Serpin 2a Is Induced in Activated Macrophages and
Conjugates to a Ubiquitin Homolog1
Jessica A. Hamerman,2* Fumitaka Hayashi,* Lea A. Schroeder,*†Steven P. Gygi,3‡
Arthur L. Haas,§Lynne Hampson,¶Paul Coughlin,?Ruedi Aebersold,†‡and Alan Aderem*†
After i.p. infection of mice with the intracellular bacterium Mycobacterium bovis bacillus Calmette-Gue ´rin, macrophages recov-
ered from the peritoneal cavity display classical signs of immune activation. We have identified a member of the serine protease
inhibitor (serpin) family which is highly induced in macrophages during bacillus Calmette-Gue ´rin infection. Serpin 2a (spi2a)
expression is also induced in macrophages in vivo during infection with Salmonella typhimurium and Listeria monocytogenes, and
in vitro by a variety of bacteria and bacterial products. The cytokine IFN-? also induces spi2a expression in macrophages, and
this induction is synergistic with bacterial products. We also demonstrate here that a ubiquitin homolog, IFN-stimulated gene of
15-kDa (ISG15), is strongly induced during in vitro and in vivo activation of macrophages and that it conjugates to spi2a in
activated macrophages. The ISG15-spi2a conjugates were identified by tandem mass spectrometry and contained spi2a conjugated
to either one or two molecules of ISG15. Whereas spi2a was induced by either bacterial products or IFN-?, ISG15 was induced
only by bacterial products. Although many protein targets have been described for ubiquitin conjugation, spi2a is the first
ISG15-modified protein to be reported. Macrophage activation is accompanied by the activation of a variety of proteases. It is of
interest that a member of the serine protease inhibitor family is concomitantly induced and modified by a ubiquitin-like
protein. The Journal of Immunology, 2002, 168: 2415–2423.
plex interplay between elements of the innate and adaptive im-
mune system (1, 2). At least three criteria must be met for the
macrophage to efficiently kill the bacterium and activate the adap-
tive immune system by a process known as Ag presentation. Mac-
rophages must increase the production of bactericidal agents in-
cluding reactive nitrogen and reactive oxygen intermediates, up-
regulate a variety of proteases that can degrade the dead bacteria,
and express higher levels of surface MHC class II molecules that
can present the resultant bacterial peptides to T cells (reviewed in
Refs. 1 and 3). Although activated macrophages are very efficient
at killing bacteria, the task of Ag presentation falls largely on the
related dendritic cells (4).
The activation of macrophages and dendritic cells in vivo is
complex (5). Using cDNA arrays, we investigated the changes in
gene expression associated with macrophage activation during in
acrophages play a critical role in host defense against
bacterial pathogens. Their bactericidal activity is in-
duced by bacterial products and is a result of a com-
vivo infection of mice with the intracellular bacterium Mycobac-
terium bovis bacillus Calmette-Gue ´rin (BCG),4a classical system
for studying immune macrophage activation. In this screen, we
identified serpin 2a (spi2a) as a protein with substantially increased
expression during BCG infection in vivo (6, 7). Serpins are a pro-
tein superfamily with conserved structure that regulate both serine
and cysteine protease function in diverse processes including co-
agulation, extracellular matrix degradation, complement activa-
tion, fibrinolysis, and apoptosis (8, 9). We report here that spi2a is
increased ?100-fold not only during in vivo activation of macro-
phages by BCG but also during infection of mice with Listeria
monocytogenes and Salmonella typhimurium. In vitro, bacteria and
bacterial products, as well as the cytokine IFN-?, induce the spi2a
promoter, and the combination of these macrophage activators is
synergistic for spi2a induction. spi2a is also induced in dendritic
cells by bacterial products. Our data suggest that spi2a regulates
intracellular proteases in activated APC.
While studying the expression of spi2a in macrophages by
Western blotting, we detected spi2a not only at its predicted mo-
lecular mass but also in more slowly migrating forms. These spe-
cies proved to be spi2a conjugated to a ubiquitin homolog known
as IFN-stimulated gene of 15-kDa (ISG15) (10). ISG15, also
known as ubiquitin cross-reactive protein, is induced by type I
IFNs in a variety of cell types (11–13) and conjugates to intracel-
lular proteins in a process analogous to that for ubiquitin (14). We
report here the first identification of a substrate for ISG15 conju-
gation. ISG15 conjugation to spi2a occurs in macrophages that
have been activated by incubation with bacterial products. Addi-
tionally, in vitro incubation with LPS as well as in vivo infection
*Department of Immunology, University of Washington, Seattle, WA 98185;†Insti-
tute for Systems Biology, Seattle, WA 98105;‡Department of Biotechnology, Uni-
versity of Washington, Seattle, WA 98195;§Department of Biochemistry, Medical
College of Wisconsin, Milwaukee, WI 53226;¶Department of Obstetrics and Gynae-
cology, St. Mary’s Hospital, University of Manchester, Manchester, United Kingdom;
and?Department of Medicine, Monash University, Melbourne, Australia
Received for publication November 20, 2001. Accepted for publication January
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1This work was supported in part by National Institutes of Health Grants AI-25032
and AI-32972 (to A.A.). J.A.H. was a predoctoral fellow of the Howard Hughes
2Address correspondence and reprint requests to Dr. Jessica A. Hamerman at the
current address: Department of Microbiology and Immunology, Box 0414, University
of California, San Francisco, CA 94143. E-mail address: email@example.com
3Current address: Department of Cell Biology, Harvard Medical School, Boston, MA
4Abbreviations used in this paper: BCG, Mycobacterium bovis bacillus Calmette-
Gue ´rin; spi2a, serpin 2a; ISG15, IFN-stimulated gene of 15-kDa; His tag, histidine
tag; HPRT, hypoxanthine phosphoribosyltransferase; AraLAM, lipoarabinomannan;
mAGP, mycolylarabinogalactan-peptidoglycan complex; iNOS, inducible NO syn-
thase; SCWP, soluble cell wall proteins.
Copyright © 2002 by The American Association of Immunologists 0022-1767/02/$02.00
with BCG potently induces the expression of both spi2a and ISG15
Materials and Methods
Cells and mice
RAW264.7 macrophages (American Type Culture Collection (ATCC),
Manassas VA) were cultured in RPMI (BioWhittaker, Gaithersburg, MD)
with 10% FCS (HyClone Laboratories, Logan, UT), glutamine, penicillin,
and streptomycin (Life Technologies, Gaithersburg, MD) at 37?C and 5%
CO2. Female ICR and C57BL/6 mice were purchased from Charles River
Breeding Laboratories (Wilmington, MA).
Bacteria and infection
Mycobacterium bovis BCG (strain Pasteur; ATCC) was a gift from Dr. S.
Smith (University of Washington, Seattle, WA). BCG was grown in
Proskauer-Beck medium with aeration to 5 ? 107CFU/ml and stored in
aliquots at ?70?C. To infect mice, an aliquot of BCG was thawed, soni-
cated three times for 30 s in a water bath sonicator, and diluted in PBS.
Mice were injected i.p. with 5 ? 106CFU. Listeria monocytogenes, strain
10403S, was grown in trypticase soy broth (Difco, Detroit, MI). Listeria in
log phase were diluted in PBS, and 1 ? 103CFU were injected i.p. S.
typhimurium SL3261 (attenuated DL1344 ?aroAhisGsylrpsL) was a gift
from Dr. B. Cookson (University of Washington). Salmonella growing at
log phase in LB (Difco) were diluted in PBS, and 1 ? 105CFU/mouse
were injected i.p.
