Hemoglobin Switching in Sheep and Goats

Document Sample
Hemoglobin Switching in Sheep and Goats Powered By Docstoc
Vol. 252, No. 6, Isue of March 25, pp. 1906-1916,   1977
                   Printed in U.S.A.

Hemoglobin                             Switching                   in Sheep and Goats
PREPARATION                 AND CHARACTERIZATION     OF COMPLEMENTARY                                                 DNAs         SPECIFIC           FOR THE              a-, /3-,

                                                                                                                            (Received for publication,           October 26, 1976)

                 EDWARD J. BENZ, JR.,*                      CRAIG E. GEIST,    ALAN          W. STEGGLES,$               JANE       E. BARKER,             AND
                 ARTHUR W. NIENHUIS
                 From the Section on Clinical     Hematology,     Molecular  Hematology     Branch, National                                           Heart,      Lung,       and
                 Blood Institute, National    Institutes   of Health, Bethesda,    Maryland     20014

    Specific complementary        DNAs (cDNAs) for the messenger                          zygous or heterozygous for Hb A are subjected to acute ane-
RNAs coding for sheep LY-, p”-, /3”-, $‘-, and y-globins were                             mia, hypoxia, or treatment with erythropoietin,          production of
prepared by thermal denaturation              of heterologous      hybrids                Hb A is replaced by the synthesis of Hb C (cy2p2(‘) (3-5).

                                                                                                                                                                                            Downloaded from by guest, on July 26, 2011
(e.g. cup’-cDNA.ay-mRNA)              followed      by hydroxylapatite                    Removal of the erythropoietic         stress results in a reciprocal
chromatography.          Each cDNA represented             a nearly full-                 change from Hb C to Hb A synthesis. Sheep homozygous for
length copy of its globin mRNA complement as determined                                   Hb B do not exhibit the property of “A-C” switching (3-5).
by electrophoretic        analysis in polyacrylamide         gels contain-                    The molecular basis of hemoglobin switching in sheep has
 ing 98% formamide.         The purity of each cDNA fraction was                          not been fully defined. Previous studies of messenger RNA
 estimated by hybridization        analysis and thermal denatura-                         (mRNA) translation in cell-free extracts have established that
tion. The j3- and y-cDNAs contained 5 to 20% contaminating                                both the F-+A (6) and the A-& (7, 8) switches are associated
cu-cDNA while the a-cDNA was 25 to 30% contaminated                    with               with concomitant changes in the content of functional (trans-
non-cy-cDNA. The melting temperatures               (T,,,) of homologous                  latable) y-, p”-, and @‘-globin mRNAs. A prerequisite                 to
duplexes between each non-a chain cDNA and its mRNA                                       further analysis of the molecular            events regulating     such
 complement ranged from 69.5-71.5” in 50% formamide while                                 changes in functional mRNA content is the availability                of
 (Y-(Yduplexes melted with a T,,, of 75-76”. The T,,, values of                           specific molecular probes for use in DNA-DNA and DNA-RNA
heterologous       duplexes formed between each non-cu-cDNA                               hybridization    assays. In this report we describe the purifica-
 and the various globin mRNAs (e.g. P’S-cDNA. Hb C mRNA)                                  tion and characterization      of radioactive full-length DNA cop-
ranged between 64.5” and 68” and thus were only 1.5-5.0                                   ies (cDNAs) complementary         to each of the mRNAs coding for
 below that of homologous          duplexes. These results suggest                         a-, p”-, PI’-, @‘-, and y-globin and present evidence indicating
that the nucleotide        sequence divergence among the various                          these can be used as unique probes for the analyses of DNA or
 non-cu-mRNAs (or cDNAs) is not greatly different from the                                RNA sequences specific for the individual globins.
 minimum       predicted from the amino acid sequence differ-
                                                                                                                      MATERIALS         AND     METHODS
 ences of the corresponding         globins. When annealing           reac-
tions were performed above the T,,, of the heterologous                 hy-                   Reagents -RNA-dependent                DNA       polymerase          isolated     by the
 brids (68”), each non-a-cDNA            hybridized      only to its own                  method of Kacian et al. (9) was obtained                  from Dr. J. W. Beard (Life
                                                                                          Sciences,     Inc., Gulfport,      Fla.) through           the office of Program            Re-
 complementary         mRNA. Thus the purified cDNAs provide                              sources and Logistics,        Viral Oncology           Branch,      National      Cancer In-
 molecular      probes for the quantitation           of (Y-, j?-, and y-                 stitute.   [“HldCTP      (specific    activity      17.3 Ci/mmol)          and [:‘Hlleucine
 globin-specific     nucleotide sequences.                                                (specific activity   400 mCi/mmol)            were products         of Amersham/Searle
                                                                                          and New England         Nuclear,      respectively.         Radioactive      and nonradio-
                                                                                          active DNA markers, prepared by restriction endonuclease digestion
                                                                                          of SV40 DNA, were purchased           from Bethesda Research Laboratories
   The control of hemoglobin synthesis in sheep is character-                             (Bethesda,    Md.). Sephadex      G-25 (coarse) and Sephadex        G-50 (coarse)
                                                                                          were products      of Pharmacia,        Ltd. Actinomycin      D, dithiothreitol,
ized by the existence of at least two major hemoglobin switch-                            dATP, dGTP, dTTP, Hepes,’ Escherichia              coli DNA, E. coli RNA, and
ing phenomena. The fetal + adult (“F-A”)      switch (l-3) occur-                         yeast RNA were obtained         from Sigma or Calbiochem.          S, nuclease of
ring just prior to birth is marked by an abrupt decline in the                            Aspergillus   oryzae and oligo(dT),P-,8       were purchased    from Miles Lab-
synthesis of Hb F (o(iyJ coupled with an increase in the                                  oratories,  Inc.; acrylamide,     bisacrylamide,       Temed, and hydroxylapa-
synthesis of adult hemoglobin (Hb A: a2/3/‘, or Hb B: LY&“, or
both). The spontaneous       and subsequently   reversed “AC”                                  ’ The abbreviations         used are: Hepes, 4-(2-hydroxyethyl-l-pipera-
                                                                                          zineethanesulfonic         acid; oligo(dT),,_,,,     deoxythymidylic      acid 12 to 18
switch (1) is observed during postnatal life in animals inherit-
                                                                                          residues long; cDNA, DNA complementary                       to messenger     RNA syn-
ing at least one gene for Hb A. Also, when adult sheep homo-                              thesized     by RNA-dependent            DNA polymerase;           HAP, hydroxylapa-
                                                                                          tite; Temed, N,N,N’,N’-tetramethylmethanediamine;                         T,,, melting
   * To whom reprint requests should be addressed.                                         temperature       (temperature        at which 50% of preformed            DNA-RNA
   $ Present address, Department of Molecular  Pathology               and Biol-          hybrids     are thermally       denatured);      EDTA, ethylenediaminetetraace-
ogy, Northeast       Medical     School,    Kent,   Ohio.                                 tic acid, disodium        salt; 01, PA, PR, p”, y, sheep globin chains.

