COX-3, a cyclooxygenase-1 variant inhibited by

acetaminophen and other analgesic/antipyretic

drugs: Cloning, structure, and expression
N. V. Chandrasekharan, Hu Dai, K. Lamar Turepu Roos, Nathan K. Evanson, Joshua Tomsik, Terry S. Elton,
and Daniel L. Simmons*
Department of Chemistry and Biochemistry, E280 Benson Science Building, Brigham Young University, Provo, UT 84602

Communicated by John Vane, William Harvey Foundation, London, United Kingdom, August 5, 2002 (received for review April 17, 2002)

Two cyclooxygenase isozymes, COX-1 and -2, are known to cata- Materials and Methods
lyze the rate-limiting step of prostaglandin synthesis and are the Unless otherwise stated all basic protocols used were from the
targets of nonsteroidal antiinflammatory drugs. Here we describe manual on molecular cloning by Sambrook and Russell (3).
a third distinct COX isozyme, COX-3, as well as two smaller
COX-1-derived proteins (partial COX-1 or PCOX-1 proteins). COX-3 Isolation of RNA and Construction of a cDNA Library. Isolation of
and one of the PCOX-1 proteins (PCOX-1a) are made from the RNA and library construction methods have been described (4).
COX-1 gene but retain intron 1 in their mRNAs. PCOX-1 proteins Human Multiple Tissue Northern blots (MTN) were purchased
additionally contain an in-frame deletion of exons 5– 8 of the from CLONTECH.
COX-1 mRNA. COX-3 and PCOX mRNAs are expressed in canine Antisense oligonucleotides to the first intron of human and
cerebral cortex and in lesser amounts in other tissues analyzed. In canine COX-1 genes were synthesized and end-labeled using
human, COX-3 mRNA is expressed as an ⬇5.2-kb transcript and is [␥-32P]dATP.
most abundant in cerebral cortex and heart. Intron 1 is conserved A canine cerebral cortex cDNA library was screened using an
in length and in sequence in mammalian COX-1 genes. This intron ⬇1.0-kb canine COX-1 fragment previously cloned in this
contains an ORF that introduces an insertion of 30 –34 aa, depend- laboratory by reverse transcription-coupled (RT)-PCR. The
ing on the mammalian species, into the hydrophobic signal peptide library was also screened with a 32P-labeled canine COX-1 intron
that directs COX-1 into the lumen of the endoplasmic reticulum and 1 antisense oligonucleotide. Two full-length clones were isolated,
nuclear envelope. COX-3 and PCOX-1a are expressed efficiently in completely sequenced, and designated COX-3 and partial
insect cells as membrane-bound proteins. The signal peptide is not COX-1a (PCOX-1a). Both were derived from the canine COX-1
cleaved from either protein and both proteins are glycosylated. gene but retain intron 1. PCOX-1a also has a 657-bp in-frame
COX-3, but not PCOX-1a, possesses glycosylation-dependent cy- deletion spanning exons 5–8.
clooxygenase activity. Comparison of canine COX-3 activity with
murine COX-1 and -2 demonstrates that this enzyme is selectively RT-PCR of Canine and Human Cerebral Cortex mRNA. Canine cerebral
inhibited by analgesic/antipyretic drugs such as acetaminophen,
cortex cDNA was synthesized, and primers were designed for
phenacetin, antipyrine, and dipyrone, and is potently inhibited by
PCR amplification. The sense primer (5⬘-CGGATCCGCCGC-
some nonsteroidal antiinflammatory drugs. Thus, inhibition of
CCAGAGCTATGAG-3⬘) corresponded to nucleotides 15–32 of
COX-3 could represent a primary central mechanism by which these
canine COX-3 sequence (submitted to GenBank under acces-
drugs decrease pain and possibly fever.
sion no. AF535138), with the 3⬘ end of the primer being two
nucleotides downstream of the initiating methionine. The anti-