Macrophages and dendritic cells
Activated macrophages were harvested by peritoneal lavage with PBS 12
days after infection with BCG and 5 days after infection with Listeria or
Salmonella. Resident peritoneal macrophages were harvested from unin-
fected ICR mice. Unless otherwise noted, macrophages were plate adhered
for 2 h and then washed several times with PBS to remove nonadherent
cells. Cells remaining were ?95% macrophages by visual inspection. Bone
marrow dendritic cells were generated by a modification of the method of
Inaba et al. (15). Briefly, bone marrow cells from C57BL/6 mice were
cultured with 20 ng/ml rGM-CSF (R&D Systems, Minneapolis, MN) for 7
days. Dendritic cells were then cultured in medium or in medium with 100
ng/ml LPS (List Biological Laboratories, Campbell, CA) for 24 h; stained
with Abs to CD11b, CD11c, and IAb; and purified by sorting on a FACS-
Vantage (BD Biosciences, San Jose, CA). Unstimulated dendritic cells
were sorted as CD11b?CD11c?IAb?cells, whereas LPS-stimulated den-
dritic cells were sorted as CD11b?CD11c?IAb?cells, using class II MHC
up-regulation as a marker of activation.
Macrophages from day 12 after BCG infection were lysed in Trizol (Life
Technologies), and then total RNA was isolated according to manufactur-
er’s instructions. mRNA was then purified using two rounds of oligo(dT)
cellulose columns (Pharmacia, Piscataway, NJ); 5 ?g of this mRNA were
used as a template to generate an unamplified oligo(dT)-primed cDNA
library in the pSPORT plasmid vector according to the manufacturer’s
directions (Life Technologies). Then 9200 individual clones in E. coli were
grown and stored in 384-well plates. These clones were then spotted using
a Q-bot onto 20- ? 20-cm nylon membranes in duplicate, with the central
spot of every grid of 9 containing a plasmid with a control cDNA. The
colonies were lysed on the membrane using proteinase K, and the plasmid
DNA was denatured, neutralized, and cross-linked to the membrane using
a Stratalinker (Stratagene, La Jolla, CA). Duplicate membranes were
probed with32P-labeled first strand cDNA generated from mRNA from
resident peritoneal or BCG macrophages primed with a mixture of
oligo(dT) and random hexamer primers. Control mRNA was added to the
labeling reactions to control for labeling and hybridization efficiency be-
tween probes and membranes, respectively. Hybridization was detected
using a Molecular Dynamics (Sunnyvale, CA) PhosphorImager, and the
location and intensity of individual spots were determined using SFV soft-
ware provided by Dr. R. Baumgartner. Average background for the mem-
brane was subtracted from the spot intensities, and intensities were nor-
malized for average control spot intensity. Duplicate spots on each
membrane for each clone were averaged, and data were discarded if the
duplicate intensities varied by ?50%. Clones that met this criterion and
that showed a ?2-fold difference in intensity were sequenced. Sequences
were analyzed by searching the nonredundant and EST databases at Gen-
Bank using the Advanced Blast search algorithm.
Total RNA from resident peritoneal and BCG-activated macrophages was
isolated as above. For in vitro activated macrophages, resident peritoneal
macrophages were plate adhered overnight, nonadherent cells were re-
moved, and cells were activated with 100 ng/ml LPS or 10 U/ml IFN-?
(R&D Systems). After 4 h (LPS) or 48 h (IFN-?), cells were lysed, and
RNA was prepared as above. Northern blot analysis of 10 ?g total RNA
was conducted using standard methods. The Northern blot was probed with
a random primed, [32P]dCTP (New England Nuclear, Boston, MA)-labeled
probe corresponding to the entire cDNA for spi2a or for EF1? as a
Expression constructs and transfection
All expression constructs were in the EF6/V5-His-TOPO vector (Invitro-
gen, San Diego, CA). The spi2a-HA construct was generated by amplifying
the spi2a open reading frame using a forward primer encoding an HA tag
N-terminal to the cDNA start (forward primer, CGGAATTCATGTAC
reverse primer, CGGGATCCTCACTGTCCAATCAGGCATAG) with the
cDNA array clone as template. The spi2a-HisHA expression construct was
generated with the identical reverse primer to the spi2a-HA construct and with
a forward primer encoding both the 6-residue histidine tag (His tag) and the
HA tag (forward primer, ATGCATCATCACCATCACCATTACCCATAC
GACGTCCCAGACTACGCTGCTGGTGTCTCCCCTGCTGTC). To gen-
erate the ISG15-V5 construct, the open reading frame of ISG15 was am-
plified from cDNA made from total RNA of RAW264.7 cells treated for
4 h with 100 ng/ml LPS (List) using a forward primer encoding a V5 tag
N-terminal to the start codon (forward primer, ATGGGTAAGCCTATC
GGTG; reverse primer, TTAGGCACACTGGTCCCCTCC). All constructs
were verified by sequence analysis. Ten micrograms DNA were transiently
transfected into between 10 and 50 ? 106RAW264.7 cells by electropo-
ration (16). Cells were plate adhered overnight and used for subsequent
experiments. To generate stable clones, transient transfectants were se-
lected in medium with 5 ?g/ml blasticidin (Invitrogen) for 10 days and
then cloned by limiting dilution.
Approximately 5 ? 105transiently transfected cells or stable clones were
plated in wells of 24-well dishes. In some cases, cells were treated over-
night with 100 ng/ml Salmonella minnesota LPS (List). The cells were
lysed in 10 mM HEPES, pH 7.4, with 150 mM NaCl and 1% Triton X-100
containing leupeptin (1 ?M, Boehringer Mannheim, Indianapolis, IN),
aprotinin (1/100, Sigma, St. Louis, MO), and PMSF (1 mM, Boehringer
Mannheim). The extracts were spun at 15,000 rpm for 15 min to remove
nuclei, and then the supernatant removed for analysis by SDS-PAGE. Res-
ident peritoneal macrophages were plated at ?1 ? 106macrophages per
well in 24-well tissue culture dishes. After overnight adherence, nonad-
herent cells were washed away, and then cells were cultured with medium
alone or medium with LPS at 100 ng/ml or IFN-? at 10 U/ml (R&D Sys-
tems). At the indicated times after activation, the cells were lysed, and
cytoplasmic extracts generated as above. For the experiment involving in
vivo activated macrophages, macrophages from uninfected mice or those
from BCG-infected mice were plate adhered for 2 h before removal of
nonadherent cells and lysis. Extracts were prepared as above. A 25-?g
protein sample was used per lane for this experiment. For detection of the
tagged proteins, the anti-HA.11 mAb (Covance, Princeton, NJ) or the anti-
PK mAb to the V5 tag (Serotech, Raleigh, NC) was used as suggested by
manufacturer with anti-mouse HRP (Zymed Laboratories, San Francisco,
CA) and detected using ECL Plus (Amersham, Arlington, Heights, IL). To
detect ISG15, Western blots were probed with affinity purified rabbit poly-
clonal antiserum to ISG15 (14) used at 1 ?g/ml and anti-rabbit HRP
(Zymed Laboratories). To detect spi2a, Western blots were probed with
affinity-purified rabbit polyclonal antiserum generated to recombinant
spi2a (E. C. Morris, T. Dafforn, S. L. Forsyth, A. J. Horvath, L. Hampson,
I. N. Hampson, R. W. Carrell and P. B. Coughlin, manuscript in prepara-
tion) used at 1/1000 dilution followed by anti-rabbit HRP.
Approximately 1 ?g total RNA from macrophage populations was reverse
transcribed using Superscript II reverse transcriptase and oligo(dT) primers
(Life Technologies). Serial 1/3 dilutions of cDNA were amplified with
primers to murine hypoxanthine phosphoribosyltransferase (HPRT) to
standardize between cDNA samples (forward primer, GATACAGGC
CAGACTTTGTTG; reverse primer, GGTAGGCTGGCCTATAGGCT).