                                                                       Complementary                       DNAs              for Sheep          Globin          mRNAs                                                                       1909
tite (“Bio-Gel            HTP” - DNA grade) from Bio-Rad,                                     formamide            from         Spectrometer.             The per cent hybridization                       was defined as the ratio of
Eastman           Kodak; and Rexyn I-300, bromphenol                                   blue, and methylene                      S, nuclease-resistant                trichloroacetic            acid-precipitable             counts per min
blue from Fisher Scientific                        Products,          Inc.                                                      to the precipitable             counts per min present in the undigested                                  parallel
     Preparation          of Sheep Reticulocytes                    -Blood        specimens           used for the              aliquot of each hybridization                       mixture.         Corrections         for self-annealing
preparation            of Hb F mRNA and Hb B mRNA were obtained,                                                respec-         of cDNA were made by subtracting                                      the percentage             of the input
tively,      from the umbilical                   vein of a normal                 loo-day-old           fetal lamb             radioactivity           which was resistant                    to S, nuclease after incubation
and from an anemic adult sheep homozygous                                           for Hb B. Anemia                 was        without       globin mRNA                (0.5 to 3%) as indicated                 in the figure legends.
induced in the adult sheep by daily treatment                                        with both phlebotomy                           Thermal          Denaturation             Analysis        of cDNA-mRNA                  Hybrids        - Com-
and phenylhydrazine                    using protocols described previously                                (7, 10). In          plementary           DNA-mRNA                duplexes were formed in 50% formamide                                 as
order to obtain Hb A mRNA,                               anemia          was induced in a 60-day-old                            described         above. Sufficient                mRNA           was added to ensure complete
 lamb homozygous                 for Hb A. On Days 3 through                           5 of treatment,            when          hybridization            of the probe (2- to lo-fold excess mRNA).                               The reaction
the first appearance                 of circulating             reticulocytes           was noted, 300 ml of                    mixtures         were incubated              at 50” for 12 to 16 h. Individual                   reactions        (10
peripheral           blood were obtained                    for RNA extraction.                    At this time,                ~11 were then incubated                      for 5 min at the desired temperature                               in a
only Hb A could be detected by cellulose acetate electrophoresis                                                      (7-       recirculating            water bath, quick-frozen                       in acetone/dry            ice, and ex-
 10); the appearance                of reticulocytes             preceded the appearance                     of circu-          pelled on thawing                 into 1.1 ml of digestion                     buffer which had been
 lating      Hb C by 2 days. Hb C mRNA                                       was extracted             from blood               maintained            at O-4” in an ice bath. The per cent of the initial cDNA
 obtained on Day 14 of treatment,                              after complete conversion                      to Hb C           remaining            double-stranded               was then determined                       by S, nuclease
 had occurred.                                                                                                                  digestion        as described           above.
     Isolation        of Reticulocvte           RNA and Globin Messenger                             RNA - Preu-                    Preparation           of Specific Globin cDNAs by Hydroxylapatite                                  Chroma-
 aration of hemolysate,                  isolation          of total reticulocyteRNA                      by phenol             tography       -Individual            reactions         (1 to 3 ml) containing                 approximately
 extraction,          and fractionation                 of RNA by sucrose gradient                             density          350 rig/ml of cDNA and 8 pg/ml of a heterologous                                      globin mRNA (e.g.
 centrifugation             were performed               by methods described in earlier com-                                   ol/3”-cDNA.ay-mRNA)                       were incubated                in aliquots        of 100 ~1 for 90
 munications            (7, 11, 12). The 6 to 15 S RNA fraction                                  which contains                 min at 50”. Each was heated to 68” for 5 min, quick-frozen                                          in acetone/
 globin mRNA                 was further              purified         by a second sucrose density                              dry ice, thawed, and rapidly                       expelled into 10 volumes of ice-cold 0.05
 gradient        centrifugation.             The final RNA preparations                           used for cDNA                 M sodium phosphate                     buffer,      pH 6.8. This was applied to a
 synthesis were 20 to 60% 10 S globin mRNA as estimated                                                  by electro-            diameter         column of hydroxylapatite                         (HAP) which had been equili-

                                                                                                                                                                                                                                                        Downloaded from by guest, on July 26, 2011
 phoresis in 6% polyacrylamide                            geis containing              98% formamide.                           brated at 60” with 0.05 M sodium phosphate,                                      pH 6.8, in a jacketed
      Translation           of Messenger               RNA -Each                globin messenger                   RNA          column at a flow rate of 1 mlimin.                          Each column contained                    1 to 1.5 mg
 fraction was incubated                   with L:‘Hlleucine               in Xenopus laevis oocytes by                          of HAPing           of cDNA. The column was washed with 20 to 50 ml of 0.05
 the method of Gurdonetal.                         (13, 14). Each oocyte was injected with 35                                   M sodium phosphate,                  pH 6.8, and then successively                        with 20 ml of 0.16
 to 70 ng of mRNA;                   10 to 20 oocytes were injected for each analysis.                                          M sodium          phosphate,          pH 6.8, and 20 ml of 0.5 M sodium phosphate,
 Globin-enriched              protein was isolated by successive chromatography                                                 pH 6.8. Three-milliliter                   fractions       were collected.            Five-microliter            ali-
 on Sephadex            G-100 and carboxymethylcellulose                               columns as described                     quots of each fraction                 were mixed with 0.5 ml of water and 10 ml of
 elsewhere           (10, 13-151.” The globin chains were then separated                                               by       Riafluour         (New England                Nuclear)        and counted (efficiency                   for :‘H =
 carboxymethylcellulose                    chromatography                   in 8 M urea, using methods                          28%). The radioactive                   fractions        eluting upon application                    of the 0.16
 described in earlier communications                                 (7, 10).                                                   and 0.5 M buffers were pooled separately                                 and mixed with 200 pg of E.
      Synthesis        of cDNA - Globin mRNA was incubated                                         at a final con-              coli RNA. The “single-stranded”                           (0.16 M) and “double-stranded’                         (0.5
 centration         of 1.0 A,,;,, unit/ml             with RNA-dependent                     DNA polymerase                     M) cDNA           fractions        were then applied separately                         to Sephadex            G-50
  (250 units/ml?           in reaction          mixtures          containing           Tris. HCl, pH 7.9 (50                    columns        (1.5 x 60 cm) previously                      equilibrated          with Buffer B (0.1 M
 mM), KC1 (50 mM), M&l,                              (10 mM), dATP, dGTP, dTTP (100 /*M                                         sodium chloride,               1 rnM EDTA, pH 7, 0.01 M TrisiHCl,                                pH 7.6, 0.5%
 each),        [“H]dCTP            (specific        activity         17.3 Ci/mmol,                50 FM),         oligo-        sodium dodecyl sulfate).                      Radioactivity             eluting     in the void volume
  (dT),r-18 (0.3 A,,,, unit/ml),                    and actinomycin                   D (0.1 mg/ml).              After         was collected             and precipitated               with ethanol            at -20”. The cDNAs
 incubation          at 37” for 30 min the reaction mixture                               was chilled, passed                   were recovered              by centrifugation                at 13,000 x g for 1 h at -20” and
 through         a dry pad of Sephadex G-25 by centrifugation,                                     and recovered                fractionated          on 5 to 20% alkaline                 sucrose gradients.              The high molec-
 from an alkaline                sucrose gradient                 as previously             described          (16-191.         ular weight fraction                 of each cDNA was recovered,                          neutralized,          pre-
 The cDNA in fractions                       sedimenting              more rapidly              than an E. coli                 cipitated       in alcohol, resuspended                    in water, desalted with dry Sepha-
 DNA marker              peak (400 to 500 bases) was recovered                                 after neutraliza-                dex G-25, and stored in liquid nitrogen                                 as described          above.
 tion by ethanol              precipitation             as previously            described           (16-19). This                  Polyacrylamide              Gel-98%          Formamide              Electrophoresis          -Analytical
 “high       molecular          weight”          cDNA fraction                 was then centrifuged                     at      polyacrylamide              gels (5 x 70 mm) were cast in 98% formamide                                         con-
  13,000 x g for 1 h at -2O”, resuspended                                 in deionized           water, desalted                taining 20 rnM barbital                  and 20 rnM NaCl as described by Forget et al.
 by centrifugation                through         dry Sephadex                G-25, and stored in liquid                         (23) and soaked overnight                     in buffered           formamide.          Pre-electrophore-
 nitrogen.                                                                                                                      sis was then performed                    according         to the method of Forget et al. (231,
      cDNA-RNA              Hybridization              Assay-Unless                 otherwise         indicated         in      except that the running                       buffer in the reservoir                  was 20 mM NaCl,
 the figure legends, each lo-p1 hybridization                                    mixture          contained:         50%        without       formamide.            In order to remove salt, all samples for electro-
 formamide,            0.5 M sodium chloride,                       0.05 M Hepes, pH 7.0, 0.5 mM                                phoresis were centrifuged                      through        dry Sephadex             G-25 and ethanol-
 EDTA, pH 7.0, 0.7 pg/ml ofE. coli DNA, 50 wg/ml ofE. coli RNA,                                                                 precipitated;          the pellet was then rinsed twice with ethanol and dried
 and the desired amounts                        of globin mRNA and cDNA. The reaction                                           in a stream of nitrogen.                   The samples were resuspended                           in deionized
 mixtures          were sealed in capillary                      tubes which had been pretreated                                water, lyophilized,              and resuspended                in 98% formamide                containing         20
 with diethylpyrocarbonate                         (Baycovinl           and coated with E. coli DNA                             mM barbital,            20 mM NaCl, 7.5% sucrose, 0.025% xylene cyanolol FF,
 as described previously                   (20). The mixtures                   were heated to 99” for 10                       and 0.025% bromphenol                     blue at a final concentration                    at 200 to 400 pg/
 min and incubated                    under conditions                  of time and temperature                       de-       ml. Twenty-five              to fifty microliters             were applied to each gel. Electro-
  scribed in the figure legends.                                                                                                phoresis was for 8 h at 100 V (10 V/cm, 1 to 1.5 mA/gel).                                             Following
      Formation          of double-stranded                  cDNA-RNA               hybrids        was measured                 electrophoresis,             the gels were stained with methylene                                   blue as de-
 by digestion           with single-stranded                    specific S, nuclease (21, 22). Each                             scribed by Forget et al. (231, destained                             in water for approximately                    12
  reaction was quick frozen in acetone/dry                                   ice and rapidly            expelled on             h, frozen in dry ice, and sliced. Each slice was placed in a scintilla-
 thawing         into 1.1 ml of a solution                    containing          Buffer A (0.1 M sodium                        tion vial, covered with 2.0 ml of NCS in water (9:1), incubated                                              at 50”
  acetate, pH 4.5, 1 rnM zinc acetate, and 10 pg/ml of denatured                                                     calf       overnight,          cooled, and mixed with 9 ml of Hydromix.                                     The samples
  thymus DNA). Two aliquots                           (0.5 ml each) were removed.                       To one was              were then cooled to 0” and counted in a Packard                                     3375 liquid scintilla-
  added 1.5 ml of Buffer A. To the other was added 1.5 ml of Buffer A                                                           tion spectrometer              with an efficiency                for :iH of approximately                 20%.
  containing          300 units of S, nuclease. Each was incubated                                     at 45” for 30
  min and chilled quickly                     in an ice water bath. Yeast RNA (400 pg)                                                                                             RESULTS
  and 60% cold trichloroacetic                        acid (1.0 ml) were then added sequen-
  tially. After 20 min at 4”, the precipitates                                were collected on nitrocel-                          Translation         of Sheep Globin      mRNA     -The   translational     ca-
  lulose filters, dried, placed in 10 ml of Econoflour,                                      and counted with                   pacity    of each mRNA           was assayed     in Xenopus      oocytes.   Mes-
  an efficiency         for tritium        of 22% in a Packard                  3385 Liquid Scintillation                       senger      RNA       obtained    from   the reticulocytes      of an anemic
                                                                                                                                animal,      heterozygous      for Hb A and B, directed       the synthesis      of
      2 A. W. Steggles,              manuscript             in preparation.                                                     PA-, pH-, @-, and cu-globin            (Fig.   1A). The proportion        of the
                                                   Complementary            DNAs          for Sheep      Globin        mRNAs

