Inheritance of Susceptibility to Bleomycin-induced Pulmonary ......mycin (clinical-grade Blenoxane; gift of Sal Lucania and of Dr. Terry Dugan, Bristol-Myers Squibb, Evansville, IN)

[CANCER RESEARCH56. 2596-2601. June I. 19961

(15), capacity to repair DNA damage (16), and extent of inflammatoryresponse (1 1). Based on the varying extent of pulmonary fibrosis infour strains of mice, Schrier et a!. (14) proposed that bieomycininduced fibrosis is under polygenic control, but this genetic basis hasnot been pursued.

Recently, a method that uses animal models to isolate the geneticbasis of a heritable trait and link it with a genetic marker has beenoutlined by Lander and Botstein (17). The development of a mousegenetic model enables measurement of the biological variability associated with a trait in addition to presenting the heritability of adisorder (18). Mathematical modeling of the data from inheritance

studies can be used to identify the presence of an X-linked factor orgenomic imprinting and to estimate the number of QTL3 controllingthe trait. Mouse genetic models have been established to investigatemorphine preference (19), radiation-induced apoptosis of thymocytes(20), and autoimmune type I diabetes (21).

In this study, a mouse model is developed, using inbred strains, todetermine the extent of a genetic component in susceptibility tobleomycin-induced pulmonary fibrosis and to characterize the inheritance of this susceptibility. The base of the model consists of twoinbred strains of mice that differ in their susceptibility to a fibrosinglesion, C57BLJ6J, which is prone (8, 10, 14), and C3Hf/Kam, which isshown here to be resistant. The model developed follows the methodfor identifying heritable traits outlined by Lander and Botstein (17)and Levitt et aL (22). initially, a standard protocol for characterizingthe flbrosing phenotype, including bleomycin dose, assay of thefibrosing phenotype, and assay time, was developed in parental mice.The inheritance of the fibrosing phenotype was then characterized inF1 and F2 (first and second filial, respectively) generationmice fromthese progenitor strains after treatment with the bleomycin standardprotocol.

MATERIALS AND METHODS

Mice. C57BLI6Jmice were purchased from The Jackson Laboratory andhoused in microisolator cages in the specific pathogen free room of the animalcolony of the Department of Veterinary Medicine. The F1 (C57BL/6J female X C3Hf/Kam male = B6C3F1; C3HI/Kam female X C,7BU6Jmale = C3B,@,F1)and F2(F1intercross) generations were bred by us and housedsimilarly in this colony. The C3Hf/Kam mice were bred and housed in thespecific pathogen free animal colony of the Department of ExperimentalRadiotherapy. At the time of treatment all mice were 8â€”10weeks old.

Bleomycin Treatment. Lung damage was elicited by administering bleomycin (clinical-grade Blenoxane; gift of Sal Lucania and of Dr. Terry Dugan,Bristol-Myers Squibb, Evansville, IN) through osmotic minipumps implanteds.c. Bleomycin was dissolved in 0.9% NaC1 solution and loaded into 7-dayminipumps (model 2001; ALZA Corp., Palo Alto, CA). Each mouse wasanesthetized with Nembutal (5.6 mg/mI sodium solution) at a dose of0.0l ml/gof mouse body weight given i.p. The pump was implanted through a smallincision in the mouse's back, and this wound was sealed with wound clips.

Breathing Rate Assay. The mice were checked daily for signs of distress.To identify mice in respiratory distress, breathing rates were collected at 4weeks after the implantation of the pumps, and then weekly to 8 weeks, which

3 The abbreviations used are: QTL, quantitative trait loci; ROl, region of interest; HP,hydroxyproline.

ABSTRACT

Based on the range of patient responses to treatment, and on animalstudies, It Is hypothesized that individual variation in sensitivity to bleomycin-Inducedpulmonaryfibrosis is controlledgenetically.A geneticmodel has been developed by (a) establishing a distinct difference inbleomycin-Induced lung damage in two inbred strains of mice [parentalgeneration: C@BU6J (fibrosis-prone phenotype) and C3Hf/Kam (fibrosisresistant phenotype)l and (b) characterizing inheritance of the fibrosingphenotype in the F1 (first filial) and F2 (F1 intercross second filial)generations derived from the parental strains. Male mice received 100mg/kgandfemalemice125mg/kgofbleomyclnvias.c.osmoticminipump.Theanimalsweresacrificed8 weeksaftertreatmentor whentheirbreathlag rate indicated respiratory distress. The percentage oflung with fibro$15for each mouse was quantified with image analysis of a hIstOlOgicalsectionof the left lung.Themeanpercentageof fibrosisfor the C5@,BU6.Jmales was 8.4 Â±0.8% (SE) and 4.4 Â±0.8% for females, and the C3Hf/Kam mice of either sex did not present the fibrosing lesion (mean score,0%). SignIficant difference (P 6 x 106) was measured In percentage offibrosis between the two strains of F1 males, but not F1 females (P 0.38),suggesting the presence ofan X-linked factor associated with the fibrosingphenotype. From an ANOVA the X-linked factor Is estimated to contrib

ute 19% of the fibrosis phenotype. A genetic model of two or three mcicontrollingthe fibrosingphenotypeis proposedfrom the data of theparental, F1, and F2 generations. The mouse model demonstrates thatsusceptibility to bleomycin-induced pulmonary fibrosis is a heritable traitcontrolled by a few genetic lad.

