with Lung Tumors Serum Levels of Surfactant ...

Serum Levels of Surfactant Protein D Are Increased in Mice with Lung Tumors Feijie Zhang, William Pao, Sarah M. Umphress, et al. Cancer Res 2003;63:5889-5894.

Updated version

Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/63/18/5889

Cited Articles

This article cites by 36 articles, 23 of which you can access for free at: http://cancerres.aacrjournals.org/content/63/18/5889.full.html#ref-list-1

Citing articles

This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/63/18/5889.full.html#related-urls

E-mail alerts Reprints and Subscriptions Permissions

Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected].

Downloaded from cancerres.aacrjournals.org on June 1, 2013. © 2003 American Association for Cancer Research.

[CANCER RESEARCH 63, 5889 –5894, September 15, 2003]

Serum Levels of Surfactant Protein D Are Increased in Mice with Lung Tumors1 Feijie Zhang, William Pao, Sarah M. Umphress, Sonia B. Jakowlew, Amy M. Meyer, Lori D. Dwyer-Nield, Larry D. Nielsen, Katsuyuki Takeda, Erwin W. Gelfand, James H. Fisher, Lening Zhang, Alvin M. Malkinson, and Robert J. Mason2 National Jewish Medical and Research Center, Denver, Colorado 80206 [F. Z., L. N., K. T., E. W. G., L. Z., R. J. M.]; Denver Health and Hospitals, Denver, Colorado 80204 [J. H. F.]; University of Colorado Health Sciences Center, Denver, Colorado 80262 [A. M. Ma., A. M. Me., L. D. D-N.]; Memorial Sloan Kettering Cancer Center, New York, New York 20021 [W. P.]; and National Cancer Institute, Rockville, Maryland 20850 [S. M. U., S. B. J.]

ABSTRACT Most murine lung tumors are composed of differentiated epithelial cells. We have reported previously that surfactant protein (SP)-D is expressed in urethane-induced tumors. Serum levels of SP-D are increased in patients with interstitial lung disease and acute respiratory distress syndrome and in rats with acute lung injury but have not been measured in mice. In this study, we sought to determine whether SP-D could be detected in murine serum and discovered that it was increased in mice bearing lung tumors. Serum SP-D concentration was 5.0 ⴞ 0.2 ng/ml in normal C57BL/6 mice, essentially absent in SP-D nulls, and 63.6 ⴞ 9.0 ng/ml in SP-D-overexpressing mice. SP-D in serum was verified by immunoblotting. Serum SP-D was increased in mice bearing tumors induced by three different protocols, and the SP-D level correlated with tumor volume. However, in mice with a single adenoma or a few adenomas, SP-D levels were usually within the normal range. SP-D was expressed by the tumor cells, and there was also a field effect whereby type II cells near the tumor expressed more SP-D than type II cells in the remainder of the lung. Serum SP-D was also increased by lung inflammation. In airway inflammation induced by aerosolized ovalbumin in sensitized BALB/c mice, the serum levels were elevated after challenge. In conclusion, serum SP-D concentration is increased in mice bearing lung tumors and generally reflects the tumor burden but is also elevated during lung inflammation.

INTRODUCTION One goal of cancer research is to identify a biomarker in serum that can be used to monitor tumor burden. Such a marker could be used to detect tumors in a relatively noninvasive manner and to evaluate the growth of tumors or their regression in response to therapy. For example, the value of the prostate-specific antigen test as a biomarker for prostate cancer and of ␤-human chorionic gonadotrophin for trophoblastic tumors is well recognized. For lung cancer, a variety of candidate serum biomarkers have been evaluated such as CA125 tumor-associated antigen, carcinoembryonic antigen, and cytokeratin fragment 21.1 (CYFRA 21.1; Ref. 1–5). However, none has proven to be clinically useful, although high levels can indicate a large tumor burden. In a recent study of urethane-induced adenomas in mice, we showed that these tumors expressed relatively high levels of SPs3 on the basis of immunostaining and in situ hybridization (6). These tumors all expressed SP-C, a marker restricted to alveolar type II epithelial cells, but not CC-10, a marker of nonciliated bronchiolar epithelial cells, commonly referred to as Clara cells. Hence, SPs could potentially serve as biomarkers of lung cancer in mice. Of the SPs,

only SP-A and SP-D are hydrophilic and readily measured in serum. Serum levels of SP-A and SP-D are increased in acute lung injury and several interstitial lung diseases in humans (7, 8). In addition, serum SP-D is a biomarker of acute lung injury and inflammation in rats (9). SP-D also has the advantage that it should theoretically be useful in tumors derived from either type II cells or Clara cells. Wikenheiser and Whitsett (10), who studied murine tumors induced by SV40 early region genes under the control of the SP-C promoter, demonstrated that these murine tumors could express either CC-10, a marker of bronchiolar epithelial cells, or SP-C. SP-D was not measured in those mice. SP-D is a calcium-binding lectin that is important in host defense (11, 12). Although SP-D can be detected at very low levels in a variety of tissues, it is most highly expressed in type II cells and Clara cells (12, 13). SP-D is a collagenous glycoprotein whose monomeric structure is composed of an NH2-terminal unit, a long collagenous portion, a neck region, and a highly conserved globular carbohydrate recognition domain that is the functional unit for attachment to microorganisms (11). SP-D binds to viruses, bacteria, yeast, and mycobacteria, and it can also alter inflammatory cell responses (11, 12). Mature SP-D is most commonly composed of four trimeric units with a molecular mass of about 516 kDa. SP-D is very similar to bovine conglutinin, a circulating protein (14). Although SP-D is conventionally referred to as a SP, it binds surfactant poorly and is not found in lamellar bodies. SP-D is probably not a true SP but part of the innate immune system. In a preliminary survey of SP-D serum levels in inflammatory lung disease, we found an extremely high level in a BALB/c mouse that had a spontaneous adenocarcinoma. We therefore sought to determine whether serum SP-D could be used to detect lung epithelial tumors in mice. MATERIALS AND METHODS

