The Use of DNA Probes for Taxonomic Study of Dictyostelium Wild Isolates

William B. Evans, Joanne E. Hughes and Dennis L. Welker Molecular BiologylBiochemistry Program, Department of Biology, Utah State University, Logan, Utah 84322-5500 Manuscript received October 9, 1987 Revised copy accepted March 3, 1988

ABSTRACT The classification of 27 wild isolatesassigned to Dictyostelium discoideum on the basis ofmorphological criteria was reexamined using probes specific for DNA sequences cloned from the type strain NC4. These probes included ones specific for ribosomal spacer DNA regions and for a ribosomal RNA coding sequence, as well as probes for two chromosomal gene families (actin and discoidin) and for the DIRS-1 transposable element. Four isolates (AC4, WS526,WS584 and ZASA)which had previously been shown to have unusual mating characteristics were distinctly different from other isolates. We interpret these differences as indicating that the four atypical isolates represent species other than D. discoideum. Probes for the ribosomal spacer DNA either did not hybridize to the DNA of these four isolates or had decreased levels of hybridization to EcoRI restriction fragments of different lengths to that observed with the type strain. With the discoidin probe, all isolates had DNA fragments that hybridized but AC4, WS526, WS584 and ZASA lacked a pair of fragments that were conserved in NC4 and other isolates. With the actin probe, AC4, WS526, WS584 and ZASA lacked numerous fragments that the other isolates shared with NC4. The DIRS-1 probes showed strong hybridization with ZASA and weak hybridization to the other threeisolates; however, the major EcoRI fragment in WS526 and WS584 was smaller than that in NC4 while ZASA and AC4 had fragments of similar size to that in NC4. The asexual isolates WS526 and WS584 were closely related to each other and appearto be isolates of the same species although there were some differences in the overall pattern of EcoRI restriction fragments from high copy number sequences and of restriction fragments hybridizing to the actin and discoidin gene probes. The homothallic isolates AC4 and ZASA appear to represent two additional species; ZASA, and possibly AC4, are more closely related to D. discoideum than the species represented by WS526 and WS584. Overall it appears that the four atypical isolates (AC4, ZASA, WS526 and WS584) are less distantly related to D. discoideum than are isolates of D.purpureum and D. mucoroides.

HE cellular slime mold Dictyosteliumdiscoideum the wild isolate mustbe crossed toa genetically T has been studied for many years by develop- marked haploidtester strain to producea diploid mental biologists. This organism differentiates dur- from which segregant haploids arethen obtained ing asexualfruiting body formationinto only two (WELKER,HIRTH and WILLIAMS 1985;GRANT, major cell types, spores and stalk cells (LOOMIS 1982; WELKERand WILLIAMS 1985;WELKER et al. 1986). In SPUDICH 1987).Molecular genetic analyses involving order to facilitate such studies wild isolates must be transformation,insertional mutagenesis, antisense properly assigned to their respective species. Like- RNA, and amplified gene expressionare now possible wise, it is importantto know that plasmid-bearing (NELLEN,SILAN and FIRTEL 1984; KNECHT and isolates are properlyidentified; particularly in the LOOMIS1987; DELOZANNE and SPUDICH 1987; REY- construction and use of transformation vectors based MOND, NELLENand FIRTEL 1985;REYMOND et al. on endogenous plasmids, so that the host range of 1986). Molecular approaches arealso being employed the plasmids can be defined and problems associated to expand the genetic linkage map of D. discoideum with it understood. (GRANT, WELKERand WILLIAMS,1985; WITKE et al. In thepast, morphological featuresof the amoebae, 1986; WELKERet al. 1986; DINGERMANNet al. 1987) asexualfruiting bodies and sexual structures have and to study nuclearplasmid DNAs (METZet al. 1983; been used for classification in the Dictyostelids (RAPER NOEGELet al. 1985; NOEGEL,METZ and WILLIAMS 1984).However, increased knowledge of the wild 1985). Many of the recent genetic studies use either isolates has made it apparent that there arelimitations proteinpolymorphisms or restriction fragment associated withclassification based solely on mor- length polymorphisms (RFLPs). Mostof the poly- phology, since it is known that morphological char- morphisms utilized have been identified by compar- acteristics are dependent to some extent on various ison of different wild isolates. For linkage analyses, aspects of culture conditions (RAPER 1984). Nucleic

