Leaders in Pharmaceutical Business Intelligence (LPBI) Group

Buffering of genetic modules involved in tricarboxylic acid cycle metabolism provides homeomeostatic regulation

Larry H. Bernstein, MD, FCAP, Reviewer and Curator

Leaders in Pharmaceutical Intelligence

This posting is the fifth in a series on metabolomics.  The first covered general principles.  Proteomics has not been covered and will be returned to.  But we have opened a door.  We have now looked at a comparison of two lymphocytic cell lines, and then the measurement of external effluxes to define internal metabolic conditions in yeast, with a view to delineating relationships between internal metabolic pathways and genetic variants under metabolic constraints.  These studies were confined to the experimental conditions, and could not measure metabolic fluxes, but I consider a study of the fluxes referred to in the comment by Dr. Jose Eduardo des Salles (JEDS) Roselino, who has brought up the concept, not yet specifically discussed – homeostasis.  It is part of a series of communications over several months.  

In the last article I might not have provided answers to some of the questions posed up front.  One of them refers to whether one finds a relationship to the Pasteur effect.  In both studies, leukemia cells and yeast, the cells are eukariotic, not prokaryotic, although the studies were preceded by other studies of bacteria.  It is important to remember that there are differences between prokaryotes and eukaryotes, and these studies encompassed aerobiuc and anaerobic glycolysis, mitochochondrial pathways, and cell death pathways, as well as energy balance, which involves ATP and a series of linked hydrogen transfers in the electron transport chain (ETC).  In the first two papers, we could infer the comparison between differences in oxidative phosphorylation and lactic aciid formation between two strains of cells, whether they be yeast or lymphocytic leukemia.  This is where the observation of Otto Warburg refers back to the work of Pasteur 60 years prior to his discovery.  

In this study we find that metabolic fluxes can be and are measured in Saccharomyces Cerevisiae, and the internal metabolites are measured extensively.  JEDS refers me to Schroedinger’s (Physics Nobel, Quantum Field Theory) classic work, “What is Life?” and his famous CATS, or Rabbits ( a Twilight Zone where objects can be two places at the same time: Schroedinger’s Rabbits: Colin Bruce, Joseph Henry Press, Washington, DC.) That is for another time.  I have also previously referred to the work of Ilya Prigogine (Chemistry Nobel; self organizing systems).  We can set up studies, but we cannot identify the initial state. These two major scientists understood the limits of our ability to study life.

The focus here is on homeostasis.  Homeostasis (Wikipedia), also spelled homoeostasis (from Greek: ὅμοιος, “hómoios”, “similar”,^[1] and στάσις, stásis, “standing still”^[2]), is the property of a system in which variables are regulated so that internal conditions remain stable and relatively constant. Examples of homeostasis include the regulation of temperature and the balance between acidity and alkalinity (pH). It is a process that maintains the stability of the human body’s internal environment in response to changes in external conditions. The concept was described by Claude Bernard in 1865.  The term was originally used to refer to processes within living organisms, it is frequently applied to automatic control systems.

Buffering of deoxyribonucleotide pool homeostasis by threonine metabolism

John L. Hartman IV *

Author Affiliations   

 Communicated by Leland H. Hartwell, Fred Hutchinson Cancer Research Center, Seattle, WA, June 4, 2007 (received for review January 21, 2007)

Synergistically interacting gene mutations reveal buffering relationships that provide growth homeostasis through their compensation of one another.

Abstract

This analysis in Saccharomyces cerevisiae revealed genetic modules involved in

• tricarboxylic acid cycle regulation (RTG1, RTG2, RTG3),
• threonine biosynthesis (HOM3, HOM2, HOM6, THR1,THR4),
• amino acid permease trafficking (LST4, LST7), and
• threonine catabolism (GLY1).

These modules contribute to a molecular circuit that

• regulates threonine metabolism and
• buffers deficiency in deoxyribonucleotide biosynthesis.

Phenotypic, genetic, and biochemical evidence for this buffering circuit was obtained

• through analysis of deletion mutants,
• titratable alleles of ribonucleotide reductase genes, and
• measurements of intracellular deoxyribonucleotide pool concentrations.