Animals. Unless stated differently, serum was collected from control mice by cardiac puncture in anesthetized mice. For each laboratory and tumor model, concomitant control samples were obtained so that genetic differences and variations in animal husbandry between control and the tumor-bearing mice were minimized. The genotype and phenotype of the SP-D transgenic and nulls were as reported previously (15, 16). For these experiments, all SP-D⫺/⫺ mice were outbred into a NIH Swiss Black background eight times (15). SP-D transgenic mice express a rat SP-D cDNA driven by the human SP-C promoter (16). These SP-D transgenic mice express a 10 –20-fold increase in SP-D in lavage fluid and lung homogenates (16). The SP-D transgenic mice were initially developed in C57BL/6 ⫻ C3H hybrids by pronuclear injection and subsequently outbred onto a NIH Swiss Black background eight times. GenReceived 3/31/03; revised 6/11/03; accepted 6/25/03. otypes for mice subject to experimentation were confirmed by both PCR and The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with DNA dot blot analysis. 18 U.S.C. Section 1734 solely to indicate this fact. Ovalbumin Sensitization and Challenge. BALB/c mice were sensitized 1 Supported by grants from the Environmental Protection Agency (Grant R825702) and exposed to aerosolized ovalbumin to produce airway inflammation and and the NIH (Grants HL-67671, HL-36577, HL-61005, CA33497, CA96133, and hyperresponsiveness, as described previously (17, 18). In this protocol there CA09712). W. P. was supported by NIH Grant K12-CA09712. 2 To whom requests for reprints should be addressed, at National Jewish Medical and are two control groups [ i.e., one immunized by i.p. injection (IP only), and one Research Center, 1400 Jackson Street, Denver, CO 80206. Fax: (303) 398-1806; E-mail: that was not sensitized but received nebulized ovalbumin on three successive [email protected]. 3 days (3 Neb only)]. The experimental group is sensitized by i.p. injection and The important abbreviations used are: SP, surfactant protein; TGF, transforming challenged by nebulization (IPN), and this is the group that manifests airway growth factor; TBS, Tris-buffered saline. 5889


