Proc. Nat. Acad. Sci. USAVol. 73, No. 3, pp. 795-799, March 1976Biochemistry

Preliminary characterization of the acrasin of the cellular slimemold Polysphondylium violaceum

(chemotaxis/cell aggregation/acrasinase)

BERND WURSTER*, PAULINE PAN, GING.- GING TYAN, AND J. T. BONNERDepartment of Biology, Princeton University, Princeton, New Jersey 08540

Contributed by J. T. Bonner, January 9, 1976

ABSTRACT Some species of cellular slime mold do notrespond to cyclic AMP as an acrasin, or chemoattractant. Inone such species, Polysphondylium violaceum, we have iso-lated and purified its acrasin and determined some of itschemical properties, which lead us to believe it is a smallmolecule of less than 1500 daltons. One possibility is that itmight be a peptide. The acrasin specifically attracts the am-oebae of P. violaceum and P. pallidum and fails to do so forsix species of Dictyostelium tested. We also have evidencefor a specific acrasinase that inactivates the Polysphondyl-ium acrasin.

Aggregation is a major event in the development of cellularslime molds: single amoebae stream together by chemotaxisand form a multicellular pseudoplasmodium. For a numberof species of Dictyostelium (D. discoideum, D. mucoroides,D. rosarium, and D. purpureum) cyclic 3':5'-AMP is knownto be the aggregation chemoattractant, or acrasin (1-3).There are, however, other species, including those of Poly-sphondylium, whose amoebae do- not orient at all to cyclicAMP, and their acrasins are not known (2, 4). It is true thatfolic acid and other pteridine derivatives attract a variety ofspecies (5, 6), but because they often are more effective at-tractants of vegetative as opposed to aggregating amoebae,and because they do not appear to be species specific, theyare not obvious candidates for acrasins.The first positive evidence that there might be a separate

Polysphondylium acrasin came from some early experi-ments in which we took supernatants from aggregating P.violaceum amoebae that had been boiled and concentrated,and shook them with aggregation competent cells of P. vio-laceum or D. discoideum. The crude extract shaken with P.violaceum cells completely lost its ability to attract P. viola-ceum amoebae in a chemotaxis test, while the extract thatwas shaken with D. discoideum cells lost little or none of itsability to attract Polysphondylium cells. Since we knewfrom previous work that D. discoideum produced enzymesthat inactivated the chemotactic activity of cyclic AMP (3,4, 7-14) and folic acid and related compounds (5, 6), weconcluded that Polysphondylium must have an acrasin of itsown. This was given further support by showing that this at-tractant, which had been identified in this crude way, wasreleased in peak amounts precisely during the aggregationperiod.

In this paper we describe the isolation of that Polysphon-dylium violaceum acrasin, its purification, and the determi-nation of some of its molecular properties. As we shall show,one possibility is that it might be a peptide.

MATERIALS AND METHODSEnzymes. Acetylcholine esterase from electric eel, ester-

ase from hog liver (type 1), and Pronase from Streptomycesgriseus (type 6) with 30% Ca(CHsCOO)2, trypsin, chymo-trypsin, prolidase, leucine aminopeptidase, elastase, and car-boxypeptidase B were purchased from Sigma Chemical Co.Elastase was also obtained from the Worthington Company.

Organisms. The cellular slime mold species used were: (i)Dictyostelium discoideum (NC-4H); (ii) D. mucoroides (no.11); (iii) D. rosarium (CC-7); (iv) D. purpureum (no. 2); (v)Polysphondylium molaceum (no. 1); (vi) P. pallidum (no.2); (vii) D. minutum (V-3); and (viii) D. lacteum (no. 1).The amoebae of these species were grown with Escherichiacoli B/r on nutrient agar. This was a buffered 1% dextroseand 1% peptone agar for the first five species listed above,and a 0.1% lactose and 0.1% peptone agar for the remainingthree species (see p. 83 of ref. 15).When the amoebae had just started to aggregate they

were harvested and washed three times in 1% (10-4 M) stan-dard salt solution (16). Species i to iv above were centrifugedfor 7 min at 75 X g, and species v to viii for 7 min at 150 X

