Effects of Low-Shear Modeled Microgravity on Cell Function, Gene Expression, and Phenotype in Saccharomyces cerevisiae

Only limited information is available concerning the effectsof low-shear modeled microgravity (LSMMG) on cell function andmorphology. We examined the behavior of Saccharomyces cerevisiaegrown in a high-aspect-ratio vessel, which simulates the low-shearand microgravity conditions encountered in spaceflight. Withthe exception of a shortened lag phase (90 min less than controls;P < 0.05), yeast cells grown under LSMMG conditions did notdiffer in growth rate, size, shape, or viability from the controlsbut did differ in the establishment of polarity as exhibitedby aberrant (random) budding compared to the usual bipolar patternof controls. The aberrant budding was accompanied by an increasedtendency of cells to clump, as indicated by aggregates containingfive or more cells. We also found significant changes (greaterthan or equal to twofold) in the expression of genes associatedwith the establishment of polarity (BUD5), bipolar budding (RAX1,RAX2, and BUD25), and cell separation (DSE1, DSE2, and EGT2).Thus, low-shear environments may significantly alter yeast geneexpression and phenotype as well as evolutionary conserved cellularfunctions such as polarization. The results provide a paradigmfor understanding polarity-dependent cell responses to microgravityranging from pathogenesis in fungi to the immune response inmammals.

INTRODUCTION

Top

Introduction

Understanding the effects of microgravity on the functions ofeukaryotic and prokaryotic cells is important for the safetyand health of astronauts during spaceflight. Mechanisms throughwhich cells sense gravity and adapt in its absence address fundamentalquestions regarding the evolution of terrestrial life and generalmechanosensory responses in living organisms. However, spaceflightstudies are technically difficult and expensive. Fortunately,low-shear fluid dynamics (referred to as simulated microgravityor low-shear modeled microgravity [LSMMG]), a key aspect ofthe microgravity environment, can be modeled in specializedground-based bioreactors. One such bioreactor called the high-aspect-ratiovessel (HARV) was used in this study (49). The HARV is a constantlyrotating culture vessel that is typically used for suspensionculture and tissue growth (21, 29, 41). Yeast cells are inoculatedinto the vessel, and all air bubbles are removed. The cellssuspended within the culture medium do not settle but, rather,revolve around a horizontal axis, continuously falling throughthe fluid at 1 x g under terminal velocity conditions. The cellsare not agitated but move sufficiently enough in the HARV toallow for the ample exchange of dissolved gases through a permeablemembrane in the device and the exchange of nutrients and wastesin the medium within the vessel. In addition, the system randomizesthe unidirectional pull of gravity and minimizes turbulence(shear), altering the net effect of gravity on the cells bycreating a state of "functional weightlessness" (21, 29, 41).The bioreactor does not remove the force due to gravity, but,rather, the gravitational vector present in the rotating deviceis averaged over time to near zero (vector-averaged gravity)to simulate low-gravity conditions (29). Thus, the HARV canbe used to examine the physiological behavior of cells in afluid environment under altered inertia, i.e., low shear, toprovide insights into how gravity affects biological systems.HARVs have been used to mimic the effects of the spaceflightenvironment on various cellular functions (22, 33, 40) and tomodel the progression of infectious diseases in low-shear environments(10, 40, 41).

Microgravity and the related low-shear stress environment mayhave a role in the regulation of gene expression, physiology,and pathogenesis (17, 24, 37). Previous studies from both in-flightand ground-based LSMMG experiments described a shortened lagphase (36), a prolonged exponential phase (28), altered secondarymetabolite production profiles (16), and enhanced pathogenesiswith increased resistance to general stress in bacteria suchas Salmonella enterica serovar Typhimurium (40, 42) and Escherichiacoli (33).

Saccharomyces cerevisiae, because of its well-defined geneticsystem, robust viability, and ease of handling, is an idealmodel organism for studying the effects of spaceflight conditionson eukaryotic cells. We previously found that global gene expressionmediated by the Msn2/4p transcriptional activators that acton stress response promoter elements is altered in S. cerevisiaein response to short-term (180 min) exposure to LSMMG (22).However, phenotypic changes associated with the altered geneticresponse to LSMMG have not been described.

