JOURNAL OF BACTERIOLOGY, July 1982, p. 429-4370021-9193/82/070429-09$02.00/0

Vol. 151, No. 1

Selective Transport of Nutrients via the Rhizoids of the WaterMold Blastocladiella emersoniiDARRYL L. KROPF* AND FRANKLIN M. HAROLD

Department of Molecular and Cellular Biology, National Jewish Hospital and Research Center, Denver,Colorado 80206, and Department ofBiochemistry, Biophysics and Genetics, University of Colorado School of

Medicine, Denver, Colorado 80262

Received 6 October 1981/Accepted 24 February 1982

Previous work in this laboratory demonstrated that the rhizoids of Blastocla-diella emersonii grow chemotropically toward a source of P1 and thus providedpreliminary evidence that, in addition to serving as a holdfast, the rhizoids absorbnutrients. To further examine the role of the rhizoids in nutrient uptake, wedevised a technique to introduce a barrier between the rhizoids and the thallus sothat these cell compartments could be studied independently. Cells were grown onpolycarbonate membrane filters in such a way that all of the thalli were on one sideof the filter and essentially all of the rhizoids were on the opposite side. Nutrientuptake into the rhizoids and the thallus was measured by floating the filtersbearing cells on radioactive medium so that only one side of the filter contactedthe label. Mineral oil was used to block the diffusion of the label through theunfilled pores in the filter. This technique permitted us to establish clearly that therhizoids absorb all seven of the nutrients tested. In addition, we found that somenutrients, specifically Pi and amino acids, appeared to be preferentially taken upvia the rhizoids, whereas K+, Rb+, and Ca + entered the thallus and rhizoidsequally. Cells grown in the presence of the microtubule synthesis inhibitorsnocodazole and carbendazim elaborated only a stunted rhizoid system, so weexamined their ability to accumulate the two classes of compounds. As expected,these cells were severely inhibited in Pi and amino acid uptake but retained normaluptake of K+, Rb+, and Ca2+.

Most eucaryotic cells are inherently polar-ized: appendages, internal organelles, and phys-iological functions occupy defined positions andconfer directionality upon the whole cell. Myce-lial fungi supply a striking example, for thehyphae that comprise the mycelium elongateonly at their tips. The materials required forgrowth, such as precursors of new plasma mem-brane and cell wall, are transported unidirection-ally down the hyphae to the terminal growthzones. We have chosen to investigate an aspectof functional polarity in the water mold Blasto-cladiella emersonii. This large, unicellular orga-nism consists of two distinct compartments, athallus and a rhizoid system, that can be studiedseparately in hope of elucidating their special-ized functions.

B. emersonii is classified as a chytridiomyceteand, like other members of its order, begins itslife cycle as a small, posteriorly flagellated zoo-spore (12). When placed in growth medium, thezoospore rounds up, retracts its flagellum, andgerminates. The most striking feature of germi-nation is the outgrowth of a germ tube that givesrise to the anucleate rhizoid system; the sporebody is destined to become the coenocytic thal-

lus. The cytoplasm is continuous throughout thecell and lacks obvious barriers such as septae.When the vegetative cell reaches a limiting sizeor is starved for nutrients, sporulation ensues;each nucleus and the cytoplasm surrounding itare reorganized into a zoospore.Our investigations focused on the localized

uptake of nutrients by the rhizoids. The sugges-tion that Blastocladiella rhizoids are capable ofnutrient uptake arose from studies on chemotro-pism conducted in this laboratory. Harold andHarold (7) showed that in phosphate-deficientgrowth medium, rhizoids grow chemotropicallytoward any source of Pi (see Fig. 1). Cells thatrespond to the gradient grow much larger thando phosphate-starved cells further from the Pisource; presumably, this growth stimulation re-sults from the ability of the rhizoids to absorb Pi.These findings encouraged us to devise a

technique to demonstrate that rhizoids are in-deed capable of transporting Pi and other nutri-ents. By growing the cells across a polycarbon-ate membrane filter so that the thalli were on oneside of the filter and nearly all of the rhizoidswere on the opposite side, we were able to showthat the rhizoids do take up nutrients and that Pi

429

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

430 KROPF AND HAROLD

and some amino acids are apparently taken uppreferentially through the rhizoid system. Bycontrast, K+ and Ca2+ enter the rhizoids andthallus equally. Experiments utilizing the antitu-bulin drugs nocodazole and carbendazim dem-onstrated that cells grown in the presence ofthese drugs possessed a very stunted rhizoidsystem (7). Concomitantly, they lost the abilityto transport the nutrients that preferentially en-ter the rhizoid, but retained the capacity totransport K+ and Ca2+.

