Proc. Natl. Acad. Sci. USAVol. 76, No. 9, pp. 4294-4298, September 1979Biochemistry

Structure of genes for human growth hormone and chorionicsomatomammotropin

(in vitro packaging/filter hybridization/intervening sequences/DNA sequences)

JOHN C. FIDDES*, PETER H. SEEBURG*t, FRANCES M. DENOTO*, ROBERT A. HALLEWELL*,JOHN D. BAXTER*t§, AND HOWARD M. GOODMAN*ftThe Howard Hughes Medical Institute Laboratory, Departments of *Biochemistry and Biophysics and §Medicine, and the IMetabolic Research Unit, Universityof California, San Francisco, California 94143

Communicated by Donald F. Steiner, June 1, 1979

ABSTRACT A 2.6-kilobase (kb) EcoRI restriction endonu-clease fragment containing human growth hormone (hGH; so-matotropin) gene sequences and a 2.8-kb EcoRI fragment con-taining human chorionic somatomammotropin (hCS; cho-riomammotropin) gene sequences have been identified by hy-bridization to cloned cDNA. Human DNA was cleaved withEcoRI and fractionated by preparative agarose gel electro-phoresis; DNA in the size range 2-3 kb was ligated to XgtWES'XB DNA and viable recombinant bacteriophage wererecovered by in vitro packaging. After infection of Escherichiacoli and screening of phage plaques, single isolates of hGH andhCS gene sequences were obtained. Restriction endonucleasemapping showed that the hGH gene contains three interveningsequences interrupting the coding sequence. Partial DNA se-quence analysis of the hGH gene, obtained by the chain ter-mination method, confirmed the location of the interveningsequences and the identity of the fragment.

Human growth hormone (hGH) and human chorionic soma-tomammotropin (hCS) are two closely related polypeptidehormones which have more than 80% of their 191 amino acidsin common (1, 2). They have different biological activities andare synthesized in different tissues: hGH in the pituitary andhCS in the placenta. The genes coding for these two hormonesprovide a good system for studying the organization of struc-turally related sequences and their tissue-specific expression.A 550-base-pair Hae III fragment cDNA clone coding for

amino acids 24-191 of the hCS sequence has been described(3). Recently, the analogous 550-base-pair Hae III fragmentcoding for hGH has also been cloned and its nucleotide se-quence has been determined (unpublished data). The nucleo-tide sequences of these two fragments are 93% homologous.

Further analysis of this system requires the isolation ofgenomic DNA fragments that contain the entire coding se-quences and the regulating elements that may be involved inthe differential expression of the genes. This report describesthe construction of genomic DNA clones for sequences of boththe hGH and hCS genes by utilizing bacteriophage Xgt WES-XB as the vector (4) and the bacteriophage X in vitro packagingsystem of Blattner et al. (5).

MATERIALS AND METHODSHuman Placental DNA. High molecular weight DNA was

extracted from human placenta obtained by caesarian section.The frozen tissue was dispersed in a blender, treated withproteinase K, extracted with phenol/chloroform, and then in-

cubated sequentially with RNase A and proteinase K (6). Afterextraction with phenol/chloroform, the DNA was precipitatedwith ethanol.

Preparative Agarose Gel Electrophoresis. EcoRI-digestedhuman placental DNA (17.5 mg) was fractionated on a con-tinuous elution, horizontal, 0.8% agarose gel, essentially as de-scribed by Polsky et al. (6) except that Seakem HGT (P) agarosewas used because this appeared to contain fewer contaminantsthat inhibit subsequent enzymatic steps. The DNA was collectedin approximately 80 fractions, and aliquots of these were as-sayed by analytical agarose gel electrophoresis followed by filterhybridization to a radioactive hGH cDNA probe.

Filter Hybridization. The filter hybridization technique ofSouthern (7) was used with the following modifications. Filters[Schleicher & Schuell, 0.45 jim (BA85) nitrocellulose; ap-proximately 15 X 15 cm] were preannealed, in sealed plasticbags, at 420C in 10 ml 50% formamide/50 mM Hepes, pH7.0/5-fold concentrated Denhardt's solution (8, 9)/0.45 MNaCl/0.045 M sodium citrate, pH 7 (3-fold concentratedNaCl/Cit) containing 18 ,jg of sheared and denatured Micro-coccus lysodekticus DNA and 40 jig of yeast tRNA per ml.Heat-denatured hybridization probe (approximately 0.5-1 X106 cpm) was added in 0.5 ml of the same solution and the filterswere hybridized at 420C for 1-2 days. The filters were washedat 500C for 1-2 hr with two or three changes (about 200-500ml per filter) of 0.1-strength NaCl/Cit and 0.1% sodium do-decyl sulfate (10) and then briefly with 0.1-strength NaCl/Citat room temperature and were exposed at -70°C to X-OmatR films and Du Pont Cronex Lightning Plus intensifyingscreens.