INTRODUCTION

Almost since its discovery 30 years ago (I), bleomycin has beenknown to induce pulmonary fibrosis (2). In a review of patientresponse to bleomycin, Weiss (3) reports that certain individuals areespecially sensitive to this side effect of the drug. Hsu et a!. (4), basedon results from a cytogenetic assay, found interindividual variation insusceptibility of cultured lymphocytes to bleomycin-induced DNAdamage. Differences in susceptibility to bleomycin-induced pulmonary fibrosis by mouse strain have also been demonstrated (5â€”7).Onthe basis of these observations in humans (2, 3) and animals (5â€”7),itis hypothesized that susceptibility to bleomycin-induced pulmonary

fibrosis is, in part, genetically regulated.Considerable work has been completed using animal models to

study the characteristics of bleomycin-induced pulmonary fibrosis.The lesion caused by bleomycin involves focal collagen deposition(8â€”10),that may occur after an inflammatory response (9â€”i1) and achange in regulation of lung cytokines (7, 12, 13). The extent offibrosis that develops depends on dose (9, 11, 14), route of administration (8), and mouse strain (5â€”7).The differences in strain responseto bleomycin have been attributed to variation in drug inactivation

Received I2/28/95; accepted 4/3/96.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 with18 U.S.C. Section 1734 solely to indicate this fact.

I Supported by National Cancer Institute Grant CA06294, the University Cancer Fund,

and a Conversation with a Living Legend-Palmer Allocation.2 To whom requests for reprints should be addressed, at Department of Experimental

Radiotherapy, The University of Texas M. D. Anderson Cancer Center, 1515 HolcombeBoulevard, Houston, TX 77030.

2596

Inheritance of Susceptibility to Bleomycin-induced Pulmonary Fibrosisin the Mouse1

Christina K. Haston, Christopher I. Amos, Tern M. King, and Elizabeth L. Travis2

Departments of E.rperimen:al Radiotherapy [C. K. H.. E. L T.J and Epidemiology [C. 1.A.. T. M. K.], The University of Texas M. D. Anderson Cancer Center, Houston, Texas77030

on April 14, 2021. © 1996 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

GENETIC MODEL OF BLEOMYCIN-INDUCEDLUNG FIBROSIS

was the end of the experiment. Respiratory distress was defined as abreathing rate more than 130% of the control value. The plethysmographbreathing rate assay developed in this laboratory (23) and recently amended(24) was used.

Histology and Fibrosis Scoring. At autopsy, the lungs were removed; theright four lobes were archived, and the left lobe was submitted for histology.The single left lobe of each mouse was perfused with 10% neutral bufferedformalin and fixed for at least 24 h in formalin. The tissue sections werestained with H&E and with Masson's trichrome to identify the site(s) ofcollagen deposition in the lung, and examined by light microscopy. Each lungsection was scored for both the fibrosis phenotype and the extent of inflammation.

A quantitative end point of the fibrotic phenotype was necessary to permitgenetic modeling based on the variance of the amount of fibrosis in parental,F1, and F2 mice. The area of the fibrosing phenotype for each mouse wasscored by image analysis of a single left lobe section. The fibrosing phenotypewas defined as the fibrotic scar appearing in lungs of C57BU6J mice afterbleomycin treatment and is shown in Fig. I. This lesion consisted of subpleuralfoci of collapsed apposed alveolar walls with collagen superimposed.

The fibrotic lesion was quantified by collecting a 24-bit color image (framegrabber; Vision plus AT, Bedford, MA) of each section with a color videocamera (model DXC-960MD, Sony Medical Systems, Montvale, NJ) attachedto a stereomicroscope and interfaced to a computer. Two ROIs were defined onthe image using OPTIMAS software (Bioscan, Edmunds, WA), as illustratedin Fig. 2. The total area (Lang RO!) of the left lobe was determinedby the cornpeter using a thresholdpixel value for tissue.A user-selectedROI was then drawnon the imageto enclosethe fibrosis(FibrosisRO,r).â€˜fl.,@percentageoffibrosis wascalculatedas:

. . Fibrosis RO!% Fibrosis = x 100

Lung ROI

Fig. 1. Histological aectiona of the loft lung from misk mice after 100 mg/kg bleomycii,,it, sectionfromC,7BL/6Jmouseeshibitinglbs fibrosingphenotype;lungdamageisaubplsuralfoci of collapsedalveolarwallawith auperimpoaedcollagen;inflammatoryresponse of cells in airspace aurrounda the fibrotic lesion. B section from C3Hf/Kammouse with no evidence of pulmonary lesion. Masson's trichrome atain; magnification,X63.

Fig. 2. Image of a histological section of the left lung of a male C,7BL/6J mousetreated with 100 mg/kg bleomycin. The fibrotic lesion is located in the subpleura and ismultifocal. The superimposed ROIs define percentage of fibrosis as 12%. Masson'strichrome stain; magnification, X 10.

To determine whether scoring one section per mouse would ensure accurateclassification of the fibrosing phenotype, 10 sections of lung tissue from eachof 6 fibrosing mice and 20 mice with zero fibrosis were analyzed.

The extent of inflammation was also assessed for each lung section topresent the total lung response to bleomycin treatment. A graded system wasused to score inflammation, because the diffuse inflammatory lesion lacks theclear boundary that was used to define the area of fibrosis. Inflammation was

(A) scored by assigning a score of 0 to 3 to each left lobe section. A score of 0indicates no inflammation; a score of 1 indicates a single limited area ofinflammation characterized by the presence of a few foamy cells and cells inthe air space; a score of 2 represents a moderate inflammation that involvesmore than one area in the lung; and a score of 3 indicates a severe inflammatory response involving more than half of the lung.