Genetics 119: 561-569 (July, 1988). 562 W. B. Evans, J. E. Hughes and D. L. Welker acid and protein homologieshave been useful for FRANCIS(University of Delaware). Isolates of the other species identification and phylogenic studies in other specieswere obtained from K. B. RAPER (University of organisms. Phylogenic work based on DNA includes Wisconsin), D. FRANCIS(University of Delaware), D. WAD- DELL (BergischeUniversitat, Wuppertal), J. C. CAVENDER determinationsof sequences, hybridization studies, (Ohio University), G. W. ERDOS (University ofFlorida), D. and analyses of RFLPs. DNA sequences present in H. O'DAY(University of Toronto) and the American Type multiplecopies per cell haveoften been utilized. Culture Collection. Strains were grownat 21" in association These include mitochondrial DNA, kinetoplast DNA, with Escherichia coli B/ron DM medium (PODGORSKIand repetitive DNA, and, in particular, the DNA encodingDEERINC 1980) or with Klebsiella aerogenes on SM medium (WELKER1986; SL'SSMAN1966). The geographicalorigins ribosomal RNA (DUTTA1986; AVISE, LANSMANand and sexual mating characteristicsof the D.discoideum isolates SHADE 1979; MOREL et al. 1980;BORST et al. 1980; are given as part of Table 1. JEFFREYS, WILSONand THEIN1985; APPELSand Dvo- DNA isolation: DNAs were isolated using a modification RAK 1982; MATTHEWSand DEBONTE1985; WOESE ofpreviously published techniques (WELKERet al. 1986). 1987). Genes encoding ribosomal RNAs are highly DNA isolated from bacteriallygrown cells was extracted with phenol and chloroform, ethanol precipitated and then conserved because of the ubiquitous requirement for further purified using cesium chloride density gradients. these nucleic acids, hence their sequences can be used After recovery from the gradients the DNA was spool- to determine the phylogenic relationships of groups precipitated with ethanol, dried, and dissolved in TE (10 that havebeen separate for long periods of time. mM Tris, 1 mM EDTA, pH 7.5). DNA sequence analysisof a ribosomal RNAgene Slot blots, electrophoresis and hybridization: The Schleicher and Schuell Minifold I1 slot blot apparatus was indicates that D.discoideum diverged early from other used for slot blots; genomicDNA (1 kg/slot) was pretreated eukaryotes (MCCARROLLet al. 1983). Characterization and loaded according to the manufacturer's instructions. of theless highly conserved spacerDNA surrounding Following application of the DNA, the nitrocellulose mem- the ribosomal RNA genes by sequencing, hybridiza- brane was rinsed in 5 X SSC, baked for 2 hr, then tion or RFLP analysis provides information on more prehybridized, hybridized, and hybridization detected according to standard techniques. DNA samples for gels were recently separated species. digested with restriction enzymes and separated on 0.8% The DNAencoding the ribosomal genes in D. agarose gels using a Tris phosphate running buffer (36 discoideum has been studied extensively. This rDNA mM Tris, 30 mM NaH2P04,1 mM EDTA). The separated is present in extrachromosomal palindromic dimers DNA fragments were blotted to nylon membranes (Zeta- of 88 kb (COCKBURN,TAYLOR and FIRTEL1978). Each bind, Bio-Rad) using the alkaline transfer protocol (REED and MANN 1985) or to nitrocellulose (Nitroplus 2000, cell contains about 90 dimers allowing easy visualiz- Micron Separations) by the SOUTHERN(1975) protocol. For ationof restriction fragments from the rDNA on useas probe DNA, E. coli plasmidswith inserts of D. agarose gels stained with ethidium bromide. Cloned discoideum DNA were nick translated using either 92P-dATP DNA is available for much of the coding and spacer or biotinylated dUTP. Biotin-labeled probe DNA was de- DNA on the rDNA palindromes(NESS et al. 1983; tected using kits and protocols supplied by Bethesda Re- search Laboratories. The probes were: pCTl (EMERYand EMERYand WEINER198 1); RFLPs affecting this DNA WEINER1981), pEcoIV and pEcoVII (NESS et al. 1983), have been identified (WELKER,HIRTH and WILLIAMS pDd812 (DEVINE,TSANG and WILLIAMS1982), ACTIN-8 1985). Hence we chose to investigate whether or not (ROMANSand FIRTEL1985), GM45A (CHUNG,ZUKER and wild isolates designated as D. discoideum on the basis LODISH1983), pB41-6 (ZUKERand LODISH1981). Probes pCTl and pEcoIV are of ribosomal spacer DNA, pEcoVII of morphological criteria were closely related based is of ribosomal gene-coding DNA; pDd812is a cDNA clone on (1) the ability oftheir DNA tohybridize with of a discoidin I gene; ACTIN-8 contains a genomicse- probes derived from the rDNA of the D. discoideum quence encoding an actin gene; GM45A and pB41-6 con- type strain NC4 and (2) where hybridization to the tain the DIRS-I internal EcoRI fragment and terminal rDNA probes occurred, on the presence of RFLPs. repeat, respectively. We also utilized probes for two protein-coding gene families (actin and discoidin) and the DIRS-1 trans- RESULTS posable element. Our findings indicate that 4 of 27 isolatespreviously designated D. discoideum are not Slot blot analysis of Dictyostelium isolates using closely related to the others. These results are con- rDNA probes: As with other groups of species, the sistentwith those obtained using assays based on spacer DNA surrounding the ribosomal RNA genes protein polymorphisms (BRISCOEet al. 1987). of Dictyostelium speciescan beused in taxonomic studies of this genus. Spacer DNA probes (pEcoIV, and pCT1) cloned from NC4-derived strains of D. MATERIALS AND METHODS discoideum did not hybridize to DNA purified from Strains and growth conditions: Most of the D.discoideum D. mucoroides (12 isolates) and D. purpureum (10 iso- wild isolates (Table 1) havebeen described previously lates). Theseprobes did not hybridize withDNAs (ERDOS,RAPER and VOCEN 1973; ROBSONand WILLIAMS from a more limited number ofisolates of the follow- 1979; ROBSONand WILLIAMS1980; POOLEand FIRTEL1984; WELKERet al. 1986; DINCERMANNet al. 1987). D. discoideum ing species: D. caveatum, D. giganteum, D. lacteum, D. isolates DD185, DD285 and DD385 were obtained from D. rosarium and D.sphaerocephalum. A probefor the Dictyosteliurn DNA Taxonomy 563

TABLE 1

Hybridization of rDNA probes to isolates identified as D. discoideum by morphological criteria