A cetaminophen is often categorized as a nonsteroidal anti-

inflammatory drug (NSAID), even though in clinical prac-
tice and in animal models it possesses little antiinflammatory
sense primer (5⬘-cgccatcctggtgggggtcaggcacacgga-3⬘) corre-
sponded to nucleotides 1865–1894, located 32 nucleotides up-
stream of the stop codon.
activity (1). Like NSAIDs, however, acetaminophen inhibits Northern blot analysis of human tissues with an intron 1 probe
pain and fever and is one of the world’s most popular analgesic/ detected an ⬇5.2-kb mRNA similar in size to one previously
antipyretic drugs. Despite acetaminophen’s long use and popu- reported (5). Marathon-ready human cerebral cortex cDNA
larity it lacks a clear mechanism of action. Flower and Vane (CLONTECH) was amplified by PCR (CLONTECH, Advan-
showed that acetaminophen inhibited cyclooxygenase (COX) tage 2 PCR enzyme system), using 5⬘ and 3⬘ primers, and an
activity in dog brain homogenates more than in homogenates ⬇4.2-kb amplified fragment was recovered and found to contain
from spleen (2). This gave rise to the concept that variants of the entire coding region of human COX-1 with intron 1 retained.
COX enzymes exist that are differentially sensitive to this drug
and that acetaminophen acts centrally. Yet, even though two Expression of COX-3 and PCOX-1a in Baculovirus. Both COX-3 and
isozymes of COX are known, neither isozyme is sensitive to PCOX-1a were cloned into the baculovirus expression vector
acetaminophen at therapeutic concentrations of the drug in pBlueBac 4.5/V5-His (Invitrogen). Sf9 cells (⬇1 ⫻ 106) were
whole cells or homogenates. Instead, COX-1 and -2 in homog- infected with viral stocks at a multiplicity of infection (moi) of
enates frequently exhibit the paradoxical property of being 3 for expression of COX-3, PCOX-1a, mouse COX-1, and mouse
stimulated by submillimolar concentrations of acetaminophen COX-2 (6).
and inhibited by supermillimolar levels of the drug (1). This In some cases, tunicamycin was added to a final concentration
finding suggests that neither isozyme is a good candidate for the
site of action of acetaminophen.
In analyzing COX-1 and -2 RNA expression in dog tissues, our Abbreviations: COX, cyclooxygenase; NSAID, nonsteroidal antiinflammatory drug; PCOX-1,
laboratory observed that the cerebral cortex of dog brain partial COX-1; RT, reverse transcription-coupled.
contains two distinct RNAs that hybridized to a canine COX-1 Data deposition: The sequences reported in this paper have been deposited in the GenBank
cDNA. One RNA was ⬇2.6 kb in size and the other was ⬇1.9 kb database (accession nos. AF535138 and AF535139).
in size, and analyses of these RNAs suggest that they encode See commentary on page 13371.
previously uncharacterized COX-1-related proteins. *To whom correspondence should be addressed. E-mail: dan㛭[email protected].

13926 –13931 兩 PNAS 兩 October 15, 2002 兩 vol. 99 兩 no. 21 www.pnas.org兾cgi兾doi兾10.1073兾pnas.162468699

of 10 ␮g/ml to insect cells 1 h after infection, and cells were
cultured and harvested after 48 h. Activity of intact cells was
determined by RIA (7).

Detection of 60-, 53-, and 50-kDa COX-1-Related Proteins in Human

Aorta Tissue. Total protein (20 ␮g) from human aorta was
analyzed by Western blotting, using COX-1 mAb (Cayman
Chemical, Ann Arbor, MI) and COX-3 antipeptide polyclonal
antibodies (pAb). Primary antibodies were either preincubated
with a mixture of human and mouse COX-1 intron 1 peptide
(described below) for 1 h at 4°C, or left unblocked. Blots were
processed with appropriate rabbit-anti-mouse secondary anti-
body (1:2,000) or goat-anti-rabbit secondary antibody (1:10,000)
from Sigma. Densitometry of the autoradiographic image was
performed using the AlphaImage 2000 Documentation and
Analysis System (Alpha Innotech, San Leandro, CA).

Drug Inhibition Assays. Sf9 cells were infected with high titer viral
stocks (moi ⫽ 3) and cultured for 48 h. Cells were preincubated
with drug for 30 min at 25°C, arachidonic acid (100 ␮l, final
concentration 5 or 30 ␮M) was then added for an additional
10-min incubation at 37°C. Supernatant was assayed for COX
activity by RIA for PGE2. Assays were performed multiple times Fig. 1. Northern blot analysis and RT-PCR. (A) Northern blot of canine
in triplicate. Inhibition curves were constructed and IC50 values cerebral cortex poly(A) RNA (1, 5.0 ␮g; 2, 2.5 ␮g) probed with (1) 32P-labeled
were determined using PRISM 3.0 (GraphPad, San Diego). canine COX-1 cDNA fragment and (2) 32P-labeled canine antisense oligonu-
cleotide to intron 1. (B) PCR amplification of PCOX-1 in canine cerebral cortex.
Production of Polyclonal Anti-COX-3 Antibodies. Peptides corre- Lane 1, ethidium bromide-stained gel of amplified products corresponding to
sponding to the first 13 aa of human (MSRECDPGARWGC) PCOX-1a containing intron 1 (upper band) and PCOX-1b (lower band) lacking
intron 1; lane 2, Southern blot of the amplified products probed with anti-
and mouse (MSREFDPEAPRNC) COX-3, as predicted by
sense oligonucleotide to intron 1; lane 3, Southern blot using COX-3 cDNA as
genomic clone sequences, were synthesized and coupled to probe. (C) Human Multiple Tissue Northern blots (MTN) probed with a 32P-
keyhole limpet hemocyanin. Peptides (a 50:50 mixture) were labeled human antisense oligonucleotide to intron 1 (HCI). The ⬇5.2-kb mRNA
injected into New Zealand White rabbits. The resulting poly- was detected in blots 1–3 (adult tissues), and 4 (fetal tissues). Am, amygdala;
clonal antibodies were then affinity purified using the above B, brain; C, cerebellum: Cc, cerebral cortex; Fl, frontal lobe; Hi, hippocampus;
peptides immobilized on a Sulfolink coupling gel (Pierce) ac- Ht, heart; K, kidney; Lu, lung; Li, liver; M, skeletal muscle; Md, medulla; Cn,
cording to the manufacturer’s instructions. caudate nucleus; Op, occipital pole; P, placenta; Pn, pancreas; Pu, putamen; Sc,
spinal cord; Th, thalamus; Tl, temporal lobe; Co, corpus callosum.
Results
Two distinct mRNA species (⬇2.6 and ⬇1.9 kb) were detected
on a Northern blot with a canine COX-1 coding region cDNA is simpler, especially if further variants are discovered. The COX
probe using RNA isolated from canine cerebral cortex (Fig. 1A, cDNA clone that harbored intron 1, lacked exons 5–8, and
lane 1). To further investigate these transcripts, a canine cerebral corresponded to the ⬇1.9-kb mRNA transcript has been desig-
cortex cDNA library was constructed and the nonamplified nated partial COX-1a or PCOX-1a (Fig. 2).
library was screened as described above. Eleven clones were RT-PCR of canine cerebral cortex RNA, as well as analysis of
isolated and subsequently characterized by automated DNA Northern blots, indicated that COX-3 mRNA is present in this
sequencing. All of the eleven clones were found to contain brain region at about 5% of the level of COX-1 mRNA (Fig. 1 A
canine COX-1 cDNA sequence. However, three clones harbored and data not shown). Interestingly, these analyses also demon-
an insertion of 90 nucleotides at, or near, the 5⬘ end of their strated that the ⬇1.9-kb mRNA corresponding to PCOX-1a was