This circuit provides experimental evidence, in eukaryotes, for the presence of a

• high-flux backbone of metabolism,

which was previously predicted from 

• in silicomodeling of global metabolism in bacteria.

This part of the high-flux backbone appears to

• buffer deficiency in ribonucleotide reductase
• by enabling a compensatory increase in
• de novopurine biosynthesis
• that provides additional rate-limiting substrates for
• dNTP production and DNA synthesis.

Hypotheses regarding unexpected connections

• between these metabolic pathways
• were facilitated by genome-wide, and
• quantitative phenotypic assessment of interactions.

Validation of these hypotheses substantiates

• the added benefit of quantitative phenotyping
• for identifying subtleties in gene interactionnetworks
• that modulate cellular phenotypes.

Keywords: genetic buffering, high-flux backbone of metabolism, protein trafficking, ribonucleotide reductase, mitochondria-to-nucleus retrograde signaling pathway

Introduction 

Cells are complex genetic systems, having evolved

• compensatory molecular networks
• that provide growth homeostasis (robustness).

Conceptually, gene interactions

• underlie robustness by buffering
• environmental or genetic perturbations (1–3).

Synergistic effects on the phenotype resulting from

• two genetic deficiencies or chemical inhibition
• in combination with a genetic deficiency

1. reveal buffering relationships
2. when the double limitation
3. is more severe than either single limitation.

Genome-wide phenotypic analysis,

• as possible with RNAi or
• use of the complete set of
  yeast gene deletion mutants,
• has enabled new approaches
• to investigate buffering relationships
  systematically (2, 4, 5).

It has been shown that

• quantitative (strength) and
• qualitative (pattern) aspects of
  gene interaction profiles reveal
  • how genes organize
  • in a pathway or cellular process (4, 6).

such sets of genes represent genetic modules that

• contribute buffering capacity to the cell,
  • providing insight into
  • how molecular circuitry

is arranged to achieve robustness (7, 8).

Comprehensive and quantitative methods

• for genotype–phenotype analysis are available
  • to gain a more global and precise understanding
  • of buffering networks (4, 6, 9).

These methods permit unbiased

• investigation of growth homeostasis,
• systematically revealing
  • how combinations of genetic and
  • environmental variables

result in phenotypic complexity.

High-throughput genotype–phenotype data

• offer an opportunity to use
• the extensive and growing genome annotations

to discover new connections between

• previously annotated genes and pathways,
• with respect to physiological homeostasis.

Systematic, experimentally derived understanding of

• genetic interaction networks

will advance efforts to

• map natural phenotypic variation,
• thereby aiding the dissection

of genetic disease complexity (10).

This work tests a model constructed after

• finding threonine biosynthesis
• to play a role in
• buffering growth inhibition with

the deoxyribonucleotide (dNTP) biosynthesis
inhibitor, hydroxyurea (HU) (4).

HU is a chemotherapy agent that

• limits cell proliferation by inhibition of
  ribonucleotide reductase (RNR)
  • leading to dNTP pool deficiency and
  • slow DNA synthesis (11).

The results provide

• genetic,
• biochemical, and
• phenotypic evidence

that growth homeostasis

• is maintained by
• a molecular circuit
  • that regulates threonine metabolism
  • to buffer depletion of dNTP pools.

These findings shed light on

• systems-level observations about
• cellular metabolism, including
  • function of a high-flux backbone of metabolism (12)
  • and gating of DNA synthesis by
    • oscillation of global transcription
    • and redox metabolism (13, 14).

FUNCTIONAL INTERACTIONS BETWEEN DNTP AND THREONINE METABOLISM.

This work focused on understanding genetic modules

• found to buffer RNR deficiency (4).

