Biology Reference
In-Depth Information
a
Drosha (
Homo sapiens
)
RNase III
1
dsRBD
1
2
1374
390
1374
His
6
-Drosha
390-1374
b
DGCR8 (
Homo sapiens
)
HBD
dsRBDs
12
CTT
NLS
1
773
WW
DD
751
751
NC1
NC9
276
498
Fig. 1
Domain structures and recombinant expression constructs of human
Drosha and DGCR8. The heme-binding domain (HBD) of DGCR8 includes a dimer-
ization (sub)domain (DD)
shown to be important for co-immunoprecipitation of DGCR8
with Drosha in human cells [
10
] and for formation of proper
higher order structure of DGCR8 upon binding pri-miRNAs [
22
].
See a recent review for more background information [
23
].
pri-miRNA processing may be analyzed in vitro using either
whole-cell or nuclear extracts [
24
,
25
], affinity-purified Micro-
processor complexes expressed in mammalian cells [
19
,
26
-
28
], or
recombinant Drosha and DGCR8 expressed in heterologous sys-
tems [
3
,
10
]. We focus on the last method in this review. Insect cell
and bacterial expression systems allow active Drosha and DGCR8
proteins to be expressed with high yield and be purified to near
homogeneity. These highly purified proteins do not contain other
human proteins typically found to associate with Drosha and
DGCR8 and enable the investigation of pri-miRNA processing
with greater control of the experimental conditions.
Gregory, Shiekhattar, and colleagues were the first to show
that Microprocessor may be reconstituted by expressing Drosha in
insect cells and DGCR8 in
E. coli
[
3
]. The N-terminal 275 amino
acids [
10
,
20
] and the C-terminal 22/23 residues [
4
,
20
] of the
773-residue DGCR8 have been shown to be dispensable for in
vitro pri-miRNA processing. A DGCR8 construct with these resi-
dues deleted (named NC1, Fig.
1b
) is the most active form in pri-
miRNA processing in vitro, whereas a further truncation called
NC9 (Fig.
1b
), containing only the two dsRBDs and the CTT, is
less active than NC1 [
6
]. The procedures for expression and puri-
fication of NC1 are described in this review.
In addition to being an obligate partner of Drosha, DGCR8 also
binds heme [
4
]. The central region of DGCR8, including the
WW motif, encodes a unique dimeric heme-binding domain [
29
].
Each DGCR8 dimer binds one heme molecule [
4
]. The heme in
native DGCR8, expressed in
E. coli
, is in the Fe(III) redox state [
5
].
1.2 DGCR8 as a
Heme Protein
