THE JOURNAL 0 1988 hy The American Society for Biochemistry and

OF BIOLOGICAL CHEMISTRY Molecular Biology, Inc

Vol. 263. No. 17, Issue of June 15, pp. 8473-8479,1988 Printed in U.S.A.

The Kinetics of Interleukin 1 Secretion from Activated Monocytes DIFFERENCES BETWEEN INTERLEUKIN la AND INTERLEUKIN I@*

(Received for publication, December 4, 1987)

Daria J. HazudaS, John C. Lees, and Peter R. YoungSV From the Departments of $Molecular Genetics and §Zmmunology and Anti-infectiues Therapy, Smith Kline and French Laboratories, King of Prussia, Pennsylvania 19406-0939

We have performed pulse-chase experiments to in- vestigate the secretion and processing of interleukin 1 (IL-1) by human peripheral blood monocytes. Poly- clonal antisera generated against either recombinant IL-la (PIS) or IL-lB (pI7) could distinguish the two isoelectric forms in lysates and supernatants of lipo- polysaccharide-activated monocytes. In agreement with previous results, no processed IL-1 (a or 8) is detected in cell lysates. Both the 31-kDa precursor and 17-kDa mature forms of IL-1 were present, however, in the culture media indicating that processing is not required for secretion. The relative amounts of the secreted 3 1- and 17-kDa forms of IL- 1 remain constant with time throughout each experiment; in addition, 3 1- kDa IL-1 added to monocyte cultures is not processed to the mature 17-kDa form. Precursor IL-la is how- ever, processed to 17 kDa by monocyte extracts. Therefore, the maturation and secretion of IL-1 are intimately coordinated processes.

The kinetics of IL- 1 secretion are unique in compar- ison with other secreted proteins; release of both IL- l a and IL-18 is delayed following synthesis, and large pools of precursor IL- 1 accumulate intracellularly. The intracellular half-lives of IL-la and IL-18 are 15 and 2.5 h, respectively. This discrepancy in half-lives is a reflection of the different kinetics with which IL-la and IL- 18 are secreted. IL- 1s is released continuously beginning 2 h after synthesis, whereas the secretion of IL-la is delayed for an additional 10 h. The distinct kinetics of secretion demonstrated for IL- la and IL- 18 suggest that the release of each PI species of IL-1 is controlled by a selective mechanism(s).

Interleukin 1 (IL-1)' is an immunoregulatory protein se- creted by activated monocytes as part of the general inflam- matory response (1). IL-1 is produced by a variety of cell types including lymphocytes (2), astrocytes (3), and kera- tinocytes (4); however, the pre-eminent source of this mon- okine is the peripheral blood monocyte (1).

Two forms of IL-1, a and @, have been identified via cDNA cloning (5-9) and correspond to two different isoelectric

* 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.

ll To whom correspondence should be addressed. The abbreviations used are: IL, interleukin; PMSF, phenylmeth-

ylsulfonyl fluoride; LPS, lipopolysaccharide; EGTA, [ethylene- bis(oxyethylenenitri1o)ltetraacetic acid; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; HEPES, 4-(2-hydroxy- ethyl)-1-piperazineethanesulfonic acid; rIL, recombinant interleukin; TNF, tumor necrosis factor.

forms, PIS 5.2 and 7.0, respectively (10,111. These two proteins are evolutionarily conserved, showing approximately 27-33% homology at the amino acid level, with nearly identical gene structures consisting of seven conserved exons (12, 13). The observation that IL-la and IL-1@ elicit a similar if not iden- tical spectrum of systemic and cellular responses (14) is consistent with the finding that both IL-la and IL-1@ bind to the same receptor on all cells tested to date (15). Further- more, both PI forms of the IL-1 protein are produced in response to a wide range of stimuli both in vitro and in vivo. In view of these similarities between IL-la and -& it is unclear why two distinct molecules have evolved which remain inti- mately conserved in function.

Both IL-la and -p are synthesized as large 31-kDa precur- sor molecules which are subsequently processed to their 17- kDa mature forms (16, 17). The temporal sequence vis-a-vis processing and secretion for both IL-la and -p as well as the mechanism by which IL-1 is secreted from the cell remain unresolved. Neither form of IL-1 encodes a hydrophobic stretch of amino acids of sufficient length to characterize a secretory signal sequence, and unlike most secretory proteins, IL-1 accumulates intracellularly (4, 5, 17, 18).

We have used pulse-chase experiments to follow the syn- thesis, processing, and secretion of both IL-la and -0 from LPS-activated human peripheral blood monocytes. The avail- ability of antisera which exclusively recognize either IL-la or IL-18 has enabled us to follow the kinetics of each of these processes for both IL-1s simultaneously. We demonstrate that the maturation of IL-1 from the 31-kDa precursor to the mature 17-kDa form is co-secretory and that the secretion of both IL-la and -/3 is delayed. The two proteins, however, differ in their rate of secretion from the cell suggesting that IL-la and IL-1p are secreted by distinct mechanisms and that some level of biological regulation for these molecules may be achieved during their release from the cell.