ExperimentalDesign: Parental Studies. Dose-responsestudies werecompleted in the C3Hf/Kam (40 males, 53 females) and C57BLI6J (43 males,42 females) strains to identify the dose of bleomycin that maximizes thedifference in the development of the fibrosis between these two strains whileminimizing acute toxicity. The acute toxicity phase was defined as deathoccurring 14â€”20days after pump implantation. The mice received a dose ofbleomycin (100 mg/kg of body weight). which has been shown to inducesignificant fibrosis in the lungs of C,7BL'6 mice, over 6 weeks of study (8),and also 125 and 150 mg/kg. The animals were sacrificed by cervical dislocation at 8 weeks or when in respiratory distress as indicated by an elevatedbreathing rate. To determine whether the C3HfIKamstrain develops fibrosis ata later time, 10 C3Hf/Kam males were given 100 mg/kg bleomycin and 10C3Hf/KaIn females were given 125 mg/kg bleomycin and sacrificed at 12weeks. Five male mice of each of the parental strains were given 125 mg/kgbleomycin and sacrificed at 18 days after pump implantation to examinewhether lung damage was associated with the acute toxicity phase. Ten of eachof the parental C3HI/Kam and C,7BL/6J mice were sham treated with salinefilled pumps.

Inheritance Studies. The standard treatment dose of 100 mg/kg for malesand 125 mg/kg for females, as determined in the parental study, was deliveredto the B6C3F1(28 males, 18 females) and C3B6F1(30 males, 20 females)strains and to F2 mice (140 males, 99 females) representing all four possible F1intercrosses. The mice were sacrificed at 8 weeks or when in respiratorydistress as indicated by an elevated breathing rate.

Data Analysis. Tests for differences in fibrotic area between groups weredone with aingle-factor ANOVA, The male and female data sets were conaldered separately, because male mice were treated with a dose of 100 mg/kg andfemale mice with a dose of 125 mg/ks,

Analysea of the area of fibrosis data set (the phenotype) were completed toevaluate the possibility of an X-linked factor or genomic imprinting or mitochondrial inheritance associated with lung fibroala.To investigate the influenceof X linkage or genomic imprinting in the F1 generation. the level of fibrosisin the two strains of F1 males was compared as was the level of fibrosis in F1females, To test for consistency with X linkage in the F2 generation. the F@

A

2597



Table 1Left lunginflammationscoresMale

inflammation scoresFemale infiammationscoresTotal0123Total0123C3Hf/Kam

C57BU6JB6C3F1C3B6F,F220

21283011615â€•

022335

0213280

91815430

12601229

1818209524

015205

0312230

11143

440

700

8a

Number of mice with zero inflammation.

100 125 150

DoseofBle@n (mgfltg)

GENETICMODELOF BLEOMYCIN-INDUCEDLUNGFIBROSIS

parental mice. Fourteen to 20 days after pump implantation, allmice exhibited an acute toxicity characterized by wet, slick fur, butwith no increases in breathing rate. Histological examination of thelungs from five male C57BL/6J mice treated with 125 mg/kgbleomycin and sacrificed during this acute phase indicates thelungs were free of fibrosis. Three of these five mice sacrificedduring the acute toxicity phase, however, exhibited mild inflammation (scores of 1), and the other two scored 0. The lungs fromfive male C3Hf/Kam mice sacrificed at the same time contained nohistological injury. The mice sham treated with the saline pumpsexhibited no acute toxicity.

To verify that fibrosis in one histological section per lung wasrepresentative of the level of fibrosis in the entire lobe, serial sectionsfrom fibrosing and nonfibrosing mice were analyzed. No fibrosis wasevident in any of the 10 sections from each of the 20 nonfibrosingmice. Fibrosis was present in all sections from the lungs of sixfibrosing mice. The amount of fibrosis varied less between sectionsfrom the same lung (the maximum coefficient of variation of these sixmice was 15.2%) than the intrastrain coefficient of variation of 44%

for C57BL/6J females and 81% for C57BLI6J males.The fibrotic lesion in the lungs of male C57BL/6J mice after a 7-day

s.c. exposure to 100 mg/kg bleomycin is shown in Fig. 1A. This lesionconsists of foci of collapsed alveolar walls located in the subpleuraand contains a cellular infiltrate and superimposed collagen. The lungsof C3HfIKam mice (Fig. 1B) at the same dose do not show fibrosis.The fibrotic lesion was always localized to the subpleura and wasoften multifocal, as shown in Fig. 2.

Because of the high incidence of toxicity after a dose of 150 mg/kg,this dose was discontinued, and the area of the fibrosis was scored inthe 100- and 125-mg/kg dose groups only. At the dose of 100 mg/kg,all C57BLJ6J male mice expressed the fibrosing phenotype, whereas itwas absent from all C3HfIKam male mice, including those sacrificedat 12 weeks. The higher dose did not increase the mean fibrotic scorein C57BL/6J mice (8.6 versus 8.4% at 100 mg/kg; P 0.90) and 4 of10 C3Hf/Kam male mice developed minimal fibrosis at a dose of 125mg/kg, with a mean area of 0. 1%. Thus, the dose of bleomycin todiscriminate C57BLI6J male mice as susceptible to the fibrosingphenotype and C3Hf/Kam male mice as resistant, with minimal acutetoxicity, was determined to be 100 mg/kg.