Probe Mating Strain pEcoVlI pCT 1 pEcoIV characteristics Origin AC4 +++ + Homothallic Mexico DD6 1 +++ +++ +++ Sexual Kansas DD185 +++ +++ +++ Not tested Delaware DD285 +++ +++ +++ Not tested Delaware DD385 +++ +++ +++ Not tested Delaware HU182 +++ +++ +++ Sexual Wisconsin HU188 +++ +++ +++ Sexual Wisconsin K10 +++ +++ +++ Not tested Ohio MFD +++ +++ +++ Not tested Delaware NC4 +++ +++ +++ Sexual North Carolina OHIO +++ +++ +++ Sexual Ohio V12 +++ +++ +++ Sexual Virginia ws7 +++ +++ +++ Sexual Wisconsin ws51 +++ +++ +++ Sexual Wisconsin WS112B +++ +++ +++ Bisexual Wisconsin WS269 +++ +++ +++ Sexual Wisconsin WS380B +++ +++ +++ Sexual Wisconsin ws472 +++ +++ +++ Sexual Wisconsin WS526 +++ Asexual Wisconsin WS576 +++ +++ +++ Sexual Wisconsin WS582 +++ +++ +++ Sexual Wisconsin WS583 +++ +++ +++ Sexual Wisconsin WS584 +++ Asexual Wisconsin WS656 +++ +++ +++ Sexual Wisconsin WS1956 +++ +++ +++ Sexual Wisconsin WS2 162 +++ +++ +++ Bisexual Wisconsin ZA3A +++ + +++ Homothallic Unknown Hybridization of probes for ribosomal spacer DNA (pCT1 and pEcoIV) and ribosomal RNA coding DNA (~Eco~~I)was scored using slot blots as follows: + + + , strong hybridization; + , weak hybridization; - no detected hybridization. Mating characteristics and origins are adapted from BRISCOEand coworkers (1987) based on the work of ERDOS,RAPER and VOCEN(1973), of RAPER (1984), and of ROBSOS and WILLIAMS(1979, 1980). Particularly relevant to the assignment of WS380B, WS576 and WS269A to the “sexual” isolates are their ability to exchange genetic material by parasexual diploid formation with D. discodeurn matA tester strains (ROBSONand WILLIAMS1980). coding region (pEcoVII)hybridized to DNA purified AC4 but appeared to hybridize extensively to ZA3A from allwild isolates, including those of the other (Figure 1, and data not shown). All four hybridized Dictyostelid species. Results with representative D. to the pEcoVIIprobe. These four isolates were purpureum and D. mucoroides isolates are shown in among isolates previously reported to have unusual Figure 1. These results establish that sequence dif- sexualmating characteristics. WS526 and WS584 ferences have accumulated in the spacer DNAs in were classed as asexual while AC4 and ZA3A were Dictyostelium species but that, as expected, the coding homothallic (ERDOS,RAPER and VOCEN1973; ROBSON regions are more conserved. Hybridization patterns and WILLIAMS 1980).Together these results suggest with DNA from 22 of 26 isolates identified as D. that the four isolates are not closely related to the discoideum on thebasis of morphological criteria were other isolates presently identified as D. discoideum. similar to that observed for the NC4 wild isolate itself Restriction fragment length analysis of Dictyos- (or its derivative AX3K). This was true whether the telium isolates usingrDNA probes:In previous work probe DNA was obtainedfrom spacer or coding using rDNA probes to identify restriction fragment sequences (Figure 1, Table 1). Of the remaining four lengthpolymorphisms in 14 D. discoideum isolates, (AC4, WS526, WS584 and ZAJA), hybridization pat- several polymorphisms affecting spacer DNA were terns of WS526 and WS584 were similar to those identified (WELKER,HIRTH and WILLIAMS1985). That observed with the isolates known to be from species work did not include the four atypical isolates (AC4, other than D. discoideum; there was no hybridization WS526, WS584 and ZASA) identified by the slot blot to pEcoIV andpCT1. More complicated patterns analysis. To further study the relatedness of these were observed with AC4 and ZASA. The pCTlprobe isolates tothe other isolates, we analyzed EcoRI did not hybridize with DNA from AC4 but limited restriction fragment patterns using ethidium bromide hybridization was seen to this probe with DNA from staining and blots of 0.8% agarose gels. This analysis ZASA; the pEcoIV probe hybridized only slightly to included NC4 and five additional isolates whose 564 M:. B. Evans, J. E. Hughes and D. L. Welker Isolates of D. mucoroides and D. purpureum analyzed as controls in these experiments gave the expected results for strains of species other than D. discoideum 1 (Figure 2). Banding patterns with ethidium bromide were distinctly different from those of the D. discoi- d deum isolates. Each of these six isolates had an EcoRI L restrictionfragment that hybridized tothe rRNA t gene probe (pEcoVII) but did not contain fragments 1 that hybridized to either of the spacer DNA probes I (pCT1 and pEcoIV). However, isolates assigned to c D. purpureum appeared to represent separate groups; x DP2 and DPi, which can mate, were similar to each 1' otherbut distinct from DP3645. All three isolates 1: assigned to D. mucoroides had a differnt pattern of x bands. 1L Restriction fragment length analysis usingprobes