Hybridization Probe. Approximately 200 ng of polyacryl-amide gel-purified 550-base-pair hGH cDNA fragment wasincubated (room temperature for 60-90 min) in 20 ,ul of a re-action mixture containing 10 mM Tris.HCl at pH 7.5, 10mMMgCl2, 1 mM dithiothreitol, 2.5 mg of calf thymus DNAoligonucleotides [prepared by DNase I digestion (11)] per ml,dATP, dGTP, and dTTP at 10 ,jM each, 0.5 jiM [a-32P]dCTP[2000-3000 Ci/mmol (1 Ci = 3.7 X 1010 becquerels); NEN orAmersham], and 1 unit of DNA polymerase I (Klenow modi-fication, Boehringer Mannheim). The cDNA fragment hadpreviously been heat denatured in the presence of the calfthymus oligonucleotides. Routinely, specific activities in theorder of 2 X 108 cpm/,jg were obtained. The unincorporated

Abbreviations: hGH, human growth hormone (somatotropin); hCS,human chorionic somatomammotropin (choriomammotropin);NaCl/Cit, 0.15 M NaCl/0.015 M sodium citrate, pH 7.0; kb, kilo-base(s).t Present address: Genentech Inc., San Francisco, CA 94080.

4294

The publication costs of this article were defrayed in part by pagecharge payment. This article must therefore be hereby marked "ad-vertisement" in accordance with 18 U. S. C. §1734 solely to indicatethis fact.

Dow

nloa

ded

by g

uest

on

June

19,

202

1

Proc. Natl. Acad. Sci. USA 76 (1979) 4295

1 2 3 4 5 6

-20.8

-72

5.7-53

3.2

FIG. 1. Autoradiograph of a filter hybridization between re-striction endonuclease-digested human placental DNA (fractionatedon a 1% agarose gel) and 32P-labeled chGH 550-base-pair DNA. Theenzymes used were: lane 1, EcoRI; lane 2, Pvu II; lane 3, BamHI; lane4, Pst I; lane 5, Sst I; lane 6, Xba I. The positions and sizes of thefragments produced by EcoRI digestion of wild-type bacteriophageX DNA are shown at the right. The two prominent EcoRI fragmentsof human DNA are 2.6 and 2.8 kilobases (kb) long; the weakly hy-bridizing species is 9.5 kb.

3 4 5

[a-32P]dCTP was then removed by gel filtration on SephadexG-50.

Bacteriophage X Packaging. The bacteriophage X in vitropackaging system was used exactly as described by Blattner etal. (5). The EK2 vector Xgt WES.AB (4) was used after eitherseparation of the B fragment from the vector arms on a 5-20%sucrose gradient (0.1 M NaCl/10 mM Tris-HCl, pH 7.5/1 mMEDTA in an SW 40 rotor at 35,000 rpm for 4 hr) or digestionof the DNA with Sst I, for which there are two recognition sitesin the B fragment but none in the vector arms.

In each packaging experiment, 5.5 gg of vector DNA wasligated to 400 ng of donor DNA in a total volume of 5.5 Atl.Packaging and subsequent cloning were performed in a P3physical containment facility in accordance with the NationalInstitutes of Health Guidelines for Recombinant DNA Re-search. All packaging components were tested for packagingof endogenous DNA as well as for am+ phage in control XgtWES.AB DNA packaging experiments. In all cases the ratioswere

4296 Biochemistry: Fiddes et al.

A B

1 2 3 4 5 6 7 8 910C

U._^., _~4

_~~~~~~~~~~~O s

.