PHARMACOLOGY
respective cDNAs, which showed 75% sequence identity to
intron 1 of either human or mouse COX-1 genes (data not
shown). This extra sequence also contained 5⬘ and 3⬘ consensus
splice sites indicative of a retained intron. In addition to the
retention of intron 1, one of the three clones had a 657-bp
in-frame deletion corresponding to exons 5–8 of the COX-1
message (sequence submitted to GenBank under accession no.
AF535139).
To determine whether the two detected COX mRNA tran-
scripts (i.e., ⬇2.6 and ⬇1.9 kb) harbored intron 1, the Northern
blot experiment was repeated using a radiolabeled antisense
canine COX-1 intron 1-specific oligonucleotide probe (Fig. 1 A,
lane 2). Both the ⬇1.9-kb and the ⬇2.6-kb transcripts were
detected, suggesting that multiple intron-1-containing splice
variants were indeed expressed in canine cerebral cortex. The
COX encoded by the ⬇2.6-kb cDNA clone with nonspliced
Fig. 2. Structure of COX-3 and PCOX-1a. A schematic representation of the
intron 1 has been designated as COX-3. We have named this domains of COX-3 and PCOX-1 in comparison to COX-1. s, signal peptide; d1,
newly discovered enzyme COX-3, even though it derives from dimerization domain/EGF-like domain 1; d2, dimerization domain 2; m, mem-
the same gene as COX-1, because experience has shown that the brane binding domain; c, catalytic domain; i, 90-bp sequence encoded by
important differences between COX-1 and -2 are pharmacolog- intron 1. PCOX-1b is identical to PCOX-1a except that PCOX-1b lacks intron 1.
ical rather than genetic. Furthermore a numerical nomenclature Amino acid numbering is according to residues in sheep seminal vesicle COX-1.

