The Levels and Biologic Action of the Human Neutrophil Granule Peptide HP-1
in Lung Tumors
A N D R E W BATEMAN, *~ AVA SINGH,* SERGE J O T H Y , t R I C H A R D FRASER,'~ F R E D ESCH:~ A N D S A M U E L SOLOMON*
*Endocrine Laboratory, Royal Victoria Hospital Departments of Medicine, Obstetrics and Gynecology ¢Department of Pathology, McGill University, Montreal Quebec H3A IA1, Canada
~:Athenna Neurosciences, Inc., San Carlos, CA 94070
Received 28 January 1991
BATEMAN, A., A. SINGH, S. JOTHY, R. FRASER, F. ESCH AND S. SOLOMON. The levels and biologic action of the human neutrophilgranulepeptide HP-I in lung tumors. PEPTIDES 13(l) 133-139, 1992.--HP- 1 is the most abundant human representative of a recently discovered class of neutrophil cystine- and arginine-rich peptides. These peptides have many potentially regulatory activities expressed at nanomolar concentrations. To establish the levels of HP-l that can accumulate in human lung tumors and nondiseased lung fragments, tissues were extracted for their peptide content. The extracts were purified on reverse phase HPLC, and HP-I and related peptides were identified by sequence analysis and their concentrations in the tissue quantitated by amino acid analysis. Immunohistochemistry was performed and strongly suggests that HP-I is confined to granulocytes under most circumstances, and indicates that the levels of HP-I measured in the tumors reflect the levels obtained when solid tissue is infiltrated by neutrophils. The maximum observed levels were 26 nanomoles per gram wet weight of tissue. Attempts were then made to correlate this level to the cytotoxic potential of HP-l by performing in vitro cytotoxicity dose-response curves on several cell lines. Most cells were killed at between I and 8 ttM, and the response depended on the growth conditions of the cells. The levels of HP-I that accumulate in tumors can exceed the in vitro cytolytic concentrations. The levels are also considerably in excess of those required to exert in vitro regulatory actions.
THE inappropriate or excessive expression of polymorphonu- clear leukocyte (PMN) function following the infiltration of tissue is involved in the pathology of many clinical conditions (18). Several molecular species may be involved in the damage to tissue including short-lived radicals, arachidonic acid metabo- lites, and the proteolytic enzymes of the primary and secondary PMN granules. It has been known for almost a quarter of a century that PMN granules contain nonenzymatic cationic pep- tides that are likely to be important in the destruction of ingested microorganisms (19). Only recently, however, have these peptides come under close structural and biological scrutiny. They have been variously called defensins (7) and corticostatins (21), and are between 29 and 33 amino acids in length. The structures vary markedly both between and within species, but all are char- acterized by a conserved cystine backbone of six cysteinyl resi- dues linked in three disulfide bonds. In man, four such peptides have been described: HP-1, HP-2, HP-3, and HP-4. HP-1, 2, and 3 differ only at the first amino terminal residue (14). HP-4, on the other hand, shows little direct sequence similarity with
HP-l, etc., except at the invariant defensin/corticostatin back- bone residues (15).
The biological actions of these peptides are diverse. At mi- cromolar levels they are highly effective as antibiotic agents against both gram negative and gram positive bacteria (9). They also lyse fungi, and, as has been reported by several groups, including ourselves, they can lyse mammalian cells in vitro (10,20). The mechanism of the antibiotic action has been studied in some detail and appears to involve an interaction of the pep- tide with the lipid component of cellular membrane (8). As well as the nonspecific toxicity reactions, peptides of the defensin/ corticostatin family have a number of other very specific bio- logical activities that are expressed at nanomolar concentrations. We first reported that some of these peptides are able to inhibit the action of the hormone adrenocorticotropin on the synthesis of glucocorticoids in adrenal cell suspensions (15,20-22). This action is very specific since other adrenal cortical stimulators, such as angiotensin II, are not inhibited. In humans, HP-4, but not HP-1, has an anti-ACTH action. It has been reported that
J Requests for reprints should be addressed to Andrew Bateman, Endocrine Laboratory, Royal Victoria Hospital, 687 Pine Ave West, Montreal, Quebec, Canada H3A 1Al.