Chemotaxis Test. The cellophane square chemotaxis testwas used (17, 18). Cellophane squares containing aggrega-tion competent amoebae were placed on non-nutrient agarin small petri dishes (50 X 12 mm). In the experimentals theagar contained the sample dissolved in 1 ml of a buffer [1%salt solution (16) in 20 mM potassium phosphate, pH 6.5];this was then added to 1 ml of 4% agar to give a total of 2 mlof agar in each petri dish. In the presence of an attractant,not only do more amoebae move off the squares than in thecontrol, but they move further out in a given length of time.The acrasinase in the agar produces a concentration gradientof acrasin, and the amoebae become more oriented in thisgradient (3) (Fig. 1).

Isolation of Attractants from Aggregating Amoebae ofP. violaceum. Washed amoebae of P. violaceum were sus-pended in 50% (5 X 10-3 M) standard salt solution (16) sothat there were 5 X 106 amoebae per ml. Fifteen millilitersof this suspension were added to each of a series of largepetri plates (150 X 15 mm; up to 120 plates) containing 2%agar. Thirty minutes later, after the amoebae had settleddown on the agar, the salt solution was carefully poured off,and the plates were air dried for 10 min and then covered.When the aggregation of the amoebae was in full swing(after 2-4 hr at 230), 10 ml of 40% ethanol in water (vol/vol)were poured on each of the plates. After 5 min the suspen-sion was decanted and centrifuged for 10 min at 1000 X g.The supernatant was concentrated to 50 ml by boiling, andsubsequently dried at 45°. This procedure is basically amodification of that of Shaffer (19) and Sussman et al. (20).

795

Abbreviation: PV acrasin, Polysphondilium violaceum acrasin.* Present address: Biozentrum der Universitat Basel, Klingelberg-strasse 70, Basel, Switzerland.

Dow

nloa

ded

by g

uest

on

June

18,

202

1

796 Biochemistry: Wurster et al.

* 0 0

1 ..

., ; : .a0. .a I :

* * a

FIG. 1. The cellophane square chemotaxis test. The amoebae of P. violaceum moving off the edge of a cellophane square placed on 2%agar. Left: control with buffer. Right: in the presence of P. violaceum acrasin (1/1000 part of one entire collection). (Width of each photo-graph is about 2 mm.)

The chemotactic activity of the crude material and thefractions obtained in the purification procedure were deter-mined by subjecting small aliquots (1/1000 to 1/100 part oftotal) to the chemotaxis test using P. violaceum amoebae.

RESULTSFractionation and purification of P. violaceum acrasinAnion Exchange Chromatography on QAE-Sephadex.

The dried raw fraction of attractants was extracted twicewith 4 ml of H20 and centrifuged for 10 min at 10,000 X g,and the combined supernatants were passed over a columnof QAE-Sephadex anion exchanger (Sigma Q-25-120, C1--form, prewashed with H20, bed volume 3 ml). The columnwas then washed with 6 ml of H20 at pH 5 or at pH 8, andmost (more than 90%) of the chemotactic activity did notbind to the anion exchanger. The active fraction was calledQ1.

However, some additional chemotactically active materialcan be eluted from the QAE-Sephadex with 20 mM HC1(fraction Q2), indicating the presence of another attractantwhich is negatively charged. Also, in earlier experiments,when the attractants were extracted from aggregating amoe-bae with 70% or 100% ethanol (a procedure which breaksthe cells), yet another attractant could be eluted from theQAE-Sephadex between 100 mM and 200 mM HC1 (frac-tion Q3). Fractions Q2 and Q3 were not purified further forthey were not species specific.