Chandrasekharan et al. PNAS 兩 October 15, 2002 兩 vol. 99 兩 no. 21 兩 13927

Fig. 3. Expression in insect cells. Western blots showing the expression of
COX-3, PCOX-1a, and COX-1 in insect cells treated with (⫹) and without (⫺)
tunicamycin (Upper). Arrows indicate glycosylated forms of COX-1 that are
not present in cells treated with tunicamycin. Polyclonal antibodies to human
and mouse COX-1 intron 1 sequence were used to probe the COX-3 and
PCOX-1a blots, whereas a monoclonal antibody to ovine COX-1 (Cayman
Chemical) was used to probe the mouse COX-1 blot. COX activity in insect cells Fig. 4. Western blot of human aorta lysate probed with COX-1 and -3
expressing COX-3, PCOX-1a, and COX-1 (Lower). Cells were treated with (⫹) antibodies. (A) Western blot (lanes 3– 8, 20 ␮g total aorta protein each lane)
and without (⫺) tunicamycin. probed with primary, secondary, or blocked antibodies as indicated. A solid
horizontal arrow indicates the 65-kDa protein, an open arrow indicates the
53-kDa proteins, and an upward diagonal solid arrow indicates the 50-kDa
protein. A single asterisk denotes unglycosylated canine COX-3, and a double
actually a mixture of two mRNAs that differed in size by ⬇90
asterisk denotes unglycosylated canine PCOX-1a. (B) Densitometric analysis of
nucleotides (Fig. 1B). One of these mRNAs was PCOX-1a and 65-, 53-, and 50-kDa proteins. Percent relative densitometric units (% rdu)
the other (PCOX-1b) was identical to PCOX-1a except that it were calculated by comparison to the signal from unblocked primary anti-
lacked intron 1. PCOX-1a and -1b are expressed in equal bodies. The 50-kDa protein is not detected (n/d) by unblocked or blocked
amounts in brain cortex (Fig. 1B). COX-3 polyclonal antibody (pAb).
To determine whether previously uncharacterized COX-1-
related mRNA transcripts were also expressed in human tissues,
human Northern blot experiments were performed using a PCOX-1a completely lacked detectable COX activity (Fig. 3
human intron-1-specific (HCI) probe. These results demon- Lower). COX activity in cells treated with tunicamycin was
strated the existence of previously uncharacterized ⬇5.2- and abolished, indicating that N-linked glycosylation is necessary for
⬇2.8-kb mRNA transcripts (Fig. 1C). Faint hybridization signals COX activity of COX-3.
were also seen around 1.9 kb (data not shown). Hybridization of RNA studies in human tissues indicated highest levels of
HCI to the ⬇5.2-kb form was tissue-specific, with highest levels COX-3 message to be in the cerebral cortex and heart. Western
present in the cerebral cortex, followed by the heart. These blot analysis of human aorta (Fig. 4) by using either COX-1
observations differ from the characterized expression patterns of monoclonal antibody or COX-3 antipeptide polyclonal antibody
COX-1 mRNA (8). detected the presence of distinct 65- and 53-kDa COX-1 related
COX enzymes are intralumenal residents of the endoplasmic proteins. Additionally, the COX-1 but not COX-3 antibody,
reticulum and depend on N-linked glycosylation for proper detected a 69-kDa protein, corresponding to glycosylated
folding and activity. Retention of intron 1 theoretically could COX-1, as well as a 50-kDa protein, which may represent a
prevent COX-3 and PCOX-1a expression by preventing export proteolytic fragment of COX-1 or PCOX-1b. Detection of both
of these mRNAs from the nucleus or by targeting these proteins of the 65- and 53-kDa proteins was selectively reduced by
to another subcellular compartment, preventing glycosylation. preincubation of the antipeptide sera with its cognate peptide,
Therefore, insect cells (Sf9) were infected with recombinant whereas detection of the same proteins by the COX-1 mono-
baculovirus expressing COX-3, PCOX-1a, and COX-1, and cell clonal antibody was unaffected by this treatment.
homogenates were assayed for protein expression by Western Common analgesic/antipyretic drugs and NSAIDs were tested
blotting. Antibodies specific for the conserved amino acid for their ability to inhibit activity of COX-1, -2, and -3. Analyses
sequence (MSREXDPXA) predicted to be encoded by intron 1 were done in the presence of exogenously added arachidonic
in mammals were used to probe for COX-3 and PCOX-1a. This acid at 30- and 5-␮M concentrations. At the higher concentra-
analysis demonstrated that both COX-3 and PCOX-1a are tion of substrate, only COX-3 was inhibited by acetaminophen
efficiently expressed in insect cells. No detectable products (Fig. 5A). Moreover, COX-3 was significantly more sensitive to
resulting from removal of intron 1 by splicing were detected acetaminophen than either COX-1 or -2 at the lower substrate
immunologically or by RT-PCR analysis of RNA extracted from concentration (Fig. 5B). Acetaminophen inhibited COX-3 with
infected Sf9 cells. Moreover, the signal peptide, which in COX-3 an IC50 value of 64 ␮M when done in the presence of 5 ␮M
and PCOX-1a contains an additional intron 1 encoded sequence, arachidonic acid, whereas IC50 values for COX-1 and -2 were 2.1-
was not removed by signal peptidase as it is in COX-1 and -2. and 92.4-fold higher, respectively.
Posttranslational N-linked glycosylation of COX-3 and Acetaminophen is considered to be the active metabolite of
PCOX-1a was compared with that of COX-1 by using tunica- phenacetin, a once popular analgesic/antipyretic drug that is no
mycin to inhibit core glycosylation. Immunoblot analysis dem- longer extensively used because of the occurrence of methemo-
onstrated a decrease in or disappearance of glycosylated forms globinemia, renal toxicity, and suspected renal and bladder
of COX-3, PCOX-1a, and COX-1 (Fig. 3 Upper; Left, Center, and carcinogenesis (9, 10). Phenacetin is rapidly O-deethylated in the
Right, respectively). Expression systems assayed for PGE2 pro- body to form acetaminophen and is further metabolized to other
duction found COX-3 to be ⬇20% of that of COX-1, whereas minor but toxic compounds. Thus, only small levels of phenac-

13928 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.162468699 Chandrasekharan et al.

Fig. 5. Drug inhibition studies. The effects of acetaminophen (A and B), phenacetin (C), and dipyrone (D) on COX-1 (»), COX-2 (F), and COX-3 (䡲) activity in
insect cells. COX activity was measured by the formation of PGE2 after exposure to exogenous 5 ␮M (B) or 30 ␮M (A, C, and D) arachidonic acid for 10 min. Data
are expressed as mean ⫾ SEM (n ⫽ 6 –9).

etin circulate in the blood. Interestingly, however, phenacetin 1 prevents translation and nuclear export of the mRNA. How-
was more potent at inhibiting COX-3 than was acetaminophen ever, both COX-3 and PCOX-1a mRNAs expressed in insect
(Fig. 5C). Under substrate conditions of 30 ␮M, phenacetin cells retain the intron and are exported from the nucleus and
inhibited COX-3 at an IC50 value of 102 ␮M, as opposed to 460 translated (Fig. 3). The polypeptides produced from COX-3 and
␮M for acetaminophen tested under similar conditions. As with PCOX-1a include sequence encoded by intron 1 and are func-
acetaminophen, phenacetin preferentially inhibited COX-3. tionally different from fully spliced COX-1. Therefore, retention
Another analgesic/antipyretic drug, dipyrone, was also signif- of intron 1 provides a mechanism by which a previously unchar-
icantly more potent at inhibiting COX-3 than either COX-1 or acterized COX enzyme, COX-3, can be produced in cells and
-2 (Fig. 5D). Dipyrone inhibited COX-3 with an IC50 value of 52 tissues. Consistent with the concept that retention of intron 1 is
␮M and COX-1 at a 6.6-fold higher concentration. No detectable important in creating COX-3 and/or regulating COX-1 is the
inhibition of COX-2 by dipyrone was observed below 1 mM. finding that the DNA sequence of intron 1 from dog, human, and
Dipyrone is a pro-drug that spontaneously breaks down in mouse COX-1 genes displays a high degree of conservation. In