133
134 BATEMAN ET AL.
<
~ r i i i i
60 120 180
TIME (rain)
FIG. 1. The HPLC purification of a human lung tumor extract (A) and an equivalent extract of undiseased tissue from the same patient (B). The tumor fragment had a wet weight of 1.35 g and the undiseased lung sample of 0.52 g.
HP-1 may be a monocyte chemotactic agent (17), and we have found that HP-1 appears to have a cytostatic effect on HL60 promyelocyte-like cells in vitro (1). Some corticostatins, includ- ing HP-4 but not HP-1, are able to stimulate enterocyte nifedi- pine-sensitive L-type calcium channels in vitro (11). These activities have EDs0s in the range of 10 -1° to 10 -8 M, which suggests that they may be physiologically relevant. In this report we have looked at the levels of liP-1 that accumulate in tumors, and we will attempt to correlate these levels with the concen- trations of HP-1 which illicit cellular growth and cell death of a number of lung tumor cell lines and related cells.
M E T H O D
Tissue Extraction and Purification
Biopsy samples of human lung tumors and, in each instance, undiseased lung tissue were obtained at the time of surgery from the same patient and stored at -40°C. The tissue was then defatted by homogenization in ice-cold acetone and the protein concentrated by centrifugation. The dried pellet was homoge- nized in 25 ml of acidic high salt medium [1 M, HC1, 5% formic acid, 1% NaC1, and 1% trifluoroacetic acid (TFA)] and centri- fuged at 2000 × g for 15 min at 4°C. The pellet was reextracted twice and the pooled supernatant passed through 30DS-silica cartridges (Sep-Pak Cl 8 cartridges, Waters, Milford, MA) and the peptides eluted from each cartridge with 4 m180% acetonitrile containing 0.1% TFA (3). Reverse phase HPLC purification of the extract was accomplished with a Cl8 Waters aBondapak column using a 3-h gradient from 0 to 80% acetonitrile in 0.1%
TFA. The fraction corresponding to HP-I was further purified using a 60-min linear gradient of 0% to 65% acetonitrile in 0.13% heptafluorobutyric acid (HFBA) at 1.5 ml/min. The purified material was then rechromatographed on the same TFA solvent system as above to give an apparently homogenous peak.
For cytotoxicity experiments, HP-1 was isolated from neu- trophils obtained from the peritoneal exudate of patients with peritonitis. Fresh fluid from a single patient was centrifuged at 10°C for 15 min at 200 × g. The pellet was washed three times in 10 ml of isotonic saline and a cell count was obtained using a hemocytometer. An aliquot of cells was also stained with Wrights stain to distinguish PMNs from mononuclear cells. The cells were then concentrated by centrifugation (200 × g) and sonicated in 50 ml of the acidic extraction medium. The extract was spun at 2000 × g for 10 min and the supernatant saved. HP- 1 was then purified from the extract by reverse phase HPLC using a 3-h liner gradient of 0 to 80% acetonitrile in 0.1% TFA throughout at 1.5 ml/min 1. After the first HPLC step, the frac- tions corresponding to HP- 1 were purified to homogeneity using a second linear HPLC gradient eluted with 28% to 38% aceto- nitrile in 0.1% TFA throughout, for 90 min at an elution rate of 1.5 ml/min. Purity was judged by amino acid analysis, and when necessary the final HPLC step was repeated until the ma- terial was homogeneous.