The objective of this study was to determine S. cerevisiae phenotypicand underlying genetic responses due to growth in LSMMG. Wehypothesized that altered gene expression in cells grown inLSMMG would result in changes in the yeast phenotypic and functionalcharacteristics that could be identified and measured by conventionalmicroscopic and fluorescent staining techniques. The resultsfrom this study will extend our current knowledge on eukaryoticcell function and behavior in low-shear environments and willprovide a paradigm for understanding the effects of weightlessnessin humans during space explorations.

MATERIALS AND METHODS

Top

Materials and Methods

Strains and growth conditions.
S. cerevisiae BY4743 homozygous diploid MATa (his3 ${Delta}$ 1 leu2 ${Delta}$ 0 met15 ${Delta}$ 0ura3 ${Delta}$ 0)/MAT ${alpha}$ (his3 ${Delta}$ 1 leu2 ${Delta}$ 0 met15 ${Delta}$ 0 ura3 ${Delta}$ 0) strain (Invitrogen, Carlsbad,CA), the parental strain of the yeast strain used to constructthe systematic mutant collection of the Yeast Genome DeletionProject (www.yeastgenome.org), was used throughout this study.Overnight cultures of yeast cells (optical density at 600 nm[OD₆₀₀] of 1.0 to 1.5), previously propagated from a singleyeast colony and incubated at 30°C on a roller drum, wereinoculated (diluted to 1:10,000) into fresh synthetic completemedium (0.67% [wt/vol] yeast nitrogen base without amino acids,0.3% [wt/vol] complete amino acid drop-out mix [1], 2% [wt/vol]glucose)and aseptically loaded into sterile HARVs. The test(LSMMG) and control (1 x g) HARVs were incubated at 30°Cand were rotated either vertically (with the vessel parallelto the gravitational vector) or horizontally (with the vesselperpendicular to the gravitational vector), respectively. Microscopicexaminations for the presence of bacterial or fungal contaminantswere performed routinely to ensure culture purity.

Growth curve generation.
Overnight cultures of yeast cells (OD₆₀₀ of 1.0 to 1.5), previouslypropagated from a single yeast colony and incubated at 30°Con a roller drum, were diluted (1:100) into fresh syntheticcomplete medium, aseptically loaded into control and test HARVs,and simultaneously incubated at 30°C. Cell density was determinedperiodically by measuring the OD₆₀₀. The doubling time (generationtime) was calculated from the data points, which indicated thatcell density increased linearly (between 150 and 460 min) onthe growth curve. The post-diauxic shift (51) was identifiedby using a glucose assay kit (Sigma, St. Louis, MO) to determinethe amount of glucose depleted in the samples.

Cell size.
Cell volume was calculated by measuring cell dimensions (widthand length) and by assuming that the cells were prolate ellipsoidalwith smooth surfaces (43). For overall population cell sizemeasurements, 100 cells consisting of individual budded (n =50) and unbudded (n = 50) cells were examined for each samplepoint. For the daughter cell measurements, 100 buds that wereat least half the size of the mother cell were scored. The overallpopulation cell size data were reported as the means ±standard errors (SEs) calculated from three replicate samplesmeasured three independent times for control or LSMMG samples.

