A Critical Role for Eosinophils in Allergic Airways
Remodeling
Alison A. Humbles, et al.
Science 305, 1776 (2004);
DOI: 10.1126/science.1100283
The following resources related to this article are available online at
www.sciencemag.org (this information is current as of April 23, 2007 ):
Supporting Online Material can be found at:
http://www.sciencemag.org/cgi/content/full/305/5691/1776/DC1
A list of selected additional articles on the Science Web sites related to this article can be
found at:
http://www.sciencemag.org/cgi/content/full/305/5691/1776#related-content
This article cites 26 articles, 21 of which can be accessed for free:
http://www.sciencemag.org/cgi/content/full/305/5691/1776#otherarticles
This article has been cited by 106 article(s) on the ISI Web of Science.
This article has been cited by 52 articles hosted by HighWire Press; see:
http://www.sciencemag.org/cgi/content/full/305/5691/1776#otherarticles
This article appears in the following subject collections:
Medicine, Diseases
http://www.sciencemag.org/cgi/collection/medicine
Information about obtaining reprints of this article or about obtaining permission to reproduce
this article in whole or in part can be found at:
http://www.sciencemag.org/about/permissions.dtl
Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in December, by the
American Association for the Advancement of Science, 1200 New York Avenue NW, Washington, DC 20005. Copyright
c 2004 by the American Association for the Advancement of Science; all rights reserved. The title SCIENCE is a
registered trademark of AAAS.
Downloaded from www.sciencemag.org on April 23, 2007
Updated information and services, including high-resolution figures, can be found in the online
version of this article at:
http://www.sciencemag.org/cgi/content/full/305/5691/1776
are also ambiguous, as they either do not completely eliminate pulmonary eosinophils or they
elicit the loss of eosinophils by mechanisms that
do not differentiate between effects on eosinophils and other potentially important cellular
targets (19–22). However, measurements of lung
function after OVA sensitization/aerosol challenge of PHIL mice (15) showed that methacholine-induced airway hyperresponsiveness was
dependent on the presence of eosinophils (Fig.
4). Moreover, the specific loss of eosinophils
also led to improvement of other pulmonary
function parameters associated with the distal
regions of the lung (fig. S4).
The lack of observable phenotypes in knockout mice deficient for the abundant secondary
granule proteins MBP-1 (18) and EPO (17)
suggests that activities other than degranulation,
including antigen presentation (23), the release
of small molecule mediators of inflammation
[e.g., the synthesis and release of eicosanoid
mediators of inflammation (24)], and immune
regulation of the pulmonary microenvironment
through either modulations of T cell activities
(21) or eosinophil-derived cytokine and/or chemokine expression (25) are likely to be the
relevant effector functions. Eosinophil-derived
cytokine and/or chemokine expression, in particular, is noteworthy as it may account for the
chronic and seemingly self-sustaining character
of allergic pulmonary inflammation, which often
leads to lung remodeling events (26, 27). Significant decreases of Th2 cytokine levels in BAL of
OVA-treated PHIL mice (28) lend support to
this hypothesis and suggest that a prominent
eosinophil effector function in the lung is localized immune regulation.
This study shows that eosinophil activities
are important contributory factors leading to
symptoms that are classically defined as hallmark features of asthma. More importantly,
these data provide validation of earlier studies
that independently concluded that a causative
link exists between eosinophils and allergic pulmonary pathologies (22, 29). The dependency of
allergen-induced pulmonary pathologies on eosinophils suggests that these granulocytes participate at a significant level in underlying inflammatory responses. Regardless of the ultimate
definition of the causative activities mediated by
eosinophils, the challenge of future studies will
be to develop confirmatory clinical studies to
unambiguously define the role(s) and extent of
eosinophil effector functions in asthma patients.
The results of such studies will not only widen
our understanding of the principle causes of
asthma, but are also likely to lead to targeted
therapeutic approaches previously dismissed
and/or overlooked.