Amino Acid Analysis and Gas Phase Sequencing
Amino acid composition analysis was performed using a Beckman 6300A High Performance Analyser, as previously de-
L E V E L S A N D B I O L O G I C A C T I O N O F HP-1 IN L U N G T U M O R S 135
The content of HP-1 (nanomoles/g wet weight) is listed with the classification of the tumor and some of its histological characteristics. The extent of necrosis, fibrosis, polymorphonuclear infiltrate and lymphocytes and plasma cells observed in histological sections of these specimens are graded on a scale of 1 to 3. 1 is considered low and 3 is high. The grading was done without prior knowledge of which samples contained elevated levels of the peptide HP- 1. There was no evidence of abeesses in the slides examined.
136 BATEMAN ET AL.
(A) (B)
FIG. 2. Immunostaining with anti-HP-I serum of (A) a peripheral blood smear and (B) a lung tissue secion. The alveolar epithelial and mesenchymal cells are unreactive. A polymorphonuclear leukocyte in the lumen of the capillary is strongly positive. The cytoplasm of the polymorphonuclear cells in the peripheral blood smear is strongly positive. In both cases the bar indicates 30 micrometers.
scribed (15). Gas phase sequencing was performed on the reduced and alkylated peptides as previously described (5).
Antiserum Against HP-1
Antiserum against HP-1 was prepared in rabbits using HP-1 conjugated to key-hole limpet hemocyanin as described else- where (13). The antiserum does not cross-react with rabbit cor- ticostatins or with HP-4. Details of the preparation and char- acteristics of this antibody have been published (2).
Fixation and Staining of Tissues
A smear of peripheral blood was air dried. Tissue blocks of normal lung, kidney, colon, and lymph nodes were obtained from surgical specimens. They were fixed in formalin, embedded in paraffin and sectioned according to standard procedures. In addition, tissue blocks from an adenocarcinoma, squamous cell carcinoma, and oat cell carcinoma were also taken for exami- nation. Embedded tissue sections were examined by immuno- cytochemistry using the HP- 1 antibody at 1/z5-1/2oo dilutions and the peroxidase-antiperoxidase technique (16). Nonimmune rab- bit serum was used as a control.
Cell Growth and Cell Number
All cell lines were obtained from the American Type Culture Collection (ATCC, Rockville, MD). Thymidine incorporation
was performed as follows (4). A total of 0.8 × 105 cells per well ofa 24-multiwell plate were incubated for 24 h in various media as recommended by ATCC and supplemented with 10% fetal calf serum (Gibco, Grand Island, NY). The medium was then removed, the monolayer was washed twice in serum-free me- dium, and the cells were then incubated with increasing con- centrations of peptide for 25 h. Then the cells were incubated with 5 ttCi of 3H-thymidine in 1 ml of serum-free medium per well (ICN Radiochemicals, Irvine, CA) for 1 h at 37°C. After incubation, the medium was removed and washed seven times with ice-cold Hank's balanced salt solution. The cells were then washed for 10 min in 1 ml/well of ice-cold 0.75 M trichloroacetic acid and were then hydrolyzed in 0.75 M trichloroacetic acid at 70°C for 2 h and the mixture counted in a scintillation counter. The effects of peptides on cell number were determined as fol- lows: cells were grown and incubated with peptide as above, the monolayers were trypsinized with trypsin-EDTA containing 0.05% trypsin and 0.02% EDTA (Gibco), and cell number was determined using a Coulter particle counter.
RESULTS
The reversed phase HPLC chromatographic profiles of pep- tides from an extract of a lung tumor and undiseased lung taken from the same patient are shown in Fig. 1A and B, respectively. The lung tumor shows a component eluting at 68 min, indicated by a solid bar, that is clearly absent in the undiseased tissue.