Budding pattern.
BY4743 is a diploid MATa/MAT ${alpha}$ yeast strain with a bipolar buddingpattern with bud scars typically clustering at both poles (Fig.1A, cell 5). We observed cells with bud scars clustered predominantlyat one pole but with at least one scar located at the oppositepole (Fig. 1A, cell 4) or where cells had a space between theindividual bud scars (Fig. 1A, cell 2); we also observed cellswith bud scars that were not located at the tips of the cells(Fig. 1A, cell 3) but were adjacent to a birth scar. Each ofthese geometric locations has been observed in wild-type strainsof S. cerevisiae (11). Thus, these clustering types were classifiedas normal budding populations. We utilized the fluorescent dyecalcofluor white M2R (catalog no. L-7009; Molecular Probes,Carlsbad, CA), which targets the chitin-rich area (i.e., budsite regions) on the yeast cell wall with a high degree of specificity(44), to evaluate the budding patterns of yeast cells. Cellswere pelleted by centrifugation (3,000 x g for 3 min at 23°C),resuspended in 500 µl of distilled H₂O and divided intotwo 100-µl aliquots. One of the aliquots was fixed with3.7% formaldehyde, followed by a 10-min incubation at 30°C.Each aliquot was washed twice with phosphate-buffered saline(8 g/liter NaCl, 0.2 g/liter KCl, 1.44 g/liter Na₂HPO₄, 0.24g/liter KH₂PO₄) and then resuspended in 100 µl of phosphate-bufferedsaline. Calcofluor white was added to a final concentrationof 25 µM and visualized by epifluorescence microscopyas described by Pringle (44). Cells with medial budding siteswere scored as random budding. For each sample, at least 200cells with at least three bud scars were counted from at leastfour separate microscopic fields. The percentage of random buddingcells was determined for each sample. The data are reportedas the means ± SEs calculated from three replicate samplestaken at three different times.

View larger version (31K):
[in this window]
[in a new window]
FIG. 1. (A) Epifluorescent micrographs of budding cells with normal (1 to 5) and random (6 to 10) budding patterns stained with calcofluor white. Thin white arrows on cell 6 demonstrate chitin-rich bud scarring (bottom), and the relatively darker protruded region indicates chitin-poor birth scarring (top). Thick white arrows indicate random budding patterns as revealed by chitin staining. The scale bar is 15 µm. (B) Percentage of random budding cells at different growth stages in LSMMG ( ${blacksquare}$ ) and in 1 x g ( ${square}$ ). Each point on the graph is a mean ± SE from at least three samples from three independent experiments. At least 200 cells were scored for each replicate experiment.

Determination of cell viability.
FUN1 yeast viability stain (catalog no. F-7030; Molecular Probes)was used, according to the manufacturer's instructions, to determinethe metabolic activity of cells and to differentiate betweenlive and dead cells. For each sample, the cell density was adjustedto 1 x 10⁸ cells/ml, and FUN1 was added to a final concentrationof 20 µM. The samples were examined with a confocal scanninglaser microscope (Leica Microsystems, Wetzlar, Germany). Cellswith clear fluorescent intravacuolar structures were scoredas live or metabolically active. Cells that lacked these structuresand that had diffuse green or yellow fluorescence were countedas dead or metabolically inactive. At least 500 cells from fiveseparate microscopic fields were counted for each sample. Resultsare expressed as the mean ± SE of three replicate samplesmeasured three independent times for each control and LSMMG.

Cellular aggregate measurements.
For mother-daughter cell separation studies, the percentageof cells connected to five or more individuals was counted ina hemacytometer using a bright-field microscope. At least 300cells were scored from 10 separate microscopic fields. Similarly,the population of cells in aggregates and cells that were connectedby chitin were counted in samples treated with 3.7% formaldehydeand stained with Calcofluor White. Results are reported as themean ± SE calculated from three replicate samples measuredthree independent times for control or LSMMG samples.

Culturing and RNA preparation.
During logarithmic (OD₆₀₀ of 1.0) and stationary (OD₆₀₀ of 2.2)growth phases, the LSMMG and control cells were harvested bycentrifugation (5,000 x g for 5 min at 24°C), and the pelletswere stored at –80°C until used. RNA was isolatedfrom the yeast pellets via the hot phenol-glass bead method(3). The quality of the recovered RNA was assessed by gel electrophoresis,and the yield was determined by measuring the OD₂₆₀.

