Presence Of Mouse Mammary Tumor Virus Specifically

Presence of mouse mammary tumor virus specifically alters the body odor of mice Kunio Yamazaki*, Edward A. Boyse†, Judith Bard†, Maryanne Curran*, David Kim‡, Susan R. Ross‡, and Gary K. Beauchamp*§ *Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA 19104; †Department of Microbiology and Immunology, University of Arizona College of Medicine, Tucson, AZ 85724; and ‡Department of Microbiology, University of Pennsylvania, Philadelphia, PA 19104 Contributed by Edward A. Boyse, February 15, 2002

It has long been recognized that various genetic and metabolic human disorders alter body odor, which is not surprising because they may alter body chemistry. Thus, it has been suggested that some human diseases may be diagnosed by odor alone. In that regard, the mouse mammary tumor virus (MMTV) and its tumors of mice, which may have human counterparts, are of special interest because of the need for basic research possible only in inbred and genetically defined animals. Accordingly, we now show that the mouse MMTV, whether obtained environmentally or genetically transmitted, alters the body odor of mice in both males and females, and regardless of the presence or absence of tumors. These observations, together with the prospect of artificial human odor discrimination, may aid in the search for early human diagnostics. chemosensation 兩 olfaction 兩 pheromones 兩 disease diagnosis

L

ittle systematic study of the potential role of odors for disease diagnosis has been conducted despite the many indications that such odors are produced (1, 2) and may be used in medical practice (for examples, see refs. 3 and 4). In part, this result may be due to the vast individual differences in human body odor because of genetic and environmental influences that confound rigorous testing. Thus, an animal model for which genetic and environmental factors are held constant and only the presence or absence of the disease vector is allowed to vary provides a valuable system to study body odor and disease. Mouse mammary tumor virus (MMTV) affords a favorable model with which to test for changes in odor profiles that arise before overt disease from infection with an oncogenic retrovirus or from its premalignant (mammary nodules) and malignant sequelae (5). Such mammary tumors are notably lacking in cachectic, metastatic, and other general systemic effects on the host that might be expected to alter body odor in a nonspecific manner (e.g., ref. 6). Infectious MMTV is acquired by newborn pups when suckling on mothers that shed virus into milk. MMTV replicates by reverse transcription of its RNA genome into DNA, leading to chromosomal integration in infected cells (5). Mouse strains that are genetically susceptible to MMTV infection have a high incidence of mammary tumors, preceded by a premalignant phase of increasing occurrence of mammary nodules. Tumor formation is hormone dependent, with the incidence of nodules and tumors greater in multiparous females (5). Because life-long productive infection is most easily induced when the virus is received during the postnatal period of immunological tolerance and because MMTV is characteristically transmitted in the milk, strains of genetically identical infected and uninfected mice can be produced at will by foster nursing. These mice will differ only in presence or absence of productive MMTV infection. As shown below, these mice also develop a distinct odor as a consequence of the infection. MMTV infection is known to require competent immune function, and MHC Class II molecules play a central role in its life cycle. During the course of an MMTV infection, a virally encoded protein termed the superantigen (Sag) is presented by 5612–5615 兩 PNAS 兩 April 16, 2002 兩 vol. 99 兩 no. 8