References and Notes
1. P. O’Byrne, J. Allergy Clin. Immunol. 102, S85 (1998).
2. W. W. Busse, R. F. Lemanske Jr., N. Engl. J. Med. 344,
350 (2001).
3. H. L. Huber, K. K. Koessler, Arch. Intern. Med. 30, 689
(1922).
4. J. Bousquet et al., N. Engl. J. Med. 323, 1033 (1990).
5. W. W. Busse, W. F. Calhoun, J. D. Sedgwick, Am. Rev.
Respir. Dis. 147, S20 (1993).
6. M. J. Leckie et al., Lancet 356, 2144 (2000).
7. K. A. Larson et al., J. Immunol. 155, 3002 (1995).
8. M. P. Macias et al., J. Leukoc. Biol. 67, 567 (2000).
9. S. A. Cormier et al., Mamm. Genome 12, 352 (2001).
10. K. A. Larson et al., Proc. Natl. Acad. Sci. U.S.A. 93,
12370 (1996).
11. M. A. Horton, K. A. Larson, J. J. Lee, N. A. Lee,
J. Leukoc. Biol. 60, 285 (1996).
12. C. C. Paul et al., J. Leukoc. Biol. 56, 74 (1994).
13. R. D. Palmiter et al., Cell 50, 435 (1987).
14. R. J. Collier, Toxicon 39, 1793 (2001).
15. Materials and methods are available as supporting
material on Science Online.
16. J. J. Lee et al., J. Exp. Med. 185, 2143 (1997).
17. K. L. Denzler et al., J. Immunol. 167, 1672 (2001).
18. K. L. Denzler et al., J. Immunol. 165, 5509 (2000).
19. W. Henderson et al., J. Clin. Invest. 100, 3083 (1997).
20. S. P. Hogan et al., J. Immunol. 161, 1501 (1998).
21. J. Mattes et al., J. Exp. Med. 195, 1433 (2002).
22. J. P. Justice et al., Am. J. Physiol. Lung Cell. Mol.
Physiol. 284, L169 (2003).
23. H. Z. Shi, A. Humbles, C. Gerard, Z. Jin, P. F. Weller,
J. Clin. Invest. 105, 945 (2000).
24. C. G. Irvin, Y. P. Tu, J. R. Sheller, C. D. Funk, Am. J.
Physiol. 272, L1053 (1997).
25. J. P. Justice et al., Am. J. Physiol. LCMP 282, L302
(2002).
26. J. Y. Cho et al., J. Clin. Invest. 113, 551 (2004).
27. P. Flood-Page et al., J. Clin. Invest. 112, 1029 (2003).
28. S. I. Ochkur, G. Cieslewicz, J. J. Lee, N. A. Lee, in
preparation.
29. H. H. Shen et al., J. Immunol. 170, 3296 (2003).
30. We wish to acknowledge the kind gift of the gene
encoding DTA by I. Maxwell. We also thank the Mayo
Clinic Arizona Core Facilities (L. Barbarisi, T. BrehmGibson, S. Savarirayan, M. Ruona, B. Schimeck, J.
Caplette, and J. Protheroe) and our administrative
staff (L. Mardel, J. Ford, and P. McGarry). Supported
by the Mayo Foundation; an American Heart Association grant to J.J.L. (no. 045580Z ); grants from NIH
to J.J.L. (no. HL065228), N.A.L. (no. HL058723), S.J.A.
(nos. AI033043 and AI025230), and C.G.I. (nos.
NCRR-COBRE P20RR15557, P01-HL67004, and HLEB67273); and by NIH postdoctoral fellowships to
MM.P.M. (no. HL10105) and S.A.C. (no. AR08545).
