DNase II: Genes, enzymes and function

Abstract

Deoxyribonuclease (DNase) II, which was discovered more than 50 years ago, is a mammalian endonuclease that functions optimally at acid pH in the absence of divalent cations. Its lysosomal localization and ubiquitous tissue distribution suggested that this enzyme played a role in the degradation of exogenous DNA encountered by phagocytosis, although the relative importance of such a role was unknown. Subsequent investigations also suggested that DNase II was important for DNA fragmentation and degradation during cell death. Within the last few years, our work and that of others has lead to the cloning of various mammalian DNase II genes as well as the identification and characterization of highly homologous genes in the invertebrates Caenorhabditis elegans and Drosophila melanogaster. Interestingly, studies of the C. elegans DNase II homolog NUC-1 were the first to suggest that DNase II enzymes were fundamentally important in engulfment-mediated DNA degradation, particularly that associated with programmed cell death, due to the presence of persistent apoptotic-cell nuclei within phagocytic cells in nuc-1 mutants. Similarly, mutation of the Drosophila DNase II-like gene was found to result in the accumulation of low-molecular-weight DNA throughout the animals. Homozygous mutation (knockout) of the DNase II gene in mice revealed a much more complex and extensive phenotype including perinatal lethality. The lethality of DNase II-knockout mice is likely the result of multiple developmental defects, the most obvious being a loss of definitive erythropoiesis. Closer examination revealed that a defect in engulfment-mediated DNA degradation is the primary defect in DNase II-null mice. In this review, we have compiled information from studies on DNase II from various organisms to provide a consensus model for the role of DNase II enzymes in DNA degradation.

Full text

Review
DNase II: genes, enzymes and function
Cory J. Evans
a
, Renato J. Aguilera
b,
*
a
Department of Molecular, Cell and Developmental Biology, University of California, Los Angeles, CA 90095, USA
b
Department of Biological Sciences, University of Texas at El Paso, El Paso, TX 79968-0519, USA
Received 30 July 2003; received in revised form 8 August 2003; accepted 26 August 2003
Received by A.J. van Wijnen
Abstract
Deoxyribonuclease (DNase) II, which was discovered more than 50 years ago, is a mammalian endonuclease that functions optimally at
acid pH in the absence of divalent cations. Its lysosomal localization and ubiquitous tissue distribution suggested that this enzyme played a role
in the degradation of exogenous DNA encountered by phagocytosis, although the relative importance of such a role was unknown. Subsequent
investigations also suggested that DNase II was important for DNA fragmentation and degradation during cell death. Within the last few years,
our work and that of others has lead to the cloning of various mammalian DNase II genes as well as the identification and characterization of
highly homologous genes in the invertebrates Caenorhabditis elegans and Drosophila melanogaster. Interestingly, studies of the C. elegans
DNase II homolog NUC-1 were the first to suggest that DNase II enzymes were fundamentally important in engulfment-mediated DNA
degradation, particularly that associated with programmed cell death, due to the presence of persistent apoptotic-cell nuclei within phagocytic
cells in nuc-1 mutants. Similarly, mutation of the Drosophila DNase II-like gene was found to result in the accumulation of low-molecular-
weight DNA throughout the animals. Homozygous mutation (knockout) of the DNase II gene in mice revealed a much more complex and
extensive phenotype including perinatal lethality. The lethality of DNase II-knockout mice is likely the result of multiple developmental
defects, the most obvious being a loss of definitive erythropoiesis. Closer examination revealed that a defect in engulfment-mediated DNA
degradation is the primary defect in DNase II-null mice. In this review, we have compiled information from studies on DNase II from various
organisms to provide a consensus model for the role of DNase II enzymes in DNA degradation.
D2003 Elsevier B.V. All rights reserved.
Keywords: Deoxyribonuclease; DNA-degradation; Engulfment-mediated; Phagocytosis; Apoptosis
1. Characterization of DNase II in mammals
Deoxyribonucleases (DNases), the enzymes that break
down DNA molecules, have over time been classified into
many different groups based upon the various biochemical
properties they exhibit (Laskowski, 1967). First observed in
the late 1940s (Catchside and Holmes, 1947), the mamma-
lian DNase activity that exhibited an acid pH optimum
(ranging from approximately pH 4.5 –5.5) was termed ‘‘acid
DNase’’. Acid DNase activity was found to have a ubiqui-
tous tissue distribution (Cordonnier and Bernardi, 1968);
however, it was traditionally analyzed in spleen because of
relatively high activity in this tissue and thus was often
referred to as ‘‘spleen acid DNase’’. The term DNase II was
subsequently suggested as an alternative to spleen acid
DNase in order to more readily distinguish it from what
was known as pancreatic DNase, which was called DNase I
(Cunningham and Laskowski, 1953).
Extensive biochemical analysis has demonstrated that
DNase II enzymes are endonucleases that hydrolyze the
phosphodiester backbone of DNA molecules by a single-
strand cleavage (nicking) mechanism (Lyon and Aguilera,
1997; Baker et al., 1998; Lyon et al., 2000) that generates 3V-
phosphate groups (Bernardi, 1971; Harosh et al., 1991). The
0378-1119/$ - see front matter D2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.gene.2003.08.022
Abbreviations: DNase, deoxyribonuclease; dDNase II Drosophila
DNase II; DLAD, DNase II-like acid DNase; PCD, programmed cell death;
C. elegans,Caenorhabditis elegans; CAD, Caspase-activated DNase; Nuc,
nuclease defective; EST, expressed sequence tag; Lo, low activity; UAS,
upstream activation sequence; Ced,cell-death defective; GMR, glass
multimer reporter; RNAi, RNA inhibition; TUNEL, Tdt-mediated dUTP-
nicked-end labeling; AIF, apoptosis-inducing factor; WAH-1, worm AIF
homologue; Endo G, mitochondrial endonuclease G; CPS-6, C. elegans
Endo G ortholog; CRN, cell death-related nucleases.
* Corresponding author. Tel.: +1-915-747-6852; fax: +1-915-747-
5808.
E-mail address: raguilera@utep.edu (R.J. Aguilera).
www.elsevier.com/locate/gene
Gene 322 (2003) 1 – 15

chemistry of the cleavage event is mediated by a catalytic
center that contains a critical histidine residue and does not
depend upon the presence of divalent cations (Oshima and
Price, 1973; MacLea et al., 2002). Although not required for
activity, certain divalent cations, such as zinc and copper,
strongly inhibit DNase II activity (Bernardi, 1971; Hevelone
and Hartman, 1988; Lyon and Aguilera, 1997). Furthermore,
monovalent cations such as sodium also have inhibitory
effects at increasing concentrations. As mentioned above,
DNase II exhibits an acid pH optimum of approximately 5.0,
although its activity can be detected over a significant pH
range (Lyon and Aguilera, 1997; Lyon et al., 2000). Thus, pH
and sequestration within the lysosomal compartment are
likely to be the primary means by which DNase II activity
is regulated in vivo.
Although DNase II functions as a general endonuclease
that cleaves DNA without apparent sequence specificity, its
physical interaction with DNA imparts at least some
cleavage preferences and limitations. Early work by Ber-
nardi (1971) used spectrophotometry to describe the action
of DNase II upon macromolecular DNA substrates, which
was shown to occur in three mechanistic phases. The initial
phase is characterized by induction of multiple single-strand
breaks (nicks) within the phosphodiester backbone. The
middle phase is characterized by the generation of acid-
soluble nucleotides and the liberation of oligonucleotides
on the order of one hundred nucleotides in length or less.
The terminal phase consists of a slow, non-linear hyper-
chromic shift due to the further reduction of oligonucleotide
size.
Cleavage-site preferences of DNase II enzymes can be
most readily detected during the middle phase of DNA
catabolism (Carrara and Bernardi, 1968; Bernardi, 1971;
Ehrlich et al., 1971). Analysis of DNA strand termini
generated during the middle degradation phase demonstrat-
ed that 72% of the 3V-phosphate terminal nucleotides were
purines, with guanine representing 44% and adenine repre-
senting 28%. With regard to 5V-hydroxyl nucleotides, the
distribution is somewhat wider with purines represented
56% of the time; however, thymine was present only 13%
of the time. In addition to the termini, the 5V-hydroxyl
penultimate nucleotide was examined. At this position,
purines were represented 75% of the time, with adenine
constituting 52% and guanine constituting 23%. These
preferences suggested that DNase II interacts with at least
three DNA base pairs during the cleavage process. Our
recent work has revealed that DNase II enzymes from
several species, including Caenorhabditis elegans,Dro-
sophila melanogaster and Bos taurus, exhibit a strong
cleavage preference for the site AGAGGA (see arrows in
Fig. 1A and Evans et al., 2002). Cleavage at this site is in
strong agreement with the preferred cleavage pattern de-
scribed for the middle phase of DNase II-mediated DNA
degradation. During the terminal phase, a drift in the
composition of the terminal nucleotides was observed that
resulted in a more random distribution of the nucleotide
constituents, which would be expected with a diminishing
number of ‘‘preferred’’ cleavage sites (Bernardi, 1971;
Ehrlich et al., 1971).
2. Identification and characterization of the mammalian
DNase II genes
Although DNase II enzymes had been extensively
characterized biochemically (Bernardi et al., 1965; Town-
end and Bernardi, 1971; Oshima and Price, 1973), the
genes encoding them were only first described in 1998
(Baker et al., 1998; Krieser and Eastman, 1998; Shiokawa
and Tanuma, 1998; Yasuda et al., 1998a). The human
DNase II gene was the first DNase II gene to be identified
and cloned (Yasuda et al., 1998a). Human DNase II was
purified from liver and subjected to N-terminal protein
sequencing. The Rapid Amplification of cDNA Ends
(RACE) technique, employing degenerate primers based
upon the peptide sequences, was used to identify the
corresponding cDNA insert from a human thyroid library.
Soon after this initial identification, another group also
reported cloning of the human DNase II gene (Baker et al.,
1998). This analysis began with the further purification (by
column chromatography) of partially purified, commercial-
ly available porcine DNase II (Sigma). SDS-PAGE analysis
Fig. 1. Cleavage of double-stranded DNA by DNase II enzymes from
various organisms. (A) Analysis of the site-specific cleavages mediated by
DNase II activities detected in crude extracts of D. melanogaster,C.
elegans and B. taurus. Arrows indicate the location of the two most
prominent cleavages detected at the sequence AGAGGA (Evans et al.,
2002). (B) Presence of low molecular-weight DNA fragments in crude
DNA extracts of wild-type (Oregon R) and DNase II
lo
hypomorphic mutant
flies. Bracket indicates the relative migration pattern of the low molecular
weight DNA detected in the DNase II mutant line. DNA extracts were
separated on a 1.0% agarose gel alongside a commercial 123-bp molecular-
weight marker (C.J. Evans, 2002).
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–152