                                                                                                                       1      E.COLIDNAfRKER                         A

                                                                                                              ;                                 POOLED FRACTIONS
         12                                                                                                   E
                                                                                                              n   1
         10                                                                                                   E


                                                                                                                   0          5
          4-                                                                                                                      FRACT::,   NV&          ‘&ToM



                                                                                                                                                                                        Downloaded from by guest, on July 26, 2011
                                                                                                                   0                    25                50
                                                                                                                            MILLIMETERS      OF MIGRATION
                                                                                               FIG. 2. Preparation       of full-length       cDNAs. A, preparative        alkaline
                                                                                            sucrose gradient.     The arrow         indicates     the center of 400 to 500 base
                                                                                            Escherichia     coli DNA marker           added to the gradient.        The gradient
                                                                                            was from 5 to 20% sucrose and contained                  0.1 M NaOH,     0.9 M NaCl,
                                                                                            and 10 mM EDTA. Centrifugation                 in a SW 41 rotor was at 40,000 ‘pm
        18.                                                                                 for 18 h. B, analytic      polyacrylamide           gel in formamide.      The pooled
                                                                                            fractions   in A were analyzed              as described      under “Materials        and
                                                                                            Methods.”      The positions    of RNA markers             included   in each gel are
        141                                                                                 indicated;    “LY” and “p” refer to the cu-mRNA                 and P-mRNA       compo-
                                                                                            nents of 10 S globin mRNA.

        ‘01                                                                                 dient centrifugation      (Fig. 2A). Complementary         DNA sedi-
                                                                                            menting in the indicated fractions was recovered and analyzed
         6/                                                                                 by polyacrylamide     gel electrophoresis in the presence of 96%
                                                                                            formamide (Fig. 2B). By comparison with mobility of DNA
                                                                                            markers of known size, in this system (data not shown), the
                                                                                            estimated chain length of the predominant         radioactive cDNA
                                                                                            was roughly 600 to 650 nucleotides. Since sheep globin mRNA
                                     FRACTION NUMBER
                                                                                             (6, 24) and other globin mRNAs are roughly 600 to 700 nucleo-
                                                                                            tides long, the cDNA preparations        appear to represent nearly
     FIG. 1. Translation      of sheep globin mRNAs.       mRNA was microin-
jetted into Xenopus        oocytes and these were incubated         for 48 h at 20”.        the entire lengths of their respective mRNA templates. Less
 Globin was prepared          and fractionated      by carboxymethylcellulose               than 3% of each cDNA became insensitive to S, nuclease when
 chromatography         as discussed    in the text. The elution positions           of     incubated in the absence of globin mRNA (self-annealing)        and
 authentic     globin markers      added prior to chromatography            are indi-       90 to 100% of each probe was protected from S, nuclease by
 cated by the appropriate         symbols.    Since the columns were not run
                                                                                            hybridization   to saturating amounts of its template mRNA at
 simultaneously       there is slight variation     in the position     of the indi-
 vidual    globin peaks (e.g. a) in A, B, and C. A, mRNA                     from an        50”.
 anemic Hb A/B heterozygote;           B, Hb F mRNA; C, Hb C mRNA.                            Hybridization           Analysis        of Unfractionated           Globin   cDNAs   -
                                                                                            When the annealing reactions were performed in 50% formam-
various   /3 chains        produced  in     the oocytes       was equivalent        to      ide at 50” there was extensive cross-hybridization      between
that synthesized          by the intact       reticulocytes      from which       the       each cDNA and the heterologous mRNAs. Annealing of Hb B
mRNA was obtained.               Thus the oocytes seemed to translate                       cDNA (a/3*) to its template and to Hb F mRNA (ocy) is illus-
each /3-globin mRNA                with      equal    facility.   Messenger    RNA          trated in Fig. 3A. Similar results were obtained with other
from    fetal   reticulocytes         directed      the synthesis      of a- and y-         cDNAs. Better discrimination   was obtained at 78” in aqueous
 globin but     very little     ,f3*-globin      (less than 5% of y) (Fig. 1B).             media containing 2.5% sodium dodecyl sulfate (Fig. 3B). These
Similarly, the Hb C mRNA contained less than 5% /P mRNA                                     conditions have been shown to be sufficiently stringent to
contamination     (Fig. 1C) and Hb A less than 5% p” mRNA                                   prevent cross-reaction between human p- and y-cDNA and
(data not shown).                                                                           their mRNAs (21,22). Nonetheless partial cross-reaction of the
  Synthesis   and     Characterization of cDNAs -Each   globin                              sheep y-mRNA with the pB sequences in the Hb B cDNA
mRNA was incubated with RNA-directed           DNA polymerase                               seemed likely since the plateau achieved at high RNA inputs
and the product fractionated by alkaline sucrose density gra-                               was greater     than  one-half of the level  of hybridization
                                                                 Complementary                    DNAs            for Sheep           Globin            mRNAs                                                                            1911
                                                      Hb B cDNA
                                                                                                                                                                             -Hb          B mRNA         la. fiBI
                                                                                                                                                       Hb B cDNA      VS.
                                                                                                                                                                             o---oHb       F mRNA        ((1, y)

                                                                                                                                       01          I           I                              1
                                                                                                                                                  50                    50                   70                     80
                                                                                                                                                                            TEMPERATURE           (“C)
                                                                                                                       FIG. 4. Melting        profiles of homologous        and heterologous       duplexes
                                                                                                                    formed      with Hb B cDNA.           Hybridization        conditions    and thermal
                                                                                                                    denaturation       techniaues     are discussed under “Materials             and Meth-
                                                                                                                    ods. The dotted line indicates          the plateau       between the two melting

                                                                                                                    transitions     and defines the relative          amounts     of 01- and fi-cDNA.