Supporting Online Material
www.sciencemag.org/cgi/content/full/305/5691/1773/
DC1
Materials and Methods
Figs. S1 to S4
References and Notes
22 April 2004; accepted 23 July 2004
A Critical Role for Eosinophils in
Allergic Airways Remodeling
Alison A. Humbles,1*† Clare M. Lloyd,2*† Sarah J. McMillan,2
Daniel S. Friend,3 Georgina Xanthou,2 Erin. E. McKenna,1 Sorina
Ghiran,1 Norma P. Gerard,1 Channing Yu,4 Stuart H. Orkin,5
Craig Gerard1
Fig. 4. In the absence of eosinophils, OVAinduced airway hyperresponsiveness does not
develop. Lung function was assessed as airway
resistance (Rn) in response to aerosolized
methacholine, in saline-treated ( WT/Saline)
and OVA sensitized/OVA aerosol challenged
( WT/OVA) wild-type mice in comparison to
saline-treated (PHIL/Saline) and OVA sensitized/OVA aerosol challenged (PHIL/OVA) PHIL
mice (n ⫽ 5 to 10 animals per group). Asterisks
indicate a significant difference (P ⬍ 0.01) between WT/OVA and either WT/Saline, PHIL/
Saline, or PHIL/OVA mice.
1776
Features of chronic asthma include airway hyperresponsiveness, inflammatory
infiltrates, and structural changes in the airways, termed remodeling. The contribution of eosinophils, cells associated with asthma and allergy, remains to be
established. We show that in mice with a total ablation of the eosinophil lineage,
increases in airway hyperresponsiveness and mucus secretion were similar to those
observed in wild-type mice, but eosinophil-deficient mice were significantly protected from peribronchiolar collagen deposition and increases in airway smooth
muscle. These data suggest that eosinophils contribute substantially to airway
remodeling but are not obligatory for allergen-induced lung dysfunction, and
support an important role for eosinophil-targeted therapies in chronic asthma.
Since its discovery by Paul Erlich in 1879,
there has been a wealth of information documenting the association between eosinophils
and parasitic or allergic diseases (1). The role
of eosinophils in allergic disease remains
controversial. Although T helper cell 2 (TH2)
lymphocytes are thought to drive asthmatic
responses, increasing evidence suggests that
eosinophils are associated with development
of lung dysfunction and subsequent immunopathology (2–4).
Asthma is a chronic disease characterized
by airway hyperresponsiveness (AHR), airway inflammation, and reversible airway ob-
17 SEPTEMBER 2004 VOL 305 SCIENCE www.sciencemag.org
Downloaded from www.sciencemag.org on April 23, 2007
REPORTS
struction. In addition, structural changes in
the airway, termed remodeling, occur as a
result of an imbalance in tissue regeneration
and repair mechanisms (5, 6). Subepithelial
fibrosis is a distinctive feature of airway remodeling and contributes to the thickened
airway walls due to the deposition of collagen
types I, III, and IV, fibronectin, and other
extracellular matrix (ECM) proteins such as
tenascin and laminin (7, 8). Increased airway
smooth muscle (ASM) mass and excessive
mucus secretion from hyperplastic goblet
cells are also features of airway remodeling
(9, 10).
To define the role of eosinophils in asthma pathophysiology, we used the recently
described eosinophil lineage–ablated line,
⌬dbl GATA mice (11). Deletion of a highaffinity GATA site in the GATA-1 promotor
results in a complete ablation of the eosinophil lineage without affecting the development of the other GATA-1– dependent lineages (erythroid, megakaryocytic, and mast
cell) (11).
We examined the extent of this mutation on
eosinophil recruitment following acute and
chronic allergen challenge in a murine model of
allergic airways disease (12, 13). Histological
examination confirmed that sham-treated ⌬dbl
GATA mice were completely devoid of eosinophils and that allergen challenge failed to induce eosinophilia in the airways and bone marrow of ⌬dbl GATA mice (Fig. 1, A to D).