l!5- " forchromosomal gene families: Additional infor- f 5- mation on the status of AC4, WS526,WS584 and 1;7- ZA3A was obtained from analysisof the actin and I& discoidin gene families. Both gene families have been FIGURE1 .-Slot blot analysis of representative Dictyskdium iso- characterized in previous work, so cloned probes lates using probes for ribosomal gene-coding (pEcoVII) andspacer were available (ROMANSand FIRTEL 1985; DEVINE, DSA (pEcoIV, pCT1) regions. Slots are as follows: 1, NC4;2, TSANGand WILLIAMS 1982). Restriction fragment AC4; 3. ZA3A; 4, WS.526; 5, WS584; 6, AXSK;7, M'S380B; 8, length polymorphisms among known D. discoideum M'SI 12B; 9, WS2l62; IO, M'S376; 11, M'S269A; 12. DP2; 13, DP7; isolates also have been identifiedfor these gene 14,DP3645; 15, DMUC2;16, DM6.1; 17. SC7120; 18, phage lambda negative control. Slots 1 through 11 are of isolates previ- families (WELKERet al. 1986; POOLEand FIRTEL1984; ously designated D.discoideum (AX3K is a derivative of NC4); lanes D. L. WELKER,unpublished results). Analysis of these 12 through 14 are of D. purpureum isolates; lanes 15 through 17 genes was expected to reveal differences both in the are of D. mucoroides isolates. presence or absence of hybridizing sequences and in thelength of restriction fragmentscarrying these sequences. hybridization patterns in the slot blots were similar The copy numbers of actin genes in thefour to that of KC4 but which also had unusual mating atypical isolates were similar to that in NC4 but the characteristics (ERDOS,RAPER and VOCEN 1973; ROB- pattern ofHind111 fragments hybridizing with the SON andWILLIAMS 19i9, 1980);these were sexual ACTIN4 probe were markedly different from that isolates withweak matA characteristics (WS269A, observed with KC4 (Figure 3). AC4 and ZA3A had WS380B and WS576) and bisexual isolates (WSl12B uniquepatterns while thepatterns inWS526 and and WS2 162). Ethidium bromide staining revealed WS584 shared numerous bands. While the other D. that the pattern of high copy number fragments of discoideum isolates, as expected,differed somewhat AC4, WS526, WS584 and ZA3A differed distinctly from NC4, most restriction fragments in these isolates from those of the D. discoideum type strain NC4; the were indistinguishable insize from those of KC4 restriction fragment patterns of the otherfive isolates (Figure 3). were more similar to that of NC4 (Figure 2A). These Isolates of D. purpureum and D. mucoroides had differences reflect polymorphisms affecting ribo- similar actin genenumbers to that found in D. somal DNA, mitochondrial DNA and other repeated discoideum but, as expected, the pattern of restriction DNAs. fragments was distinctly different from that in NC4 Confirming the results of the slot blots, the four (Figure 3). As on the ethidium bromide stained gel, atypical isolates each contained an EcoRI restriction fragment patterns of D. purfmreum isolates DP2 and fragment that hvbridized to the pEcoVII probe (Fig- DPi were related; DP3645 hada pattern distinct ure 2B). However, this fragment was larger in ZA3A from that of DP2 and DPi. The pattern of each of and smaller in AC4 than in NC4. The pCTl probe the three D. mucoroides isolates differedfrom each detected nofragment in AC4, WS526 and WS584 other. while a larger fragment than that of KC4 was ob- The discoidin probe (pDd8 12) detected more ex- served in ZA3A (Figure2C). The pEcoIV probe tensive polymorphism among the D. discoideum iso- detected no fragment in WS526 or WS584 but frag- lates. However, each of the fiveisolates (WS112B, ments that hvbridized to a lesser degree and were of WS269A, WS380B, WS576, and WS2162) we tested different size to that in NC4 were observed in AC4 apparentlyshared a pair of fragments withNC4 and ZA3A (Figure 2D). (Figure 4). This pair of fragments was missing from Dictwstelium DNA Taxonomy 363

1 2 3 4 5 6 7 8 9 lOll1213l415~~ I

1 2 3 4 5 6 7 8 9 lOn 121314151617

B D FIGURE2.-EcoRI restriction fragment length analysis of representative Dict?oste/ium isolates using ethidium-bromide staining (A) and probes for ribosomal gene-coding (pEcoVII, B) and spacer DSA (pCT1, C; pEcoIV, D) regions. Lanes on all panels are as follows: 1, SC4; 2, AC4; 3, ZA3A; 4, MrS526; 5, WS584; 6,M’S380B; 7, M’SI 12B; 8, M‘S2162; 9, MrS376; 10, WS269A: 1 I, DP2; 12. DP7; 13, DP364.i; 14, DMUC2; 15, DM6.1; 16, SC7120; 17, phage lambda negative control. Lanes 1 through 10 are of isolates previously designated D. discoideum; lanes 11 through 13 are of D.purpureum isolates: lanes 14 through 16 are of D. mucoroides isolates. The sizes of EcoRI fragments of the SC4 ribosomal palindrome are indicated on the left of the panels: in panel D the homologous fragments in AC4 are indicated by the open circles.