SERUM SP-D

hyperresponsiveness to methacholine, airway inflammation, and epithelial reaction was carried out for 5 min at room temperature in a darkened room and metaplasia (17). Serum samples were collected after the third day of nebulized was stopped by the addition of 100 ␮l of 2N sulfuric acid. Absorbance was ovalbumin or 24 h after the third nebulization. This experimental protocol was measured with Microplate Autoreader EL-311s (Bio-Tek Instruments, Inc., approved by the Institutional Animal Care and Use Committee of National Winooski, VT). SP-D Protein Identification. SP-D was partially purified from serum, as Jewish Medical and Research Center. Urethane-induced Adenoma. Male A/J mice at 6 weeks of age were described previously (9). SP-D binds Saccharomyces cerevisiae in a calciumpurchased from Jackson Laboratory (Bar Harbor, ME). Breeder pairs from dependent manner. Aliquots of serum from normal, tumor-bearing, and SP-C/ B6.A chromosome substitution strains (CSS or consomic strains; C57BL/6J TNF-␣ transgenic mice were dialyzed against 5 mM TBS (pH 7.4) containing mice bred to contain one A/J chromosome) were a gift from Dr. Joseph Nadeau 1 mM EDTA to remove glucose, a competitive inhibitor. After dialysis, the sera (Case Western Reserve University, Cleveland, OH) and were bred to provide were adjusted to 5 mM Ca2⫹ with CaCl2 and incubated with paraformaldehydemice for experimental use at the University of Colorado Health Sciences fixed S. cerevisiae cells (strain SEY6210) for 2 h at room temperature by Center Center for Laboratory Animal Care. Five A/J mice, 21 male B6.AY gentle shaking. After incubation, the cells were washed three times with TBS mice, and 7 male and 6 female B6.A14 CSS mice (each at 8 weeks old) all containing 5 mM CaCl2. The cells were then incubated with 50 mM inositol for received i.p. injection with 1 mg/g urethane in saline as described previously 1 h at room temperature to remove bound SP-D. The cells were centrifuged, (19). Twenty weeks after urethane treatment, mice were euthanized with a and the supernatant was removed, pooled, and analyzed for SP-D by immulethal dose of pentobarbital, and plasma was collected by cardiac puncture. noblotting. The samples were placed in Laemmli reducing sample buffer, Lung tumors were dissected and counted under a dissecting microscope (Wild boiled, and layered onto precast 8 –16% Tris-glycine polyacrylamide slab gels Inc., Heerbrugg, Switzerland). Plasma was also collected from four age- (Novex), and proteins were separated by electrophoresis in a Novex X cell Mini-cell (Invitrogen, Carlsbad, CA). Proteins were transferred to nitrocellumatched A/J control mice (26 weeks of age). TGF-␤ Deletion and K-ras Mutation. C57BL/6NCr mice heterozygous lose membranes using the Novex X cell blot module according to the manufor deletion of one TGF-␤1 allele (TGF-␤1⫹/⫺; Ref. 20) were mated with facturer’s instructions. Nonspecific binding sites on the nitrocellulose memC57BL/6/129/SV mice heterozygous for latent mutation of one K-ras allele branes were blocked by incubating the blots in 5% nonfat dry milk in 20 mM (K-ras⫹/⫺; Ref. 21). This mating generated four genotypes (TGF-␤1⫹/⫹/K- Tris-HCl (pH 7.5), 137 mM NaCl, and 0.05% Tween 20 (TTBS; all from ras⫹/⫹, TGF-␤1⫹/⫺/K-ras⫹/⫹, TGF-␤1⫹/⫹/K-ras⫹/⫺, and TGF-B1⫹/⫺/K-ras⫹/ Sigma, St. Louis, MO) at 4°C overnight. The blots were incubated with rabbit antimouse SP-D IgG in 5% nonfat milk in TTBS (final concentration, 0.35 ⫺). Confirmation of the genotypes was performed on DNA isolated from the tails of each offspring by using PCR amplification. Specific TGF-␤1 and K-ras ␮g/ml) for 1 h at room temperature. Horseradish peroxidase-conjugated secoligonucleotide primers were designed to distinguish heterozygous from wild- ondary antibodies (Jackson ImmunoResearch Laboratories, West Grove, PA) type littermates (20, 21). All mice were maintained in a pathogen-free barrier were added for 30 min at room temperature, and the antigens were detected by facility under conditions of constant room temperature (22 ⫾ 1°C) and on a enhanced chemiluminescence (ECL-plus; Amersham Pharmacia Biotech, Pis12-h light/12-h dark cycle. Water and rodent food were available ad libitum. cataway, NJ) on Hyperfilm (Amersham Pharmacia Biotech). Immunostaining. This was done with polyclonal rabbit IgG antimurine Wild-type littermates and heterozygous mice were humanely sacrificed, serum was collected by cardiac puncture, and the lungs were removed quickly. Total SP-D as detailed previously (6, 23). In Situ Hybridization. Tissues were fixed with 4% paraformaldehyde, tumor number was determined, and the sizes of individual tumors were measured with a metric caliper. At all times, mice were treated in accordance embedded in paraffin, and sectioned. The sections were dewaxed, hybridized with sense or antisense riboprobes to rat SP-D or SP-C, and processed as with the guidelines for animal care and use established by the NIH. Inducible Activated K-ras Mouse Model of Lung Cancer. Mice that reported previously (6, 24, 25). Data Analyses. All statistical tests were done with the SAS statistical form lung tumors due to expression of a tetracycline-regulated K-ras oncogene were generated as described previously (22). Bitransgenic (CCSP-rtTA/tet-op- analysis package (version 8.2). Significance level was set to 0.05 for all K-ras4bG12D) animals and littermate controls were bred on wild-type (FVB), comparisons. For Table 1 and Fig. 1, we compared multiple groups with a p53-deficient (FVB), or INK4a-Arf-deficient (C57BL/6) backgrounds. Geno- single control group. Because the groups had different variances, Dunnett’s typing was performed as published. The expression of activated K-ras was multiple comparison test was used with the Proc Mixed procedure where each group has its own variance. The nonparametric Wilcoxon’s rank-sum test was induced in lung epithelia by feeding adult animals (4 – 6 weeks of age) doxycycline-impregnated food pellets (Harlan-Teklad, Madison, WI) for vary- used for the comparisons in Fig. 2, A and B, and Fig. 5. For Fig. 4, both tumor ing amounts of time, ranging from 3 weeks to 11 months. Serum was collected volume and SP-D levels were first transformed to their logarithmic scale of from mice by retroorbital eye bleed in anesthetized mice before sacrifice. Half base 10 because both variables were not normally distributed. The transformed of the left lung was prepared for H&E staining. Multiple sections were data were then fitted to a linear regression equation using Proc GLM procedure. For the sensitivity and specificity analysis, tumor data were combined examined by light microscopy to determine the presence or absence of tumor. with normal control data. Because tumor volumes for normal group were all This experimental protocol was approved by the animal care committee at zero, logarithm (base 10) transformation was performed on (tumor volMemorial Sloan-Kettering Cancer Center. ELISA for Mouse SP-D. Serum SP-D levels were measured by ELISA. ume ⫹ 1). A linear regression equation was derived to predict either tumor Initially, the samples were coded, and the ELISA was performed without volume or SP-D if only one of them is known. We used whether or not the knowledge of the tumor status of the mice. When it was clear that some of the values for tumor-bearing mice were very high, appropriate dilutions of the serum were made based on the presence or absence of a tumor so as to avoid Table 1 Serum SP-D concentration in control mice wasting serum. Recombinant mouse SP-D produced in Chinese hamster ovary No. of SP-D (ng/ml) cells was used as the SP-D standard. Polyclonal antirecombinant mouse SP-D Animals Strain animals (mean ⫾ SE) rabbit IgG (10 ␮g/ml in 0.1 M sodium bicarbonate) was bound to wells in Normal microtiter plates (Immulon 1 plates; Dynatech Laboratories, Alexandria, VA) 84 4.8 ⫾ 0.2 C57BL/6a overnight at room temperature. The wells were then incubated for at least 30 FVB/N 12 5.0 ⫾ 0.6 b 9 8.3 ⫾ 1.8 A/J min at room temperature with a 4% (w/v) solution of nonfat dry milk in PBS Swiss Webster 5 4.9 ⫾ 1.0 containing 1% Triton X-100 (blocking buffer) to block nonspecific binding. BALB/c 4 4.2 ⫾ 1.5 The wells were washed with blocking buffer, and 100 ␮l of purified mouse SP-D transgenic SP-D (0 –20 ng/ml) for standards or appropriately diluted mouse serum samSwiss Webster 3 63.6 ⫾ 9.0c SP-D null ples were added to each well. Plates were incubated for 90 min at 37°C and Swiss Webster 5 0.04 ⫾ 0.03c then washed with blocking buffer. One hundred ␮l of horseradish peroxidasea The C57BL/6 mice are derived from Jackson Laboratories stock at National Jewish conjugated anti-SP-D antibody (10 ␮g/ml) were added to each well, and the and Research Center as well as C57BL/6 from the National Cancer Institute. plates were incubated for 90 min at 37°C. After washing with 1% Triton X-100 Medical b Plasma instead of serum was measured in the A/J mice. The ELISA was performed in PBS, 100 ␮l of the color-developing agent [0.1% O-phenylenediamine and as stated in “Materials and Methods.” c 0.015% hydrogen peroxide in 0.1 M citrate buffer (pH 4.6)] were added. The Groups that differ from normal control mice [all strains (P ⬍ 0.05)]. 5890