Eosinophil peroxidase (EPO) analysis (13) of
lung tissue (fig S2) and bone marrow (Fig. 1E)
confirmed the absence of eosinophils in these
tissues. During acute and chronic phases, wildtype (WT) mice showed significant increases in
pulmonary eosinophils and lymphocytes. Allergen challenge induced similar numbers of alveolar macrophages and lymphocytes in both WT
and ⌬dbl GATA mice, confirming that the ⌬dbl
GATA mutation was selective for eosinophils
[bronchoalveolar lavage (BAL) cell counts and
lung EPO are shown in figs. S1 and S2)]. Given
the role of GATA-1 in mast cell differentiation
(14), histological analysis of chloroacetate esterase–stained tissue sections demonstrated that
this mutation had no effect on mast cell numbers (15).
Department of Pediatrics, Children’s Hospital, Harvard Medical School, Boston, MA 02115, USA. 2Leukocyte Biology Section, Division of Biomedical Sciences, Faculty of Medicine, Imperial College, London SW7
2AZ, UK. 3Department of Pathology, Brigham and
Women’s Hospital, Harvard Medical School, Boston,
MA 02111, USA. 4Department of Medicine, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114, USA. 5Department of Pediatric Oncology, Dana Farber Cancer Institute and Children’s
Hospital, Harvard Medical School and the Howard
Hughes Medical Institute, Boston, MA 02115, USA.
Airway function was assessed using
whole-body plethysmography (13). Pulmonary conductance (GL) and compliance
(Cdyn) (16), and Penh [a calculated value that
correlates with measurement of airway resistance, obstruction, and intrapleural pressure
in the same mouse (17)] were assessed on day
25 after acute challenge on days 21 to 24. WT
allergen-challenged mice developed a significantly enhanced response to methacholine
(Mch) when compared to WT sham-treated
animals. In the absence of eosinophils, allergen-challenged ⌬dbl GATA mice displayed
enhanced responses to Mch relative to baseline sham controls that were comparable to
those displayed by challenged WT mice (Fig.
2, A and B). Similarly, the Penh responses of
ovalbumin (OA)-challenged ⌬dbl GATA
mice to Mch were almost identical to those of
their WT-challenged counterparts (fig. S3).
During the chronic phase, mice were assessed
for changes in lung function weekly until day
55. Although the enhancement following allergen challenge was lower than that seen
during the acute phase (at day 25), allergenchallenged WT and ⌬dbl GATA mice displayed similar enhanced responses to cholinergic stimulation relative to baseline sham
controls (fig. S4). Thus, eosinophil deficiency conferred no protection against Mch-induced AHR (during acute and chronic allergen challenge), suggesting that eosinophils
are not obligatory for allergen-induced
changes in airway physiology.
Given the role of TH2 cells in the allergic
response, we examined TH2 cytokine expression in the lungs of acute and chronically
challenged WT and ⌬dbl GATA mice. TH2
responses in ⌬dbl GATA mice appeared normal and were similar to those of their WT
littermates. ⌬dbl GATA mice displayed increased BAL and lung interleukin-4 (IL-4),
Fig. 1. Acute allergen challenge induces
eosinophil accumulation in the lung and
bone marrow of WT (A and C) but not
⌬dbl GATA mice (B and D). Original
magnification, ⫻20 ␮m. (E) EPO analysis of bone marrow confirmed that ⌬dbl GATA mice (solid bars) are devoid of eosinophils compared
to WT controls (open bars) before (sham) and after (OA) allergen challenge. Results are means ⫾
SEM (Sham, n ⫽ 4 mice; OA, n ⫽ 5 mice). Significant differences between respective sham-treated
and sensitized/challenged WT or ⌬dbl GATA mice are indicated as *P ⬍ 0.03 and **P ⬍ 0.008.
1
*These authors contributed equally to this work.
†To whom correspondence should be addressed.
E-mail: [email protected], c.lloyd@
imperial.ac.uk
Fig. 2. AHR following acute allergen challenge. Sham-treated (dashed lines) WT (䡬) or ⌬dbl GATA
mice (䢇) mice were exposed to aerosolized saline, and OA-sensitized (solid lines) WT (e) and ⌬dbl
GATA mice (f) were exposed to aerosolized OA on days 21 to 24. About 21 to 24 hours after the
last aerosol challenge, mice were anesthetized, intubated, and mechanically ventilated, and airway
responses to increasing concentrations of intravenous Mch were assessed. The dose-response
curves for (A) pulmonary conductance (GL) and (B) pulmonary compliance (Cdyn) are shown.