Cation Exchange Chromatography on SP-Sephadex.Fraction Q1 was subsequently applied to a column of SP-Sephadex cation exchanger (Sigma SP-C25-120, H+-form,prewashed with H20, bed volume 3 ml). The chemotacticactivity passes the column with H20 (fraction SP). In a simi-lar experiment the cation exchanger was equilibrated with20 mM HC1, and fraction Q1 in 20 mM HCl was applied.The chemoattractant did not bind to the cation exchangerand passed the column with 20 mM HCl.

Gel Filtration on Sephadex G-15. Fraction SP was con-centrated to 2.4 ml at 450 and applied to a column of Sepha-dex G-15 equilibrated with H20 (Pharmacia, diameter 2.6cm, height 57 cm). The compounds were eluted from thecolumn with H20. The flow velocity was 120 ml/hr, andfractions of 10 ml were collected. The void volume of thecolumn was 105 ml; the attractant was eluted between 170

and 190 ml (fraction G), and inorganic ions, like[Co(H2O)6]++ or [Cu(H2O)4]++, peaked at 210 ml. FractionG was concentrated to dryness at 45°.The elution volume of the chemotactically active material

was not changed if elution was conducted with 20 mM HClinstead of H20, showing that none of the P. violaceum at-tractant absorbed to the Sephadex matrix in H20 (21).

Because the attractant eluted far behind the exclusion vol-ume but before the inorganic ions, its molecular weight isprobably less than 1500.

Partition Chromatography on Cellulose Thin Layers.Dried fraction G was dissolved in 1 ml of H20 and appliedto a cellulose thin-layer sheet (Eastman Chromogram Sheet6065 Cellulose). After the chromatogram was developedwith 1-butanol saturated with 0.5 M HCI, it was divided intoseven horizontal fractions. Each fraction was scraped off thesheet, extracted twice with 3 ml of H20, and dried.The fraction with chemotactic activity moved to a RF be-

tween 0.70 and 0.86. We call this fraction Polysphondyliumviolaceum or PV acrasin, and it was used in most of the ex-periments described below. (It should be mentioned that cy-clic AMP and folic acid stayed close to the origin, with RFvalues below 0.1.)

Adsorption Chromatography on Activated Silica GelThin Layers. The PV acrasin could also be chromato-graphed on activated silica gel thin layers (Baker-flex SilicaGel 1B) using pure ethanol as a solvent (22). Fractions wereagain scraped off the sheet and extracted with H20. Thechemotactic activity moved to the fraction betwn RF 0.44and 0.56.

Further chemical characterizationIn the following experiments a solution of purified PV acras-in was divided into several samples of equal volume anddried at 45°. The chemotactic activity was tested in aliquotsof the sample before and after the experiment. The reactionmixture was also rechromatographed on cellulose thin layer(see above) and then tested to see if the acrasin molecule wasstill active.

Heat Stability. Dry samples of PV acrasin were incubatedfor 10 min in an oven at 100, 120, 140, 160, 180, and 2000.In no case was there any loss of chemotactic activity.

Solubility. Dry samples of PV acrasin were extractedtwice with 3 ml of one of the following solvents: water,

Proc. Nat. Acad. Sci. USA 73 (1976)

Dow

nloa

ded

by g

uest

on

June

18,

202

1

Proc. Nat. Acad. Sci. USA 73 (1976) 797

methanol, ethanol, pyridine, acetone, 1-butanol, dioxane,ethylacetate, chloroform, and ether. The extracts were cen-trifuged; the supernatants were dried at 450, subsequentlydissolved in buffer, and applied to the chemotaxis test.The PV acrasin is soluble in water, methanol, and ethanol;

slightly soluble in pyridine, acetone, 1-butanol, and dioxane;and insoluble in ethylacetate, chloroform, and ether. Clear-ly, the attractant is polar.

Ionic Properties. PV acrasin does not bind to an anion ex-changer (equilibrated with 20 mM imidazole-HCl buffer,pH 8), nor to a cation exchanger, either in H20 (pH 5) or in20 mM HC1. (These experiments were performed both withcrude extracts and purified PV acrasin.) It is therefore likelythat PV acrasin is not an anion, a cation, or a zwitterion (thelatter because of the result with 20 mM HCl).

