cDNA (2337 bp) for VTC4 was amplified with primers VTC4-NdeIfw (GACATATGAAGTACGGAAAGTATATTGAG) and VTC4-EcoRIrev2 (CGAATTCTAGAACAGCGACAGCGTC-CC) using 0.4 U of Phusion Hot Start II DNA Polymerase with 0.42 M dimethyl sulfoxide and 0.6 M betaine. PCR conditions were those recommended by the manufacturer, and the annealing temperature was 63°C. The cDNA was cloned into the Nde I-Eco RI sites of pBleJM43. To introduce the native promoter and the first intron of VTC4, the promoter (920 bp), 5′ untranslated region (5′UTR), first two exons and 91 bp of the second intron were amplified with the primers VTC4-PHindIIIIfw (AAGCTTGCCGGCGGGACCA-CAGCTACGAGG) and VTC4g-653rev (CCGCCTACTTGCCAA GCACCTACG) using the same conditions as described above.
This amplified fragment was cloned into Hind III-Afl II sites in the vector containing the full-length cDNA. Last, this construct was digested with Hind III and Not I, and the fragment generated cloned into pSP124SH. The final construct pSP124SH+P+VTC4 bore the native promoter, 5′UTR, the cDNA of VTC4 with the first intron, and the 3′UTR of the PSAD gene from pBleJM43. The vtc4-1 mutant was transformed by electroporation with 2 g of pSP124SH+ P+VTC4 linearized with Not I. GeneArt MAX Efficiency Transformation Reagent for algae (Invitrogen) was used for electroporation following the recommendations of the manufacturer. Transformants were selected on solid TAP medium containing hygromycin (20 g ml −1 ; Enzo Life) and then assayed for ARS activity.

PolyP quantification
A volume of Chlamydomonas culture containing 50 g of chlorophyll was collected by centrifugation for 1 min (3000g), and polyP was extracted by the neutral phenol/chloroform and the ethanol precipitation method previously described (55). The lysates containing polyP were diluted (1:1000 to 1:4000) and quantified using the MicroMolar Polyphosphate Assay Kit (ProFoldin, PPD1000). For negative staining with 4′,6-diamidino-2-phenylindole (DAPI) (56), polyP samples were mixed with loading buffer [6× tris boric acid-EDTA (TBE), 15% Ficoll, 0.025% xylene cyanol FF, and 0.025% bromophenol blue] and electrophoresed into a 15% Criterion TBE-urea precast gel (Bio-Rad). The gel was rinsed with a fixative solution (50 mM tris base, 25% methanol, and 5% glycerol) and incubated for 30 min in staining solution [fixative solution + PureBlu DAPI (2 g/ml), Bio-Rad]. After the gel was incubated for 1 hour in fixative solution to eliminate excess DAPI, it was exposed to ultraviolet light for 5 to 10 min. DAPI bound to polyP bleaches in minutes, while the fluorescence of free DAPI or DAPI bound to nucleic acid is stable.

Quantification of ATP and ATP/ADP ratio
Chlamydomonas cells containing 50 g chlorophyll were pelleted from cultures and resuspended in 925 l of 0.5 M tris-HCl buffer (pH 7.5). A total of 75 l of 70% perchloric acid (Sigma-Aldrich) was added to the cell samples, which were then vortexed for 15 s to both lyse the cells and deproteinize the lysate. The samples were then incubated on ice for 1 min, neutralized with 4 M KOH, mixed thoroughly by vortexing, incubated for an additional minute on ice, and then centrifuged for 1 min (3200g). A total of 500 l of supernatant was transferred to a prechilled tube that was then flash-frozen in liquid N 2 and stored at −80°C. A total of 25 l of the supernatant was used for ATP quantification based on the "ATP Fluorometric Assay Kit" from Sigma-Aldrich (MAK190). Fluorescence was quantified in a Tecan Safire plate reader. For measuring the ATP/ADP ratio, we diluted the same samples (1:250) and used the "ADP/ATP Ratio Assay kit" (Sigma-Aldrich, MAK135). Bioluminescence was determined using the TD-20/20 luminometer from Turner Designs.