MATERIALS AND METHODS

Organism and media. Previous publications havedescribed the maintenance of B. emersonii L-17 asresistant sporangia, the harvesting of zoospores fromsporangia on peptone-yeast extract-glucose plates,and the growth of the organism on a defined medium(DM2 medium) (7, 11, 21, 25). The concentrations ofnutrients in modified DM2 medium (17) were furtheradjusted as follows: amino acids-2.72 mM glutamicacid, 435 p.M methionine, 400 p.M each leucine,isoleucine, lysine, threonine, and valine, 200 p.M eacharginine, glycine, phenylalanine, serine, and tyrosine,100 p.M histidine, and 40 p.M tryptophan; salts-10mM MgSO4-7H20, 1.0 mM CaC12, 5.0 mM NH4NO3,10 mM NaCl, and 1 mM Na2HPO4; and trace metals-4.0 ptM FeSO4 7H20, 0.6 p.M CuSO4-5H20, 1.2 p.MZnSO4-7H20, 1.3 pFM MnSO4 H20, and 0.12 p.Mthiamine-HCI. The medium was made up in 1.0 mMTris-maleate buffer. The pH was adjusted to 6.8 withKOH, after which KCI was added to a total of 25 mMK+. Finally, 10 mM glucose was added. In manyexperiments the medium was made up to contain areduced concentration of a particular nutrient, asindicated by a subscript; for example, DM2P25 standsfor DM2 medium containing 25 p.M Pi.

Chemotropism. Granules of insoluble calcium phos-phate (hydroxylapatite) were sterilized in water, and afew drops of this slurry were added to 10 ml of sterile0.2 M poly-L-lysine, which was then poured into a 100-mm tissue culture dish (Falcon 3003). The slurry wasswirled and poured off, and the dish was allowed todry. DM2Po (DM2 medium lacking added phosphate;10 ml) was added to the dish and inoculated with either1.8 x 105 or 3.6 x 105 zoospores. The dishes wereincubated overnight at 24°C without agitation. Thisentire procedure was performed aseptically.

Oriented growth on membrane filters: the floatinggarden. To interpose a barrier between thalli and rhi-zoids, cells were grown through the pores in mem-brane filters. Again, sterile conditions were main-tained throughout the procedure.An inoculator was constructed by drilling 40 wells,

25 mm in diameter and 10 mm deep, in a sheet ofPlexiglas. Short lengths of aluminum tubing were cutfrom a shower curtain rod (25-mm diameter) andmachined until they fit snugly into the wells in thePlexiglas. A polycarbonate membrane filter (Nucle-pore Corp.; 25-mm diameter, 5.0-p.m pores, 10-p.mthickness) was placed in the bottom of each well andheld in place with an aluminum stack.A concentrated inoculum of fresh zoospores was

diluted to 5 X 104 zoospores per ml with DM2P25. A

portion (2 ml) of this suspension was immediatelypipetted onto the top of each of the 40 filters. Theinoculator was placed in the incubator at 24°C for 2 h,during which time the zoospores settled, germinated,and attached to the filters. The medium was thenaspirated, the stacks were removed, and the filterswere floated (germlings up) on the surface of DM2P25(90 ml in a 150-mm petri dish). Ten filters were placedin each dish. These polycarbonate filters float on thesurface with only a very thin layer of medium abovethe filter, hence the term "floating garden." Thedishes were incubated overnight (14 to 16 h) at 24°C,permitting the germlings on the upper surface of thefilters to grow rhizoids down through the pores intothe medium. The thalli remained above the plane ofthe filter.

Diffusion measurements. As the uptake of nutrientsinto rhizoids or thalli was measured by exposing oneside of the filter or the other to medium containingradioactive label, it was necessary to prevent the labelfrom diffusing through the pores to the opposite side.A thin layer of mineral oil was shown to block diffu-sion.A hole (1-cm diameter) was cut in the bottom of a

35-mm tissue culture dish (Falcon 2002), and a filterbearing cells was glued (General Electric silicone glueand sealant) to the bottom of the dish so that the holewas covered and the rhizoids dangled beneath thedish. A second 35-mm dish was filled to the brim withradioactive medium, and 40 p.l of mineral oil (Nujol)was pipetted onto the surface; the oil spreads rapidlyover the medium, forming a film approximately 40 p.mthick. The first dish with the attached filter was placedon top of the second dish so that the bottom of thefilter came into contact with the oil layer, creating ajunction free of air bubbles. To begin the experiment,0.5 ml of nonradioactive medium was pipetted onto theexposed upper surface of the filter (1-cm diameter).Samples were taken by withdrawing 10 p.1 of themedium above the filter and counting the radioactiv-ity. Any label found above the filter had to diffusethrough the pores.A similar procedure was used to measure diffusion

through the filter when the thalli were placed downinto the radioactive medium. However, as all measure-ments of nutrient uptake into the thallus were made byspreading the 40 p.1 of mineral oil directly onto therhizoids (see below), diffusion through the filter had tobe measured in an analogous fashion. The mineral oilwas spread onto the rhizoids, and the filter wasattached to the diffusion apparatus in the thallus-downorientation; 0.5 ml of nonradioactive medium wasplaced above the oil layer to initiate the experiment.Again, 10-p.l samples were withdrawn from the medi-um in the upper chamber and counted.