FIG. 3. Autoradiographs of filter hybridizations using the chGH550 probe with the 2.6-kb hGH fragment (A) and the 2.8-kb hCSfragment (B). The digests were fractionated on two separate 1%agarose gels. Lanes 1-5 were obtained with the hGH-Xgt WES re-combinant digested with: lane 1, EcoRI; 2, EcoRI and Xba I; 3, EcoRIand Bgl II; 4, EcoRI and Pst I; 5, EcoRI and Pvu II. Lanes 6-10 wereobtained with the hCS-Xgt WES recombinant digested with: lane 6,EcoRI; 7, EcoRI and Pst I; 8, EcoRI and Pvu II; 9, EcoRI and Bgl II,10, EcoRI and Xba I.

a unique Xba I site. The fact that the 2.6-kb fragment containeda Bgl II but no Xba I site whereas the 2.8-kb fragment had anXba I but no Bgl II site supports the identification of the frag-ments given above.

Isolation and Identification of Recombinant DNA Con-taining hGH and hCS Gene Sequences. In vitro packagingsystems in which bacteriophage X DNA is ligated to exogenousDNA and used to generate mature infective phage particleshave been described (5, 17-19). The DNA packaged into phagecan be used to infect E. coli with a considerably higher effi-ciency than is obtained by transfection with the DNA itself. Inthe experiments described here, the in vitro packaging systemof Blattner et al. (5) was used.

In several packaging experiments, a total of approximately106 recombinant phages were obtained by using bacteriophageXgt WES-XB DNA as the vector and human DNA in the sizerange 2-3 kb. Single isolates of the hGH 2.6-kb fragment andof the hCS 2.8-kb fragment were obtained.To confirm the identity of the cloned human DNA frag-

ments, their restriction endonuclease digestion pattern was

Pvu EjtI

compared with that of total human DNA (Figs. 2 and 3A andB). The hGH recombinant contained a 2.6-kb fragment thathybridized to the chGH 550 probe and had a restriction enzymepattern for the enzymes EcoRI, Pvu II, Bgl II, Pst I, and XbaI (Fig. 3A) that was identical to that observed in total humanDNA (Fig. 2). Similarly, the digestion pattern of the 2.8-kb hCSfragment (Fig. 3B) corresponded to that of total human DNA.This restriction enzyme data and the specific hybridizationobserved confirms the identities of the recombinants.

In the case of the hGH isolate, digestion of plaque-purifiedrecombinant phage DNA with EcoRI revealed that three EcoRIfragments of human DNA had been cloned in the same phage,presumably as the result of multiple ligation of the humanEcoRI DNA fragments. Two of these fragments were 2.6 kband one was less than 2 kb. The 2.6-kb fragment that hybridizedto the chGH 550 probe was transferred to the EcoRI site of theplasmid vector pBR322 by using X 1776 as the host strain. Thisrecombinant plasmid has been used for most of the subsequentrestriction analysis and DNA sequencing of the hGH gene andis designated pBR322-ghGH(2.6).The hGH Gene Contains at Least Two Intervening Se-

quences. A restriction enzyme cleavage map of the hGH genehas been determined for the enzymes EcoRI, BamHI, Pst I,Pvu II, Sma I, and Bgl II (Fig. 4). Certain features of this maplead to the conclusion that the hGH coding sequence is inter-rupted by at least two intervening sequences.The locations of the intervening sequences have been con-

firmed by DNA sequence analysis. The nucleotide sequenceof the junction between intervening sequence B and the codingregion at approximately nucleotide 1200 was obtained by thechain terminator method (14) using as primer a restrictionfragment corresponding to amino acid residues 45-56 of thehGH protein. This was obtained by Pst I and HinfI digestionof the chGH 550 recombinant. Template DNA was preparedfrom the pBR322-ghGH(2.6) recombinant by first digestingit with HindIII to produce a linear molecule and then con-verting it to single-stranded DNA with E. coli exonuclease III(15). An autoradiograph of a sequencing gel is shown in Fig.5 and the sequence of the junction between the coding andintervening sequence is shown in Fig. 6.