                                                                                                                     mRNA           was then accomplished                              by incubation                for 5 min at 68”.
                                                                                                                     The y- or P-cDNA                      was then separated                        from cu-cDNA-a-mRNA

                                                                                                                                                                                                                                                    Downloaded from by guest, on July 26, 2011
                I;I-r-r                                                                 ,
                                                                                                                     duplexes         by batch chromatography                                on HAP as shown in Fig. 5.
                                                                                                                     In each preparation                         90 to 99% of the initial                              cDNA         was re-
                                                                                                                     covered         (Table            I); the single-stranded                             and double-stranded
                                                                                                                     cDNA         components                  eluted         abruptly           from         the column              as two
                                                                                                                     distinct       peaks after application                             of the appropriate                   buffer     (Fig.
                                         1.0                   2.0                    3.0                             5). The separated                    peaks         were chromatographed                              on Sephadex
                                                  ng mRNA      added                                                  G-50 to remove                  sodium          phosphate,              precipitated               with ethanol,
   FIG. 3. Hybridization     of Hb B cDNA to Hb B mRNA            and Hb F                                           and then fractionated                          on alkaline               sucrose           density         gradients.
mRNA. A, in 50% formamide           containing  0.5 M NaCl, at 50” for 48 h.                                         About       70% of the radioactivity                             sedimented              in the region            of the
B, in 0.2 M sodium       phosphate,      pH 6.8, containing   0.5% sodium
dodecyl sulfate at 78” for 48 h.                                                                                     gradient          used to isolate                    full-length             cDNA            (cf. Fig. 2A). As
                                                                                                                     shown        by Fig. 6, each cDNA                            isolated         from the high molecular
                                                                                                                     weight        region          of the gradients                  consisted          predominantly                 of full-
 achieved       with Hb B mRNA                    (Fig. 3B). Since prolonged                       incuba-
                                                                                                                      length       molecules.               Similar           results        were obtained                   in analyses
tion at this temperature                       results       in extensive              breakdown             of
                                                                                                                     performed             in phosphate                buffer          (data not shown)                    under      condi-
 RNA,3       we did not attempt                    to exploit          the partial            specificity
                                                                                                                     tions shown                by Maniatis                et al. (25) to prevent                         zone disturb-
achieved        in the aqueous              hybridization             system.
                                                                                                                     ances sometimes                      observed           in barbital-buffered                       gels.
     The thermal          stability       of duplexes          between         Hb B cDNA               (@)
                                                                                                                         Table I summarizes                        the fractionation                     of each cDNA                by this
and Hb F mRNA                 (ay) in 50% formamide                      at 50” were compared
                                                                                                                     procedure.             Aliquots          of each reaction                  mixture            were assayed              for
to that of homologous                   duplexes         formed         between          Hb B cDNA
                                                                                                                     S, nuclease               resistance             after        the annealing                   at 50” and after
 and its template               mRNA           (Fig.      4). The T, (temperature                            at
                                                                                                                     thermal         denaturation                  at 68”. Complete                    (92 to 100%) hybridiza-
which 50% of the duplexes                     remain)         of Hb B cDNA-Hb                   B mRNA
                                                                                                                     tion of each cDNA                       to the heterologous                        mRNA            was routinely
duplexes         was 72.8” and only a single                         melting         transition           was
                                                                                                                     observed;          40 to 60% of the hybrids                             were denatured                    by incuba-
 apparent.        In contrast,          heterologous             duplexes        (Hb B cDNA-Hb
                                                                                                                     tion at 68”. In certain                         experiments                 the percentage                   of cDNA
F mRNA)           consisted         of two distinct           classes differing              in thermal
                                                                                                                     protected          from S, nuclease                       after incubation                    at 68” was some-
stability.        Denaturation              of the less stable,                or “low          melting”
                                                                                                                     what       higher           than expected                   on the basis of data obtained                                as
 component          (T,,, = 65.6”) was essentially                       complete          at 67”, but,
                                                                                                                     shown         in Fig. 4. This discrepancy                                    was likely              due to rapid
as indicated          by the hybrid             plateau         between        67 and 70”, dena-
                                                                                                                    reannealing                 of some of the cDNA                               after        incubation            at 68”.
turation        of the more              stable,       or “high          melting”           component
                                                                                                                     Because         of the large volumes                           (1 to 3 ml) required                    for prepara-
 (T,,z = 75”) did not commence                             until      the temperature                     was
                                                                                                                    tive reactions,                  instantaneous                    freezing           of the mixture                 after
raised to 70”. This plateau,                    which separated               these two melting
                                                                                                                    thermal          denaturation                   could         not be achieved                     even when            the
transitions,         defined       the relative         proportions           of the low melting
                                                                                                                     mixtures         were incubated                     in lOO-~1 aliquots.                     Some reannealing
 (60%) and high melting                      (40%) components.                  It seemed             likely
                                                                                                                    during        the period of cooling                        probably           occurred           since the cDNA
that the low melting                   component            represented            the imperfectly
                                                                                                                    and mRNA                     were        present             in high             concentrations.                 Conse-
matched         duplexes         between          Pa-cDNA           and y-mRNA                 while the
high melting             component            represented            perfectly          matched           a-~(      quently,         most of the cDNA purifications                                     were characterized                   by
                                                                                                                    a reduced             yield of single-stranded                            (/3 or y) cDNA                   and some-
                                                                                                                    what higher                level of contamination                          of the double-stranded                         o(-
    Fractionation            of Sheep Globin                  cDNAs        -Preparative                 reac-
                                                                                                                    cDNA         fraction           by p- or y-cDNA                       (Table        II).
tions between             each mixed             cDNA         and a heterologous                    mRNA
                                                                                                                         Purity      of a-, /%, and y-cDNAs                              -The         purity        of the individual
were performed               at 50” in 50% formamide.                         Selective          thermal
                                                                                                                     y- and fi-cDNA                  preparations                was estimated                  by two techniques.
denaturation            of p- or y-cDNA                 sequences            which         had formed
                                                                                                                     First,     each cDNA                   was annealed                   to the mRNA                  from which             it
partially        mismatched             duplexes        with the heterologous                        non-a-
                                                                                                                     was derived              (e .g p” -cDNA                  to Hb A mRNA)                      and to the heterol-
     s Unpublished            observations                                                                           ogous mRNAs                      (e.g.       /3”-cDNA               to Hb B, Hb C, and Hb F
1912                                                     Complementary                      DNAs      for Sheep         Globin    mRNAs

                                                                           STRANDED”       (al                 ,2                                A
                                                   STRANDED”                       i                                             ADULT   cDNA                          FETAL cDNA

                                                   0.16M                   0.5M

                                                       t,,                    J-
                                                                      4                                             r                            s     O.*r.


                                      FRACTION     NUMBER
   FIG.   5. Hydroxylapatite         chromatography         of Hb B cDNA          after
partial thermal      denaturation.        Conditions    for preparative  hybridiza-
tion, selective thermal        denaturation,        and HAP chromatography           are
described     under “Materials       and Methods.”

                                                                                                                                                                                              Downloaded from by guest, on July 26, 2011
                                           TABLE   I                                                                                             C

                     Fractionation       of sheep plobin cDNA                                                                     ,lB cDNA
                                      Per cent hybridization
                              __~                                                        Per cent
  cDNA        mRNA                                               68                        recovery
                               50” s, nu-                                               from HAP’             1.2
                                 ClGiSP            3eZl                   HAPb

  HbA          Hb    F                92               52                  57.1                  97
  Hb B         Hb    F               100               43                  49                    99
  Hb C         Hb    F               100               40                  30                    92
  HbF          Hb    B                95               50                  53                    90              L
                                                                                                               00                 32             64
     U Per cent hybridization      measured   by S, nuclease was defined as                                                                     m m MIGRATION
the per cent of the input cDNA resistant             to S, nuclease.                                        FIG. 6. Analytical   formamide      gel analysis of fractionated        cDNAs.
    Ir Per cent hybridization        by HAP represents       the per cent input                          Each cDNA      was subiected      to electrophoresis      as described       under
cDNA eluting in the double-stranded            fraction.                                                 “Materials   and Methods.”    Marker      RN&      were included    in each gel.
     ’ Per cent recovery      is defined as the percentage        of input cDNA                          The profiles shown were normalized          so that the positions     of the 28 S,
eluting in the single- and double-stranded             peaks (cf. Fig. 5).                               18 S, and 10 S RNA markers         were in alignment.

