- Email: [email protected]
Anti-HIV drugs: Comparative toxicities in murine fetal liver and bone marrow erythroid progenitor cells
Life Sciences, Vol. 45, pp. iii-vii Printed in the U.S.A.
Pergamon Press
AIDS RESEARCH COMMUNICATIONS
ANTI-HIV DRUGS: COMPARATIVETOXICITIES IN MURINE FETAL LIVER AND BONEMARROWERYTHROIDPROGENITORCELLS
Sudhir R. Gogu, Barbara S. Beckman and Krishna C. Agrawal
Department of Pharmacology, Tulane University School of Medicine New Orleans, LA 70112 (Received in final form May 19, 1989)
Summary The t o x i c i t y of anti-HIV drugs, 3'-azido-3'-deoxythymidine (zidovudine, AZT), 2',3'-dideoxycytidine (DDC), 2',3'-dideoxy2',3'-didehydrothymidine (d4T) and ribavarin was studied i n vitro in murine fetal liver cells (FLC) and in bone marrow cells. These studies indicate that d4T is the least toxic drug and ribavarin is the most toxic agent in both models. However, the murine FLC system was found to be a more sensitive model for the assessment of t o x i c i t y of anti-HIV agents towards erythroid progenitor cells as indicated by the ICso,values. Drug-induced hematopoietic toxicity has been a limiting factor in the administration of zidovudine (AZT) to AIDS patients ( I ) . Similar toxicities have been reported to occur with other anti-HIV agents in clinical t r i a l s suc~ as 2',3'-dideoxycytidine (DDC), 2',3'-dideoxy-2',3'-didehydrothymidine (d4T) and ribavarin (2). Anemia and neutropenia commonly develop in patients receiving different doses of these drugs (1,3). Bone marrow megaloblastic changes observed in some cases were attributed to AZT-induced depletion of normal pyrimidine pools (4). Hematopoietic growth factors such as human recombinant granulocytemacrophage colony stimulating factor (rGM-CSF) have been employed to overcome AZT-mediated biochemical perturbations. The in vitro studies have suggested that rGM-CSF p a r t i a l l y corrects AZT-induced cytotoxicity in normal human bone marrow myeloid progenitor cells (5). In i n i t i a l clinical t r i a l s rGM-CSFhas been shown to stimulate erythropoiesis (6). Similarly, recombinant erythropoietin, at present in clinical t r i a l s for anemia of end stage renal disease may have potential in inducing bone marrow regeneration in hypoplastic anemia whether primary or as a result of cytotoxic or other therapy (7). Therefore, i t was deemed desirable that an assay system that would be sensitive to the use of such growth factors and hormones would be a more appropriate model to assess the toxicities of various anti-AIDS drugs towards erythroid progenitor cells. Murine fetal l i v e r cells (FLC) are known to express erythropoietin receptors on the erythroid cells and were found to be highest at the CFU-E stage (8). We have therefore employed this technique to assess the toxicity of various anti-AIDS drugs towards erythroid progenitor cells in FLC which have been shown to express erythropoietin receptors. The results were simultaneously compared with toxicity data on murine bone marrow cells. 0024-3205/89 $3.00 + .00 Copyright (c) 1989 Pergamon Press plc
iv
Toxicity of Anti-HIV Drugs on BM and FLC
Vol. 45, No. 4, 1989
Materials and Methods
Mice - CD-I white mice (Charles River), 8-10 weeks old, and 13-14 days old f e t a l - l i v e r cells were employed in this study. All mice were cagedunder v i r a l - f r e e conditions. Drugs - Azidothymidine and d4T were synthesized in our laboratory and DDC and ribavarin were obtained from the National Cancer Institute (Bethesda MD). Erythropoietin (TCEpo) was supplied by Amgen Corp. (Thousand Oaks, CA.). Preparation of Murine Fetal Liver Cells - Cells were prepared aseptically from 13-14 day old fetuses by a modified technique (8). Cells were collected from each fetus and made into a single cell suspension by aspiration through pipettes followed by a 25 X 5/8 gauge needle. Cells were washed with alpha medium by centrifuging at 200 x g for 10 min. Cells were resuspended in alpha medium and counted in Turk's solution. The final cell concentration was adjusted to Ix1OQ/ml. Preparation