MATERIALS AND METHODS

35S Labeling of rIL-a and -+The genes encoding both the precur- sor and mature forms of IL-la and IL-lS were cloned into an inducible Escherichia coli expression vector under control of the X pL promoter (19): Overnight cultures grown at 32 "C in modified M56 minimal medium with biotin and ampicillin (50 pg/ml) were diluted 1:10 in fresh medium and grown to an Am of 0.6, at which time the temper- ature was raised to 42 "C by the addition of 0.5 volume 65 "C medium. Cells were incubated for 1 h at 42 "C, then centrifuged and resus- pended in an equal volume of unmodified M56 supplemented with 0.2% glucose, 1 pg/ml thiamine, 2 pM amino acids minus cysteine and methionine, biotin, ampicillin, and 20 pCi of trans label (85% methionine, 15% cysteine, ICN, >lo00 Ci/mmol). The cells were labeled for 20 min at 42 "C, centrifuged, and the pellets frozen on dry ice.

Prior to use, the pellets were resuspended in 0.2 volume of lysis

M.-J. Chen and P. R. Young, unpublished studies.

a473

8474 Secretion and Processing of Interleukin l a and 1/3 in Monocytes

buffer (40 mM NaCl, 50 mM Tris-HC1, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol) plus 2 mM PMSF and sonicated for 30 s using a Branson Sonifier. Sonicates were centrifuged in an Eppendorf micro- centrifuge for 30 min. The supernatants were then used in immuno- precipitation reactions essentially as described below for monocyte supernatants.

Preparation and Metabolic Labeling of Human Peripheral Blood Monocytes-Human peripheral blood monocytes were prepared from Red Cross buffy coats by centrifugation through Ficoll and Percoll gradients (20). In any single experiment, monocytes were used from an individual donor in an attempt to control for variation between donors. 5 X lo6 cells/60-mm tissue culture dish were allowed to adhere for 1 h in RPMI medium (GIBCO) containing 1% human AB serum (Irvine Scientific). After extensive washing with phosphate-buffered saline (137 mM NaC1, 2.7 mM KC1, 1.3 mM KH,PO,, 9.5 mM Na2HP04, pH 7.5), the attached monocytes were activated with 10 ng/ml E. coli LPS. Following activation for 1.5 h, the cells were starved for 20 min in methionine-free cysteine-free Dulbecco's mod- ified Eagle's medium (GIBCO) containing LPS. By this time, the monocytes had been activated with LPS for a total of approximately 2 h.

LPS-activated monocytes were labeled with 200 pCi of [35S]methi- onine (Amersham Corp., >800 Ci/mmol) and 100 pCi of [35S]cysteine (Amersham Corp., >600 Ci/mmol). Metabolic labeling was done for 1 h in the presence of LPS using methionine-free cysteine-free DMEM containing 1% dialyzed fetal bovine serum (GIBCO). After the 1-h radioactive pulse, cells were washed as above and then chased for various times in nonradioactive medium containing LPS.

Monocytes were lysed directly on tissue culture dishes using 0.5 ml of RIPA buffer (150 mM NaCl, 10 mM Tris-HC1, pH 7.5, 1% deoxy- cholate, 1% Triton X-100, 0.1% SDS, 10 mM EDTA, 2 mM PMSF) and then quickly frozen on dry ice. Culture supernatants were cen- trifuged for 5 min at 1000 rpm in a Beckman model TJ6 table-top centrifuge to remove cells and/or cell debris prior to freezing. The lysates and supernatants were stored at -20 "C.

Immunoprecipitation-Rabbit anti-rIL-la and anti-rIL-l@ anti- sera were generated by immunizing rabbits with purified E. coli- produced 17-kDa recombinant IL-la or -@, respectively (191.3 The resultant antisera were demonstrated to be neither cross-reactive in Western blots using recombinant protein nor cross-neutralizing in the EL4 thymoma activity assay.4

For immunoprecipitation experiments, monocyte lysates were made 2% milk by the addition of an equal volume of 4% nonfat dry milk (Carnation) in 1 X RIPA. The supernatant fractions contained a high percentage of protein due to the serum content and conse- quently required a lower percentage of the blocking agent to eliminate nonspecific interactions. Supernatants were, therefore, made 0.5% milk by using an equal volume of 1% milk in 2 X RIPA. The fractions were preincubated with Protein A-Sepharose beads (50 pl, half- packed volume, Bio-Rad) for 1 h at 4 'C. The beads were then removed by centrifugation for 10 min at 1000 rpm in a Beckman model TJ6 table-top centrifuge. Aliquots of these supernatants were subsequently used in immunoprecipitation reactions with antisera specific for IL-la, IL-la, or TNF-a as described. In some experiments, the antiserum was preincubated with purified antigen for 30 min at 20 "C in phosphate-buffered saline before introduction into the indi- cated reaction.

Incubation with antiserum proceeded for 1 h at 4 "C, after which 50 pl of Protein A-Sepharose beads were added, and the reaction continued for an additional 1 h. The immunoprecipitates were cen- trifuged as above and then washed twice for 10 min in RIPA. The beads were resuspended in 45 pl of SDS loading buffer and boiled for 5 min. One-third of each reaction was electrophoresed on a 15% SDS- polyacrylamide gel (21). Gels were fixed for 1 h in 40% methanol, 10% glacial acetic acid, prepared for autoradiography with Amplify (Amersham Corp.), and then dried and exposed to Kodak XAR x-ray film at -70 "C with an intensifying screen. Quantitation of the relative amounts of IL-la and -@ at each time was done by densito- metric analysis of the autoradiograms (LKB Ultrascan).