Results are means ⫾ SEM (Sham, n ⫽ 4 mice; OA, n ⫽ 8 or 9 mice) of the percentage minimal
decrease in pulmonary conductance or compliance obtained after Mch challenge compared with
the baseline value just before challenge. Significant differences between respective sham-treated
and sensitized/challenged WT or ⌬dbl GATA mice are indicated as *P ⬍ 0.05 to P ⬍ 0.01.
www.sciencemag.org SCIENCE VOL 305 17 SEPTEMBER 2004
1777
Downloaded from www.sciencemag.org on April 23, 2007
REPORTS
IL-5, and IL-13 protein following acute allergen challenge. Likewise, IL-4 and IL-5 expression during the chronic phase were comparable for WT and ⌬dbl GATA mice (fig.
S5). Numbers of TH2 cells, determined by
staining lungs for the TH2 surface marker
T1/ST2, were found to be comparable between WT and ⌬dbl GATA mice (15). Moreover, serum-specific OA–immunoglobulin
E was similar (acute OA WT ⫽ 3713 ⫾ 539
ng/ml versus OA ⌬dbl GATA mice ⫽
4059 ⫾ 789 ng/ml; chronic OA WT ⫽
5204 ⫾ 716 ng/ml versus OA ⌬dbl GATA
mice ⫽ 5635 ⫾ 741 ng/ml; n ⫽ 6 to 9
mice). These data demonstrate that allergen-driven TH2 responses develop in the
absence of eosinophils.
TH2 cytokines (IL-4, IL-5, IL-9, and IL13) and transforming growth factor–␤ (TGF␤) have been shown to induce subepithelial
fibrosis (18–24). Recent reports support a
potential role for eosinophils in the development of airway remodeling (2–4). Increased
eosinophils in the bronchial mucosa of severe
asthmatics have been associated with basement membrane thickening (25), and eosinophils are capable of secreting an array of
profibrotic mediators (22, 23). However,
studies in which IL-5 activity was inhibited,
although associated with a decrease in eosinophil numbers, could theoretically be operating on a number of pathways independent of
the eosinophil (2–4). In light of this ambiguity, we examined the effects of specific eosinophil deficiency on airway remodeling following chronic challenge.
Increased mucus secretion from hyperplastic
goblet cells, shown by periodic acid-Schiff
(PAS)–positive cells in the bronchial epithelium
(13), was similar in WT and ⌬dbl GATA mice
compared to sham controls after acute challenge,
and these increases were sustained throughout
chronic challenge (fig. S6). Thus, enhanced mucus secretion occurs in allergic airways independent of eosinophils.
Increased subepithelial deposition of
ECM proteins, specifically collagen, is a
prominent feature of airway remodeling. We
examined matrix deposition (collagen and fibrin) in lung sections stained with Martius
scarlet blue (MSB) (13, 26). Sham mice
showed a thin uniform layer of matrix in
peribronchiolar subepithelial regions (Fig. 3,
A and B), whereas acute challenge marginally increased fine matrix in both WT and ⌬dbl
GATA mice within some infiltrates (15). Prolonged challenge of WT mice significantly
increased matrix deposition in the subepithelial layer of the bronchioles and perivascular
regions. Dense fibrils were seen in the subepithelial and submucosal areas and in between the inflammatory cells. In marked contrast, matrix deposition in these same regions
was consistently reduced in ⌬dbl GATA
mice when compared with that in WT mice
(Fig. 3, A to D; fig. S7, A to D). Quantitative
image analysis of MSB-stained lung sections
and biochemical measurement of total collagen in lung tissue (13) confirmed that prolonged allergen challenge of WT mice provoked a marked increase (up to threefold) in
matrix deposition, as compared with that seen
in sham mice, and levels were significantly
reduced in challenged ⌬dbl GATA mice (Fig.