In addition to locating the boundary between the codingsequence and intervening sequence B, this confirms the identityof the 2.6-kb fragment as hGH and not hCS because the se-quence encompasses a region where the two hormones differin amino acid sequence.The sequence of the junction between the coding sequence

and intervening sequence C at approximately nucleotide 1300

Elg n RyullSmaI

L A--DLAB _ A

EcoRI BamHI RsgI

L0 500

I1000

\\. T/) t fpoly A siteS.mgIttlBaramHI\\ 3' end coding region

550 H-wI cDNA

1500 2000base pairs

FIG. 4. Restriction enzyme map of the 2.6-kb EcoRI fragment containing the hGH gene sequences. Those sites written above the line arepresent in the cDNA; those below the line are unique to the gene fragment. The region containing sequences homologous to the 550-base-paircDNA fragment used to probe the gene is indicated. This extends from amino acid 24 to 48 bases into 3' untranslated region. The 3' end of thecoding region is marked. The poly(A) addition site is located 109 bases from the 3' end of the coding region. Coding sequences are indicatedby shading. The intervening sequences are labeled A, B, and C. The precise boundary at the 3' end of region A is not clear (see text).

Ecg RI

2500

Proc. Natl. Acad. Sci. USA 76 (1979)

Dow

nloa

ded

by g

uest

on

June

19,

202

1

Proc. Natl. Acad. Sci. USA 76 (1979) 4297

A G C T(+) _(+)A T G CG C _ A TT A G2 C2A2T2 A TC G TAA2 T2 C GT A OG..G3 C3 T2A2A T C GG C T2 A2C G C GT A G2 C2G 2C A TT A \ T AC G t IA TT A T AC2 G2 A TA T -G CG C / § IG C~A TV C G_T A T2 A2A T C2 G2G C I T AA T OGC G2 G CT2 A2 T AG2 C2 C GT A T2 A2G3 C3 C GC G A TG2 C T AT A A2 T2C2 G2 G CT2 A2 Ua TA

FIG. 5. Autoradiograph of an 8% polyacrylamide DNA se-quencing gel used in the chain termination method with 2'3'-dideoxyATP, 2'3'-dideoxy CTP, 2'3'-dideoxy GTP, and 2'3'-dideoxy TTP aschain-terminating analogues. The primer was a 29-nucleotide-longfragment obtained by digestion of the cloned 550-base-pair cDNAfragment of hGH with Pst I and Hinfl. The template was preparedby treating HindIII-cut pBR322-ghGH(2.6) with E. coli exonucleaseIII. The sequence read from the gel is written adjacent to the auto-radiograph with the complementary sense strand (+) written along-side that. The arrows indicate the two possible splice points, and theasterisk marks the codon that differs from the cDNA.

(Fig. 4) has been obtained by a similar strategy (data not shown),thus confirming the existence of a second intervening sequencein the hGH gene.

An 800-base-pair "full-length" cDNA clone of hGH (chGH800) has been obtained recently (20). This contains, in additionto the entire coding sequence, 29 nucleotides of the 5' untran-slated region and all 109 nucleotides of the 3' untranslated re-gion preceding the poly(A) sequence. When this 800base-pairfragment is used as a hybridization probe with pBR322-ghGH(2.6), the PstI/Pvu II fragment between positions 600and 1000 hybridizes strongly. T6is fragment does not hybridizeto the chGH 550 probe, thus indicating that a large portion ofthe 5' end of the hGH gene is contained in this region (Fig. 4).A very low level of hybridization of the chGH 800 probe to the5'-terminal EcoRI/Pst I fragment has also been detected, in-dicating the possibility of a third intervening sequence (A) inthe hGH gene. This will require further DNA sequence analysisfor confirmation.DNA sequence analysis has also established that the 2.6-kb

hGH fragment contains the poly(A) addition site (Fig. 4).

DISCUSSIONEcoRI restriction endonuclease fragments 2.6 and 2.8 kb longcontaining coding sequences for the hGH gene and the hCSgene, respectively, have been identified by hybridization to a550-base-pair cloned Hae III fragment of hGH cDNA. In ad-dition to these two fragments, a third fragment of about 9.5 kb,that hybridizes weakly to the chGH 550 probe, is also detectedroutinely (see Figs. 1 and 2). The origin of this is not clear be-cause the other known sequence related to hGH and hCS is thepituitary hormone prolactin which only has 16% amino acidsequence homology with hGH and 13% homology with hCS(21). The hybridization conditions used would not detect pro-lactin genes (assuming that the nucleotide sequences had similarlevels of homology).