of this purified DNase II revealed the presence of a major
band at approximately 35 kDa and a few minor bands
between 10 and 15 kDa, consistent with previous obser-
vations (Liao, 1985; Yasuda et al., 1998a). Peptide se-
quencing of these protein bands yielded three unique
sequences that exhibited high homology to a corresponding
human protein encoded by the placenta-derived expressed
sequence tag (EST) T53394. Oligonucleotide probes based
on the T53394 sequence were used to retrieve putative
human DNase II clones from a human placental cDNA
library. The human DNase II cDNA sequence was then
utilized to identify the murine DNase II gene (Baker et al.,
1998).
The human DNase II genomic locus is located within the
p13.2 region of chromosome 19 (Baker et al., 1998; Krieser
and Eastman, 1998; Shiokawa and Tanuma, 1998; Yasuda
et al., 1998a,b), which is consistent with a previous pedi-
gree analysis of a human DNase II expression polymor-
phism indicating autosomal inheritance (Yasuda et al.,
1992). Specifically, the DNase II gene corresponded to
the predicted gene R31240_2 previously identified by the
Human Genome Project (Krieser and Eastman, 1998; Shio-
kawa and Tanuma, 1998). Interestingly, our laboratory had
previously identified and cloned R31240_2asagene
encoding the human homolog of an acid endonuclease
called Endonuclease involved in Somatic/Switch Recombi-
nation (Endo-SR) (see below and Lyon, 1997; Lyon and
Aguilera, 1997).
In an effort to identify a novel nuclease activity involved in
immunoglobulin isotype switch recombination, work from
our laboratory identified Endo-SR, which preferentially
cleaved DNA within switch recombination sequences (Lyon
et al., 1996; Lyon and Aguilera, 1997). Furthermore, Endo-
SR exhibited an acid pH optimum and functioned indepen-
dently of divalent cations. Endo-SR was subsequently puri-
fied to homogeneity from bovine spleen and subjected to
peptide microsequencing (Lyon et al., 1996; Lyon, 1997;
Lyon and Aguilera, 1997). Three separate peptide sequences
were obtained from this analysis and compared to database
sequences. No apparent homology was observed with any
known proteins; however, all three peptides were virtually
identical to regions within the putative human protein
R31240_2. Using primers corresponding to the database
sequence, the human Endo-SR (R31240_2) cDNA was am-
plified by RT-PCR from the Nalm-6 human preB-lymphocyte
line (Lyon, 1997). Because switch recombination had pri-
marily been studied in the mouse and because any future
genetic analyses would likely utilize the murine system, the
putative human Endo-SR cDNA was then used to screen a
murine genomic DNA cosmid library, which led to the
cloning of the murine Endo-SR cDNA (Lyon, 1997; Lyon
and Aguilera, 2002, US Patent #6,455,250). Based on the
biochemical characteristics of the enzyme, as well as the
identification of R31240_2 as human DNase II, it became
clear that Endo-SR and DNase II were one and the same
enzymes.
3. Analysis of the human DNase II gene and
corresponding enzyme
The human DNase II gene contains six exons and spans
approximately six kilobases of genomic sequence (Shio-
kawa and Tanuma, 1998). Analysis of the human DNase II
cDNA confirmed that all predicted exons are contained
within a 1,080 base pair open reading frame (ORF) encoding
a protein of 360 amino acids (Lyon, 1997; Baker et al., 1998;
Yasuda et al., 1998a). The amino terminus contains a
hydrophobic signal sequence that is cleaved after amino
acid 16 (Ala), presumably upon insertion into the endoplas-
mic reticulum (ER; Baker et al., 1998). Additionally, four
potential N-glycosylation sites (consensus N-X-S/T; Gavel
and von Heijne, 1990) were detected within the protein
sequence that may serve as mannose-6-phosphate addition
sites (Baker et al., 1998; Yasuda et al., 1998a). Both of these
structural features are required for the predicted lysosomal
localization of DNase II enzymes (De Duve and Wattiaux,
1966; Dulaney and Touster, 1972; Detwiler and MacIntyre,
1980). Additionally, peptide sequence obtained from each
purified DNase II subunit (Baker et al., 1998; Yasuda et al.,
1998a) was found within the R31240_2 open reading frame,
indicating that a single gene encoded DNase II. This
suggested that DNase II is actually synthesized as a zymo-
gen that is cleaved into the observed a(f35 kDa) and h
(f10 kDa) subunits upon maturation. Such a structural
organization and predicted maturation mechanism was con-
sistent with that of other lysosomal enzymes, such as
cathepsins B and D (Faust et al., 1985; Chan et al., 1986).
Based on the predicted mature DNase II aand hpolypep-
tides, recombinant peptides were produced and purified
from Escherichia coli (Yasuda et al., 1998a). The polypep-
tides were reactive to DNase II-specific antibody in Western
blot analyses, confirming that these sequences encoded
DNase II; however, none of the fragments, either alone or
in combination, exhibited DNase II activity. This lack of
DNase II activity was interpreted as consistent with the
previous observation that once DNase II subunits are sepa-
rated, activity cannot be reconstituted (Liao, 1985). Thus, it
was suggested that the structure of the large pro-DNase II
polypeptide was important to the generation of the mature
DNase II enzyme.
In an alternative approach, the full-length human DNase
II cDNA was cloned into a C-terminal poly-histidine-tag
expression vector to generate a tagged recombinant protein
(Baker et al., 1998). Western blot analysis using anti-His
antibodies identified a 43-kDa protein in transfected cell
extracts but failed to identify any low-molecular-weight
forms that would be indicative of processing DNase II into
subunits. The purified His-tagged DNase II was catalytically
active, exhibited an acid pH optimum and was insensitive to
the presence of EDTA, similar to endogenous DNase II
(Baker et al., 1998). Using this system, it was also demon-
strated that DNase II functioned mechanistically by endo-
nucleolytic nicking of DNA (Baker et al., 1998). Interest-
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 3

ingly, 20– 30% of the total DNase II present was secreted
into the media in these assays, although the mechanism of
secretion or function of the secreted nuclease is not currently
known.
In 1973, Oshima and Price, using biochemical techni-
ques, predicted that DNase II catalysis likely involved a
histidine side chain. Subsequently, Liao (1985) demonstrat-
ed that, for porcine DNase II, this histidine is found within
the aforementioned ‘‘asubunit’’ and is contained within the
amino acid sequence Ala-Thr-Glu-Asp-His-Ser-Lys-Trp.
The subsequent cloning of porcine DNase II (Shiokawa
and Tanuma, 1998; Wang et al., 1998) has demonstrated
that this histidine is amino acid 297, which corresponds to
His295 of human DNase II. Replacement of His295 with
an alanine residue in recombinant human DNase II resulted
in a loss of enzymatic activity, indicating that this amino
acid likely participates in catalysis (MacLea et al., 2002).
Additionally, glycosylation is apparently required for max-
imal DNase II activity because pretreatment of cells with
tunicamycin, an inhibitor of N-linked glycosylation, virtu-
ally abolished DNase II activity even though the protein is
still present (MacLea et al., 2002). This treatment also
shifted the molecular weight of DNase II from 43 kDa to
approximately 37 kDa, confirming that DNase II is glyco-
sylated (MacLea et al., 2002). The observed non-glycosy-
lated size is in excellent agreement with the predicted
molecular weight of 38 kDa after the proteolytic removal
of the leader peptide upon entry into the secretory path-
way. Based on expression of recombinant protein (see
previous section) and the aforementioned results, it is very
likely that DNase II functions as a single polypeptide
in vivo.
Previous investigations have observed DNase II enzy-
matic activity in all tissues examined (Cordonnier and
Bernardi, 1968). As expected, survey of tissues by RT-
PCR (Yasuda et al., 1998a), Northern hybridization (Baker
et al., 1998; Shiokawa and Tanuma, 1998) and tissue-
expression-array hybridization (Krieser et al., 2001) revealed
variable, though clearly ubiquitous DNase II expression
patterns.
Fig. 2. Sequence alignment of DNase II proteins isolated from several species. Alignment of human DNase II, mouse DNase II, C. elegans NUC-1 and
Drosophila DNase II polypeptides demonstrates that these proteins are highly evolutionarily conserved with multiple conservative substitutions (grey boxes)
and a significant number of conserved residues (black boxes). As might be expected, the sequences within the catalytic domain are highly conserved (see boxed
residues). The similarities (dots) and identities (asterisks) shown below the sequence were detected with the ClustalW alignment program.
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–154