Determination of Free Amino Groups. When a DowexAg 50WX8 (HW-form, 100-200 mesh) column was used, theactive PV acrasin passed through in 0.1 M phosphate buffer(pH 7.0), indicating the absence of a free amino group. As acontrol, a similar column held 0.01 M of tyrosine when elut-ed with the buffer, but was released with 0.1 M ammoniumhydroxide.

Gel Electrophoresis. This lack of any significant overallcharge on the molecule was confirmed by gel electrophore-sis, first in 0.015 M barbital-0.015 M Tris (pH 7.5) and thenin 0.03 M Tris (pH 6.0). Bromophenol blue (0.1%) was usedas the tracking dye in the first instance and pyronin B (0.1%)in the second. In experiments lasting 2 hr at 40 with 3-4 mAper gel, there was essentially no movement of the PV acrasinin the polyacrylamide gels.

Periodate Oxidation. A dry sample of PV acrasin was dis-solved in 2 ml of water, and 0.75 ml of 100 mM sodium per-iodate was added. The reaction mixture was kept in the darkfor 2 hr and then passed over a column of AG 1-XB (200-400mesh, Cl-form, Bio-Rad, bed volume 3 ml) in order to re-move iodate and excess periodate (23). The eluate was con-centrated at 450 and then rechromatographed on a cellulosethin-layer plate.

After periodate oxidation the chemotactic activity movedto the same RF region as before, indicating that the acrasinmolecule is neither inactivated nor altered by periodate. Thepresence of vicinal hydroxyl groups in the PV acrasin cantherefore be excluded, which in turn rules out all commonsugars as a functional part of the molecule.

Alkali and Acid Lability. Dry PV acrasin was dissolvedin 5 ml of water and divided into five equal parts. One partwas boiled at neutral pH for 30 min. By addition of 1 ml of40 mM KY3P04 the second part was brought to pH 12 and in-cubated for 1 hr at 250. The third part was brought to pH 12in the same way and heated for 20 min at 80°. The fourthpart was mixed with 1 ml of 2 M HC1 (yielding a final con-centration of 1 M HCI) and incubated for 1 hr at 250, andthe fifth part was mixed with 1 ml of 2 M HC1 and boiledfor 30 min. After the incubation period each sample was ad-justed to pH 6.5 for the chemotaxis test.The PV acrasin is an alkali-labile molecule; incubation at

pH 12 for 1 hr at 25° caused a decrease of the chemotacticactivity, and heating at this pH for 20 min at 800 lead tocomplete inactivation. Incubation of the acrasin in 1 M HC1for 1 hr at 250 did not lead to a loss of chemotactic activity.However, boiling in 1 M HC1 for 30 min caused completeinactivation. Boiling at neutral pH for 30 min did not affectthe acrasin.

Esterases. Dry PV acrasin was dissolved in 4 ml of 20 mMimidazole-HCl buffer pH 8.0 and divided into four samples

of equal volume. The samples were incubated for 1 hr at 250with one of the following preparations dissolved in 1 ml ofthe same buffer: (i) 0.5 mg of acetylcholine esterase; (ii) thesame inactivated by boiling; (iii) 1 mg of hog liver esterase;and (iv) the same inactivated by boiling. All the sampleswere boiled after the incubation period and centrifuged;each sample was chromatographed on cellulose thin-layerplates as described previously. In all instances the full che-motactic activity was recovered from the appropriate RF re-gion of the plates. These enzymes did not affect the acrasin(even though they inactivated acetylcholine in control ex-periments).