SERUM SP-D

Fig. 1. Serum SP-D levels in mice after ovalbumin sensitization and challenge. BALB/c mice were divided into five groups. One group received two i.p. injections of ovalbumin (2 IP only; n ⫽ 8), a second group received nebulized ovalbumin on three successive days (3 Neb only; n ⫽ 6), a third group received i.p. sensitization and nebulized challenge and was sacrificed immediately after challenge (IPN 0 hr; n ⫽ 7), a fourth group was given the same treatment as the third group but sacrificed 24 h after the third nebulized challenge (IPN 24 hr; n ⫽ 12), and a fifth group was sacrificed at 48 h after challenge (IPN 48 hr; n ⫽ 12). SP-D levels were measured by ELISA as stated in “Materials and Methods.” ⴱ, P ⬍ 0.05.

mouse had a tumor as the outcome variable and calculated the sensitivity and specificity for confidence limits of 95% or 99% for serum SP-D.

RESULTS Normal Animals. Serum SP-D levels in normal mice (4 – 8 ng/ml) were lower than values determined previously in rats (40 – 80 ng/ml) and humans [70 –110 ng/ml (7, 9)]. The serum level did not vary significantly with age, gender, or strain of mice. As shown in Table 1, SP-D was increased in transgenic mice that overexpress SP-D in type II cells driven by the SP-C promoter and was essentially zero in SP-D nulls. Ovalbumin Sensitization and Challenge. In a preliminary study, we measured SP-D in sera of mice sensitized and challenged with nebulized ovalbumin in a well-characterized model of airway inflammation and hyperresponsiveness (17). We knew from previous work that lavage levels of SP-D were elevated after Aspergillus fumigatus allergen challenge in sensitized mice (26) and in mice challenged with ovalbumin.4 Serum SP-D levels increased after challenge (Fig. 1). However, one value was excluded as an outlier because the value was 1420 ng/ml. This animal had a large tumor, which stained positive for SP-D (data not shown), and provided the rationale for the subsequent experiments. Urethane-induced Adenoma. Because urethane-induced lung adenomas express SP-D (6), we evaluated serum from mice that had few [1–2 adenoma(s); CSS mice] or many (30 –50 adenomas; A/J) adenomas. Serum SP-D was elevated in A/J mice bearing multiple tumors over their age-matched controls (Fig. 2A). All CSS mice received injection with urethane, and SP-D plasma levels were grouped according to whether the mouse did or did not develop lung tumors. Most CSS mice with 1 or 2 tumors had serum SP-D levels within the normal range, similar to that in animals bearing no tumors (Fig. 2B). Serum SP-D levels in mice without tumors were slightly higher than those in other cohorts. This could be because all of these animals were exposed to urethane, these CSS strains might exhibit higher basal SP-D concentrations (basal levels have not been measured in these 4

K. Takeda, unpublished observations.

strains), or they may contain microscopic adenomas undetectable under the dissecting microscope. We had reported previously that urethane-induced tumors expressed SP-D based on immunostaining (6). Here, we used in situ hybridization to demonstrate the relative mRNA levels in the tumors. Although these tumors expressed SP-D as expected (Fig. 3), there was also enhanced expression of SP-D mRNA in the type II cells surrounding the tumor. This proximity or field effect was not observed for SP-C expression. The implication is that the tumors produce a substance(s) that increases the expression of SP-D in neighboring normal type II cells. Tumors Induced by Heterozygous Deletion of TGF-␤ and Activated K-ras. The next goal was to determine whether the rise in serum SP-D correlated with tumor volume. For this study, we chose a different mode of carcinogenesis. Heterozygous TGF-␤1⫹/⫺ mice (C-57Bl/6NCr TGF-␤1⫹/⫺) were mated with K-ras⫹/⫺ latent activational mice (C57 BL/6/129/sv K-ras⫹/⫺) to produce four genotypes. Wild-type (TGF-␤1⫹/⫹/K-ras⫹/⫹) mice produced no tumors. Heterozygotes produce an intermediate number of tumors, and double heterozygotes (TGF-␤1⫹/⫺/K-ras⫹/⫺) produced many tumors. The mice were sacrificed at 2–14 months of age, and the tumor diameter and number were recorded, and serum was collected. All tumors were assumed to be spherules for calculating tumor volumes. The remarkable observation was that the serum SP-D levels correlated with tumor volume on the logarithmic scale with an R2 value of 0.84 (Fig. 4). If one uses tumor volume as the response variable and SP-D as the predictor, the linear equation is Log10 (mm3 volume ⫹1) ⫽ ⫺0.656 ⫹ 1.003 ⫻ Log10 (ng/ml SP-D). Only animals with tumors were included in this analysis to prevent anchoring the curve at the low end due to the large number of normal samples. Mice whose

Fig. 2. Serum SP-D levels in mice with urethane-induced adenoma. Because urethaneinduced lung adenomas express SP-D (6), we evaluated serum from mice that had many adenomas (30 –50 adenomas; A/J mice; A) and mice that had few adenomas (1–2 adenomas; CSS mice; B). Serum SP-D was elevated in A/J mice bearing multiple tumors over their age-matched controls. Shown in A are the results from four normal mice (3.8 ⫾ 0.2 ng/ml) and five tumor-bearing mice (50.8 ⫾ 6.3 ng/ml; P ⬍ 0.05). All CSS mice received injection with urethane, and the SP-D plasma levels were grouped according to whether the mouse did or did not develop lung tumors. Most CSS mice had only 1 or 2 tumors (n ⫽ 14) and had serum SP-D levels within the normal range, similar to that in animals bearing no tumors (n ⫽ 20; B). Serum SP-D levels in mice without tumors were slightly higher than those in other cohorts.

5891


SERUM SP-D

serum values were greater than 11.4 ng/ml all had tumors in this data set. All samples from tumor-free mice had serum values of ⬍11.4 ng/ml (specificity is 100%), and 80% of the tumor-bearing mice had serum values of ⬎11.4 ng/ml (sensitivity is 80%). A serum level of 11.4 ng/ml corresponds to a total tumor burden of 1.84 mm3. Conditional Induction of Lung Tumors with Activated K-ras. To extend these observations to a third experimental system, we measured SP-D levels in serum from mice bearing lung adenomas and adenocarcinomas due to the induced expression of an activated K-ras (22). This protocol used bitransgenic animals (on wild-type or tumor suppressor-deficient backgrounds) in which the expression of murine K-ras4bG12D in lung epithelial cells is under the control of doxycycline. Tumors express SP-C, a type II cell marker, but not CC-10 (10-kDa Clara cell-associated protein), a Clara cell marker. This protocol was chosen because of the rapid induction of tumors and their propensity for malignant transformation. In our studies, mice on wild-type or tumor suppressor-deficient backgrounds were tested after the addition of doxycycline for 3 weeks to 11 months (22). Similar to the other inductive protocols, serum SP-D level differentiated the control mice from the tumor-bearing mice (Fig. 5). Verification of SP-D in Serum. SP-D has been isolated from human and rat serum, and its identity has been verified by immunoblotting (8, 9). We isolated SP-D from serum by binding to yeast and eluting with inositol, as we reported previously (9). As shown in Fig.

Fig. 4. Relationship of serum SP-D levels to tumor volume. For this figure, tumors developed spontaneously in mice with a latently activated K-ras with or without a heterozygous deletion of TGF-␤1. Only the 40 mice with tumors are shown; in this cohort there were also 80 mice without tumors. There is a linear relationship of Log10 serum SP-D to Log10 tumor volume.

Fig. 5. Serum SP-D levels in mice with regulated expression of an activated K-ras oncogene. Tumor-bearing mice had significantly elevated levels of serum SP-D (157.7 ⫾ 41.6 ng/ml; n ⫽ 12) in comparison with littermate controls that did not have tumors (5.9 ⫾ 0.6 ng/ml; n ⫽ 10; P ⬍ 0.01). Serum was derived from bitransgenic (CCSP-rtTA/tet-op-K-ras4bG12D) animals and littermate controls on various genetic backgrounds (see “Materials and Methods”) that had been given doxycycline via the diet for varying amounts of time. ⴱ, P ⬍ 0.01.

6, a 43-kDa protein was isolated from serum, which has the same molecular mass as recombinant SP-D, and was recognized by a rabbit polyclonal IgG specific for murine SP-D. We chose serum from TNF-␣-overexpressing mice as a positive control because we knew that the serum level in these mice was 343 ⫾ 28 ng/ml (n ⫽ 4). These mice have significant chronic lung inflammation (27). Under nonreducing conditions, both serum SP-D and recombinant SP-D showed higher-order oligomers characteristic of SP-D (data not shown). DISCUSSION

Fig. 3. In situ hybridization for SP-D. Strain A/J mice that had received injection with urethane were sacrificed, and the lungs were fixed for in situ hybridization with 4% paraformaldehyde. In situ hybridization was performed as described in “Materials and Methods.” The tumor clearly expresses SP-D. A shows a bright-field image, and B shows a dark-field image. Note that there is also increased expression in the alveolar type II cells near the tumor as opposed to those further away. There was no neighborhood effect in the expression of SP-C (C). The original magnification was ⫻100. Sense riboprobe controls were negative as reported previously (6).