PHARMACOLOGY
aqueous solutions to structurally related pyrazolone compounds
that differ in their potency as analgesic/antipyretic agents.
Antipyrine and dimethylaminopyrene are related to dipyrone, Table 1. IC50 values of selected analgesic兾antipyretic drugs
and possess markedly reduced therapeutic potency and inhibi- and NSAIDs
tion of COX-3 (Table 1). However, these compounds, like other
IC50, ␮M
analgesic/antipyretic agents, preferentially inhibit COX-3.
COX-3 was also found to differ in its sensitivity to inhibition Drug COX-1 COX-2 COX-3
by a selection of NSAIDs. Diclofenac was the most potent
inhibitor of COX-3 tested and diclofenac, aspirin, and ibuprofen Acetaminophen ⬎1,000 ⬎1,000 460
preferentially inhibited COX-3 over COX-1 and -2. Thalidomide Aminopyrine* ⬎1,000 ⬎1,000 688
and caffeine, both of which have been described as having Antipyrine ⬎1,000 ⬎1,000 863
analgesic properties, did not inhibit COX-3. The overall results Aspirin 10 ⬎1,000 3.1
Diclofenac 0.035 0.041 0.008
indicate that COX-3 possesses COX activity that differs phar-
Dipyrone 350 ⬎1,000 52
macologically from both COX-1 and -2, but is more similar to
Ibuprofen 2.4 5.7 0.24
COX-1.
Indomethacin 0.010 0.66 0.016
Discussion Phenacetin ⬎1,000 ⬎1,000 102
Caffeine ⬎1,000 ⬎1,000 ⬎1,000
Both COX-3 and PCOX-1a are formed by intron retention, a
Thalidomide ⬎1,000 ⬎1,000 ⬎1,000
poorly understood form of alternative splicing. We have previ-
ously shown that COX-2 in chicken is regulated by intron 1 All assays were carried out in the presence of 30 ␮M arachidonic acid.
retention (11). In the case of chicken COX-2, retention of intron *4-dimethylaminoantipyrine.