Fractionation of Monocytes-Monocytes were fractionated using the method described by Matsushima et al. (22) as follows. Cells were activated with LPS and labeled with 35S essentially as outlined above except that no chase was required. 5 X lo7 monocytes were harvested from each of 4 plates using a cell scraper. The cells were washed twice with phosphate-buffered saline and then resuspended in 0.5 ml of

P. L. Simon, unpublished data. ' P. L. Simon and P. R. Young, unpublished data.

buffer A (25 mM HEPES, pH 7.5, 2 mM EDTA, 5 mM EGTA, 1 mM PMSF). Monocytes were allowed to swell on ice for 10 min in this hypotonic buffer before homogenization using a 1-ml Dounce homog- enizer. Following the addition of 0.5 ml of 0.66 M sucrose in buffer A, nuclei and undisrupted cells were removed from the lysate by centri- fugation at 200 X g for 10 min.

Monocyte lysates were then fractionated into cytosolic, membrane, and particulate components on a discontinuous sucrose density gra- dient. The cleared lysates were layered on a 1-ml cushion of 40% sucrose in buffer A and then centrifuged at 32,000 rpm for 50 min at 4 "C using a Beckman model TJ6 table-top ultracentrifuge equipped with the SW 55 swinging bucket rotor. The top 750 pl representing the cytosolic fraction was carefully removed without disturbing the interface. The pellet representing the particulate fraction was resus- pended in 1 ml of buffer A and spun as above to repellet. The particulate components were dissolved in 750 pl of buffer A.

RESULTS

Characterization of Immunoreactivity of Rabbit Anti-IL-1 Polyclonul Antisera-The ability to follow the kinetics of IL- 1 secretion from activated monocytes by pulse-labeling studies is contingent upon the availability of antisera highly specific to either PI form of IL-1. Rabbit polyclonal antisera inde- pendently raised against purified recombinant IL-la or IL- I@, and which had been previously shown to react with the cognate proteins in enzyme-linked immunosorbent assays, Western blots, and activity neutralization ~ tudies ,~ were tested for their ability to immunoprecipitate 35S-labeled pre- cursor and mature IL-la and IL-1@ expressed in E. coli.

As shown in Fig. 1, both antisera are monospecific, i.e. the IL-la antiserum recognized both the 31- and 17-kDa labeled forms of the IL-la synthesized in E. coli (Fig. VI, lanes 1 and 2) but did not immunoprecipitate either form of IL-lp (Fig. lA, lanes 3 and 4) , and vice versa (Fig. 1B). Furthermore, the identity of each IL-1 species immunoprecipitated from labeled monocytes was confirmed by preincubating antisera with the appropriate purified recombinant IL-1 (Fig. IC). These prop- erties, therefore, enabled simultaneous analysis of IL-la and -8, as well as discrimination of the molecular state of each of these proteins. This is a particularly important issue with regard to IL-la where previous studies monitoring activity could not distinguish between the precursor and mature forms of this protein as both are biologically active (23).

IL-la and -@ Are Synthesized as Cytosolic Precursor Pro- teins with Different Intracellular Half-lives-Previous studies indicated that LPS induced rapid synthesis of both IL-la and -p mRNAs. Maximum levels of mRNA were reached 2-4 h after the addition of LPS after which time synthesis remained constant or decreased (22).5 At this time, IL-1@ mRNA con- stitutes approximately 5% of total poly(A+) mRNA.' We, therefore, selected to pulse the monocytes with %labeled methionine and cysteine 2 h after treatment with LPS to achieve the greatest levels of incorporation. The label was chased via subsequent incubation with unlabeled medium.

IL-1 was detected in labeled monocyte cell lysates by im- munoprecipitation with antisera specific for either IL-la or -@. The results, shown in Fig. 2, confirm that both IL-la and IL-1@ are synthesized by monocytes as larger precursor mol- ecules of approximately 31 kDa. There is no detectable intra- cellular mature 17-kDa form of either IL-la or IL-1p nor any evidence of processing intermediates. These experiments sug- gest that the maturation of IL-1 to 17 kDa is either co- or postsecretory.

It should be noted that IL-la appears in two forms of slightly different mobility (Fig. 2A and also Fig. 441, suggest- ing a post-translational modification. There is, however, no

B. Dalton, unpublished data. M. Smith, Unpublished data.

Secretion and Processing of Interleukin l a and l p in Monocytes

A. 0 I 2 4 7.5 12.5 16 Hr MW

43.0 -

25.7 - 18.4 -

14.3 - 6.2 - 3.0 -

43.0 -

25.’7 - 18.4 - 14.3 - 6.2 - 3.0 -

ant i - r1L l -C ant i - r IL I ts c 1 2 3 4

- 43.0

- 25.7

- 18.4

- 14.3

- 6.2

- 3.0

anti-rIUa

B. 0 I 2 4 7.5 125 16 Hr MW

- 43.0

-25.7

- 184

- 14.3

- 6.2

-30

anti-rILlp

a475

“

r l L l a rlL1P Anti- Anti-

FIG. 1. Specificity of anti-IL-la and anti-IL-1fl antisera in immunoprecipitation reactions. 35S-Labeled recombinant precur- sor IL-la, rIL-la, recombinant precursor IL-10 and rIL-lj3 were immunoprecipitated with anti-rIL-la or anti-rIL-l@ antisera (A and B, respectively) and analyzed by 15% SDS-PAGE as described under “Materials and Methods.” Lane 1, recombinant precursor IL-la; lane 2, rIL-la; lane 3, recombinant precursor IL-lp; lane 4 , rIL-lj3. In C, 35S-labeled monocyte cell lysates were immunoprecipitated with anti- rIL-la (lanes 1 and 2) or anti-rIL-lP antisera (lanes 3 and 4 ) . The band corresponding to either IL-la or IL-lj3 was confirmed by pre- blocking the antisera with 2 pg of the appropriate IL-1 (lanes 2 and 4).

apparent kinetic relationship between these two forms, The higher molecular weight immunoprecipitated products seen in Fig. 2 A , lanes 5 and 6, were not observed in all experiments, but based on their mobility, may represent dimers of IL-la.