Nutrient uptake. By taking advantage of the abilityof a mineral oil layer to block diffusion through thefilter, we were able to measure uptake into eitherrhizoids or thalli. In all filter assays, the nutrient ofinterest was present in limiting concentrations. Uptakeinto the rhizoids was measured by dangling thesefilaments down through an oil layer into radioactivemedium. A filter bearing vegetative cells was liftedfrom the floating garden and placed, rhizoids down, ontissue paper for 10 s to absorb the medium trapped inthe rhizoid mass. The filter was then floated, rhizoidsdown, on a film of mineral oil approximately 40 p.m

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

NUTRIENT UPTAKE VIA RHIZOIDS 431

thick (40 p.l of oil) that was spread over the surface of3.0 ml of radioactive medium in a 35-mm tissue culturedish. Rhizoids are approximately 150 to 200 p.m long,so we assume that they extended down through themineral oil into the radioactive medium. In addition topreventing the label from contacting the thalli, themineral oil probably caused a slight underestimation ofuptake into the rhizoids, as a 40-p.m length of eachrhizoid was immersed in oil and out of contact with themedium. Each filter was washed twice with nonradio-active medium on a Millipore filtration apparatus, andthe radioactivity was counted.Attempts to measure uptake into the thallus in the

same fashion did not give satisfactory results. Thalliwere typically 50 to 80 p.m long, so less than half thesurface could penetrate the oil layer to contact theradioactive medium below. A thinner oil layer allowedsome label to diffuse across the filter, so we dismissedthis approach. To circumvent this problem, 40 p.1 ofmineral oil was placed directly onto the rhizoids andgently spread over the filter with a rubber policeman;the filter was immediately floated, thallus down, onradioactive medium. The oil on the upper surfaceprevented label from diffusing into the rhizoid mass.However, any rhizoids within the pores came intocontact with the medium, resulting in some overesti-mation of uptake into the thallus.The possibility that the cells were damaged by

placing the filters on tissue paper to absorb the medi-um from the rhizoid mass or by directly spreadingmineral oil onto the rhizoids was examined. To dem-onstrate that placing the filters, rhizoids down, ontissue paper did not disrupt these filaments, we mea-sured the uptake of Pi into cells with and withoutmedium absorption. Filters bearing cells either wereremoved from the gardens and floated, rhizoids down,on DM2P25 containing 32p, or were removed from thegardens, placed, rhizoids down, on tissue paper for 10s, and then floated, rhizoids down, on the radioactiveDM2P25. Painting the rhizoids with a rubber policemanwas also shown not to harm the cells. Control filterswere simply placed on tissue paper for 10 s andfloated, thallus down, on the radioactive DM2P25. Thetest filters were mock painted with DM2P25 as follows:after the medium from the rhizoid mass was absorbed,40 p.l of DM2P25 was spread over these filaments witha rubber policeman; this medium was then absorbed asusual, and the filter was floated, thallus down, on theradioactive medium.Uptake into the entire cell was also measured to

confirm that it equalled the sum of the uptake intorhizoids and thalli. This measurement was made bysubmerging an entire filter in 3 ml of radioactivemedium in a 35-mm tissue culture dish.

Control experiments were conducted to measure thebinding of radioactive tracers to the filters. Filterswere inoculated with DM2P25 lacking zoospores andwere incubated overnight in floating gardens. Theseblank filters were subjected to the radioactive uptakeassays exactly as were the experimental filters bearingcells; the time course of filter binding was followed for12 min under conditions analogous to rhizoids-down,thallus-down, and whole-cell assays. Regardless of thetechnique used to measure nutrient uptake, the filterbinding of all radioactive tracers except 45Ca2+ wasnegligible for the duration of the assays. 45Ca2+ bind-ing equalled approximately 20% of the label taken up

by the cells, and this background was subtracted fromall 45Ca2" uptake measurements.

Microtubule inhibitor studies. A liquid culture wasstarted in a Spinner flask containing 500 ml of DM2medium inoculated with 4 x 104 zoospores per ml. Allglassware was pretreated with silane (Prosil-28). Thisculture was grown for 6 h at 24°C with continuousaeration, after which three 20-ml aliquots were pipet-ted into 125-ml Erlenmeyer flasks. Nocodazole andcarbendazim were added in 20 p.1 of dimethyl sulfoxide(Me2SO) to give final concentrations of 5 and 10 p.M,respectively; controls received 20 p.1 of Me2SO. Thecultures were then grown for an additional 12 h on areciprocating shaker. Cells were collected on Sargent-Welch filter paper (grade S32915G), washed twice, andresuspended in 20 ml of fresh medium containing areduced concentration of the nutrient of interest. Anti-tubulins and Me2SO were again added to the cultures,and samples were taken for protein measurement. At 1h after resuspension, the radioactive tracer was addedto each culture, and 1.0-ml samples were filtered on0.45-p.m membrane filters (Millipore Corp.). The fil-ters were soaked in a solution of the substrate beforeuse to reduce background binding. Each filter waswashed once with 4 ml of nonradioactive medium, andthe radioactivity was counted.

Protein assay. Cells were collected on Sargent-Welch paper, washed, and extracted with 0.1 N NaOHin a boiling-water bath. The extracts were acidified,and protein was determined by the Coomassie bluemethod of Bradford (4).

Materials. Reagents were purchased from standardsuppliers, often Sigma Chemical Co. Nocodazole andcytochalasins were purchased from Aldrich ChemicalCo., and carbendazim was a gift from H. C. Van derPlas, Landbouwhogeschool, Wageningen, The Neth-erlands; inhibitors were dissolved in Me2SO. Radioac-tive compounds were purchased from New EnglandNuclear Corp.

RESULTSChemotropism. Our first experiments were

aimed at confirming the chemotropic growth ofrhizoids toward a source of Pi. Cells grown inthe absence of Pi were able to detect a Pigradient emanating from hydroxylapatite gran-ules and to orient rhizoid growth (Fig. 1). Thecells with chemotropic rhizoids were significant-ly larger than cells further from the granules,strongly suggesting that rhizoids take up Pi.Furthermore, some rhizoids that initially grewaway from the granules were able to redirecttheir growth so that the filament bent and begangrowing toward the granule. As expected, thepresence of sufficient Pi in the medium blockedrhizoid chemotropism. In control experiments,neither rhizoids nor thalli grew chemotropicallytoward the granules when these were made toserve as the sole source of Ca2+ in calcium-limited medium.