mRNAs). Increasing amounts of mRNA were added and the                                                   contaminated          with 25 to 30% y-cDNA sequences. Only 70 to
reactions were allowed to incubate until equilibrium              was                                   75% of the duplexes melted at 75-76”. The T,,, of the remaining
achieved (saturation analysis). Incubation in 50% formamide                                             25 to 30% was 70” when the duplexes were formed with Hb F
was at 68”; we reasoned that cross-reaction of heterologous                                             mRNA, and 65” when the duplexes were formed with Hb B
species (e.g. p’-cDNA to p“-mRNA) would not happen if the                                               mRNA. Since these T,,, values are characteristic of y-y and y-
temperature    at which annealing occurred was above that of                                            /P duplexes, respectively, the contaminant is almost certainly
the T,,, of the heterologous        duplexes. Hence the plateau                                         y cDNA. Conversely, CPCDNA prepared for Hb B cDNA also
achieved when a given cDNA was saturated with a heterolo-                                               contained 20 to 30% non-a-cDNA               sequences; the T,,, values of
gous mRNA containing a different non-oc species represents                                              the contaminating             components were those expected of pB-
the amount    of ol-cDNA contaminating       that preparation.   Rep-                                   cDNA.
resentative data for y- and PR-cDNA are shown in Fig. 7 and                                                 Characterization          of Thermal  Stability   of Homologous      and
the results are summarized in Table II.                                                                 Heterologous         Duplexes     Formed  with Purified     cDNAs    -Each
   The second method employed to estimate the purity of the                                             purified cDNA was annealed at 50” in 50% formamide to each
individual cDNAs is illustrated by the results depicted in Fig.                                         of the various sheep mRNA preparations                   and the thermal
8. Heterologous duplexes were formed at 50” (e.g. between /3”-                                          stability of the resulting duplexes was determined (Fig. 8 and
cDNA and Hb F (a + -y)-mRNA) and a melting profile was                                                  Table III). The T,,, values of the homologous p and y duplexes
obtained. As was observed with mixed (unfractionated)           cDNA                                    were 70-71” whereas the T,,, values of the heterologous /3 and y
(Fig. 41, two components     of differing thermal      stability were                                   duplexes were 65-66”. Very similar results were obtained
present but now the higher melting component (a+ duplexes)                                              when reciprocal tests were performed (Table III). For example,
was only 5 to 20% of the total. The results obtained with /3”-                                           PB-cDNA annealed to Hb F cay)-mRNA gave duplexes with a
and y-cDNA are shown in Fig. 8, A and B, respectively, and                                              T,,, of 66.5” while y-cDNA annealed to Hb B (@)-mRNA                    gave
the data are summarized in Table III.                                                                   duplexes with a T,,, of 65.5”. This small difference is within the
   Only the second method of estimating purity was applicable                                           range of experimental            error. WCDNA formed duplexes having
for the c*-cDNAs since no mRNA lacking cx sequences was                                                 T, values of 75-76” regardless of which mRNA preparation
available. The thermal denaturation        profiles shown in Fig. 8C                                    was included in the reaction (Fig. 8C and Table III). The cy-
suggest that the a-cDNA prepared from Hb F cDNA was                                                     cDNA thus annealed to perfectly complementary                     mRNA se-
                                                                         Complementary                             DNAs        for Sheep    Globin          mRNAs                                              1913

                                                TABLE             II                                                                               BETAB     cDNA
                              Purity           of p       and      y-cDNAs                                                                                                                           A.
                                                                                                                                                   M          Hb B mRNA       (a, /IBI
                                                                              a cDNA           Content
     cDNA                                                                                                                                          +-+,Hb           F mRNA   (a, 7)
                     tested                       Saturation                  hybridiza-                 Thermal       dena-
                                                                       tima                                  turationb
                                                                              %                                       %
                     Hb B                             Cross-hybridized                                          5-8
                     Hb C                                   5-10                                                5-10
                     Hb F                                   N.T.’                                               5-8

                     Hb A                             Cross-hybridized                                          5-8
                     Hb C                                   8-10                                                5-8
                     Hb F                                         5-10                                          S-10

                     Hb A                                         10-15                                        15-20
                     Hb B                                         15-20                                        15-20
                     Hb F                                         15-20                                        E-20

                     Hb A                                         N.T.                                          5-10
                     Hb B                                         S-10                                          S-10                                            TEMPERATURE              (“C)
                     Hb C                                         S-10                                          S-10

    a The cu-cDNA measured       by saturation        hybridization       represents
the maximum       per cent hybridization       attained       by hybridization         to

                                                                                                                                                                                                                        Downloaded from by guest, on July 26, 2011
                                                                                                                                                   GAMMA        cDNA
excess heterologous    mRNA at 68” (cf. Fig. 7).                                                                                                                                                     8.
    b o( cDNA content    measured       by thermal       denaturation        analysis
was defined as the per cent cDNA in the high melting component                      (cf.
Fig. 8).
    ’ N.T. = not tested.


                                                                                           .   (1 cDNA
                                                      o-------o                                flB &NA

                              AeS---                       _____                           A ycDNA
                                                                                                                                                   50                  60                       70

                                                                                                                                                               TEMPERATURE            I°C)

                                                                                                                                                   ALPHA      cDNA
                                Hb F mRNA             Ia, “I
                   100 c

                                                                                               a cDNA
                                                                                               7 cDNA

                              --o----------o                                                   pB cDNA
                                                   I                                    1
                     0             2.5            5.0                                 10.0
                                          NANOGRAMS                     OF RNA
   FIG.  7. Saturation hybridization       with p”- and y-cDNA.     Incuba-
tions in 50% formamide   containing      0.5 M NaCl were conducted     at 68”
for 60 h. A, Hb B mRNA          uers’sus (Y-, p”-, and y-cDNAs;   B, Hb F

quences  present in each of the globin mRNAs; in contrast, the                                                                                                       TEMPERATURE             1%
non-cu-cDNAs annealed to perfectly complementary sequences                                                                         FIG.    8. Thermal    denaturation      of duplexes     formed    with o(-, p”-.
present only in homologous globin mRNA.                                                                                          and y-cDNAs.       The procedure      for melting     temperature      analysis   is
   The thermal stability of homologous and heterologous du-                                                                      described    under “Materials      and Methods.”
1914                                                               Complementary                 DNAs            for Sheep    Globin          mRNAs
                                               TABLE         III                                                                                            SHEEP y cDNA

  Melting       temperatures            of homologous              and heterologous           duplexes
                               HbA                    HbB                   Hb C                 Hb F
         PA                    70.5"                  68.8                  66.0                 64.8
         ic B                  69.0
                               66.2                   71.5
                                                      66.0                  65.5
                                                                            70.3                 65.3

         Y                     64.8                   65.5                  66.0                 70.8
         UFb                   76.0                   76.5                  75.5                 76.8
         a,*                   76.5                   75.0                  76.0                 75.8

    a Values given are T, values for the major component        and were
derived from melting profiles obtained     as described under “Materials
and Methods”      (cf. Fig. 8).                                                                                                                 TEMPERATURE     (“Cl
    * up, CQ: (Y-cDNA      isolated from Hb F cDNA and Hb B cDNA,                                                      FIG. 9. Thermal    denaturation      of hybrids   formed with y-cDNA.
respectively.                                                                                                       Incubations    were performed      at 65” for 48 h in 0.2 M sodium phos-
                                                                                                                    phate, pH 6.8, containing      0.5% sodium dodecyl sulfate.     0- - -0, Hb
                                                                                                                    F mRNA;     O-O,      Hb B mRNA;        O----O,    Hb C mRNA.
plexes formed with y-cDNA in aqueous solution at 65” was also
determined (Fig. 9). The y-cDNA.y-mRNA        duplexes melted
                                                                                                                    component purified from mixed cDNA anneals to all mRNA
with a sharp transition     at 93.5”, whereas the majority of
duplexes formed between y-cDNA and Hb C (cupC‘)-mRNA and                                                            preparations, a result compatible with it being predominantly
Hb B (@‘)-mRNA       melted in each case with a T,,, of 88”. Also,                                                  oc-cDNA. In experiments described elsewhere4 each of the p-
                                                                                                                    and y-globin cDNAs specifically recognized only homologous