Chandrasekharan et al. PNAS 兩 October 15, 2002 兩 vol. 99 兩 no. 21 兩 13929

fact intron 1 is more conserved in these species than is exon 1. (24, 25), and some new COX-1-selective inhibitors, such as
COX-1 and -2 genes differ in the placement of intron 1. COX-1 SC-560, have proven ineffective at inhibiting LPS-induced fever
has ten introns, whereas COX-2 has nine. The additional intron in animal models (26, 27). Clinically, rofecoxib, a COX-2-
in the COX-1 gene is intron 1, which is retained in COX-3. selective inhibitor, inhibits naturally occurring fever (28) and
Highly conserved elements of intron 1 may also regulate intron also inhibits the maintenance of fever in animal models. Yet
retention. aspirin, a COX-1 preferential inhibitor, is one of the most
COX-3 shares all of the catalytic features and important effective antipyretic NSAIDs, and inhibits fever at doses ranging
structural features of COX-1 and -2. However, the insertion of from 5–15 mg/kg (29), far below the 60–80 mg/kg used to treat
intron 1 two amino acids downstream from the initiating me- inflammatory disease (30, 31). Furthermore, nimesulide, a
thionine would result in the addition of 30 aa to the signal COX-2 preferential inhibitor, was found to be antipyretic in dogs
peptide. Despite having a signal peptide and intron-1-encoded only at plasma concentrations that would also inhibit COX-1
sequence retained, COX-3 comigrates with COX-1 in SDS/ (32). Thus, a role for COX-1 in fever may exist.
PAGE gels. It also appears to enter the endoplasmic reticulum The mechanism of action of acetaminophen has been un-
where it is glycosylated, and its glycosylation is required for known and postulated to be through inhibition of a brain COX
activity. In insect cells, COX-3 shows approximately 20% of that has never been identified (1, 2). Northern blot analysis and
the activity of COX-1, which in turn exhibits about 20% of the cDNA cloning show that COX-3 is expressed in canine brain.
activity of COX-2. COX-1, COX-2, COX-3, and PCOX-1a all COX-3 also appears from Northern blot studies to be expressed
show equivalent expression in our baculovirus system, and so a in specific regions of the human brain, in particular cerebral
lowered ability of insect cells to express active COX-1 relative to cortex (Fig. 1C). Moreover, our studies using ectopically ex-
COX-2 may be due to the inability of insect cells to posttrans- pressed COX-3 in insect cells demonstrate that COX-3 is
lationally process COX-1 correctly. Whatever the mechanism, significantly more sensitive to acetaminophen than COX-1 or -2.
COX-3 also exhibits this problem and to a greater extent than The steady-state concentrations of acetaminophen after thera-
COX-1. Subcellular localization studies done by differential peutic dosage are approximately 100 ␮M, at which concentration
centrifugation demonstrate that COX-3 and PCOX-1a are mem- only COX-3 is appreciably inhibited (Fig. 5B). Thus, inhibition
brane-bound (data not shown). of COX-3 in brain and the spinal cord may be the long
Retention of intron 1 could alter folding and may affect sought-after mechanism of action of acetaminophen. This pro-
dimerization and the active site. These effects could be through posed mechanism of action also appears to extend to pyrazolone
structural changes or altered protein targeting. COX-1 site- drugs such as dipyrone and related compounds aminopyrine and
directed mutagenesis of either Cys-313 or -540, both of which antipyrine. Dipyrone’s active breakdown product, 4-methylami-
are more than 25 Å from the heme iron, was observed to reduce
noantipyrine, reaches concentrations of 104 ␮M and 86 ␮M in
the activity of the enzyme by 80–90% (12). Therefore, although
plasma and the central nervous system, respectively (33). Thus,
COX-3 contains all of the COX-1 sequence, the retained intron
COX-3 inhibition occurs at known physiological concentrations
sequence could significantly alter its enzymatic properties. In-
of pyrazolone drugs.
hibition studies of COX-3 indicate this to be the case.
Analgesic/antipyretic drugs penetrate the blood–brain barrier
Our studies show COX-3 to be sensitive to drugs that are
well and accumulate in the CNS at high enough concentrations
analgesic/antipyretic, but which have low antiinflammatory ac-
to inhibit COX-3. Carboxylate-containing NSAIDs, on the other
tivity. Pain and fever have many etiologies that employ complex
cellular and biochemical pathways. The finding that COX-3 is hand, cross the blood–brain barrier poorly. Still, central anal-
sensitive to analgesic/antipyretic drugs suggests that the COX-1 gesic mechanisms of action for carboxylate NSAIDs have been
gene plays an integral role in pain and/or fever. Depending on proposed in brain or spinal cord (34, 35). Because COX-3 is so
the physiological context, pain pathways involve products from sensitive to some carboxylate NSAIDs, COX-3 in the CNS may
either the COX-1 or -2 genes. COX-2-selective drugs, for be an essential target of both analgesic/antipyretics and standard
example, are clinically useful in inhibiting inflammatory pain in NSAIDs. Furthermore, the sensitivity of COX-3 to analgesic/
humans (13) and are more potent than COX-1-selective antipyretic drugs and NSAIDs observed in these studies (Fig. 5,
NSAIDs at inhibiting pain induced by proinflammatory agents Table 1) suggests that highly selective inhibitors can be made for
(e.g., carrageenan) in some paw inflammation assays in rodents COX-3.
(14, 15). COX-1-selective drugs, in contrast, are superior to Human COX-3 is mainly expressed as an ⬇5.2-kb mRNA and
COX-2-selective agents at inhibiting visceronociception caused has a tissue-specific pattern of expression (Fig. 1C). This ⬇5.2-kb
by a variety of chemical pain stimulators (16–18). Moreover, mRNA is an alternatively polyadenylated human COX-1 mes-
Ballou and colleagues (19) found that visceronociception was sage reported and partly characterized in its 3⬘ region (5). It
greatly decreased in COX-1 but not COX-2 knockout mice. Both appears, therefore, that the retention of intron 1 may influence
COX-1 and -2, on the other hand, have been implicated in the site at which the mRNA is polyadenylated. This finding
nociception models that measure analgesia outside the gut, such suggests that the 3⬘ untranslated regions of the mRNA may play
as in formalin and urate crystal tests (18–20). A role for COX-1 a functional role in expression of COX-3 and perhaps PCOX-1a.
in pain is further supported by the fact that COX-1-selective The functional significance and the mechanism by which intron
NSAIDs [e.g., as identified by Warner and colleagues (21)]— retention and alternative polyadenylation are coordinated need
such as aspirin, ketorolac, ketoprofen, ibuprofen, and supro- to be elucidated. It is also interesting to note that the ⬇5.2-kb
fen—are clinically important analgesic agents in humans and mRNA has been shown to be regulatable (36) and hence may be
animals. Despite their relative exclusion from the brain, these regulated in response to physiological stimuli and signal trans-
drugs may reach sufficient concentration to effect COX-3 in the duction. Indeed, the levels of COX-3 mRNA in human and
brain. Furthermore, the analgesic effects of these drugs often canine cerebral cortex are relatively low. This may be due to cell
occur at significantly lower doses than those needed to inhibit type-specific expression such as has been shown for COX-1
inflammation (22). Clinical and experimental association of immunoreactive protein in a subpopulation of putative nocicep-
COX-1 and pain may be functionally explained by the finding tor neurons (23). However, COX-3 in human will require further
that COX-1 is a marker for subpopulations of putative nocicep- experimentation because some of the published sequences differ
tor neurons in the dorsal root ganglion (23). by one nucleotide in intron 1 and hence are out of frame. These
With regard to pyresis, COX-2 but not COX-1 knockout mice may constitute genuine polymorphisms or sequencing errors.
demonstrate reduction in LPS- and interleukin-1-induced fevers Alternatively, intron 1 may be out of frame in humans, requiring