Respiration measurements
Dark respiration and gross O 2 uptake were measured using a Clark-type electrode with the Chlorolab 2 + System (Hansatech). Chlamydomonas cells in midlog growth phase were pelleted for 2 min (2300g), dispersed in TAP or TAP-S at 10 g ml −1 chlorophyll, and then grown for 6 hours in the light. The cells were then dark-acclimated for 10 min, and 2 ml of culture was used to measure O 2 uptake. Respiration values were calculated from the initial slope of dark O 2 uptake. When indicated, gross O 2 uptake was measured directly from light-incubated cells without dark acclimation. Cells were illuminated in the O 2 chamber with light-emitting diode (LED) lighting (white light of 4100 K provided with the Chlorolab 2 + System) of the intensities indicated in the text.

Photosynthetic electron transport
Cells grown in TAP or TAP-S media to a chlorophyll concentration of 10 g ml −1 were dark-acclimated for 20 min. Chlorophyll fluorescence measurements were performed using a DUAL-PAM-100 (Walz). Fv/Fm was measured following dark acclimation. Cells were then exposed to 120 mol photons m −2 s −1 for 15 s to quantify the yield of PSII. PET was calculated according to the formula PET = 0.5 × PSII yield × 0.84 × photosynthetically active radiation (in mol photons m −2 s −1 ).

Chlamydomonas genetic crosses
For the N deprivation experiments, the vtc4-1 mutant was crossed, according to a previously described method (57), with a WT strain able to use ammonium, nitrate, or nitrite as sole N sources (strain 21gr; fig. S10). Instead of tetrad analysis, the zygotes were germinated in paromomycin (25 g ml −1 ) and 7.5 mM nitrate as the sole N source, providing a positive selection for segregants that are drug resistant (containing the introduced cassette) and that can grow on nitrate; these segregants contain the vtc4-1 lesion in the 21gr genetic background with respect to N utilization.

Protein extraction and immunoblot analysis
Total cells in a volume of culture containing 50 g of chlorophyll were pelleted for 2 min by centrifugation (2300g). The pellet was resuspended in protein extraction buffer [5 mM Hepes (KOH; pH 7.5), 10 mM EDTA (NaOH; pH 7.5), 1 mM phenylmethylsulfonyl fluoride, 1 mM -amino-n-caproic acid, and 1 mM benzamidine HCl], collected again by centrifugation for 2 min (14,000g), and resuspended in extraction buffer containing 100 mM dithiothreitol, 100 mM Na 2 CO 3 , 2% (w/v) SDS, and 12% (w/v) sucrose. Samples were boiled for 1 min, briefly incubated on ice, and then subjected to SDS-polyacrylamide gel electrophoresis (12% acrylamide gels); the loading of the samples were normalized to chlorophyll content (and similar levels of tubulin). The polypeptides were then transferred to polyvinylidene difluoride membranes using the Trans-Blot Turbo Midi and the Trans-Blot Turbo Transfer Systems from Bio-Rad following the manufacturer's instructions. Membranes were blocked for 1 hour at room temperature with a 5% (w/v) suspension of powdered milk in tris-buffered saline containing 0.1% (w/v) Tween 20 (TBS-T). The membranes were then incubated for 2 hours with the primary LAO1 antibody (1:10,000, 3% milk in TBS-T) at room temperature followed by three 5-min washes in TBS-T and incubated with the secondary antibody (3% milk in TBS-T) for 1 hour at room temperature. Secondary antibodies used for LAO1 and tubulin were horseradish peroxidase (HRP)-conjugated anti-rabbit immunoglobulin G (IgG; Life Technologies, 656120) and anti-mouse IgG (Life Technologies, G2104), respectively. The membranes were then each washed three times for 5 min with TBS-T, and peroxidase activity was assayed by chemiluminescence using the WesternBright ECL HRP substrate (Advansta).

SUPPLEMENTARY MATERIALS
Supplementary material for this article is available at http://advances.sciencemag.org/cgi/ content/full/6/40/eabb5351/DC1 View/request a protocol for this paper from Bio-protocol.