Matched 3-fold dilutions of cDNA from each sample were then amplified
2416 REGULATION OF spi2a DURING MACROPHAGE ACTIVATION
with primers to spi2a (forward primer, GGAATGGCAGGTGTCGGATG;
reverse primer, GGTCAGGAACCTGATTTCGTC). These primers were
chosen to minimize cross-hybridization with other serpins that may be
expressed in macrophages; the forward primer encompasses the reactive
site loop of spi2a, and the reverse primer is in the 3?-UTR. Amplified
products were separated on 1% agarose gels and visualized with ethidium
Real time PCR
cDNA from dendritic cells purified by cell sorting was generated as above
and amplified with probe and primer sets for murine HPRT and spi2a as
indicated below using TaqMan Universal PCR master mix and an ABI
Prism 7700 Sequence Detector (Applied Biosystems, Foster City, CA) fol-
lowing the manufacturer’s instructions. Genomic DNA contamination was
measured by including template that had been mock reverse transcribed
and at all times accounted for ?10% of the signal. Fold induction was
determined from Ctvalues normalized for HPRT expression and then nor-
malized to the value derived from medium-treated dendritic cells. Primers
and probes used were as follows. Spi2a: forward, CCAAATGGTGAGG
GTGCTTCT; reverse, GCATAGCGGATCACCAAAACA; probe, CCCA
ACGGCTGGAATCTAAGCGTTTAT. HPRT primers: forward, TGGAA
AGAATGTCTTGATTGTTGAA; reverse, AGCTTGCAACCTTAACCA
TTTTG; probe, CAAACTTTGCTTTCCCTGGTTAAGCAGTACAGC.
RAW 264.7 cells (1 ? 107) were transiently transfected by electroporation
as above with 10 ?g pXP-2 plasmid (17) containing ?259 to ?46 nucle-
otides of the spi2a promoter driving the firefly luciferase gene (18) and
plated into one 96-well dish. Cells were adhered overnight, washed once
with PBS, and treated as indicated with IFN-? at 10 U/ml (R&D Systems)
or medium for 8 h. Bacteria or their products were then added for 14 h at
the indicated final concentration. All stimuli except LPS and S. minnesota
were treated with polymyxin B (10 ?g/ml) for 1 h before addition to cells
to ensure that the data did not result from LPS contamination. Cells were
lysed, and luciferase activity was read using the Luciferase Reporter Assay
System (Promega, Madison, WI) according to the manufacturer’s instruc-
tions. Heat-killed Mycobacterium tuberculosis H37Rv (100 ?g/ml) and M.
tuberculosis products lipoarabinomannan (AraLAM; 10 ?g/ml), mycoly-
larabinogalactan-peptidoglycan complex (mAGP; 100 ?g/ml), and soluble
cell wall proteins (SCWP; 1 ?g/ml) were obtained from J. Belisle (National
Institute of Allergy and Infectious Diseases, Bethesda, MD) through the
TB Research Materials and Vaccine Testing Contract. Heat-killed Staph-
ylococcus aureus (clinical isolate) was used at 2 ? 106bacteria/well, and
heat-killed S. minnnesota R595 (ATCC) was used at 2 ? 103bacteria per
well. Zymosan (3 ? 106particles/well) was obtained from Molecular
Probes (Eugene, OR), S. minnnesota LPS (100 ng/ml) was obtained from
List, Staphylococcus aureus peptidoglycan (10 ?g/ml) was from Fluka
(Buchs, Switzerland), and the synthetic lipopeptide PAM3CSK4(100 ng/
ml) was from Boehringer Mannheim.
Purification of spi2a-HisHA and identification of complexed
For affinity purification, CNBr-activated Sepharose beads conjugated to the
HA.11 mAb at 1 mg Ab/ml beads following the manufacturer’s instruc-
tions (Pharmacia) were used followed by Ni2?beads (Invitrogen). Cyto-
plasmic extracts generated in 20 mM phosphate buffer, pH 7.8, with 150
mM NaCl, 1% Triton X-100 (lysis buffer), and protease inhibitors as above
were pooled from 2 ? 109RAW264.7 macrophages transiently transfected
with the spi2a-HisHA construct. The extracts were incubated with HA.11
beads for 1 h at 4°C with rocking. The HA.11 beads were washed three
times with lysis buffer without Triton X-100 and eluted by boiling in this
buffer with 0.3% SDS to prevent precipitation of proteins. This eluate was
then incubated with Ni2?beads at room temperature for 1 h, washed five
times in 20 mM phosphate buffer, pH 7.8, with 150 mM NaCl and then
three times with 20 mM phosphate buffer, pH 6.0, with 150 mM NaCl. The
protein was eluted by boiling in SDS sample buffer containing 2-ME. Nine-
ty-five percent of the eluted protein was run in one lane of a 10% SDS-
PAGE gel for silver stain, and the remaining 5% was run in one lane for
Western blotting with the HA.11 Ab. Silver staining was performed as
To prepare samples for tandem mass spectrometry, silver-stained bands
were excised from the gel, cut into ?1-mm cubes, and subjected to in-gel
tryptic digests (19). Peptides generated were identified by nanoscale mi-
crocapillary liquid chromatography-tandem mass spectrometry techniques
using an LCQ Classic ion trap mass spectrometer (ThermoFinnigan, San
Jose, CA) (20). Spectra were searched against the OWL nonredundant
protein sequence database as well as the EST databases using the program
SEQUEST (21), which matches theoretical and acquired tandem mass
Spi2a is up-regulated in macrophages during bacterial infection
We used cDNA arrays to identify a large number of genes that are
specifically induced in murine peritoneal macrophages during in
vivo infection with BCG (data not shown). One of these genes
encoded serpin 2a (spi2a), a member of the serine protease inhib-
itor (serpin) family. By Northern blot analysis, mRNA for spi2a
was not detected in resident peritoneal macrophages from unin-
fected mice but was detected strongly in macrophages from mice
infected for 12 days with BCG (Fig. 1A). The strong induction in
BCG-activated macrophages was also seen at the protein level;
Fig. 1B shows a Western blot of macrophage lysates probed with
polyclonal antiserum recognizing spi2a. The specificity of the
polyclonal antiserum to spi2a is also shown in Fig. 1B. The poly-
clonal serum detects HA-tagged spi2a when transfected into the
bacterial infection. A, Top, Northern blot analysis of total RNA from res-
ident peritoneal macrophages from uninfected mice (RPM?) and perito-
neal macrophages from mice infected 12 days previously with BCG
(BCGM?) probed with the complete cDNA for spi2a; bottom, the same
blot stripped and reprobed for the housekeeping gene EF1?. B, Top, West-
ern blot analysis of spi2a expression in cytoplasmic extracts from perito-
neal macrophages as in A or from vector control (control) and HA-tagged
spi2a (spi2a-NHA) expression construct-transfected RAW264.7 macro-
phages. spi2a protein is detected using polyclonal antiserum generated to
recombinant spi2a protein. Bottom, The same blot stripped and reprobed
with a mAb to the HA epitope tag. The slower mobility of the transfected
spi2a in comparison with endogenous spi2a is due to the addition of the HA
epitope tag. C, Semiquantitative RT-PCR analysis of cDNA generated
from total RNA from resident peritoneal macrophages, macrophages iso-
lated 5 days after Salmonella and Listeria infection, and 12 days after BCG
infection. Three-fold serial dilutions of cDNA were matched for HPRT
expression (bottom) and then amplified with primers for spi2a (top). PCR
products were separated on agarose gels and detected by ethidium bromide
Spi2a is up-regulated in macrophages induced in vivo by
2417The Journal of Immunology
macrophage cell line, RAW264.7, but not in vector control-trans-
fected cells, which do not express any endogenous spi2a mRNA
(Fig. 1B and data not shown). The band detected by the polyclonal
serum is identical with that detected by an anti-HA Ab on the same
blot after stripping (Fig. 1B).