LEVELS A N D BIOLOGIC A C T I O N OF HP-1 IN L U N G T U M O R S 137
TABLE 3
PERCENTAGE CHANGE IN 3H-THYMIDINE INCORPORATION AND CELL NUMBER IN RESPONSE TO HP-I
Peptide Concentration nmol/ml
Cell Line N 0.25 0.5 1.0 2.0 4.0 8.0
A. Percentage Change in 3H-Thymidine Incorporation in Response to HP-I
The cytotoxic effect of HP-1 on 3H-thymidine incorporation relative to controls and cell number relative to untested controls. Numbers in parentheses are + SEM; N is the number of experiments. The cells were incubated in medium containing 10% fetal calf serum and the incubation with peptide was performed in serum-free medium for 24 h as described in the Method section.
SKMES-1: human squamous carcinoma of the lung. CHO-KI: Chinese hamster ovary. A549: human adenocarcinoma of the lung. WI-26: human lung--fibroblast-like. WI-38: human lung--virus transformed. CALU-I: epidermoid carcinoma--human lung. CALU-3: adenocarcinoma--human lung. CALU-6: anaplastic carcinoma--lung. 3T3: Swiss albino mouse.
Amino acid analysis of this peptide is shown in Table 1, where it is compared with the analysis for HP-1 isolated from human neutrophils. To confirm that both peptides were identical, they were sequenced by gas phase microsequencing which established that the tumor peptide had the same sequence as human neu- trophil HP- 1 and consisted of a 3:1 mixture of HP- 1 (1-30) and des-alanine HP-l(2-30) . Fourteen lung tumors and fragments of undiseased lung removed with the tumor were screened in this way. Table 2 summarizes the HP- I content of these tumors. Four tumors had appreciable amounts of HP-1, at levels between 10 and 26 nanomoles per gram wet weight. The level of HP-I in the undiseased lung fragments was never sufficiently high to be detected with these techniques.
Immunostaining was used to identify cells containing HP-1. Peripheral blood showed a strong cytoplasmic reaction in the cytoplasm of all polymorphonuclear cells, including eosinophils, which were readily identified on the basis of their granule mor- phology (Fig. 2). The immunohistochemieal testing of the various normal tissue specimens displayed no evidence of staining in the various types of epithelial and mesenchymal cells. All sub- populations of lymphocytes present in the lymph node were
unreactive. As expected from the results obtained with the pe- ripheral blood, there was a strongly positive reaction of the PMNs which were present in the lumens of the blood vessels and cap- illaries (Fig. 2). There was a positive reaction with the PMNs when the squamous cell carcinoma and oat cell carcinoma sec- tions were examined, and no other cells were seen to be positive. We have earlier reported that HP-1 is cytotoxic to mammal ian cell lines in vitro (20). The value of 26 nanomoles per gram of tissue sets an experimentally determined upper limit on the amount of HP-1 available to elicit potential in vivo cytotoxicity. We have extended our initial cytotoxicity studies to obtain a better understanding of the dose-response characteristics of HP- l cytotoxicity. Over half the cell lines tested showed a marked decrease in cell number and 3H-thymidine incorporation at concentrations between 2 and 8 tzM. These results are sum- marized in Table 3. The dose-response curve for the death of cells in response to HP-1 is very sharp, in some cases going from no cell death to almost complete cell death in as little as a dou- bling of the HP- l concentration. In our initial studies, we ob- served that NIH 3T3 cells differed in their response to HP-I from other cells tested in showing a marked increase in 3H-
138 BATEMAN ET AL.
120 ill c -
oo ~° 8 0 " , , ~ ~ , .c .c_ E
~ 4 0 I -I-
i m (
0 0 . ? 5 025 I 2 4 ; concentration of HP-1 (nmol/ml)
FIG. 3. The effect of different media formulations on the response of A549 cells to HP- I. Cells were grown in RPMI- 1640 (open circles), Dul- becco's Modified Eagles Medium (open squares), and McCoy's Medium (closed circles). The effect of HP-1 on 3H-thymidine incorporation was determined as described in the Method section. Results are the mean of four experiments in duplicate _+ SEM.
thymidine incorporation at concentrations which killed other cell lines. One other cell line was found to respond in this way, namely CALU-6. Two cell lines, Y-1 and HEC-IA, did not re- spond to HP-1 even at concentrations of 16 #M (data not shown).