Floating gardens. To prove that rhizoids takeup Pi, it was necessary to separate the rhizoidsfunctionally from the thalli so that each structurecould be studied independently. Severing the

VOL . 1 51, 1982

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

432 KROPF AND HAROLD

80

E 60S

E20

20L

FIG. 1. Chemotropic growth of rhizoids toward a

source of Pi. The calcium phosphate granule in themiddle of the field was the only source of Pi for cellsgrown in DM2Po. The background spots are poly-L-lysine attached to the tissue culture dish. The thick,elongated rhizoids grew directly toward the granule.Bar, 100 FLm.

FIG. 2. Oriented growth of B. emersonii on poly-carbonate membrane filters. Cells were grown as

described in the text. This edge-on view of a filter (10,um thick) shows that all the thalli remained above thefilter, whereas all the rhizoids grew through the poresinto the medium below the filter. Bar, 100 p.m.

Time (min) Time (min)

FIG. 3. Diffusion of radionuclides through filterpores in the presence and absence of mineral oil.Filters bearing cells were placed, rhizoids down, on

the surface of radioactive DM2 medium (0) or on a

layer of mineral oil (40 ,um deep) spread over thesurface of the radioactive medium (0). Samples (10 [lIeach) were taken from the DM2 medium above thefilter (initially nonradioactive). (A) 32Pi diffusion. TheDM2 medium below the filter contained 50 nCi of 32p,per ml. (B) 42K' diffusion. The DM2 medium belowthe filter contained 35 nCi of 42K' per ml.

cells near the thallus-rhizoid junction proved tobe lethal, so we introduced a physical barrierbetween the two halves of the cell, much asRobinson and Jaffe did with germinating Pelve-tia eggs (16). When a filter bearing 2-h-oldgermlings attached to one surface was floated onPi-limited DM2 medium, the rhizoids grew downthrough the pores into the medium below, leav-ing the thalli on the upper surface; thalli are toolarge to fit through the pores (Fig. 2). Therhizoids presumably grew down into the mediumto procure nutrients.

Since the pores in the filter permitted thediffusion of material from one side of the filter tothe other, we employed a mineral oil layer tomake the barrier between the two halves of thecells more complete. A film of mineral oil only40 ,um thick blocked the diffusion of radionu-clides through the pores in the filter. Figure 3shows the effect of the mineral oil layer on thediffusion of 42K' and 32p,. When a filter bearingoriented, vegetative cells was placed, rhizoidsdown, on top of radioactive medium, the labeldiffused through the pores and was recoveredabove the filter in significant amounts after 10min. The presence of the thin film of mineral oilbetween the medium and the filter reduced thediffusion of both radionuclides to nearly unde-tectable levels. Similar results were obtainedwhen the mineral oil was spread directly ontothe rhizoids and diffusion was measured in thethallus-down orientation.

Nutrient uptake. By use of mineral oil as a

diffusion barrier, the ability of the rhizoids to

A: 32Pi B:42K

10 2

)o0/

0oI

00.

0 5 10 15 20 2'5 0 5 10 15 2'0 25 30

J . BACTERIOL .

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

NUTRIENT UPTAKE VIA RHIZOIDS 433

concentrate nutrients was clearly demonstrated(Fig. 4). Potassium, Pi, and glutamic acid weretaken up by the rhizoids at constant rates for atleast 10 min. At most, 10% of the label taken upby the rhizoids could be attributed to diffusionthrough the filter followed by absorption bythalli. This figure was determined by estimatingthe maximum amount of substrate available tothe thalli by diffusion through the filter (Fig. 3).Note that, although the thalli also accumulat-

ed all three nutrients, both Pi and glutamic acidapparently entered the cell preferentiallythrough the rhizoid system. Potassium enteredrhizoids and thallus at equal rates. Under ourconditions, the surface areas of the thallus andthe rhizoids were approximately equal (deter-mined from a series of measurements on repre-sentative cells). It therefore appears that the rateof potassium uptake per unit of surface area wasroughly the same in the two cell compartments,whereas the rates for glutamic acid and Pi weregreater in the rhizoids than in the thallus.Phosphate uptake into the entire cell was

determined by submerging filters bearing cells inmedium containing 32p;. The total uptake wasapproximately equal to the sum of the uptakeinto the rhizoids and into the thallus (Fig. 4B).The total uptake was measured for all the nutri-ents under consideration and, in all cases, wasfound to be greater than the uptake into the

A: K B: Pi1.6- - 8.0

1.2 / 6.0

10/0~~~~~~~

0 0.8 C/ 4.0

0.4 0:. 2.0-

00

0*/0@

rhizoids or into the thallus alone (data notshown). The total uptake did not always equalthe sum of the uptake into the rhizoids andthallus, but, as the three measurements weremade by using different techniques, quantitativeagreement was not to be expected. Neverthe-less, the fact that no serious discrepancies werefound reassures us that our methods for measur-ing differential uptake into the thallus and rhi-zoids were reasonably valid.