The mean percentage of fibrosis in the lungs of C57BLI6J femalesat a dose of 100 mg/kg was not significantly different from that at 125mg/kg (2.7 versus 4.4%; P 0.10), but at the lower dose, 10 of 32mice did not develop fibrosis. At a dose of 125 mg/kg, all C57BLI6Jfemale mice developed the fibrosing phenotype. In C3HI/Kam females, only 1 of 29 mice developed any fibrosis at 8 weeks atthe higher dose of 125 mg/kg, and 1 of 10 developed fibrosis(area = 0.4%) by 12 weeks. Thus, the dose of bleomycin to distinguish C57BL/6J female mice as susceptible to the fibrosing phenotypeand C3Hf/Kam female mice as resistant, with minimal acute toxicity,was determined to be 125 mg/kg. No control or saline-treated miceshowed any evidence of fibrosis.

The scores of inflammation in the lungs are presented by mousestrain in Table 1. The mice sacrificed before the end of the experiment

20

.@

U-

Fig. 3. Lethal acute toxicity in male (top) and female (bottom) mice of parental strains.For C3HfJKamfemale mice at the doses of 100 and 150 mg/kg and for C3Hf/Kam malemice at the doses of 100 and 125 mg/kg, zero lethal acute toxicity was measured.

females were grouped by F1 sire and the level of fibrosis compared byANOVA. This comparison is of mice with 75% X@ and 25% XC3H versusmice with 75% &3H and 25% [email protected] investigate mitochondrial inheritance,the phenotypic data in offspring of a B6C3F1dam and offspring of a C3B@,F1dam were compared. ANOVA was also used to identify any difference in thephenotype among the four F2 types defmed by the F1 cross used in derivation.

To estimate the number of quantitative trait loci associated with susceptibility to fibrosis, Wright's formula, modified by Lander et aL (17) for intercross F2s, was used as given below.

- (@tC57- @xC3H)2

kâ€” 8aG2

In this formula, k is the number of genetic factors or QTL; pC3H and gAC57are the means of fibrosis percentages in the parental strains; oQ2 is the geneticvariance defined as @2 @y@12where o@F22and @,.p2@ the pooledvariances of the F2 and F1 generations, respectively. The number of geneticfactors was estimated for both the male and female data sets.

The breathing rate data of the two strains of F1 males and of the two strainsof F1 females were also compared using single-factor ANOVA.

RESULTS

Fig. 3 shows the fraction of lethality measured up to 20 daysafter pump implantation in males and females of both strains of the

2598



Table 2 Acute toxicity,respiratory distress, andmean percentage offibrosis inparental and inheritancestudiesMales

(100 mg/kg)Females (125mg/kg)Acute

toxicityRespiratory distressMean % fibrosisAcute toxicityRespiratory distressMean %fibrosisnn(%)n (%)(variance)nn (%)n(%)(variance)C3HI/Kam20000

(0)334(12%)00.0(0)C57BIJ6J283(1 1%)16 (76%)8.4 (13.5)224 (18%)7 (39%)4.4(12.8)B6C3F1

(C57X C3H)28013 (46%)4.2 (15.4)18000.4(0.3)C3B@,F1(C3H X C57)303 (10%)00.5 (0.6)201(5%)00.2(0.4)F214024

(17%)23 (20%)1.7 (13.9)994 (4%)17 (18%)1.6 (8.5)

A. Malesâ€¢$...-I-S

III IIII

B. FemalesIIISS-.-f#@I

25

20-

@ 15-@:@ 5.

0-

25

20-

@ 15-

!@ 5.

0-


8.4%; P = 0.0004). To determine whether the response of the F1generation was due in part to X linkage or genomic imprinting, thepercentage of fibrosis data were compared by parental strain dam. Theamount of fibrosis in the B6C3F1 (C57BL/6J dam) mice was greaterthan that in C3B6F1 (C3Hf/Kam dam) mice for the males(P = 6 X l0_6) but not the females (P = 0.38), which is consistentwith X linkage. There was no evidence of X linkage in the F2 femalemice (P = 0.63).

The contribution of an X-linked factor was estimated as the relativeinterstrain variance in the F1 compared to the parental (F0) generation.This approach considers the between-strain variation in the parentalgeneration, where the strains differ in all the factors that constitutesusceptibility to fibrosis, and the between-strain variation in the F1generation, because the F1 males only differ in their sex chromosomes. The relative variance between the strains of F1 males is anestimate of the contribution of the X factor to the total geneticvariance. Using the following equation and the phenotypic data set ofthe males, the contribution of an X-linked factor is estimated to be19% of total genetic variance.

ir@F1 (X@,c@ri XFI + (Xo@F1 XFI) x 100%;@ â€”(Xc57Bll@@ @)2+ (@HflKes@

The F2 data are not consistent with mitochondrial inheritance asdetermined by comparing the results in offspring of a B6C3F1 dam

10@

I19- 1 I II I I

I II

ii@I@@ I

C3HI/Kam C57BL/6J B6C3FI C3B6FI F2

Fig. 4. Final breathing rate of male mice after 100 mg/kg bleomycin.

all had inflammation scores of 2 or 3. Mice of the C57BLI6J strainalways presented moderate to severe inflammation (scores of 2 or 3),and most of the C3Hf/Kam mice did not develop inflammation. Theincidence of respiratory distress (Table 2) also indicates the straindifference in response to bleomycin. The B6C3F1 males were the onlyF1 type relatively sensitive to the bleomycin, as indicatedby respiratory distress and inflammation. As shown in Table 2, a similar fractionof F2 males and F2 females were sacrificed because of functionaldifficulty. Minimal acute toxicity was measured in the inheritancestudy as indicated in Table 2. Mice exhibiting lethal acute toxicitywere excluded from the analyses.