Real-time RT-PCR.
RNA samples were treated immediately prior to reverse transcription-PCR(RT-PCR) set up with DNase I (catalog no. 18068-015; Invitrogen)to remove residual DNA contamination. A quantitative RT-PCR(QRT-PCR) analysis was performed by using the SuperScript One-StepQRT-PCR System with Platinum Taq Polymerase (Invitrogen). QRT-PCRswere performed in a reaction volume of 25 µl containinga 0.2 mM concentration of each deoxynucleotide triphosphate,0.2 µM (each) sense and antisense primers, 0.3 µlof RT-Taq mixture, and an appropriate amount of poly(A) RNA.The 20- to 25-bp primers were designed with primer3 software(http://frodo.wi.mit.edu/cgi-bin/primer3/primer3) with ampliconsizes ranging from 100 to 150 bp (Table 1). Fragments were amplifiedby incubation at 50°C for 30 min (reverse transcription);94°C for 2 min (pretreatment);and 40 cycles of 94°Cfor 30 s, 55°C for 30 s, 72°C for 90 s, followed bya final extension at 72°C for 5 min. QRT-PCR was performedon RNA samples from three independent replicate experimentsfor each data point, and the relative expression levels werequantified by the 2^–^CT method (5). Briefly, the thresholdvalue (C_T) value of PDA1 (encodes the El ${alpha}$ subunit of the pyruvatedehydrogenase complex) was subtracted from that of the geneof interest to obtain a ${Delta}$ C_T value. The ${Delta}$ C_T value of a controlsample (i.e., sample at 1 x g) was subtracted from the ${Delta}$ C_T valueof an LSMMG sample to obtain a ${Delta}$ ${Delta}$ C_T value [ ${Delta}$ ${Delta}$ C_T = ${Delta}$ C_T _(LSMMG) – ${Delta}$ C_T₍_control₎]. The gene expression level in LSMMG for each primerset relative to the control was expressed as 2^–^CT andwas reported as the relative difference in expression (n-fold)of cells in LSMMG versus the control. We chose PDA1 as an internalstandard for normalization because it is a more suitable internalstandard for QRT-PCRs for both exponentially growing and stationarycells than the commonly used ACT1 gene (57). We confirmed thatPDA1 gene expression was not affected by LSMMG by comparingits expression level to that of GRS1, a gene previously usedas a standard for gene expression analysis in LSMMG (22). Wefound no statistical difference between the expression levelsof GRS1 and PDA1 in four independent experimental samples (datanot shown).

View this table:
[in this window]
[in a new window]
TABLE 1. List of primer sets used for QRT-PCR quantification

Gene selections for QRT-PCR.
For gene expression studies we chose eight genes involved inthe budding and cell separation processes (Table 2). Genes RAX1and RAX2 encode functionally characterized budding proteinsthat regulate the establishment and/or maintenance of the corticallandmarks specific for bipolar budding (20, 25). We chose BUD5because it is involved in the establishment of polarity duringbudding and because it is required for both axial and bipolarbudding patterns (8, 26, 27). BUD25 also typifies the 22 genesidentified by Ni and Snyder (39) that have a role in determiningthe bipolar budding phenotype. Deletion of any of these fourgenes results in cell clumps in diploid yeasts (15). Genes DSE1,DSE2, DSE4, and EGT2 encode glucanases and are responsible forcell separation cleavage at the mother-daughter cell juncture.Deletion of any of these four genes results in cell clumps indiploid yeasts (15).

View this table:
[in this window]
[in a new window]
TABLE 2. Relative differences in expression of genes involved in random budding and a cell-cell separation phenotype for cultures grown in LSMMG

Statistical analysis.
Statistical comparisons were made with a one-way analysis ofvariance test by using Minitab (version 13.3; Minitab, Inc.,State College, Pa.) software, and the differences were reportedas significant for P values of $≤$ 0.05.

RESULTS

Top

Results

Growth kinetics in LSMMG.
There were no statistically significant differences betweenthe growth rates in control and LSMMG cultures, with a doublingtime of 95 ± 5 min for both conditions (Fig. 2A). However,the lag phase of cells grown in LSMMG was shorter than thatof the control cells by approximately 88 ± 7 min (Fig.2A) with a slightly higher final density (P < 0.05). Fromthe growth curve, we determined that an OD₆₀₀ of 0.5 correspondedto exponential growth, an OD₆₀₀ of 1 corresponded to late-exponentialgrowth, an OD₆₀₀ of 1.5 corresponded to early-stationary growth,and an OD₆₀₀ of 2.2 corresponded to late-stationary growth.