Both the 2.6-kb hGH and the 2.8-kb hCS fragments havebeen isolated by cloning in the bacteriophage Xgt WES-XBvector with the in vitro packaging approach. A restriction en-zyme cleavage map (Fig. 4) derived for the hGH gene indicatesthat, in common with other eukaryotic genes (22-26), the hGHgene contains intervening sequences that interrupt the codingsequences.One of the two identified intervening sequences (sequence

B) is about 220 base pairs long; the other (sequence C) is about300 base pairs. A third intervening sequence (sequence A) veryclose to the 5' end of the gene may also exist. The possibility thatother very small intervening sequences are present-i.e., thatsome of the intervening sequences are interspersed within shortcoding regions-has not yet been excluded.

Preliminary DNA sequence analysis of the pBR322-ghGH(2.6) recombinant has confirmed the existence of theintervening regions shown in Fig. 4. The sequence at the 3' endof intervening sequence B (at approximately position 1200) isshown in Figs. 5 and 6. As in other systems, the 3' end is markedby an A-G doublet (27).A detailed restriction enzyme map has not yet been deter-

mined for the hCS gene but the preliminary evidence suggeststhat the pattern of intervening sequences may be similar to thatof the hGH gene. Total human placental DNA, when digestedwith both EcoRI and Pvu II (Fig. 2), contains a single 980-

Intervening Sequence: C CTT GGT GGG CGG TCC TTC TCC TAG~~~~~~~~~~~Coding Sequence: . GAG TTT GAA GAA GCC TAT ATC CTG AAG GAG CAG AAG TAT TCA * -

Amino Acid: . GLU PHE GLU GLU ALA TYR ILE LEU LYS GLU GLN LYS TYR SER30 9 35 40

FIG. 6. Partial sequence of an intervening region in the hGH gene. The junction between the 3' end of the intervening sequence B and thecoding region is shown. The codon marked with an asterisk is CCA (proline) in the cDNA. The amino acids are numbered. The coding sequenceto the 5' side of the intervening sequence is taken from the cDNA. The splice point could be as shown or one nucleotide to the right.

Biochemistry: Fiddes et al.

Dow

nloa

ded

by g

uest

on

June

19,

202

1

Proc. Natl. Acad. Sci. USA 76 (1979)

base-pair fragment that hybridizes to the chGH 550 probe. Thisprobe also detects hCS sequences. Because this fragment con-tains intervening and coding sequences in the hGH gene, theappearance of6a single band implies that the hCS gene may alsocontain a similar structure in this region.

Several nucleotide sequence differences have been detectedbetween the ghGH(2.6) gene and the cloned cDNA fragment.One of these is shown in Fig. 6: a CCA proline codon at aminoacid position 37 in the cDNA is changed to a CTG leucine codonin the gene. The cDNA proline codon corresponds to the aminoacid sequence established for the hGH protein (28). This codonchange is in a region where tryptic mapping data of the hGHprotein indicates that amino acid sequence variants occur (29).The ghGH(2.6) fragment may therefore correspond to an allelicvariant of growth hormone. Alternatively, the differences may-be the result of the different sources of material used: the hGHcDNA came from a pool of pituitary tumors and the hGH genecame from a single placenta.The amino acid data in Fig. 6 exclude the possibility that a

growth hormone variant of molecular weight 20,000 (29), ratherthan the normal 22,000, has been cloned. This variant, whichexists as 5-10% of the total growth hormone, lacks the trypticpeptide located between amino acids 38 and 41. The DNA se-quence analysis presented in Fig. 6 shows that the coding regioncorresponding to this peptide is present in ghGH(2.6). Nucle-otide sequence differences between the cDNA and ghGH(2.6)have also been detected in the 3' untranslated region (unpub-lished data).An independently isolated 2.6-kb fragment has been recently

cloned (unpublished data). This fragment gives rise to anidentical hybridization pattern as the first isolate when theenzymes Pvu II, Pst I, Bgl II, and Sma I arepused. However, adifference between the two is noticed with the enzyme Hae III.The possibility therefore exists that this second isolate corre-sponds to the expressed growth hormone gene. DNA sequenceanalysis of this isolate will be required in order to determinewhether this is the case.

Note Added in Proof. The existence of the third intervening sequence(A) has been confirmed by DNA sequence analysis. The splice pointis located in the second codin of the pre-sequence of hGH.

J.C.F. is the recipient of a postdoctoral fellowship from the AnnaFuller Fund. R.A.H. is the recipient of a postdoctoral fellowship fromthe British Science Research Council. This work was supported byGrants CA14026 to H.M.G. and AM19997 to H.M.G. and J.D.B.;H.M.G. and J.D.B. are Investigators of the Howard Hughes MedicalInstitute.