The half-lives of the intracellular IL-1 precursors were determined by quantitating the relative amount of IL-la and -B from densitometric measurements of autoradiograms as a function of time (Fig. 3). In three separate experiments, the measured half-life of IL-la in LPS-activated human periph- eral blood monocytes was 3-4 times longer than that of IL- ID: 15 h versus 3-4 h.

The half-life of IL-1 in monocytes could be determined by at least two distinct parameters: degradation and secretion. Degradation, in turn, might be influenced by subcellular lo- calization. Several groups have reported that IL-1 activity may be found in both the cytosolic and noncytosolic fractions of monocytes (22, 29). To examine whether the difference in half-lives may be accounted for by differences in the com- partmentalization of each IL-1 species, we fractionated mon- ocyte lysates into cytosolic and particulate components, the latter representing subcellular organelles including the plasma membrane and lysosomes. Immunoprecipitation analysis of these crude fractions using polyclonal anti-IL-la or anti-IL- l p antisera demonstrates that both PI forms are found in the cytosol and in the particulate fraction. Although the relative amount in either fraction differs for IL-la (Fig. 4A) and -p (Fig. 4B, compare lanes 1 and 2), the difference is insufficient to account for the difference in the half-lives of these two proteins. Moreover, in pulse-chase experiments, the half-life

FIG. 2. Pulse labeling of intracellular IL-la and IL-lfl. LPS- activated monocytes were metabolically labeled with [35S]methionine as outlined under “Materials and Methods.” After 1 h the cells were washed extensively, and the radiolabel was chased with cold media for 0, 1, 2, 4, 7.5, 12.5, and 16 h as indicated. Monocyte lysates were analyzed for radiolabeled IL-la (A) and IL-16 ( B ) by immunoprecip- itation with specific antisera as in Fig. 1.

- c C P W

0 - L .- - L m 3 0

- - - m C

L

- I

C

4 E

% I , , , E 1

0 5 10 15 20 Hr. After Pulse

FIG. 3. Intracellular half-life of IL-la and IL-lfl. The rela- tive amounts of intracellular IL-la and IL-lj3 (shown in Fig. 2) were quantitated by densitometry and then expressed as a mole percent of the radiolabeled IL-la or -j3 present at the 1-h chase time (maximal incorporation). The decay of IL-la (M) and IL-lj3 (W) in LPS-activated monocytes is graphed as a function of chase time. From three such experiments, the half-lives of intracellular IL-la and IL-lj3 were determined to be 15 and 3.5 h, respectively.

measured for IL-la and IL-lp in each fraction is identical (data not shown). It should also be noted that we could not detect significant IL-1 in the membrane fraction.

IL-la and IL-lp Are Released from Monocytes with Dra- matically Different Kinetics-To assess whether the difference in the half-lives of IL-la and $3 is related to secretory differ- ences between the proteins, we investigated the kinetics of IL-la and -p secretion. Culture fluids derived from the same pulse-chase experiment above were immunoprecipitated with IL-la and -p specific antisera.

As shown in Fig. 5, A and B, the secretion of both IL-la

8476 Secretion and Processing of Interleukin 1 (Y and 1 p in Monocytes

A. I 2 8. 1 2

i

e : i

FIG. 4. Immunoprecipitation of fractionated monocytes. One-h pulse-labeled LPS-activated monocytes were fractionated into cytosolic and particulate fractions (22). Fractions were divided in half and then immunoprecipitated with either anti-IL-la ( A ) or anti-IL- 18 ( B ) specific antisera as described previously. Immunoprecipitated reactions were analyzed by SDS-PAGE. Lane I, cytosol; lane 2, particulate fractions immunoprecipitated with the antisera as indi- cated.

A. 0 I 2 4 6.5 7.5 12.5 Hr MW

- 43.0 - 2 5 7

- 18.4

- 14.3 - 6.2

- 3.0

anti-rILla

B 0 I 2 4 7.5 12.5 16 Hr MW

- 430

- 257

- 184 I - I4 3

5B, lane 3) . In contrast, IL-la is released from monocytes in significant quantities after 12.5 h (Fig. 5A, lane 7). The observation that IL-1P is secreted some 8 h prior to IL-la is consistent with the previously demonstrated differences in their respective intracellular half-lives. Moreover, secretion of IL-1P is continuous as is evident in experiments where labeled monocytes were repeatedly washed and refed during the course of the chase and the supernatants harvested at each time immunoprecipitated with IL-1 antisera. Starting at 2 h, and for several hours thereafter, the IL-1P which was synthesized during the initial 1-h pulse continues to be se- creted (data not shown). Similar experiments for IL-la con- firmed the longer delay in its release from cells. These data imply a distinct mechanism of secretion for each PI species of IL-1.

The Release of IL-1 from Human Monocytes Is Delayed in Comparison to Other Secreted Monokines-Given the unusu- ally slow secretion of IL-1 from monocytes in comparison to what has been observed for signal sequence-directed secreted proteins in other cells (24), we wished to examine whether this was a general phenomenon for activated monocytes or peculiar to IL-1. We, therefore, examined the secretion of another monokine, tumor necrosis factor a (TNF-a).

TNF-a was chosen as a control for several reasons. TNF-a is produced by monocytes in response to LPS (25), and like IL-1, TNF-a is synthesized as a larger precursor protein (26); however, unlike IL-1, the precursor for TNF-a encodes a putative hydrophobic signal sequence (26).