To eliminate the possible harmful effect of residual serum proteases which could reduce the effective concentration of HP- 1, we did the following control experiment. A549 and CHO-KI cells were incubated for 24 h in Hams FI2 medium supple- mented with 10 #g insulin, 20 mg transferrin, 3.52 mg ascorbic acid and 40/~M sodium selinite per liter. To this medium was either added or omitted 100 mg fatty acid-free BSA. Wells were covered with or without 500 #1 fibronectin which is equivalent to 0.5 mg protein (8 mM/ml). The results obtained are as follows:
The conclusion drawn from this study is that fibronectin has no effect on HP-1 action and BSA inhibits HP-1 action.
The differences in the response of cells to HP-1 may be in- trinsic, or reflect the conditions under which the assay was con- ducted. To investigate this further, A549 cells were assayed in
different media (Fig. 3); the dose-response curve shifts repro- ducibly according to which media was used, but the type of response remains the same. When the thymidine incorporation experiment was repeated using 3T3 cells synchronized by me- dium depletion, the increase in thymidine incorporation over untreated controls was about ten-fold (data not shown). The possibility was therefore considered that the increase in thymi- dine incorporation seen in 3T3 cells and CALU-6 might reflect their low DNA synthesis under the conditions of the assay. This was tested by depleting CALU-I cells of nutrient by preincu- bating them for 72 h prior to the assay, which halves their thy- midine incorporation per hour. CALU-1 cells were chosen be- cause of all the cells killed by HP-1 they had the lowest intrinsic rate of thymidine incorporation. Under these conditions there is a greater increase in relative 3H-thymidine incorporation in the medium-depleted cells (Fig. 4).
DISCUSSION
Defensin/corticostatins are PMN peptides that are likely to be involved in the nonoxidative killing of ingested microorgan- isms. Several extracellular activities have also been attributed to these peptides, but the physiological role of these activities is less clear. PMNs are implicated in many clinical conditions so it is important to clarify which, if any, of these activities are likely to occur in vivo. We have begun by asking the question, how much HP-I accumulates in tissues infiltrated by neutro- phils? Lung tumors were chosen because it was relatively easy to obtain control samples of undiseased tissue removed from the same patients at the time of biopsy.
The maximum tissue level of HP-1 that we observed was 26 nanomoles per gram wet weight of tissue. The identity of this peptide with HP-1 extracted from PMNs was established by mi- crosequencing. Immunocytochemical studies suggest that the only likely source of this peptide in the tumor is from PMN infiltration. During phagocytosis, between 2 and 8% of the PMN HP-1 is released into the extracelluar medium (12). In a tissue
C A L U - 1 500-
z o_
400
0 300 0 Z IJJ _Z - 200
100 .....
0 i i i i 1 i
0.25 0.5 1.0 2.0 4.0 8.0 nmols/rnl
FIG. 4. The effect of HP-I on LH thymidine in CALU-1 cells. Cells were plated and grown for 3 days (open circles) or 24 h (closed circles) prior to the addition of HP-1. Results are the mean of 5 experiments __. SEM.
LEVELS AND BIOLOGIC ACTION OF HP-I IN L U N G TUMORS 139
sample containing 25 nanomoles of HP-1, and with actively phagocytosing PMNs, these figures suggest the m i n i m u m release of between 0.5 and 2.0 nanomoles of HP-1. If cell lysis is taken into account this value may be higher. The reported monocyte chemotactic effect of liP-1 is observable at 10 -1° M HP-1 in the presence of serum (17). The amounts of l iP- 1 likely to be released are therefore considerably in excess of the amounts required to elicit an in vitro paracrine monocyte chemotactic response. HP- 4 occurs at about 1% the level of HP-1, and activates calcium channels in vitro with an EDso of about 10 -9 M in the presence of serum albumin (11). Thus, on the basis of the calculations above, it is feasible that sufficient HP-4 is released from PMNs infiltrating tissue to elicit a paracrine effect on calcium channels. It should be pointed out, however, that the fate of defensins/ corticostatins released from PMNs in tissue is not known, and mechanisms may exist to immobilize or inactivate them.