Additional experiments provided further evi-dence that the techniques employed did notdamage the cells. Cells in which the medium wasabsorbed from the rhizoid mass by placing thefilters on tissue paper accumulated as much Pi ascontrol cells not exposed to the absorption tech-nique. Furthermore, cells whose rhizoids hadbeen mock painted with DM2P25 also took up P1as well as did controls when the cells wereplaced, thallus down, on radioactive medium.Hence, neither medium absorption nor paintingof the rhizoids reduced the ability of the cells totransport Pi.There is some reason to believe that most of

the P1 uptake into the rhizoids occurred near therhizoid tip. When the depth of the oil layer,through which the rhizoids had to penetrate tocontact the radioactive medium, was increasedfrom 40 to 100 p.m, the uptake of P1 was unaffect-ed (data not shown). Hence, the upper 100 ,um of

Time (min) Time (min) Time (min)

FIG. 4. Nutrient uptake into rhizoids (O), thalli (0), and whole cells (A). After overnight growth on DM2P25,filters were removed from the floating gardens and placed, rhizoids down or thallus down, on radioactivemedium in the presence of mineral oil. Whole-cell uptake was measured by submerging the filters in radioactivemedium. After being washed, each filter was counted for radioactivity; each point represents the uptake into onefilter. (A) K+ uptake from DM2P25K25 containing 35 nCi of 42K' per ml. DM2P25K25 was made by setting the pHwith NaOH and then adjusting the monovalent salt concentration with NaCl. (B) Pi uptake from DM2P25containing 5 nCi of 32p, per ml. (C) Glutamic acid uptake from DM2P25 (no amino acids except 10 p.M glutamicacid) containing 10 nCi of [14C]glutamic acid per ml.

VOL. 151, 1982

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

434 KROPF AND HAROLD

the rhizoid probably transported very little Pi.Figure 4 suggests that there may be two

classes of metabolites, those that preferentiallyenter the rhizoids and those that are absorbeduniformly over the entire surface. To test thishypothesis, we assayed the uptake of morenutrients. In Fig. 5, the patterns of uptake for anumber of metabolites are expressed as therhizoid preference ratio (R/T; see legend to Fig.5), which serves as an index of the degree towhich uptake is localized. Pi and the amino acidsall had R/T values greater than 1, but within thisgroup the R/T values ranged from a low of 1.5for lysine to a high of 3 for glutamic acid. Theuptake of Rb+ (an analog of K+) and Ca2+mirrored that of K+ (R/T = 1). If more metabo-lites had been examined, the graph in Fig. 5might have been transformed into a continuum.Nevertheless, the results of studies with antitu-bulin drugs confirmed the existence of two dis-crete patterns of uptake.

Inhibition of nutrient uptake by antitubulins.Harold and Harold (7) found that cells grown inthe presence of the microtubule synthesis inhibi-tors nocodazole and carbendazim are small andpossess short, stubby rhizoids. We thereforeexamined the uptake of the various metabolitesby cells grown in the presence of antitubulins;uptake was expressed per milligram of cell pro-

3.01

2.0

1.0

K+ Ca++ Rb+ Lys

"I

Pi Glu Met

FIG. 5. Rhizoid preference ratio for the two classesof nutrients. R/T, Ratio of the rate of uptake intorhizoids to the rate of uptake into thalli. Rates ofuptake were determined from the slopes of the lines ingraphs similar to those in Fig. 4. Floating gardens weregrown on DM2P25 and transferred to the followingmedia for uptake measurements: K+, DM2P25K25 con-taining 35 nCi of 42K' per ml; Ca2+, DM2P25CajOcontaining 200 nCi of 45Ca2+ per ml; Rb+,DM2P25RbjO (all K+ replaced with Na+) containing 25nCi of 86Rb+ per ml; Lys, DM2P25 (no amino acidsexcept 10 ,uM lysine) containing 10 nCi of ['4C]lysineper ml; Pi, DM2P25 containing 5 nCi of 32p, per ml;Glu, DM2P25 (no amino acids except 10 ,uM glutamicacid) containing 10 nCi of [14C]glutamic acid per ml;Met, DM2P25 (no amino acids except 10 puM methio-nine) containing 25 nCi of [35S]methionine per ml.

tein to allow for the inhibition of growth. Asexpected, cells grown overnight in micromolarconcentrations of nocodazole or carbendazimaccumulated very little Pi or glutamic acid (Fig.6B and C). The inhibition of Pi uptake wasindependent of Pi concentration over the entirerange tested (50 ,uM to 1.0 mM). Potassiumuptake, however, was the same in treated cellsas in the controls (Fig. 6A). The distinctionbetween the two classes of compounds, suggest-ed by the floating garden experiments, wasmagnified in these antitubulin experiments (Fig.7).We would emphasize that antitubulins inter-

fered with transport only if the cells were grownin the presence of the drugs. Exposure of 15-h-old cells to nocodazole (20 ,uM) or carbendazim(80 ,uM) for 2 h just before use did not affect theuptake of 32p;. Cytochalasins A and B wereentirely ineffective in blocking uptake, whetherthe cells were grown in their presence overnightor exposed to higher concentrations for 2 h.Taken together, these findings suggest that elab-oration of a normal rhizoid system is requiredfor the uptake of Pi; once formed, the rhizoidsretain their transport capacity even in the pres-ence of the inhibitors.