Fig. 6 demonstrates that under conditions of the previous experiments, monocytes are competent to release TNF-a with kinetics which are consistent with co-translational transport into the endoplasmic reticulum and subsequent secretion. TNF-a is seen exclusively outside the cells (Fig. 6, compare lanes I and 2). In contrast to IL-1, TNF-a is secreted from both monocytes rapidly, and we can conclude that the delay in the release of IL-1 from monocytes is a difference in secretion mechanism.

Monocytes Release a Variety of Proteins with Kinetics Sim- ilar to IL-1-When the culture fluid from pulse-labeled acti- vated monocytes is electrophoresed without immunoprecipi- tation, two classes of proteins are distinguishable solely on the kinetics of their release from the cell (see Fig. 7A). A large

cell supernatant t TNF

CHASE: 101 10 I 2 4 1 m LANE: I 2 3 4 5 6 7 MW

43.0 - 25.7

- 18.4

- 14.3

FIG. 5. Pulse labeling of extracellular IL-la and IL-1s. The supernatants from pulse-labeled LPS-activated monocytes were har- vested at 0, 1, 2, 4, 6.5, 7.5, and 12.5 h for IL-la ( A ) and 0, 1, 2, 4, 7.5, 12.5, and 16 h for IL-18 (B). Supernatants were analyzed for secreted radiolabeled IL-la and IL-lp, respectively, by immunopre- cipitation as in Fig. 2.

and IL-lP from monocytes is delayed for several hours follow- ing their synthesis despite the high levels of IL-1 mRNA and intracellular protein rapidly induced upon LPS stimulation. This delay in release is quite unlike that observed for signal sequence-directed secretion (24).

Although both forms of IL-1 are delayed in their secretion from monocytes, they differ in their detailed release kinetics. Precursor (31 kDa) and mature (17 kDa) IL-lP are released in significant amounts beginning 2 h following synthesis (Fig.

- 6.2

- 3.0

anti-rTNF

FIG. 6. Pulse labeling of TNF-CY. LPS-activated monocytes were metabolically labeled with [35S]methionine and cysteine essen- tially as described in previous experiments except that [35S]cysteine was supplemented to 100 pCi/ml total. After a 1-h pulse, the label was chased for 0, 1, 2, and 4 h as indicated. The cell lysate at zero- chase and supernatants a t four chase times were analyzed for TNF- a by immunoprecipitation with anti-recombinant TNF-a antisera. The band corresponding to mature TNF-a was confirmed by pre- blocking the antisera with purified cold TNF-a. Lane I, cell lysate after 1-h pulse, no chase; lanes 2-5, supernatant after 0, 1,2, and 4 h of chase; lane 6, same as lane 1 but the TNF antisera was preblocked with 2 pg of cold recombinant TNF-a; lane 7, same as lane 2, but the TNF antisera was preblocked as in lane 6.

Secretion and Processing of Interleukin 1 a and 1 p in Monocytes a477 A 0 I 2 4 7.5 125 16Hr MW

- 25.7

- 18.4

- 14.3

- 6.2

- 3.0

I I I I 0'; 5 10 15 20

Hr. After Pulse

FIG. 7. '%-Labeled proteins secreted from LPS-activated monocytes. Supernatants from 35S pulse-labeled activated mono- cytes after 0, 1, 2,4, 7.5, 12.5, and 16 h of chase with cold media were subjected to SDS-PAGE ( A ) without prior immunoprecipitation. The bands indicated (A, B, and C) were quantitated by densitometry and

-, C). graphed as a function of time of chase ( B ) (U, A; M, B

number of proteins, like TNF, appear to be released either simultaneous with or very soon after synthesis, appearing in the extracellular media at the zero chase time (Fig. 7A, lane 1; see also Fig. 7B, quantitation of "C"). These proteins are secreted with kinetics typical of the classical secretory path- way. In contrast, a variety of other proteins are delayed in their release from monocytes, appearing in the supernatant from one to several hours after synthesis (Fig. 7A, lanes 2-7). When the secretion of three proteins (A, B, C in Fig. 7A) was quantitated as a function of time it is evident that C is secreted rapidly whereas A and B are delayed in their secretion from monocytes in a manner similar to IL-lp (compare Fig. 7B and Fig. 8). Such kinetics may be characteristic of a secretory mechanism for these proteins which is distinct from the classical pathway previously described (24).

There Is No Apparent Precursor-Product Relationship be- tween Secreted 31- and 17-kDa IL-lp-Although no 17-kDa IL-lP was detected in immunoprecipitations of labeled mon- ocyte lysates (Fig. 2), both 17- and 31-kDa IL-1p were found in monocyte supernatants (Fig. 5)) suggesting that processing of IL-1 is either co- or postsecretory. To assess whether 17- kDa IL-lp was processed from the secreted 31-kDa precursor, i.e. whether maturation of IL-1 was a postsecretory process, the relative amount of the precursor and mature IL-lp se- creted in pulse-chase experiments was quantitated as a func- tion of time. As shown in Fig. 8A, there was no discernible difference in the rate of appearance of either form of IL-1p outside the cell. Moreover, the ratio of 31- to 17-kDa secreted IL-1p did not change significantly for up to 16 h after the pulse, whether monitored continuously (Fig. 8B) or sampled at different intervals after the pulse (data not shown). These data suggest that maturation of IL-1 occurs prior to or upon release from the cell.