Fig. 4 shows the distribution of breathing rates at sacrifice for malemice of the parental, F1, and F2 generations after treatment with 100mg/kg of bleomycin, and includes mice sacrificed when sick and atthe end of the experiment. There is a clear separation of breathing ratedata for the two parental strains, as the C3FIf/Kam mice have breathing rates near 4â€”5breaths/s, and the C57BL/6J breathe at 6â€”9breaths/s. Mean breathing rates for control mice range from 4â€”5breaths/s for C3HI/Kam and 5â€”6breaths/s for C57BLI6J mice (datanot shown). The final breathing rate of the B6C3F1 males was significantly higher than that of the C3B@,F1males [6.5 Â±0.2 (SE) versus5.8 Â±0.1 breaths/s; P = 0.002], and this difference also occurredbetween the F1 females (B6C3F1, 5.8 Â±0.1; C3B6F1, 5.5 Â±0.1;P = 0.0005; data not shown).

The distribution of the fibrosing phenotype in F1 and F2 male andfemale mice is given in Fig. 5, and these data are summarized in Table2. The F2 data set includes 30 (32%) of 95 female mice and 46 (40%)of 116 male mice that did not develop fibrosis. Using Wright'sformula (Eq. B) and the data presented in Table 2, the number of QTLfor the male data set is 3.0, and for the female data set it is 0.3.

The mice of the F1 generation displayed minimal fibrosis in response to bleomycin, with the exception of the higher fibrotic response of the B6C3F1 males. The mean area of fibrosis in B6C3F1males is significantly lower than that of C57BL/6J males (4.2 versus

C3HI/KamC57BL/6J 86C3F1 C3B6FI F2

Fig. 5. Distribution of percentage of fibrosis in males (top) and females (bottom). Bar,mean score.

2599




with those of a C1B6FJ dam for either the males (P = 0.69) or thefemales (P = 0.1 1). To investigate the effect of an F1 cross (F1 damand sire) on the percentage of fibrosis data, the data of the F2generation were grouped by F1 cross as presented in Table 3. ANOVAshowed no difference among the four F2 groups for males (P = 0.94)or females (P = 0.34); therefore, the extent of fibrosis in the F2generation does not depend on the F, cross.

DISCUSSION

The foundation of this genetic model of bleomycin-induced lungfibrosis is the presence of the distinctive fibrosing phenotype in thelungs of C57BL/6J mice and the lack of fibrotic response in lungs fromC3Hf/Kam mice after the same dose of bleomycin. The inheritancestudies of this clear fibrosing phenotype in parental, F1, and F2generations suggest that susceptibility to bleomycin-induced lungfibrosis is a heritable trait controlled by two or three genetic factors,one of which may be X-linked.

The susceptibility to bleomycin-induced lung damage is unequivocally presented in the histology of each mouse. The imaging assay ofpercentage of fibrosis was developed so that these histological datacould be quantified, and the assay has been used by others (10, 25,26). Quantifying fibrosis in one section per lung was determined to berepresentative of fibrosis in the whole lung, in agreement with VanRongen et a!. (25).

The fibrotic phenotype is the same lesion Harrison et a!. (8)produced in the lungs of C57BL16 mice with bleomycin administeredthrough osmotic minipumps. In that study (8) and others (6, 9, 11, 12,14, 25), HP content was used to assay tissue collagen as a measure oflung fibrosis. The HP assay has been used to discern differences insusceptibility to bleomycin-induced fibrosis among mouse strains (6,8, 14) but not to categorize individual mice. In our preliminary studiesto develop the quantitative assay of fibrosis, the amount of HP in theright lungs of mice from both parental strains was quantified using astandard technique (27). Higher levels of HP per lung from C57BLI6Jmice than from C1HfIKam mice were recorded, but the amount oflung tissue HP did not correlate with the severity of the fibrosinglesion measured by quantitative histology (data not shown). In areview of the pathology of pulmonary fibrosis, Burkhardt (28) alsoobserved that in some studies in which tissue sections showed obviousfibrosis, the HP assay demonstrated normal amounts of collagen. TheHP measure,therefore, is not a suitable assay to representthe fibrosing phenotype in each F2 mouse, as would be necessary to developa genetic model.

The C57BU6J fibrosis-prone phenotype and the C3HI/Kamfibrosis-resistant phenotype after bleomycin treatment reported hereare consistent with what others have reported in these two strains inresponse to other cytotoxic insults, including radiation (29), ozone(30),andhyperoxia(31).In addition,thepresentstudyagreeswellwith the many investigations (6, 7, 10, 14) that have classifiedC57BU6J as highly fibrosingafterbleomycin treatment,althoughone

Table 3 Mean percentage offibrosis in the F2 generation by F1 cross

Males

n avgâ€•(var)b

26 1.8(16)25 2.0(21)38 1.8(12)27 1.4 (8.9)

116 1.7(14)

study (5) has described C3H/He mice as fibrosing in response tobleomycin.