Fig. 3. In the absence of allergen challenge, WT (A) and
⌬dbl GATA (B) mice exhibited minimal subepithelial
MSB staining (blue). In contrast, chronic OA challenge
induced a significant increase in MSB staining (C), which
was markedly reduced in challenged ⌬dbl GATA mice (D). Data are representative of 8 to 12 mice
per group; original magnifications, ⫻40. (E) Image analysis of MSB-stained lung sections from sham
or chronically challenged WT (open bars) and ⌬dbl GATA mice (solid bars) confirmed that
challenged ⌬dbl GATA mice were significantly protected from collagen deposition (**P ⬍ 0.0001).
Results are means ⫾ SEM (n ⫽ 8 to 12 mice per group). Significance between Sham WT and OA
WT is indicated (*P ⬍ 0.0001). (F) Lung collagen was measured in sham and chronically challenged
⌬dbl GATA mice. Individual values and means (solid lines) for each group are shown (n ⫽ 10 to 16
mice per group). Significant differences between Sham WT and OA WT, and between OA WT and
⌬dbl GATA mice, are indicated as *P ⬍ 0.001 and **P ⬍ 0.003, respectively.
1778
3, E and F, respectively). These results conclusively demonstrate that eosinophils contribute to allergen-induced subepithelial collagen deposition.
The effects of eosinophil deficiency on
ASM hyperplasia and proliferation were determined by counting the numbers of total and
proliferating cell nuclear antigen (PCNA)–positive smooth muscle cells along the basement
membrane of three or four bronchioles per animal (13). Prolonged allergen challenge of WT
mice induced a significant increase in the total
and proliferating number of ASM cells compared with that seen in sham mice. This increase was absent from airways of chronically
challenged ⌬dbl GATA mice (Fig. 4 and fig.
S8). Prolonged allergen challenge induces phenotypic changes in ASM cells (27), which
could conceivably induce secretion of a number
of growth factors, like TGF-␤, which contribute
to ECM formation. We investigated expression
of active TGF-␤1 in WT and ⌬dbl GATA mice
and consistently found no differences in TGF␤1 expression between chronically challenged
WT and ⌬dbl GATA mice (either protein or
mRNA). These data suggest that reduced subepithelial fibrosis in our model is independent
of TGF-␤1 expression.
Our work contrasts with a recent report
demonstrating that IL-5– deficient mice are
protected from collagen deposition because
of a reduction in TGF-␤–positive eosinophils
(4). We have previously shown that mononuclear cells, presumably macrophages, and not
Fig. 4. Eosinophil deficiency protects against
increases in ASM. (A) Number of ASM cells
(round and elongated) and (B) proliferating
(PCNA-positive) ASM cells along the basement
membrane of three or four bronchioles per
mouse were determined in sham and chronically challenged WT (open bars) and ⌬dbl
GATA mice (solid bars). Results are means ⫾
SEM for each group (Sham, n ⫽ 4 mice; OA, n ⫽
6 mice). Significant difference between OA WT
and ⌬dbl GATA mice is indicated (*P ⬍ 0.004).
17 SEPTEMBER 2004 VOL 305 SCIENCE www.sciencemag.org
Downloaded from www.sciencemag.org on April 23, 2007
REPORTS
REPORTS
References and Notes
1. G. J. Gleich, C. R. Adolphson, K. M. Leiferman, Annu.
Rev. Med. 44, 85 (1993).
2. D. I. Blyth, T. F. Wharton, M. S. Pedrick, T. J. Savage, S.
Sanjar, Am. J. Respir. Cell Mol. Biol. 23, 241 (2000).