To localize the site of processing, 35S-labeled recombinant

IL-1p precursor was incubated with cultures or lysates of activated monocytes. As demonstrated in Fig. 9A, 35S-labeled IL-1p precursor was not processed to 17 kDa in the extracel- lular media of activated monocytes within times at which processing of endogenously labeled IL-1 was observed (Fig.

A 3 O C 0 1

I I I

Hr. After Pulse 5 10 15 20

r-

5 1 Y 0

O O % 5 10 15 Hr. After Pulse

FIG. 8. Extracellular 31- and 17-kDa IL-lB. In A the relative amount of extracellular 31-kDa IL-lj3 (o"--o) and 17-kDa IL-la (U) was quantitated by densitometry from pulse-chase experi- ments shown in Fig. 5. The amount of extracellular IL-1j3 is expressed as a mole percent of the labeled intracellular IL-lj3 at the time of 1- h chase as in Fig. 2. The lack of precursor-product relationship between the secreted 31-kDa IL-1j3 and secreted 17-kDa IL-1j3 is further demonstrated in B by the relative constancy of the ratio of the two secreted forms.

n MW O 5 1 2 H r

4 3 0 - 0"-

2 5 7 - 184 - 143 - 6 2 -

3 0 -

-mature IL-~B

1 2 3 4 1 2 3 4

FIG. 9. IL-la is processed by cell-associated enzyme(s). %- Labeled recombinant precursor IL-la was added to the media of LPS- activated monocytes and incubated with the cells as indicated, at which time the media was collected and samples analyzed by immu- noprecipitation. In B, labeled samples (31-kDa precursor or 17-kDa mature IL-10) were added to activated monocytes, and extracts were prepared as described under "Materials and Methods" except that EDTA, EGTA, and PMSF were omitted. The resultant cytosolic (c) and particulate ( p ) fractions were analyzed as in A .

8478 Secretion and Processing of Interleukin la and 18 in Monocytes

5A). In contrast, the precursor was processed to the mature 17-kDa mature protein upon incubation with activated mon- ocyte lysates (Fig. 9B, lane 2). The processed product comi- grates with the 17-kDa IL-lp which has been shown to cor- respond exactly to the monocyte-derived protein (19) (Fig. 9B, lanes 3 and 4) . It is interesting to note that the correctly processed IL-lB precursor is only observed in the cytosolic compartment (lane 2) and not in the particulate fraction (lane 3), where there are a number of other processed forms which do not comigrate with the authentic mature IL-1p (Fig. 9A, lanes 1 and 2). These data suggest that the processing occurs outside of the particulate fraction. Since we did not detect intracellular processing of endogenous IL-10 precursor (Fig. 2B), we can conclude that processing is coordinated with secretion. Our inability to detect 17-kDa IL-la in monocytes is not due to instability during isolation. Fig. 9 (lanes 3 and 4 ) shows that the mature protein is stable in these extracts for greater than 1.5 h. Together these data provide strong evidence suggesting that the maturation of IL-la is accom- plished via a co-secretory mechanism.

Two other points are worth noting. The secretion of IL-lp from monocytes is relatively efficient, with 50% of the syn- thesized intracellular protein being released into the medium either in precursor or mature form (Fig. EM). However, in other experiments the amount of IL-1 released varied as did the ratio of precursor to mature IL-1. The observed differences between experiments and preparations of monocytes may be attributed to variability among donors. In any single experi- ment, cells were, therefore, prepared from a single donor. Changing the concentration of fetal bovine serum (0.1-10%) had little or no effect on either the secretion or processing of IL-1.

DISCUSSION

We have studied the secretion and processing of the two PI forms of IL-1 (a and 8) from LPS-activated human peripheral blood monocytes. Polyclonal antisera which recognize both the 31-kDa precursor and 17-kDa mature forms of either IL- la or -p were used in immunoprecipitations of metabolically labeled monocyte cultures. These antisera are highly specific for each PI form of IL-1, thus allowing us to follow the secretion and processing of both 1L-la and IL-lp simultane- ously.

This is the first study to analyze in detail the kinetics of IL-la and -/3 processing and secretion from activated human peripheral blood monocytes in radiolabeled pulse-chase ex- periments. With the exception of studies in the mouse mac- rophage cell line P388D1 (16) previous secretion studies have relied on measurements of IL-1 activity (27, 28), which is more sensitive but less informative. IL-la and -0 cannot be distinguished in activity assays without additional fractiona- tion on the basis of PI or neutralization with monospecific antisera. The interpretation of actibity data is further com- plicated by the fact that precursor IL-la is active while that for IL-lp is inactive (23). Activity measurements, therefore, fail to detect an entire subpopulation of secreted IL-1, namely precursor IL-lP, which our data indicates may be a significant proportion (in some experiments more than 50%) of the IL- I@ secreted from monocytes. Finally, activity measurements reflect the steady state kinetics of the system, whereas de- tailed pulse-chase experiments can monitor the dynamics of the Secretion and maturation processes.

Both IL-la and -/3 are synthesized and accumulate in monocytes as larger precursor proteins of 31 kDa. We, as well as others (17, la), have failed to detect any evidence of intracellular mature 17-kDa IL-1 or other intracellular proc-

essing intermediates. Our experience indicates that the pres- ence of lower molecular weight products of IL-1 varies with the purity and/or quality of the monocyte preparation and the appropriate use of protease inhibitors during lysis (cf. Ref. 18). Since processing and secretion of IL-1 occurs in the absence of such processed intracellular forms, it is unlikely that such forms are intermediates in either the maturation or secretion of IL-1.