Synergistic interactions between HU and

• threonine biosynthesis genes, were uncovered ( 1a).
• but not genes that function in the synthesis of other amino acids

Deletion ofAAT2 (aspartate aminotransferase) was also synergistic,

• suggesting buffering by tricarboxylic acid (TCA) cycle flux
• as AAT2converts the TCA cycle intermediate,
  • oxaloacetate è aspartate
  • the substrate for synthesis of homoserine
  • and ultimately threonine (yeastgenome.org). 

RTG1, RTG2, and RTG3, transcription factors

• regulating transcription of TCA cycle genes
• in response to mitochondrial stress
  • were also synergistic with hydroxyurea growth limitation
• further implicating TCA cycle involvement ( 1b).

The RTG and threonine biosynthesis modules were

• independently confirmed to buffer HU-induced stress
  by Panet al. (17).

Synergistic interaction between HU and

• deletion alleles ofLST4 and LST7 
• implicated extracellular uptake of threonine
• as an alternative mechanism
• to augment threonine flux ( 1cand ​and22a)
  • becauseLST4 and LST7 regulate
  • delivery of amino acid permeases
    • between the vacuole and
    • plasma membrane compartments of the cell (18).

To confirm that chemical–genetic

• interactions with HU
• were caused by its known inhibitory effect
• on dNTP biosynthesis,
  • a more specific method was used.

Integrating plasmids were

• introduced into mutant strains
  • to placeRNR1 or RNR2 
• under transcriptional control by doxycycline (19).

Deletion of homoserine or threonine biosynthesis genes

• was found to be synergistic with
• repression of RNR activity by
• using doxycycline in these mutants ( 1d),
  • confirming that interactions with HU
  • were caused by its inhibitory effect on RNR.

 Media supplementation with amino acids

• tested whether uptake of extracellular threonine
  • suppresses the growth limitation of mutations
• in threonine biosynthesis in the presence of HU.

Threonine was found to selectively suppress

• interaction between HU and
• disruption of threonine biosynthesis,
• in a concentration-dependent manner ( 2b–d).

This finding led to the prediction that

• disabling both threonine biosynthesis and
• threonine uptake
  • would be synergistic
• in the presence of HU growth limitation.

 Double deletion mutants of

• the four possible combinations of
• thr1or thr4 and lst4 or lst7 were created.

 All combinations were synthetic lethal

• even in the absence of HU

This is consistent with the hypothesis that lst4 and lst7 

• compensate threonine biosynthetic deficiency
• through regulation of extracellular uptake ( 3).

A slow-growth phenotype observed

• for thehom6 deletion mutant
  • was exacerbated by extracellular threonine,
• even in the absence of HU (Fig. 2b).

 The hom6deletion mutant is unique

• among threonine biosynthesis mutants
• in that the resulting intermediate metabolite is toxic (20)
  (aspartate β-semialdehyde)

Whether this toxicity could be related to its different phenotype

• in the context of HU perturbation is unexplained.

BIOSYNTHESIS AND EXTRACELLULAR UPTAKE OF THREONINE CONTRIBUTE TO DNTP POOL HOMEOSTASIS.

To test the effect of threonine metabolism on dNTP pools

• pools were measured in threonine metabolism deletion mutants
• perturbed by doxycycline-conditional repression
  • of RNR2transcription ( 4).

Although rnr2 deletion is lethal in a haploid,

• repression of RNR2transcription
• only reduced the growth rate

when an otherwise non-growth-inhibitory concentration (5 mM) of HU was present (data not shown).

In contrast, RNR1repression was

• growth-limiting without HU (see  1d) and
• did not sensitize growth to 5 mM HU
  (data not shown).

The specificity of low-dose HU for

• growth inhibition in combination with
• repression of RNR2
• is explained by the mechanism of action of HU. 

HU scavenges a tyrosyl radical

• that is present on the Rnr2p and
• that is required as a cofactor
• for ribonucleotide reduction (21).

 Non-growth-inhibitory concentrations of HU

• paradoxically induced increased steady-state
• dNTP pool concentrations.

The increase in pools was sustained over time

• and additive with the effect of modulating RNR
  transcriptional levels ( 4a).
• As a result, growth inhibition,
• caused byRNR2 repression
• combined with low-dose HU treatment,
• occurred with dNTP levels slightly higher than
• those in the untreated wild-type (WT) control strain
  (endogenousRNR2 promoter).