                                                                                                                                                                                                  Downloaded from by guest, on July 26, 2011
approximately   20% of the y-cDNA became sensitive to S,
nuclease at temperatures    between 65-85”.                                                                         globin mRNAs in sheep reticulocyte samples obtained from
                                                                                                                    animals heterozygous for Hb A and Hb B or during different
                                               DISCUSSION                                                           stages of fetal ontogeny and hemoglobin           switching.     In all
                                                                                                                    cases, estimates of mRNA content by saturation                analyses
   We have described a method by which preparation        of sheep
                                                                                                                    correlated well with the translational    capacities of each globin
cy-, PA-, pR-, p“-, and y-globin cDNAs can be accomplished
                                                                                                                    mRNA preparation.         We thus conclude that the fractionated
with satisfactory yield of approximately      full-length product.                                                  cDNAs represent highly purified DNA sequences specific in
This method was developed because other techniques availa-                                                          annealing reactions for the individual globin mRNAs.
ble for purification of specific mRNAs (e.g. preparation       of (Y-                                                   Our analysis of the sheep globin mRNAs and their cDNAs
mRNA from postribosomal         reticulocyte supernatant   (12, 26-                                                 has yielded two unexpected observations. The first of these is
281, generation of heavy (/3) and light (01) polysomes by incuba-                                                   the difference in T,,, of (Y- and non-oc-cDNA-mRNA           duplexes.
tion    of reticulocytes              with     the isoleucine            antagonist,          O-methyl-
                                                                                                                    Homologous duplexes formed with non-cr-globin             cDNAs had
threonine (12, 28-32) or recovery of separate CP and p-mRNA                                                         T,,, values between 69.5 and 71.5” (Table III), whereas duplexes
from polyacrylamide      gels run in 98% formamide (23, 24, 33-                                                     formed with ol-cDNA had T,,, values of 75-76” regardless of the
37)) either were not applicable or resulted in a poor yield of                                                      mixed cDNA used as a source of a probe. These differences in
purified mRNA. Human y-globin cDNA has been purified by                                                             T,n were observed with every sheep globin mRNA preparation
hybridization  of fetal reticulocyte mRNA ((u, /3, and y) to adult                                                  tested including       mRNA fractions not used to synthesize
reticulocyte  mRNA (o(, p) under stringent conditions followed                                                      cDNA. Thus there appears to be an intrinsic difference in
by recovery of the single-stranded      c-y) and double-stranded
                                                                                                                    thermal stability among duplexes formed by (Y-cDNA and
cDNA (a and p) by hydroxylapatite       chromatography      (38-40).
                                                                                                                    duplexes formed with the various p- or y-cDNAs. We have not
Human           o(- or /3-cDNA               has    also     been      prepared         using     similar
                                                                                                                    yet elucidated a structural basis for this phenomenon. How-
techniques (40-46). For the sheep globin cDNAs however, it                                                          ever, one very plausible explanation is that the G + C content
 was more convenient and reproducible to perform the anneal-                                                        of a-cDNA may be higher than that of the p- or y-cDNAs since
ing reaction at lower temperature          and generate single-
                                                                                                                    duplexes rich in G-C base pairs are known to have increased
 stranded globin-specific    cDNA by brief incubation above the
                                                                                                                    thermal stability (47, 48).
T,,, of heterologous    duplexes. The separation of the desired
                                                                                                                        The second unexpected observation was the relatively small
 non-a (/3 or y)-globin cDNA from the a~-cDNA was facilitated
                                                                                                                    difference in thermal stability between homologous and heter-
by the unexpected finding that homologous duplexes formed                                                           ologous duplexes formed with p- or y-cDNAs (Table III). The
 with oc-cDNA had T,,# values considerably higher than homolo-
                                                                                                                    AT,?, values range from roughly 2” for PA-P” duplexes to 5.0-6.0
 gous duplexes formed with non-a-cDNA       (Fig. 8 and Table III).
                                                                                                                    for PB-/3c, pB-r, and /Y--y duplexes. These small differences
    That the low melting component formed in each case when
                                                                                                                    were not greatly dependent on the conditions used for the
 cDNA was annealed to a heterologous mRNA represents du-
                                                                                                                    hybridization    reaction. Thus when heterologous and homolo-
plexes between the non-a species was demonstrated by the fact                                                       gous duplexes were formed with y-cDNA in aqueous solution
 that purification of this fraction of the mixed cDNA yielded a                                                      (Fig. 9) the AT,,, was also only 5”. In general the AT,,, values
probe       which,     under            stringent          conditions,             annealed       only      to
                                                                                                                    paralleled the differences in amino acid sequence observed for
RNAs which might be anticipated        to contain homologous                                                        the corresponding globin chains. Thus, for example, the small-
mRNA sequences (Fig. 7 and Table II). Cross-hybridization                                                           est AT,,, (1.7-2.5”) was obtained with /3”-/3” hybrids; p”- and
between cDNA and mRNA sequences specific for PA- and pB-
                                                                                                                     /P-globin chains differ in only 7/145 positions (Table IV).
globin was routinely observed (Table II). Since the T,,! values
                                                                                                                        Several laboratories (49-55) have shown that AT,,, is related
of PA- and PB-specific nucleotide sequences differed by only                                                        to the number of nucleotide differences (sequence divergence)
1.5”, this finding was not surprising. In all other cases, com-
plete specificity was obtained. In contrast the high melting                                                           4 E. J. Benz,   Jr.,    manuscript     in preparation.
                                                                       Complementary                        DNAs       for Sheep Globin                 mRNAs
                                                  TABLE          IV                                                      chains     having          similar          amino         acid      sequences.            For     example,         as
 Estimated          nucleotide         sequence        diverpence             of sheep plobin               mRNA         shown in Table IV, we observed a AT,,, of 15” for human               p-y
                                 No.                                                                                     hybrids   using the same hybridization        system in which the AT,,,
                                                   Minimum                     Minimum
    Hybridizing               a!m”o
                             acids As             per cent se-                 predicted              Observed           values of sheep y-/3i1 and r-p duplexes were only 5”. The
                                                   quence di-                                           ATmd
                              in glo-              vergencc?                      AT,’                                   amino acid sequence divergence of human /3- and y-globins
                                                                                                                         (39/146 differences   (57,581) is nearly    comparable    to those of the
PA versus 6”                       6-7              0.92-1.1                  1.5-     3.7            1.7-2.5”           sheep p’=y (27/145)- and p-y (34/145)-globins. Also, Dalgleish
PA versus /3”                    16-18                2.4-3.0                 3.8-10.0                3.5-4.1            et al. (691, using unfractionated        (a + p)-cDNAs observed a
/3” versus y                     22-24                3.3-4.0                 5.2-13.6                5.7-6.0            AT, of 5” for human and baboon Hb A mRNAs under annealing
p” versus /3’                    19-22                2.9-3.6                 4.6-12.2                4.3-6.0
                                                                                                                         conditions nearly identical with the formamide system de-
pB versus y                      25-27                3.8-4.5                 6.0-15.3                5.0-5.3
PC versus y                      29-34                4.4-5.6                 7.0-19.0                4.8-5.0
                                                                                                                         scribed in this report. This figure corresponds to the largest A,,,
fF vemus y                       25-27                3.8-4.5                 6.0-15.3                5.0
                                                                                                                         values observed with sheep mRNAs; however, there are only
   (aqueous)                                                                                                             nine amino acid differences in each globin chain (cu and p)
PC versus y                      29-34               4.4-5.6                  7.0-19.0                5.5                between the human and baboon globins. Finally, when other
    (aqueous)                                                                                                            mixed (a + /?) mammalian and avian mRNAs were studied by
Human       p*                   39                  6.0-6.5                  9.6-22.1               15.5                interspecies cross-hybridization       (12,22,55,56),   the AT,,, values
   versus Y                                                                                                              were     considerably              greater        than       expected           on the basis           of amino
   (1 The     number        of amino      acid     differences           was estimated            by compari-            acid differences alone. Possibly, the rather complex hemoglo-
sons of published      sequences (57, 58).                                                                               bin switching mechanisms which exist in sheep erythroid cells
     b Minimum    per cent sequence divergence          was calculated   assum-                                          may require strict conservation of the secondary structure of
ing that each amino acid difference          arose by single nucleotide   substi-                                        the p- and y-mRNAs; thus, the nucleotide sequences of these
tution; unique globin mRNA           and cDNA sequences were assumed to                                                  mRNAs may be subject to more severe constraints, than those