13930 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.162468699 Chandrasekharan et al.

other mechanisms such as ribosomal frame shifting to produce Although the peroxidase activity of cyclooxygenase is needed
a functional COX-3 protein. to create the protein radical used in the cyclooxygenase reaction,
We have immunologically identified a 65-kDa protein in continued peroxidase activity is not essential for continued
human aorta, which we postulate is COX-3, and ⬇53 kDa cyclooxygenase activity (39). Because only one turnover of the
proteins, which we postulate to be PCOX-1a. These proteins are peroxidase active site is required for cyclooxygenase activity in
detected by both COX-3 antipeptide polyclonal antibody and a COX-1 and -2, there may be enough residual peroxidase activity
COX-1 monoclonal antibody and appear to be present at about in PCOX-1 proteins to prime them.
25% of the level of COX-1. The 65-kDa protein is smaller than We have previously found that cyclooxygenases bind to
would be predicted if the protein is glycosylated to the same nucleobindin (40). Nucleobindin is a candidate for binding to
extent as COX-1, suggesting that hypoglycosylation or other PCOX-1 proteins as well. Additionally, a form of COX-1 has
differences exist between the 65-kDa protein and COX-1. The been described that colocalizes with prostacyclin synthase in
53-kDa proteins are present as a doublet, and are of a higher filamentous structures of cultured endothelial cells (41). This
molecular weight than that predicted by the PCOX-1a protein filamentous form of COX-1 has no cyclooxygenase activity, and
is a candidate for being a PCOX-1 protein.
primary sequence. This suggests that, like canine PCOX-1a
We previously identified an acetaminophen-inhibited COX
expressed in insect cells, the human protein may be glycosylated,
enzyme activity in a murine macrophage cell line (J774.2)
and that different glycosylation states may exist, giving rise to the treated with diclofenac (42). We proposed that this activity was
doublet observed. A 50-kDa protein is also detected only by the a variant of COX-2, because a protein immunoreactive with
COX-1 monoclonal antibody, and is a candidate for being anti-COX-2 sera was co-induced with the activity. However, this
PCOX-1b. It appears to be present at about 15% of the level of activity was insensitive to aspirin and showed reduced sensitivity
COX-1. to diclofenac, indomethacin, and flurbiprofen—properties not
PCOX-1a is identical to COX-3 except for a deletion of 219 aa shared with COX-3. This finding suggests that additional distinct
in the catalytic domain of the protein, corresponding to exons acetaminophen-inhibitable COX forms exist that are likely
5–8. It lacks detectable cyclooxygenase activity, as shown by its derived from COX-2.
inability to make prostaglandins from arachidonic acid. The
deleted portion contains structural helices HE, H1, H2, H3, H5, We thank and acknowledge the assistance of Marcus Rampton, John
and part of H6 defined for COX-1 and -2. Of these helices, H2 Clinger, Jenny Taylor, Joel Wilson, and Eme Ekpo for assistance with
and H5 form part of the core peroxidase catalytic site. Because Western blots and radioimmunoassays. We thank Kenneth D. Westover
of the lack of H2 and H5, PCOX-1 proteins most likely lack for suggesting the acronym PCOX. We gratefully acknowledge the
financial support of the National Institutes of Health (Grant AR46688),
detectable peroxidase activity. In this way they are similar to the Brigham Young University Cancer Research Center, the Brigham
plant pathogen-induced oxygenase (PIOX) enzymes and Gaeu- Young University Technology Transfer Office, the Office of Research
mannomyces graminis linoleate diol synthase, which also lack and Creative Work, and the College of Physical and Mathematical
peroxidase activity (37, 38). Sciences.

1. Botting, R. M. (2000) Clin. Infect. Dis. 31, 8202–8210. 22. Buckley, M. M. & Brogden, R. N. (1990) Drugs 39, 86–109.
2. Flower, R. J. & Vane, J. R. (1972) Nature (London) 240, 410–411. 23. Chopra, B., Giblett, S., Little, J. G., Donaldson, L. F., Tate, S., Evans, R. J. &
3. Sambrook, J. & Russell, D. (2001) Molecular Cloning: A Laboratory Manual Grubb, B. D. (2000) Eur. J. Neurosci. 12, 911–920.
(Cold Spring Harbor Lab. Press, Plainview, NY), 3rd Ed. 24. Li, S., Wang, Y., Matsumura, K., Ballou, L. R., Morham, S. G. & Blatteis, C. M.
4. Simmons, D. L., Levy, D. B., Yannoni, Y. & Erikson, R. L. (1989) Proc. Natl. (1999) Brain Res. 825, 86–94.
Acad. Sci. USA 86, 1178–1182. 25. Li, S., Ballou, L. R., Morham, S. G. & Blatteis, C. M. (2001) Brain Res. 910,
5. Hla, T. (1996) Prostaglandins 51, 81–85. 163–173.
6. Simmons, D. L., Xie, W., Chipman, J. G. & Evett, G. E. (1991) in Prostaglan- 26. Steiner, A. A., Li, S., Llanos, Q. J. & Blatteis, C. M. (2001) Neuroimmuno-
dins, Leukotrienes, Lipoxins and PAF, ed. Bailey, J. M. (Plenum, New York), pp. modulation 9, 263–275.
67–78. 27. Dogan, M. D., Ataoglu, H. & Akarsu, E. S. (2002) Brain Res. Bull. 57, 179–185.
7. Salmon, J. A. (1978) Prostaglandins 15, 383–397. 28. Schwartz, J. I., Chan, C.-C., Mukhopadhyay, S., McBride, K. J., Jones, T. M.,
8. O’Neill, G. P. & Ford-Hutchinson, A. W. (1993) FEBS 330, 156–160. Adcock, S., Moritz, C., Hedges, J., Grasing, K., Dobratz, D., et al. (1999) Clin.
9. Moyad, M. A. (2001) Semin. Urol. Oncol. 19, 280–293. Pharmacol. Ther. 65, 653–660.
10. Cohen, S. M., Shirai, T. & Steineck, G. (2000) Scand. J. Urol. Nephrol. Suppl. 29. Vigano, A., Dalla Villa, A., Cecchini, I., Biasini, G. C. & Principi, N. (1986) Eur.
205, 105–115.