AC4, WS526, WS584, ZASA and isolates of D. pur- the isolates of D.purpureum and D. mucoroides are to pureum and D. mucoroides (Figure 4). In other work, D. discoideum. There was little orno hybridization the 1.8 kb fragment was shown to vary in sizein with D.purpureum and D.mucoroides isolates (data not some D. discoideum isolates butno polymorphisms shown) but there was some hybridization with each affecting the 1.4-kb fragment have been observed of the fouratypical isolates. ZAJA hybridized strongly (D. L. WELKER,unpublished results). D. purpureum with bothprobes andthe major hybridizing band and D. mucoroides isolates contained fragments that was indistinguishable in size from that observed with hybridized with the discoidin geneprobe but the D. discoideum isolates. AC4 hybridized weakly to both weaker hybridization tothese fragments indicated probes but the majorhybridizing band observed with that there were differences in the DIVA sequence. these probes was again similar in size to that seen in WS526 and WS584 had related patternsof fragments NC4. WS526 and WS584 also hybridized weakly to as did the D.purpureum isolates DP2 and DP7. AC4, both probes. However, the size of the major hybrid- ZASA and the remaining D. purpureum and D. mu- izing band in each case was smaller than in NC4. coroides isolates had unique banding patterns. The sizes of these bands in WS526 and WS584 were Restriction fragment length analysis using probes similar. for the DIRS-1 transposable element: Information obtained using probes forthe internal DIRS-1 coding DISCUSSION sequence and for the DIRS-1 terminal repeats was consistent with thedata obtained with theother Our results indicate that 4 of 27 wildisolates probes (Figure 3). However, results suggest that the classified as D.discoideum on the basis of morpholog- four atypical isolates (AC4, ZASA,WS326 and ical criteria (AC4,WS526, WS384, ZASA) differ WS584) are more closely related to D.discoideum than extensively from the D. discoideum type strain NC4 366 M’. B. Evans, J. E. Hughes and D. L. M’elker 1 2 3 4 5 6 7 8 9 10111213l415l6 l7 12345678910

c a-.

Io- c

52-

3.5-

A I m c 12345678910

FIGURE3,”HindIII restriction fragment length analysis of representative Dictystelium isolates using the ACTIS-8 probe. Lanes are as follows: I, SC4; 2, AC4; 3, ZA3A: 4, WS526; 5, CVS584; 6, IVSSXOB: 7, WSIISB;8, M’S2162; 9, M’S576; IO, WS269A; 11, DP2; 12, DP7; IS, DP364.i; 14, DMUCZ; I.?, DM6.1: 16, SC7120; 4- 17, phagelambda negative control. Lanes 1 through IO areof isolates previously designated D. dlscoideurn: lanes 11 through 13 are of D.pttrpure~tn isolates: lanes 14 through 16 are of D. mucoroides isolates. Approximate size markers are indicated on the left.

1 2 3 4 5 6 7 8 9 loll 1213141516

FIGURE5.-EcoRI restriction fragment length analysis of rep- lo resentative isolates previously designated D. di.ccoidettm using the DIRS-1 probes pB41-6 (A) and GSI45A (B). Lanes are as follows: 72, 66 I, SC4; 2, AC4; C, USA; 4, WS.526; 5, WS584; 6, IVS38OB; 7, WSI 12B; 8. WS2162; 9, WS576; IO, WS269A. Approximate size of the major band is indicated on the left.

discoideum isolates can potentially exchangegenetic 149 material by either sexual or parasexual means. All 1.4 those isolates experimentally shown to be capable of macrocyst formation or ofparasexual diploid formation with known D. disc0ideu.m tester strains (NC4 and V12 derivatives) have extensive sequence simi- FIGURF. 4,“EroRI restriction fragmen: length analysis of representative Dict~octrliumisolates using the pDcIXI2 discoidin I gene larity that is readily detected using a variety of probes probe. Lanes are as follows: 1, SC4; 2, AC4; 3, ZA3A: 4. iVS.526; for both noncoding and coding sequences. All four 5, M’S584: 6, \VSSXOB: 7, I\’S112B: 8, M’S2162; 9. WS576; IO, isolates for which the test of genetic exchange fails, iVS269A: 11, DP2: 12. DP7; 13, DP3645; 14, DSIUC2; 15, DM6.1; eitherdue to an asexualresponse (WS526 and 16, SC7120. Lanes I through IO are of isolates previously desig- WS584) or due to an homothallic response (AC4 and nated D. discoidrum: lanes 11 through 13 are of D. purpureum isolates: lanes 14 through 16 are of D. mucoroides isolates. Size of ZA3A) have extensive differences from the type D. the SC4 fragments are indicated on the left. discoideum strain (Table 2). If we consider together all restrictionfragments recognized by the probes used, isolates of D. discoideum (WS380B,WS112B, and other D. discoideum isolates. These findings are WS2162, WS576 and WS269A) share a minimum of consistent with data based on a set of allozvme and 80% of the fragments with NC4. The four atypical cell surface proteins (BRISCOEet al. 198’i). Our inter- isolates share only 12-36Yi of fragments with NC4; pretation is that these four isolates are of species this is comparable to the numberof fragments shared other than D.discoideum. The concept of a biological by isolates of D.purpureum (12-24%) or of D. muco- species being “the members in aggregate of a group roides (20-40%) with NC4. These values are similar of populations that interbreed or potentially inter- to those obtained in a comparison of allozyme elec- breed with each otherunder natural conditions” trophoretic mobilities with these isolates (BRISCOEet (FUTUYJIA1986) is applicable to D. discoidwm. D. al. 1987). Dictyostelium DNA Taxonomy 567

TABLE 2 Number of fragments apparently shared with NC4 in other isolates

No. of Isolate NC4 Probe fragments AC4 ZA3A WS526 WS584 WS380B WSllPB WS2162 WS576 WS269A DPZ DP7 DP3643 DMUCZ DM6.1 NC7120 pEcoVII 1 0011 1 1 1 1 1 111 1 1 1 pCT 1 1 000 0 1 1 1 1 1 00 0 0 00 pEcoIV 1 000 0 1 1 1 1 1 00 0 0 00 ACTIN-8 16 276 5 14 14 14 14 13 42 5 8 43 pDd812 5 011 0 3 3 2 3 4 00 0 1 01 GM45A" 1 110 0 1 1 1 1 1 00 0 0 00 Data for central DIRS-1 fragment only.