In addition to the identification of the human DNase II
gene, several groups have identified and described the
murine (Lyon, 1997; Baker et al., 1998), porcine (Shiokawa
and Tanuma, 1998; Wang et al., 1998) and bovine (Krieser
and Eastman, 1998) DNase II genes. Each of the
corresponding proteins exhibited extensive conservation
with human DNase II at the amino acid level, as well as
various structural features including overall size, the pres-
ence of leader peptides, intron– exon structure, disulfide
bridging, N-linked glycosylation sites and a highly con-
served catalytic domain. As mentioned previously, these
DNase II proteins showed no obvious similarity to any
known sequences; however, they were highly homologous
to three open reading frames in C. elegans (C07B5.5,
K04H4.6 and F09G8.2; Lyon, 1997; Lyon et al., 2000;
Baker et al., 1998; Krieser and Eastman, 1998; Wang et al.,
1998) and a single open reading frame in Drosophila
(CG7780; Evans et al., 2002). The characterizations of the
C. elegans and Drosophila DNase II homologs are de-
scribed further below.
Other putative genes encoding proteins homologous to
DNase II have recently been observed in the genomes of
many species including Gallus gallus (chicken), Fugu
rugripes (pufferfish), Xenopus laevis (frog) and Anopheles
gambiae (mosquito; MacLea et al., 2003). Interestingly,
proteins of high homology were also found in the slime
mold Dictyostelium discoideum, the protozoan Trichomonas
vaginalis and the bacterium Burkholderia pseudmallei
(MacLea et al., 2003). Analysis of these homologs, along
with those already known, has revealed a striking conserva-
tion of cysteine residues as well as the amino acid residues
surrounding the catalytic site, indicating their functional
importance (MacLea et al., 2003; see Fig. 2). Whether the
putative non-metazoan homologs exhibit DNase II activity
or function in some other fashion is not known. The presence
of DNase II-like genes in these species suggests a very early
origin of a DNase II ancestral gene, however this is con-
founded by a lack of DNase II-like genes in yeasts, plants or
archaebacteria (MacLea et al., 2003). It should also be noted
that a protein of high homology to DNase II, called DNase II-
like acid DNase (DLAD), also known as DNase IIh(Shio-
kawa and Tanuma, 1999; Krieser et al., 2001), has been
found in mammals. DLAD appears to be a divergent form of
DNase II that presumably arose through gene duplication.
The biochemical characteristics and primary protein struc-
ture of DLAD are similar to that of DNase II, although
DLAD is functional over a much broader pH range (Shio-
kawa and Tanuma, 1999). Unlike the ubiquitous tissue
distribution of DNase II expression, DLAD appears to be
expressed only in a few tissues, which vary between species.
For example, mice express DLAD specifically in liver while
human DLAD is absent from the liver but is highly expressed
in the salivary gland (Krieser et al., 2001). Because DLAD
enzymes are distinct from DNase II and appear to be limited
to mammals (Krieser et al., 2001), they will not be consid-
ered further here.
4. DNase II enzymes and DNA degradation
4.1. C. elegans
Despite its extensive characterization in mammals, stud-
ies from C. elegans provided the first indication of how
DNase II enzymes functioned in vivo (Sulston, 1976; Hedge-
cock et al., 1983; Lyon et al., 2000; Mukae et al., 2002).
DNase II enzymes were, of course, predicted to degrade
DNA, but how cells and tissues utilized DNase II was not
known. Furthermore, nothing was understood about the
relative physiological importance of DNase II-mediated
DNA degradation. Sequence analysis of the C. elegans
genome revealed three putative DNase II-like genes,
C07B5.5, K04H4.6 and F09G8.2 (Lyon, 1997; Lyon et al.,
2000). Previous biochemical analysis demonstrated the pres-
ence of a single detectable acid endonuclease in C. elegans
extracts with functional similarities, including an acid pH
optimum, to mammalian DNase II (Hevelone and Hartman,
1988). This activity was shown to be deficient in a mutant
strain called nuc-1; however, it was unclear whether nuc-1
encoded the acid endonuclease or was a regulator of its
function (Sulston, 1976; Hedgecock et al., 1983). Of the
three putative open reading frames, only C07B5.5 mapped
near the nuc-1 genetic locus, suggesting that C07B5.5 might
encode an acid endonuclease corresponding to NUC-1 (Lyon
et al., 2000; Wu et al., 2000). This hypothesis was subse-
quently confirmed by molecular and biochemical analysis as
well as genetic complementation (Lyon et al., 2000; Wu et
al., 2000).
Recently, the K04H4.6 DNase II-like gene was implicated
as a cell death-related nuclease (crn gene)byanRNA
inhibition (RNAi)-based functional genomic approach (Par-
rish and Xue, 2003; described in more detail below).
Although both NUC-1 and K04H4.6 (crn-6)appearto
function at a later stage of the apoptotic DNA degradation
pathway, crn-6 inhibition did not result in the persistence of
bacterial DNA in the gut lumen or pycnotic bodies in the
cytoplasm of phagocytic cells, typical of the nuc-1 pheno-
type (Parrish and Xue, 2003). It is therefore likely that CRN-
6 plays a different role than NUC-1 in the process of DNA
degradation, although what role that may be remains unclear.
The identification of nuc-1 as a DNase II homolog (see
Fig. 2) was interesting not only because it definitively
showed that the DNase II family of enzymes extends to
invertebrates, but also because of other phenotypic aspects
associated with the nuc-1 mutation. In nuc-1 mutants, the
nuclei of dead cells persist as pycnotic masses within
phagocytic cells (Sulston, 1976; Hedgecock et al., 1983;
Lyon et al., 2000; Wu et al., 2000). Interestingly, analysis of
cell corpses in engulfment-defective mutants revealed that
DNA is not degraded within the dead cells despite these
animals exhibiting normal NUC-1 activity (Hedgecock et al.,
1983). Taken together, these observations suggested that
NUC-1-mediated DNA degradation was somehow depen-
dent upon the process of engulfment. Another characteristic
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 5

of nuc-1 mutants is the failure to degrade ingested bacterial
DNA, which accumulates in the gut lumen (Hedgecock et al.,
1983; Lyon et al., 2000) and indicates that NUC-1 is secreted
to degrade DNA in an extracellular environment, similar to
mammalian DNase II (Baker et al., 1998). As expected,
DNA replication, recombination and repair were normal in
nuc-1 mutant strains (Hevelone and Hartman, 1988; Hen-
gartner, 1997).
Using TdT-mediated dUTP-nicked-end labeling
(TUNEL; Gavrieli et al., 1992), it was possible to follow
DNA fragmentation that occurs in vivo during C. elegans
programmed cell death (Wu et al., 2000). Analysis of 1.5-fold
stage wild-type embryos revealed that an average of 1.7
nuclei were TUNEL-reactive among an average of 14 dying
cells. Similar analysis of nuc-1 embryos revealed a dramatic
increase in the number of TUNEL-reactive nuclei (47.8
nuclei for the nuc-1 mutant strain e1392) without any change
in the observable number of cell corpses. This data indicated
that in apoptotic cells, initial genomic fragmentation occurs
independent of nuc-1 function, followed rapidly by further
processing by nuc-1. This cell-autonomous role is inferred
from the accumulation of TUNEL-reactive DNA ends in nuc-
1mutant animals (Wu et al., 2000). That NUC-1 functions in
the elimination of TUNEL-reactive DNA ends is consistent
with the generation of 3V-phosphate groups (which are not
substrates for the TdT enzyme) by DNase II enzymes. Further
analysis demonstrated that the engulfment genes ced-2,ced-
5,ced-6 and ced-10 (or the process of engulfment, in general)
were neither required for the generation of TUNEL-reactive
DNA ends nor NUC-1-mediated elimination of TUNEL-
reactive ends (Wu et al., 2000). Thus, it was proposed that
the lack of complete corpse DNA degradation in engulfment
mutants was due to a block of a downstream nuclease
normally activated or contributed by the engulfing cell.
Whether this was a novel nuclease present in either cell or
NUC-1 functioning again in the phagocytic cell was not
addressed.
The simplest interpretation of the above data is that NUC-
1 functions in a two-step mechanism. Initially, NUC-1
functions in dying cells to fragment nuclear DNA (Fig. 3).
However, how NUC-1 gains access to the nucleus is un-
known. Subsequently, after engulfment, lysosomal NUC-1
of the phagocytic cells completely degrades the pre-frag-
mented DNA (Fig. 3). This argument, is based on the fact
that undegraded nuclear DNA persists in nuc-1 mutants after
engulfment of dead or dying cells (Hedgecock et al., 1983;
Lyon et al., 2000). An alternative interpretation is that NUC-
1 in the dying cell functions to generate a competent DNA
substrate for another downstream nuclease, the function of
which is engulfment-dependent. While possible, genetic
evidence argues against this mechanism because mutation
of a downstream nuclease, if not required for viability, would
likely phenocopy nuc-1. Despite numerous genetic screens
and several independent isolations of nuc-1 alleles, no gene
that can phenocopy the nuc-1 mutation has so far been
described.
Fig. 3. Role of NUC-1 in PCD and engulfment-mediated DNA degradation in C. elegans. During the early stages of PCD, TUNEL-positive DNA is generated
via the activation of an unknown nuclease. Prior to engulfment of apoptotic DNA, NUC-1 (open circles) further degrades the partially cleaved TUNEL-positive
DNA into a TUNEL unreactive state (see text for additional details). At this point, NUC-1 may be released from destabilized lysosomes (see open arrows) and/
or activated by a transient lowering of intracellular pH. After engulfment of an apoptotic cell, NUC-1 is released into the phagosome after fusion of lysosomes
and proceeds to degrade ingested DNA.
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–156