Although the precursors for both IL-la and IL-lB accu- mulate intracellularly, the intracellular half-lives of the two proteins are distinct: 15 and 3 h, respectively. The differences in the half-lives of IL-la and IL-lP do not reflect differences in the apparent intracellular compartmentalization of the two proteins since both PI forms of IL-1 are found in association with cytosolic and noncytosolic (particulate) fractions of monocytes, and the half-life of either IL-1 in each fraction is indistinguishable. Because we do not see either PI form of IL- 1 chase between the two compartments, our data do not support a clearly defined pathway of secretion through the particulate fraction, as had been suggested from activity data (29).

No direct evidence for a membrane-associated form of IL- 1 was seen in our fractionation studies of labeled monocytes. If present, membrane IL-1 must exist in minute amounts detectable by activity but not by metabolic labeling (30, 31). The amounts of IL-la and $3 associated with the particulate fraction are not insignificant and could easily be followed if these proteins were to have chased to the membrane. These data, therefore, are also not consistent with the proposed hypothesis that association of IL-1 with the lysosome leads to maturation of surface-associated IL-1 (29).

The difference in intracellular half-lives of IL-la and IL- 1 B is in part a consequence of the difference in kinetics with which the two proteins are secreted. As demonstrated in pulse- chase experiments, IL-1p is secreted from monocytes 2 h following synthesis, whereas the appearance of IL-la is de- layed further, first appearing in the culture medium of LPS- activated monocytes 12.5 h after the pulse label. This latter observation is consistent with data published by Giri et al. (16) using the P388D1 mouse macrophage cell line. After a 3- h pulse label of induced P388D1 cells, IL-la secretion was maximal at chase times of 9 and 15 h. Considering that the pulse time in these experiments was 3 h uersus the 1 h used in our studies, this time frame is nearly identical to the 12 h which we have observed for the secretion of IL-la from LPS- activated human monocytes. A similar analysis of secretion of IL-lB from monocytic cell lines has not been described.

Our data indicate monocytes secrete a number of proteins, including TNF-a, rapidly with kinetics which are consistent with the classical signal sequence-directed mechanism (24). A variety of other proteins are released slowly, with delayed kinetics similar to IL-1. It has been suggested that IL-1 leaks out of monocytes following activation due to compromised cell viability (28). Although these data do not formally pre- clude this mechanism for the release of these proteins, the observation that IL-la is secreted significantly later than IL- 1p suggests that distinct selective mechanisms exist for the secretion of the two pl forms of IL-1. The fact that 50% of the intracellular IL-la can be chased from the cell to the supernatant implies a relatively efficient process that is not compatible with a loss of viability throughout the experiment. Furthermore, with regard to the data on the secretion of IL- la, we do not see a disproportionate increase in the release of other 35S-labeled proteins concurrent with the secretion of IL- la. One would expect this to be the case if the release of IL- la were due to decreased cell viability or all lysis. However,

Secretion and Processing of Interleukin 1 (Y and 1 p in Monocytes a m we cannot completely rule out that IL-la release is due to a small number of lysed cells.

Both the 31- and 17-kDa forms of IL-lP appear in the supernatant of activated monocytes with little or no obvious preference. Maturation of IL-1, therefore, is not required for secretion. This might suggest that processing is a postsecre- tory event; however, we found no evidence of a precursor- product relationship between the two secreted forms of IL- lp. Incubation of labeled IL-lP precursor with cell lysates or supernatants indicated that processing activity resided in the cells. Because we failed to observe any mature IL-1P inside cells in the pulse-labeling studies, these results imply that the maturation of IL-1 occurs during secretion.

Preliminary fractionation studies indicate that IL-1 proc- essing activity is not in the particulate fraction, which would be expected if secretion occurred via a lysosomal or endo- plasmic reticulum pathway. Only the IL-1 retained in the cytosolic fraction is processed correctly, suggesting that the processing activity may be a cytosolic or peripheral membrane protein.

Apart from specifically localizing the processing activity to either of these fractions, future studies will have to address how this activity is regulated so that only the IL-1 that is being secreted is processed.

The observation that the kinetics of the secretion for IL- la and IL-1p differ so dramatically is somewhat surprising given their broad spectrum of similar properties (14). The regulation of IL-la and IL-lp activity in vivo may, therefore, be partly achieved at the level of their release from the cell. It has remained an issue why two separate molecules should have evolved independently, yet remain functionally con- served. The difference in the secretion of IL-la and IL-lP is one of a few characterized distinctions between these two monokines. Other differences have been observed at the level of transcription (32), membrane localization (33), and the activity of the precursor forms of these molecules (23). Further studies will be required to assess the relative importance of these differences in the biology of IL-1 and the reason for the evolution of two distinct gene products.

Acknowledgments-We would like to thank Philip Simon for pro- viding the IL-la and IL-lS antisera, Mann-Jy Chen for the TNF-a and TNF-a antiserum, Jay Lillquist and Linda Rebar for technical assistance, Elaine Jones for careful reading of the manuscript, and Deborah Murdoch for typing of the manuscript.

REFERENCES 1. Dinarello, C. A. (1984) Rev. Infect. Dis. 6 , 51-95 2. Matsushima, K., Procopio, A., Abe, H., Scala, G., Ortaldo, J. R.,

and Oppenheim, J. J. (1985) J. Zmrnunol. 135,1132-1136 3. Fontana, A., Kristensen, F., Dubs, R., Gemsa, D., and Weber, E.

(1982) J. Zmmuml. 129 , 2413-2419 4. Luger, T. A., Stadler, B. M., Katz, S. I., and Oppenheim, J. J.