Geneticexchange between strains of a species other species, primarily because one is unsure that modulates theeffects of genetic changes, for example fragments of similar size come from homologous by facilitating their spread throughout the species. regions of the genome rather than having a similar The presence of a set of similar restriction fragments size due to chance. The data presented in Table 2 is in two strains is a direct indication of shared genetic consequently an underestimate of the genetic change. material. The absence of similar fragments in strains The approach of BRISCOEand coworkers (1987) presumed to have a common ancestor is an unequiv- provides a betterbasis for comparison of polymorph- ocal indication of the accumulationof genetic change. isms in Dictyostelium to those in other species; we The evidence for the accumulation of extensive ge- refer the reader to their discussion for such a com- neticchange uncovered in our work andthat of parison. For quantitation it would be better to use a BRISCOEand coworkers (1987) coupled with the in- collection of uniquesequence probes so thatno ability of AC4, ZASA, WS526and WS584 to exchange ambiguity exists in the identification of homologous geneticmaterial with known D. discoideum strains fragments. Of the probes we used the rDNA spacer provides a compelling argument for the removal of and rDNA gene coding probes as well as the DIRS- these four isolates from the species D. discoideum. 1 central fragment probe best fit this criterion. We Conversely, our data also support the findings of wish to point out that the different probes pick out BRISCOEand coworkers (1987) that the four atypical different degrees of polymorphism. The sequences isolates are more related to D. discoideum than are surrounding different gene families in D. discoideum isolates identified as D.purpureum and D. mucoroides. have diverged to different degrees. Restriction frag- All four isolates retained sequences sufficiently sim- ments bearing actin genes are moreconserved among ilar to hybridize with some or all of the probes used, D. discoideum isolates thanare fragments bearing while with isolates of D.purpureum and D. mucoroides discoidin genes (this work, WELKERet al. 1986; D. L. hybridization to fewer of the probes was found. The WELKER,unpublished data). Restriction fragments extent of hybridization and the sizes of the homolo- bearing tRNA genes or copies of the Tdd-2 element gous fragments observed with the rDNA and DIRS- are even less conserved (DINGERMANNet al. 1987; D. 1 probes suggest thatZAJA is most closely related to L. WELKER unpublishedresults). These highly poly- D. discoideum, AC4 is somewhat less related,and morphic fragments may be useful for studying in- WS526 and WS584 are least related to D.discoideum. traspecific relationships. Preliminary results with This sequence similarity is consistent with the prior uniquegene probes such as those for myosin or classification of these isolates as D. discoideum; these calmodulin indicate that sequences surrounding es- isolates are expectedto have some affinity to D. sential single copy genes, not unexpectedly, have a discoideum unless their similar morphologicalchar- much higher degree of conservation (D. L. WELKER, acteristics aroseindependently. We interpret these unpublishedresults). These differences in conser- results as implying descent of D. discoideum and the vation of fragmentlength reflect the nonessential four atypical isolates from a common ancestor not nature of specific members of multigene families shared by the species D.purpureum or D.mucoroides. versus the requirementto maintain a functionalsingle Our investigation was set up tobe a qualitative copy gene. Differences in the apparent rateof change study of genetic relatedness among these isolates. For observed with probes for different gene families are this, probes that identify restriction fragments bear- not unexpected considering that characters in other ing members of multigene families are excellent. The species have also been shown to evolve at different differences observed with the actin probe and the rates. discoidin probe are striking. However, it is difficult Of the four atypical isolates, WS526 and WS584 to quantitate the data for comparison to work with were closely related to each other. Except for a few 568 W. B. Evans, J. E. Hughes and D. L. Welker differences in the sizes of high copy number EcoRI However,numerous differences were detected by restriction fragments and of fragments hybridizing ethidiumbromide staining and the gene family to the actin and discoidin probes,patterns were probes in the small set of D. purpureum and D. identical for this pair of isolates. Using probes and mucoroides isolates that we analyzed in more detail. monoclonal antibodies to study a-actinin, WITKEand These findings and supporting data of BRISCOEand coworkers(1986) observed that these two isolates coworkers (1987) suggest that isolates presently des- were similar to each other but distinct from a set of ignated D.purpureum and D.mucoroides each may also D. discoideum isolates. Likewise BRISCOEand cowork- represent several