Arrangement of Base Sequences in Deoxyribonucleic Acid' · ment of nucleotides in deoxyribonucleic acid (DNA) may be discussed. At the gross level, a considerable body of information

BACTERIOLOGICAL REvIws, Dec. 1967, p. 215-229 Vol. 31, No.4Copyright © 1967 American Society for Microbiology Printed in U.S.A.

Arrangement of Base Sequences inDeoxyribonucleic Acid'

BRIAN J. MCCARTHYDepart nents of Microbiology and Genetics, University of Washington, Seattle, Washington 98105

INTRODUCrION...........O....STRAND AsSOCIATION BY DNA.................................................

General Considerations.......................................................

Viral DNA.................................................................Bacterial DNA..............................................................Mammalian DNA...........................................................Partially Redundant Nucleotide Sequences ....................................

COmERVATSM IN THE EVOLUnON OF BASE SEQUENCES............................INTEACION OF OLIGONUCLEOTIDES wrrTH DNA..................................

Specificity ofInteraction ...................................................Estimation ofthe Degree ofMismatching ofBases...............................

EVOLUTION AND 1ECOMBINATION oF DNA MOLECULEs ........................LIERATURE CMD............................................................

215216216217218219220221223223224225227

INTRODUCrION

There are several levels at which the arrange-ment of nucleotides in deoxyribonucleic acid(DNA) may be discussed. At the gross level, aconsiderable body of information exists con-cerning the overall base composition of DNA.This has proven valuable for taxonomic pur-poses, at least in microorganisms, and providessome comparative information as to the relativevariability of the composition of DNA amongvarious groups of organisms. In higher plantsand metazoans, the base composition is rela-tively constant, so that these measurements arenot sensitive enough for comparative purposes.At a higher level of resolution, nearest-neighborrelationships may be determined with the aid ofDNA polymerase (21), and this may be extendedto larger nucleotide sequences by allowing theincorporation of a mixture of ribonucleotidesand deoxynucleotides into the product in thepresence of Mn++ (4). In the case of nearest-neighbor frequencies, 16 parameters are pro-vided for the characterization of a given DNA.The method has been used for evolutionarystudies with viral DNA (2, 37), although it isclearly limited in sensitivity.The ultimate description of a DNA molecule or

of a gene is in its complete base sequence. Com-plete nucleotide sequence data are available onlyfor a few soluble ribonucleic acid (sRNA) mole-cules (18, 41, 57). There is no general method forthe isolation of a single gene, except through thepurification of a single messenger which in turn is

Eli Lilly Award Address (1967).

possible only in simple viral systems (1). Whenthis is achieved, sequence determination is muchmore difficult than amino acid sequence deter-mination in proteins, for a number of obviousreasons. There are three times the number ofresidues in the messenger ribonucleic acid(mRNA) or a single strand of the DNA ascompared to the gene product, and only fourkinds of bases compared to 20 kinds of aminoacids. The existence of minor bases in sRNA hasbeen as valuable to the determination of sequenceas the small number of residues. Again, highlyspecific enzyme degradation methods other thanlimited digestion are not available.At the moment, therefore, the problem of

complete sequence determination of even thesmaUest viral nucleic acids is far from solution.Fortunately, however, the great power of nucleo-tide sequence recognition which nucleic acidsthemselves possess may be used as a tool fordetailed comparison of related DNA and ribonu-cleic acid (RNA) molecules. After the discoveryby Doty et al. (12) and Marmur and Lane (24)that the two strands of DNA may be separatedand reannealed, it became clear that this reactioncould be extremely valuable for comparisons ofrelated nucleic acids as well as the analysis ofRNA produced from the DNA. Thus, the forma-tion of heteroduplex DNA molecules by theseparated DNA strands originating from twoorganisms has become an accepted method intaxonomy (19, 26, 31). This method is appro-priate with any group of DNA molecules frombacteria to primates. An example of these resultsis shown in Table 1, in which various enterobac-

215

on August 30, 2020 by guest

http://mm

br.asm.org/

Dow

nloaded from

http://mmbr.asm.org/

BACrERIOL. REV.

TABLE 1. Reaction among heterologous DNApreparations

DNA prepn Reaction

NoBacteriala

Escherichia coli............. 100Salmonella typhimurium ........... 71Aerobacter aerogenes.............. 51Proteus vulgaris................... 14Serratia marcescens ............... 7Pseudomonas aeruginosa........... 1Bacillus subtilis.................... 1

MammalianbRhesus monkey ............ 100Human ........................... 76Chimpanzee....................... 76Owl monkey ...................... 68Potto............................ 50Mouse............................ 22Chicken .......................... 11Salmon........................... 5E. coli............................ 0

o From McCarthy and Bolton (31). Reactionsexpressed relative to the homologous reaction byE. coli DNA.

b From Hoyer et al. (20). Reactions expressedrelative to the homologous reaction by rhesusmonkey DNA.

terial DNA preparations are ordered accordingto the similarity of their base sequence with thatof Escherichia coli DNA (31), and primate DNApreparations are compared with those of therhesus monkey (20).The ensuing paragraphs will be concerned not

so much with these applications of the techniques,but rather with the arrangement of differentnucleotide sequences in a genome. As will beclear later, the reaction of two strands ofhomologousDNA can provide useful informationas to the frequency of occurrence of a givensequence in the whole genome and the incidenceof similar, but not identical, sequences. Theformation of structures other than true renaturedDNA is a function of the kinds of related se-quences which exist in that particular DNA.

STRAND ASsoCIATION BY DNAGeneral Consideration

The reassociation of two DNA strands to formrenatured DNA may be studied by a variety ofmethods. These include changes in hypochromic-ity and other optical parameters, buoyantdensity, and biological activity (26). None of theseis suited to studies of the interaction of twostrands originating from two organisms or of

interactions which form products other thanrenatured DNA. The density gradient techniquemay be modified to study interspecific DNAreactions by labeling one DNA with a heavyisotope (12). However, more convenient means ofstudying all of these various types of reassociationreactions are provided by methods in which aradio-labeled DNA is allowed to interact with ahomologous or heterologous DNA immobilizedin a solid phase. The latest and most convenientversion of this approach is offered by Denhardt's(10) modification of the membrane filter tech-nique of Gillespie and Spiegelman (15), originallydevised for assay of DNA-RNA hybrids.At reasonably low concentrations of DNA