3. P. Flood-Page et al., J. Clin. Invest. 112, 1029 (2003).
4. J. Y. Cho et al., J. Clin. Invest. 13, 551 (2004).
5. P. J. Jeffery, A. J. Wardlaw, F. C. Nelson, J. V. Collins,
A. B. Kay, Am. Rev. Respir. Dis. 140, 1745 (1989).
6. J. Bousquet, P. K. Jeffery, W. W. Busse, M. Johnson,
A. M. Vignola, Am. J. Respir. Crit. Care Med. 161,
1720 (2000).
7. J. A. Elias, Z. Zhu, G. Chupp, R. J. Homer, J. Clin. Invest.
104, 1001 (1999).
8. W. Busse, J. Elias, D. Sheppard, S. Banks-Schlegel,
Am. J. Respir. Crit. Care Med. 160, 1035 (1999).
9. C. E. Brewster et al., Am. J. Respir. Cell Mol. Biol. 3,
507 (1990).
10. A. Wanner, Chest 97, 11S (1990).
11. C. Yu et al., J. Exp. Med. 195, 1387 (2002).
12. S. McMillan, C. M. Lloyd, Clin. Exp. Allergy 34, 1
(2004).
13. Materials and methods are available as supporting
material on Science Online.
14. A. R. Migliaccio et al., J. Exp. Med. 197, 281 (2003).
15. A. A. Humbles et al., unpublished observations.
16. T. R. Martin, N. P. Gerard, S. J. Galli, J. M. Drazen,
J. Appl. Physiol. 64, 2318 (1988).
17. E. J. Hamelmann et al., Am. J. Respir. Crit. Care Med.
156, 766 (1997).
18. J. A. Rankin et al., Proc. Natl. Acad. Sci. U.S.A. 93,
7821 (1996).
19. J. J. Lee et al., J. Exp. Med. 185, 2143 (1997).
20. U. A. Temann, G. P. Geba, J. A. Rankin, R. A. Flavell, J.
Exp. Med. 188, 1307 (1998).
21. Z. Zhu et al., J. Clin. Invest. 103, 779 (1999).
22. I. Ohno et al., Am. J. Respir. Cell Mol. Biol. 15, 404
(1996).
23. E. M. Minshall et al., Am. J. Respir. Cell Mol. Biol. 17,
326 (1997).
24. A. M. Vignola et al., Am. J. Respir. Crit. Care Med.
156, 591 (1997).
25. S. E. Wenzel et al., Am. J. Respir. Crit. Care Med. 160,
1001 (1999).
26. A. C. Lendrum, J. Clin. Pathol. 15, 401 (1962).
27. L. M. Moir et al., Am. J. Physiol. Lung Cell. Mol.
Physiol. 284, L148 (2003).
28. H. A. Chapman, J. Clin. Invest. 113, 148 (2004).
29. T. C. Beller et al., Proc. Natl. Acad. Sci. U.S.A. 101,
3047 (2004).
30. C. Bandeira-Melo, P. F. Weller, Prostaglandins Leukot.
Essent. Fatty Acids, 69, 135. (2003).
31. We thank J. Brewer for technical assistance and the
staff of Animal Resources Children’s Hospital for
animal care. This work was supported by the NIH
(grants AI39759 and HL10463). C.M.L. and S.J.M.
were supported by the Wellcome Trust (award
057704) and G.X. by the European Molecular Biology
Foundation.
Supporting Online Material
www.sciencemag.org/cgi/content/full/305/5691/1776/
DC1
Materials and Methods
Figs. S1 to S8
References
13 May 2004; accepted 21 July 2004
Children Creating Core Properties of
Language: Evidence from an
Emerging Sign Language in Nicaragua
Ann Senghas,1* Sotaro Kita,2 Aslı Özyürek3,4,5
A new sign language has been created by deaf Nicaraguans over the past 25
years, providing an opportunity to observe the inception of universal hallmarks
of language. We found that in their initial creation of the language, children
analyzed complex events into basic elements and sequenced these elements
into hierarchically structured expressions according to principles not observed
in gestures accompanying speech in the surrounding language. Successive cohorts of learners extended this procedure, transforming Nicaraguan signing
from its early gestural form into a linguistic system. We propose that this early
segmentation and recombination reflect mechanisms with which children learn,
and thereby perpetuate, language. Thus, children naturally possess learning
abilities capable of giving language its fundamental structure.