species. This is not surprising con- ers (1987) found nodifferences affectingthe mobility sidering the almost worldwide distribution of isolates of enzymes and no differences in the reaction with and the rangeof sexual mating characteristics attrib- monoclonalantibodies to cell surfaceproteins be- uted to these species (RAPER1984). We are presently tween WS526 and WS584. On the basis of the evi- studying these species in more detail and our results dence of close genetic relatedness, we conclude that to date are consistent with this conjecture. WS526 and WS584 are isolates of a single species. We thank K. L. WILLIAMS(MacQuarie University, Australia) for Indeed WS526 and WS584 are closer genetically to communication of unpublished results and M. LARCENTfor pre- each otherthan most of the sexual D. discoideum paring the manuscript. This work was supported by Public Health isolates are to NC4. Service grant GM35167 from the National Institutes of Health AC4 and ZA3A were differentfrom all other and by Utah Agricultural Experiment Station project UTAGOIC. isolates examined andappear to represent single This paper is published with the approval of the Director, Utah Agricultural Experiment Station, Utah State University, Logan, isolates of two additional species. This pair of isolates Utah, as journal paper No. 3491. had distinctly different patterns of high copy number DNA fragments from other isolates, and the sizes of EcoRI fragments that hybridized to specific probes LITERATURECITED for regions of the rDNA palindrome were unique. APPELS,R., and J. DVORAK,1982 Relative rates of divergence of In particular, sizes of fragments hybridizing with spacer and gene sequences within the rDNA region of species pEcoVII, which carries portions of the NC4 rRNA in the Triticeae: implications for the maintenance of homoge- genes,differed from sizes observed inall other neity of a repeated genefamily. Theor. Appl. Genet. 63: 361- isolates. Unlike WS526 and WS584, AC4 and ZA3A 365. AVISE,J. C., R.A. LANSMAN andR. 0. SHADE,1979 The use of had sequences that hybridized with the rDNA spacer restriction endonucleases to measure mitochondrial sequence probes from NC4. Both AC4 and ZA3A hybridized relatedness in natural populations. I. population structure and to pEcoIV; ZA3A also hybridized to pCT1. But again evaluation in the genus Peromyscw. Genetics 92: 279-295. the sizes of the hybridizing fragments differed from BORST,P., F. FASE-FOWLER,J. HOEIJMAKERSand A. FRASCH, those in all other isolates. ZA3A and AC4 also had 1980 Variations in maxi-circle and mini-circle sequences in kinetoplast DNAs from different TTfalzosoma brucei strains. unique patterns of fragments that hybridized to the Biochim. Biophys. Acta 610: 197-210. actin and discoidin probes. The data obtained with BRISCOE,D. A,, A.A. GOOLEY,R. L. BERNSTEIN,G. M. MCKAY the DIRS-1 probes are consistent with the hypothesis and K. L. WILLIAMS,1987 Genetic diversity in cellular slime that ZA3A and AC4 represent separate species. With molds: allozyme electrophoresis and a monoclonal antibody these probes the amount of hybridization to ZA3A reveal cryptic species among D. discoideum strains. Genetics 117: 213-220. was similar to that of D.discoideum isolates, as was the CAPPELLO,J., S. M. COHEN andH. F. LODISH,1984 Dictyosteliurn size of the major hybridizing band. While AC4 re- transposable element DIRS-1 preferentially inserts into DIRS- tained hybridization to DIRS-1 probes and the size 1 sequences. Mol. Cell. Biol. 4: 2207-2213. of the major hybridizing band was similar to that in CHUNG,S., C. ZUKERand H. F. LODISH,1983 A repetitive and apparently transposable DNA sequence in Dictyostelium discoi- NC4, theamount of hybridization was decreased. deum associated with developmentally regulated RNA’s. Nu- The differences observed with the DIRS-1 probe are cleic Acids Res. ll: 4835-4852. in keeping with previously characterizedsequence COCKBURN,A. F., W. C. TAYLORand R. A. FIRTEL, divergence of DIRS- 1 elementsisolated from a single 1978 Dictyostelium rDNA consists of nonchromosomal palin- D. discoideum strain; these included both single base dromic dimers containing 5s and 36s coding regions. Chro- mosoma 70: 19-29. pair changes as well as deletions (ZUKERet al. 1984; DELOZANNE,A. and J. A. SPUDICH,1987 Disruption of the Dic- CAPPELLO,COHEN andLODISH 1984). The variation tyostelium myosin heavy chain gene by homologous recombi- in DIRS-1 copy number is also consistent with varia- nation. Science 236: 1086- 1091. tions in copy numbers of transposable elements seen DEVINE,J. M., A. S. TSANCand J. G. WILLIAMS,1982 Differential with other species. expression of the members of the discoidin I multigene family during growth and development of Dictyostelium discoideum. Isolates presently designated D.purpureum and D. Cell 28: 793-800. mucoroides hadno detectable hybridization to the DINGERMANN,T., E. AMON,K. L. WILLIAMS andD. L. WELKER, DIRS-1 probes or therDNA spacer probes but as 1987 Chromosomal mapping of tRNA genes from Dictyoste- expecteddid hybridize tothe rRNA gene probe. lzum discoideum. Mol. Gen. Genet. 207: 176-187. Dictyostelium DNA Taxonomy 569