(<10 Ag/ml), renaturation of DNA in solutionfollows second-order kinetics, as would be ex-pected from the interaction of separated DNAstrands (49). At higher concentrations or in highionic strength, aggregation of renatured DNAcomplicates the kinetics (48, 49). The rate of rena-turation is clearly dependent upon the number ofdifferent base sequences in the DNA. Thus,DNA representing a simple genome, such as thatof a virus, renatures more rapidly than that of amore complex bacterial or mammalian genomewhen incubated at the same concentration inaccordance with the higher concentration ofparticular complementary elements (6, 25). Underthe nonideal, but more convenient, conditions ofreaction of labeled DNA in solution with un-labeled filter-bound DNA, the kinetics are morecomplex. As shown in Fig. 1, the rate of reactionis proportional to the concentration of radio-labeled DNA in solution. With small amounts oflabeled DNA in solution where the reactionbetween two labeled DNA strands is unimportant,the reaction rate is proportional to the amountof DNA adsorbed on the filter over a limitedrange. With larger amounts of filter-bound DNA,the reaction rate is not limited by the quantity ofimmobilized DNA (Fig. 2). These facts com-plicate the interpretation of reaction rate data,although, as will be discussed later, the rate ofreaction does depend on the relative homogeneityof the DNA sample in question.Two important kinds of measurement applied

to various types of DNA will be discussed. Therate of reassociation of two strands at varioustemperatures is a measure of the number ofelements or sequences in that particular DNAwhich are similar enough to form a duplexstructure. When true renaturation is being meas-ured, the rate of reaction is proportional to thecomplexity of the DNA or the genome sizemeasured in nucleotide residues. Under lessstringent conditions, such as lower temperatures,the rate depends upon the incidence of similar,

216 McCARTHY


http://mm

br.asm.org/

Dow

nloaded from


BASE SEQUENCES IN DNA

solution at the optimum temperature, 56 C, to0.15 give structures indistinguishable from native

DNA (48, 49). This process has a broad optimumo.I0-/ temperature and does not take place above 75 C

or below 40 C. Similar effects can be seen in the0.025~14 ~Lg/mI reaction of labeled DNA with filter-bound DNA.0.05 A The rate of reaction shows an optimum at about

0.10 60 to 65 C and falls off at lower or higher tempera-

Az.^/ tures (Fig. 4). At about 40 C, there is a suggestiont 0.08 ~ jug5 10 15 of a second component, although, as shown by the

,i±g DNA /mI rate of reaction with heterologous DNA, this isnot species-specific. In fact, the rate of reaction

0.06 ~~~~~~6.5kug/ml at 30 C is faster with B. subtilis or mouse DNAthan with homologous DNA, since their greater

complexity permits an increased possibility of0.04- 3.5/ig/m complementary pair formation over short regions.

/ A characterization of the reaction products is0.02 >/ _g , o illustrated in Fig. 5. At 40 C or above, the thermal

0.65 pg/ml stability of the products formed are identical,having a melting temperature (Tm) of approxi-

10 20 30 40 50 60 mately 82 C. This is close to that expected on theTime ( minutes) basis of the Tm for native DNA, 85 C, and a

decrease of about 3 C resulting from the lowerFIG. 1. Dependence orthe rate ofreaction ofsheared meculare of the sheared DNA (hendenatured DNA with filter-bound denatured DNA on molecularwedght of the sheared DNA (6). Whenthe concentration ofDNA in solution. Various quantities incubated at 25 C, only complexes of low stabilityof 2P-labeled sheared T4 DNA were incubated at 60C are formed, and these are also formed withwith a S-mm filter containing 12,ug of denatured T4DNA in 0.2 ml ofI X SSC (see Fig. 3). The amount ofcomplex formed was assayed by counting filters re- 0.15moved at various times. /

/but not identical, sequences distributed through- 0.10I -0.10out the genome. C3%

In addition, the properties of the complex 4_formed under various conditions may be assessed 0.05-by means of its thermal stability. As in the case 36kLgof denaturation of DNA in solution, the tempera- 0.075 20 3 40ture at which the duplex dissociates is a function DNA (0 g) /of its base composition and the degree of base 248Lgpairing which exists. The thermal dissociation of r3i

duplexes formed by incubating labeled denaturedBacillus subtilis DNA with homologous filter- 005 12 gbound DNA is illustrated in Fig. 3. The tempera-ture of dissociation depends on the ionic strength kof the solution. In the following pages, examplesof these two types of measurement will be given

0

with simple and complex DNA. The guanosine 0.025 3/igplus cytosine (GC) percentage values of thethree DNA preparations to be discussed mostfully are quite similar, so that differences are

ascribable to complexity rather than to basecomposition. Values usually given for the three l0 20 30 40 50 60are: T4 bacteriophage DNA, 36% GC; B. subtilis Time ( minutes)DNA, 42%; mouse DNA, 41 %. FIG. 2. Dependence ofthe rate ofreaction ofsheared,

Viral DNA denaturedDNA withfilter-bound denaturedDNA on theamount of filter-bound DNA. 2P-labeled, sheared T4

Under conditions of low concentration and DNA (1.2ug) was incubated with various quantities ofmoderate ionic strength, T4 DNA renatures in filter-boundDNA at 60 C in 0.2 ml of1 X SSC.

217VOL.31, 1967

r


http://mm

br.asm.org/

Dow

nloaded from


MCCARTHY

loo0 heterologous DNA. There is no evidence forC 14 B. subtilis < w >/ complexes of intermediate stability which would

result from partial homology of regions in80/ different parts of the phage chromosome. In fact,>*/ */ { f / it appears that no two regions of the phage

genome are any more similar in base sequenceE60 ~ 0.1 0.25 0.5 0.O 2.0 than would be expected on a random basis.

Associations between these give structures with aTm of about 35 C (Fig. 5).

q40 0Bacterial DNA

20/ Experiments similar to those described in theprevious section but with B. subtilis DNA givesomewhat different results (Fig. 6). In the higher

60 70 80 90 100) temperature range, reaction of labeled DNA with60 mpe70 tu80 90 1 the filter-bound DNA shows an optimum at theTemperature °C

FIG. 3. Thermal denaturation profiles of duplexformed by incubating I ,ug of '4C-labeled, sheared, anddenatured Bacillus subtilis DNA with filters (5 mm in 100iodiameter) containing 10 ,ug of denatured B. subtilis T4 Phage DNADNA. Incubations were carried out in 0.20 ml of 2 X oX fSSC at 67 Cfor 15 hr. The duplex was dissociated by 80 25Cheating the filters in 2-ml volumes ofsolutions ofsaline- 4600Ccitrate ofthe indicated strength to various temperatures. a 60 / *50°CThe radioactive DNA removed at each temperature was ocollected by precipitation with trichloroacetic acid and /*counted. The numbers on each line indicate the salt cconcentrations in terms ofstandard saline-citrate (SSC), ,i.e., 0.15 M NaCl, 0.015 m sodium citrate. Thus, 0.1 indi- 20cates a solution ofone-tenth these concentrations. A

30 40 50 60 70 80 90 100Temperature IC

5 T4 DNA FIG. 5. Thermal dissociation profiles of duplexesformed by incubating 1.2 ,ug of OP-4abeled T4 DNAwith 12 ,g offilter-bound homologous DNA at various

4 temperatures.

