G‘e/w, 103 (1991) 263-268 rci 1991 Elsevier Science Publishers
GENE
B.V. All rights reserved.
263
0378-I 119:91/$03.50
0502 I
Structure and expression of the human gene encoding testicular HI histone (Hlt) (Chromatin;
spermatogenesis;
genomic
library;
Northern
blots;
Sl nuclease
assay)
Birgit Drabent, Efterpi Kardalinou and Detlef Doenecke Institut fir Biochenzie, Abteilung Molekulurbiologie, Received by H.G. Zachau: 21 December Revised: 25 February 1991 Accepted: 12 April 1991
Georg-August-Universitiit,
D-3400
Giittingen (Germart~)
1990
SUMMARY
The gene coding for the human H 1t histone, a testis-specific H 1 subtype, was isolated from a genomic library using a human somatic HI gene as a hybridization probe. The corresponding mRNA is not polyadenylated and encodes a 206-amino-acid protein. Sequence analysis and S 1 nuclease mapping of the human Hlt gene reveals that the 5’ flanking region contains several consensus promoter elements, as described for somatic, i.e., S-phase-dependent HI subtype genes. The 3’ region includes the stem-and-loop structure necessary for mRNA processing of most histone mRNAs. Northern blot analysis with RNAs from different human tissues and cell lines revealed that only testicular RNA hybridized with this gene probe.
INTRODUCTION
Histone Hl forms the most divergent class within the highly conserved histone family (Cole, 1987). Various Hl subtype patterns may modulate the cooperative interactions between H 1 histones and result in an altered local chromatin structure (Weintraub, 1984). An extensive change of the HI histone pattern occurs during mammalian spermatogenesis. A testis-specific variant, Hlt (Seyedin et al.. 1981) is first detected at meiosis in pachytcne spermatocytes, when it replaces other H 1 subtypes and represents the major part of the total H 1 histonc complement. Hlt then persists until the stage of round spcrmatids and is lost when nuclear condensation occurs (Meistrich et al., 1985).
Dr. D. Docnecke, Institut 23, D-3400 Giittingcn (Germany)
Corrt~s~Jorlclen~e 10 :
Humboldtallec
Tel. (49)551:395972; Abbrcviatlons: HI histone;
fiir
Biochemie,
Fax(49)551:‘395960.
aa, nmino acid(s); bp, base pair(s): HI/. gem encoding
nuclcotide(s): ORF, open reading /.s/>,transcription start point.
Hlt; frame;
Hlt, human testicular
kb, kilobase SDS, sodium
or 1000 bp; nt, dodecyl
sulfate;
Cole et al. (1986) and Grimes et al. (1990) have isolated and sequenced the rat HI t gene. Its scquencc differed substantially from somatic Hl subtype genes, which may reflect a specific function of H It in the organization of chromatin during meiosis and during the later steps of chromatin condensation until histone replacement in late spcrmatids (Meistrich et al., 1985). The meiotic expression of the Hit gene (Grimes et al.. 1990) may depend on different factors than the S-phaserelated synthesis of somatic H I histones and the cell-cycleindependent synthesis of the subtypes H5 and H I ‘. Thus. the synthesis of H 1~ may depend on a third type of histonc gene regulation. A detailed knowledge of Hlt genes and flanking sequences is a prerequisite for an analysis of this varied mode of histone gent expression.
EXPERIMENTAL
AND
DISCUSSION
(a) Isolation and characterization of a human Hit gene A human genomic DNA library in /IEMBL.? bacteriophage was screened under low stringency conditions using a hybridization probe from the human Hf.2 gene (Eick et al., 1989). Two overlapping clones were obtained. One
264 co\wcd the coding portion plus 750 nt upstream and ;~bout 14 kb downstream of the HI t gcno. whcrc:~s the second
clont: axered just part of the coding region and rcachcd further downstream. of the respccti\rc fragments Mapping and sequencing from both clones yielded identical results.