An amount of 2 nanomoles of HP- 1 per ml is sufficient to lyse some cell lines in vitro. The lytic effect of HP-1 is inhibited by serum (data not shown), so that if HP-I is involved in cell necrosis in vivo it most likely requires close contact between the target cell and the neutrophil. The effect of HP-1 on cells in vitro depends on the conditions of the assay (Figs. 2 and 3). This could be quite important in vivo if the ability of a cell to with- stand HP-1 is also dependent on its condition, for example, its nutritive state, whether it is actively dividing, etc.
Immunocytochemical studies revealed the presence of HP- 1 immunoreactivity in eosinophils as well as neutrophils. This surprising finding contradicts earlier reports from other workers
(7) that HP-I is confined to neutrophils, but is in agreement with the results of hybridization experiments with HP-1 cDNA (12). Several explanations for this apparent discrepancy can be suggested. It is possible that HP- 1 is processed differently in neu- trophils and eosinophils, or that eosinophils contain a peptide different but similar to HP-1. Control experiments with non- immune rabbit serum eliminate the possibility that the staining seen in eosinophils is an artifact of eosinophilic peroxidase ac- tivity. We have also isolated and purified eosinophil cationic peptide, the major low molecular weight cationic polypeptide of eosinophils, and shown that it does not cross-react with our anti-HP-1 antiserum. Resolving the identity of the eosinophilic HP-l-like immunoreactivity will be difficult because the slightest contaminat ion with neutrophils will introduce uncertainty, but it is an important problem because of the role eosinophilic cy- tolytic polypeptides are thought to have in many clinical dis- orders.
In conclusion, it can be stated that the levels of HP-1 that accumulate in the infiltrated tissue can exceed the amounts re- quired to elicit a cytolytic effect in vitro, and it is possible, al- though not proven, that HP-1 may contribute to cell necrosis in these tissues. It should be pointed out that not all lung tumors contain HP- 1.
ACKNOWLEDGEMENTS
The authors wish to acknowledge the expert technical assistance of Neola Matusiewicz. This work was supported by operating grants MT- 1658 from the Medical Research Council of Canada and HDO-4365 from the National Institute of Child Health and Human Development.
REFERENCES
1. Bateman, A.; Singh, A.; Congote, L. D.; Solomon, S. The effect of HP-1 and related granule peptides on DNA synthesis in HL60 cells. Regul. Pept. 35:135-143; 1991.
2. Bateman, A.; Singh, A.; Shustick, C.; Mars, W. M.; Solomon, S. Multiple forms of neutrophil granule peptides in human leukemic cells. J. Biol. Chem. 266:7524-7530; 1991.
3. Bennett, H. P. J.; Browne, C. A.; Solomon, S. Purification of the two major forms of rat pituitary corticotropin using only reversed- phase liquid chromatography. Biochemistry 20:4530-4538; 1981.
4. Congote, L. F. Similarities in structure and function of bovine serum erythrotropin and human insulin-like growth factor II: Two fetal erythroid cell stimulating factors. Biochem. Biophys. Res. Commun. 126:653-659; 1985.
5. Esch, F. Polypeptide microsequence analysis with the commercially available gas-phase sequenator. Anal. Biochem. 136:39-47; 1984.
6. Ganz, T. Extracellular release of antimicrobial defensins by human polymorphonuclear leukocytes. Infect. Immun. 55:568-572; 1987.