Antitubulins may well have multiple effects oncell function, and the inhibition of rhizoidgrowth may not be solely responsible for thereduced uptake of Pi and amino acids. Cellsgrown in the presence of antitubulins were prob-ably not energy depleted, as they respired asrapidly as did the controls (Q[021 [protein] = 1.2and 1.3 ml of 02 h-' mg-1, respectively). Fur-thermore, the treated cells continued to synthe-size protein for the duration of the experiment,but at slower rates than untreated cells (data notshown). Nevertheless, the cells grown in thepresence of the drugs had severely altered mor-phologies, and an investigation of all the possi-ble effects of the antitubulins on nutrient uptakeis not within the scope of the present research.

DISCUSSIONIn the laboratory B. emersonii can grow at-

tached to solid surfaces, and we presume thatthe same mode of growth occurs in nature. Therhizoids serve as the holdfast and have beenassumed to absorb nutrients (5). However, priorto this report there has been little direct evi-dence to support this hypothesis.To our knowledge, the only relevant data have

come from studies on nutrient uptake by thehaustoria of fungi that are parasitic on plants.The haustorium is a specialized structure, re-sembling a rhizoid, that penetrates the cell wallof the host plant, bringing host and parasite intoclose contact (6). By growing host cells in la-beled medium and then infecting the host with

J . BACTERIOL .

I--, XXX

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

NUTRIENT UPTAKE VIA RHIZOIDS 435

.5120 0 600 / 10 0~~~~~~~~~~~~~

E 90- o*. 400- 400- o

200- 200-30

0 0 +

0 5 10 15 20 25 0 5 10 15 0 5 10 15Time (min) Time (min) Time (min)

FIG. 6. Effect of antitubulins on nutrient uptake. Cells were grown overnight in DM2 medium containing 5,uM nocodazole (0), 10 ,uM carbendazim (A), or Me2SO (0). Stock solutions of the antitubulins in Me2SO werediluted 1:1,000 into the medium. To measure uptake, cells were collected by filtration and suspended in freshmedia as indicated below; radioactive substrate was added 1 h later. (A) K+, DM2K25 containing 50 nCi of 42K+per ml. The pH of DM2K25 was adjusted with NaOH, and the monovalent salt concentration was manipulatedwith NaCl. (B) Pi, DM2P50 containing 10 nCi of 32p, per ml. (C) Glu, DM2Gluioo containing 10 nCi of[14C]glutamic acid per ml.

fungal spores, transfer of labeled material fromthe host to the fungus has been demonstrated for32P and 35S (13), carbohydrates (20), and aminoacids (15). Transfer was assumed to occur viathe haustoria, but release of label into the extra-cellular space followed by fungal uptake was notruled out. Extending these earlier findings, weprovide direct evidence that rhizoids do indeedtake up nutrients and that they appear to do soselectively.

This conclusion rests on two lines of evi-dence: floating garden assays and experimentsutilizing antitubulin drugs. The floating gardenassays, conducted in the presence of mineral oil,provided evidence that both rhizoids and thalliare capable of accumulating all the nutrientstested. Moreover, the results suggest that Pi andamino acids are preferentially taken up via therhizoids, whereas K+ and Ca2+ uptakes aremore uniformly distributed.For several reasons, we are reasonably confi-

dent that the floating garden assays measuredthe active transport of nutrients across the plas-malemma of the thallus and rhizoids. The possi-bility that the radioactive label was bound toextracellular sites seems unlikely; controls dem-onstrated that blank filters bound negligibleamounts of label (except 45Ca2+), and experi-ments utilizing respiratory inhibitors indicatedthat the binding of 32Pi and 42K' to nonex-changeable sites on the cell surface was mini-mal. To wit, hydroxyquinoline-N-oxide and so-dium azide severely reduced the uptake of thesetwo radioactive tracers in vegetative cells grown

in suspension culture (data not shown). Thiseffect of the respiratory inhibitors also suggeststhat the transports of Pi and K+ are activeprocesses. Furthermore, we found that after 30min of uptake in cells grown on filters, both cellcompartments and whole cells continued to ac-cumulate all the nutrients (except Ca2+) againstconcentration gradients (data not shown), pro-

100ae

2 755L

*_5

tr.

25-

o

7-

K + Ca++ Rb Lys Pi Glu Met

FIG. 7. Effect of antitubulins on nutrient uptake.This is a composite based on several experiments ofthe kind illustrated in Fig. 6. Relative uptake is theratio of the rate of uptake by antitubulin-grown cells tothe rate of uptake by control cells, multiplied by 100.Cells were grown on DM2 medium except as notedbelow. Media used for uptake: K+, DM2K25 contain-ing 50 nCi of 42K' per ml; Ca2+, cells grown onDM2Ca2OO, uptake in DM2Ca2OO containing 1.0 ,uCi of45Ca2+ per ml; Rb+, DM2P25Rb2oo (all K+ replacedwith Na+) containing 20 nCi of 86Rb+ per ml; Lys,DM2 medium containing 100 nCi of ['4C]lysine per ml;Pi, DM2P50 containing 10 nCi of 32p, per ml; Glu, cellsgrown on DM2Glu1oo, uptake in DM2Glujoo containing10 nCi of [14C]glutamic acid per ml; Met, DM2 Met0oocontaining 100 nCi of [35S]methionine per ml.