We investigated whether induction of spi2a was unique to in-
fection with Mycobacterium or whether infection with other intra-
cellular pathogens would also induce this transcript. Mice were
infected i.p. with L. monocytogenes, a Gram-positive bacterium, or
an attenuated strain of S. typhimurium, a Gram-negative bacterium,
and peritoneal macrophages were isolated after 5 days. spi2a
mRNA levels were assessed by semiquantitative RT-PCR and
compared with resident peritoneal macrophages and BCG macro-
phages. Fig. 1C shows that both infection with S. typhimurium and
infection with L. monocytogenes induced spi2a mRNA. This sug-
gests that induction of spi2a is a general response of macrophages
to in vivo infection with intracellular bacteria and that it is not
specific to mycobacterial infections.
spi2a is induced in vitro by bacterial products and IFN-?
We assessed whether the induction of spi2a was in response to the
bacteria directly or indirectly via a cytokine induced in vivo by
bacterial infection. To address this, we initially performed North-
ern blot analysis on total RNA from resident peritoneal macro-
phages that were treated in vitro with LPS, a molecule from the
coat of Gram-negative bacteria, or with IFN-?, a cytokine that is
secreted by NK cells and T cells in response to bacterial infection.
These molecules are both potent activators of macrophages al-
though they act on distinct pathways. Both LPS treatment for 4 h
and IFN-? treatment for 48 h induced spi2a mRNA in resident
peritoneal macrophages to a greater extent than culture alone (Fig.
2A). We also investigated whether spi2a is similarly regulated in
myeloid dendritic cells, a cell highly related to the macrophage.
Real time PCR analysis showed that spi2a mRNA is also induced
by LPS treatment in bone marrow-derived dendritic cells, a stim-
ulation that also induces maturation of these cells and up-regula-
tion of cell surface class II MHC levels (Fig. 2B). These data
suggest that spi2a is regulated similarly in macrophages and den-
The finding that LPS, a Gram-negative bacterial product, can
activate transcription of spi2a in vitro led us to ask whether other
classes of bacteria and their products can also induce spi2a. To do
so, we used the RAW264.7 macrophage cell line transfected with
a plasmid containing base pairs ?259 to ?46 of the spi2a pro-
moter driving the firefly luciferase gene as a reporter (18). LPS
treatment of RAW264.7 cells induces spi2a mRNA as is seen in
resident peritoneal macrophages (data not shown). A variety of
whole bacteria as well as purified bacterial cell wall components
and yeast cell walls directly induced the spi2a promoter in vitro
(Fig. 2C). This induction varied between 2.5-fold (araLAM) and
4.6-fold (zymosan) over basal levels. As found with in vivo infec-
tion, Gram-negative (S. minnesota) and Gram-positive (Staphylo-
coccus aureus) bacteria as well as Mycobacterium (M. tuberculo-
sis) activated this promoter. The bacterial products tested were
derived from Gram-negative bacteria (LPS), Gram-positive bacte-
ria (peptidoglycan) and Mycobacterium (araLAM, MAGP,
SCWP). The doses used in this experiment have been previously
shown to induce maximal production of TNF-? from RAW264.7
cells (data not shown), and all stimuli except LPS and Salmonella
were treated with polymyxin B to ensure the induction did not
result from LPS contamination.
IFN-? is known to sensitize macrophages to respond to bac-
terial stimuli. Therefore, we were interested in how the pres-
ence of IFN-? affected the ability of macrophages to induce
spi2a in response to these stimuli. To test this, we pretreated
RAW264.7 macrophages transfected with the spi2a promoter
luciferase construct with IFN-? and then activated with bacteria
treatment with bacterial products and IFN-?. A, Northern blot analysis of res-
ident peritoneal macrophages that had been plate adhered (medium) or acti-
vated with LPS for 4 h (LPS 4 h) or IFN-? for 48 h (IFN-? 48 h). Spi2a and
EF1? transcripts were detected as in Fig. 1A. B, Real time PCR analysis of
spi2a expression in bone marrow-derived dendritic cells. Dendritic cells from
a 7-day culture of GM-CSF-treated bone marrow were either cultured in me-
that were either IAb?(medium) or IAb?(LPS). cDNA from these populations
was subjected to real time PCR using primer/probe sets specific for spi2a and
HPRT and normalized to HPRT levels. Data are represented as the fold in-
duction of spi2a over medium-treated cells and are the average of triplicate
bases ?259 to ?46 of the spi2a promoter driving the firefly luciferase gene.
The cells were stimulated with bacteria or their products as indicated for 14 h
and lysed, and the luciferase activity was measured. The amount of each stim-
uli used is listed in Materials and Methods. The activity of the promoter is
represented as the fold induction over untreated cells (none) on the y-axis. D,
Cells were transfected and treated as in B but were first pretreated with IFN-?
(10 U/ml) for 8 h before the addition of stimuli. The data in C and D are
ylococcus aureus; MTb, M. tuberculosis.
spi2a is induced in macrophages and dendritic cells by in vitro
2418 REGULATION OF spi2a DURING MACROPHAGE ACTIVATION
or bacterial components as in Fig. 2C before assaying luciferase
activity. As shown in Fig. 2D, IFN-? alone induces the spi2a
promoter in this system as it does in primary macrophages (Fig.
2A). The induction by IFN-? is 16-fold in the RAW264.7 sys-
tem, which is greater than the extent of induction by bacterial
products alone seen in Fig. 2C. Interestingly, pretreatment with
IFN-? caused a large increase in the induction of the spi2a
promoter by the bacteria and their components (Fig. 2D). This
induction ranged from 27-fold (MAGP) to 68-fold (LPS) over
background and was much greater than the additive effects of
IFN-? and the bacteria/components alone. These results were
confirmed when looking at protein production by primary mac-
rophages treated in vitro with LPS either with or without IFN-?
pretreatment for 14 h (data not shown). Therefore, as with other
proteins important in the macrophage response to pathogens
such as inducible NO synthase (iNOS) and IL-12, there is pro-
found induction of both spi2a mRNA and protein by a combi-
nation of IFN-? and bacterial products.
Identification of spi2a-ISG15 conjugates in activated
While studying the expression of spi2a protein in the RAW264.7
macrophage cell line, we observed that higher molecular mass
forms of spi2a were detected (Fig. 3A). An N-terminally HA-
tagged version of spi2a (spi2a-HA) was transiently expressed in
the RAW264.7 macrophage cell line, cytoplasmic extracts were
separated by SDS-PAGE, and spi2a was detected with a mAb to
the HA tag. The majority of the spi2a-HA was present in a band
with the electrophoretic mobility predicted for the tagged protein,
50 kDa (Fig. 3A). Slower migrating forms of ?65 and 80 kDa
were also detected. The 42-kDa band corresponds to a spi2a-HA
To identify the protein components of the 65- and 80-kDa com-
plexes, we purified spi2a from transiently transfected RAW264.7
cells using a version of spi2a with both an N-terminal HA and a
six-histidine tag (spi2a-HisHA). Spi2a-HisHA was first affinity pu-
rified from transiently transfected RAW264.7 cells with Sepharose
beads conjugated to a mAb to the HA tag, eluted, and then affinity
purified with Ni2?beads that bind the His tag. The protein was
eluted from the Ni2?beads, separated by SDS-PAGE, and silver
stained (Fig. 3B). Two of the silver-stained bands that matched
precisely with the 65- and 80-kDa forms of spi2a detected by
Western blotting (Fig. 3B, arrows) were cut out, and the protein
content was identified by tandem mass spectrometry. The 80-kDa
band contained only two proteins, spi2a and murine ISG15, also
known as ubiquitin cross-reactive protein; Fig. 3C shows the se-
quence of ISG15 with the three independently identified peptides
boxed. The ISG15 peptides identified covered 30.6% of the protein
sequence. The 65-kDa band also contained spi2a and ISG15, as
well as an apparent contaminant, TCP-1?, a 66-kDa chaperone.