1. Niall, H. D., Hogan, M. L., Sauer, R., Rosenblum, I. Y. &Greenwood, F. C. (1971) Proc. Natl. Acad. Sci. USA 68, 866-869.

2. Bewley, T. A., Dixon, J. S. & Li, C. H. (1972) Int. J. Pept. ProteinRes. 4, 281-287.

3. Shine, J., Seeburg, P. H., Martial, J. A., Baxter, J. D. & Goodman,H. M. (1977) Nature (London) 270,494-499.

4. Leder, P., Tiemeier, D. & Enquist, L. (1977) Science 196,175-177.

5. Blattner, F. R., Blechl, A. E., Denniston-Thompson, K., Faber,H. E., Richards, J. E., Slightom, J. L., Tucker, P. W. & Smithies,0. (1978) Science 202, 1279-1284.

6. Polsky, F., Edgel, M. H., Seidman, J. G. & Leder, P. (1978) Anal.Biochem. 87,397-410.

7. Southern, E. M. (1975) J. Mol. Biol. 98, 503-517.8. Denhardt, D. T. (1966) Biochem. Biophys. Res. Commun. 23,

641-646.9. Jeffreys, A. J. & Flavell, R. A. (1977) Cell 12, 429-439.

10. Tilghman, S. M., Tiemeier, D. C., Polsky, F., Edgell, M. H.,Seidman, J. G., Leder, A., Enquist, L. W., Norman, B. & Leder,P. (1977) Proc. Natl. Acad. Sci. USA 74, 4406-4410.

11. Taylor, J. M., Illemensee, R. & Summers, S. (1976) Biochim.Biophys. Acta 442,324-330.

12. Grunstein, M. & Hogness, D. C. (1975) Proc. Natl. Acad. Sci. USA72,3961-3965.

13. Benton, W. D. & Davis, R. W. (1977) Science 196, 180-182.14. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad.

Sci. USA 74,5463-5467.15. Smith, A. J. H. (1979) Nucleic Acids Res. 6, 831-848.16. Sanger, F. & Coulson, A. R. (1978) FEBS Lett. 87, 107-110.17. Becker, A. & Gold, G. M. (1975) Proc. Natl. Acad. Sci. USA 72,

581-585.18. Hohn, B. & Murray, K. (1977) Proc. Natl. Acad. Sci. USA 74,

3259-3263.19. Sternberg, N., Tiemeier, D. & Enquist, L. (1977) Gene 1,

255-280.20. Martial, J. A., Hallewell, R. A., Baxter, J. D. & Goodman, H. M.

(1979) Science, in press.21. Shome, B. & Parlow, A. F. (1977) J. Clin. Endocrinol. Metab.

45, 1112-1115.22. Weinstock, R., Sweet, R., Weiss, M., Cedar, H. & Axel, R. (1978)

Proc. Natl. Acad. Sci. USA 75, 1299-1303.23. Lai, E. C., Woo, S. L. C., Dugaiczyk, A., Catterall, J. F. &

O'Malley, B. W. (1978) Proc. Natl. Acad. Sci. USA 75, 2205-2209.

24. Jeffreys, A. J. & Flavell, R. A. (1977) Cell 12, 1097-1108.25. Breathnach, R., Mandel, J. L. & Chambon, P. (1977) Nature

(London) 270, 314-318.26. Tilghman, S. M., Teiemeier, D. C., Seidman, J. C., Peterlin, B.

M., Sullivan, M., Maizel, J. V. & Leder, P. (1978) Proc. Natl. Acad.Sci. USA 75, 725-729.

27. Catterall, J. F., O'Malley, B. W., Robertson, M. A., Staden, R.,Tanaka, Y. & Brownlee, G. G. (1978) Nature (London) 275,510-513.

28. Li, C. H., Dixon, J. S. & Liu, W. K. (1969) Arch. Biochem. Bio-phys. 133, 70-91.

29. Lewis, U. J., Dunn, J. T., Bonewald, L. F., Seavey, B. K. & VanderLaun, W. P. (1978) J. Biol. Chem. 253, 2679-2687.

4298 Biochemistry: Fiddes et al.

Dow

nloa

ded

by g

uest

on

June

19,

202

1