A possible explanation is that increases in dNTP pools

• are required for growth fitness
• in the setting of DNA damage, which is
• known to involve RNR regulation (22).

 Fig. 1.

HU chemical–genetic interactions. Interactions for three genetic modules are depicted.
The WT control is compared with deletion strains representative of each module,
with respect to their area under the growth curve (AUGC) vs. perturbing drug …

Fig. 2.

Extracellular threonine suppresses interactions between HU and threonine biosynthesis.
(a) A model explaining interactions between HU and genes involved in threonine metabolism.
In the context of dNTP pool deficiency, threonine metabolism is up-regulated …

To confirm that chemical–genetic interactions with HU

• were caused by its known inhibitory effect on dNTP biosynthesis,
• a more specific method was used.

Integrating plasmids were introduced

• into a variety of mutant strains to place 
• RNR1or RNR2 under transcriptional control
  by doxycycline (19).

Deletion of homoserine or threonine biosynthesis

• genes was found to be synergistic with
• repression of RNR activity by using doxycycline
• in these mutants ( 1d), confirming that
  • interactions with HU were caused by
    its inhibitory effect on RNR.

Media supplementation with amino acids was used

• to test whether uptake of extracellular threonine
• suppresses the growth limitation of mutations
• in threonine biosynthesis in the presence of HU.

Threonine selectively

• suppresses interaction between HU and
• disruption of threonine biosynthesis,
  • in a concentration-dependent manner
    ( 2b–d).

This finding led to the prediction that disabling

• both threonine biosynthesis and threonine uptake
• would be synergistic with HU growth limitation.

To test this hypothesis, double deletion mutants of

• the four possible combinations of
  • thr1or thr4 and lst4 or lst7 were created.

All combinations were synthetic lethal

• even in the absence of HU,
• consistent with the hypothesis that
  • lst4andlst7 compensate threonine
    biosynthetic deficiency

through regulation of extracellular uptake (Fig. 3).

A slow-growth phenotype observed for

• the hom6deletion mutant was exacerbated
• by extracellular threonine, even
• in the absence of HU ( 2b).

The hom6deletion mutant is unique among

• threonine biosynthesis mutants in that
• the resulting intermediate metabolite is toxic (20)
  (aspartate β-semialdehyde), although
  • whether this toxicity could be related
    to its different phenotype
• in the context of HU perturbation is unexplained.

Fig. 3.

Deficiency in amino acid permease trafficking (LST4 or LST7) is synthetic lethal
with deletion of threonine biosynthesis (THR1 or THR4). Interactions between
permease trafficking (LST4 and LST7) and threonine biosynthesis (THR1 and THR4)
were assessed …

BIOSYNTHESIS AND EXTRACELLULAR UPTAKE OF
THREONINE CONTRIBUTE TO DNTP POOL HOMEOSTASIS.

To test the effect of threonine metabolism on dNTP pools,

• pools were measured in threonine metabolism deletion mutants
• perturbed by doxycycline-conditional repression of RNR2transcription ( 4).

Although rnr2 deletion is lethal in a haploid,

• repression of RNR2transcription only
• reduced the growth rate when an otherwise
  • non-growth-inhibitory concentration (5 mM)
• of HU was present (data not shown).

In contrast, RNR1repression was

• growth-limiting without HU (see  1d)
  • and did not sensitize growth
  • to 5 mM HU (data not shown).

The specificity of low-dose HU for growth inhibition

• in combination with repression of RNR2
• is explained by the mechanism of action of HU.

HU scavenges a tyrosyl radical

• present on the Rnr2p
• that is required as a cofactor
  • for ribonucleotide reduction (21).

Fig. 4.

Effect of threonine metabolism on dNTP pool homeostasis.
(a) Intracellular dNTP pool concentrations are depicted 90 (black)
and 360 (gray) min after exposure to the perturbations  indicated
by each block. Block 1 is the unperturbed WT (BY4741) strain in …

Non-growth-inhibitory concentrations of HU

• paradoxically induced increased steady-state
  dNTP pool concentrations.

The increase in pools was sustained over time

• and additive with the effect of modulating
  RNR transcriptional levels ( 4a).

As a result, growth inhibition, caused by RNR2 repression

• combined with low-dose HU treatment, occurred
  • with dNTP levels slightly higher than those
  • in the untreated wild-type (WT) control strain
    (endogenous RNR2promoter).

A possible explanation is that increases in dNTP pools are

• required for growth fitness in the setting of DNA damage,
• which is known to involve RNR regulation (22).

However, production of DNA damage
(requiring increased dNTP pools for DNA repair)

• would have been expected only at HU concentrations
• high enough to arrest DNA synthesis in the first place (23).

The observation that low concentrations of HU

• led to increased pools could be explained
• if DNA damage occurs by a mechanism
  • independent of the effect of HU
  • on cytoplasmic pools.

A possible mechanism could involve

• dNTP pool concentrations at replication forks
• being affected differentially from
  • cytoplasmic pools;
• this is not thought to occur
  in eukaryotic cells (24).

Thus, the paradoxical effect of

• low HU concentrations
• on increasing dNTP pools
• remains unexplained.

dNTP pools were increased

• by expression ofRNR2  from the Tet promoter
  (in the absence of repression with doxycycline),
• presumably because of overexpression
• relative to the endogenousRNR2 

Dox-conditional repression of RNR2 

• reduced dNTP pools in a
• concentration-dependent fashion ( 4a)
  • so that synergism from deletion of
    threonine metabolism genes could be tested.

The rtg2, hom2, thr1, and lst4 deletion mutants

• all exacerbated the reduction in dNTP pools
  • afterRNR2 repression ( 4b).

The contribution of RTG2 for

• dNTP pool maintenance
• was less than that of HOM2,THR1, or LST4,
• consistent with their effects on growth ( 1).

 Consistent with low dNTP pools

• causing cell cycle arrest in each of the mutants,
  • median cell size increased
  • the relative number of cells and
  • total cell volume
    (median cell size × median cell volume)
  • decreased as pools became depleted ( 4c).

The scs7 (functions in sphingolipid metabolism) deletion strain

• maintained dNTP pools comparable with WT,
• despite a greater fitness defect ( 4b and c),
  • indicating a specific role of threonine metabolic genes
  • in homeostatic regulation of dNTP pools.

THREONINE ALDOLASE IS RATE-LIMITING FOR DNTP METABOLISM IN SACCHAROMYCES CEREVISIAE.

The genetic, phenotypic, and biochemical results

• are consistent with a model whereby
  • TCA cycle regulation (RTG genes),
  • threonine biosynthesis (HOM and THR genes), and
  • permease trafficking (LST genes) pathways
  • coordinately buffer dNTP pool depletion by
    • up-regulating threonine metabolism.

The model postulates that threonine catabolism

• contributes glycine to augment
• de novopurine synthesis ( 2a).

HU has been shown to preferentially deplete

• dATP pools in mammalian cells (25, 26), and
• there was a tendency for purine pools to fluctuate
  (particularly dATP)
  • acutely whenever threonine metabolism and
  • RNR activity were perturbed in combination
    ( 4aand b).

However, allosteric regulation of RNR would

• distribute this effect across all pools (21).

Threonine aldolase, encoded by GLY1 (EC 4.1.2.5),

• cleaves threonine into glycine and acetaldehyde (27).

Notably, the gly1 deletion mutant exhibited slow growth
(data not shown)

• even with glycine supplementation.

This phenotype was found to be the result of

• limitation of dNTP metabolism.

Basal dNTP pools were reduced

• in the gly1deletion mutant,

Pools fell dramatically

• after treatment with 10 mM HU,

and normal homeostatic increases in dNTP concentrations

• after treatment with 50 mM HU were delayed,
• particularly dATP pools (Fig. 5). 