the MHC Class II on B cells to T cells bearing specific V␤ chains of the T cell receptor (7, 8). After their initial activation, these T cells are deleted from the immune repertoire through apoptosis. Because MMTV infection has such a profound effect on the T cell repertoire of infected animals, it is possible that the viral phenotypic odor we document below is related to this alteration in the immune system mechanisms. MMTV can also be transmitted genetically as an endogenous provirus. Most mouse strains have one or more endogenous proviruses, but they rarely produce viral particles that can be transmitted exogenously. As with exogenous MMTV, endogenous proviruses cause specific deletion of T cell subpopulations during the neonatal shaping of the immune repertoire. Thus, mice with endogenous MMTVs, rather than showing a gradual deletion of superantigen-cognate T cells, are essentially T cell deleted from birth (reviewed in ref. 7). Similarly, transgenic mice that express MMTV Sags also show neonatal deletion of cognate T cells (7). Consequently, if the effects of exogenous MMTV are due to activity of the viral genes on immune function, endogenous or transgene-encoded MMTV should also be characterized by a specific odor. As further shown below, this result is what we found. Materials and Methods Mice. Three classes of mice were used as follows: (i ) Odor donor mice with infectious MMTV. These mice with infectious MMTV were bred from stock obtained from J. Hilgers (The Netherlands Cancer Institute, Amsterdam) of infected (MMTV⫹) and uninfected control (MMTV⫺) strains (BALB兾cHeAC3H-MMTV:5). (ii ) Odor donor mice with endogenous MMTV. Donor mice with MMTV as a transgene were C3H兾HeN mice that contained a MMTV provirus that expresses the C3H exogenous envelope and Sag proteins (10). These mice are termed HYB PRO⫹ and the nontransgenic control littermates are HYB PRO⫺. Transgenic mice were maintained in the heterozygous state, and, thus, the cross (HYB PRO⫺兾HYB PRO⫺(C3H兾HeN) ⫻ (HYB PRO⫺兾HYB PRO⫹) yielded half transgenic heterozygotes and half nontransgenic homozygotes. Affected and nonaffected littermates served as odor donors. After weaning, these mice were typed for presence兾absence of the transgene by using standard PCR procedures. (iii ) Trained sensor mice. These mice were either C57BL兾6Boy or congenic B6-H-2k adult females. The numbers, age ranges, and gender of odor donors are tabulated in Table 1. Procedures for collecting odor stimuli are described in Table 2. Abbreviations: MMTV, mouse mammary tumor virus; Sag, superantigen. §To

whom reprint requests should be addressed. E-mail: [email protected].

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

www.pnas.org兾cgi兾doi兾10.1073兾pnas.082093099

Table 1. Characteristics and number (N) of mice serving as odor donors N (per panel)*

Donor type†

Group

Training

Odor source

Gender

1 2 3 4 5 6

Infectious MMTV Infectious MMTV Infectious MMTV Infectious MMTV Rederived infectious MMTV Rederived infectious MMTV

Mice Urine Mice Urine Mice Urine

Female Female Male Male Female Female

7

Genetically transmitted MMTV

Urine

Male

Generalization

Age range, weeks‡

⫹

⫺

⫹

⫺

⫹

⫺

16 34 20 34 24 44 ⫹兾⫺ 13

19 38 12 38 24 42 ⫺兾⫺ 4

12 14 12 32 12 12 ⫹兾⫺ 10

12 14 12 24 12 13 ⫺兾⫺ 11

12–32 12–36 12–32 12–32 12–40 20–44 ⫹兾⫺ 16–24

12–32 12–36 12–32 12–32 12–40 20–44 ⫺兾⫺ 20–24

Apparatus and Procedures. Y-maze training and testing procedures.

Trained mice, previously deprived of water, were rewarded with a drop of water for choosing one of two arms of a Y-maze containing either MMTV⫹ or MMTV⫺ odors. Which of the two alternatives to be rewarded for any one trained mouse was arbitrarily determined; experience with MHC distinctions indicates that any mouse can be trained to make a given distinction: it makes little difference which of any two alternative odors presented is chosen for reinforcement. Detailed procedures for training and testing are described elsewhere (11, 12). Brief ly, air is conducted through two odor boxes, containing either animals (groups 1, 3, and 5) or urine samples exposed in Petri dishes (all other groups), to the two arms of the maze. Each odor box has a hinged lid to admit odor sources: either mice temporarily housed in wire mesh containers or a Petri dish containing urine. The air currents then pass

to the left and right arms of the maze, which have hinged transparent lids and are exhausted to the exterior. Each arm of the maze is fitted with a plastic tube perforated at the bottom to make one drop of water available. Each water tube is guarded by a fence that is raised only if the mouse enters the arm scented by the odor concordant with its training. Each arm of the maze is fitted with a gate that is lowered once the mouse has entered. The time interval in the starting compartment is set at 30 s to allow for changing the odor sources in the odor boxes and for replacing the drop of water (if indicated); after this interval, on a timed signal, all three gates are raised to commence the next trial. Left-right placing is decided by a series of random numbers suited to the sample size. The time taken for a trained mouse to make a choice is 2 or 3 s; the choice is made without pause, or after sniffing at the entrance