Certain properties of language are so central to
the way languages operate, and so widely observed, that Hockett termed them “design features” of language (1). This study asks whether
these properties can arise naturally as a product
of language-learning mechanisms, even when
they are not available in the surrounding language environment. We focus here on two particular properties of language: discreteness and
combinatorial patterning. Every language consists of a finite set of recombinable parts. These
basic elements are perceived categorically, not
continuously, and are organized in a principled,
hierarchical fashion. For example, we have discrete sounds that are combined to form words,
that are combined to form phrases, and then
sentences, and so on. Even those aspects of the
world that are experienced as continuous and
Department of Psychology, Barnard College of Columbia University, 3009 Broadway, New York, NY
10027, USA. 2Department of Experimental Psychology, University of Bristol, 8 Woodland Road, Bristol
BS8 1TN, UK. 3F. C. Donders Center for Cognitive
Neuroimaging, Nijmegen University, Adelbertusplein
1, 6525 EK Nijmegen, Netherlands. 4Max Planck Institute for Psycholinguistics, Wundtlaan 1, 6525 XD
Nijmegen, Netherlands. 5Department of Psychology,
Koç University, Rumeli Feneri Yolu, 34450, Sariyer,
Istanbul, Turkey.
1
*To whom correspondence should be addressed.
E-mail: [email protected]
holistic are represented with language that is
discrete and combinatorial. Together, these properties make it possible to generate an infinite
number of expressions with a finite system. It is
generally agreed that they are universal hallmarks of language, although their origin is the
subject of continued controversy (2–7).
Humans are capable of representations
that lack these properties. For example, nonlinguistic representations such as maps and
paintings derive their structure iconically,
from their referent. That is, patterns in the
representation correspond, part for part, to
patterns in the thing represented. In this way,
half a city map represents half a city. Unlike
language, such nonlinguistic representations
are typically analog and holistic.
The present study documents the emergence of discreteness and combinatorial patterning in a new language. Over the past 25
years, a sign language has arisen within a
community of deaf Nicaraguans who lacked
exposure to a developed language. This situation enables us to discover how fundamental
language properties emerge as the nonlinguistic becomes linguistic.
Before the 1970s, deaf Nicaraguan children
and adults had little contact with each other.
Societal attitudes kept most deaf individuals at
home, and the few schools and clinics available
www.sciencemag.org SCIENCE VOL 305 17 SEPTEMBER 2004
1779
Downloaded from www.sciencemag.org on April 23, 2007
eosinophils are the main secretors of TGF-␤1
protein during chronic challenge (12). The
reason for this disparity is unclear (4), but
variability in protocols may account for the
differences seen in the cell source and expression levels of TGF-␤. A number of other
factors have been demonstrated to be profibrotic in the lung, notably the chemokine
MCP-1, thrombin, endothelin-1, and plasminogen activator inhibitor 1 (28). It is difficult to link the presence of these factors to
eosinophils specifically. However, the cysteinyl leukotrienes have been shown to be
linked to both profibrotic remodeling responses and eosinophils (29, 30). In fact, the
eosinophil may be a major source of leukotrienes, often overlooked.
Of importance is that these animal studies
are in accordance with observations made in
humans. Mild asthmatic patients pretreated
with IL-5–specific antibody exhibited significant reduction in tenascin, lumican, and procollagen III (3). Our results independently
demonstrate that eosinophils are in part responsible for both collagen and smooth muscle changes in a chronic model of asthma.
Further, although the contribution of eosinophils to lung dysfunction has been controversial, we show here that eosinophils are not
obligatory for airway physiology changes associated with this disease. Taken together,
these data provide a rationale for anti-eosinophil– based therapeutics in chronic allergic
airways disease.