View larger version (17K):
[in this window]
[in a new window]
FIG. 2. (A) S. cerevisiae growth kinetics. The percentage of metabolically inactive cells counted in FUN1-stained cells (B) and volume of individual cells (µm³) grown in LSMMG ( ${blacksquare}$ ) or 1 x g ( ${square}$ ) at different stages of the growth cycle (C). Each data point on panel A is a mean ± SE of three samples from three independent replicate experiments, with at least 500 and 100 individual measurements for each replicate for panels B and C, respectively.

Cell viability in LSMMG.
Both control and LSMMG samples were examined at logarithmic,late-logarithmic, early-stationary, and late-stationary growthphases. The mean population of dead cells in the microgravitycondition was higher than in control cells during the late stagesof growth (Fig. 2B), but these differences were not statisticallysignificant (P = 0.05).

The effect of LSMMG on cell size.
There was no significant difference in the size of individualcells over the course of growth under control and LSMMG conditions(Fig. 2C). Excessive apical or isotropic growth can cause abnormalmorphology characterized by either highly elongated or sphericalbuds, respectively (45). If growth is completely undirected,then bud enlargement ceases, and large, round unbudded mothercells are observed (45). Based on the cell dimensions, therewere no such irregularities in overall cell shape or size (Fig.2C) or budding populations (data not shown) under either LSMMGor control conditions.

Effect of LSMMG on budding pattern.
We observed cells with an atypical, nonpolar pattern (Fig. 1,cells 6 to 10) that we classified as a random budding population(the criteria used for normal budding population are describedin Materials and Methods). Under LSMMG conditions, the percentageof the random budding cells was more than double that foundin the control cells during the early-stationary and late-stationaryphases (Fig. 1B).

Effect of LSMMG on cell clumping.
Cell samples from LSMMG cultures had more clumps of five ormore yeast cells (Fig. 3B) than did the control samples (Fig.3A). The differences between control and LSMMG cells were mostobvious in the late-logarithmic growth phase, during which $~$ 5%of the LSMMG cell population was found in multicellular aggregates,whereas no such clumps were observed in the control cultures(Fig. 3C). During early-stationary growth, there were 10 timesas many clumps in the LSMMG cultures as in the control culture,with the percentage of the cellular aggregates decreasing from15% to 5% during the late-stationary phase. These cellular aggregatescontained metabolically active live cells (not shown) with anormal distribution of chitin, based on calcofluor white staining(not shown).

View larger version (66K):
[in this window]
[in a new window]
FIG. 3. Increased frequency of clumped cells was observed when cells were grown in LSMMG compared to when they were grown under the control condition 1 x g. Transmitted images of control (1 x g) cells (A) and LSMMG cells (B) during the early-stationary phase (OD₆₀₀ of 1.5) are shown. A representative cellular clump consisting of $≥$ 5 individual cells observed in LSMMG is indicated by a large white arrow, and two individual clumps consisting of five individual cells are indicated by two smaller arrows (B). Scale bar, 35 µm (A and B). (C) The percentage of cells that are found in the multiple cell arrangement ( $≥$ 5) in LSMMG ( ${blacksquare}$ ) and 1 x g ( ${square}$ ) at different growth stages. Each point on the graph is a mean ± SE from at least three independent experiments with at least 300 cell counts for each replicate.

Following treatment with 3.7% formaldehyde, the cellular aggregatesbroke down into chains composed primarily of 3 to 4 cells (Fig.4A). Quantitatively, the differences in the percentage of aggregatedcells were the greatest between the two gravity conditions duringearly-stationary phase growth, with 29.5 ± 4.9% and 10.9± 1.6% in LSMMG and the control cultures, respectively,followed by a decrease to 11 ± 0.5% and 4 ± 2%in these cultures, respectively (Fig. 4B). Of the aggregatedcells, 60 ± 10% of the cells in LSMMG cultures and 40± 12% of the cells in the control cultures were connectedvia chitin. Daughter cells could be linked to either opposite(Fig. 4A, cell 1) or the same (Fig. 4A, cells 3 and 4) polesof the mother cells or to random positions on mother cells (Fig.4A, cell no. 2 and 5).