7. Ganz, T.; Selsted, M. E.; Szklareg, D.; Harwig, S. S. L.; Daher, K.; Bainton, D. F.; Lehrer, R. I. Defensins, natural peptide antibiotics of human neutrophils. J. Clin. Invest. 76:1427-1435; 1985.
8. Kagan, B. L.; Selsted, M. E.; Ganz, T.; Lehrer, R. I. Antimicrobial defensin peptides from voltage-dependent ion-permeable channels in planar lipid bilayer membranes. Proc. Natl. Acad. Sci. USA 87: 210-214; 1990.
9. Lehrer, R. I.; Ganz, T.; Selsted, M. E.; Babior, B. M.; Curnette, J. T. Neutrophils and host defense. Ann. Intern. Med. 109:127-145; 1988.
10. Lichtenstein, A.; Ganz, T.; Selsted, M. E.; Lehrer, R. I. In vitro tumor cell cytolysis mediated by peptide defensins of human and rabbit granulocytes. Blood 68:1407-1410; 1986.
11. MacLeod, R. J.; Hamilton, J. R.; Bateman, A.; Belcourt, D.; Hu, J.; Bennett, H. P. J.; Solomon, S. Corticostatic peptides cause ni- fedipine sensitive volume reduction in jejunal villus enterocytes. Proc. Natl. Acad. Sci. USA 88:552-556; 1991.
12. Mars, W. M.; Stellrecht, C. M.; White, J. W.; Saunders, G. F. Lo- calization of an abundant myeloid-related sequence. Acta Haematol. (Basel) 78(Suppl. 1):11-17; 1987.
13. Richardson, C. D.; Berkovich, A.; Rozenblatt, S.; Bellini, W. J. Use of antibodies directed against synthetic peptides for identifying cDNA clones, establishing reading frames, and deducing the gene order of measles virus. J. Virol. 54:186-193; 1985.
14. Selsted, M. E.; Harwig, S. S. L.; Ganz, T.; Schilling, J. W.; Lehrer, R. 1. Primary structures of three human neutrophil defensins. J. Clin. Invest. 76:1436-1439; 1985.
15. Singh, A.; Bateman, A.; Zhu, Q.; Shimasaki, S.; Esch, F.; Solomon, S. Structure of a novel human granulocyte peptide with anti-ACTH activity. Biochem. Biophys. Res. Commun. 155:524-529; 1988.
16. Sternberger, L. A.; Hardy, P. H.; Cuculis, J. J.; Meyer, H. G. The unlabelled antibody enzyme method of immunohistochemistry. J. Histochem. Cytochem. 18:315-333; 1970.
17. Territo, M. C.; Ganz, T.; Selsted, M. E.; Lehrer, R. I. Monocyte chemotactic activity ofdefensins from human neutrophils. J. Clin. Invest. 84:2017-2020; 1989.
18. Willoughby, D. A. Inflammation--Mediators and mechanisms. London, UK: Churchill Livingstone; 1987.
19. Zeya, H. I.; Spitzanagel, J. Arginine-rich proteins of polymorpho- nuclear leukocyte lysosomes. J. Exp. Med. 127:927-941 ; 1968.
20. Zhu, Q.; Singh, A.; Bateman, A.; Esch, F.; Solomon, S. The corti- costatic (anti-ACTH) and cytotoxic activity of peptides isolated from fetal, adult, and tumor-bearing lung. J. Steroid Biochem. 27:1017- 1022; 1987.
21. Zhu, Q.; Hu, J.; Mulay, S.; Esch, F.; Shimasaki, S.; Solomon, S. Isolation and structure of corticostatin peptides from rabbit fetal and adult lungs. Proc. Natl. Acad. Sci. USA 85:592-595; 1987.
22. Zhu, Q.; Bateman, A.; Singh, A.; Solomon, S. Isolation and biological activity of corticostatic peptides (anti-ACTH). Endocr. Res. 15:129- 150; 1989.