ISG15 is a 15-kDa protein that contains two ubiquitin homology
domains. It has been shown to conjugate to cellular proteins in a
process analogous to that for ubiquitin, using homologous, but
distinct, enzymes (22–24). Thus, the 65-kDa form of spi2a likely
represents spi2a (50 kDa) covalently bound to one molecule of
ISG15, whereas the 80-kDa form represents spi2a bound to two
molecules of ISG15. The other bands seen on the silver-stained gel
contained serum proteins, including serum albumin (69 kDa) and
?2-macroglobulin (165 kDa), and were most likely derived from
the FCS in the tissue culture medium.
To confirm that the spi2a 65 and 80 kDa bands contained ISG15,
we affinity purified spi2a-HisHA from RAW264.7 cells and
probed Western blots with Abs to the HA tag or to endogenous
ISG15. Polyclonal Abs raised against purified human ISG15 iden-
tified the 65-kDa and 80-kDa bands that also reacted with the
anti-HA Ab (Fig. 4) but did not label the 50-kDa band that con-
tained the unconjugated spi2a.
We next coexpressed spi2a-HA and V5-tagged ISG15 (ISG15-
V5) in RAW264.7 cells and assayed for the presence of ISG15-
V5-spi2a conjugates. The 65- and 80-kDa spi2a conjugates were
detected when spi2a-HA was transfected alone or cotransfected
with the ISG15-V5 construct (Fig. 5, top). As expected, the con-
jugates were slightly larger in the cells cotransfected with the
ISG15 containing the 14-aa V5 tag. A replicate blot was probed
with Abs to the V5 tag present on ISG15. When ISG15-V5 was
transfected alone, we detected a ladder of bands that presumably
corresponds to ISG15 conjugated to a variety of cellular proteins,
including spi2a (Fig. 5, middle). When spi2a-HA and ISG15-V5
were coexpressed, ISG15 was found in four principal bands, two
taining ISG15 in RAW264.7 macrophages. A, Cytoplasmic extract from
RAW264.7 cells transiently transfected with an expression construct for
spi2a-HA was analyzed by Western blot after separation by SDS-PAGE.
spi2a expression was detected using a mAb to the HA tag. B, spi2a protein
was affinity purified from RAW264.7 cells transiently transfected with
spi2a-HisHA using Sepharose beads conjugated to an anti-HA mAb fol-
lowed by Ni2?beads to bind the six-His tag. The eluate from the Ni2?
beads was separated by SDS-PAGE and either silver stained (left) or West-
ern blotted (right) with detection by the anti-HA Ab. The two bands des-
ignated with arrows were identified by both Western blot and silver stain.
The two bands running at 65 and 80 kDa (arrows) were cut out of the gel,
digested with trypsin, and subjected to tandem mass spectrometry to de-
termine the peptide sequence. C, Amino acid sequence of ISG15 with the
three tryptic peptides identified by mass spectrometry of the 80-kDa band
(upper arrow, B) shown in the boxes.
spi2a is present in higher molecular mass conjugates con-
2419The Journal of Immunology
of which corresponded to the spi2a complexes detected with the
HA Ab. It is striking that overexpression of spi2a along with
ISG15 caused the redistribution of ISG15 into four bands as op-
posed to the numerous bands seen when ISG15-V5 was expressed
alone. This may reflect a rate-limiting step in the conjugation pro-
cess exposed by overexpression of both proteins. The lower panel
in Fig. 5 shows that unconjugated ISG15-V5 was not limiting re-
gardless of whether spi2a was overexpressed.
ISG15 expression is induced during macrophage activation
Macrophages are strongly activated by bacterial DNA, and tran-
sient transfection using bacterial vectors results in the enhanced
expression of many proteins associated with the activated state
(25). We therefore examined the conjugation of spi2a and ISG15
in stably transfected macrophages, where the transient activation
induced by the bacterial plasmid DNA has subsided. Surprisingly,
stable clones of RAW264.7 cells expressing spi2a-HA did not con-
tain the 65- and 80-kDa spi2a-ISG15 conjugates that were present
in transiently transfected cells (Fig. 6A). However, LPS induced
these conjugates in the macrophages (Fig. 6A), implying that LPS
had induced the expression of ISG15, which could now conjugate
with stably transfected spi2a. This was confirmed by Western blot
analysis demonstrating that ISG15 was essentially undetectable in
untreated RAW264.7 cells and was very strongly induced after
LPS stimulation (Fig. 6B). The hypothesis that bacterial DNA was
inducing ISG15 during transient transfections was similarly con-
firmed (Fig. 6B).
Although the RAW 264.7 cell line is useful for transfection
studies, it was important to demonstrate the regulation of ISG15 in
primary macrophages, both in vitro and in vivo. LPS induced
ISG15 in resident peritoneal macrophages; the protein was de-
tected within 1 h and reached maximal levels of expression 4 h
after stimulation with LPS (Fig. 7A). LPS also induced the expres-
sion of spi2a in resident peritoneal macrophages with similar ki-
netics to ISG15. Interestingly, although IFN-? alone strongly in-
duced spi2a, it had no effect on the expression of ISG15 (Fig. 7A),
demonstrating that the regulation of these two proteins could be
We identified spi2a as a protein induced in response to macro-
phage activation by BCG infection in vivo (Fig. 1), and we there-
fore investigated whether ISG15 is also induced under these con-
ditions. Infection of mice i.p. with BCG resulted in the very strong
induction of both spi2a and ISG15 in macrophages in vivo (Fig.
7B). This demonstrates that the induction of ISG15 and spi2a in
vitro accurately reflects the more complex process of macrophage
activation during bacterial infection in vivo.
Using cDNA array analysis, we have identified spi2a as a protein
that is induced in macrophages activated during infection with in-
tracellular bacteria. The cDNA for spi2a was originally cloned
from a chondrocyte cell line (6) and subsequently demonstrated to
be expressed in hemopoietic progenitor cells (7). Interestingly,
Hampson et al. (7) found spi2a mRNA to be down-regulated dur-
ing granulocyte macrophage differentiation of a multipotential he-
mopoietic progenitor cell line and during macrophage differenti-
ation of bipotential granulocyte/macrophage precursor cells
isolated from mouse bone marrow. They found that spi2a mRNA
was absent from mature macrophages differentiated from the pro-
genitor cell line, which is consistent with our results showing that
spi2a mRNA and protein are undetectable in unstimulated resident
Up-regulation of spi2a appears to be a general response of mac-
rophages to bacterial infection both in vivo and in vitro. In addition
to BCG, both Gram-positive (L. monocytogenes) and Gram-
negative (S. typhimurium) infections in vivo induced macrophages
that had up-regulated spi2a mRNA. These bacteria, along with
M. bovis BCG, used to initially identify spi2a, are all intracellular
pathogens that can live within macrophages in the infected host.
Extracellular pathogens, including yeast (zymosan) as well as the
Gram-positive bacterium Staphylococcus aureus, were also able to
activate the spi2a promoter in vitro in RAW264.7 macrophages.
Indeed, the spi2a promoter was induced by a variety of pathogens,
and their products, that are known to activate macrophages for
production of proteins important in the antimicrobial immune re-
sponse, such as TNF-? and IL-12. This suggests that induction of
spi2a is a general response of macrophages to infection with patho-
gens and is not specific to mycobacterial infections.
spi2a mRNA and protein are induced in vitro not only by bac-
terial products but also by the cytokine IFN-?. This cytokine, pro-
duced by activated natural killer cells and T cells, is critical for a
tein from RAW264.7 cells transiently transfected with vector alone or vec-
tor encoding spi2a-HisHA was affinity purified as in Fig. 3 and separated
by SDS-PAGE; Western blots were probed with affinity-purified poly-
clonal serum to ISG15 (right). This blot was then stripped and reprobed
with an anti-HA mAb (left).