Proteases. A number of proteases were tested in the samemanner, and similar negative results were obtained withtrypsin, chymotrypsin, elastase, prolidase, leucine amino-peptidase, and carboxypeptidase B. Also, there was no activ-ity loss of PV acrasin when incubated with insolubilizedtrypsin, chymotrypsin, and papain. However, clear inactiva-tion was obtained with peptidase from hog intestinal mucosaand Pronase from Streptomyces griseus. The latter was in-vestigated carefully, and it was possible to show, in four sep-arate experiments (using 5 mg of Pronase), that the chemo-tactic activity was totally eliminated and not recoverable inany part of the thin-layer plates, while the normal recoveryappeared in the controls with the boiled enzyme. In order tobe certain that the PV acrasin was not simply binding to theenzyme, the mixture was incubated at 40 instead of 22°,producing no acrasin inactivation. Finally, the Pronase wasdialyzed against 0.05 M EDTA to eliminate neutral protein-ase activity (24); it then was no longer capable of inactivat-ing the PV acrasin. Therefore, one could assume that somespecific neutral proteinases in S. griseus Pronase are able todestroy the acrasin activity.

Weight/Protein Analysis. One entire collection of PV ac-rasin (from 100 large petri plates; see Materials and Meth-ods) was purified through to the cellulose thin-layer step. Itstotal weight was 2.1 mg. With the Folin-Ciocalteau test forprotein, a yield of 650 ,ug was obtained, using bovine serumalbumin as a standard.The material was hydrolyzed under reduced pressure

with 1 ml of 12 M HCI for 20 hr at 1100. In some prelimi-nary experiments the hydrolysate was examined in anamino-acid analyzer. There were considerable backgroundamino acids, but the active material contained some proline,valine, and leucine. There was less clearcut evidence forthreonine, glutamine, alanine, and isoleucine.SpecificityWhen we had isolated and purified the attractant it was im-portant to determine if it was specific for P. violaceum. Inthe following experiments it was possible to show that in-deed it was specific for both P. violaceum and P. pallidum,but did not attract the amoebae of any of the Dictyosteliumspecies tested.

(i) Extracellular material was isolated and purified fromaggregating amoebae of D. discoideum using the proceduredescribed for P. violaceum. None of the material from anyof the fractions attracted sensitive P. violaceum cells; D. dis-coideum does not give off PV acrasin.

(Hi) Purified PV acrasin, when incubated with aggregationcompetent amoebae of P. violaceum, lost its activity, mostlikely by the action of an acrasinase (see below). However,when a similar experiment was done with D. discoideumamoebae there was no loss of activity; D. discoideum doesnot inactivate PV acrasin.

Biochemistry: Wurster et al.D

ownl

oade

d by

gue

st o

n Ju

ne 1

8, 2

021

798 Biochemistry: Wurster et al.

Table 1. Chemotactic response ofaggregation competent amoebae of eight species

of cellular slime molds to PV acrasin and cyclic AMP

Chemoattraction by

Species PV acrasin cyclic AMP

P. violaceum (no. 1) + 0P. pallidum (no. 2) + 0D. discoideum (NC-4H) 0 +D. rosarium (CC-7) 0 +D. mucoroides (no. 11) 0 +D. purpureum (no. 2) 0 +D. minutum (V-3) 0 0D. lacteum (no. 1) 0 0

For each species at least two chemotaxis tests were performed.+, Strong chemotactic response. 0, None; same as buffer control.

(iii) Purified PV acrasin was tested on sensitive amoebaeof a number of different species (Table 1), and it is clear thatit only attracts the two Polysphondylium species. Chemo-tactic results with cyclic AMP are also included in Table 1.Note that two small species, D. minutum and D. lacteum,are attracted neither by cyclic AMP nor by the PV acrasin.AcrasinaseThese experiments were designed to determine whether ornot P. violaceum produced an enzyme (or enzymes) capableof inactivating PV acrasin, and whether the enzyme waspresent outside or inside the amoebae, or bound to a particu-late fraction.