Bleomycin administration produced differences in pulmonary function between the parental strains and between the F1 strains. Anelevated breathing rate was found to be an accurate determinant offunctional difficulty, in good agreement with Jaeger et a!. (32),because all animals sacrificed due to respiratory distress in the parental and inheritance studies had histological evidence of pulmonaryinflammation and/or fibrosis. The use of breathing rate measures, afunctional assay, in this study allowed the identification of mice withsevere respiratory insufficiency such that fewer than 10% of C57BU6Jmice died compared to 25% in a study by Harrison et a!. (8).Breathing rate measures did not predict fibrosis, however, becausemice presented fibrotic areas of up to 6% of the lung at the end of theexperiment, without any indications of functional difficulty. Changesin breathing rate after bleomycin were due to all lung damage,including both fibrosis, a focal lesion, and inflammation, a morediffuse lesion. The effect of a diffuse versus a focal lesion on breathing rate was demonstrated most clearly in F1 females. B6C3F1 femaleshad higher breathing rates at sacrifice, and more mice had higherinflammation scores than C3B6F1 females, but there was no differencein the level of fibrosis measured.

The area of fibrosis was measured in three generations of mice toidentify the inheritance of this phenotype. A number of programs areavailable to analyze phenotypic data. None of these programs, however, are appropriate to evaluate the data set of this study. The lungfibrosis measured is an induced response for which there is nobaseline level. The C3Hf/Kam mice did not develop fibrosis, and thisnull response and associated zero variance, along with X-linkage, hasnot been addressed in statistical packages. The BCROSS program ofSAGE (33) permits the evaluation of genetic models, but this programis not suitable for the present phenotypic data, because BCROSSanalysis requires normality in the data set, nonzero variance measures,and autosomal inheritance, all of which are inconsistent with thepresent data.

Despite statistical limitations, a model of inheritance can be proposed from the data set presented in Fig. 5 and Table 2. First, the malemice of the two F1 strains developed significantly different levels offibrosis in response to bleomycin, which is suggestive of either anX-linked factor or genomic imprinting associated with the phenotype.An X-linked factor is indicated when offspring display differentsensitivity to treatment and can be categorized by the X chromosomethey receive. Alternatively, genome imprinting arises when the phenotype produced by a gene differs when that gene is maternallyinherited from when that gene is paternally inherited (34). Thegenomic imprinting interpretation of the F1 data was rejected as thetwo strains of F1 females developed the same amount of bleomycininduced pulmonary fibrosis. The contribution of an X-linked factor isestimated to be 19% of the total genetic variance using the equationpresented in the results. Because the level of fibrosis is low in F1females, the X-linked factor is assumed to be recessive. No effects ofx linkagewereidentifiedintheF2females,whichmightindicatethatepistatic interactions are required for the expression of this factor.

Secondly, both F1 strains in this study developed fibrosis, indicating that one of the proposed loci must be autosomal and associatedwith a dominant factor (see Table 2). The level of fibrosis in malemice of the B6C3F1 strain is significantly lower than that of males ofthe C57BU6J strain, although both strains have the proposed X-linkedand autosomal dominant factors. This difference could be due to asecond autosomal factor (recessive) contributing to the fibrotic phenotype in the C57BU6J mice or due to incomplete dominance of thefirst autosomal factor in B6C3F1 males.

The confirmation and identity of the proposed factors can only be

Females

n avg (var)

28 2.2(18)7 2.7 (7.9)

36 1.0(3.5)24 1.6(5.495 1.6 (8.5)

Crossâ€•

B6C3/B6C3F2B6C3/C3B6F2C5B4@/C3B6F2C1B,@/B6C3F2Total F2

a femaleof crosslistedfirst.bavg.meanpercentageof fibrosis;var,variance.

2600




associated with mouse strain variation in susceptibility to pulmonary fibrosis. Reg.Immunol., 5: 207â€”217, 1993.

8. Harrison, J. H. Jr., and Lazo, J. S. High dose continuous infusion of bleomycin inmice: a new model for drug-induced pulmonary fibrosis. J. Pharmacol. Exp. Ther.,243: 1185â€”1194,1987.

9. O'Neill, C. A., Gin, S. N., Wang, Q., Perricone, M. A., and Hyde, D. M. Effects ofdibutyrylcyclic adenosine monophosphate on bleomycin-induced lung toxicity inhamsters. J. Appl. Toxicol., 12: 97â€”111, 1992.

10. Herman, E. H., Hasinoff, B. B., Zhang, J., Raley, L. G., Zhang, T-M., Fukuda, Y., andFerrans, V. J. Morphologic and morphometric evaluation of the effect of ICRF-l87 onbleomycin-induced pulmonary toxicity. Toxicology, 98: 163â€”175,1995.

11. Schrier, D. J., and Phan, S. H. Modulation of bleomycin-induced pulmonary fibrosisin the BALB/c mouse by cyclophosphamide-sensitive T cells. Am. J. Pathol., 116:270â€”278, 1984.

12. Hoyt, D. G., and Lazo, J. S. Alterations in pulmonary mRNA encoding procollagens.fibronectin and transforming growth factor-j3precede bleomycin-induced pulmonaryfibrosis in mice. J. Pharmacol. Exp. Ther., 246: 765â€”771,1988.

13. Phan,S. H., and Kunkel,S. L. Lungcytokineproductionin bleomycin-inducedpulmonary fibrosis. Exp. Lung Res., 18: 29â€”43,1992.