Serum SP-D is elevated in mice with lung tumors, and the values correlate with tumor burden in this cross-sectional study. The ELISA measurement was reasonably sensitive and detected tumors greater than 1.84 ␮m3 (diameter, 1.52 ␮m). Serum SP-D was elevated in all types of tumors tested, which included chemically induced tumors, genetically induced tumors, and conditionally regulated mutant K-rasinduced tumors. However, an elevated serum SP-D is not specific for lung tumors and was also observed with lung inflammation. For example, serum SP-D was elevated in mice that overexpressed

5892


SERUM SP-D

Fig. 6. Immunoblotting of SP-D in serum. SP-D was partially purified from serum by adsorption to S. cerevisiae, eluted with inositol, and immunoblotted as described previously (9). Lanes 1–3 contain 5, 10, and 20 ng of recombinant murine SP-D, respectively; Lanes 4 – 6 contain SP-D isolated from serum of mice with tumors (estimated to be 18, 9, and 5 ng by ELISA); Lane 7 contains SP-D from mice overexpressing TNF-␣ under the SP-C promoter (estimated to be 4 ng by ELISA); and Lane 8 contains SP-D isolated from serum of normal C57BL/6 mice. All samples gave one band under denatured and reduced conditions that was identified by the polyclonal rabbit antimurine SP-D with an apparent molecular mass of 43 kDa. The isolation of SP-D from serum was performed twice with identical results.

TNF-␣, during respiratory syncytial virus infections, after ozone exposure, and after ovalbumin challenge in sensitized mice (some data not shown). Serum SP-D measurement is a relatively simple means of distinguishing normal mice from those with tumors or those with chronic inflammation. Serum SP-D will likely increase with tumor growth and decrease with regression, although this has not been tested in a prospective manner. Because measurement of tumor volume is difficult, and imaging techniques are not widely available, serum SP-D measurements should be useful to identify mice with tumors or to monitor response to therapy over time. The importance of this finding for human lung cancer remains to be defined. Some human adenocarcinomas and large cell carcinomas express SP-D or SP-A, whereas small cell tumors and squamous cell tumors do not (28 –32). However, about 75% of pulmonary adenocarcinomas express thyroid transcription factor 1, and about 50% of pulmonary adenocarcinomas express SP-A (32). Expression of SP-D should be similar to SP-A in adenocarcinoma (28). Although a large study of serum SP-D in human lung cancers has not been done, serum values of SP-D are elevated in some patients with adenocarcinoma and increase further if patients develop radiation pneumonitis (33, 34). To define the role of serum SP-D in human lung cancer, serum values of SP-D will need to be determined in a large cross-sectional study with multiple tumor histological types. In our murine tumor study, only mice with adenomas or adenocarcinomas were tested. Because normal human serum levels are higher and more variable than those in mice (7), serum SP-D levels may not be as distinctive a biomarker for lung tumors in humans. Serum SP-D levels may only be useful in that subset of patients with adenocarcinomas, such as the bronchioloalveolar carcinoma subtype. The mechanism responsible for the increased serum level of SP-D is not known. The most likely reason for this elevation is the production of SP-D by epithelial cells within the tumors. Tumor cells express SP-D mRNA and have detectable SP-D by immunostaining (6). Presumably, SP-D is secreted, enters the extracellular space, and reaches the systemic circulation via local lymphatic drainage. It is unlikely that the bronchoalveolar lavage fluid levels of SP-D would vary greatly depending on the presence or absence of tumors, although these values have not been measured to date. However, we cannot eliminate a contribution from the neighboring type II cells that overexpress SP-D. Although SP-D can be expressed on many mucosal surfaces, the highest concentration is found in the lung and bronchoalveolar lavage fluid (12, 13, 35). It is highly unlikely that the increased serum SP-D came from an extrapulmonary source. It is also unlikely that the volume of distribution or clearance of SP-D from the vascular compartment is altered in animals with tumors, although decreases in either parameter would theoretically increase serum concentrations. The cause of the increased expression of SP-D in the type II cells surrounding the tumor is unknown. In vitro, SP-D expression is altered by factors that affect the state of differentiation of type II cells

but, in general, is less responsive to alterations in culture conditions than expression of other SPs (36, 37).5 In vivo, pulmonary SP-D expression is increased by acute inflammation (endotoxin, antigen challenge, and so forth) and by overexpression of interleukin 4 (38, 39). We presume that some factor is expressed in the tumor that diffuses to the neighboring type II cells and increases SP-D expression. The factor could be produced by inflammatory cells within the tumors (40). However, this factor(s) has not been identified and will be the subject of future studies. In summary, murine lung tumors express SP-D, and the tumor burden is reflected in the serum level of SP-D. However, serum SP-D is also elevated with lung inflammation, and SP-D is not specific for lung tumors. ACKNOWLEDGMENTS We thank Mandy Evans, Dr. Dennis Voelker, and Dr. Erika Crouch for the Chinese hamster ovary cell line that expresses murine SP-D and the rabbit antisera and Jennifer Doherty, Galen Fisher, and Harold Varmus for the conditional induction of lung tumors with activated K-ras.