Table 2. Mice with MMTV can be identified by scent in the absence of tumors Training trials‡ Group 1 2 3 4 5 6 7

Generalization trials§

Odor Source†

C

NC

C

NC

Female mice Females’ urine Male mice Males’ urine Rederived female mice Rederived females’ urine Genetically transmitted virus; male urine

316 191 201 315 205 402 370

96*** 58*** 59*** 84*** 58*** 54*** 58***

40 24 28 44 29 58 54

18** 11* 13* 20** 13* 18*** 26***

Training and testing procedures are described in text and refs. 11 and 12. source mice are described in Table 1. Mice were used either as odor sources themselves (groups 1, 3, and 5) or to provide urine (groups 2, 4, 6, and 7). For mice as odor donors, one mouse was housed in a wire mesh cage that fit comfortably within the odor box. For each trial, different pairs of individual mice, differing in MMTV status, served as odor generators. Urine odor stimuli were obtained from mice by gentle abdominal pressure (males) or in metabolic cages (females). Usually, a single mouse provided enough urine (0.2– 0.3 ml) to cover the bottom of a 3.5-cm-diameter Petri dish, but sometimes two mice were needed. Fresh samples from different donors were used for each run. Generally, on each day of testing, a given combination of two numbered donors was not repeated. Virtually all MMTV⫹ female odor donors, and no MMTV⫺ female odor donors, developed mammary tumors after being bred following completion of these trials. ‡Trained (sensor) mice were either C57BL兾6Boy or congenic B6-H-2k adult females; four sensors were assigned to the distinctions of groups 1, 2, and 3; three sensors to group 4; and five sensors to groups 5, 6, and 7. Approximately half the trained sensors were reinforced in the Y-maze for MMTV⫹ odor, the remainder for MMTV⫺ odor. C, Concordant with training (‘‘correct’’); NC, Non-concordant (‘‘not correct’’). §Generalization trials were conducted with odors of MMTV⫹ and MMTV⫺ mice or urines not previously encountered by the sensors. Sensors were never reinforced for concordant responses in these trials, allowing for blind testing, and thus providing a critical test of discrimination between presence and absence of MMTV. *, P ⬍ 0.05 (binomial test); **, P ⬍ 0.01; ***, P ⬍ 0.001. †Odor

Yamazaki et al.

PNAS 兩 April 16, 2002 兩 vol. 99 兩 no. 8 兩 5613

MEDICAL SCIENCES

*Number of individual mice making up each donor panel. For any given trial, one mouse or urines of one to three mice selected randomly from the donor panel served as the stimulus. ⫹, mice positive for infectious MMTV; ⫺, mice without infectious MMTV; ⫹兾⫺, mice heterozygous for MMTV provirus; ⫺兾⫺, mice with no MMTV provirus alleles. †Infectious MMTV, MMTV transmitted from mother to offspring via virus particles in mouse milk. Rederived infectious MMTV, offspring of previously noninfected female mice fostered onto infected nursing mothers. Genetically transmitted MMTV, mice carrying an endogenous MMTV provirus. ‡Post hoc examination of the data showed no indication of any differences in response as a function of age of donors.