                                                                                                                                                                                                                                                 Downloaded from by guest, on July 26, 2011
be 600 to 650 bases long.
                                                                                                                         of globin mRNAs from other species.
     c Minimum    predicted    AT,,, was calculated    as discussed in the text:
                                                                                                                            Hemoglobin    synthesis in sheep and goat erythroid cells
AT, = 1.6-3.4”      x % sequence      divergence.
     d Observed  AZ’,,, were derived from Table III.                                                                     constitutes an attractive system for the study of the phenom-
                                                                                                                         ena important in the regulation of the expression of individual
                                                                                                                         globin genes at the molecular level. In addition, the similari-
between        nonidentical         hybridizing       sequences.       The reduction              in                     ties between F+A switching in sheep and humans suggest
T,,, has been variously                estimated       at 1.6-3.4”      for each 1% se-                                  that the sheep may be a useful model system for studying the
quence       divergence.         The minimum              per cent sequence               diver-                         analogous process in human erythroid cells. The purified
gence between            each pair of sheep globin                mRNAs can be pre-                                      cDNAs we have prepared provide specific molecular hybridiza-
dicted     from the number                of amino       acid differences,          listed       in                      tion probes for DNA-RNA       and DNA-DNA      hybridization   as-
Table     IV, between           the corresponding             globins.     These       figures                           says designed      to enhance     the precise     analysis                                of the molecular
represent       conservative          estimates      of the amino        acid differences                                events  regulating      hemoglobin      switching.
since portions           of the sheep globin              sequences       have been as-
signed by interspecies               homology        with goat and bovine                hemo-                              Acknowledgments-We                             are grateful    to Dr.                   W. French   An-
globins      (56,571.     The majority          of the amino acid differences                 can                        derson,    for his generous                    support,   encouragement,                      and helpful
be explained          on the basis of single              base substitutions             in the                          discussions. We wish to thank Dr. Joseph E. Pierce for his
corresponding           mRNA.                                                                                            assistance in procuring sheep blood samples, and Dr. Bernard
    One generally           expects     the nucleotide        sequence       divergence            to                    Forget for much helpful advice concerning polyacrylamide   gel
be somewhat greater than the predicted minimum             because                                                       electrophoresis          in formamide.       We are endebted     to Ms. Patricia
many amino acids can be encoded by two or more codons (58).                                                              H. Turner,          who provided       excellent   technical    assistance,    and
Moreover,   globin  mRNA is 600 to 700 nucleotides long (6, 23-                                                          Mr. Bernard           Kefauver,     who performed       the carboxymethylcel-
25, 33-37). Allowing 50 to 70 bases for the poly(A) tail (59-681,                                                        lulose chromatography;    to Mr. Mac McCaskill and Mrs. James
each globin mRNA contains 90 to 150 “untranslated           nucleo-                                                      Hoes, who assisted with animal surgery; to Mr. Leonard
tides” in addition to the 450 needed to code for the amino acid                                                           Stuart and his associates at Poolesville Animal Center, and to
sequence of the globin chain (64-68). Assuming a unique base                                                             Mr. Max S. Foltz, Mr. Jesse N. Judy, and Mrs. Donna Ma-
sequence length of 600 to 650 bases, the absolute minimum                                                                thews at the Max Foltz Farms, who cared for the animals. We
base sequence divergence for any pair of mRNAs is thus given                                                             are especially   grateful      to Mrs.                     Exa Murray    and Mrs.                     Margaret
by the formula: min % = number of amino acid differences +                                                               Motter   for expert     preparation                        for the manuscript.
600 to 650. Therefore, using the aforementioned          estimates                                                                                                     REFERENCES
from the literature, the minimum predicted AT,,, can be calcu-
                                                                                                                          1. Huisman,          T. H. J., Lewis,                J. P., Blunt, M. H., Adams,    H. R.,
lated: minimum AT,,, = 1.6 to 3.4 x min %. We have applied                                                                     Miller,        A., Dozy, A. M.,                and Boyd, E. M. (1969) Pediatr.   Res.
the above formula    to estimate the sequence homologies     among                                                                3, 189
sheep globin mRNAs. As shown in Table IV, the measured                                                                    2. Hammerbere.         B.. Brett. I.. and Kitchen.   H. (1974) Ann. N. Y.
AT,,, values were close to the predicted minimum          in every                                                             Acad. Sci.-241,‘672-682
                                                                                                                          3. Bard, H., Battaglia,        F. C., Makrowski,   E. L., and Meschia, G.
case, suggesting that sequence divergence among the mRNAs
                                                                                                                                1972. Proc.    Sot. Exp. Biol. Med. 139, 1148-1150
coding       for    sheep      globins           is very         close    to the           minimum            neces-      4. Thurmon,            T. F.,       Boyer,      S. H.,          Crosby,      E. F.,      Shephard,          M. K.,
sary to generate the known amino acid differences. A corollary                                                                    Noyes,      A. N., and Stohlman,                         F. (1970) Blood            36, 598-606
of our observation is that the minimum       estimates relating                                                           5. Gabuzda,          T. G., Schuman,         M. A., Silver,                    R. K.,      and    Lewis,     H. B.
AT,,, to sequence divergence (1.6” AT,,, per 1% divergence)                                                                     (1968)       J. Clin.  Invest. 47, 18951904
                                                                                                                          6. Kazazian,          H. H., Jr., Silverstein,      A. M.,                     Snyder,         P. G., and       Van
appear to be the most accurate in the two hybridizing systems.                                                                    Beneden,          R. J. (1976) Nature                    20, 67-70
  Sheep globin mRNAs     appear  to be more closely  homologous                                                           7. Nienhuis,         A.     W.,      and     Anderson,             W.     F. (1972)      Proc. Natl.        Acad.
than        other      mammalian                  globin              mRNAs              coding      for      globin              Sci.     U. S. A. 69, 2184-2188
1916                                                          Complementary DNAs for Sheep Globin mRNAs
 8. Baldy, M., Gaskill,    P., and Kabat, D. (1972) J. Biol. Chem.   247,                                             N&l.      Acad.       Sci.   U. S. A.       72, 258-262
       6665-6670                                                                                             39. Old, J., Clegg, J. B., Weatherall,             D. J., Ottolenghi,     S., Compi,
 9. Kacian,    D., Watson,   K., Burny,  A., and Spiegleman,    S. (1971)                                           P., Giglioni,     B., Mitchell,     M., Tolstoshev,     P., and Williamson,
         B&him.            Biophys.    Acta      246,   365-383                                                     R. (1976) Cell 8, 13-18
10. Barker,    J. E., Anderson,                W. F., and Nienhuis,               A. W. (1975) J.            40. Kan, Y. W., Holland,           J. P., Dozy, A. M., and Varmus,                N. E.
       Cell Biol.  64, 515-527                                                                                      (1975) Proc. Natl.      Acad.     Sci. U. S. A. 72, 5140-5144
11. Nienhuis,     A. W., Falvey,                  A. K.,      and Anderson,           W. F. (1974)           41. Taylor, J. M., Dozy, A., Kan, Y. W., Varmus,                  H. E., Lie-Injo,     L.
         Methods       Enzymol.         30, 621-630                                                                 E., Ganesan,       J., and Todd, D. (1974) Nature           251, 392-393
12. Velez, R., Kantor,       J. A., Picciano,  D. J., Anderson,     W. F., and                               42. Ottolenghi,      S., Lanyon,      W. G., Paul, J., Williamson,        R., Weath-
       Nienhuis,      A. W. (1973) J. Biol. Chem.      250, 3193-3198                                               erall, D. J., Clegg, J. B., Pritchard,              J., Pootrakul,      S., and
13. Lane, C. D., Marbaix,        G., and Gurdon,   J. B. (1971)J.     Mol. Biol.                                    Boon, W. H. (1974) Nature             251, 389-392
       61, 73-91                                                                                             43. Kan, Y. W., Dozy, A. M., Varmus,               H. E., Taylor, J. M., Holland,
14. Gurdon,      J. B., Lane, C. D., Woodland,       H. R., and Marbaix,         G.                                 J. P., Lie-Injo,     L. E., Ganeson,       J., and Todd, D. (1975) Nature
       (1971) Nature     233, 177-182                                                                               255. 255-256
15. Chan, L., Kohler,        P. O., and G’Malley,       B. W. (1976) J. Clin.                                44. Ottolenghi,      S., Lanyon,       W. G., Williamson,       R., Weatherall,        D.
         Inuest.     57, 576-585                                                                                             --.
                                                                                                                    J.. Cleaa, J. B., and Pitcher. C. S. (1975)Proc.             Natl. Acad.     Sci.
16. Falvey,   A. K., Kantor,    J. A., Robert-Guroff,       M. G., Picciano,     D.                                    U. S. A.         72, 2294-2299
       J., Weiss, G., Vavich,     J., and Anderson,        W. F. (1974) J. Biol.                             45. Tolstoshev,        R., Mitchell,         J., Lanyon,           G., Williamson,             R., Otto-
       Chem. 249. 7049-7056                                                                                          lenghi,      O., Comi, P., Giglioni,                   B., Masers,          G., Modell,          B.,
17. Weiss, G. B.; Wilson, G. N., Steggles, A. W., and Anderson,                 W.                                   Weatherall,         D., and Clegg, J. (1976) Nature                       259, 95-98
       F. (1974) J. Biol.   Chem.    249. 3425-3431                                                          46. Ramirez.        F.. Natta. C.. O’Donnell.                 J. V.. Canale. V.. Bailev. G..
18. Wilson,   G. N., Steggles, A. W., Kantor,          J. A., Nienhuis,    A. W.,                                     Sanguensermsri,’           T.: Maniatis;              G. M., Marks,              ‘P. A.; ‘and
       and Anderson,      W. F. (19751 J. Biol.      Chem.    250, 8604-8613                                         Bank, A. (1975)Proc.             Natl.Acad.            Sci. U.S.A.          72.1550-1554
19. Wilson, G. N., Steggles, A. W., and Nienhuis,             A. W. (1975) Proc.                             47. Nygaard,        A. P., and Hall, B. D. (1964) J. Mol. Biol.                         9125-142
         Natl.     Acad.       Sci.   U. S. A.      72, 4835-4839                                            48. Britten,      R. J., and Kohne, D. E. (1968) Science 161, 529-540
20. Steggles,       A. W., Wilson,          G. N., Kantor,          J. A., Picciano,             D. J.,      49. Laird,      C. D., McConaughy,                 B. L., and McCarthy,                   B. J. (1969)