5 B. subtilis DNA

13-

2 I,Zs

|| ~~~~~~~~~~I I 0

20 40 60 80 100Temperature IC

FIG. 4. Rate ofDNA duplex formation by sheared, 20 40 60 80 100denatured DNA of phage T4 with homologous and Temperature OCheterologous DNA at various temperatures. DenaturedDNA (1.2 ,ug) was incubated with 12 ,g offilter-bound FIG. 6. Rate of DNA duplex formation by labeledDNA. Initial reaction rates are expressed as percentages Bacillus subtilisDNA with homologous and heterologousofinput labeledDNA reacting per hour. (A) T4 DNA. DNA at various temperatures. Conditions as in Fig. 2.(0) Bacillus subtilis DNA. (0) Mouse DNA. (@) B. subtilis DNA (0) Mouse DNA. (A) T4 DNA.

218 BACrERioL. REV.


http://mm

br.asm.org/

Dow

nloaded from



60 70 80 90 100Temperature °C

FiG. 7. Thermal denaturation profiles of duplexesformed by incubating labeled Bacillus subtilis DNA withfilter-bound B. subtilis DNA at various temperatureconditions (as in Fig. 3).

expected temperature (25). As the temperature isreduced below 45 C, however, the rate of reactionincreases to a second optimum at 30 to 35 C. Bycomparison with reactions of labeled B. subtilisDNA with heterologous DNA, it appears thatthis component is at least partially species-specific.Thus, the reaction at 40C with homologous DNAis at least 10 times more rapid than that withmouse or phage T4 DNA. At even lower tempera-tures, the heterologous reaction becomes as rapidas the homologous reaction.The thermal stability of the various reaction

products formed is shown in Fig. 7 and 8. Afterincubation at 50, 60, 67, or 75 C, most of thereaction product has the high stability expectedof well-renatured DNA, i.e., with a Tm similar tothat of sheared native DNA (Fig. 7). With in-cubation at 40 C, some of the product has thishigh stability (Fig. 8), but most is of muchlower stability. With incubation at 25 C, thereaction is not species-specific and the stability islow. Thus, there is some evidence for a species-specific recognition of a very distant intragenomehomology unlike that in T4 phage DNA. How-ever, the stability is quite low and the degree ofbase sequence homology existing among differentparts of the bacterial chromosome must be quitelimited.

Mammalian DNAIt is not possible to demonstrate the renatur-

tion of mammalian DNA by the CsCl method.On the other hand, specific strand associationmay be demonstrated by the DNA-agar (19) or

membrane-filter method. It has been clear forsome time, however, that these strand associations,athough species-specific, do not represent theformation of genuine renatured DNA. This isevident from considerations both of the rates ofsuch reactions and of the properties of the com-plexes formed. These facts are illustrated by thedata in Fig. 9 and 10. As is shown in Fig. 9,the rate of reaction at the optimum temperatureis closely similar to that for B. subtilis DNA. Inview of the fact (29) that the DNA content perhaploid genome is some 500 x higher than forB. subtilis (Table 2), this reaction rate is muchgreater than theoretical and cannot represent theassociation of two complementary strands orig-inating from the same genetic site. Either con-siderable duplication of sequences exists or thereactions are not completely specific. Otherdifferences between B. subtilis and mouse DNAoccur in the dependence of reaction rates ontemperature. Although the base compositionsare almost identical, the optimum temperaturefor reaction is lower for mouse DNA. In addition,the rate falls to extremely low levels above 70C, whereas the rate of reactions of B. subtilisDNA falls only slowly.The thermal dissociation profiles for the com-

plexes formed are highly dependent upon thetemperature of incubation (Fig. 10). Higherincubation temperatures give rise to complexesof higher stability. Thus, a great variety of species-specific complexes may be formed whose relative

B. subti/is DNA100 _

1 50-^7jo5q 500~~A

20 40 60 80 100Temperature IC

FiG. 8. Thermal dissociation profiles of duplexesformed by incubating labeled Bacillus subtilis DNAwith filter-bound homologous DNA at low-temperatureconditions (as in Fig. 3).

VOL. 31, 1967 219


http://mm

br.asm.org/

Dow

nloaded from


McCARTHY

Temperature OC

FIG. 9. Rate of DNA duplex formation by labeledmouse DNA with homologous and heterologous DNA asa function of temperature. Conditions as in Fig. 2. (0)Mouse DNA. (AL) Bacillus subtilis DNA.

proportion depends on the chosen reaction con-ditions. It should be noted that none of thesecomplexes has the stability expected for native orrenatured DNA, i.e., Tn = 85 C. These results

are analogous to those reported by Walker andMcClaren (53), Martin and Hoyer (27), andBritten and Kohne (6), obtained from studies ofstrand association by the DNA-agar or hydroxyapatite methods.

Partially Redundant Nucleotide SequencesThe experiments reported with viral and bac-

terial DNA are consistent with the absence ofclosely related nucleotide sequences distributedaround the genome. With T4 DNA, only twokinds of reaction product are formed, one with astability close to that of native DNA and theother with a stability no higher than that ofduplexes formed between two strands of un-related DNA molecules. Thus, one may concludethat no two nucleotide sequences exist whichhave homology greater than that expected on arandom basis. The situation with bacterial DNAis similar except for the evidence for some distanthomologies significantly greater than those ex-isting between two DNA preparations of thesame base composition chosen at random.

In contrast, the experiments reported withmouse DNA show that the predominant reactionproduct formed is one between strands which areonly partially complementary, representing differ-ent regions of the genome which are partially,

TABLE 2. Nucleotide content of various genomes

Organism Mol wt (daltons) Nucleotide pairs Reference

VirusespX174.............................. 0.9 X 106 5 X 10oa Sinsheimer (46)T5 coliphage ............... 84 X 106 130 X 103 Hershey et al. (17)T2 coliphage......... 130 X 106 200 X 103 Rubenstein et al. (44)

BacteriaMycoplasma gallisepticum ............. 0.2 X 109 0.3 X 106 Morowitz et al. (36)Haemophilus influenzae ............... 0.7 X 109 1 X 106 Berns and Thomas (5)Escherichia coli...................... 2.6 X 109 4 X 106 Cairns (7)Pseudomonas sp...................... 2.4 X 109 4 X 106 Park and DeLey (40)Bacillus subtilis .. ................... 2 X 109 3 X 106 Dennis and Wake (9)

1.3 X 109 2 X 106 Massie and Zimm (28)

YeastSaccharomyces cerevisiae............. 13 X 109 20 X 106 Ogur et al. (39)

InvertebratesDrosophila melanogasterb ............. 0.12 X 1012 0.2 X 109 Ritossa and Spiegelman'(43)Sea urchin (Lytechinus) sperm........ 0.6 X 1012 0.8 X 109 Mirsky and Ris (34)

VertebratesChickenb............................. 1.3 X 1012 2.1 X 109 Mirsky and Ris (34)Mouseb .............................. 3 X 1012 4.7 X 109 Vendrely and Vendrely (52)Humanb.............................. 3 X 1012 5.6 X 109 Vendrely and Vendrely (52)

a For double-stranded replicative form.bDiploid cell; all others haploid.

220 BACTERIOL. REV.


http://mm

br.asm.org/

Dow

nloaded from



Tempera tu re °C

FIG. 10. Thermal dissociation profiles of duplexesfonned by incubating labeled mouse DNA with filter-bound homologousDNA at various temperatures.

but not completely, similar in base sequence.Moreover, it is clear that many degrees of sim-ilarity of base sequence exist among many partsof the mouse genome, so that duplexes of variousdegrees of mismatching may be formed. Thisleads to the concept of families of related basesequences in which the individual members arerelated to one another to various degrees (6).It is important to note that not all of the basesequences in mouse DNA are partially redundantin this sense. Britten and Kohne (6) have shownthat about half of the base sequences are uniqueand not members of families of related genes.Thus, if the concentration is high enough and theincubation time long enough, this fraction of theDNA may be completely renatured.

Thus, it is apparent that the interpretation ofstudies ofDNA duplex formation or DNA-RNAhybrid formation requires some caution. Forexample, the reactions among primate DNA prep-arations (Table 1) are not measures of basesequence differences in individual genes of thevarious species but of relationships among fam-ilies of base sequences. The reactions are notcompletely locus-specific, although as demon-strated by these results they do show considerablespecies specificity.

Correspondingly, although DNA-RNA hybridformation is generally thought of as the recon-struction of the association between a gene andits primary gene product, this concept is justifiableonly in viral and bacterial systems (8, 30). Inmammalian systems, the reaction is one betweenan RNA molecule and a DNA molecule similarin base sequence to the DNA site which pro-duced that RNA molecule. Results similar tothose of Fig. 10 are obtained if hybrids are formed

at different temperatures and then dissociated(8). This does not mean, of course, that hybridiza-tion reactions are not useful in the analysis ofRNA synthesis in mammalian systems, but merelythat the results are easily over-interpreted. Forexample, the degree of difference demonstrablebetween the population of RNA molecules in twotissues is highly dependent on the reaction con-ditions. The more stringent the conditions, themore differences appear. Thus, mouse liver andmouse kidney RNA appear quite similar incompetition reactions carried out at 60 C andalmost completely different when assayed at 72C (Church and McCarthy, in preparation). This iseasily interpretable, by comparison with the dataof Fig. 10, as resulting from different degrees ofcross-reaction between the RNA molecules andrelated DNA sequences. This means that thedifferences demonstrated between two popula-tions of RNA molecules of mammalian origin arealways minimal estimates of the actual differences,since the existence of related sequences limitsresolution. In the same way, titration or satura-tion experiments, in which a determination of themaximal proportion of DNA sites which willreact is made (55), are often misinterpreted.Since RNA molecules will react with sites otherthan those responsible for their synthesis, thisassay will provide an overestimate of the fractionof the DNA sites actively synthesizing RNA inthe situation in question. Notwithstanding thesequalifications, the analysis of RNA by hybridiza-tion methods provides the most sensitive meanspresently at hand for comparison of populationsof RNA molecules. Even in the limiting case ofno locus specificity in the reaction of RNA withDNA, the method may be considered as a formof chromatography in which the DNA columncontains a vast variety of different sites, each withits own affinity for a given group of RNA mole-cules. From this point of view it is apparent thatdifferences among RNA populations may oftenbe demonstrable, although the failure to detectdifferences does not establish identity.

CONSERVATISM IN THE EVOLUTION OFBASE SEQUENCES

The evolution of base sequences takes place atdifferent rates at different sites in the genome. Thisis evident from comparative studies of individualproteins in various groups of organisms. Inbirds, for example, certain structural proteinsevolve more rapidly than certain enzymes (45,54). From more direct experiments with DNA, ithas been shown that certain nucleotide sequencesretain some homology between organisms asdistant as fish and primates, while the majorityof the genome has undergone much greater

22 1VOL. 31, 1967


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

b

65 70 75 80 85 90 95 100Temperature IC

.T-

75 80 85 90 95Temperoture IC

FiG. 11. Comparison of the thermal dissociation profiles of hybrids for ned with either Escherichia coli rRNAor E. coli pulse-labeled RNA. E. coli 2P-labeled 23S or 16S RNA (10 ,ug; 9.4 X 105 counts per min per pg) wasincubated with filters containing 50 ,ug of the various DNA preparations. The proffles for 14C (pulse) -lJabeled RNA(179 counts per min per lug) were obtained by incubating 33.3 jug ofRNA with filters containing 50 p&g ofDNAfrom various sources. Pulse-labeled RNA (a); 23S RNA (A); 16S RNA (0). (a) Escherichia coli DNA filters;(b) Serratia marcescens; (c) Aerobacter aerogenes; (d) Proteus morgani. From Moore and McCarthy (35).

change in base sequence (19). In addition, a coreof relatively conserved sequences exists among allmammals (19).Ribosomal RNA (rRNA) and sRNA cistrons

are examples of conservative regions of thegenome. These regions are represented in theDNA of many genomes by some 0.3 and 0.02%of the total, respectively (55, 56). Interspecifichybridization between E. coli sRNA and otherenterobacterial DNA molecules occurs to a muchgreater extent than corresponding hybridizationof pulse-labeled or mRNA (16). In Bacillus,enterobacteria, and myxobacteria, base sequencesin rRNA are conservative relative to the total(11, 13, 35, 50). This is exemplified by the data ofFig. 11, showing a comparison between thethermal stability profiles of hybrids with DNAformed by E. coli pulse-labeled 16S and 23SrRNA (35). In each case, heterologous hybridsformed by rRNA show smaller differences instability, compared with homologous hybrids,than is the case for pulse-labeled RNA (35).

Similar results were obtained with Myxococcusxanthus RNA. Since, in this case, the base com-position of the mRNA is richer in GC than isrRNA, the higher cross-reaction of rRNA ismore readily attributable to conservatism in basesequence in these cistrons than to its intrinsicbase composition (35).Comparisons of the hybridization properties

of rRNA of mammals and bacteria provideresults quite similar to those obtained from studiesofDNA duplex formation described above. Thus,the Tm of E. coli rRNA-DNA hybrids is close tothat expected from its base composition (Mooreand McCarthy, in preparation). The same thermalstability is obtained if the hybrids are treatedwith ribonuclease to remove grossly unpaired re-gions (15). The thermal stability of the hybrid isnot dependent on the temperature of incubation(Fig. 12). On the other hand, the thermal stabilityof rabbit rRNA-DNA hybrids is lower, eventhough the base composition is higher in GC andthe stability of the remaining complex is increased

100

222 MCCARTHY


http://mm

br.asm.org/

Dow

nloaded from



E. co// Rabbit

90 95 100 60 65 70 75Temperature 'C

80 85 90 95 100

FIG. 12. Thermal dissociation profies ofDNA-RNA hybrids formed by Escherichia coli 23S ribosomal RNAor rabbit 28S ribosomal RNA with homologous DNA by incubating at various temperatures. Profiles were de-termined with or without treatment with ribonuclease to eliminate some mispaired structures.

after ribonuclease treatment. Moreover, as withDNA-DNA duplexes of mammalian origin, theproperties of the hybrid depend on the reactionconditions employed during its formation (Fig.12). These facts suggest that the many rRNAcistrons (about 500) in mammalian DNA aresimilar, but not identical, in sequence, so thathybridization occurs between rRNA having onebase sequence and a DNA base sequence repre-senting another rRNA species, a situation entirelyanalogous to that observed with DNA-DNAduplex formation.

IERACrIoNoF OLIGoNucLEOrIDEs wrri DNA

Specificity of InteractionOne of the obvious questions which arises from

these experiments is: how does one correlatedifferences in the stability of duplexes with thenumbers of complementary and noncomplemen-tary bases in them? As mentioned at the outset,our aim is to define differences and similarities inbase sequence between two genes of one organismor between corresponding genes in related or-ganisms. This would furnish valuable informationregarding the rate of evolution of base sequencesin various groups of organisms and in variousparts of the genome.

One-approach involves the study of the inter-action of fragments of DNA or oligodeoxyribo-nucleotides with DNA. Chemical methods ofdegradation and fractionation allow the prepara-tion of oligonucleotides of known chain lengthrepresenting all the base sequences of the parentalDNA of that length (32). These may be used inreactions with single-stranded DNA to determinethe stability and specificity of such interactions.An example of this approach is given in Fig. 13.

A mixture of mouse and E. coli oligonucleotidesofmean chain length, n = 33, were incubated withboth single-stranded parental DNA preparations.The complexes were then subjected to increasingtemperature to dissociate them in the usual way.As is clear from the data, some specificity existsin the amount of complex formed (as representedby the 25 C points), and this increases as lesswell-paired structures are dissociated; at hightemperatures, only species-specific complexes re-main stable. Thus, in this size range, it is quitepossible to demonstrate specificity.Another experiment demonstrated that small

fragments (of various sizes) of T4 phage DNAreact nonspecifically and that specificity increasesvery rapidly with size (Fig. 14). Oligonucleotidesderived from phage T4 DNA, in the size range

VOL. 31, 1967 223


http://mm

br.asm.org/

Dow

nloaded from


BACTER1oL. REV.

Mouse DNA.tI

q)I.

0I80- 0

0

60-

40 \

20-

20 40 60

igonucleotidesMouseE co/i

E. co/i DNA

40 60Temperature 0C

FIG. 13. Thermal dissociation profiles of complexesformed by slow cooling 0.2 ,ug of a mixture of 14C.labeled mouse oligodeoxynucleotides and 32P-labeledEscherichia coli oligodeoxynucleotides of chain lengthn = 33 with 100 ,g of mouse or E. coli DNA. FromMcConaughy and McCarthy (32).

n = 11 to 28, were incubated with parental DNAand two types of unrelated DNA (32), and melt-ing profiles of the complexes formed are shown.With the lowest size shown, n = 14, and below,all the points superimpose on one line, indicatinga lack of specificity. At n = 17 and above, specificinteractions dominate the reaction, and thisspecificity increases with increasing chain length.These results show that, as would be expectedfrom elementary statistical considerations (32),oligonucleotides of chain length greater thanabout 14 to 15 show high specificity. Similarresults from studies of hybrids between oligonu-cleotides of the ribose series and DNA have beenreported (38a). This is clearly due to the failure tofind highly complementary base sequences inheterologous DNA molecules. Interactions do, ofcourse, occur with any DNA, but these do notoccupy all of the bases contained in the chain.It is clear from these considerations that, withhigh-molecular-weight DNA, at least this extentof base-pairing between the two strands mustoccur for the interaction to be a specific one.

Estimation of the Degree of Mismatching of BasesThe same experiment may be used to provide

information as to the number of base sequencesimilarities which occur between similar sites inthe genome or corresponding sites in differentDNA molecules. The principle is based on thefact that, in the size range, n = 3 to n = about50, the stability of the duplex depends upon itslength. Thus, the T,, increases with increasingchain length. This has been shown for manymodel systems of synthetic polynucleotides (23,38, 42), and for the interactions of oligonucleo-tides and DNA (32).

Figure 15 illustrates a preliminary experimentaimed at using this principle to determine theextent of mismatching in the duplexes formed by

mammalian DNA. The data are taken from adouble-label experiment, of the type shownearlier, in which a differentially labeled mixture ofoligonucleotides from the two species are incu-bated with both parental DNA preparations. Thesubtraction plot, in which the percentage reactionof heterologous oligonucleotides is subtractedfrom the homologous reaction, is a measure of thespecies-specific complexes. Even though the basecompositions are the same, B. subtilis complexesare considerably more stable, indicating a lack offidelity in matching for the mouse DNA. Thedisplacement to lower temperatures is a measureof the degree of mismatching. Current experi-ments have shown that B. subtilis fragments ofsize n = 20 give a curve close to that for mouseDNA at size 25. We may, therefore, draw thetentative conclusion that about five base pairs aremismatched. This effect must be due to the ar-rangement of base sequences in mouse DNAitself and would be expected from the data derivedfrom interactions of high-molecular-weight DNA.Because of the fact that there are many sitesin DNA of related, but not identical, base se-

01 nI,,I4

5-

b 60-St

K 50-CZq.)CZ40-

10-

n 22DNA

* T4a MouseA E.co/i

100 20 60 100Temperature IC

FIG. 14. Melting profiles of complexes formed byincubating 8 Ug of 32P-labeled T4 phage oligodeoxy-nucleotides of various chain lengths with 100 pug of T4phage, Escherichia coli, or mouse DNA. From Mc-Conaughy and McCarthy (32).

224 McCCARTHY


http://mm

br.asm.org/

Dow

nloaded from



I00~- -u-

80- \

n=15/ mouse DN B subt//is DNAmouse DNAo600 \

b 40k

20H

0 20 40 60 8CTemperature °C

FIG. 15. Relative stability of complexesformed by amixture of oligodeoxyribontucleotides originatinig frommouse or Bacillus subtilis DNA (with n = 15 or 25) andthe correspondinig single-stranzded DNA. Tlhe data rep-resent the stability of species-specific complexes, ob-taimied by subtracting the reaction2 of heterologous oligo-,iucleotides from that of homologous oligontucleotides.Aln equimolar mixture of 3H-labeled mouse and 32p_labeled B. subtilis oligoniucleotides (0.07 ,ug) was slowlycooled with 12 ,ug offilter-bound DNA.

quence, these somewhat mispaired structures willdominate the spectrum of reaction products.

It should be emphasized that this estimate forthe extent of mismatching is very crude andpreliminary. It assumes that mismatched baseshave no influence on the stability of the inter-action complex. This is almost certainly an over-simplification. Nevertheless, it is clear that thisapproach can be highly rewarding and can giveus quantitative information. Artificial in vitromodification of the bases to produce known de-grees of mismatching can also be used to calibratethe method, to permit quantitative estimates ofbase sequence divergence.The possibility of estimating base sequence

divergence may be illustrated by a considerationof interspecific reactions. Figure 16 shows theinteraction of a mixture of oligonucleotides, ofn = 25 or n = 40, derived from mouse and B.subtilis DNA with various rodent DNA prepara-tions. It is possible to determine whether eachDNA is able to recognize the mouse oligonucleo-tides as those of the same or a related organism.Differentially labeled oligonucleotides originatingfrom B. subtilis DNA are included as a controlfor nonspecific interactions. Thus, when thecurves superimpose, no discrimination is made.Obviously, specific reactions are given with bothsizes by parental DNA. On the other hand, basesequence complementarity offered by fragmentsof mouse DNA of n = 25 is not sufficient to givespecific interactions with rat or hamster DNA.

When n is increased to 40, the information contentis sufficiently high to allow specific reactions withrat DNA. Even at this size, however, sufficientsimilarity in sequence does not exist for specificinteractions to occur with hamster DNA. ThisDNA is related in base sequence, as demonstratedby reactions between high-molecular-weightmouse and hamster DNA (20) and by taxonomicconsiderations. Clearly, we can make estimatesof the degree of pairing and therefore the extentof base sequence similarity between mouse andrat DNA by subtraction plots as shown earlier.These experiments could be carried out amongany group of related DNA types to provide in-formation as to extents of sequence divergenceamong species or within the DNA of one or-ganism.

EVOLUTION AND RECOMBINATION OF DNAMOLECULES

During the evolution of life on earth, twokinds of changes in the DNA have accompaniedthe development of more complex forms. Oneinvolves changes in base sequence and the othera progressive increase in the DNA content of agenome. The former changes are developed bymutation leading to base substitution, or insertionand deletion of bases leading to a more radicalchange in base sequence. This diversification ofbase sequence is most easily seen in bacteria inwhich the overall base composition varies be-tween 25 and 75'(, GC. This variation in base

40 Mouse DNA0.

30

20\

25 50 75

n=40

Rat DNA - Hamster DNA

25 50 75 25 50 75

Temperature °C

FIG. 16. Tlhermal dissociation profiles of complexesformed by slow coolintg a mixtutre of mouse anid Bacillussubtilis oligodeoxyribonucleotides with mouse, rat, orhamster DNA bounld to filters; 0.07 Mg of anl equimolarmixture of oligoniucleotides (mt = 25 or 40) ofmouse alidB. subtilis was slowly cooled with 12 ,ug oflfilter-boumudDNA. (0) 3H-labeled mouse oligommucleotides; (A)32P-labeled B. subtilis DNA.

225VOL. 31, 1967


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

composition within a group alsoexists inprotozoa,the more primitive metazoans, and higher plants,although the overall base composition of theDNA of vertebrates is quite similar. In all thesecases, however, a considerable variation in basecomposition of segments within a genome isapparent.As a general rule, the DNA content per haploid

cell has increased progressively through theevolution of more complex forms. Viruses, withthe smallest number of gene functions, containthe least DNA, and this quantity increases pro-gressively through bacteria, protozoa, and meta-zoans to the vertebrates (Table 2). There are,however, some exceptions to this generalization,the meaning of which remains obscure. The mostwell-known example is that of the Anuranamphibians, whose DNA complement is manytimes that of other amphibia (29, 34). Other well-known examples have been documented invarious families of higher plants (47). Whateverthe meaning of these anomalies, it is reasonableto conclude that the predominant pattern is oneof a steady increase in the DNA content of agenome during evolution, closely correlated withthe structural and the behavorial complexity ofthe organism.There are several possibilities that might be

suggested for the mechanism by which moreDNA is created to specify newly evolving func-tions. Such new base sequences may possibly begenerated de novo by end addition to existingDNA molecules or by reiterative synthesis (22).Such base sequences would presumably be ran-dom, although useful sequences might con-ceivably evolve during an extended period.Alternatively, an increase in the total DNAmight be accomplished through the acquisition ofready-made sequences from within the genomeor other external sources. In bacteria, transduc-tion and sexduction lead to an increase in theamount of DNA. The latter process, in particular,is known to operate across species boundariesand would therefore allow the acquisition ofgenetic material from different species. After sex-duction or transduction, the integrated episomalgenes may become a stable part of the bacterialchromosome. It is quite possible that similarmechanisms may operate in higher organisms,since clear evidence exists that the genome of tumorviruses may enter into a stable association withthe host cell nucleus (3).

Notwithstanding these other possibilities, itseems most likely that gene duplication followedby translocation is a major contributor to theincreasing DNA content in the genome. Muchevidence for this process derives from determina-tions of amino acid sequence in similar proteins.

A particularly persuasive example is provided bythe various hemoglobin chains (58). It is evidentfrom the studies reported here and those of otherworkers (6) that this mechanism has been animportant one during the evolutionary history ofmammalian DNA.

In a very simple DNA, as represented by T4phage DNA, there is no evidence for the existenceof duplications. Thus, it may be that this processdoes not occur or is rapidly obscured by diver-gence, or that the generation of such duplicatedbase sequences is for some reason disadvantage-ous. In this regard, the arguments set forth byThomas (51) would tend to rationalize thisfinding. If, as seems most likely, a basic stage inthe recombination of DNA molecules involvedthe association of complementary base sequences(33) between the two DNA molecules, the fidelityof recombination would be disturbed by theexistence of very similar base sequences aroundthe chromosome. Since the unequal crossing-overgenerated by this effect would, in most cases, belethal (51), the persistence of gene duplicationswould be unlikely unless the duplicated genes canrapidly be made dissimilar by mutation andselection.

Similar conclusions may be drawn from theexperiments with B. subtilis DNA. No closelysimilar base sequences exist, although there isevidence for very distant intragenome homologies.The rate of the duplex-forming reaction underthese conditions suggests that base sequencesimilarities at this level are shared by many re-gions of the DNA. This may well be the result ofhistorically remote gene duplications. Freese andYoshida (14) presented evidence for homologyamong three B. sibtilis dehydrogenases, suggest-ing that these have evolved from a commonDNA region during bacterial evolution. Otherexplanations may be offered; for example, it maybe that certain sequences of codons occur quitefrequently as a result of the existence of similaractive sites in many enzymes. In either case,apart from the rRNA cistrons, it seems clearthat base sequence relationships among differentcistrons above a certain level do not exist. Again,it is possible that they are, in fact, precluded bythe requirements of recombination events.

In mouse DNA, the situation is quite different,so that a very complex set of related base se-quences exist. If the existence of similar basesequences along the chromosomes does indeedpresent a problem for recognition of events pre-ceding recombination, it seems most likely thatthis is surmounted by the compartmentalizationof DNA among chromosomes. The synapsis ofchromosomes would serve to restrict the spatialdistribution of potentially recombining DNA

226 MCCARTHY


http://mm

br.asm.org/

Dow

nloaded from



base sequences so that strand recombinationcould take place only between allelic segments.At the moment, the relationship between the

chromosome complement and the distribution ofbase sequences in the DNA is far from clear,since few DNA preparations have been studied indetail. The question as to whether related basesequences are randomly distributed amongchromosomes cannot be answered until individualchromosomes may be prepared. Nor is it knownwhether there is any correlation between eucaryo-tic chromosome structure and the presence ofrelated base sequences in the DNA. Whether thediscontinuity between procaryotic and eucaryoticcells is also represented as a boundary betweensimple and complex partially redundant DNAmolecules remains to be seen.

ACKNOWLEDGMENTSThis research was supported by grants from the

National Science Foundation (GB 6099) and from theNational Institutes of Health, Public Health Service(GM 12449). This research would not have beenpossible without the eager cooperation of my studentsand colleagues, R. B. Church, C. D. Laird, B. L.McConaughy, and R. L. Moore.

LITERATURE CITED1. BAUTZ, E. K. F., AND E. REILLY. 1966. Gene-

specific messenger RNA: isolation by the dele-tion method. Science 151:328-330.

2. BELLETT, A. J. D. 1967. Numerical classificationof some viruses, bacteria and animals accordingto nearest-neighbor base sequence frequency. J.Mol. Biol. 27:107-112.

3. BENJAMIN, T. L. 1966. Virus specific RNA in cellsproductively infected or transformed by poly-oma virus. J. Mol. Biol. 16:359-373.

4. BERG, P., H. FANCHER, AND M. CHAMBERLIN. 1963.The synthesis of mixed polynucleotides con-taining ribo- and deoxyribonucleotides bypurified preparations of DNA polymerase fromE. coli, p. 467-483. Int V. Bryson and H. J.Vogel [ed.], Informational macromolecules.Rutgers Symposium. Academic Press, Inc.,New York.

5. BERNS, K. I., AND C. A. THOMAS, JR. 1965. Isola-tion of high molecular weight DNA fromHemophilus influenzae. J. Mol. Biol. 11:476-490.

6. BRITTEN, R. J., AND D. E. KOHNE. 1966. Nucleo-tide sequence repetition in DNA. Ann. Rept.Carnegie Inst. Wash. 65:78-106.

7. CAIRNS, J. 1963. The chromosome of E. coli. ColdSpring Harbor Symp. Quant. Biol. 28:43-45.

8. DEN1S, H. 1966. Gene expression in amphibiandevelopment. I. Validity of the method used:interspecific and intraspecific hybridization be-tween nucleic acids. Properties of mRNA syn-thesized by developing embryos. J. Mol. Biol.22:269-283.

9. DENNIS, E. S., AND R. G. WAKE. 1966. Auto-radiography of the Bacillus subtilis chromosomeJ. Mol. Biol. 15:435-439.

10. DENHARDT, D. T. 1966. A membrane-filter tech-nique for the detection ofcomplementary DNA.Biochem. Biophys. Res. Commun. 23:641-646.

11. Doi, R. H., AND R. T. IGARASHI. 1965. Conserva-tion of ribosomal and messenger ribonucleicacid cistrons in Bacilluis species. J. Bacteriol.90:384-390.

12. DOTY, P., J. MARMUR, J. EIGNER, AND C. SCHILD-KRAUT. 1960. Strand separation and specificrecombination in DNA's. II. Physical chemicalstudies. Proc. Natl. Acad. Sci. U.S. 46:461-476.

13. DUBNAU, D., I. SMITH, P. MORELL, AND J.MARMUR. 1965. Gene conservation in Bacillusspecies. I. Conserved genetic and nucleic acidbase sequence homologies. Proc. Natl. Acad.Sci. U.S. 54:491-498.

14. FREESE, E., AND A. YOSHIDA. 1965. The role ofmutations in evolution, p. 341-355. In V. Bry-son and H. J. Vogel [ed.], Evolving genes andproteins. Rutgers Symposium. Academic Press,Inc., New York.

15. GILLESPIE, D., AND S. SPIEGELMAN. 1965. A quan-titative assay for DNA/RNA hybrids withDNA immobilized on a membrane. J. Mol.Biol. 12:829-842.

16. GOODMAN, H. M., AND A. RICH. 1962. Formationof a DNA soluble RNA hybrid and its relationto the origin, evolution and degeneracy ofsoluble RNA. Proc. Natl. Acad. Sci. U.S. 48:2101-2109.

17. HERSHEY, A. D., E. BURGI, AND L. INGRAHAM.1962. Sedimentation coefficient and fragilityunder hydrodynamic shear as measures ofmolecular weight of the DNA of phage T5.Biophys. J. 2:423-431.

18. HOLLEY, R. W., J. APGAR, G. A. EVERETT, J. T.MADISON, M. MARQUISEE, S. H. MERRILL, J. R.PERSWICK, AND A. ZAMIR. 1965. Structure of aribonucleic acid. Science 147:1462-1465.

19. HOYER, B. H., B. J. MCCARTHY, AND E. T. BOLTON.1964. A molecular approach in the systematicsof higher organisms. Science 144:959-967.

20. HOYER, B. H., E. T. BOLTON, B. J. MCCARTHY,AND R. B. ROBERTS. 1965. The evolution ofpolynucleotides, p. 581-590. In V. Bryson andH. J. Vogel [ed.], Evolving genes and proteinsRutgers Symposium. Academic Press, Inc.,New York.

21. JOSSE, J., A. D. KAISER, AND A. KORNBERG. 1961.Enzymatic and synthesis of DNA. VIII. Fre-quencies of nearest-neighbor base sequences inDNA. J. Biol. Chem. 236:864-871.

22. KORNBERG, A. 1965. Synthesis of DNA-likepolymers de nIovo or by reiterative replication,p. 403-417. In V. Bryson and H. J. Vogel [ed.],Evolving genes and proteins. Rutgers Sympo-sium. Academic Press Inc., New York.

23. LiPSETT, M. N. 1964. Complex formation betweenpolycytidylic acid and guanine oligonucleotides.J. Biol. Chem. 239:1256-1260.

24. MARMUR, J., AND D. LANE. 1960. Strands separa-tion and specific recombination in DNA's. I.Biological studies. Proc. Natl. Acad. Sci. U.S.46:453-461.

227VOL. 31, 1967


http://mm

br.asm.org/

Dow

nloaded from


BACTERIOL. REV.

25. MARMUR, J., AND P. DOTY. 1961. Thermal re-naturation of DNA. J. Mol. Biol. 3:585-594.

26. MARMUR, J., R. ROWND, AND C. L. SCHILDKRAUT.1963. Denaturation and renaturation of DNA.Progr. Nucleic Acid Res. 1:231 300.

27. MARTIN, M. A., AND B. H. HOYER. 1966. Thermalstabilities and species specificities of reannealedanimal DNA's. Biochemistry 5:2706-2713.

28. MASSIE, H. R., AND B. H. ZIMM. 1965. Molecularweight of the DNA in the chromosomes of E.coli and B. subtilis. Proc. Natl. Acad. Sci. U.S.54:1636-1641.

29. MCCARTHY, B. J. 1965. The evolutionl of poly-nucleotides. Progr. Nucleic Acid Res. 4:129160.

30. MCCARTHY, B. J. 1967. General aspects of homol-ogy studies with nucleic acids. Intern. Symp.Tumor Viruses, 1st, Subviral Carcinogenesis, p.43 -61.