REFERENCES 1. Sawabata, N., Ohta, M., Takeda, S., Hirano, H., Okumura, Y., Asada, H., and Maeda, H. Serum carcinoembryonic antigen level in surgically resected clinical stage I patients with non-small cell lung cancer. Ann. Thorac. Surg., 74: 174 –179, 2002. 2. Buccheri, G., and Ferrigno, D. Serum biomarkers facilitate the recognition of earlystage cancer and may guide the selection of surgical candidates: a study of carcinoembryonic antigen and tissue polypeptide antigen in patients with operable non-small cell lung cancer. J. Thorac. Cardiovasc. Surg., 122: 891– 899, 2001. 3. Salgia, R., Harpole, D., Herndon, J. E., II, Pisick, E., Elias, A., and Skarin, A. T. Role of serum tumor markers CA 125 and CEA in non-small cell lung cancer. Anticancer Res., 21: 1241–1246, 2001. 4. Kim, Y. C., Lim, S. C., Bom, H. S., Park, K. O., Na, K. J., Park, H. K., and Hwang, J. H. Different cutoff values of Cyfra 21-1 for cavitary and noncavitary lung cancers. Lung Cancer, 30: 187–192, 2000. 5. Watine, J. Prognostic evaluation of primary non-small cell lung carcinoma patients using biological fluid variables. A systematic review. Scand. J. Clin. Lab. Investig., 60: 259 –273, 2000. 6. Mason, R. J., Kalina, M., Nielsen, L. D., Malkinson, A. M., and Shannon, J. M. Surfactant protein C expression in urethane-induced murine pulmonary tumors. Am. J. Pathol., 156: 175–182, 2000. 7. Greene, K. E., King, T. E., Jr., Kuroki, Y., Bucher-Bartelson, B., Hunninghake, G. W., Newman, L. S., Nagae, H., and Mason, R. J. Serum surfactant proteins-A and -D as biomarkers in idiopathic pulmonary fibrosis. Eur. Respir. J., 19: 439 – 446, 2002. 8. Kuroki, Y., Takahashi, H., Chiba, H., and Akino, T. Surfactant proteins A and D: disease markers. Biochim. Biophys. Acta, 1408: 334 –345, 1998. 9. Pan, T., Nielsen, L. D., Allen, M. J., Shannon, K. M., Shannon, J. M., Selman, M., and Mason, R. J. Serum SP-D is a marker of lung injury in rats. Am. J. Physiol. Lung Cell Mol. Physiol., 282: L824 –L832, 2002. 10. Wikenheiser, K. A., and Whitsett, J. A. Tumor progression and cellular differentiation of pulmonary adenocarcinomas in SV40 large T antigen transgenic mice. Am. J. Respir. Cell Mol. Biol., 16: 713–723, 1997. 11. Crouch, E., and Wright, J. R. Surfactant proteins A and D and pulmonary host defense. Annu. Rev. Physiol., 63: 521–554, 2001. 12. Mason, R. J., Greene, K., and Voelker, D. R. Surfactant protein A and surfactant protein D in health and disease. Am. J. Physiol., 275: L1–L13, 1998. 13. Stahlman, M. T., Gray, M. E., Hull, W. M., and Whitsett, J. A. Immunolocalization of surfactant protein-D (SP-D) in human fetal, newborn, and adult tissues. J. Histochem. Cytochem., 50: 651– 660, 2002. 14. Hansen, S., and Holmskov, U. Lung surfactant protein D (SP-D) and the molecular diverted descendants: conglutinin, CL-43 and CL-46. Immunobiology, 205: 498 – 517, 2002. 15. Korfhagen, T. R., Sheftelyevich, V., Burhans, M. S., Bruno, M. D., Ross, G. F., Wert, S. E., Stahlman, M. T., Jobe, A. H., Ikegami, M., Whitsett, J. A., and Fisher, J. H. Surfactant protein-D regulates surfactant phospholipid homeostasis in vivo. J. Biol. Chem., 273: 28438 –28443, 1998. 16. Fisher, J. H., Sheftelyevich, V., Ho, Y. S., Fligiel, S., McCormack, F. X., Korfhagen, T. R., Whitsett, J. A., and Ikegami, M. Pulmonary-specific expression of SP-D corrects pulmonary lipid accumulation in SP-D gene-targeted mice. Am. J. Physiol. Lung Cell Mol. Physiol., 278: L365–L373, 2000. 17. Tomkinson, A., Cieslewicz, G., Duez, C., Larson, K. A., Lee, J. J., and Gelfand, E. W. Temporal association between airway hyperresponsiveness and airway eosinophilia in ovalbumin-sensitized mice. Am. J. Respir. Crit. Care Med., 163: 721–730, 2001. 5