DUTTA,S. K. (Editor), 1986 DNA Systematics. CRC Press, Boca RAPER,K. B., 1984 The Dictyostelids. Princeton University Press, Raton, Fla. Princeton, N.J. EMERY,H. S., and A. M. WEINER,1981 An irregular satellite REED,K. C., and D. A. MANN,1985 Rapid transfer of DNA from sequence is found at the termini of the linear extrachromo- agarose gels to nylon membranes. NucleicAcids. Res. 13: somal rDNA in Dictyostelium discoideum. Cell 26: 41 1-419. 7207-722 1. ERDOS,G. W., K. B. RAPER and L. K. VOCES, 1973 Mating types REYMOND,C. D.,W. NELLENand R.A. FIRTEL,1985 Regulated and macrocyst formation in Dictyostelium discodeurn. Proc. Natl. expression of rcls gene constructs in Dictyostelium transformants. Acad. Sci. USA 70: 1828-1830. Proc. Natl. Acad. Sci. USA 85: 7005-7009. FUTUYMA,D. J., 1986 EvolutionaryBiology, Ed. 2 Sinauer Associ- REYMOND,C. D., R. H.COMER, W. NELLEN,A. THEIBERT,P. ates, Sunderland, Mass. DEVREOTESand R. A. FIRTEL,1986 Phenotypic changes GRANT,W. N.,D. L. WELKERand K.L. WILLIAMS,1985 A induced by a mutated rcls gene during the development of polymorphic prespore-specific cell surface glycoprotein is Dictyostelium transformants. Nature 323: 340-343. present in the extracellular matrix of Dictyostelium discoideum. ROBSON,G. E., and K. L. WILLIAMS,1979 Vegetative incompat- Mol. Cell. Biol. 5: 2559-2566. ibility and the mating-type locus in the cellular slime mold JEFFREYS,A. J., V. WILSONand S. L. THEIN,1985 Hypervariable Dictyostelium discodeurn. Genetics 93: 861-875. minisatellite regions in human DNA. Nature 314: 67-73. ROBSON,G. E., and K. L. WILLIAMS,1980 The mating system of KNECHT,D. A,, and W.F. LOOMIS,1987 Antisense RNA inacti- the cellular slime mould Dictyostelium discoideum. Curr. Genet. vation of myosin heavy chain gene expression in Dictyostelium 1: 229-232. discoideum. Science 236: 1081-1086. ROMANS,P., and R.A. FIRTEL,1985 Organization of the actin LOOMIS,W. F. (Editor), 1982 Dictyostelium discoideum: A Develqp- multigene family of Dictyostelium discoideum and analysis of mental System. Academic Press, New York. variability in the protein coding regions. J. Mol.Biol. 186: MATTHEWS,B.F., and L. R. DEBONTE,1985 Chloroplast and 321-335. mitochondrial DNAs of the carrot and its wild relatives. Plant SOUTHERN,E.M., 1975 Detection of specific sequences among Mol. Biol. Report. 3: 12-16. DNA fragments separated by gel electrophoresis. J. Mol. Biol. MCCARROLL,R., G. J. OLSEN,Y. D. STAHL,C. R. WOESEand M. 98: 503-517. L. SOGIN,1983 Nucleotide sequence of the Dictyostelium dis- SPUDICH,J. A. (Editor), 1987 Dictyostelium discoideum: molecular coideum small subunit ribosomal ribonucleic acid inferred from approaches to cell biology. Methods Cell Biol. 28. the gene sequence: evolutionary implications. Biochemistry SUSSMAN,M., 1966 Biochemical and genetic methods in the study 22: 5858-5868. of cellular slime mold development. Methods Cell Physiol. 2: METZ, B. A., T. E. WARD,D. L. WELKERand K. L. WILLIAMS, 397-410. 1983 Identification of an endogenous plasmid in Dictyostelium WELKER,D. L., 1986 Linkage analysis of nystatin resistance discoideum. EMBO J. 2: 515-519. mutations in Dictyostelium discoideum. Genetics 113: 53-62. MOREL,C., E. CHIARI,E. CAMARGO,D. MATTEI,A. ROMANHA and WELKER,D. L., K. P. HIRTHand K. L. WILLIAMS, L. SIMPSON,1980 Strains and clones of Trypanosoma cruzi can 1985 Examination of the inheritance during the asexual life be characterized by pattern of restriction endonuclease prod- cycle of extrachromosomal ribosomal DNA in Dictyostelium ucts of kinetoplast DNAminicircles. Proc. Natl. Acad.Sci. discoideum using DNA polymorphisms. Mol. Cell. Biol. 5: 273- USA 77: 6810-6814. 280. NELLEN,W., C. SILANand R.A. FIRTEL,1984 DNA-mediated WELKER,D. L., K. P. HIRTH,P. ROMANS,A. NOEGEL,R. A. FIRTEL transformation inDictyostelium discoideum: regulated expression and K. L.WILLIAMS, 1986 The use of restriction fragment of an actin gene fusion. Mol. Cell. Biol. 4 2890-2898. length polymorphisms and DNA duplications to study the NESS, P. J., P. LABHART,E. BANZ,T. KOLLERand R.W. PARISH, organization of the actin multigene family in Dictyostelium 1983 Chromatin structure along the ribosomal DNA of discoideum. Genetics 112: 27-42. Dictyostelium. Regional differences and changes accompanying WITKE, W., M. SCHLEICHER,F. LOTTSPEICHand A. NOEGEL, cell differentiation. J. Mol. Biol. 166: 361-381. 1986 Studies on the transcription, translation, and structure NOEGEL, A., A.B. METZ and K. L. WILLIAMS, of a-actinin in Dictyostelium discoideum. J. Cell. Biol. 103: 969- 1985 Developmentally regulatedtranscription of Dictyoste- 975. lium discoideum plasmid Ddpl. EMBO J. 4: 3797-3803, WOESE,C. R., 1987 Bacterial evolution. Microbiol. Rev. 51: 221- NOEGEL,A., D.L. WELKER,B. A. METZ and K. L. WILLIAMS, 271. 1985 Presence of nuclear associated plasmids in the lower ZUKER,C., and H. F. LODISH,1981 Repetitive DNA sequences eukaryote Dictyostelium discoideum. J. Mol. Biol. 185: 447-450. cotranscribed with developmentally regulated Dictyostelium dis- PODGORSKI,G., and R. A. DEERING,1980 Quantitation of induced coideum RNAs. Proc. Natl. Acad. Sci. USA. 78: 5386-5390. mutation in Dictyostelium discoideum. Characterization and use ZUKER,C., J. CAPPELLO,H. F. LODISH,P. GEORGEand S. CHUNG, of a methanol-resistance mutation assay. Mutat. Res. 74: 459- 1984 Dictyostelium transposable element DIRS- 1 has 350-base- 468. pairinverted terminal repeats that contain aheat shock POOLE,S. J.,and R. A. FIRTEL,1984 Genomic instability and promoter. Proc. Natl. Acad. Sci. USA. 81: 2660-2664. mobile genetic elements in regions surrounding two discoidin Communicating editor: S. L. ALLEN 1 genes of Dictyostelium discozdeum. Mol. Cell. Biol. 4: 67 1-680.