A recent search for genes involved in DNA degradation
during C. elegans PCD identified seven novel cell death-
related nucleases (crn genes) in addition to the previously
identified nuc-1 and cps-6 genes (Parrish and Xue, 2003).
CPS-6 and WAH-1, the nematode homologues of Endo-G
and apoptosis-inducing factor (AIF), respectively, were pre-
viously shown to cooperate to promote DNA degradation
during PCD (Parrish et al., 2001; Wang et al., 2002).
Interestingly, several of the newly identified crn genes have
known nuclease homologs such as the mammalian FEN-1
endonuclease (crn-1), the E. coli TatD magnesium-dependent
nuclease (crn-2), the mammalian Cyclophilin E protein (cyp-
13) and 3Vto 5Vexonucleases (crn-4 and crn-5;Parrish and
Xue, 2003). Based on genetic and biochemical assays, the
authors concluded that there are at least two cell-autonomous
DNA degradation pathways with crn-2 and crn-3 functioning
in a distinct pathway from that including crn-1,crn-4,crn-5
and cyp-13. The proteins in the latter set appear to interact
with each other as well as CPS-6 perhaps to form part of a
degradation complex that the authors refer to as a ‘‘degrado-
some’’ (Parrish and Xue, 2003). The two pathways identified
by Parrish and Xue (2003) appear to contribute independently
to cell death while the DNase II-like enzymes, NUC-1 and
CRN-6, appear to function at a subsequent stage of DNA
degradation, consistent with the model presented in Fig. 3.It
is important to mention that the putative nuclease responsible
for creating TUNEL-positive DNA breaks at the onset of
PCD (Fig. 3) was not identified by the aforementioned RNAi-
basedapproachindicatingthattheremayyetbemore
nucleases left to be uncovered. Lastly, Parrish and Xue
(2003) report that inhibition of apoptotic DNA degradation
negatively affects cell corpse engulfment. This evidence
comes from the observation that inhibition of crn-2/crn-3
and cps-6 results in persistent un-engulfed cell corpses
(Parrish and Xue, 2003). How phagocytic cells are influenced
by the state of DNA degradation in a dying cell is an
interesting question that will likely be resolved in the near
future.
4.2. D. melanogaster
Similar to C. elegans, early work in Drosophila identified
an acid DNase activity in crude extracts that exhibited similar
biochemical characteristics to mammalian DNase II (Boyd
and Mitchell, 1965). Recent analysis of Drosophila acid
DNase cleavage preferences has shown that the enzyme
recognizes DNA substrates in a very similar manner to C.
elegans NUC-1 and mammalian DNase II enzymes (Evans et
al., 2002; see Fig. 1A). Previous analyses of acid DNase
identified three electrophoretic variants of the protein (rep-
resenting three wild-type alleles), the gene for which was
referred to as DNase 1 (Grell, 1976; Detwiler and MacIntyre,
1978). Additionally, DNase 1 activity was found to be
developmentally controlled, reaching a maximum during
metamorphosis (Detwiler and MacIntyre, 1978). An acid
DNase hypomorphic variant called DNase 1
lo
was subse-
quently generated that exhibits very low levels of acid DNase
activity (Grell, 1976; Evans et al., 2002). We and others have
detected an array of low-molecular-weight DNA species in
DNase 1
lo
flies, particularly in the ovaries, that is not present
in wild-type flies (see Fig. 1B;Mukae et al., 2002). This
observation was consistent with the previously described
phenotype of the DNase 1
lmh
mutant (which unfortunately
no longer exists) that was shown to accumulate DNA
molecules on the order of f200 bp in length (Stone et al.,
1983). This accumulation was found to be ubiquitous, al-
though mature ovaries accumulated DNA at a relatively
higher rate with DNA-containing vesicles found near the
chorionic appendages of oocytes and in the lateral oviducts
(Stone et al., 1983). It was suggested that the presence of this
low-molecular-weight DNA was the result of deficient deg-
radation of nurse cell nuclei during oocyte maturation (Stone
et al., 1983; Mukae et al., 2002). Because the accumulated
DNA was neither high-molecular-weight nor on the order of
oligonucleotides, a model was proposed in which DNase 1
functions in a secondary step to degrade low-molecular-
weight DNA into oligo-and mononucleotides (Fig. 4B).
Interestingly, such a model is consistent with the recent
identification of a mammalian caspase-activated DNase
(CAD) homolog in Drosophila (dCAD; Yokoyama et al.,
2000). CAD appears to be the primary cell-autonomous
nuclease in mammals responsible for internucleosomal
DNA fragmentation during apoptotic cell death, a process
that liberates DNA multimers of f180 bp (Enari et al.,
1998).
As with C. elegans, the genome of Drosophila was found
to potentially encode a DNase II-like enzyme called
CG7780 (Fig. 2;Vernooy et al., 2000; Evans et al., 2002).
Interestingly, the physical location of CG7780 was coinci-
dent with the DNase 1 genetic locus, suggesting that
CG7780 might encode acid DNase in Drosophila (Evans
et al., 2002). Sequence analysis of CG7780 in the DNase 1
lo
line revealed a missense mutation within the coding region
converting a serine into an asparagine codon (Evans et al.,
2002). To confirm that CG7780-encoded Drosophila DNase
1 (DNase II, see below), transgenic DNase 1
lo
lines were
generated that ubiquitously expressed wild-type CG7780.
Complementation experiments revealed that expression of
the wild-type CG7780 ORF rescued the acid DNase defi-
ciency (Evans et al., 2002).
A unique feature of Drosophila DNase II (DNase 1 has
been renamed DNase II to be consistent with the mamma-
lian terminology; Evans et al., 2002; see Flybase, http://
flybase.bio.indiana.edu) is that it has only one intron, which
contrasts with the multiple-intron structures of C. elegans
nuc-1 (six introns) and human DNase II (five introns; Evans
et al., 2002). Drosophila DNase II is similar in size to other
DNase II enzymes (366 amino acids; C. elegans NUC-1,
375 amino acids; human DNase II, 360 amino acids) and
shares 33% identity and 46% similarity with the NUC-1
enzyme (Fig. 2;Evans et al., 2002). Like other DNase II
enzymes, Drosophila DNase II has a leader peptide for entry
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 7

into the secretory pathway and has a highly conserved
catalytic domain. It also contains a potential consensus N-
linked glycosylation site at position 109 (NGT), which is
consistent with the previous detection of mannose-6-phos-
phate with highly purified Drosophila acid DNase (Gaszner
and Udvardy, 1991; Evans et al., 2002).
As mentioned above, Drosophila expresses a DNase
activity (dCAD) that functions similarly to mammalian
CAD (Yokoyama et al., 2000). Upon apoptotic activation,
dCAD begins to fragment genomic DNA by cleaving the
linker region between nucleosomes. Examination of oligo-
nucleosomal DNA fragmentation (laddering) revealed that,
while wild-type embryos or dissected mature ovaries exhibit
a robust activity, dCAD-deficient mutants fail to fragment
their DNA (Mukae et al., 2002).Incontrast,DNase 1
lo
mutants exhibit an enhanced fragmentation phenotype. In
mutants deficient for both enzymes, again no laddering was
observed (Mukae et al., 2002). These observations indicate a
mechanism of DNA degradation similar to that proposed by
Stone et al. (1983), where dCAD is required to generate low-
molecular-weight nucleosomal multimers and DNase II
functions to remove this DNA (Fig. 4). Further analysis of
these DNase deficiencies revealed that lack of DNase II
activity specifically resulted in the upregulation of the anti-
bacterial peptides attacin A and diptericin, but not the
antifungal gene drosomycin (Mukae et al., 2002). The ex-
pression of these genes was enhanced by dCAD-deficiency,
although dCAD-deficiency alone did not result in any ob-
servable upregulation. Thus, DNase II deficiency results in
increased antibacterial-gene expression, presumably due to
increased susceptibility to bacterial infection. It is not known
whether this is a direct or indirect effect although it is likely
that DNase II-deficiency negatively impacts macrophage
function thus impairing cellular immunity (Elrod-Erickson;
see the following sections). Importantly, this also indicates
that defects in DNase II function can have a significant impact
upon systems other than DNA degradation.
4.3. Mice
It is known that phagocytes, particularly macrophages
and related cell types, contain enzymatic activities capable
Fig. 4. DNA degradation during PCD and engulfment-mediated DNA degradation. (A) Schematic representation of the cell autonomous and non-cell
autonomous (engulfment-mediated) steps leading to the degradation of apoptotic DNA. Persistent undegraded nuclear DNA (see pycnotic nucleus) is one of the
hallmark features of the nuc-1 mutation in C. elegans, and similar DNA accumulation has been shown to occur in the murine phagocytes defective for DNase II.
(B) Nucleases implicated in apoptotic DNA degradation. Note that in C. elegans, a CAD-homologue has not been identified (?) and, although several death-
related nucleases (crn genes) have been recently identified, it remains unclear which nuclease is responsible for the initiation of DNA degradation (Wu et al.,
2000; Parrish and Xue, 2003). It is, however, highly likely that the mammalian homologues of the recently identified crn genes, some of which had previously
been implicated in PCD, will play similar roles in the progression of DNA degradation in mammals. It has been well established that CAD of mammals and
Drosophila generates the cell autonomous nucleosomal-laddering commonly observed during PCD (Mukae et al., 2002; Enari et al., 1998). In CAD-deficient
mice and flies (cad /), this early DNA degradation step is not detected although apoptotic nuclear DNA is still destroyed after phagocytosis (see text for
details). Disruption of DNase II activity in homozygous mutant cells (dnase II /) or via inhibition of DNase II function (McIlroy et al., 2000) results in a
block of the final DNA degradation stage that occurs after engulfment.
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–158