Fig. I indicates
the sccluencing stratcg! and Fig. 3 presents the nt scqucncc of the coding and flanking portions of the //I gcnc and the derived aa scquencc. It show.s an ORF
encoding 206 aa.
which is in the rmgc of most H I histones. Tho nt scquencc: of the coding region shows 67”,, homolog! to the rat Hlr pie
et 31.. 1986) but .iual %I”,, compared with the HI I and 5X”,, \vilh the /II..? gene (the initial
(Cole
human
xmxning
probe).
The 5’ flanking
sequence
contains
consensus
promotes
elements including the TATA-box
( -30) and ;I CC’:2:2Tbox (-52. rclativc to the putatiw cap site. + I ). T\vo clcrnents were found. \\hich arc characteristic for most ccllqclc-dcpendcnt HI gcncs. a Spl-binding site ;II -78 and the ‘t! I bou’ at position -10.1 (Cslcs and \Vella. I S-45). The 3’ untranslated region contains the histonc gencspecific dqad symnlctry elcmcnt. which is chxrctcristic fm cull-c!,clc-dependent histonc gcncs (Birnsticl et al., 1085) but is missing in the solitary tfl ’ (Kress ct al., I 986) or H5 (Kricg ct al.. 1982; Doenccke and Tiinjcs. 198-I) gents. This clancnt is follomxd b\ ACCC/\ and a purinc-rich motif, \vhich is located do\\nstrcam of the mRNA site (Kirnsticl the ORF
3’ processing
ct al., 1% 1). Thus, the scqucnce \uggcsts that
and its flanking parts rcprcsent ;I
bearing most of the regulatory clomenls
I‘~~nc~icm~I
gene
of S-phasc-dcpen-
dent ifI
gales. An alignment of the deduced aa sequence and the knomm scclucnccs of rat H It (Cole ct al., lW6). boar H 1t (Colt
ct al.. 1981) and [he Herman If I. 1 and H I .7 histonch ct al..
19X9).
(kick
is shown
in Fig. 3. It demonstrates that ~hc sequence 01‘ the putati\,c h~mian I I It is strongl) related to the rat and boar H It proteins (62 and OO”,,. rcspectivcly). but differs from the IILII~XI somatic H I subtypes (Table I ). SC\cr:ll XI. \I hich had been mtcrprctcd b> C‘olc ct al. ( 19X6) ax ff It-specific at certain posilionj. wcrc again found in ~LIIII;I~I H It, I\ hcrcas others acre replaced. Dcspitc thcsc diffcrcnccs. the newly detcctcd HI pcnc \V;IS intcrprctcd .I\ the h~~m:m !-llr
subt!pc gene.
I‘hi~ conclusion \I as supportal bl ;I compariwn of the noncoding regions ofthis gent \\ ilh the rat 1111gcnc and the ~LIII~~II Ifl.1 and H/.2 gcncs (Fig. -la). ‘1‘1~~ 5’ flanking regions of all li)ur gcncs share xc\cral promoter clcmcnts. Ket\vccn these motifs. human and rat 111, genes shop par-
tial similarities,
whercaj the III.
I :md HI.2
genes differ in
this rapcct. T\\o wc~ucnc~clc‘nicnt\ \\liicli
nwe
suggcstcd
(Grimes
26.5
SET”PAASA~AGLAAMEKL~TKKRGilKPA~LISASRKtP~LS”SKLITEA50 :::A::::STL”P:P”::PA::R::K::G--MATA::PRGF::::::P:: A::A:: : : :
:
:P:DS”P:S”::P:A::::K::“::TGT:::A:SA::::::::: :
: :P:P:ASA-:PEKP:AG: A:::P:A:--PPA::A:“::KAA:K::--GTP::ASGPP::E:::K:
:
:AKKl’AKAAAASKK:PAGP:::E::“Q:
LS”SQER”GMSL”ALKKALAAAGYDVEKNNSRIKLSLKSL”NKGlL”QTR’oo ::M::::A::::A::::::::::::::::::::::A::H:::: : : :A: : : :A: : : : : : A:S:K:
:G:“::A::::::::::::::::::::::GI::::S.:T:::.K
“AA:K:
:S:“::A::::::::::::::::::::::G:::::S::T::::K
:
:
:
G T : :
G A : :
S :
G S :
:
F :
K :
I. :
S :
T KKVI PKSTRSKAKKS : :AASGNDKG:G:::A:::A::-:G:::A:R-----:::
:
:
:
:
:
:
:
:
:
:
:
:
::::::::::N: ::::::::::N:
:
:
:
:
:
:
:
:
:
:
:
:
:G:
”
S
A
i
T
K
K
-
L
”
: : : :I;:
I.