of Bone Marrow Cells - Bone marrow cells were obtained from mouse femurs. Femurs were removed aseptically after anesthetizing the mouse with ether; marrow was flushed with a 25 X 5/8 gauge needle. A single cell suspension was made after repeated aspirations and counted in ~urk's solution. The final cell concentration was adjusted to I x 10°/ml. CFU-E Assay - Methylcellulose (1.3%) was prepared in RPMI-1640 containing L-glutamine (2.0 mM), fetal bovine serum (20%), 2-mercapto-ethanol (0.1 mM), p e n i c i l l i n and streptomycin (~00 U/ml and 100ug/ml) and erythropoietin (50 mU/ml). Cells (Ix10°/ml) were suspended in the methylcellulose mixture. One m i l l i l i t e r of methylcellulose containing cells was plated in duplicate and counted for erythroid colony formation ~CFU-E) after incubation in a humidified atmosphere of 5% C02-95% air at 37uC for 48 hr. The plates were stained with 3,3'-diaminobenzidTne and colonies containing 8 or more benzidine-positive cells were counted under an inverted microscope. Results and Discussion Hematological t o x i c i t y has been reported to be one of the major adverse effects of the various agents being evaluated for anti-HIV a c t i v i t y . AZT-induced anemia in patients treated for AIDS has been well documented. Similar toxic effects have been observed with other drugs such as DDC, d4T and ribavarin which are currently in clinical t r i a l s . Previous studies have demonstrated the toxic effects of these drugs both in v i t r o and in vivo on bone marrow hematopoietic progenitor cells. The mode of action for antiviral a c t i v i t y of AZT, DDC and d4T seem to be similar at the level of inhibition of reverse transcriptase a c t i v i t y since after being incorporated into DNA these agents act as chain terminators. Ribavarin on the other hand has been shown to block mRNAcapping. In an attempt to understand the mechanism of toxicitY of these drugs on bone marrow hematopoietic stem cells, we have compared the inhibitory effect on erythroid progenitor cells in bone marrow with the inhibition of erythropoietin-mediated stimulation of erythropoiesis in FLC which is d i r e c t l y related to heme synthesis. The data in Figures I-4 show that all four drugs i n h i b i t CFU-E formation in a dose dependent manner in both bone marrow and FLC systems. The IC50 values for AZT, DDC, d4T and ribavarin in bone marrow cells calculated from these data were 2.1, 8.0, 9.0 and 0.62UM respectively. In contrast, the corresponding ICso values in the FLC system were 1.0, 1.0, 2.6 and 0.23uM indicating that ~ e FLC system was more sensitive to the effect of these drugs
Vol.
45, No. 4, 1989
I.d
I
Toxicity
of A n t i - H l V
Drugs
on BM and FLC
1°°1
;:.) It,,. (.3
75-
"6
=,,.
0
50 ¢..
'E
25.
Q.,
00.1
0.5
1
5
10
50
Concentration of AZT (uM)
FIG. 1 Effect of various concentrations of AZT on murine bone marrow ( [] ) and fetal ~iver progenitor cells ( m ) by CFU-E assay. Results are mean + SE of 4 different experiments with duplicate determination for each point. 1 O0 W
I
"0
75
"6
.lr=
0
50 c-
•
25
III O.
0.1
0.5
1
5
10
50
Concentration of DDC (uM)
FIG. 2 Effect of various concentrations of DDC on murine bone marrow ( [] ) and fetal l i v e r progenitor cells ( [] ) by CFU-E assay. Results are mean ÷ SE of different experiments with duplicate determination for each point.
v
vi
Toxicity of Anti-HIV Drugs on BM and FLC
Vol. 45, No. 4, 1989
1 O0
','
1
I
0
"
75 l
E o
~-- 50 .E L,-
25.
0 0.1
0.5
1
5
10
100
Concentration of d4T (uM)
FIG, 3 Effect of various concentrations of d4T on murine bone marrow ( [] ) and fetal l i v e r progenitor cells ( • ) by CFU-E assay. Results are mean + SE of 4 different experiments with duplicate determinations for ~ch point. 1 O0 -
bJ
I
"0
75.
¢..
50c-
0
~
111
25-
i
0.05
0.1
0.5
1
10
Concentration of Ribavarin (uM)
FIG. 4 Effect of various concentrations of ribavarin on murine bone marrow ( [] ) and fetal l i v e r progenitor cells ( • ) by CFU-E assay. Results are mean + SE of 4 different experiments with duplicate determination for-each point.