14. Schrier,N. J., Kunkel,R.G., andPhan,S. H.Theroleof strainvariationin murinebleomycin-induced pulmonary fibrosis. Am. Rev. Respir. Dis., 127: 63â€”66,1983.

15. Filderman, A. E., and Lazo, J. S. Murine strain differences in pulmonary bleomycinmetabolism. Biochem. Pharmacol., 42: 195â€”198,1991.

16. Harrison,J. H. Jr., Hoyt, D. G., and Lazo, J. S. Acute pulmonarytoxicityofbleomycin: DNA scission and matrix protein mRNA levels in bleomycin-sensitiveand -resistant strains of mice. Mol. Pharmacol., 36: 231â€”238,1989.

17. Lander, E. S., and Botstein, D. Mapping Mendelian factors underlying quantitativetraits using RFLP linkage maps. Genetics, 121: 185-199, 1989.

18. Crabbe, J. C., Belknap, J. K., and Buck, K. J. Genetic animal models of alcohol anddrug abuse. Science (Washington DC), 264: 1715-1723, 1994.

19. Ben@ettini,W. H., Ferraro, T. N., Alexander, R. C., Buchberg, A. M., and Vogel,W. H. Quantitative trait loci mapping of three loci controffing morphine preferenceusing inbred mouse strains. Nat. Genet.. 7: 54â€”58,1994.

20. Mod, N., Okumoto, M., Morimoto, J., Imai, S., Matsuyama, T., Takamori. Y., andYagasaki, 0. Genetic analysis of susceptibility to radiation-induced apoptosis ofthymocytes in mice. Int. J. Radial. Biol., 62: 153â€”159.1992.

21. Todd, J. A., Aitman, T. J., Cornall, R. J., Ghosh, S., Hall, J. R. S., Hearne, C. M.,Knight, A. M., Love, J. M., McAleer, M. A., Pains, J-B., Rodrigues, N., Lathrop,M., Pressey, A., DeLarato, N. H., Peterson, L. B., and Wicker, L. S. Geneticanalysis of autoimmune type I diabetes mellitus in mice. Nature (Lond.), 351:542â€”547,1991.

22. Levitt, R. C., Mitzner, W., and Kleeberger, S. R. A genetic approach to the study oflung physiology: understanding biological variability in airway responsiveness. Am.J. Physiol., 258: Ll57â€”Ll64, 1990.

23. Travis, E. L., Vojnovic, B., Davies, E. E., and Hirst, D. G. A plethysmographicmethod for measuring lung function in locally irradiated mouse lung. Br. J. Radiol..52: 67â€”74,1979.

24. Haston,C. K.,Newcomb,C. H.,Grant,K.,Hill,R. P., and VanDyk,J. Ultrasonicmeasurements of breathing rate in rats and computer assisted analysis. mt. J. Radiat.Oncol. Biol. Phys., 27: 651â€”657,1993.

25. van Rongen, E., Tan, C. H. 1.. and Durham, S. K. Late functional, biochemical andhistological changes in the rat lung after fractionated irradiation to the whole thorax.Radiother. Oncol., JO: 23 1â€”246,1987.

26. Habib,M.P.,Lackey,D. L.,Lantz.R.C., Sobonya,R. E.,Grad,R.,Earnest,D. L.,and Bloom, J. W. Vitamin A pretreatment and bleomycin induced rat lung injury. Res.Commun. Chem. Pathol. Pharmacol., 81: 199â€”209,1993.

27. Travis, E. L., Meistrich, M. L., Finch-Neimeyer, M. V., Watkins, T. L., and Kiss, I.Late functional and biochemical changes in mouse lung after irradiation: differentialeffects of WR-2721. Radiat. Res., 103: 219â€”231,1985.

28. Burkhardt, A. Alveolitis and collapse in the pathogenesis of pulmonary fibrosis. Am.Rev. Respir. Dis., 140: 513â€”524,1989.

29. Sharplin, J., and Franko, A. J. A quantitative histological study of strain-dependentdifferences in the effects of irradiation on mouse lung during the early phase. Radiat.Res., 119: 1â€”14,1989.

30. Kleeberger,S.R.,Bassett,D.J. P.,Jakab,G.J.,andLevitt,R.C.Ageneticmodelforevaluation of susceptibility to ozone-induced inflammation. Am. J. Physiol., 258:L313â€”L320,1990.

31. Hudak, B. B., Zhang. L-Y., and Kleeberger, S. R. Inter-strain variation in susceptibility to hyperoxic injury of murine airways. Pharmacogenetics, 3: 135â€”143,1993.

32. Jaeger, R. J., Sussman, R. G., Gearhart, J. M., Bowers, A. B., Shore, R. E., andSellakumar, A. R. Repeated pulmonary function evaluation following bleomycintreatment. J. Toxicol. Environ. Health, 11: 415-423, 1983.

33. Elston, R. C. BCROSS module. Statistical Analysis for Genetic Epidemiology,Release 2.2. New Orleans: Department of Biometry and Genetics, Louisiana StateUniversity Medical Center, 1994.

34. Sapienza,C.Genomeimprintingandcarcinogenesis.Biochim.Biophys.Acta,1072:51â€”61,1991.

35. Dietrich, W. F., Miller, J. C.. Steen, R. G., Merchant, M., Damron, D., Nahf, R.,Gross, A., Joyce, D. C., Wessel, M., Dredge, R. D., Marquis. A., Stein, L. D.,Goodman, N., Page, D. C., and Lander, E. S. A genetic map of the mouse with 4006simple sequence length polymorphisms. Nat. Genet., 7: 220â€”245,1994.