of the total degradation of the objects they engulf, which are
usually other cells or cell fragments. Early work on lyso-
somal cellular fractions, made possible by the development
of analytical centrifugation, suggested that acid DNase
likely played a degradative scavenging role (De Duve and
Wattiaux, 1966; Bernardi, 1971; Dulaney and Touster,
1972); however, there was little evidence for this in vivo.
Recent in vitro studies have revealed that a single lysosomal
acid DNase of approximately 43 kDa was absolutely re-
quired for the process of engulfment-mediated DNA degra-
dation (Odaka and Mizuochi, 1999). The identified nuclease
was inhibited by zinc and iodoacetate, suggesting that the
identified enzyme was DNase II (Odaka and Mizuochi,
1999).
Further support for DNase II being involved in engulf-
ment-mediated DNA degradation came indirectly from
studies of CAD (Enari et al., 1998).Toassessthe
importance of cell-autonomous DNA fragmentation during
cell death, transgenic mice incapable of activating CAD
were generated (McIlroy et al., 2000). Despite this defi-
ciency, CAD-deficient mice were viable and appeared to
develop normally. As expected, apoptotic cells (thymo-
cytes) from these animals died with normal kinetics,
however no nucleosomal laddering was detected. In con-
trast, when tissue sections were examined for DNA frag-
mentation by TUNEL, DNA fragmentation was observed
in CAD-deficient tissues. Furthermore, this DNA fragmen-
tation pattern was virtually identical to that observed in
wild-type tissues. Closer microscopic examination of
CAD-deficient tissues revealed that TUNEL-reactive cells
were found only within macrophages, suggesting that
while cells in CAD-deficient mice underwent apoptosis
in a normal fashion, DNA fragmentation was occurring in
and was dependent upon an activity supplied by phago-
cytes (McIlroy et al., 2000). When these experiments were
recapitulated in vitro using apoptotic thymocytes and
macrophages, inhibitors of lysosomal acidification sup-
pressed the macrophage-specific DNA fragmentation.
Thus, DNase II was again implicated as the primary
enzyme responsible for the phagocyte-mediated DNA deg-
radation (McIlroy et al., 2000). Significantly, these results
suggested that cell-autonomous DNA fragmentation may be
dispensable as long as the mechanism of phagocyte-medi-
ated DNA degradation is intact, a conclusion that is
supported by observations in C. elegans (Wu et al., 2000;
Fig. 4).
The subsequent targeted disruption (‘‘knock-out’’) of the
DNase II gene in mice confirmed the requirement of DNase
II activity within phagocytic cells (Kawane et al., 2001;
Krieser et al., 2002). Mice heterozygous (+/ ) for the
DNase II disruption were healthy and fertile; however,
homozygous DNase II-mutant mice ( /) died either in
utero or just after birth (Kawane et al., 2001; Krieser et al.,
2002). Southern blot analysis of DNA from embryos of
timed pregnancies indicated that DNase II /embryos
were present at a frequency of approximately 25%, which is
expected for Mendelian inheritance (Kawane et al., 2001).
Homozygous mutant embryos develop normally through
day 12.5 but by day 17.5 their pale coloration was indic-
ative of severe anemia. Analysis of peripheral blood cells
revealed that DNase II /embryos contained approxi-
mately 10% of the mature erythrocytes observed in hetero-
zygous and wild-type littermates (Kawane et al., 2001).
Additionally, the peripheral blood of homozygous mutant
animals contained what appeared to be erythroblasts that
had not enucleated. The presence of other blood cell types,
including neutrophils and macrophages, appeared normal in
homozygous DNase II /embryos. Examination of
embryos just prior to birth (day 18.5) revealed that,
although severely anemic, DNase II /animals were
still viable indicating that these mice died at or soon after
birth (Krieser et al., 2002).DNase II /mice were also
shown to have diaphragmatic defects as well as uninflated
lungs (Krieser et al., 2002). Thus, DNase II /mice
probably died of asphyxiation, although any contribution of
other factors including anemia could not be ruled out.
In mice, embryonic hematopoiesis occurs primarily in the
fetal liver (Cumano and Godin, 2001). Extraction of blood-
cell progenitors from the fetal liver and subsequent in vitro
culture in the presence of lineage-specific cytokines revealed
that DNase II /hematopoietic precursors were similar to
wild-type in their differentiation potential (Kawane et al.,
2001). Transplantation of bone marrow cells from DNase
II /mice into lethally irradiated wild-type mice (strain
B6.SJL) demonstrated that mature erythrocytes could be
derived from DNase II /cells, indicating that the anemia
observed in DNase II /embryos was not due to a cell-
autonomous defect.
Histological analysis of DNase II /tissue sections
revealed the presence of many foci of various sizes, which
were shown to be reactive to DNA stains (such as Feulgen
and hematoxylin; Kawane et al., 2001; Krieser et al., 2002).
These foci were found in many tissues but were most
prominent in the liver. Examination of these liver foci by
electron microscopy revealed that macrophages were resi-
dent at the center of each focus and contained several
inclusions that appeared to be engulfed pycnotic nuclei
(Kawane et al., 2001). Furthermore, cells of various types
including erythroblasts, mature erythrocytes and reticulo-
cytes often surrounded these macrophages. During normal
erythropoiesis, which occurs in what are commonly referred
to as blood islands, macrophages take up ejected erythro-
blast nuclei as they mature into enucleated erythrocytes
(Sadahira and Mori, 1999). The specificity and timing of
the enucleation event likely requires extensive signaling
between these cells to ensure that erythroblast nuclei are
not lost prematurely or inappropriately to the extracellular
environment. Thus, it is likely that the engorged macro-
phages observed in the fetal liver are blood-island-macro-
phages that were incapable of engulfment-mediated DNA
degradation. Because DNase II /cells were capable of
normal differentiation in a wild-type host, it indicated that
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 9

deficient erythropoiesis in DNase II /mice was likely
the result of dysfunctional interaction between macrophages
and enucleating erythroblasts, and not a direct cell-autono-
mous effect of DNase II-deficiency in the erythrocyte
lineage (Kawane et al., 2001).
A combinatorial analysis of CAD and DNase II function
in mice has recently been reported (Kawane et al., 2003),
similar to that performed in Drosophila (Mukae et al., 2002).
As in previous experiments (McIlroy et al., 2000),in vitro
analysis revealed that wild-type macrophages were capable
of efficient degradation of the DNA of engulfed wild-type or
CAD-deficient apoptotic thymocytes with equal efficiency
(Kawane et al., 2003). In contrast, DNase II /macro-
phages failed to degrade the DNA of either type, although
the accumulation of DNA in these cells was less pronounced
when wild-type thymocytes were engulfed. This indicates
that while DNase II is principally important for engulfment-
mediated DNA degradation of apoptotic cells, CAD and
DNase II normally function in a cooperative manner to
degrade the DNA of apoptotic cells (see Fig. 4). This is
supported by electron microscopy analysis that revealed a
less fragmented appearance of engulfed DNA within DNase
II /macrophages when CAD /apoptotic cells were
used as targets compared to wild-type apoptotic cells
(Kawane et al., 2003).
In general, the gross phenotype of CAD /DNase
II /embryos was very similar to that of DNase II /
embryos; however, the development of the thymus and
kidney were significantly impaired (Kawane et al., 2003).
The thymi of DNase II /and CAD /DNase II /
embryos was significantly smaller and contained significant-
ly fewer cells than wild-type and CAD /embryos.
Analysis of various differentiation markers, including Thy-
1, CD4, CD8, CD44 and CD25, demonstrated that in DNase
II /and CAD /DNase II /embryos, T cell
development was blocked at an early progenitor stage
(Kawane et al., 2003). In particular, DNase II mutant lines
contained a higher percentage of immature CD44 + CD25
precursors and fewer more-mature CD44 CD25 cells.
While lower numbers of double positive cells were detected
in the DNase II /embryo thymi, this reduction was more
pronounced (1/3 of normal) in CAD /DNase II /
embryos. As in other tissues, the thymi of DNase II /and
CAD /DNase II /embryos exhibited significant
accumulation of DNA within macrophages. Additionally,
the thymus of DNase II /and CAD /DNase
II /embryos expressed 5- to 8-fold and 70- to 100-
fold higher levels of interferon-h(IFN-h), respectively,
compared to wild-type thymi. In addition to its role in
immunity, interferon-hhas been reported to have signif-
icant function in the process of thymic atrophy (Su et al.,
1997; Vidalain et al., 2002). Thus, as in Drosophila, lack
of DNase II function in mice not only resulted in
defective DNA degradation but also detrimentally affected
other developmental processes including erythropoiesis
and thymopoiesis.
5. DNase II enzymes and the induction of cell death
Cell-autonomous DNA fragmentation has long been a
hallmark of programmed cell death and as such, a large effort
has been made to identify the nuclease or nucleases respon-
sible for this effect (Hewish and Burgoyne, 1973; Wyllie,
1980; Wyllie, 1998). It is known that DNA damage can elicit
cell death (Rich et al., 2000) and so it has been reasoned that
a directed nucleolytic attack on genomic DNA might be one
mechanism of apoptotic induction. Subsequently, many
different activities, including DNase II, have been implicated
in possibly having a role in this process (Barry and Eastman,
1993; Nagata, 2000). The identification of DNase II
appeared especially relevant in view of the fact that intra-
cellular acidification was a characteristic of many apoptotic
cell deaths (Famulski et al., 1999; Matsuyama et al., 2000;
Pervaiz and Clement, 2002) and because DNase II was
associated with lysosomes, known repositories of hydrolytic
enzymes. Additionally, some studies have reported the
presence of DNase II in nuclear extracts and that the enzyme
was possibly capable of inducing nucleosomal laddering
(Gottlieb et al., 1995; Lyon et al., 1996).
In 1998, Krieser and Eastman reported the identification
of the human DNase II gene, work that was driven by their
previous implication of DNase II in the cell-autonomous
DNA fragmentation associated with apoptosis (Barry and
Eastman, 1993). Subsequently, they were the first to examine
whether overexpression of DNase II was sufficient to induce
cell death. Transient expression of human DNase II in
Chinese hamster ovary cells, along with a GFP-expressing
vector to monitor transfection efficiency, resulted in f20%
of GFP positive cells becoming apoptotic, as assessed by
chromatin condensation. In contrast, less than 5% of control
cells became apoptotic by this criterion indicating that over-
expression of human DNase II was sufficient to induce cell
death, at least in a subset of transfected cells. Whether or not
this induction of cell death was specifically due to DNase II
enzymatic activity was not addressed.
In contrast to the above results, loss-of-function analysis
of C. elegans nuc-1 animals has indicated that NUC-1 has no
effect upon the initiation of cell death. This evaluation was
based upon the timing of the appearance of refractive cell
corpses by differential interference contrast (DIC) microsco-
py (Sulston, 1976; Hedgecock et al., 1983). A recent and
more thorough characterization of the nuc-1 mutation by
TUNEL analysis also failed to identify any statistically
observable alteration in the onset of cell deaths (Wu et al.,
2000). Rather, the nuc-1 mutation only delays the subsequent
elimination of TUNEL-reactive DNA-ends within apoptotic
cells. During our characterization of nuc-1, transgenic nuc-1
lines containing the wild-type gene were generated (Lyon et
al., 2000). These lines were shown to express elevated levels
(fthree-fold) of NUC-1 activity compared to wild-type;
however, no ectopic cell death was observed, suggesting that
overexpression is not sufficient to induce cell death in C.
elegans.
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–1510