Vol. 45, No. 4, 1989
Toxicity of Anti-HlV Drugs on BM and FLC
vii
on CFU-E formation than the bone marrow. The difference in the toxicities of these agents towards erythroid progenitor cells between the two models was significantly increased in the case of DDC which was found to be 8-fold more toxic to FLC. The increased t o x i c i t y shown by ribavarin in both the models may reflect the specificity of this drug in inhibiting the translation process perhaps at the level of mRNAand delaying the maturation process of late phase erythroblast. The clinical t r i a l s with ribavarin in patients with Lassa fever have shown that anemia occurred in about half of these patients at the dose levels of 20-60 mg/kg/day (9). Thus, the in vitro erythroid t o x i c i t y described in this report seem to correlate with c l i n i c a l l y observed erythroid myelosuppression. Although other reports (10-12) have been published on the t o x i c i t y of AZT to hematopoietic progenitor cells (both CFU-GM and BFU-E) the effects of AZT and other anti-HIV drugs on the CFU-E compartment, either in human or murine systems have not been f u l l y ascertained. We have examined the comparative toxicities of the four anti-HIV drugs which are in clinical t r i a l s except AZT, the only drug approved by FDA so far for the treatment of AIDS. The data suggests that d4T is the least toxic drug and ribavarin is the most toxic agent in both FLC and bone marrow cells. However, the murine FLC system was found to be a more sensitive model as indicated by the IC50 values for the assessment of t o x i c i t y of anti-HIV agents towards erythroid progenitor cells. This system should provide an appropriate test system for distinguishing the absolute toxicities of anti-HIV drugs for the erythroid progenitor cell compartment. Furthermore, the FLC system provides a model to further study the mechanism of t o x i c i t y which may be at the level of heme synthesis. Acknowledgement This work was supported by PHS grant number A125909 awarded by the NIAID, DHHS. References
I.
2. 3. 4. 5.
R. YARCHOAN,K.J. WEINHOLD, H.K. LYERLY, E. GELMANN, R.M. BLUM, G.M. SHEARER, H. MITSUYA, J.M. COLLINS, C.E. MYERS, R.W. KLECKER, P.D. MERKHAM, D.T. DURACK, S.N. LEHRMAN, D.W. BARRY, M.A. FISCHL, R.C. GALLOW, D.P. BOLOGNESI, S. BRODER, Lancet. I 575-580 (1986). M. BABA, R. PAUWELS, P. HERDEWIJN, E. CLERCQ, J. DESMYER, and M. VANDEPUTTE. Biochem. Biophys. Res. Commun. 142 128-134, (1987). R. YARCHOAN, S. BRODER. N. Engl. J. Med. 316 557-564, (1987). D.D. Richman, M.A. FISCHL, M.H. GREICO, M.S. GOTTLIEB, P.A. VOLBERDING, O.L. LASKIN, J.M. LEEDOM, J.E. GROOPMAN, D. MILDVAN, M.S. HIRSCH, G.G. JACKSON, D.T. DURACK, S.N. LEHRMAN. N. Engl. J. Med. 317 192-197 (1987). K. BHALLA, M. BIRKHOFER, S. GRANTand G. GRAHAM. Exp. Hematol. 17 17-20 (1989).
6.
7. 8. 9. 10. 11. 12.
J.E. GROOPMAN, R.T. MITSUYASU, M.J. DELEO, D.H. OETTE, D.W. GOLDE. N. Engl. J. Med. 317 593-598 (1987). P.M. COTES. Erythropoietin: The developing story. Br. Med. J. 296 805-806 (1988). H. FUKAMACHI, T. SAITO, A. TOJO, T. KITAMURA, A. URABE, and F. TAKAKU. Exp. Hematol. 15 833-837 (1987). H. FERNANDEZ, G. BANKS, and R. SMITH. Eur. J. Epidemiol. 2 1-14 (1986). J.P. SOMMADOSSI and R. CARLISLE. Antimicrob. Agents Chemother. 31 452-454 (1987). R.M. RUPRECHT, L.G. O'BRIEN, L.D. ROSSONI, S. NUSINOFF-LEHRMAN. Nature 323 467-469 (1987). J.P. SOMMADOSSI, and R. CARLISLE. Antimicrob. Agents Chemother. 31 23-26 (1988).