to the arms, or sometimes with brief retracing from one arm to the other. During training, if the choice is discordant (incorrect), the fence is not raised, and the mouse is returned to the starting compartment. If the choice is concordant (correct), the fence is raised to give access to the drop of water. Generalization testing. Critical trials involve generalization without reward. The purpose of this procedure is to test new urine samples without reward and thus obviate the possibility that incidental cues were being learned. Test mice are trained until they achieve concordance scores of greater than 80% in distinguishing the two donor panels. In subsequent training, reward for correct trials is withheld on average every fourth trial. Concordance for unrewarded trials is consistently as high as for rewarded trials. Next, odors from two other generalization panels of donors are presented, on average, every fourth trial. Correct responses on this discrimination are never rewarded. Test mice have never before been trained with urines from any individuals in these generalization panels. When concordance in these generalization tests is greater than chance level, it is concluded that a commonality exists in the factors differentiating the training and generalization pairs of odor stimuli. A marked advantage of this technique is that it lends itself to blind testing of coded samples; the operators of the maze do not know the identity of the generalization odors because they are never required to supply reward. Results Trained mice discriminated female or male mice with and without MMTV viral infection as well as urines collected from these classes of mice (Table 2, groups 1–4). It is likely that only MMTV infection was responsible for this discrimination because we have found that adventitious genetic differences between maternal genomes do not influence odortypes of offspring (e.g., female offspring of the reciprocal crosses AKR ⫻ B6.YAKR and B6.YAKR ⫻ AKR do not differ in odortype; ref. 13). Nevertheless, because these tested mouse strains had been maintained separately for some generations, and may have undergone genetic drift, we selected the MMTV⫺ strain to rederive the two lines. Half these offspring were fostered on the MMTV⫹ strain, producing two groups of genetically identical mice differing only for MMTV infection. These mice, and their urines, were then tested in the Y-maze and were found to be odor disparate by the trained sensor mice (Table 2, groups 5 and 6). To further verify the generality of MMTV-specific odor and to provide insights into possible mechanisms of odor production, we next tested whether mice could be trained to discriminate the urine odors of transgenic mice containing functional intact proviral genomes on a C3H兾HeN genetic background (MMTVHN). These mice transmit MMTV genetically (males and females) or as an infectious, tumorigenic agent through milk (females) (14, 15). To prevent transmission of this infectious virus to offspring, nontransgenic females were bred with heterozygous transgenic HYB PRO males, and their transgenic and nontransgenic offspring were used. Significant discrimination of urines from male heterozygotes vs. nontransgenics indicated that genetically transmitted MMTV also confers a distinctive odor (Table 2, group 7). 1. Penn, D. & Potts, W. K. (1998) Trends Ecol. Evol. 13, 391–396. 2. Kavaliers, M., Colwell, D. D., Ossenkopp, K.-P. & Perrot-Sinal, T. S. (1997) Behav. Ecol. Sociobiol. 40, 373–384. 3. Moncrieff, R. W. (1967) The Chemical Senses (Morgan-Grampian Books, London), 3rd Ed., pp. 15–16. 4. Senol, M. & Fireman, P. (1999) Cutis 63, 107–111. 5. Luther, S. A. & Acha-Orbea, H. (1997) Adv. Immunol. 65, 139–243. 5614 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.082093099

Discussion These studies show that mice can be trained to discriminate female or male mice or their urine odors based on the presence vs. absence of MMTV, acquired either through infection or genetically. Furthermore, odor distinction based on the presence of virus occurs in the absence of overt disease. The mechanism for distinction odor production is not known. The life cycle of infectious MMTV, a B-type retrovirus, is tightly linked to the immune response. After initial ingestion in infected milk, MMTV crosses the intestinal barrier of neonates and invades the lymphoid cells of Peyer’s patches and spreads to all lymphoid organs before arriving at the epithelial cells of the mammary glands, its jumping-off point to the next generation (5). Because there is a Sag encoded in the virus, infection is accompanied by deletions in the T cell repertoire; deletions also occurs in genetically transmitted MMTV. Thus, the odor differences observed between MMTV⫹ and MMTV⫺ mice may be attributable to MMTV-associated perturbations of the immune system (including those related to MHC molecules, which induce strong responses in the Y-maze and other test systems; refs. 16–19), rather than to the virus itself. Whether the odor difference is due to the presence of infectious virus or to Sag-mediated alterations of the immune system can be determined by using C3H兾HeN transgenic mice that contain only the MMTV sag gene. A number of studies (1) have demonstrated that body odors of animals infected with certain parasites (e.g., protozoa and nematodes) and viruses are avoided. Generally, these studies have evaluated odors of animals with acute illness. It would be of interest to determine whether mice harboring latent exogenously transmitted MMTV infection are also avoided. There are indications that endogenous MMTV provides protection against exogenous infection (20). Consequently, a mating preference for mice with genetically based MMTV might be expected. Whether these odors are specific to different types of MMTV (5) or to other viruses, and the extent to which viral and other diseases can be diagnosed before any overt symptoms in mice or other organisms such as humans, should be investigated. There have been several anecdotal reports of dogs’ seeming ability to detect certain skin cancers (21), and there is a recent report supporting this ability.¶ It should be possible to verify this finding in a controlled study such as we have described here. Our current model system is particularly timely because humans may harbor endogenous Sags (22), and several recent studies (23–25) have implicated MMTV-like genes in some human breast cancers. Also, there is a wide variety of other viral diseases (e.g., HIV), for which obvious symptoms are slow to develop, that could be investigated for unique odor production. Finally, our data suggest that it would be profitable to investigate further the potential for diagnosis of viral infection, as well as other disease vectors, by using not only biological sensors as we have done here, but also some of the so-called artificial noses or e-noses (26).