During our analysis of Drosophila DNase II, we sought to
analyze whether directed overexpression of this gene would
elicit ectopic cell death and lead to developmental defects
(Evans et al., 2002). It was expected, based upon the data
described above for NUC-1, that overexpression of Dro-
sophila DNase II would not result in an overt deleterious
effect. Using the ubiquitous actin 5C promoter and the
bipartite UAS-GAL4 system (Brand and Perrimon, 1993),
wild-type DNase II was expressed in transgenic lines at
levels greater (more than five-fold) than that observed in
wild-type animals (Evans et al., 2002). This ubiquitous
overexpression had no obvious effect on viability or time
to eclosion (CJE and RJA, unpublished observations). Ad-
ditionally, examination of the gross morphology, size and
fecundity of these flies revealed no discernable abnormali-
ties. This analysis was extended by expressing DNase II at
higher levels specifically in the developing eye using the
construct glass multimer reporter (GMR), which gives very
strong expression in the eye and has been utilized extensive-
ly in the assay of apoptotic induction by various proteins
(Hay et al., 1997). Examination of newly enclosed flies
revealed no discernable difference between the eyes of
control and DNase II-overexpressing flies (Evans et al.,
2002), with the overall size, shape and color of the eye fields
appearing similar and wild-type (see http://utminers.utep.
edu/raguilera/ for color figures). Thus, significant overex-
pression of DNase II either does not induce cell death in
Drosophila or does so very inefficiently.
6. General model of DNase II function
In C. elegans, it has been demonstrated that NUC-1 can
gain access to the nucleus during cell death (Wu et al.,
2000); however, it seems that NUC-1 has little, if anything,
to do with the decision of the cell to carry out the cell death
program. Whether other DNase II enzymes gain access to
the nucleus in dying cells has not been definitively shown.
DNase II loss-of-function analyses in C. elegans,Drosoph-
ila and mice have failed to demonstrate any requirement for
DNase II in the induction or procession of apoptosis. A
similar conclusion can be reached when considering the
DNase II-overexpression analyses described above. Work in
C. elegans indicates that cell-autonomous NUC-1 enzymat-
ic function in apoptotic cells is a likely consequence of prior
activation of the cell death pathway rather than an inductive
mechanism itself. Accordingly, the appearance of refractive
cell corpses and TUNEL-reactivity clearly occur prior to
NUC-1 function (Wu et al., 2000). With regard to DNA
degradation, the cell-autonomous role of NUC-1 appears
relatively minor and possibly not required because of its
likely non-autonomous role subsequent to phagocytosis.
In contrast, work in mice has conclusively shown that
DNase II enzymes have a primary function in non-cell-
autonomous (engulfment-mediated) DNA degradation
(Odaka and Mizuochi, 1999; McIlroy et al., 2000; Kawane
et al., 2003, 2001; Krieser et al., 2002). Significant evidence
in C. elegans also supports such a role for the DNase II-like
enzyme, NUC-1 (Hedgecock et al., 1983; Lyon et al., 2000;
Wu et al., 2000). DNase II has ubiquitous tissue distribution
consistent with the fact that the vast majority of cells, if not
all cells, have some phagocytic capacity. Phagocytes recog-
nize signals presented by dying cells and efficiently begin the
process of attachment and engulfment (Fadok and Chimini,
2001). Engulfment places dying cells within a vesicle called
a phagosome. Lysosomes fuse with phagosomes to form an
acidified intermediate compartment called a phagolysosome,
in which hydrolytic enzymes such as DNase II function to
degrade the internalized materials (Tjelle et al., 2000; see
Fig. 4A). Interestingly, work from both Drosophila and mice
has demonstrated that DNase II is sufficient to degrade the
genomic DNA of engulfed apoptotic cells despite a lack of
prefragmentation by cell-autonomous activities such as CAD
(McIlroy et al., 2000; Mukae et al., 2002). This is likely
assisted by lysosomal proteases that destroy DNA-associated
histones and other chromatin proteins allowing DNase II to
completely digest the engulfed DNA. Furthermore, DNase II
is absolutely required for this process because phagocytes
accumulate engulfed DNA in its absence (Hedgecock et al.,
1983; Stone et al., 1983; Lyon et al., 2000; McIlroy et al.,
2000; Wu et al., 2000; Kawane et al., 2003, 2001; Krieser et
al., 2002; Mukae et al., 2002). This accumulation of DNA is
known to be functionally significant in both mice and
Drosophila as it negatively affects development and immu-
nity (Kawane et al., 2003, 2001; Krieser et al., 2002; Mukae
et al., 2002), and undoubtedly other systems and processes as
well.
7. Future directions
7.1. C. elegans
As mentioned previously, sequencing of the C. elegans
genome identified, in addition to C07B5.5 (nuc-1), two other
putative DNase II homologs, ORFs F09G8.2 (YLS2_
CAEEL) and K04H4.6 (YMV6_CAEEL), located on the
third chromosome (Krieser and Eastman, 2000).The
F09G8.2 ORF exhibits 35% identity and 53% similarity to
human DNase II while ORF K04H4.6 exhibits 32% identity
and 52% similarity (Krieser and Eastman, 2000). RT-PCR
analysis with a mixed-stage population of worms subse-
quently revealed that F09G8.2 and K04H4.6 were indeed
expressed, although when, where or to what extent is
unknown (Krieser and Eastman, 2000). The generation of
reporter constructs utilizing the promoter regions of F09G8.2
and K04H4.6 should reveal their respective expression
patterns. As mentioned above, biochemical analysis has
demonstrated that nuc-1 mutants exhibit virtually no acid
endonuclease activity (less than 0.1% of wild-type), with the
remainder thought to be attributable to mRNA translational
read-through (Lyon et al., 2000). Recently, it was reported
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 11

that one of these ORFs, K04H4.6 (CRN-6), encodes a
nuclease activity as assessed by RNAi (Parrish and Xue,
2003). However, it was also reported that inhibition of crn-6
expression did not result in a nuc-1 phenotype. To examine
whether F09G8.2 and K04H4.6 are functional DNase II-like
enzymes, it should be possible to generate transgenic animals
in which these genes are coupled to constitutive, ubiquitous
promoters. Placing the transgenes in the nuc-1 genetic
background should facilitate the detection of any attributable
activities.
To address where NUC-1 primarily functions to facilitate
the complete removal of DNA, genetic mosaic analysis
should be performed. In fact, a method has already been
devised to specifically distinguish between autonomous and
non-autonomous gene function between apoptotic and
engulfing cells (Liu and Hengartner, 1998; Wu and Horvitz,
1998). With this system, it should be relatively easy to
identify, score and determine where NUC-1 primarily func-
tions in the complete removal of apoptotic-cell DNA.
7.2. D. melanogaster
In previous sections, it was noted that only the DNase 1
lo
hypomorphic allele of Drosophila DNase II is still available
for study (Grell, 1976; Evans et al., 2002). In order to fully
understand the impact of DNase II deficiency in Drosophila,
it will be important to find mutations representing a complete
loss of DNase II activity. An attempt was previously made to
isolate null mutations in Drosophila and several mutations
were recovered that resulted in extremely low levels of
DNase II activity; however, no true null alleles were found
(Detwiler, 1979). Based on these results, it is not clear
whether null alleles of DNase II can be generated in
Drosophila or whether a low level of DNase II is absolutely
required for viability.
Drosophila has an effective pathogen surveillance system
composed of both humoral and cellular immune responses
(Hoffmann and Reichhart, 2002). The humoral response is
mediated by the rapid induction and secretion of anti-
microbial peptides (mainly by the fat body, which is analo-
gous to the mammalian liver) that directly neutralize and
likely opsonize pathogens. The cellular response is mediated
by hemocytes, which consist primarily of plasmatocytes
(macrophage-like cells) and crystal cells (which are involved
in melanization) in adults (Evans and Banerjee, 2003). It has
recently been demonstrated that the cellular and humoral
immunity in Drosophila function cooperatively in the re-
moval of invading pathogens. Drosophila plasmatocytes
readily engulf polystyrene beads until they reach their
‘‘phagocytic limit’’ and cannot further engulf bacteria sub-
sequently introduced into their systems (Elrod-Erickson et
al., 2000). The pre-application of plastic beads to phagocytes
results in a measurable impact on viability subsequent to
bacterial challenge but only in flies mutant for humoral
responses (Elrod-Erickson et al., 2000). It is unclear, how-
ever, whether the observed increase in mortality is directly
due to a phagocytic defect or indirectly due to improper cell
interactions. Interestingly, the accumulation of plastic beads
in Drosophila macrophages appears similar to the accumu-
lation of apoptotic nuclei in macrophages of DNase II /
mice (Kawane et al., 2003). As such, DNase II mutation in
flies may negatively impact immune responses due to
accumulation of DNA within macrophages. As mentioned
earlier, DNase 1
lo
mutants exhibit an enhanced fragmenta-
tion phenotype and most interestingly they also exhibit an
increase expression of the antibacterial genes, diptericin and
attacin A, which are generally induced by bacterial infection
(Mukae et al., 2002). Activation of the antibacterial genes
was found to be further enhanced in flies deficient for both
the CAD and DNase II pathways indicating that DNA-
degradation defects activate the fly humoral immune re-
sponse, although the mechanism of activation is not known.
Therefore, it should be expected that animals deficient in
both DNA degradation and humoral immune response genes
(such as imd) should result in severely immunocompromised
animals. We are currently evaluating whether DNase II
deficiency negatively affects viability and/or the normal
induction of immune responses and are attempting to gen-
erate null Drosophila DNase II alleles using a variety of
modern methodologies.
7.3. Mice
Targeted ablation of DNase II gene expression in mice
leads to death during embryogenesis (Kawane et al., 2001),
which restricts studies of DNase II function to this period of
development. Thus, a more thorough analysis of DNase II
function may benefit from the use of technical innovations,
such as the Cre/loxP system (Lewandoski, 2001).This
system allows embryonic lethal phenotypes to be circum-
vented by the conditional ablation of gene function (by Cre-
mediated deletion of specific DNA sequences) at later stages
of development or in particular tissues or cells. Many
different Cre expression systems have been devised and
some of them, including the LysM –Cre system that ablates
gene function in the vast majority of macrophages and
granulocytes, should be of particular use in studying the
function of DNase II (Clausen et al., 1999; Lewandoski,
2001).
The inefficient clearance of nuclear antigens (particularly
DNA and associated proteins) can predispose mice and
humans to the autoimmune disease known as systemic lupus
erythematosis (SLE; Robson and Walport, 2001). Classical
symptoms of SLE include moderate to high levels of anti-
DNA antibodies (to both double-and single-stranded DNA),
excessive protein in the urine (proteinuria), immune-com-
plex deposition in kidney glomeruli and glomerulonephritis
(inflammatory damage of glomeruli; Robson and Walport,
2001). Using mouse models that are predisposed to the onset
of SLE, several genes have been identified that normally
facilitate the removal of potentially immunogenic cellular
constituents (Walport, 2000). Until recently, however, no
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–1512