VOL. 151, 1982

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

436 KROPF AND HAROLD

viding further evidence that these nutrients areactively transported. Ca2+ uptake rapidlyreached a plateau at an internal concentration 1/10 that of the extracellular medium, which isconsistent with the hypothesis that Ca2+ is ac-tively excluded from the cells. However, somenutrients, especially Pi, may be sequestered inan insoluble form or in organelles, significantlylowering their free internal concentrations. VanBrunt et al. recently provided additional evi-dence supporting an active transport mechanismfor K+ uptake in B. emersonii (24).The suggestion that the nutrients tested in the

floating garden assays fall into two discreteclasses was reinforced by microtubule inhibitorstudies. The treated cells took up K+ and Ca2+as well as did untreated cells but were markedlydeficient in their ability to transport Pi andamino acids. There is now good evidence thatantitubulin drugs interfere with the assembly offungal microtubules (8, 14, 18). Nocodazole andcarbendazim also cause B. emersonii to elabo-rate only short, stubby rhizoids in comparison tothe bushy rhizoid mass of control cells, and webelieve that the abnormal rhizoids are at leastpartially responsible for the reduction in Pi andamino acid uptake. However, nocodazole andcarbendazim may cause additional lesions in thecell, not directly related to the reduced rhizoidmass, that result in the selective inhibition ofuptake. For example, the few remaining rhizoidsmay not be competent to translocate nutrients tothe thallus, as intact microtubules are requiredfor intracellular trafficking in many diverse celltypes (9). The inhibition of Pi and amino aciduptake is so severe (Fig. 6 and 7) as to suggestthat microtubules may be required even fortransport of these metabolites into the thallus.We are reluctant to speculate about possiblereasons but would draw attention to previousobservations that suggest a relationship betweenmicrotubules and metabolite uptake (23, 26).Although the experiments utilizing microtubuleinhibitors are inconclusive with respect to thecause of the selective inhibition of nutrient up-take, the results are consistent with the sugges-tion that the rhizoids are the major site of Pi andamino acid uptake.

This localization of functions appears to makeecological sense. In ponds, most of the Pi andamino acids are likely to be tied up in macromol-ecules contained in organic debris. Potassium,on the other hand, is soluble and present in pondwater at a concentration near 30 ,uM; calciumlevels are 10-fold higher, whereas the concentra-tion of Pi is only 1 ,uM (2). We therefore specu-late that the rhizoids serve the complementaryfunctions of attachment and procurement ofthose scarce nutrients that are bound up inorganic matter.

The localization of transport processes in par-ticular regions of a single cell is a well-docu-mented but poorly understood aspect of physiol-ogy. Epithelial cells, for example, are wellknown to effect the movement of NaCl from themucosal to the luminal side of the cell. Themucosal membrane is thought to possess Na+channels that increase the local Na+ permeabili-ty, whereas the luminal membrane has a highconcentration of the Na+-K+ ATPase that ex-trudes Na+, thus creating a Na+ flow throughthe cell. We know of no clear precedent from thefungal world, with one exception: Slayman andSlayman (19) reported that the membrane poten-tial declines toward the tip of Neurospora hy-phae, suggesting differential localization of anion pump or an ion leak. In all cases, themechanisms by which pumps and leaks becomefunctionally segregated remain unknown.We believe that tip growth provides the means

whereby transport systems for P1 and aminoacids are concentrated in the rhizoids. Fungal tipgrowth depends on membrane-bound vesiclesthat are transported down the hyphae and accu-mulate in the apical region; the fusion of thevesicles with the plasma membrane adds newmaterial to the hyphal tip (5). Rhizoids alsoelongate at the tip, and the characteristic apicalvesicles have been identified (3). We speculatethat the proteins which comprise the P1 andamino acid transporters are selectively incorpo-rated into the membranes of vesicles destinedfor the rhizoid tip and thus become localized.The Pi and amino acids that enter the rhizoid

must be translocated to the thallus and incorpo-rated into macromolecules. Fungal hyphaetranslocate nutrients at rates of 2 to 3 cm/h viacytoplasmic streaming (5). If a similar mecha-nism operates in rhizoids, the Pi and amino acidscould be moved from the rhizoid tips to thethallus in a matter of minutes. Although cyto-plasmic streaming has recently been observed inB. emersonii (R. L. Harold, personal communi-cation), we would note that the uptake of P1 wasresistant to cytochalasins which inhibit stream-ing in many organisms (1). The mechanism bywhich nutrients are carried through the rhizoidsinto the thallus remains to be clarified. Wewonder whether the transcellular electrical cur-rent observed in B. emersonii by Stump et al.(22) has any bearing on nutrient transport or onthe localization of transport systems at the rhi-zoid tip (10).

ACKNOWLEDGMENTSWe are indebted to Donald Heefner, who initiated these

experiments, and to Mayer Goren, who suggested the use ofmineral oil.

This research was supported by National Science Founda-tion grant PCM-8009439 from the Metabolic Biology Programand Public Health Service grant AI-03568 from the NationalInstitute of Allergy and Infectious Diseases.

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.

NUTRIENT UPTAKE VIA RHIZOIDS 437

LITERATURE CITED

1. Allen, R. D., and N. S. Allen. 1978. Cytoplasmic streamingin ameboid movement. Annu. Rev. Biophys. Bioeng.7:469-495.

2. Altman, P. L., and D. S. Dittmer (ed.). 1966. Environmen-tal biology, p. 508-510. Federation of American Societiesfor Experimental Biology, Bethesda, Md.