ISG15 Abs detect the 65- and 80-kDa forms of spi2a. Pro-
cells were transiently transfected with vector alone, vector encoding spi2a-
HA, and/or ISG15-V5. Cytoplasmic extracts were separated by SDS-
PAGE, and Western blots were probed with anti-HA (top) or anti-V5 (mid-
dle and bottom) mAbs. The bottom panel is a shorter run of the middle
panel to show free ISG15-V5 protein.
Epitope-tagged ISG15 conjugates to spi2a. RAW264.7
2420REGULATION OF spi2a DURING MACROPHAGE ACTIVATION
successful immune response to intracellular pathogens, including
those used in this study (2, 26). During activation of macrophages,
IFN-? is essential for the induction of bactericidal mechanisms
including stimulation of reactive oxygen intermediate production
and induction of iNOS for the generation of reactive nitrogen in-
termediate (3, 27, 28). IFN-? also induces the expression of class
II MHC molecules that allow activated macrophages to present Ag
to CD4 T cells (2, 27). Pretreatment of macrophages with IFN-?
sensitized them for induction of the spi2a promoter by bacterial
products and for the production of spi2a protein by LPS. This
activation of the spi2a promoter by IFN-? and bacteria, or their
products, is clearly more than additive. This is reminiscent of the
induction of iNOS and the IL-12 p40 subunit in macrophages; both
are slightly induced by LPS, but the combination of LPS with
IFN-? potently induces these proteins (29–32). The high levels of
spi2a mRNA and protein in macrophages from BCG-infected mice
may be due to the fact that both IFN-? and bacteria and their
products are available to activate macrophages in this setting.
Both bacterial products and IFN-? regulated spi2a at the tran-
scriptional level. The induction of the spi2a promoter by LPS and
other bacterial products is consistent with the presence of a con-
sensus NF-?B binding site in the promoter region (?259 to ?46)
used to drive the luciferase reporter in these experiments. Hamp-
son et al. (18) found that this NF-?B binding site was critical for
maximal induction of the spi2a promoter in a multipotential he-
mopoietic progenitor cell line and in primary murine splenocytes,
presumably measuring expression in T cells. NF-?B translocation
to the nucleus is a well-documented consequence of macrophage
activation through Toll-like receptors, which have been shown to
signal downstream of bacteria and their products (33). Interest-
ingly, the spi2a promoter used in these studies also contains a
predicted STAT binding site consensus sequence at bases ?128 to
?137 (J. A. Hamerman, L. Hampson, and A. Aderem, unpublished
observations). This STAT binding site may explain the respon-
siveness of the spi2a promoter to IFN-?. spi2a can be added to the
list of other IFN-?-responsive genes important in macrophage
function during infection.
The role of spi2a in activated macrophages is not yet clear.
Serpins are a protein superfamily with conserved structure and
have been demonstrated to regulate serine and cysteine protease
function both extracellularly and intracellularly (8). Serpins have
been shown to participate in diverse processes mediated by pro-
teases including complement activation, coagulation, fibrinolysis,
extracellular matrix degradation, and apoptosis (9). Although some
members of the serpin family, such as OVA and angiotensin, are
not functional protease inhibitors (8), spi2a possesses the serpin
proximal hinge motif indicative of a functional inhibitor (34) (J. A.
Hamerman and A. Aderem, unpublished observations). We there-
fore propose that spi2a is a functional protease inhibitor that reg-
ulates protease activity in activated macrophages and that this ac-
tivity is involved in the function of macrophages during infection
with intracellular bacteria. Interestingly, spi2a is up-regulated in
activated CD8?T cells as well as activated macrophages (7) and
therefore during infection may be expressed in both cell types.
spi2a may play a similar role in both or may regulate different
proteases in the each cell type.
Although the majority of well-characterized serpins are se-
creted, some function intracellularly (8). spi2a is primarily ex-
pressed intracellularly by several criteria. spi2a lacks an N-termi-
nal signal sequence, it shows diffuse cytoplasmic staining by
immunofluorescence in LPS-treated resident peritoneal macro-
phages, and it cannot be detected in supernatants from LPS-treated
resident peritoneal macrophages or stably transfected RAW264.7
cells, whereas cell lysates are strongly positive (J. A. Hamerman,
L. A. Schroeder, and A. Aderem, unpublished observations). Sev-
eral cytoplasmic serpins have been shown to regulate apoptosis, a
process dependent on proteolytic cascades. This includes inhibi-
tion of caspase 1 by the cowpox virus serpin crmA (35) and the
inhibition of granzyme B by the human serpin PI-9 (36). Other
fection and LPS treatment. A, Cytoplasmic extracts from RAW264.7 cells
either transiently transfected with the spi2a-HA construct or stable clones
made with vector alone or the spi2a-HA construct were analyzed as in Fig.
1A. The stable clones were either untreated or treated overnight with 100
ng/ml LPS. B, Cytoplasmic extracts from RAW264.7 cells that were un-
treated, transiently transfected with empty vector (?transfection), or
treated with 100 ng/ml LPS overnight were analyzed by Western blot with
affinity-purified polyclonal Abs to ISG15.
Spi2a-ISG15 conjugates and ISG15 are induced by trans-
vated primary murine macrophages. A, Cytoplasmic extracts were gener-
ated from resident peritoneal macrophages that were treated with 100
ng/ml LPS or 10 U/ml IFN-? from 0 to 24 h as indicated. ISG15 was
detected as in Fig. 4 and spi2a was detected as in Fig. 1B. B, Cyto-
plasmic extracts from peritoneal macrophages from uninfected mice
(RPM?) or mice infected 12 days previously with BCG (BCG?) were
analyzed as in A.
ISG15 and spi2a are induced in in vitro and in vivo acti-
2421The Journal of Immunology
cytoplasmic serpins have been shown to protect cells from their
own proteases. PI-6 has been shown to inhibit cathepsin G, a neu-
trophil granule protease (37). Presumably, the cytoplasmic PI-6
protects the neutrophil from granule rupture before release from
the cell. PI-9, described above, is expressed by cytotoxic T cells
and NK cells, which also express the PI-9 target, granzyme B, in
their cytotoxic granules (36). It has been proposed by Bird et al.
(38) that PI-9 protects T cells and NK cells against misdirected
granzyme B after degranulation or leakage of granzyme B from
cytotoxic granules within the cell. spi2a may have functions anal-
ogous to those of these well-characterized serpins. Macrophages
have an extensive lysosomal system containing a variety of pro-
teases that are up-regulated in response to IFN-? treatment (39–
41). This enables the macrophage to degrade bacteria and other
ingested pathogens once they have been killed. The potential re-
lease of lysosomal enzymes into the cytoplasm could result in
macrophage cell death, and therefore a mechanism involving spi2a
may exist to protect against this risk.
We have also demonstrated that spi2a forms conjugates with
ISG15, a ubiquitin homolog. Two complexes are detected, 65- and
80-kDa conjugates. This suggests that the 50-kDa spi2a is com-
plexed with either one or two molecules of the 15-kDa ISG15.
ISG15 is a member of a small family of proteins that demonstrate
significant sequence similarity to ubiquitin and that covalently
modify other cellular proteins (42). It contains two ubiquitin-like
domains with 43 and 62% homology to ubiquitin (10, 14). The
mechanism of conjugation of ISG15 to cellular substrates has been
proposed to be analogous to that for ubiquitin involving homolo-
gous but not identical enzymes (22, 23, 43, 44). Ubiquitin can form
polyubiquitin chains from one lysine residue in a target protein,
and it is possible that the spi2a-ISG15 complex with two ISG15
molecules is the result of a di-ISG15 chain. Alternatively, this
80-kDa complex may reflect conjugation of a single ISG15 mol-
ecule to two distinct sites on spi2a. The pattern of ISG15 conju-
gates within cells is distinct from that of ubiquitin-modified pro-
teins, suggesting that ubiquitin and ISG15 have different target
proteins (14). Although many protein targets have been described
for ubiquitin conjugation, spi2a is the first ISG15-modified protein
to be reported.