View larger version (28K):
[in this window]
[in a new window]
FIG. 4. (A) Representative images of 3.7% formaldehyde-treated and calcofluor-stained cell clumps that are linked through chitin. Scale bar, 15 µm. (B) The percentage of 3.7% formaldehyde-treated cells that were in arrangements of 3 to 4 cells at different growth stages.

Gene expression analysis.
We examined four genes related to budding pattern—RAX1,RAX2, BUD5, and BUD25—and four genes that are responsiblefor the cell separation—DSE1, DSE2, DSE4, and EGT2. Expressionlevels of all four genes associated with budding (BUD5, BUD25,RAX1, and RAX2) and three of the four genes associated withcell separation (DSE1, DSE2, and EGT2) were significantly affectedby growth in LSMMG (Table 2). Seven of eight genes displayedchanges in expression of close to twofold or more. Expressionof BUD5 and BUD25 under LSMMG conditions was consistently higherthan the control, and expression of RAX1, RAX2, DSE1, DSE2,and EGT2 under LSMMG conditions were no more than half thatobserved in the controls.

DISCUSSION

Top

Discussion

The ability of laboratory-based bioreactors, e.g., HARV, tomimic the low-shear conditions present during spaceflight providesan inexpensive and rapid means to study the effects of low-shearenvironments on microbial cell function. In terms of growthkinetics, we found no difference in the doubling time but ashorter lag phase (by approximately 90 min) and a slightly higherconcentration yield relative to the control cells. These resultsand the random budding data are consistent with most of thespaceflight studies performed on microorganisms (28, 36), suggestingthat HARV cultivation provides a time- and cost-effective wayto study the effects of microgravity and low-shear-related environmentson microbial physiology and morphology. McPherson et al. (34,35) postulated that the altered microbial growth kinetics inspace resulted from anomalous mass transfer processes due tothe low fluid shear environment of weightlessness. They suggestedthat the increased accumulation of cellular by-products in themicroenvironment surrounding the cells may trigger a rapid responsefor growth that results in a shorter lag phase and that thehigher final cell density may result simply from a more evendistribution of cells in the low-shear microgravity of space(34, 35). In this study we found, however, that the overallmetabolic activity of cells in LSMMG and control cells was similarthroughout the experiment (Fig. 2B). Thus, differences in overallmetabolic activity are not sufficient to explain the phenotypicand genetic changes we observed.

The frequency of aberrant budding patterns in cells growingunder LSMMG conditions (LSMMG, 20 ± 2.5%; control, 8.3± 0.5%) increased even though there were no significantchanges in cell volume, shape, or metabolic activity. Theseobservations are consistent with those from previous in-flightexperiments in which there were more randomly distributed budscars in the experimental (17%) than in the ground-based control(5%) cells but no differences in growth rate or volume (56).The abnormal distribution of bud scars may be an indirect resultof alterations in cytoskeletal organization (19), such as thoseobserved during spaceflight experiments with human epidermoidcells (46) and lymphocytes (14). However, the authors of theprevious studies did not determine if the abnormalities in thecytoskeleton of these cell types occurred because of the hypergravityexperienced during the launch of the space shuttle or becauseof their subsequent exposure to microgravity.

We observed no abnormalities in the volume or shape of cellsgrown in LSMMG compared to the controls, which is consistentwith normal polarized growth. Thus, we hypothesize that therandom budding pattern in LSMMG cells may result from changesin the expression of genes with roles in upstream budding processessuch as bud site selection and polarity establishment. Therewere significant differences in expression levels of all genesinvolved in polarity establishment (BUD5) and bipolar buddingphenotype (RAX1, RAX2, and BUD25) (Table 2). In a study by Niand Snyder (39), a deletion of a previously uncharacterizedgene, which was termed BUD25, resulted in a random budding phenotype.In our study, enhanced expression of BUD25 in LSMMG comparedto the control initially appears to be counterintuitive. Itis not, however, unusual for a gene deletion or overexpressionto result in a similar phenotype in the yeast as random buddingphenotype can occur in both BUD5 deletion (www.yeastgenome.com)and overexpression (27) mutants. RAX1 and RAX2 encode transmembraneproteins that mark nascent budding sites in a bipolar-specificmanner (13, 20, 25, 59). Significant reduction in the expressionof both RAX1 and RAX2 in experimental cells is consistent withthe aberrant budding phenotype. Identification of the upstreamregulators of RAX1 and RAX2 might provide important clues tothe signaling pathways involved in the LSMMG response.