36. Taylor,A. M. R., Rosney,C. M.,andCampbell.J. B. Unusualsensitivityof ataxiatelangiectasia cells to bleomycin. Cancer Res., 39: 1046â€”1050.1979.

2601

achieved with genomic mapping. With Wright's formula (Eq. B),which doesn't allow for sex-linked effects, the number of factorscontrolling the fibrotic phenotype is one to three. The linkage map ofthe mouse (35) is of sufficient density to permit the mapping of theseloci with the present data.

Possible genes associated with the loci controlling bleomycininduced pulmonary fibrosis include those of the MHC (5, 7), thoseinvolved in DNA repair (16), and bleomycin hydrolase (15). Ekimotoet aL (5) have proposed the MHC role based on differences theyobserved in the level of pulmonary fibrosis in two congenic mousestrains differing only in h-2 haplotype. The response of the parentalstrains used in the present study is consistent with MHC influences asthe strains displayed different levels of bleomycin-induced lung fibrosis and differ in h-2 haplotype; C57BU6J mice are h@2b,andC3HI/Kam mice are h-2â€•.

Bleomycin is known to produce DNA strand breaks (4, 36); thus,sensitivity to bleomycin, measured as lung damage, could involve adeficient response in repairing or minimizing this DNA damage.Harrison et aL (16) reported a DNA repair deficiency in C57BL16mice, a fibrosing strain, as compared to BALB/c mice, a nonfibrosingstrain, after a dose of bleomycin that induced the same initial DNAdamage in both strains.

A thirdpossibility for the difference in susceptibilityto bleomycininduced pulmonary fibrosis is a difference in the lung level of bleomycin hydrolase, which inactivates bleomycin, between the parentalstrains. The lung tissue level of bleomycin hydrolase has been shownto be inversely correlated with the development of bleomycin-inducedpulmonary fibrosis in three mouse strains, including the fibrosingC57BL/6 and nonfibrosing BALB/c (15). Although the level of bleomycin hydrolase in C3Hf/Kam mice was not determined, this strainhas been shown to be nonfibrosing in response to bleomycin in amanner similar to the BALB/c mice, which have a high level ofbleomycin hydrolase in the lung (15).

Approximately 10% ofmice were lost early in these studies because ofacute toxicity. In a study administering bleomycin by different mutes,Harrison et aL (8) reported acute toxicity associated with i.v. and s.c.injections but none with the osmotic minipump. Death in the acutetoxicity phase, in this investigation, was attributed to severe dehydration

and is not related to the development offibmsis, as death occurred 1 weekbefore even the earliest fibrotic responders, and the lungs of mice sacriflood while in this phase showed no fibrotic injury histologically.

In summary, this three-generation mouse model demonstrates thatsusceptibifity to bleomycin-induced pulmonary fibrosis is a heritable traitthat can be quantified by image analysis of histological sections. Themodel suggested from the data is of two or three loci, one of which maybe X linked, associated with susceptibility to the flbmsing phenotype.

REFERENCES

1. Umezawa, H., Maeda, K., Takeuchi, T., and Okami, Y. New antibiotics: bleomycinA and B. J. Antibiot. (Tokyo) 5cr. A, 19: 200â€”209,1966.

2. Ichikawa, T. The clinical effect of bleomycin against squamous cell carcinoma andfurther developments. In: Proceedings of the Sixth International Congress of Chemotherapy, Tokyo, Japan, pp. 288-290. Vol. 2. Baltimore, MD, 1969.

3. Weiss, R. B. Hypersensitivity reactions. Semin. Oncol., 19: 458â€”477, 1992.4. Hsu, T. C., Johnston, D. A., Cherry, L. M., Ramkissoon, D., Schantz, S. P., Jessup,

J. M., Winn, R. J., Shirley, L., and Furlong, C. Sensitivity to genotoxic effects ofbleomycin in humans: possible relationship to environmental carcinogenesis. mt. J.Cancer, 43: 403â€”409,1989.

5. Ekimoto, H., Takada, K., Ohnuki, T., Takahashi, K., Matsuda, A., Takita, T., andUmezawa, H. Different sensitivity to bleomycin-induced pulmonary fibrosis amongvarious strains of mice. J. Clin. Biochem. Nutr., 2: 25â€”31,1987.

6. Harrison, J. H. Jr., and Lazo, J. S. Plasma and pulmonary pharmacokinetics ofbleomycin in murine strains that are sensitive and resistant to bleomycin-inducedpulmonary fibrosis. J. Pharmacol. Exp. Ther., 247: 1052â€”1058,1988.

7. Baecher-Allan, C. M., and Barth, R. K. PCR analysis of cytokine induction profiles



1996;56:2596-2601. Cancer Res Christina K. Haston, Christopher I. Amos, Terri M. King, et al. Fibrosis in the MouseInheritance of Susceptibility to Bleomycin-induced Pulmonary

Updated version

http://cancerres.aacrjournals.org/content/56/11/2596

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

[email protected] at

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/56/11/2596To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Documents

Inheritance of Susceptibility to Bleomycin-induced Pulmonary ......mycin (clinical-grade Blenoxane; gift of Sal Lucania and of Dr. Terry Dugan, Bristol-Myers Squibb, Evansville, IN)