(1981) J. Zmmunol. 127, 1493-1498 5. March, C. J., Mosley, B., Larsen, A., Cerretti, D. P., Braedt, G,,

Price, V.. Gillis, S., Hennev. C. S.. Kronheim. S. R.. Grabstein.

K., Conlon, P. J., Hopp, T. P., and Cosman, D. (1985) Nature

6. Auron, P. E., Webb, A. C., Rosenwasser, L. J., Mucci, S. F., Rich, A., Wolff, S. M., and Dinarello, C. A. (1984) Proc. Nutl. Acad. Sci. U. S. A. 8 1 , 7907-7911

7. Gray, P. W., Glaister, D., Chen, E., Goeddel, D. V., and Pennica, D. (1986) J. Zmmunol. 137, 3644-3648

8. Furutani, Y., Notake, M., Yamayoshi, M., Yamagishi, J.-I., No- mura, H., Ohue, M., Ruruta, R., Fukui, T., Yamada, M., and Nakamura, S. (1985) Nucleic Acids Res. 13,5869-5882

9. Lomedico, P. T., Gubler, U., Hellmann, C. P., Dukovich, M., Giri, J. G., Pan, Y-C. E., Collier, K., Semionow, R., Chua, A. O., and Mizel, S. (1984) Nature 312,458-462

10. Bayne, E. K., Rubb, E. A., Limjuco, G., Chin, J., and Schmidt, J. A. (1986) J. Exp. Med. 163,1267-1280

11. Schmidt, J. A. (1984) J. Exp. Med. 160, 772 12. Clark, B. D., Collins, K. L., Gandy, M. S., Webb, A. C., and

13. Furutani, Y., Notake, M., Fukui, T., Ohue, M., Nomura, H.,

315,641-647

Auron, P. E. (1986) Nucleic Acids Res. 14, 7897-7914

Yamada, M., and Nakamura, S. (1986) Nucleic Acids Res: 14 , 3167-3179

14. Oppenheim, J. J., Kovacs, E., Matsushima, K., and Durum, S. K.

15. Dower, S. K., and Urdal, D. L. (1987) Zmmunol. Today 8, 46-51 16. Giri, J. G., Lomedico, P. T., and Mizel, S. B. (1985) J. Zmmunol.

17. Limjuco, G., Galuska, S., Chin, J., Cameron, P., Boger, J., and Schmidt, J. A. (1986) Proc. Nutl. Acad. Sci. U. S. A. 8 3 , 3972- 3976

18. Auron, P. E., Warner, S. J. E., Webb, A. C., Cannon, J. G., Bernheim, H. A., McAdam, K. J. P. W., Rosenwasser, L. J., LoPreste, G., Mucci, S. F., and Dinarello, C. A. (1987) J. Zmmunol. 138,1447-1456

19. Meyers, C. A., Johanson, K. O., Miles, L. M., McDevitt, P. J., Simon, P. L., Webb, R. L., Chen, M-J., Holskin, B. P., Lillquist, J. S., and Young, P. R. (1987) J. Bwl. Chem. 262,11176-11181

20. Colotta, F., Peri, G., Villa, A., and Mantovani, A. (1984) J. Immunol. 132,936-944

21. Laemmli, U. K. (1970) Nature 227,680-685 22. Matsushima, K., Taguchi, M., Kovacs, E. J., Young, H. A., and

Oppenheim, J. J. (1986) J. Zmmunol. 136,2883-2891 23. Mosley, B., Urdal, D. L., Prickett, K. S., Larsen, A., Cosman, D.,

Conlon, P. J., Gillis, S., and Dower, S. K. (1987) J. Biol. Chem.

(1986) Zmmunol. Today 7, 45-56

134,343-349

262,2941-2944 24. Tartakoff, A. M. (1986) EMBO J. 5 , 1477-1482 25. Wang, A. M., Creasey, A. A., Ladner, M. B., Lin, L. S., Strickler,

J., Van Arsdell, J. N., Yamamoto, R., and Mark, D. F. (1985) Science 228,149-154

26. Pennica, D., Nedwin, G. E., Hayflick, J. S., Seeburg, P. H., Derynck, R., Palladino, M. A., Kohr, W. J., Aggarwal, B. B., and Goeddel, D. V. (1984) Nature 312 , 724-729

27. Mizel, S. B., and Rosenstreich, D. L. (1979) J. Zmmunol. 122 , 2173-2179

28. Geri, I., and Lepe-Zuniya, J. L. (1983) in Lymphokines (Pick, E., ed) Vol. 9, pp. 109-126, Academic Press, Orlando, FL

29. Bakouche, O., Brown, D. C., and Lachman, L. B. (1987) J. Zmmunol. 138,4249-4255

30. Kurt-Jones, E. A., Beller, D. I., Mizel, S. B., and Unanue, E. R. (1985) Proc. Natl. Acad. Sci. U. S. A. 8 2 , 1204-1208

31. Kurt-Jones, E. A., Virgin, H. W., IV, and Unanue, E. R. (1985) J. Zmmunol. 135,3652-3654

32. Lomedico, P. T., Killian, P. L., Gubler. U.. Stern. A. S.. and Chizzonite, R. (1986) Cold Spring Harbor Symp. Quant.'Biol. 5 1 - 681 -689 ", --- ---

33. Conlon, P. J., Grabstein, K. H., Alpert, A., Prickett, K. S., Hopp, - . T. P., and Gillis, S. (1987) J. Zmmunol. 139, 98-102