                                                                                                                                                                                                                              Downloaded from by guest, on July 26, 2011
       Falvey,       A. K., and Anderson,              W. F. (1974) Proc. Natl.                  Acad.              Nature       224, 149-154
       Sci. U. S. A. 71, 1219-1223                                                                           50. Ullman,        J. S., and McCarthy,                  B. J. (1973) Biochim.                 Biophys.
21. Housman.         D., Fornet.       B. F.. Skoultchi.         A.. and Benz. E. J.. Jr.                           Acta     294, 405-415
        (1973) Proc: Nati:         Acad.     Sci. U. S. A’. 76, 180991813                           ’        51. Southern,        E. M. (1971) Nature              New Biol.           232, 82-83
22. Housman,          D., Skoultchi,       A.. Forget.       B. G.. and Benz. E. J.. Jr.                     52. McCarthy,          B. J., and Duerksen,                 J. (1970) Cold Spring                 Harbor
        (1974) Ann. N. Y. Acad. .‘?ci. 241,‘280-290                                                                 Symp.       Qua&.      Biol.    35, 621-629
23. Forget, B. G., Housman,                D., Benz, E. J., Jr., and McCaffrey,                         R.   53. Morrison,        M. R., Paul, J., and Williamson,                            R. (1972) Eur.             J.
       P. (1975) Proc. Natl. Acad.                Sci. II. S. A. 72, 984-988                                        Biochem.        27, l-9
24. Kazazian,        H. H., Ginder,        G. D., Snyder, P. C., Van Beneden,                          K.    54. Leder, P., Aviv, H., Gielen, J., Ikawa,                           Y., Packman,            S., Swan,
       J., and Woodhead,            A. P. (1975)Proc.          Natl.Acad.            Sci. U. S. A.                   D., and Ross, J. (1973) Cold Spring                      Harbor        Symp.      Quant.      Biol.
       72, 567-571                                                                                                   38, 753-761
25. Maniatis,       T., Jeffrey,      A., and Van deSande, H. (1975) Biochem-                                55. Gummerson,            K. S., and Williamson,                   R. (1974) Nature            247, 265-
       istry   14, 3787-3794                                                                                         267
26. Jacobs-Lorena,          M., and Baglioni,          C. (1972)Proc.          Natl.      Acad.      Sci.    56. Dayhoff,        M. 0. (1969) Atlas            ofProtein          Sequence and Structure,
       U. S. A. 69, 1425-1429                                                                                        Vol. 4, National         Biomedical          Research Foundation                  Publication,
27. Gianni,      A. M., Giglioni,          B., Ottolenghi,         S., Comi, P., and Gui-                            Silver Spring,         Md.
       dotti, G. G. (1972) Nature               New Biol.       240, 183-185                                 57. Wilson,       J. B., Edwards,           W. C., McDaniel,                  M., Dobbs, M., and
28. Temple, G. F., and Housman,                    D. E. (1972)Proc.           Natl.      Acad. Sci.                 Huisman,        T. H. J. (1969)Arch.              Biochem.         Biophys.       115.385-400
       U. S. A. 69, 1574-1578                                                                                58. Watson, J. (1976) Molecular                   Biology         of the Gene, 3rd Ed, W. A.
29. Hori, M., and Rabinovitz,               M. (1968)Proc.           Natl.    Acad.        Sci. U. S.                Benjamin,        New York, N.Y.
      A. 59, 1349-1355                                                                                       59. Gaskill,      P., and Kabat, D. (1971)Proc.                     Natl.     Acad. Sci. U. S. A.
30. Smulson,        M., and Rabinovitz,           M. (1968)Arch.           Biochem.         Biophys.                 68,72-75
        124,306-313                                                                                          60. Labrie,      F. (1969) Nature          221, 1217-1222
31. Rabinovitz,        M., Freedman,           M. L., Fisher, J. M., and Maxwell,                            61. Pemberton,          R. E., and Baglioni,              C. (1972) J. Mol. Biol.                65, 531-
       C. R. (1969) Cold          Spring      Harbor       Symp.      Quant.        Biol.     34, 567-               535
       578                                                                                                   62. Williamson,          R. (1972) Am. J. Med. Genet.                       9, 348-355
32. Kacian,      D L., Gambino,             R., Dow, L. W., Grossbard,                    B., Natta,         63. Burr, H., and Lingrel,             J. B. (1971)NatureNewBiol.                        233,41-43
       C., Ramirez,         F., Spiegleman,           S., Marks,        P. A., and Bank, A.                  64. Lim, L., and Canellakis,                 E. S. (1970) Nature                227, 710-712
        (1973) Proc. Natl.         Acad.     Sci. U. S. A. 70. 1886-1890                                     65. Marotta,        C. A., Forget,        B. G., Weissman,                  S. M., Verma,           I. M.,
33. Morrison,        M. R., Brinkley,          S. A., Gorski,         J., and Lingrel,             J. B.             McCaffrey,        R. P., and Baltimore,                  D. (1974) Proc. Natl. Acad.
        (1974) J. Biol.       Chem.      249, 5290-5295                                                               Sci.    U. S. A.         71, 2300-2304
34. Orkin,     S. H., Swan, D., and Leder, P. (1975) J. Biol. Chem.                                  250,    66. Forget,     B. G., Marotta,         C. A., Weissman,        S. M., Verma,       I. M.,
       8753-8760                                                                                                    McCaffrev.       R. P.. and Baltimore.         D. (1974)Ann.      N. Y. Acad.
35. Kazazian,        H. H., Snyder,           P. G., and Tu-Chen,                  C. (1974) Bio-                  Sci. 241,%0-309'
      them.      Biophys.      Res. Commun.             59, 1053-1061                                        67. Forget,    B. G., Marotta,          C., Weissman,      S. M., and Cohen-Solal,
36. Gould, H. J., and Hamlyn,                P. H. (1973)FEBSLett.                  30,301-304                      Tvf. (1975) Proc.     Natl.     Acad.  Sci. U. S. A. 72, 3614-3618
37. Shearman,         J. J., Hamlyn,          P. N., and Gould, H. J. (1974) FEBS                            68. Proudfoot,     N. J., and Brownlee,          G. G. (1974)Nature     252,359-362
      Lett.    47, 171-176                                                                                   69. Dalgleish,      R., Williamson,            R., and Tolstoshev,           P. (1976)
38. Lanyon,      W. G., Ottolenghi,             S., and Williamson,              R. (1975) Proc.                   Biochim.      Biophys.       Acta 435, 76-81

Shared By: