Induction of Lymphomas in Mice by Busulfan and...
Transcript of Induction of Lymphomas in Mice by Busulfan and...
[CANCER RESEARCH 41, 3478-3482, September 1981]0008-5472/81 /0041-OOOOS02.00
Induction of Lymphomas in Mice by Busulfan and Chloramphenicol1
E. Robin, M. Berman, N. Bhoopalam, H. Cohen, and W. Fried2
Departments ol Medicine ¡E.ft.. M. B., W. F.] and Pathology [H. C.], Michael Reese Hospital and Medical Center. Chicago, Illinois 60616, and Department ofMedicine, HiñesVeterans Administration Hospital, Hiñes,Illinois 60141 [N. B.j
ABSTRACT
Busulfan causes long-lasting defects in the hematopoietic
stem cells and in the immune system of mice. We designedstudies to determine whether Chloramphenicol further damagedthe already defective hematopoietic stem cells of mice thatwere pretreated with busulfan, and we unexpectedly observedthat mice given injections of the combination of busulfan andChloramphenicol developed lymphomas in relatively high incidence. The disease is invariably associated with splenomegalyand enlargement of the thymus. Leukocytosis with lympho-
blasts in the peripheral blood occurred in some affected mice.The malignant cell is a lymphoblast of thymic origin. Thirteenof 37 mice which received both busulfan and Chloramphenicoldeveloped lymphomas. An additional five of the remaining 24mice without proven lymphoma died and were not autopsied.Twelve of the 13 proven lymphomas developed within 280days from the start of the experiment. Four of the 35 micetreated with busulfan alone developed lymphomas, and anadditional five of the remaining 31 died but were not autopsied.Two of 41 mice treated with only Chloramphenicol developedlymphomas. Of the mice treated with either busulfan or Chloramphenicol alone that developed lymphomas, all did so morethan 280 days from the start of the experiment. None of thecontrol mice developed lymphoma.
We conclude that both busulfan and Chloramphenicol mayinduce lymphomas in mice that are not known to develop themspontaneously. The combination of both busulfan and Chloramphenicol increased the frequency and accelerated the onsetof the disease.
INTRODUCTION
Busulfan, an alkylating agent commonly used in therapy ofpatients with myeloproliferative disorders, is notorious for itspropensity to cause long-term marrow hypoplasia. Morley and
Blake (5), Morley et al. (6), and Pugsley era/. (8) have reportedthat the administration of busulfan to mice results in a prolonged suppression of the colony-forming unit (spleen) andcolony-forming unit (cell) populations and impairment of T-
lymphocyte function; yet, the peripheral blood counts remainnearly normal.
Chloramphenicol, a broad-spectrum antibiotic that impairs
mitochondrial protein synthesis, has been implicated as acause of aplastic anemia and leukemia in some patients (2, 4).However, the blood counts of the majority of patients areunaffected by moderate doses of the drug. One explanation ofthis phenomenon is that an occult preexisting stem cell defectpredisposes some persons to the hematopoietic toxicity of
' This research was supported by USPHS Grant AM 19455 and Grant CH-
179 from the American Cancer Society.2 To whom requests for reprints should be addressed.
Received January 5. 1981 ; accepted June 12. 1981.
Chloramphenicol. Chloramphenicol does not affect the bloodcounts of normal mice.
We designed an experiment to determine whether the hematopoietic functions of mice, the stem cell compartments ofwhich were damaged by exposure to busulfan, would be affected by Chloramphenicol. In the course of these studies, weserendipitously observed that a significant number of micetreated with busulfan and Chloramphenicol developed lymphomas. We shall report here on the incidence of lymphoma andsome of the features of the disease.
MATERIALS AND METHODS
Male BALB/c x A f, (hereafter called CAP,) mice ages 6 to 8weeks were used. One hundred mg busulfan were dissolved in 10 mlacetone and then diluted with 40 ml distilled water. One g Chloramphenicol was dissolved in 100 ml 0.9% NaCI solution.
WBC, platelet counts, and hematocrits were performed on bloodfrom the retroorbital sinus. WBC counts were performed with the ModelB Coulter Counter, hematocrits were by the microcapillary method, andplatelet counts were by phase microscopy.
Peripheral blood smears were air dried, fixed in methanol, andstained with Wright-Giemsa stain. Touch preparations of the spleen
and smears of tibial bone marrow were air dried, fixed in methanol,and stained with Wright-Giemsa.
Organs (spleen, thymus, liver, kidney, lymph nodes) were fixed inBouin's solution, dehydrated through ethyl alcohol and toluene, embed
ded in paraffin, sectioned at 4 ^m thickness, and stained with hema-
toxylin and eosin. Bones were decalcified in EDTA after fixation withBouin's solution and then treated as other organs.
Peripheral blood for cell surface marker determination was collectedfrom retroorbital venous sinuses into heparinized capillary tubes. Aftersacrificing the mice by cervical dislocation, single cell suspensions ofspleen and thymus were prepared by passing the tissue (finely cut)through a stainless steel screen and flushing through a 26-gaugeneedle. The cells were subjected to Ficoll-Hypaque (1 ) gradient sepa
ration to remove dead cells and to isolate the mononuclear cell population. Rabbit antisera to mouse T-cell antigen, IgG, IgM, and IgA,
conjugated with fluorescein isothiocyanate, were obtained from CappelLaboratories (Cochranville, Pa.). The staining procedure used was thatof Predhomme and Labaume (7). The mononuclear cell suspensionswere adjusted to a concentration of 1 x 107 cells/ml in phosphate-
buffered saline [0.01 M sodium phosphate 0.15 M sodium chloride (pH7.2)] with 0.2% sodium azide and a solution containing 10 mg of bovineserum albumin per ml. One hundred /il of the suspension were transferred to a 10- x 75-mm test tube. One hundred ^l of the antibody
conjugate of appropriate dilution were added and mixed well (thedilutions were determined previously). This was incubated at roomtemperature for 30 min. The cell suspension was then washed 3 timeswith phosphate-buffered saline-bovine serum albumin. The resulting
pellet was gently suspended after the final wash, transferred onto aslide, and covered with a coverslip. The cells were examined with aZeiss microscope. At least 200 to 500 lymphocytes were counted, andthe number of cells with positive fluorescence were counted.
Mice were initially divided into 4 groups, each with 45 mice. Twogroups (B/C and B/O) were given injections of 0.5 mg (0.25 ml) ofbusulfan i.p. every 2 weeks for 4 doses, i.e., on Days 1,15, 29, and
3478 CANCER RESEARCH VOL. 41
on July 8, 2018. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
Induction of Lymphomas in Mice
43 of the experiment. The other 2 groups (O/C and O/O) receivedinjections only of the diluent, i.e., acetone plus distilled water. After a20-week rest period, on Day 183 of the experiment, there were 78
mice left in the B/C and B/O groups and 88 mice left in the O/C andO/O groups. The remaining mice died within 24 hr of injection, apparently of complications of the injections. The groups designated B/Cand O/C then received 2.5 mg (0.25 ml) of chloramphenicol 5 daysper week for 5 weeks, i.e., from Day 183 to Day 218 of the experiment.Groups B/O and O/O received injections of only 0.9% NaCI solutionin the same frequency. Prior to the development of lymphoma, 3 micefrom each group were removed for cell surface marker determinationsof peripheral blood, thymus, and spleen, leaving the following numberof mice in each group: B/C, 37; B/O, 35; O/C, 41 ; and O/O, 41.
Blood was to be obtained from the retroorbital sinus of 5 mice pergroup every week for determination of hematocrit, WBC, platelet, anddifferential counts (no single mouse was to be sampled more than 2times during the study). The original plan was to continue the study fora total of 20 weeks after the initial injection of chloramphenicol, i.e.,Day 323 of the experiment. However, because of the development oflymphomas, this plan was slightly modified. After the initial observationof lymphoma on Day 235 of the experiment in one mouse of the B/Cgroup, we continued the initial protocol. However, when 2 more micein the B/C group developed lymphomas on Day 240 of the experiment,we introduced the following modifications. All mice were inspecteddaily (Monday through Friday). Mice that appeared obviously ill or haddistinctly protuberant abdomens and/or palpably enlarged spleens hadblood counts performed, were sacrificed, and were autopsied. Thespleens were weighed, touch preparations were made, and the following tissues were submitted for histological sections: spleen; thymus;bone marrow; axillary lymph nodes; liver; and kidney. Histologicalsections of spleen were also obtained from randomly selected micethat did not apparently have lymphomas. The T- and B-lymphocyte
content of the peripheral blood, thymus, and spleen of 6 lymphomatousmice (3 B/C and 3 B/O) as well as 6 control mice (O/O) wasdetermined after the onset of lymphomas. As indicated above, the T-and B-lymphocyte content of peripheral blood on nonlymphomatous
mice from each of the treated groups was determined prior to the onsetof lymphomas.
In a total of 8 mice with lymphomas, one-half of the spleen wasremoved immediately postmortem; suspensions were made, and 10sor 106 nucleated cells in 0.5 ml Hanks' solution were injected into each
of 10 male CAP, mice, ages 6 to 8 weeks, to determine whether thelymphoma was transplantable.
On Day 350 of the experiment, all surviving mice were sacrificedafter obtaining blood for determination of the hematocrit and WBC. Thespleens were weighed, and any suspicious organs were fixed, sectioned, and stained for histological examination.
As indicated in "Results," although an attempt was made to identify
moribund mice and to determine whether they had developed lymphomas. a total of 10 mice (5 in B/C and 5 in B/O) died and were notautopsied.
RESULTS
Table 1 shows the results of hematocrit, WBC, and plateletcounts prior to treatment with chloramphenicol, immediatelyafter chloramphenicol, and 6 weeks after completing treatmentwith chloramphenicol (mice in B/O and O/O received diluentinstead of chloramphenicol). There was considerable variabilityof the WBC in the 4 groups, with no statistically significantdifferences among the experimental groups. There also wereno consistent differences in the hematocrits or platelet countsof mice in the various experimental groups.
Table 2 shows the blood counts of lymphomatous mice. Allhad significant decrease in hematocrits and platelet counts.
Table 1Blood counts of mice that did not develop lymphomas
Table 1 was made in reference to when chloramphenicol was given.
Prior to chloramphenicolB/C"B/OO/CO/OImmediately
postchloramphenicolB/CB/OO/C0/06
wkpostchloramphenicolB/CB/OO/CO/OHematocrit
(%)47
±0.9°46
±0.950±1.249±1.147
±0.648±0.749±0.649±1.248
±0.949±0.847±1.151±0.4WBC
X1035.2
±0.63.5±0.46.8±0.97.6±1.08.2
±1.03.9±0.75.5±0.84.0±0.68.2
±0.93.9±0.45.5±0.45.4±0.6Platelet
countx1061.21
±0.250.97±0.131.56±0.281.80±0.391.48
±0.141.35±0.212.46±0.292.07±0.321.5
±0.181.5±0.192.4
±0.211.6±0.21
B/C, busulfan/chloramphenicol; B/O, busulfan/diluent; O/C, diluent/chloramphenicol; O/O, diluent/diluent.
b Mean ±S.E. based on 5 min/group.
Table 2
Blood counts of 19 mice with lymphomas
HematocritWBCPlateletsMean
±S.E.33.0
± 0.3830,500 ±2,950
440,000 ±3,155Range17-472,600-202,000
30,000-830,000Median34.010,000
440,000
Table 3
Number of mice in each treatment group that developed lymphoma
GroupB/C8
B/OO/CO/OProven
lymphoma13/374/35
2/410/41Died
without proof oflymphoma5
500
B/C, busulfuan/chloramphenicol; B/O, busulfan/diluent; O/C; diluent/chloramphenicol; O/O. diluent/diluent.
The WBC of 7 of 19 lymphomatous mice (5 in B/C and 2 in O/C) were greater than 30,000.
Table 3 shows the incidence of lymphoma in the 4 experimental groups. No lymphomas were detected in any of thecontrol mice. Five mice each in the B/C and B/O groups diedwithout verification of the cause of death.
Chart 1 demonstrates the time that lymphoma was detectedin the various groups. It is of note that 12 of the lymphomas inthe B/C group became apparent within 280 days from the startof the experiment whereas, in the B/O group, all 4 mice thatdeveloped lymphoma did so more than 320 days from the timethat the experiment began. The 2 mice in the O/C group thatdeveloped lymphoma did so on Day 290 of the experiment.
All mice with proven lymphomas had enlarged spleens (0.7to 1.0 g) when compared to those of nonlymphomatous mice(0.09 to 0.12 g). Both the white pulp and red pulp of the splenictissue were diffusely invaded by neoplastic cells, resulting ineffacement of normal landmarks. The pattern was usually diffuse, although focal nodularity was occasionally seen. Thecapsule and pericapsular fat were focally infiltrated by neoplastic cells. Moderate numbers of normal hematopoietic cellsremained, chiefly megakaryocytes and erythroid precursors.Granulocytic precursors were essentially absent. Fig. 1 showsphotomicrographs of normal and lymphomatous spleens.
SEPTEMBER 1981 3479
on July 8, 2018. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
E. Robin et al.
TIME Of ONSET OF LYMPHOMA
Table 4
Surface markers on cells of lymphomatous and control mice
Source of lymphocytes
T/Ba lymphomaT/B control
DAYS FROH START OF EXPERIMENT
Chart 1. Note the increased frequency and rapidity of development of lym-phoma in the group receiving both busulfan and chloramphenicol compared tothe groups that received only one or the other. D, B/C; O, B/O; •,O/C.
Fig. 2 shows the morphology in lymphoma cells. The neo-plastic cells are larger than the lymphocytes in normal splenicwhite pulp. In tissue section, the nuclei of lymphoma cells areleptochromatic, while in smear preparations, the chromatin isdelicate and finely distributed. The nucleoli generally are notevident; however, when seen, they are usually single and smallto medium in size. The ratio of nucleus to cytoplasm is high,and the cytoplasm is basophilic and agranular.
All mice with lymphomatous spleens had enlarged thymusand lymph nodes, which histologically were diffusely invadedby neoplastic cells resulting in effacement of normal architecture.
The liver and kidney were found to be involved in somelymphomatous mice. The liver contained a diffuse infiltration ofneoplastic cells, that was most marked in the portal areas, andless prominent in the sinuses of the hepatic lobules. In thekidney, the interstitial tissue was diffusely infiltrated by neoplastic cells.
Table 4 shows the results of cell surface marker determinations on lymphomatous and control mice. In the mice withlymphomas, 87 to 100% of the lymphocytes in the blood,spleen, and thymus had cell surface markers, indicative of T-cell lineage. T- and B-cell markers done on peripheral blood of
treated, nonlymphomatous mice revealed markers similar tothose of control mice.
Transplantation of spleen cell suspensions from 8 of 8 micewith lymphoma resulted in clinically evident lymphomas in100% of the recipients within 3 weeks when injected with 106cells and within 4 weeks when injected with 10s cells. All micethat received 106 cells were dead within 5 weeks, and all micethat received 105 cells were dead within 7 weeks.
DISCUSSION
The studies reported in this paper were undertaken initiallyto determine whether in mice chloramphenicol further influences hematopoiesis already altered by busulfan. Serendipi-
tously, we observed that mice treated with both chloramphenicol and busulfan developed lymphomas earlier and more frequently than did those treated with either agent alone. Nolymphomas were detected in control CAFi mice, and this hybridstrain is not known to develop any tumor in particularly highincidence.
PeripheralbloodSpleenThymus87/1195/3100/095/1100/1100/0100/0100/066/1680/1276/1480/1467/3153/4660/5766/5790/2100/2100/0100/0
T/B, percentage of T-cells/percentage of B-cells.
Morley and Blake (5) noted that mice treated with busulfanalone developed lymphomas but did not describe these neoplasms in detail. It should also be noted that their reportconcerned studies done with Swiss Webster mice, a totallyunrelated strain to the CAFt mice used in our studies. Thelymphomas which developed in CAFi mice treated with busulfan and chloramphenicol were of T-cell type and almost invariably presented with splenomegaly. Lymphocytosis with a leu-kemic peripheral blood picture was present in some but not allmice. The thymus, lymph nodes, liver, kidney, and bone marrowwere infiltrated to varying extents in many autopsied animalswith lymphoma. The mechanism by which busulfan and chloramphenicol exert their lymphomagenic effects on CAP, mice isnot known. Possibly, by interfering with immune function, theypermit the propagation of virally transformed cells (8). It remains, however, to be demonstrated that the lymphomas of thebusulfan- and chloramphenicol-treated mice are transmissibleby a cell-free extract. We are currently evaluating this possi
bility.Another possible mode of action is that these agents directly
cause or facilitate the transformation of lymphocytes to amalignant growth pattern.
Yunis ef al. (10) demonstrated that conversion of chloramphenicol to a nitroso derivative creates cell cycle changessimilar to those of nitrosoureas. They then speculate that thiscompound may be responsible for the chloramphenicol-in-duced aplastic anemia and acute leukemia seen in some patients (2,4). That busulfan predisposes a host to the conversionof chloramphenicol to its nitroso derivative is intriguing in lightof our findings but not demonstrated as of now.
The protocol described here for induction of lymphoma inmice may prove to be an excellent model for studying spontaneously occurring lymphomas, since the incidence of lymphoma in our treated mice is relatively high. Also, since patientstreated with chemotherapeutic agents have been shown todevelop lymphomas several years after induction of remission(3, 9), it is possible that further studies of the mechanism bywhich busulfan and chloramphenicol induced lymphomas willhelp to improve our understanding of this phenomenon. Finally,chloramphenicol has been shown to produce not only aplasticanemia but, in some patients, also an acute nonlymphocyticleukemia (2, 4). Perhaps, further study of this phenomenon inmice will shed some light on the occurrence of this event inhumans.
3480 CANCER RESEARCH VOL. 41
on July 8, 2018. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
REFERENCES
1. Bloom, B. R., Cohen, Z. A., and David, J. R. Evaluation of in vitro methodsfor characterization of lymphocytes and macrophages. In: B. R. Bloom andJ. R. David (eds.), In vitro Methods in cell mediated and tumor immunity, p.3. New York: Academic Press, Inc., 1976.
2. Brayer, M.. and Damesheck, W. Hypoplastic anemia and myeloblastic leukemia following chloramphenicol therapy. N. Engl. J. Med., 277. 1003-1005, 1967.
3. Chabner, B. A. Second neoplasm—a complication of cancer chemotherapy.N. Engl. J. Med., 297. 213-215, 1977.
4. Cohen, T., and Creger, W. Acute myeloid leukemia following seven years ofaplastic anemia induced by chloramphenicol. Am. J. Med., 43: 762-770,1967.
5. Morley. A., and Blake, J. An animal model of chronic aplastic marrow failure.
10.
Induction of Lymphomas in Mice
I. Late marrow failure after busulfan. Blood, 44: 49-57, 1974.Morley, A., Trainor, K., and Blake, J. A primary stem cell lesion in experimental chronic hypoplastic marrow failure. Blood, 45: 681-688, 1975.Predhomme, J. L, and Labaume, S. Detection of surface immunoglobulinson human cells by direct immunofluorescence. In: R. B. Blood and J. R.David (eds.). In Vitro Methods in cell mediated and tumor immunity, p. 155.New York: Academic Press, Inc., 1976.Pugsley, C. A., Forbes, I. J., and Morley, A. A. Immunologie abnormalities inan animal model of chronic hypoplastic marrow failure induced by busulfan.Blood, 5Õ. 601-610, 1978.
Reimer, R. R., Hoover, R., and Fraumeni, J. F. Acute leukemia after alkylatingagent therapy of ovarian cancer. N. Engl. J. Med., 297: 117-181, 1977.Yunis, A. A., Miller, A. M., Corbe«, M. D., and Arimura, G. K. Nitrosochloramphenicol: possible mediator in chloramphenicol-induced aplasticanemia. J. Lab. Clin. Med., 796. 36-46, 1980.
SEPTEMBER 1981 3481
on July 8, 2018. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
E. Robin et al.
CAP••j
^s*5ÄÄS
Fig. 1. Photomicrograph of normal Wl and lymphomatous (S) spleens. A. normal spleen showing red pulp (HP) with its megakaryocytes (arrows) and othermyeloid cells. Lymphoid nodule of white pulp (WP). B, red and white pulp diffusely occupied by lymphomatous cells; some megakaryocytes (arrows) remain. CAP,capsule, x 100.
Fig. 2. Photomicrograph of lymphocytes of white pulp of normal spleen (A) and lymphoma cells of involved spleen (B), x 1000.
3482 CANCER RESEARCH VOL. 41
on July 8, 2018. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
1981;41:3478-3482. Cancer Res E. Robin, M. Berman, N. Bhoopalam, et al. ChloramphenicolInduction of Lymphomas in Mice by Busulfan and
Updated version
http://cancerres.aacrjournals.org/content/41/9_Part_1/3478
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://cancerres.aacrjournals.org/content/41/9_Part_1/3478To request permission to re-use all or part of this article, use this link
on July 8, 2018. © 1981 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from