CHA1 (EC 4.2.1.13) and ILV1 (EC 4.3.1.19) are

• deaminases that convert threonine to 2-oxybutanoate
• or other metabolic intermediates such as
  • homoserine, cystathionine, or propionyl-CoA.

However, deletion of neither CHA1 nor ILV1 

• modified the growth response to HU (4).

Fig. 5.

Threonine aldolase contributes to normal dNTP metabolism. Intracellular dNTP pools are shown for the WT control strain (BY4741) and gly1 (threonine aldolase) deletion mutant before (Left) and 120 or 360 min after perturbation with 10 mM (Center) or 50 …

DISCUSSION

Computational analysis of global metabolism

• inEscherichia coli has suggested
• that threonine flux is of particular importance.

These studies propose that

• threonine synthesis and
• its degradation to glycine
  • for purine biosynthesis
• are part of a high-flux backbone (HFB)
• of metabolism (12).

The HFB was defined by a

• subset of all metabolic reactions
• found to have sufficient flux for
  • providing growth homeostasis
• in response to growth-limiting perturbations
  (such shifting to a poor carbon source).

Utilization of threonine for buffering

• dNTP metabolism and growth homeostasis
• provides experimental evidence for
• the presence of the HFB in eukaryotes.

Discovery of new connections between

• dNTP and threonine metabolism

1. demonstrates the value of quantitative high-throughput
   cellular phenotyping for identifying 
2. functional redundancies in gene networks
   • by measuring interactions between
   • genetic module metabolism.

The ability to detect relatively small effects

• of individual modules and
• to order their relative quantitative impact
  • aided hypotheses about how
  • these modules might relate to one another (4).

By this approach, genes involved in

1. TCA cycle regulation,
2. threonine biosynthesis,
3. amino acid permease trafficking,
4. threonine catabolism, and
5. ribonucleotide reduction

• were found to function as a modular circuit
• to maintain robust dNTP pools
• for DNA synthesis
  • even though these modules
  • appear to function independently
    in other contexts (1, 8, 35).

In natural (outbred) populations,

• compensatory networks also buffer
• genetic and chemical growth perturbations;

however, the amount of genotypic and phenotypic variation

• renders dissection of interactions relatively intractable.

By contrast, systematic analysis of yeast deletion mutants

• exposes interactions on a fixed genetic background
  • but does not survey natural variation.

Recently, segregants from a cross of S288C
(the background used for systematic gene deletion)

• and a natural isolate
• have been genotyped at high resolution (36, 37).

Quantitative high-throughput cellular phenotyping,

• applied in parallel to these strains and
• the comprehensive collection of
• yeast gene deletion mutants, would
  • provide a dual strategy to
  • deconstruct gene networks
  • that buffer growth perturbations, by
  • systematic analysis of all deletion mutants
  • in parallel with surveying for natural occurrence.

Quantitative genetic dissection of

• buffering networks in yeast thus
• provides a way to model genotype–phenotype variation
• on a genomic scale, providing insight into
  • functional interactions between conserved pathways
  • that potentially modulate human disease.

Cell Proliferation Measurements. 

Experiments represented in Figs. 1 and ​and22 were performed in Hartwell complete agar medium. High-throughput kinetic phenotyping (by imaging and image analysis) and area under the growth curve (AUGC) calculations were performed as described previously (4). AUGC encapsulates the overall growth phenotype of a strain with respect to time under a particular condition. AUGC is affected by initial population size (no. of cells transferred in a spot culture), lag time (delay before log-linear growth), maximum specific rate (actual log-linear rate), total efficiency (saturation density), and duration of the assay. For assessing the strength of a genetic interaction, the change in the AUGC conferred by a particular deletion allele relative to its WT control allele is considered with respect to perturbation intensity, e.g., concentration of HU, as depicted in Fig. 1. AUGC values for all mutants perturbed with 0, 50, and 150 mM HU are available at (http://genomebiology.com/2004/5/7/R49/additional).

dNTP Pool Sample Collections. 