Pergamon Press
AIDS RESEARCH COMMUNICATIONS
ANTI-HIV DRUGS: COMPARATIVETOXICITIES IN MURINE FETAL LIVER AND BONEMARROWERYTHROIDPROGENITORCELLS
Sudhir R. Gogu, Barbara S. Beckman and Krishna C. Agrawal
Department of Pharmacology, Tulane University School of Medicine New Orleans, LA 70112 (Received in final form May 19, 1989)
Summary The t o x i c i t y of anti-HIV drugs, 3'-azido-3'-deoxythymidine (zidovudine, AZT), 2',3'-dideoxycytidine (DDC), 2',3'-dideoxy2',3'-didehydrothymidine (d4T) and ribavarin was studied i n vitro in murine fetal liver cells (FLC) and in bone marrow cells. These studies indicate that d4T is the least toxic drug and ribavarin is the most toxic agent in both models. However, the murine FLC system was found to be a more sensitive model for the assessment of t o x i c i t y of anti-HIV agents towards erythroid progenitor cells as indicated by the ICso,values. Drug-induced hematopoietic toxicity has been a limiting factor in the administration of zidovudine (AZT) to AIDS patients ( I ) . Similar toxicities have been reported to occur with other anti-HIV agents in clinical t r i a l s suc~ as 2',3'-dideoxycytidine (DDC), 2',3'-dideoxy-2',3'-didehydrothymidine (d4T) and ribavarin (2). Anemia and neutropenia commonly develop in patients receiving different doses of these drugs (1,3). Bone marrow megaloblastic changes observed in some cases were attributed to AZT-induced depletion of normal pyrimidine pools (4). Hematopoietic growth factors such as human recombinant granulocytemacrophage colony stimulating factor (rGM-CSF) have been employed to overcome AZT-mediated biochemical perturbations. The in vitro studies have suggested that rGM-CSF p a r t i a l l y corrects AZT-induced cytotoxicity in normal human bone marrow myeloid progenitor cells (5). In i n i t i a l clinical t r i a l s rGM-CSFhas been shown to stimulate erythropoiesis (6). Similarly, recombinant erythropoietin, at present in clinical t r i a l s for anemia of end stage renal disease may have potential in inducing bone marrow regeneration in hypoplastic anemia whether primary or as a result of cytotoxic or other therapy (7). Therefore, i t was deemed desirable that an assay system that would be sensitive to the use of such growth factors and hormones would be a more appropriate model to assess the toxicities of various anti-AIDS drugs towards erythroid progenitor cells. Murine fetal l i v e r cells (FLC) are known to express erythropoietin receptors on the erythroid cells and were found to be highest at the CFU-E stage (8). We have therefore employed this technique to assess the toxicity of various anti-AIDS drugs towards erythroid progenitor cells in FLC which have been shown to express erythropoietin receptors. The results were simultaneously compared with toxicity data on murine bone marrow cells. 0024-3205/89 $3.00 + .00 Copyright (c) 1989 Pergamon Press plc
iv
Toxicity of Anti-HIV Drugs on BM and FLC
Vol. 45, No. 4, 1989
Materials and Methods
Mice - CD-I white mice (Charles River), 8-10 weeks old, and 13-14 days old f e t a l - l i v e r cells were employed in this study. All mice were cagedunder v i r a l - f r e e conditions. Drugs - Azidothymidine and d4T were synthesized in our laboratory and DDC and ribavarin were obtained from the National Cancer Institute (Bethesda MD). Erythropoietin (TCEpo) was supplied by Amgen Corp. (Thousand Oaks, CA.). Preparation of Murine Fetal Liver Cells - Cells were prepared aseptically from 13-14 day old fetuses by a modified technique (8). Cells were collected from each fetus and made into a single cell suspension by aspiration through pipettes followed by a 25 X 5/8 gauge needle. Cells were washed with alpha medium by centrifuging at 200 x g for 10 min. Cells were resuspended in alpha medium and counted in Turk's solution. The final cell concentration was adjusted to Ix1OQ/ml. Preparation of Bone Marrow Cells - Bone marrow cells were obtained from mouse femurs. Femurs were removed