S
R
6
S
K
-
:“:::.K : :
:
:
:
-
-
-
S
-
:
I’
K=’
:AA:EPRKG:“::PAA:::::-::::::::-----::: :ASS”E:KPG:S: :AASGEAKP:“::AGGT:P::P”GAAKKF:-----KAA
-:AT:::A-TGA:KKL:KATGA:K:
TAKTNKR-AKKPRATTP-KT”RSGRKAKGAKGKQQQK--SP-----”KAR~~~ ss::-:“-v:: :K::PT-:GSG:R::T:::::L::R:--::-----A::: S.,A:::-.: :S:T.AAO:AA:::::T:E::“:::R:--::-~~~-A::’
iv:
:P:K-::::-:A:R-:SSKNPK:P:T”:F:KVA:--::AKAKA::PK
GGA:F:KS:::TPKKAK-:PAAAT”TK:“::SPKKA:”AK:KKAAKSA.K
Fig.
3. Comparison
et 81. (1984; necessary
ASKSKLTQHHEV----N”RKATSKK
Hit
:TN:NSGKSKM:MQKTDL:::AGR: :A:P:AGNPKLTQQKT:P::::NR: :A:AR”:KPKTA----KPK::AP::K :“:P:AAKPK”:----KPK::AP::K
Hlt lilt HI. 1 111.2
of histone
1986), human
for alignment
.4rrowheads
H It aa sequcncca
of sequences
mark borders
of human
H 1.1 and HI.2 sequences of protein
rat and boar with human
from Eick et al. (1989). Colons
(see also Table 1). Numbering domains.
Computer
refers to general
alignment
0.06) (207) (211) (214) (212)
HI.1 and HI.2 data. The rat and boar data are taken from Colt indicate
identical
frame,
numbers
\vas done with homology
aa (as in first lint), and hyphens in parentheses
search
programme
indicate Gencprc
actual
describe
sequence
gaps
lengths.
3.20.
from different human tissues and cell lines was analyzed electrophoretically, blotted and hybridized with probes derived from the putative Hit gene and the S-phase-dcpendent human HZ.2 gene (Fig. 5). The autoradiogram shows that the putative Hlt probe hybridized with just one mRNA species in human testicular RNA. The Hlt mRNA is not polyadenylated. This varies from the testicular expression of somatic, replication-dependent core histone genes, bvhich are encoded by poly(A)+ mRNA due to an alternative pathway of mRNA 3’ end formation during spermatogenesis (Challoner et al., 1989). No hl,bridization was detcctablc in human spleen RNA or in RNA isolated from lcukcmia. hepatoma and cervix carcinoma cell lines. Even rat or bull testicular RNA failed
et al., 1990) as regulatory sites for testicular expression (5’-CTAGGGAT between Spl site and CAAT; oligo(dT) bctwccn cap site and ATG) are partly preserved in the hullian scquencc. The first motif reads 5’-CTAGGTGAT, whereas the (dT),,, dccanucleotide is essentially missing. All four HI genes contain the consensus histone mRNA processing signals (Birnstiel et al., 1985) at their 3’ ends, are again more pronounced but sequence similarities between the Hlt gene sequences compared \vith other HI gene spccics (Fig. 4b). (h) Identification of the human Hlr gene transcript Northcrn blot hybridizations were done to prove the testicular expression of the putative Hlr gene. Total RNA
TABLE
human rat boar human human
I
Comparison
of human
Hlt an sequence
\\ith other testicular
Identical,’
Mismatched
(a)
(b)
and somatic h
histones
Rat
Hit
131
79
14
Hlt H I. 1 HI .2
147 105 I02
64 112 113
5 I6 II
.’ Identical
aa, when compared aa. including
with human
Gaps