Strains were grown overnight in liquid medium at 30°C to a concentration of ≈3 × 10⁶ cells/ml and diluted to prewarmed medium with HU or doxycycline to achieve the desired cell and drug concentrations in a final volume of 30 ml. Each time point was grown separately and harvested when the cell concentration was ≈3 × 10⁶ cells per ml. Twenty milliliters of culture was collected by vacuum filtration and immediately washed with ice-cold medium, and filters were transferred to 2 ml of ice-cold medium (dNTP concentrations remain stable in iced medium for several hours). Cells were removed from the filter by vortexing, the sample was divided in half for duplicate readings, cells were pelleted, medium removed by aspiration, and cells were lysed with 40 μl of 0.1 M perchloric acid and then snap-frozen.

Cell Volume Measurements. 

Cell volumes were measured by size analysis with a Coulter Counter (Beckman–Coulter, Fullerton, CA). The total cell volume of each culture (median cell size × total cell number) was used for calculating intracellular dNTP pool concentrations. Before vacuum filtration and lysis of each culture for mass spectrometry analysis, 200 μl was collected into 10 ml of ice-cold isoton (Beckman–Coulter). Samples were sonicated at low power to separate nonspecifically adherent cells. To calculate relative changes in total cell volume (Fig. 4c), values for each strain were first normalized against self at time zero and then divided by the corresponding normalized WT (BY4741) values.

HPLC. 

Samples were thawed by microcentrifugation (18,000 × g) for 15 min at 4°C. Sixteen microliters of lysate was added to 8 μl of 3× mobile-phase buffer [60 mM acetic acid/0.075% dimethylhydroxylamine (Sigma, St. Louis, MO)/pH adjusted to 7 with ammonium hydroxide], and 10 μl was injected onto an Agilent C-8 Zorbax column (part 883700-906) with a linear 5–30% methanol gradient from 2 to 11 min, 30–50% from 11 to 12 min, with final reequilibration for 5 min in 5% methanol (flow rate of 0.3 ml/min). Retention times of 4.5 (dTTP), 7.5 (dGTP and dTTP), and 9.5 min (dATP) were observed. dNTP-depleted lysate was obtained by lysis of saturation-density cultures after 30-min incubation in room temperature water. Dilution of standards in this lysate improved dCTP chromatography. Trace amounts of dNTPs remaining in the diluent were subtracted for standard curve calculations.

Mass Spectrometry. 

Mass spectrometry was performed with electron spray ionization in negative ion mode. Two instruments were used: (i) an Agilent 1100 MSD [dNTPs were monitored as single ions at m/z 466 (dCTP)], 481 (dTTP), 490 (dTTP), and 506 (dGTP). The drying gas was N₂ at 340°C at 10 liters/min, and nebulizing pressure 25 psi (1 psi = 6.89 kPa). The fragmentor was set at 90 eV and capillary voltage 3500. (ii) An ABI API-4000 Q-trap triple quadrupole instrument was used [mass transition to a 189 fragment was monitored for each of the dNTP species, as described previously (39); N₂ gas was used for nebulization, drying, and collision and the ionization chamber temperature was 250°C]. New standard curves were created for every assay.

Calculation of Intracellular dNTP Concentrations. 

Sample concentrations were determined from standard curves and adjusted to account for dilution by lysis and total cell volume [volume added for lysis + 2(tcv)] μL /tcv (μL). Standard curves showed high linear correlation (R² > 0.998), and variation from duplicate mass spec measurements was generally <10%.

ABBREVIATIONS: AUGC, area under the growth curve; dNTP, deoxyribonucleotide; HU, hydroxyurea; RNR, ribonucleotide reductase; TCA, tricarboxylic acid..

The requirement reported here of mitochondrial-to-nucleus retrograde signaling for dNTP pool homoeostasis in yeast may be of importance to a recent report that mutations in p53R2 cause human mitochondrial depletion syndromes (MDS) (40, 41). If compensatory/buffering relationships between RNR and retrograde signaling in yeast are evolutionarily conserved, then genetic variation in retrograde signaling may modulate MDS disease phenotypes resulting from deficiency in p53R2 activity.