aseptically after anesthetizing the mouse with ether; marrow was flushed with a 25 X 5/8 gauge needle. A single cell suspension was made after repeated aspirations and counted in ~urk's solution. The final cell concentration was adjusted to I x 10°/ml. CFU-E Assay - Methylcellulose (1.3%) was prepared in RPMI-1640 containing L-glutamine (2.0 mM), fetal bovine serum (20%), 2-mercapto-ethanol (0.1 mM), p e n i c i l l i n and streptomycin (~00 U/ml and 100ug/ml) and erythropoietin (50 mU/ml). Cells (Ix10°/ml) were suspended in the methylcellulose mixture. One m i l l i l i t e r of methylcellulose containing cells was plated in duplicate and counted for erythroid colony formation ~CFU-E) after incubation in a humidified atmosphere of 5% C02-95% air at 37uC for 48 hr. The plates were stained with 3,3'-diaminobenzidTne and colonies containing 8 or more benzidine-positive cells were counted under an inverted microscope. Results and Discussion Hematological t o x i c i t y has been reported to be one of the major adverse effects of the various agents being evaluated for anti-HIV a c t i v i t y . AZT-induced anemia in patients treated for AIDS has been well documented. Similar toxic effects have been observed with other drugs such as DDC, d4T and ribavarin which are currently in clinical t r i a l s . Previous studies have demonstrated the toxic effects of these drugs both in v i t r o and in vivo on bone marrow hematopoietic progenitor cells. The mode of action for antiviral a c t i v i t y of AZT, DDC and d4T seem to be similar at the level of inhibition of reverse transcriptase a c t i v i t y since after being incorporated into DNA these agents act as chain terminators. Ribavarin on the other hand has been shown to block mRNAcapping. In an attempt to understand the mechanism of toxicitY of these drugs on bone marrow hematopoietic stem cells, we have compared the inhibitory effect on erythroid progenitor cells in bone marrow with the inhibition of erythropoietin-mediated stimulation of erythropoiesis in FLC which is d i r e c t l y related to heme synthesis. The data in Figures I-4 show that all four drugs i n h i b i t CFU-E formation in a dose dependent manner in both bone marrow and FLC systems. The IC50 values for AZT, DDC, d4T and ribavarin in bone marrow cells calculated from these data were 2.1, 8.0, 9.0 and 0.62UM respectively. In contrast, the corresponding ICso values in the FLC system were 1.0, 1.0, 2.6 and 0.23uM indicating that ~ e FLC system was more sensitive to the effect of these drugs
Vol.
45, No. 4, 1989
I.d
I
Toxicity
of A n t i - H l V
Drugs
on BM and FLC
1°°1
;:.) It,,. (.3
75-
"6
=,,.
0
50 ¢..
'E
25.
Q.,
00.1
0.5
1
5
10
50
Concentration of AZT (uM)
FIG. 1 Effect of various concentrations of AZT on murine bone marrow ( [] ) and fetal ~iver progenitor cells ( m ) by CFU-E assay. Results are mean + SE of 4 different experiments with duplicate determination for each point. 1 O0 W
I
"0
75
"6
.lr=
0
50 c-
•
25
III O.
0.1
0.5
1
5
10
50
Concentration of DDC (uM)
FIG. 2 Effect of various concentrations of DDC on murine bone marrow ( [] ) and fetal l i v e r progenitor cells ( [] ) by CFU-E assay. Results are mean ÷ SE of different experiments with duplicate determination for each point.
v
vi
Toxicity of Anti-HIV Drugs on BM and FLC
Vol. 45, No. 4, 1989
1 O0
','
1
I
0
"
75 l
E o
~-- 50 .E L,-
25.
0 0.1
0.5
1
5
10
100
Concentration of d4T (uM)
FIG, 3 Effect of various concentrations of d4T on murine bone marrow ( [] ) and fetal l i v e r progenitor cells ( • ) by CFU-E assay. Results are mean + SE of 4 different experiments with duplicate determinations for ~ch point. 1 O0 -
bJ
I
"0
75.
¢..
50c-
0
~
111
25-
i
0.05
0.1
0.5
1
10
Concentration of Ribavarin (uM)
FIG. 4 Effect of various concentrations of ribavarin on murine bone marrow ( [] ) and fetal l i v e r progenitor cells ( • ) by CFU-E assay. Results are mean + SE of 4 different experiments with duplicate determination for-each point.
Vol. 45, No. 4, 1989
Toxicity of Anti-HlV Drugs on BM and FLC
vii
on CFU-E formation than the bone marrow. The difference in the toxicities of these agents towards erythroid progenitor cells between the two models was significantly increased in the case of DDC which was found to be 8-fold more toxic to FLC. The increased t o x i c i t y shown by ribavarin in both the models may reflect the specificity of this drug in inhibiting the translation process perhaps at the level of mRNAand delaying the maturation process of late phase erythroblast. The clinical t r i a l s with ribavarin in patients with Lassa fever have shown that anemia occurred in about half of these patients at the dose levels of 20-60 mg/kg/day (9). Thus, the in vitro erythroid t o x i c i t y described in this report seem to correlate with c l i n i c a l l y observed erythroid myelosuppression. Although other reports (10-12) have been published on the t o x i c i t y of AZT to hematopoietic progenitor cells (both CFU-GM and BFU-E) the effects of AZT and other anti-HIV drugs on the CFU-E compartment, either in human or murine systems have not been f u l l y ascertained. We have examined the comparative toxicities of the four anti-HIV drugs which are in clinical t r i a l s except AZT, the only drug approved by FDA so far for the treatment of AIDS. The data suggests that d4T is the least toxic drug and ribavarin is the most toxic agent in both FLC and bone marrow cells. However, the murine FLC system was found to be a more sensitive model as indicated by the IC50 values for the assessment of t o x i c i t y of anti-HIV agents towards erythroid progenitor cells. This system should provide an appropriate test system for distinguishing the absolute toxicities of anti-HIV drugs for the erythroid progenitor cell compartment. Furthermore, the FLC system provides a model to further study the mechanism of t o x i c i t y which may be at the level of heme synthesis. Acknowledgement This work was supported by PHS grant number A125909 awarded by the NIAID, DHHS. References
I.
2. 3. 4. 5.
R. YARCHOAN,K.J. WEINHOLD, H.K. LYERLY, E. GELMANN, R.M. BLUM, G.M. SHEARER, H. MITSUYA, J.M. COLLINS, C.E. MYERS, R.W. KLECKER, P.D. MERKHAM, D.T. DURACK, S.N. LEHRMAN, D.W. BARRY, M.A. FISCHL, R.C. GALLOW, D.P. BOLOGNESI, S. BRODER, Lancet. I 575-580 (1986). M. BABA, R. PAUWELS, P. HERDEWIJN, E. CLERCQ, J. DESMYER, and M. VANDEPUTTE. Biochem. Biophys. Res. Commun. 142 128-134, (1987). R. YARCHOAN, S. BRODER. N. Engl. J. Med. 316 557-564, (1987). D.D. Richman, M.A. FISCHL, M.H. GREICO, M.S. GOTTLIEB, P.A. VOLBERDING, O.L. LASKIN, J.M. LEEDOM, J.E. GROOPMAN, D. MILDVAN, M.S. HIRSCH, G.G. JACKSON, D.T. DURACK, S.N. LEHRMAN. N. Engl. J. Med. 317 192-197 (1987). K. BHALLA, M. BIRKHOFER, S. GRANTand G. GRAHAM. Exp. Hematol. 17 17-20 (1989).
6.
7. 8. 9. 10. 11. 12.
J.E. GROOPMAN, R.T. MITSUYASU, M.J. DELEO, D.H. OETTE, D.W. GOLDE. N. Engl. J. Med. 317 593-598 (1987). P.M. COTES. Erythropoietin: The developing story. Br. Med. J. 296 805-806 (1988). H. FUKAMACHI, T. SAITO, A. TOJO, T. KITAMURA, A. URABE, and F. TAKAKU. Exp. Hematol. 15 833-837 (1987). H. FERNANDEZ, G. BANKS, and R. SMITH. Eur. J. Epidemiol. 2 1-14 (1986). J.P. SOMMADOSSI and R. CARLISLE. Antimicrob. Agents Chemother. 31 452-454 (1987). R.M. RUPRECHT, L.G. O'BRIEN, L.D. ROSSONI, S. NUSINOFF-LEHRMAN. Nature 323 467-469 (1987). J.P. SOMMADOSSI, and R. CARLISLE. Antimicrob. Agents Chemother. 31 23-26 (1988).