DNase defects had been correlated with the disease. Targeted
disruption of the murine DNase I gene (which is unrelated to
DNase II;Napirei et al., 2000) resulted in mice exhibiting
classical SLE symptoms, including the presence of anti-
DNA antibodies in the serum and immune-complex deposi-
tion in the kidneys (Napirei et al., 2000). Interestingly,
disruption of a single gene, DNase I, was sufficient to elicit
SLE symptoms in otherwise wild-type mice (i.e. DNase I-
null-dependent onset of SLE did not require a genetic
predisposition), suggesting that DNase deficiencies may
heavily influence the induction of inappropriate autoimmune
responses. These results were supported by the recent
correlation of DNase I deficiencies in human patients with
SLE (Yasutomo et al., 2001).
Because DNase II is directly implicated in engulfment-
mediated DNA, it is very likely that a generalized deficiency
in DNA clearance due to the disruption of DNase II
expression will lead to some form of autoimmune dysfunc-
tion, in particular SLE. Unfortunately, such a phenotype is
impossible to analyze in DNase II-null mice because they
die in utero due to developmental defects. Thus, the use of
conditional or hypomorphic DNase II mutant mice, gener-
ated as described above, to bypass this lethality may allow
study of DNase II dysfunction in later stages. Additionally,
it may be possible to correlate low DNase II activity with
SLE in human patients, similar to what was observed with
DNase I.
8. Summary
DNase II enzymes are a family of highly homologous
DNases that function primarily in engulfment-mediated
DNA degradation. These enzymes have ubiquitous tissue
distribution, function optimally at acidic pH, do not require
cofactors for efficient catalysis and are associated with
lysosomes. DNase II enzymes were first identified in
mammals; however, proteins with high homology have
recently been identified in invertebrates and non-metazoans.
Furthermore, DNase II homologs in C. elegans and Dro-
sophila have been shown to function similarly to mamma-
lian DNase II. Defects in DNA degradation associated with
mutations in DNase II enzymes have highlighted the phys-
iological importance of efficient DNA removal, particularly
in mammals. Additionally, it has become clear that engulf-
ment-mediated DNA degradation, as opposed to DNA
fragmentation that occurs cell-autonomously, is the primary
mechanism of DNA removal and is critical for proper
development and homeostasis.
Acknowledgements
We thank Dr. William Balwin for critically reading this
review. This work was supported by NIH grants
2G12RR08124 and 2S06-GM008012-33.
References
Baker, K.P., Baron, W.F., Henzel, W.J., Spencer, S.A., 1998. Molecular
cloning and characterization of human and murine DNase II. Gene
215, 281 – 289.
Barry, M.A., Eastman, A., 1993. Identification of deoxyribonuclease II as an
endonuclease involved in apoptosis. Arch. Biochem. Biophys. 300,
440 – 450.
Bernardi, G., 1971. Spleen acid deoxyribonuclease. In: Boyer, P.D. (Ed.),
Hydrolysis. Academic Press, New York, pp. 271 – 287.
Bernardi, G., Appella, E., Zito, R., 1965. Biochemistry 4, 1725 – 1729.
Boyd, J.B., Mitchell, H.K., 1965. Identification of deoxyribonucleases in
polyacrylamide gel following their separation by disk electrophoresis.
Anal. Biochem. 13, 28 – 42.
Brand, A.H., Perrimon, N., 1993. Targeted gene expression as a means of
altering cell fates and generating dominant phenotypes. Development
118, 401 – 415.
Carrara, M., Bernardi, G., 1968. Studies on acid deoxyribonuclease: V. The
oligonucleotides obtained from deoxyribonucleic acid and their 3V-phos-
phate termini. Biochemistry 7, 1121 – 1131.
Catchside, D.G., Holmes, B., 1947. The action of enzymes on chromo-
somes. Symp. Soc. Exp. Biol. 1, 225.
Chan, S.J., San Segundo, B., McCormick, M.B., Steiner, D.F., 1986. Nu-
cleotide and predicted amino acid sequences of cloned human and
mouse preprocathepsin B cDNAs. Proc. Natl. Acad. Sci. U. S. A. 83,
7721 – 7725.
Clausen, B.E., Burkhardt, C., Reith, W., Renkawitz, R., Forster, I., 1999.
Conditional gene targeting in macrophages and granulocytes using
LysMcre mice. Transgenic Res. 8, 265 – 277.
Cordonnier, C., Bernardi, G., 1968. A comparative study of acid deoxy-
ribonucleases extracted from different tissues and species. Can. J. Bio-
chem. 46, 989 – 995.
Cumano, A., Godin, I., 2001. Pluripotent hematopoietic stem cell develop-
ment during embryogenesis. Curr. Opin. Immunol. 13, 166 – 171.
Cunningham, L., Laskowski, M., 1953. Presence of two different desoxy-
ribonucleodepolymerases in veal kidney. Biochim. Biophys. Acta 11,
590 – 591.
De Duve, C., Wattiaux, R., 1966. Functions of lysosomes. Annu. Rev.
Physiol. 28, 435 – 492.
Detwiler, C.R., 1979. A genetic and developmental analysis of DNase-1, an
acid deoxyribonuclease in Drosophila melanogaster. Dissertation, Cor-
nell University, Ithica, NY.
Detwiler, C., MacIntyre, R., 1978. A genetic and developmental analysis of
an acid deoxyribonuclease in Drosophila melanogaster.Biochem. Genet.
16, 1113– 1134.
Detwiler, C.R., MacIntyre, R.J., 1980. A subcellular localization of the
gene product of the DNase-1 locus in Drosophila melanogaster. Insect
Biochem. 10, 255 – 263.
Dulaney, J.T., Touster, O., 1972. Isolation of deoxyribonuclease II of rat
liver lysosomes. J. Biol. Chem. 247, 1424–1432.
Ehrlich, S.D., Torti, G., Bernardi, G., 1971. Studies on acid deoxyribonu-
clease: IX. 5V-hydroxy-terminal and penultimate nucleotides of oligo-
nucleotides obtained from calf thymus deoxyribonucleic acid. Biochem-
istry 10, 2000 – 2009.
Elrod-Erickson, M., Mishra, S., Schneider, D., 2000. Interactions between
the cellular and humoral immune responses in Drosophila. Curr. Biol.
10, 781 – 784.
Enari, M., Sakahira, H., Yokoyama, H., Okawa, K., Iwamatsu, A., Nagata,
S., 1998. A caspase-activated DNase that degrades DNA during apop-
tosis, and its inhibitor ICAD. Nature 391, 43 – 50.
Evans, C.J., 2002. The identification and characterization of mammalian
DNase II homologs in C. elegans and Drosophila. Dissertation, Uni-
versity of California, Los Angeles, CA.
Evans, C.J., Banerjee, U., 2003. Transcriptional regulation of Drosophila
hematopoiesis. Blood Cells Mol. Diseases 30, 223 – 228.
Evans, C.J., Merriam, J.R., Aguilera, R.J., 2002. Drosophila acid DNase is
a homolog of mammalian DNase II. Gene 295, 61 – 70.
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 13