R. J. Mason, unpublished observations.

5893


SERUM SP-D

18. Takeda, K., Haczku, A., Lee, J. J., Irvin, C. G., and Gelfand, E. W. Strain dependence of airway hyperresponsiveness reflects differences in eosinophil localization in the lung. Am. J. Physiol. Lung Cell Mol. Physiol., 281: L394 –L402, 2001. 19. Malkinson, A. M., and Beer, D. S. Major effect on susceptibility to urethane-induced pulmonary adenoma by a single gene in BALB/cBy mice. J. Natl. Cancer Inst. (Bethesda), 70: 931–936, 1983. 20. Kang, Y., Mariano, J. M., Angdisen, J., Moody, T. W., Diwan, B. A., Wakefield, L. M., and Jakowlew, S. B. Enhanced tumorigenesis and reduced transforming growth factor-␤ type II receptor in lung tumors from mice with reduced gene dosage of transforming growth factor-␤1. Mol. Carcinog., 29: 112–126, 2000. 21. Johnson, L., Mercer, K., Greenbaum, D., Bronson, R. T., Crowley, D., Tuveson, D. A., and Jacks, T. Somatic activation of the K-ras oncogene causes early onset lung cancer in mice. Nature (Lond.), 410: 1111–1116, 2001. 22. Fisher, G. H., Wellen, S. L., Klimstra, D., Lenczowski, J. M., Tichelaar, J. W., Lizak, M. J., Whitsett, J. A., Koretsky, A., and Varmus, H. E. Induction and apoptotic regression of lung adenocarcinomas by regulation of a K-Ras transgene in the presence and absence of tumor suppressor genes. Genes Dev., 15: 3249 –3262, 2001. 23. Xu, X., McCormick-Shannon, K., Voelker, D. R., and Mason, R. J. KGF increases SP-A and SP-D mRNA levels and secretion in cultured rat alveolar type II cells. Am. J. Respir. Cell Mol. Biol., 18: 168 –178, 1998. 24. Deterding, R. R., Shimizu, H., Fisher, J. H., and Shannon, J. M. Regulation of surfactant protein D expression by glucocorticoids in vitro and in vivo. Am. J. Respir. Cell Mol. Biol., 10: 30 –37, 1994. 25. Deterding, R. R., and Shannon, J. M. Proliferation and differentiation of fetal rat pulmonary epithelium in the absence of mesenchyme. J. Clin. Investig., 95: 2963– 2972, 1995. 26. Haczku, A., Atochina, E. N., Tomer, Y., Chen, H., Scanlon, S. T., Russo, S., Xu, J., Panettieri, R. A., Jr., and Beers, M. F. Aspergillus fumigatus-induced allergic airway inflammation alters surfactant homeostasis and lung function in BALB/c mice. Am. J. Respir. Cell Mol. Biol., 25: 45–50, 2001. 27. Fujita, M., Shannon, J. M., Irvin, C. G., Fagan, K. A., Cool, C., Augustin, A., and Mason, R. J. Overexpression of tumor necrosis factor-␣ produces an increase in lung volumes and pulmonary hypertension. Am. J. Physiol. Lung Cell Mol. Physiol., 280: L39 –L49, 2001. 28. Betz, C., Papadopoulos, T., Buchwald, J., Dammrich, J., and Muller-Hermelink, H. K. Surfactant protein gene expression in metastatic and micrometastatic pulmonary adenocarcinomas and other non-small cell lung carcinomas: detection by reverse transcriptase-polymerase chain reaction. Cancer Res., 55: 4283– 4286, 1995. 29. Mizutani, Y., Nakajima, T., Morinaga, S., Gotoh, M., Shimosato, Y., Akino, T., and Suzuki, A. Immunohistochemical localization of pulmonary surfactant apoproteins in

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

various lung tumors. Special reference to nonmucus producing lung adenocarcinomas. Cancer (Phila.), 61: 532–537, 1988. Bejarano, P. A., Baughman, R. P., Biddinger, P. W., Miller, M. A., Fenoglio-Preiser, C., al-Kafaji, B., Di Lauro, R., and Whitsett, J. A. Surfactant proteins and thyroid transcription factor-1 in pulmonary and breast carcinomas. Mod. Pathol., 9: 445– 452, 1996. Abutaily, A. S., Addis, B. J., and Roche, W. R. Immunohistochemistry in the distinction between malignant mesothelioma and pulmonary adenocarcinoma: a critical evaluation of new antibodies. J. Clin. Pathol., 55: 662– 668, 2002. Zamecnik, J., and Kodet, R. Value of thyroid transcription factor-1 and surfactant apoprotein A in the differential diagnosis of pulmonary carcinomas: a study of 109 cases. Virchows Arch., 440: 353–361, 2002. Sasaki, R., Soejima, T., Matsumoto, A., Maruta, T., Yamada, K., Ota, Y., Kawabe, T., Nishimura, H., Sakai, E., Ejima, Y., and Sugimura, K. Clinical significance of serum pulmonary surfactant proteins A and D for the early detection of radiation pneumonitis. Int. J. Radiat. Oncol. Biol. Phys., 50: 301–307, 2001. Takahashi, H., Kuroki, Y., Tanaka, H., Saito, T., Kurokawa, K., Chiba, H., Sagawa, A., Nagae, H., and Abe, S. Serum levels of surfactant proteins A and D are useful biomarkers for interstitial lung disease in patients with progressive systemic sclerosis. Am. J. Respir. Crit. Care Med., 162: 258 –263, 2000. Akiyama, J., Hoffman, A., Brown, C., Allen, L., Edmondson, J., Poulain, F., and Hawgood, S. Tissue distribution of surfactant proteins A and D in the mouse. J. Histochem. Cytochem., 50: 993–996, 2002. Shannon, J. M., Pan, T., Nielsen, L. D., Edeen, K. E., and Mason, R. J. Lung fibroblasts improve differentiation of rat type II cells in primary culture. Am. J. Respir. Cell Mol. Biol., 24: 235–244, 2001. Mason, R. J., Lewis, M. C., Edeen, K. E., McCormick-Shannon, K., Nielsen, L. D., and Shannon, J. M. Maintenance of surfactant protein A and D secretion by rat alveolar type II cells in vitro. Am. J. Physiol. Lung Cell Mol. Physiol., 282: L249 –L258, 2002. Ikegami, M., Whitsett, J. A., Chroneos, Z. C., Ross, G. F., Reed, J. A., Bachurski, C. J., and Jobe, A. H. IL-4 increases surfactant and regulates metabolism in vivo. Am. J. Physiol. Lung Cell Mol. Physiol., 278: L75–L80, 2000. McIntosh, J. C., Swyers, A. H., Fisher, J. H., and Wright, J. R. Surfactant proteins A and D increase in response to intratracheal lipopolysaccharide. Am. J. Respir. Cell Mol. Biol., 15: 509 –519, 1996. Beer, D. G., and Malkinson, A. M. Genetic influence on type 2 or Clara cell origin of pulmonary adenomas in urethane-treated mice. J. Natl. Cancer Inst. (Bethesda), 75: 963–969, 1985.

5894