31. MCCARTHY, B. J., AND E. T. BoLTON. 1963. Anapproach to the measurement of genetic related-ness among organisms. Proc. Natl. Acad. Sci.U.S. 50:156 164.

32. MCCONAUGHY, B. L., AND B. J. MCCARTHY. 1967.The interaction of oligodeoxynucleotides withdenatured DNA. Biochim. Biophys. Acta 149:180-189.

33. MESELSON, M. 1964. On the mechanism of geneticrecombination between DNA molecLiles. J.Mol. Biol. 9:734 745.

34. MIRSKY, A. E., AND H. Ris. 1951. The DNA con-tent ol animal cells and its evolutionary signifi-cance. J. Gen. Physiol. 34:451 462.

35. MOORE. R. L., AND B. J. MCCARTHY. 1967. Com-parative study of ribosomal cistrons in entero-bacteria and myxobiacteria. J. Bacteriol. 94:1066- 1074.

36. MOROWITZ, H. J., M. E. ToURTELIOTTE, W. R.GUILD, E. CASTRO, C. WOESE, AND R. C.CLEVERDON. 1962. The clhemical compositionand submicroscopic morphology of Ml (co-plCasin galliseptiien. J. Mol. Biol. 4:93 -103.

37. MORRISON, J. M., H. M. KEIR, H. SUBAK SHARPE,AND L. V. CRAWFORD. 1967. Nearest-neighborbase sequence analysis of the DNA of a furtherthree mammaliain viruses. Simian virus 40,human papilloma and adeniovirus type 2. J.Gen. Virol. 1:101-108.

38. NAYLOR, R., AND P. T. GILHANI. 1966. Studies onsome interactions and reactions of oligonucleo-tides in aqueous solution. Biochemistry 5:2722-2728.

38a. NIYoGI, S. K., AND C. A. THOMAS, JR. 1967. Thespecific association of ribo-oligonucleotides ofknown chain length with denatured DNA.Biochem. Biophys. Res. Commun. 26:51 57.

39. OGUR, M., S. MINCKLER, G. LINDERGREN, ANDC. C. LINDERGREN. 1952. The nucleic acids in apolyploid series of Sacchlaroin -cce.%. Arch. Bio-chem. Biophys. 40:175 184.

40. PARK, I. W., AND J. DELEX. 1967. Anicestral rem-nants in DNA from Pseuionomonas and Xoa-thIo,nonas. Antonie vani L nseUllweloek J Micro-biol. Serol. 33:1 16.

41. RAJBHANDARY, U. L., A. STUART, R. D. FAULK-NER, S. H. CHANG, AND H. G. KHORANA. 1966.Nucleotide sequence studies on yeast phenylala-nine sRNA. Cold Spring Harbor Sym. Quant.Biol. 31:425-434.

42. RICH, A. 1960. A hybrid helix containing bothdeoxyribose and ribose polynucleotides and itsrelation to their transfer of information betweenthe nucleic acids. Proc. Natl. Acad. Sci. U.S.46:1044 -1053.

43. RITOSSA, F. M., AND S. SPIEGELMAN. 1965. Loca-lization of DNA complementary to ribosomalRNA in the nucleolus organizer region ofDrosophilc me/laniogaer.. Proc. Natl. Acad. Sci.U.S. 53:737 745.

44. RUBENSTEIN, I., C. A. THOMAS, JR., AND A. D.HERSHEY. 1961. The molecular weights of T2bacteriophage DNA and its first and secondbreakage products. Proc. Natl. Acad. Sci. U.S.47:1113 1122.

45. SIBLEY, C. G. 1965. The characteristics of specificpeptides from single proteins as data for classifi-cation. p. 435-450. 1I V. Bryson and H. J.Vogel led. ], Evolving genes and proteins.Rutgers Symposium. Academic P'ress Inc..New York.

46. SINSHEIMEIZ, R. L. 1959. A siigle-stranided DNAfrom bacteriophage 0 X174. 1. Mol. Biol. 1:43 -53.

47. STEBBINS, G. L. 1966. Clhromosomal variation andevolution. Science 152:1463- 1469.

48. SUBIRANA, J. A. 1966. Kinetics of renaturation ofdenatured DNA. 11. Products of the reaction.Biopolymers 4:189-200.

49. SUBIRANA, J. A., AND P. DOTY. 1966. Kinetics ofrenaturationi of denatured DNA. 1. Spectro-plhotometric results. Biopolymers 4:171 187.

50. TAKAHASHI, M., M. SAITO, AND Y. IKEDA. 1967Species specificity of the ribosomal RNA cis__trolls in bacteria. Biochim. Biophys. Acta 134:124 -133.

51. THOMAS, C. A. JR., 1966. Recombin1ation of DNAmolecules. I'rogr. Nucleic Acids Res. Mol. Biol.5:315-348.

52. VENDREL\, R., AND C. VENDRELY. 1949. La teneurdu noyau cellulaire en acide desoxyribonuclei-que ai travers les organes, les individus et lesespeces animales. Experientia 5:327-329.

53. WALKER, P. M. B. AND A. MCCLAREN. 1965.Specific duplex formation in vitro of mammalianDNA. J. Mol. Biol. 12:394 409.

54. WILSON, A. C., AND N. 0. KAPLAN. 1965. Enzymestructure and its relation to taxonomy, p. 321 -

346. In V. Brysoin and H. J. Vogel led.], Evolv-ing genes and proteins. Rutgers Symposium.Academic Press Inc.. New York.

55. YANKOFSKY, S., AND S. SPIEGELMAN. 1962. Theidentification of the ribosomal RNA cistron b\sequence complementarity. 1I. Saturation ofand competitive interaction at the rRNA cis-tron Proc. Natl. Acad. Sci. U.S. 48:1466-1472.

56. YANKOF.SK\, S. AND S. SPIEGFL\IAN. 1963. Dis-

228 MCCARTHY


http://mm

br.asm.org/

Dow

nloaded from



tinct cistrons for the two ribosomal RNAcomponents. Proc. Natl. Acad. Sci. U.S. 49:538-544.

57. ZACHAU, H. G. D. DUTrING, H. FELDMANN,F. MELCHERS, AND W. KARAU. 1966. Serinespecific transfer RNA's. XIV. Comparison ofnucleotide sequences and secondary structure

models. Cold Spring Harbor Symp. Quant.Biol. 31:417-424.

58. ZUCKERKANDL, E., AND L. PAULING. 1965. Evolu-tionary divergence and convergence in proteins,p. 97-166. In V. Bryson and H. J. Vogel [ed.],Evolving genes and proteins. Rutgers Sympo-sium. Academic Press Inc., New York.

VOL. 31, 1967 229


http://mm

br.asm.org/

Dow

nloaded from


Documents

Arrangement of Base Sequences in Deoxyribonucleic Acid' · ment of nucleotides in deoxyribonucleic acid (DNA) may be discussed. At the gross level, a considerable body of information