ISG15 and spi2a are both induced during in vitro and in vivo
macrophage activation, although their induction in this process can
be uncoupled. Thus, whereas spi2a is profoundly induced by either
LPS or IFN-?, ISG15 is induced only by bacterial products such as
LPS or bacterial DNA, which both signal through Toll-like recep-
tors (33). ISG15 conjugation to spi2a correlates with induction of
these two proteins in activated macrophages. Previously, ISG15
had been shown to be induced by type 1 IFNs, and this induction
was attributed to the presence of an IFN-stimulated response ele-
ment in the ISG15 promoter (45). Interestingly, LPS induces not
only ISG15 expression but also production of IFN-? from macro-
phages (46). It is therefore possible that LPS-induced IFN-? acts
in an autocrine manner on macrophages resulting in ISG15 pro-
duction, although LPS could also induce ISG15 directly. However,
we demonstrate here that ISG15 protein can be detected in resident
peritoneal macrophages after 1 h of LPS treatment, suggesting that
the effect of LPS is most likely direct. This does not preclude that
the higher levels seen at later times are not due, in part, to auto-
crine IFN-? production.
The functional consequences of spi2a conjugation to ISG15 are
unclear. Although ubiquitin modification targets proteins for deg-
radation via the proteasome, there is no evidence that ISG15 has
this function. Indeed, treatment of spi2a-transfected RAW264.7
cells with proteasome inhibitors had no effect on the accumulation
of spi2a-ISG15 conjugates (J. A. Hamerman, L. A. Schroeder and
A. Aderem, unpublished observations), whereas this accumulates
ubiquitin-protein conjugates (47, 48). These conditions have also
been shown to have no effect on the half-life of total ISG15 con-
jugates in a lung carcinoma cell line while producing significant
effects on total protein degradation (J. Narasimhan and A. Haas,
unpublished observations). Loeb and Haas (24) have reported that
ISG15 conjugates colocalize with intermediate filaments in a lung
carcinoma cell line, but we have not been able to see this in res-
ident peritoneal macrophages or in RAW264.7 cells (J. A. Hamer-
man, L. A. Schroeder, and A. Aderem, unpublished observations).
We have also not detected spi2a colocalized with intermediate fil-
aments in these cells. Despite this negative result, it is known that
LPS regulates intermediate filaments. It induces the reorganization
of the vimentin network into bundles in microglia and fibroblasts
(49, 50), whereas IFN-?, which is induced by LPS treatment, stim-
ulates transcription of the vimentin gene in epithelial cells (51).
ISG15 conjugation to spi2a may target spi2a to intermediate fila-
ments allowing for regulation of protease activity at this site by
this serpin. Interestingly, it has recently been shown that the in-
fluenza B virus NS1 protein inhibits conjugation of ISG15 to cel-
lular proteins (44). This suggests that ISG15 conjugation is an
effective part of the host response to viral infection, because patho-
gens often target pathways that decrease their ability to survive and
In summary, we have identified both spi2a and ISG15 as pro-
teins which are induced in activated macrophages and that phys-
ically interact in these cells. spi2a expression is up-regulated by
interactions with a variety of bacterial pathogens and their prod-
ucts as well as by IFN-?, an abundant cytokine during bacterial
infection; whereas ISG15 is induced by bacterial products, but not
IFN-?. This is the first demonstration of a target for ISG15 con-
jugation, and its identification will help advance the understanding
of both the biochemical mechanism by which ISG15 conjugation
occurs and the functional consequences of ISG15 modification of
We thank Drs. Peter Nelson and Leroy Hood for help with the cDNA
arrays, Kathy Allen for cell sorting, Robert Alaniz for bone marrow-de-
rived dendritic cells, Drs. Adrian Ozinsky and Ananda Goldrath for critical
review of the manuscript, and Dr. Kelly Smith for helpful discussions
regarding the data. We are also grateful to Dr. Sherilyn Smith for providing
M. bovis BCG and Dr. Brad Cookson for providing S. typhimurium.
1. Unanue, E. R. 1997. Inter-relationship among macrophages, natural killer cells
and neutrophils in early stages of Listeria resistance. Curr. Opin. Immunol. 9:35.
2. Trinchieri, G. 1997. Cytokines acting on or secreted by macrophages during
intracellular infection (IL-10, IL-12, IFN-?). Curr. Opin. Immunol. 9:17.
3. Nathan, C., and M. U. Shiloh. 2000. Reactive oxygen and nitrogen intermediates
in the relationship between mammalian hosts and microbial pathogens. Proc.
Natl. Acad. Sci. USA 97:8841.
4. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of
immunity. Nature 392:245.
5. Hamerman, J. A., and A. Aderem. 2001. Functional transitions in macrophages
during in vivo infection with Mycobacterium bovis bacillus Calmette-Gue ´rin.
J. Immunol. 167:2227.
6. Inglis, J. D., M. Lee, D. R. Davidson, and R. E. Hill. 1991. Isolation of two
cDNAs encoding novel ?1-antichymotrypsin-like proteins in a murine chondro-
cytic cell line. Gene 106:213.
7. Hampson, I. N., L. Hampson, M. Pinkoski, M. Cross, C. M. Heyworth,
R. C. Bleackley, E. Atkinson, and T. M. Dexter. 1997. Identification of a serpin
specifically expressed in multipotent and bipotent hematopoietic progenitor cells
and in activated T cells. Blood 89:108.
8. Potempa, J., E. Korzus, and J. Travis. 1994. The serpin superfamily of proteinase
inhibitors: structure, function, and regulation. J. Biol. Chem. 269:15957.
9. Wright, H. T. 1996. The structural puzzle of how serpin serine proteinase inhib-
itors work. Bioessays 18:453.
10. Haas, A. L., P. Ahrens, P. M. Bright, and H. Ankel. 1987. Interferon induces a
15-kilodalton protein exhibiting marked homology to ubiquitin. J. Biol. Chem.
2422 REGULATION OF spi2a DURING MACROPHAGE ACTIVATION
11. Farrell, P. J., R. J. Broeze, and P. Lengyel. 1979. Accumulation of an mRNA and
protein in interferon-treated Ehrlich ascites tumour cells. Nature 279:523.
12. Knight, E., Jr., D. Fahey, B. Cordova, M. Hillman, R. Kutny, N. Reich, and
D. Blomstrom. 1988. A 15-kDa interferon-induced protein is derived by COOH-
terminal processing of a 17-kDa precursor [published erratum appears in 1988
J. Biol. Chem. 263:10040]. J. Biol. Chem. 263:4520.
13. Korant, B. D., D. C. Blomstrom, G. J. Jonak, and E. Knight, Jr. 1984. Interferon-
induced proteins: purification and characterization of a 15,000-dalton protein
from human and bovine cells induced by interferon. J. Biol. Chem. 259:14835.
14. Loeb, K. R., and A. L. Haas. 1992. The interferon-inducible 15-kDa ubiquitin
homolog conjugates to intracellular proteins. J. Biol. Chem. 267:7806.
15. Inaba, K., M. Inaba, N. Romani, H. Aya, M. Deguchi, S. Ikehara, S. Muramatsu,
and R. M. Steinman. 1992. Generation of large numbers of dendritic cells from
mouse bone marrow cultures supplemented with granulocyte/macrophage colo-
ny-stimulating factor. J. Exp. Med. 176:1693.
16. Gold, E. S., D. M. Underhill, N. S. Morrissette, J. Guo, M. A. McNiven, and
A. Aderem. 1999. Dynamin 2 is required for phagocytosis in macrophages.
J. Exp. Med. 190:1849.
17. Nordeen, S. K. 1988. Luciferase reporter gene vectors for analysis of promoters
and enhancers. BioTechniques 6:454.