ARTICLE INFORMATION

Proc Natl Acad Sci U S A. Jul 10, 2007; 104(28): 11700–11705.

Published online Jul 2, 2007. doi:  10.1073/pnas.0705212104

PMCID: PMC1913885

Genetics

John L. Hartman, IV^*

Department of Genetics, University of Alabama at Birmingham, Birmingham, AL 35294

*To whom correspondence should be addressed. E-mail: ude.bau@namtrahj

Communicated by Leland H. Hartwell, Fred Hutchinson Cancer Research Center, Seattle, WA, June 4, 2007.

Author contributions: J.L.H. designed research, performed research, contributed new reagents/analytic tools,
analyzed data, and wrote the paper.

Received January 21, 2007

Buffering of deoxyribonucleotide pool homeostasis by threonine metabolism

John L. Hartman, IV

Additional article information

NOTE ADDED IN PROOF.

Note Added in Proof.

The requirement reported here of mitochondrial-to-nucleus retrograde signaling for dNTP pool homoeostasis in yeast may be of importance
to a recent report that mutations in p53R2 cause human mitochondrial depletion syndromes (MDS) (40, 41). If compensatory/buffering
relationships between RNR and retrograde signaling in yeast are evolutionarily conserved, then genetic variation in retrograde signaling
may modulate MDS disease phenotypes resulting from deficiency in p53R2 activity.

FOOTNOTES

The author declares no conflict of interest.

ARTICLE INFORMATION

Proc Natl Acad Sci U S A. Jul 10, 2007; 104(28): 11700–11705.

Published online Jul 2, 2007. http://dx.doi.org:/10.1073/pnas.0705212104

PMCID: PMC1913885

Genetics

John L. Hartman, IV^*

Department of Genetics, University of Alabama at Birmingham, Birmingham, AL 35294

*To whom correspondence should be addressed. E-mail: ude.bau@namtrahj

Communicated by Leland H. Hartwell, Fred Hutchinson Cancer Research Center, Seattle, WA, June 4, 2007.

Author contributions: J.L.H. designed research, performed research, contributed new reagents/analytic tools, analyzed data, and wrote the paper.

Received January 21, 2007  Copyright © 2007 by The National Academy of Sciences of the USA

This article has been cited by other articles in PMC.

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

 REFERENCES

1. Hartman JL, IV, Garvik B, Hartwell L. Science. 2001;291:1001–1004. [PubMed]
2. Tong AH, Lesage G, Bader GD, Ding H, Xu H, Xin X, Young J, Berriz GF, Brost RL, Chang M, et al. Science. 2004;303:808–813. [PubMed]
3. Lehner B, Crombie C, Tischler J, Fortunato A, Fraser AG. Nat Genet. 2006;38:896–903.[PubMed]
4. Hartman JL, IV, Tippery NP. Genome Biol. 2004;5:R49. [PMC free article] [PubMed]
5. Parsons AB, Brost RL, Ding H, Li Z, Zhang C, Sheikh B, Brown GW, Kane PM, Hughes TR, Boone C. Nat Biotechnol. 2004;22:62–69. [PubMed]
6. Collins SR, Schuldiner M, Krogan NJ, Weissman JS. Genome Biol. 2006;7:R63.[PMC free article] [PubMed]
7. Csete ME, Doyle JC. Science. 2002;295:1664–1669. [PubMed]
8. Hartwell LH, Hopfield JJ, Leibler S, Murray AW. Nature. 1999;402:C47–C52. [PubMed]
9. Shah NA, Laws RJ, Wardman B, Zhao LP, Hartman JL., IV BMC Syst Biol. 2007;1:3.[PMC free article] [PubMed]
10. Badano JL, Katsanis N. Nat Rev Genet. 2002;3:779–789. [PubMed]
11. Krakoff IH, Brown NC, Reichard P. Cancer Res. 1968;28:1559–1565. [PubMed]
12. Almaas E, Kovacs B, Vicsek T, Oltvai ZN, Barabasi AL. Nature. 2004;427:839–843.[PubMed]

 …..