Fadok, V.A., Chimini, G., 2001. The phagocytosis of apoptotic cells.
Semin. Immunol. 13, 365 – 372.
Famulski, K.S., Macdonald, D., Paterson, M.C., Sikora, E., 1999. Activa-
tion of a low pH-dependent nuclease by apoptotic agents. Cell Death
Differ. 6, 281 – 289.
Faust, P.L., Kornfeld, S., Chirgwin, J.M., 1985. Cloning and sequence anal-
ysis of cDNA for human cathepsin D. Proc. Natl. Acad. Sci. U. S. A. 82,
4910 – 4914.
Gaszner, M., Udvardy, A., 1991. Purification of a lysosomal DNase from
Drosophila melanogaster. Biochem. Biophys. Res. Commun. 181,
44 – 50.
Gavel, Y., von Heijne, G., 1990. Sequence differences between glycosy-
lated and non-glycosylated Asn-X-Thr/Ser acceptor sites: implications
for protein engineering. Protein Eng. 3, 433 – 442.
Gavrieli, Y., Sherman, Y., Ben-Sasson, S.A., 1992. Identification of pro-
grammed cell death in situ via specific labeling of nuclear DNA frag-
mentation. J. Cell Biol. 119, 493 – 501.
Gottlieb, R.A., Giesing, H.A., Engler, R.L., Babior, B.M., 1995. The acid
deoxyribonuclease of neutrophils: a possible participant in apoptosis-
associated genome destruction. Blood 86, 2414 – 2418.
Grell, E.H., 1976. Genetics of some deoxyribonucleases of Drosophila
melanogaster. Genetics 83, s28 – s29.
Harosh, I., Binninger, D.M., Harris, P.V., Mezzina, M., Boyd, J.B., 1991.
Mechanism of action of deoxyribonuclease II from human lympho-
blasts. Eur. J. Biochem. 202, 479 – 484.
Hay, B.A., Maile, R., Rubin, G.M., 1997. P element insertion-dependent
gene activation in the Drosophila eye. Proc. Natl. Acad. Sci. U. S. A.
94, 5195 – 5200.
Hedgecock, E.M., Sulston, J.E., Thomson, J.N., 1983. Mutations affecting
programmed cell deaths in the nematode Caenorhabditis elegans. Sci-
ence 220, 1277 – 1279.
Hengartner, M.O., 1997. Cell death. In: Preiss, J.R. (Ed.), C. elegans II.
Cold Spring Harbor Laboratory Press, Plainview, NY, pp. 383 – 415.
Hevelone, J., Hartman, P.S., 1988. An endonuclease from Caenorhabditis
elegans: partial purification and characterization. Biochem. Genet. 26,
447 – 461.
Hewish, D.R., Burgoyne, L.A., 1973. Chromatin sub-structure. The diges-
tion of chromatin DNA at regularly spaced sites by a nuclear deoxy-
ribonuclease. Biochem. Biophys. Res. Commun. 52, 504 – 510.
Hoffmann, J.A., Reichhart, J.M., 2002. Drosophila innate immunity: an
evolutionary perspective. Nat. Immunol. 3, 121 – 126.
Kawane, K., Fukuyama, H., Kondoh, G., Takeda, J., Ohsawa, Y., Uchiya-
ma, Y., Nagata, S., 2001. Requirement of DNase II for definitive eryth-
ropoiesis in the mouse fetal liver. Science 292, 1546 – 1549.
Kawane, K., Fukuyama, H., Yoshida, H., Nagase, H., Ohsawa, Y., Uchiya-
ma, Y., Okada, K., Iida, T., Nagata, S., 2003. Impaired thymic develop-
ment in mouse embryos deficient in apoptotic DNA degradation. Nat.
Immunol. 4, 138 – 144.
Krieser, R.J., Eastman, A., 1998. The cloning and expression of human
deoxyribonuclease: II. A possible role in apoptosis. J. Biol. Chem. 273,
30909 – 30914.
Krieser, R.J., Eastman, A., 2000. Deoxyribonuclease: II. Structure and
chromosomal localization of the murine gene, and comparison with
the genomic structure of the human and three C. elegans homologs.
Gene 252, 155 – 162.
Krieser, R.J., MacLea, K.S., Park, J.P., Eastman, A., 2001. The cloning,
genomic structure, localization, and expression of human deoxyribonu-
clease IIbeta. Gene 269, 205 – 216.
Krieser, R.J., MacLea, K.S., Longnecker, D.S., Fields, J.L., Fiering, S.,
Eastman, A., 2002. Deoxyribonuclease IIalpha is required during the
phagocytic phase of apoptosis and its loss causes perinatal lethality. Cell
Death Differ. 9, 956 – 962.
Laskowski Sr., M., 1967. DNases and their use in the studies of primary
structure of nucleic acids. Adv. Enzymol. Relat. Areas Mol. Biol. 29,
165 – 220.
Lewandoski, M., 2001. Conditional control of gene expression in the
mouse. Nat. Rev. Genet. 2, 743 – 755.
Liao, T.H., 1985. The subunit structure and active site sequence of porcine
spleen deoxyribonuclease. J. Biol. Chem. 260, 10708 – 10713.
Liu, Q.A., Hengartner, M.O., 1998. Candidate adaptor protein CED-6
promotes the engulfment of apoptotic cells in C. elegans. Cell 93,
961 – 972.
Lyon, C.J., 1997. Characterization of a candidate switch endonuclease
activity. Dissertation, Molecular Biology Institute, University of Cali-
fornia, Los Angeles.
Lyon, C.J., Aguilera, R.J., 1997. Purification and characterization of the
immunoglobulin switch sequence-specific endonuclease (Endo-SR)
from bovine spleen. Mol. Immunol. 34, 209 – 219.
Lyon, C.J., Aguilera, R.J., 2002. Endonuclease compositions and methods
of use. US Patent # 6,455,250 issued September 24.
Lyon, C.J., Miranda, G.A., Piao, J.S., Aguilera, R.J., 1996. Characteriza-
tion of an endonuclease activity which preferentially cleaves the G-rich
immunoglobulin switch repeat sequences. Mol. Immunol. 33, 157 – 169.
Lyon, C.J., Evans, C.J., Bill, B.R., Otsuka, A.J., Aguilera, R.J., 2000. The
C. elegans apoptotic nuclease NUC-1 is related in sequence and activity
to mammalian DNase II. Gene 252, 147 – 154.
MacLea, K.S., Krieser, R.J., Eastman, A., 2002. Revised structure of the
active form of human deoxyribonuclease IIalpha. Biochem. Biophys.
Res. Commun. 292, 415 – 421.
MacLea, K.S., Krieser, R.J., Eastman, A., 2003. A family history of deoxy-
ribonuclease: II. Surprises from Trichinella spiralis and Burkholderia
pseudomallei. Gene 305, 1 – 12.
Matsuyama, S., Llopis, J., Deveraux, Q.L., Tsien, R.Y., Reed, J.C., 2000.
Changes in intramitochondrial and cytosolic pH: early events that mod-
ulate caspase activation during apoptosis. Nat. Cell Biol. 2, 318 – 325.
McIlroy, D., Tanaka, M., Sakahira, H., Fukuyama, H., Suzuki, M., Yama-
mura, K., Ohsawa, Y., Uchiyama, Y., Nagata, S., 2000. An auxiliary
mode of apoptotic DNA fragmentation provided by phagocytes. Genes
Dev. 14, 549 – 558.
Mukae, N., Yokoyama, H., Yokokura, T., Sakoyama, Y., Nagata, S., 2002.
Activation of the innate immunity in Drosophila by endogenous chro-
mosomal DNA that escaped apoptotic degradation. Genes Dev. 16,
2662 – 2671.
Nagata, S., 2000. Apoptotic DNA fragmentation. Exp. Cell Res. 256,
12 – 18.
Napirei, M., Karsunky, H., Zevnik, B., Stephan, H., Mannherz, H.G., Mo-
roy, T., 2000. Features of systemic lupus erythematosus in DNase1-
deficient mice. Nat. Genet. 25, 177 – 181.
Odaka, C., Mizuochi, T., 1999. Role of macrophage lysosomal enzymes in
the degradation of nucleosomes of apoptotic cells. J. Immunol. 163,
5346 – 5352.
Oshima, R.G., Price, P.A., 1973. Alkylation of an essential histidine residue
in porcine spleen deoxyribonuclease. J. Biol. Chem. 248, 7522 – 7526.
Parrish, J.Z., Xue, D., 2003. Functional genomic analysis of apoptotic
DNA degradation in C. elegans. Mol. Cell 4, 987 – 996.
Parrish, J., Li, L., Klotz, K., Ledwich, D., Wang, X., Xue, D., 2001.
Mitochondrial endonuclease G is important for apoptosis in C. elegans.
Nature 412, 90 – 94.
Pervaiz, S., Clement, M.V., 2002. A permissive apoptotic environment:
function of a decrease in intracellular superoxide anion and cytosolic
acidification. Biochem. Biophys. Res. Commun. 290, 1145 – 1150.
Rich, T., Allen, R.L., Wyllie, A.H., 2000. Defying death after DNA dam-
age. Nature 407, 777 – 783.
Robson, M.G., Walport, M.J., 2001. Pathogenesis of systemic lupus eryth-
ematosus (SLE). Clin. Exp. Allergy 31, 678 – 685.
Sadahira, Y., Mori, M., 1999. Role of the macrophage in erythropoiesis.
Pathol. Int. 49, 841 – 848.
Shiokawa, D., Tanuma, S., 1998. Cloning of cDNAs encoding porcine and
human DNase II. Biochem. Biophys. Res. Commun. 247, 864 – 869.
Shiokawa, D., Tanuma, S., 1999. DLAD, a novel mammalian divalent
cation-independent endonuclease with homology to DNase II. Nucleic
Acids Res. 27, 4083 – 4089.
Stone, J.C., Dower, N.A., Hauseman, J., Cseko, Y.M., Sederoff, R.,
1983. The characterization of a mutant affecting DNA metabolism
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–1514

in the development of D. melanogaster. Can. J. Genet. Cytol. 25,
129 – 138.
Su, D.M., Wang, J., Lin, Q., Cooper, M.D., Watanabe, T., 1997. Interferons
alpha/beta inhib it IL-7- induced pr oliferation of CD4 CD8 CD3 
CD44 + CD25 + thymocytes , but do not in hibit that of CD4 CD8 
CD3 CD44 CD25 thymocytes. Immunology 90, 543 – 549.
Sulston, J.E., 1976. Post-embryonic development in the ventral cord of
Caenorhabditis elegans. Philos. Trans. R. Soc. Lond., B Biol. Sci.
275, 287 – 297.
Tjelle, T.E., Lovdal, T., Berg, T., 2000. Phagosome dynamics and function.
Bioessays 22, 255 – 263.
Townend, R., Bernardi, G., 1971. Studies on acid deoxyribonuclease: X.
Molecular weight in denaturing solvents. Arch. Biochem. Biophys. 147,
728 – 733.
Vernooy, S.Y., Copeland, J., Ghaboosi, N., Griffin, E.E., Yoo, S.J., Hay,
B.A., 2000. Cell death regulation in Drosophila: conservation of mech-
anism and unique insights. J. Cell Biol. 150, F69 – F76.
Vidalain, P.O., Laine, D., Zaffran, Y., Azocar, O., Servet-Delprat, C., Wild,
T.F., Rabourdin-Combe, C., Valentin, H., 2002. Interferons mediate
terminal differentiation of human cortical thymic epithelial cells. J.
Virol. 76, 6415 – 6424.
Walport, M.J., 2000. Lupus, DNase and defective disposal of cellular deb-
ris. Nat. Genet. 25, 135 – 136.
Wang, C.C., Lu, S.C., Chen, H.L., Liao, T.H., 1998. Porcine spleen deoxy-
ribonuclease: II. Covalent structure, cDNA sequence, molecular clon-
ing, and gene expression. J. Biol. Chem. 273, 17192 – 17198.
Wang, X., Yang, C., Chai, J., Shi, Y., Xue, D., 2002. Mechanisms of AIF-
mediated apoptotic DNA degradation in Caenorhabditis elegans. Sci-
ence 298, 1587 – 1592.
Wu, Y.C., Horvitz, H.R., 1998. The C. elegans cell corpse engulfment
gene ced-7 encodes a protein similar to ABC transporters. Cell 93,
951 – 960.
Wu, Y.C., Stanfield, G.M., Horvitz, H.R., 2000. NUC-1, a Caenorhabditis
elegans DNase II homolog, functions in an intermediate step of DNA
degradation during apoptosis. Genes Dev. 14, 536 – 548.
Wyllie, A.H., 1980. Glucocorticoid-induced thymocyte apoptosis is as-
sociated with endogenous endonuclease activation. Nature 284,
555 – 556.
Wyllie, A., 1998. Apoptosis. An endonuclease at last. Nature 391, 20 – 21.
Yasuda, T., Nadano, D., Sawazaki, K., Kishi, K., 1992. Genetic polymor-
phism of human deoxyribonuclease II (DNase II): low activity levels in
urine and leukocytes are due to an autosomal recessive allele. Ann.
Hum. Genet. 56, 1 – 10.
Yasuda, T., Takeshita, H., Iida, R., Nakajima, T., Hosomi, O., Nakashima,
Y., Kishi, K., 1998a. Molecular cloning of the cDNA encoding human
deoxyribonuclease II. J. Biol. Chem. 273, 2610 – 2616.
Yasuda, T., Takeshita, H., Iida, R., Nakajima, T., Hosomi, O., Nakashima,
Y., Mogi, K., Kishi, K., 1998b. Chromosomal localization of a human
deoxyribonuclease II gene (DNASE2) to 19p13.2 – p13.1 using both the
polymerase chain reaction and fluorescence in situ hybridization anal-
ysis. Biochem. Biophys. Res. Commun. 244, 815 – 818.
Yasutomo, K., Horiuchi, T., Kagami, S., Tsukamoto, H., Hashimura, C.,
Urushihara, M., Kuroda, Y., 2001. Mutation of DNASE1 in people with
systemic lupus erythematosus. Nat. Genet. 28, 313 – 314.
Yokoyama, H., Mukae, N., Sakahira, H., Okawa, K., Iwamatsu, A., Nagata,
S., 2000. A novel activation mechanism of caspase-activated DNase
from Drosophila melanogaster. J. Biol. Chem. 275, 12978 – 12986.
C.J. Evans, R.J. Aguilera / Gene 322 (2003) 1–15 15