BUD5 modulates the activity of a small GTPase, Cdc42, and playsan important role in the establishment of cell polarity duringbudding (27). BUD5 expression increased fourfold relative tothe control during exponential growth in LSMMG (Table 2). Ectopicoverexpression of BUD5 in yeast also results in a random buddingpattern (27). Deletion of BUD5 stops pseudohyphal developmentin diploid yeast cells (32), perhaps due to the lack of thepolarized division required for cell filamentation.

In addition to a random budding phenotype, cells grown in LSMMGwere also more likely to be found in aggregates containing fiveor more cells (Fig. 3B). These multicellular aggregates weremetabolically active live cells (not shown) with normal chitindistributions (not shown). We observed the clumps only in experimentalsamples prior to 3.7% formaldehyde fixation, which reduces thenumber of cells per aggregate to 3 to 4 (Fig. 4A and B). Thisphenotypic change in S. cerevisiae has not been reported previously(56), perhaps due to the limitations associated with spaceflightexperiments in which the extensive use of fixatives is requiredand longer incubation times occur than are necessary for ground-basedanalyses.

Most cell clumping phenotypes in yeast result from defects inthe mother-daughter cell separation process due to mutationsin enzymes such as chitinases and glucanases (4, 15, 18, 54).We found no statistical difference in the number of cells linkedthrough chitin when the LSMMG and control cell populations werecompared, suggesting that normal chitinase activity occurredin the experimental cells. Thus, we followed the expressionlevels of the genes that are involved in the production of glucanasesthat function at the mother-daughter cell juncture (15).

We found significantly reduced expression of three of the fourgenes tested (EGT2, DSE1, and DSE2) (Table 2). DSE2 has a stressresponse promoter element motif for the binding of Msn2/4p transcriptionalactivators (52) that are responsive to various environmentalstresses (48, 53) including LSMMG (22). DSE1, DSE2, and EGT2are all part of a daughter-specific expression program (15)and are dependent on the fidelity of the cell polarization process(2, 12). Thus, the LSMMG-induced defect in cell polarity couldbe the underlying mechanism for both the random bud scarringand the formation of the large cellular aggregates we observed.

Polarization is a fundamental property of living cells and occursin both yeast and higher eukaryotes, where it is mediated viahighly conserved signaling pathways such as those regulatedby Cdc42 (12, 23, 47). Cell functions involving Cdc42, or itshomologs, include morphogenesis (7, 58) and virulence (31, 55)in the opportunistic pathogen Candida albicans (30); efficiencyand fidelity of the immune response in mammals (9); and cellfate in embryonic (6) and neuronal (38) development. Thus, wespeculate that two recurring problems encountered in space flight,i.e., a weakened immune response in astronauts (50) and enhancedvirulence of microbial pathogens (41), may be linked throughLSMMG-induced effects on a common, evolutionary conserved pathwaysuch as cell polarization.

ACKNOWLEDGMENTS

This work was supported by NASA grant NAG9-1559. We thank PaulStoodley for financial support (RO1 GM60052) for the use ofthe Image Analysis Facility at the Center for Biofilm Engineering,Montana State University.

We thank Laura Richert and Miranda Orr for technical assistance,Kate McInnerney for assistance in analysis of gene expression,William Baricos for critically reviewing the manuscript, andBarry Pyle for experimental discussions.

FOOTNOTES

* Corresponding author. Mailing address: Division of Health Sciences, WWAMI Medical Program, Montana State University, 308 Leon Johnson Hall, P.O. Box 173080, Bozeman, MT 59717. Phone: (406) 994-4411. Fax: (406) 994-4398. E-mail: lhyman{at}montana.edu.

REFERENCES

Top