3. Barstow, W. E., and J. S. Lovett. 1974. Apical vesicles andmicrotubules in rhizoids of Blastocladiella emersonii:effects of actinomycin D and cycloheximide on develop-ment during germination. Protoplasma 82:103-117.

4. Bradford, M. M. 1976. A rapid and sensitive method forthe quantitation of microgram quantities of protein utiliz-ing the principle of protein-dye binding. Anal. Biochem.72:248-254.

5. Burnett, J. H. 1976. Fundamentals of mycology. E. Ar-nold and C. Russak, London.

6. Bushnell, W. R. 1972. Physiology of fungal haustoria.Annu. Rev. Phytopathol. 10:151-176.

7. Harold, R. L., and F. M. Harold. 1980. Oriented growth ofBlastocladiella emersonii in gradients of ionophores andinhibitors. J. Bacteriol. 144:1159-1167.

8. Howard, R. J., and J. R. Aist. 1980. Cytoplasmic microtu-bules and fungal morphogenesis: ultrastructural effects ofmethyl benzimidazole-2-ylcarbamate determined byfreeze-substitution of hyphal tip cells. J. Cell Biol. 87:55-64.

9. Hyams, J. S., and H. Stebbings. 1979. Microtubule associ-ated cytoplasmic transport, p. 487-530. In K. Roberts andJ. S. Hyams (ed.), Microtubules. Academic Press, Inc.,New York.

10. Jaffe, L. F. 1979. Control of development by ionic cur-

rents, p. 199-231. In R. A. Cone and J. E. Dowling (ed.),Membrane transduction mechanisms. Raven Press, NewYork.

11. Lovett, J. S. 1967. Aquatic fungi, p. 341-358. In F. H. Wittand N. K. Wessels (ed.), Methods in developmentalbiology. Thomas Crowell Co., New York.

12. Lovett, J. S. 1975. Growth and differentiation of the watermold Blastocladiella emersonii: cytodifferentiation andthe role of ribonucleic acid and protein synthesis. Bacter-iol. Rev. 39:345-404.

13. Mount, M. S., and A. H. Ellingboe. 1969. 32P and 35Stransfer from susceptible wheat to Erysiphe graminis f.sp. tritici during primary infection. Phytopathology59:235.

14. Quinlan, R. A., C. I. Pogson, and K. Gull. 1980. Theinfluence of the microtubule inhibitor, methyl benzimida-zol-2-yl-carbamate (MBC) on nuclear division and the cellcycle in Saccharomyces cerevisiae. J. Cell Sci. 46:341-352.

15. Reisener, H. J., and E. Ziegler. 1970. Uber den Stoffwech-sel des parasitischen Mycels und dessen Beziehungenzum Wirt bei Puccinia graminis auf Weizen. Angew. Bot.44:343-346.

16. Robinson, K. R., and L. F. Jaffe. 1975. Polarizing fucoideggs drive a calcium current through themselves. Science177:70-72.

17. Selitrennikoff, C. P., and D. R. Sonneborn. 1977. Alkalinephosphatase of Blastocladiella emersonii: partial purifica-tion and characterization. J. Bacteriol. 130:249-256.

18. Sheir-Neiss, G., M. H. Lai, and N. R. Morris. 1978.Identification of a gene for 3-tubulin in Aspergilllus nidu-lans. Cell 15:639-647.

19. Slayman, C. L., and C. W. Slayman. 1962. Measurementsof membrane potentials in Neurospora. Science 136:876-877.

20. Smith, D., L. Muscatine, and D. Lewis. 1969. Carbohy-drate movement from autotrophs to heterotrophs in para-sitic and mutualistic symbiosis. Biol. Rev. 44:17-90.

21. Soil, D. R., R. Bromberg, and D. R. Sonneborn. 1969.Zoospore germination in the water mold, Blastocladiellaemersonii. I. Measurement of germination and sequenceof subcellular morphological changes. Dev. Biol. 20:183-217.

22. Stump, R. F., K. R. Robinson, R. L. Harold, and F. M.Harold. 1980. Endogenous electrical currents in the watermold Blastocladiella emersonii during growth and sporu-lation. Proc. Natl. Acad. Sci. U.S.A. 77:6673-6677.

23. Tauber, R., and W. Reutter. 1980. A colchicine-sensitiveuptake system in Morris hepatomas. Proc. Natl. Acad.Sci. U.S.A. 77:5282-5286.

24. Van Brunt, J., J. H. Caldwell, and F. M. Harold. 1982.Circulation of potassium across the plasma membrane ofBlastocladiella emersonii: K' channel. J. Bacteriol.150:1449-1461.

25. Van Brunt, J., and F. M. Harold. 1980. Ionic control ofgermination of Blastocladiella emersonii zoospores. J.Bacteriol. 141:735-744.

26. Walker, P. R., and J. F. Whitfield. 1978. Inhibition bycolchicine of changes in amino acid transport and initia-tion of DNA synthesis in regenerating rat liver. Proc.NatI. Acad. Sci. U.S.A. 75:1394-1398.

VOL. 151, 1982

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

16

Nov

embe

r 20

21 b

y 22

0.11

8.13

.235

.