Aggregating amoebae and surrounding extracellular ma-terial were washed off of 12 large petri dishes (2% agar; 150X 15 mm) with 30 ml of 20 mM potassium phosphate bufferin 1% salt solution (pH 7.0) (16). To separate the variousfractions, we centrifuged the cell suspension for 10 min at1000 X g, yielding a supernatant and a precipitate of cells.The supernatant was subsequently centrifuged for 20 min at30,000 X g, resulting in a clear supernatant (extracellularfraction). The precipitate of cells was washed twice withbuffer and centrifuged for 10 min at 1000 X g. Precipitatedcells (0.7 ml) were suspended in 5 ml of buffer, treated in aBranson sonifier 5 times for 10 sec, and subsequently centri-fuged for 10 min at 1000 X g. The supernatant was thencentrifuged for 20 min at 30,000 X g, yielding a new super-natant (intracellular fraction) and a precipitate of cell frag-ments. The precipitate was washed by suspending it in 5 mlof buffer, centrifuging for 10 min at 1000 X g, and subse-quently centrifuging the supernatant for 20 min at 30,000 Xg. Again the precipitate was suspended in 5 ml of buffer.(particulate fraction).The three fractions were divided into two parts, and one

part of each fraction was boiled. PV acrasin samples wereincubated with these fractions or with buffer for 30 min at250. Subsequently the reaction mixtures were boiled andcentrifuged, and the supernatants tested for chemotactic ac-tivity.

All three fractions have the capacity to inactivate the ac-rasin. Furthermore, the inactivating principle is heat labileand loses its ability to inactivate upon boiling. Therefore, itis likely that the inactivator in all three fractions is an en-zyme, an acrasinase. This situation is similar to that of D.discoideum, where the occurrence of extracellular (7-14),intracellular (4, 25), and membrane-bound cyclic cAMP-phosphodiesterases (12-14) have been demonstrated.

DISCUSSIONTo summarize the results briefly, we have found a substancethat appears to act as a specific acrasin for Polysphondyl-ium; there is also a specific acrasinase produced by Polys-phondylium violaceum which, along with the acrasin, is notproduced by Dictyostelium discoideum. No known speciesresponds chemotactically to both cyclic AMP and PV acras-in.

This Polysphondylium molaceum acrasin has a molecularweight of less than 1500 and is heat stable. It is polar but un-charged; it is not an anion, or a cation, or a zwitterion. Theinactivation of acrasin by alkali and acid in comparativelymild conditions suggests the presence of an ester bond, anesterized carboxyl group of an amino acid, or a labile pep-tide bond [such as the y-glutamyl bond in glutathione (26)].One possibility is that the substance is a peptide. The best

evidence comes from the Pronase results. If it is a peptide, itmust have a blocked amino group. This raises the possibilityof a cyclic peptide, or the group could be blocked in someother way, as Schiffmann et al. (27) have shown for leuko-cyte attractants.

Besides the further identification of the PV acrasin mole-cule, another approach will be important. It is well knownfrom earlier work that Polysphondylium species synthesizeand release cyclic AMP, as well as a cyclic AMP-phosphodi-esterase (4, 10). What is the biochemical relation betweenthe PV acrasin and the cyclic AMP? Does PV acrasin act as aprimary messenger and cyclic AMP as a secondary messen-ger in a manner similar to what one finds in many mamma-lian hormone systems? This question, which is being activelypursued, is of interest both from the point of view of cellphysiology and from that of phylogeny when one considersthat in Dictyostelium, cyclic AMP seems to serve both func-tions.

We thank Drs. C. Gilvarg and 0. Shimomura for helpful discus-sions, and Drs. E. C. Cox and A. Newton for critiques of the manu-script. Also we thank Ms. B. Bamman and Mr. L. R. Hyde for run-ning the preliminary amino-acid analyses. We are especially grate-ful to Ms. M. Murphy-Caine for her invaluable technical assistancein the latter part of this study. This investigation was supported byResearch Grant GM 33439 from the National Science Foundation,by Grant GM 17856 from the National Institute of General MedicalSciences, and by a research fellowship from the Deutsche For-schungsgemeinschaft to B. W. which is also gratefully acknowl-edged. We benefited from the central equipment facilities of theBiology Department, Princeton University, supported by the Whi-tehall Foundation.