18. Hampson, L., I. N. Hampson, C. K. Babichuk, L. Cotter, R. C. Bleackley,
T. M. Dexter, and M. A. Cross. 2001. A minimal serpin promoter with high
activity in haemopoietic progenitors and activated T cells. Hematol. J. 2:150.
19. Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric
sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 68:850.
20. Gygi, S. P., Y. Rochon, B. R. Franza, and R. Aebersold. 1999. Correlation be-
tween protein and mRNA abundance in yeast. Mol. Cell. Biol. 19:1720.
21. Eng, J., A. McCormack, and J. R. Yates. 1994. An approach to correlate tandem
mass spectral data of peptides with amino acid sequences in a protein database.
J. Am. Soc. Mass Spectrom. 5:976.
22. Narasimhan, J., J. L. Potter, and A. L. Haas. 1996. Conjugation of the 15-kDa
interferon-induced ubiquitin homolog is distinct from that of ubiquitin. J. Biol.
23. Potter, J. L., J. Narasimhan, L. Mende-Mueller, and A. L. Haas. 1999. Precursor
processing of pro-ISG15/UCRP, an interferon-?-induced ubiquitin-like protein.
J. Biol. Chem. 274:25061.
24. Loeb, K. R., and A. L. Haas. 1994. Conjugates of ubiquitin cross-reactive protein
distribute in a cytoskeletal pattern. Mol. Cell. Biol. 14:8408.
25. Aderem, A., and D. A. Hume. 2000. How do you see CG? Cell 103:993.
26. Jouanguy, E., R. Doffinger, S. Dupuis, A. Pallier, F. Altare, and J. L. Casanova.
1999. IL-12 and IFN-? in host defense against mycobacteria and Salmonella in
mice and men. Curr. Opin. Immunol. 11:346.
27. Dalton, D. K., M.-S. Pitts, S. Keshav, I. S. Figari, A. Bradley, and T. A. Stewart.
1993. Multiple defects of immune cell function in mice with disrupted interfer-
on-? genes. Science 259:1739.
28. Boehm, U., T. Klamp, M. Groot, and J. C. Howard. 1997. Cellular responses to
interferon-?. Annu. Rev. Immunol. 15:749.
29. Ma, X., J. M. Chow, G. Gri, G. Carra, F. Gerosa, S. F. Wolf, R. Dzialo, and
G. Trinchieri. 1996. The interleukin 12 p40 gene promoter is primed by interferon
? in monocytic cells. J. Exp. Med. 183:147.
30. Hayes, M. P., J. Wang, and M. A. Norcross. 1995. Regulation of interleukin-12
expression in human monocytes: selective priming by interferon-? of lipopo-
lysaccharide-inducible p35 and p40 genes. Blood 86:646.
31. Ding, A. H., C. F. Nathan, and D. J. Stuehr. 1988. Release of reactive nitrogen
intermediates and reactive oxygen intermediates from mouse peritoneal macro-
phages. Comparison of activating cytokines and evidence for independent pro-
duction. J. Immunol. 141:2407.
32. Lorsbach, R. B., W. J. Murphy, C. J. Lowenstein, S. H. Snyder, and
S. W. Russell. 1993. Expression of the nitric oxide synthase gene in mouse
macrophages activated for tumor cell killing. Molecular basis for the synergy
between interferon-? and lipopolysaccharide. J. Biol. Chem. 268:1908.
33. Aderem, A., and R. J. Ulevitch. 2000. Toll-like receptors in the induction of the
innate immune response. Nature 406:782.
34. Hopkins, P. C., and J. Whisstock. 1994. Function of maspin. Science 265:1893.
35. Ray, C. A., R. A. Black, S. R. Kronheim, T. A. Greenstreet, P. R. Sleath,
G. S. Salvesen, and D. J. Pickup. 1992. Viral inhibition of inflammation: cowpox
virus encodes an inhibitor of the interleukin-1? converting enzyme. Cell 69:597.
36. Sun, J., C. H. Bird, V. Sutton, L. McDonald, P. B. Coughlin, T. A. De Jong,
J. A. Trapani, and P. I. Bird. 1996. A cytosolic granzyme B inhibitor related to
the viral apoptotic regulator cytokine response modifier A is present in cytotoxic
lymphocytes. J. Biol. Chem. 271:27802.
37. Scott, F. L., C. E. Hirst, J. Sun, C. H. Bird, S. P. Bottomley, and P. I. Bird. 1999.
The intracellular serpin proteinase inhibitor 6 is expressed in monocytes and
granulocytes and is a potent inhibitor of the azurophilic granule protease, cathep-
sin G. Blood 93:2089.
38. Bird, C. H., V. R. Sutton, J. Sun, C. E. Hirst, A. Novak, S. Kumar, J. A. Trapani,
and P. I. Bird. 1998. Selective regulation of apoptosis: the cytotoxic lymphocyte
serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis
without perturbing the Fas cell death pathway. Mol. Cell. Biol. 18:6387.
39. Lafuse, W. P., D. Brown, L. Castle, and B. S. Zwilling. 1995. IFN-? increases
cathepsin H mRNA levels in mouse macrophages. J. Leukocyte Biol. 57:663.
40. Lah, T. T., M. Hawley, K. L. Rock, and A. L. Goldberg. 1995. ?-Interferon
causes a selective induction of the lysosomal proteases, cathepsins B and L, in
macrophages. FEBS Lett. 363:85.
41. Lesser, M., J. C. Chang, and M. Orlowski. 1985. Cathepsin B and D activity in
stimulated peritoneal macrophages. Mol. Cell. Biochem. 69:67.
42. Yeh, E. T., L. Gong, and T. Kamitani. 2000. Ubiquitin-like proteins: new wines
in new bottles. Gene 248:1.
43. Hershko, A., and A. Ciechanover. 1998. The ubiquitin system. Annu. Rev. Bio-
44. Yuan, W., and R. M. Krug. 2001. Influenza B virus NS1 protein inhibits conju-
gation of the interferon (IFN)-induced ubiquitin-like ISG15 protein. EMBO J.
45. Kessler, D. S., D. E. Levy, and J. E. Darnell, Jr. 1988. Two interferon-induced
nuclear factors bind a single promoter element in interferon-stimulated genes.
Proc. Natl. Acad. Sci. USA 85:8521.
46. Gessani, S., F. Belardelli, A. Pecorelli, P. Puddu, and C. Baglioni. 1989. Bacterial
lipopolysaccharide and gamma interferon induce transcription of ? interferon
mRNA and interferon secretion in murine macrophages. J. Virol. 63:2785.
47. Ward, C. L., S. Omura, and R. R. Kopito. 1995. Degradation of CFTR by the
ubiquitin-proteasome pathway. Cell 83:121.
48. Leitch, V., P. Agre, and L. S. King. 2001. Altered ubiquitination and stability of
aquaporin-1 in hypertonic stress. Proc. Natl. Acad. Sci. USA 98:2894.
49. Chakravortty, D., and K. S. Nanda Kumar. 2000. Bacterial lipopolysaccharide
induces cytoskeletal rearrangement in small intestinal lamina propria fibroblasts:
actin assembly is essential for lipopolysaccharide signaling. Biochim. Biophys.
50. abd-el-Basset, E., and S. Fedoroff. 1995. Effect of bacterial wall lipopolysaccha-
ride (LPS) on morphology, motility, and cytoskeletal organization of microglia in
cultures. J. Neurosci. Res. 41:222.
51. Alldridge, L. C., M. K. O’Farrell, and G. B. Dealtry. 1989. Interferon ? increases
expression of vimentin at the messenger RNA and protein levels in differentiated
embryonal carcinoma (PSMB) cells. Exp. Cell Res. 185:387.
2423The Journal of Immunology