PHARMACOLOGY
J. Clin. Pharmacol. 31, 359–361.
11. Xie, W., Chipman, J. G., Robertson, D. L., Erikson, R. L. & Simmons, D. L.
30. Doughty, R. A., Giesecke, L. & Athreya, B. (1980) Am. J. Dis. Child. 134,
(1991) Proc. Natl. Acad. Sci. USA 88, 2692–2696.
461–463.
12. Kennedy, T. A., Smith, C. J. & Marnett, L. J. (1994) J. Biol. Chem. 269,
31. Gunsberg, M., Bochner, F., Graham, G., Imhoff, D., Parsons, G. & Cham, B.
27357–27364.
(1984) Clin. Pharmacol. Ther. 35, 585–593.
13. Matheson, A. J. & Figgitt, D. P. (2001) Drugs 61, 833–865.
32. Toutain, P. L., Cester, C. C., Haak, T. & Laroute, V. (2001) J. Vet. Pharmacol.
14. Riendeau, D., Percival, M. D., Brideau, C., Charleson, S., Dube, D., Ethier, D.,
Ther. 24, 43–55.
Falgueret, J. P., Friesen, R. W., Gordon, R., Greig, G., et al. (2001) J. Phar-
macol. Exp. Ther. 296, 558–566. 33. Cohen, O., Zylber-Katz, E., Caraco, Y., Granit, L. & Levy, M. (1998) Eur.
15. Smith, C. J., Zhang, Y., Koboldt, C. M., Muhammad, J., Zweifel, B. S., Shaffer, J. Clin. Pharmacol. 54, 549–553.
A., Talley, J. J., Masferrer, J. L., Seibert, K. & Isakson, P. C. (1998) Proc. Natl. 34. Bovill, J. G. (1997) Eur. J. Anaesthesiol. Suppl. 15, 9–15.
Acad. Sci. USA 95, 13313–13318. 35. Jurna, I. & Brune, K. (1990) Pain 41, 71–80.
16. Kusuhara, H., Matsuyuki, H. & Okumoto, T. (1998) Prostaglandins Other Lipid 36. Plant, M. H. & Laneuville, O. (1999) Biochem. J. 344, 677–685.
Mediat. 55, 43–49. 37. Sanz, A., Moreno, J. I. & Castresana, C. (1998) Plant Cell 10, 1523–1537.
17. Goto, K., Ochi, H., Yasunaga, Y., Matsuyuki, H., Imayoshi, T., Kusuhara, H. 38. Oliw, E. H., Su, C. & Sahlin, M. (1999) Adv. Exp. Med. Biol. 469, 679–685.
& Okumoto, T. (1998) Prostaglandins Other Lipid Mediat. 56, 245–254. 39. Landino, L. M., Crews, B. C., Gierse, J. K., Hauser, S. D. & Marnett, L. J.
18. Ochi, T., Motoyama, Y. & Goto, T. (2000) Eur. J. Pharmacol. 391, 49–54. (1997) J. Biol. Chem. 272, 21565–21574.
19. Ballou, L. R., Botting, R. M., Goorha, S., Zhang, J. & Vane, J. R. (2000) Proc. 40. Ballif, B. A., Mincek, N. V., Barratt, J. T., Wilson, M. L. & Simmons, D. L.
Natl. Acad. Sci. USA 97, 10272–10276. (1996) Proc. Natl. Acad. Sci. USA 93, 5544–5549.
20. Martinez, R. V., Reval, M., Campos, M. D., Terron, J. A., Dominguez, R. & 41. Liou, J. Y., Shyue, S. K., Tsai, M. J., Chung, C. L., Chu, K. Y. & Wu, K. K. (2000)
Lopez-Munoz, F. J. (2002) J. Pharm. Pharmacol. 54, 405–412. J. Biol. Chem. 275, 15314–15320.
21. Warner, T. D., Giuliano, F., Vojnovic, I., Bukasa, A., Mitchell, J. A. & Vane, 42. Simmons, D. L., Botting, R. M., Robertson, P. M., Madsen, M. L. & Vane, J. R.
J. R. (1999) Proc. Natl. Acad. Sci. USA 96, 7563–7568. (1999) Proc. Natl. Acad. Sci. USA 96, 3275–3280.

Chandrasekharan et al. PNAS 兩 October 15, 2002 兩 vol. 99 兩 no. 21 兩 13931