¶Pickel,

D. P., Cognetta, A. B., Manucy, G. P., Walker, D. B., Hall, S. B. & Walker, J. C. (2001) in Association for Chemoreception Sciences XXIII Annual Meeting Abstracts (A Chem S, Sarasota, FL), p. 25.

This work was supported by National Science Foundation Grants 9728787 and 0112528. 6. Penn, D., Schneider, G., White, K., Slev. P. & Potts, W. (1998) Ethology 104, 685–694. 7. Ross, S. R. (1997) Adv. Pharm. 39, 21–46. 8. Hsu, P.-N., Bryant, P. W., Sutkowski, N., McLellan, B., Ploegh, H. L. & Huber, B. T. (2001) J. Immunol. 166, 2209–3314. 9. Golovkina, T. V., Chervonsky, A., Dudley, J. P. & Ross, S. R. (1992) Cell 69, 637–645.

Yamazaki et al.

18. Singh, P. B., Brown, R. E. & Roser, B. (1987) Nature (London) 327, 161–164. 19. Wedekind, C., Seebeck, T., Bettens, F. & Paepke, A. J. (1995) Proc. R. Soc. London Ser. B 260, 245–249. 20. Golovkina, T. V., Prescott, J. A. & Ross, S. R. (1993) J. Virol. 67, 7690 –7694. 21. Church, J. & Williams, H. (2001) Lancet 358, 930. 22. Posnett, D. N. & Yarilina, A. A. (2001) Immunity 15, 503–506. 23. Wang, Y., Go, V., Holland, J. F., Meland, S. M. & Pogo, B. G-T. (1998) Clin. Cancer Res. 4, 2565–2568. 24. Etkind, P., Du, J., Kahn, A., Pillitteri, J. & Wiernik, P. H. (2000) Clin. Cancer Res. 6, 1273–1278. 25. Liu, B., Wang, Y., Melana, S. M., Pelisson, I., Najfeld, V., Holland, J. F. & Pogo, B. G. (2001) Cancer Res. 61, 1754–1759. 26. Montag, S., Frank, M., Ulmer, H., Wernet, D., Gopel, W. & Rammensee, H.-G. (2001) Proc. Natl. Acad. Sci. USA 98, 9249–9254.

MEDICAL SCIENCES

10. Golovkina, T. V., Chervonsky, A., Prescott, J. A., Janeway, C. A. & Ross, S. R. (1994) J. Exp. Med. 179, 439–446. 11. Yamaguchi, M., Yamazaki, K., Beauchamp, G. K., Bard, J., Boyse, E. A. & Thomas, L. (1981) Proc. Natl. Acad. Sci. USA 78, 5817–5820. 12. Boyse, E. A., Beauchamp, G. K. & Yamazaki, K. (1987) Trends Genet. 3, 97–102. 13. Yamazaki, K., Beauchamp, G. K., Matsuzaki, O., Bard, J., Thomas, L. & Boyse, E. A. (1986) Proc. Natl. Acad. Sci. USA 83, 4438–4440. 14. Golovkina, T. V., Dudley, J. P. & Ross, S. R. (1998) J. Immunol. 161, 2375–2382. 15. Hook, L. M., Agafonova, Y., Ross, S. R., Turner, S. J. & Golovkina, T. V. (2000) J. Virol. 74, 8876–8883. 16. Yamazaki, K., Beauchamp, G. K., Kupniewski, D., Bard, J., Thomas, L. & Boyse, E. A. (1988) Science 240, 1331–1332. 17. Penn, D. & Potts, W. K. (1998) Physiol. Behav. 64, 235–243.

Yamazaki et al.

PNAS 兩 April 16, 2002 兩 vol. 99 兩 no. 8 兩 5615