1. Konijn, T. M., van de Meene, J. G. C., Bonner, J. T. & Bark-ley, D. S. (1967) Proc. Nat. Acad. Sci. USA 58, 1152-1154.

2. Konijn, T. M. (1973) in Proceedings of the Tenth Interna-tional Congress of Microbiology, Mexico City, Mexico, ed.Perez-Mirauete, A. (Plenum Press, New York), pp. 48-61.

3. Bonner, J. T., Barkley, D. S., Hall, E. M., Konijn, T. M.,Mason, J. W., O'Keefe, G., III & Wolfe, P. B. (1969) Dev. Biol.20,72-87.

4. Bonner, J. T., Hall, E. M., Noller, S., Oleson, F. B., Jr. & Rob-erts, A. B. (1972) Dev. Biol. 29, 402-407.

5. Pan, P., Hall, E. M. & Bonner, J. T. (1972) Nature New Biol.237, 181-182.

6. Pan, P., Hall, E. M. & Bonner, J. T. (1975) J. Bacteriol. 122,185-191.

7. Chang, Y. Y. (1968) Science 160, 57-59.8. Riedel, V. & Gerisch, G. (1971) Biochem. Biophys. Res. Com-

mun. 42, 119-124.

Proc. Nat. Acad. Sci. USA 73 (1976)

Dow

nloa

ded

by g

uest

on

June

18,

202

1

Biochemistry: Wurster et al.

9. Riedel, V., Gerisch, G., Muller, E. & Beug, H. (1973) J. Mol.Biol. 74,573-585.

10. Gerisch, G., Maichow, D., Riedel, V., Miller, E. & Every, M.(1972) Nature New Biol. 235,90-92.

11. Chassy, B. M. (1972) Science 175, 1016-1018.12. Maichow, D., Niigele, B., Schwartz, H. & Gerisch, G. (1972)

Eur. J. Biochem. 28, 136-142.13. Pannbacker, R. G. & Bravard, L. J. (1972) Science 175,

1014-1015.14. Malkinson, A. M. & Ashworth, J. M. (1973) Biochem. J. 134,

311-319.15. Bonner, J. T. (1967) The Cellular Slime Molds (Princeton

Univ. Press, Princeton, N.J.), 2nd ed.16. Bonner, J. T. (1947) J. Exp. Zool. 106, 1-26.17. Bonner, J. T., Kelso, A. P. & Gillmor, R. G. (1966) Biol. Bull.

130,28-42.18. Bonner, J. T., Hall, E. M., Sachsenmaier, W. & Walker, B. K.

(1970) J. Bacteriol. 102, 682-687.

Proc. Nat. Acad. Sci. USA 73 (1976) 799

19. Shaffer, B. M. (1956) Science 123, 1172-1173.20. Sussman, M., Lee, F. & Kerr, N. S. (1956) Science 123, 1171-

1172.21. Determann, H. (1967) Gelchromatographie (Springer-Verlag,

Berlin).22. Bobbit, J. M., Schwarting, A. E. & Gritter, R. J. (1968) Intro-

duction to Chromatography (Van Nostrand Reinhold Co.,New York).

23. Kohn, P., Lerner, L. M. & Kohn, B. D. (1967) J. Org. Chem.32,4076.

24. Narahashi, Y., Shibuya, K. & Yaganita, M. J. (1968) J. Bio-chem. (Tokyo) 64,427-432.

25. Rossomando, E. F. & Sussman, M. (1973) Proc. Nat. Acad. Sci.USA 70, 1254-1257.

26. Kendall, E. C., Mason, H. L. & McKenzie, B. F. (1930) J. Biol.Chem. 88,409-423.

27. Schiffmann, E., Showell, H. V., Corcoran, B. A., Ward, P. A.,Smith, E. & Becker, E. L. (1975) J. Immunol. 114, 1831-1837.

Dow

nloa

ded

by g

uest

on

June

18,

202

1