The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Serreze, D. V.
Right arrow Articles by Cox, G. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Serreze, D. V.
Right arrow Articles by Cox, G. A.
The Journal of Immunology, 2003, 170: 2742-2749.
Copyright © 2003 by The American Association of Immunologists

Paralytic Autoimmune Myositis Develops in Nonobese Diabetic Mice Made Th1 Cytokine-Deficient by Expression of an IFN-{gamma} Receptor {beta}-Chain Transgene1

David V. Serreze2,*, Melissa A. Pierce*, Cristina M. Post*, Harold D. Chapman*, Holly Savage*, Roderick T. Bronson*, Paul B. Rothman{dagger} and Gregory A. Cox*

* The Jackson Laboratory, Bar Harbor, ME 04609; and {dagger} Department of Microbiology, College of Physicians and Surgeons, Columbia University, New York, NY 10032


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Nonobese diabetic (NOD) mice and some human type 1 diabetes (T1D) patients manifest low to high levels of other autoimmune pathologies. Skewing their cytokine production from a Th1 (primarily IFN-{gamma}) to a Th2 (primarily IL-4 and IL-10) pattern is a widely proposed approach to dampen the pathogenicity of autoreactive diabetogenic T cells. However, it is important that altered cytokine balances not enhance any other autoimmune proclivities to dangerous levels. Murine CD4 T cells are characterized by a reciprocal relationship between the production of IFN-{gamma} and expression of the {beta}-chain component of its receptor (IFN-{gamma}RB). Thus, NOD mice constitutively expressing a CD2 promoter-driven IFN-{gamma}RB transgene in all T cells are Th1-deficient. Unexpectedly, NOD.IFN-{gamma}RB Tg mice were found to develop a lethal early paralytic syndrome induced by a CD8 T cell-dependent autoimmune-mediated myositis. Furthermore, pancreatic insulitis levels were not diminished in 9-wk-old NOD.IFN-{gamma}RB Tg females, and overt T1D developed in the few that survived to an older age. Autoimmune-mediated myositis is only occasionally detected in standard NOD mice. Hence, some manipulations diminishing Th1 responses can bring to the forefront what are normally secondary autoimmune pathologies in NOD mice, while also failing to dependably abrogate pancreatic {beta} cell destruction. This should raise a cautionary note when considering the use of protocols that induce alterations in cytokine balances as a means of blocking progression to overt T1D in at-risk humans.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Due to generalized tolerance induction defects, nonobese diabetic (NOD)3 mice fail to delete a wide array of autoreactive CD4 and CD8 T cells, most notably those that cause type 1 diabetes (T1D) (reviewed in Refs. 1 and 2). However, the numbers and functional activities of other autoimmune effectors in NOD mice, some of which are normally only present at subclinical levels, can be greatly enhanced by certain pharmacological or genetic interventions. These include the development of lupus- and multiple sclerosis-like syndromes following respective treatment of NOD mice with bacillus Calmette-Guérin vaccine or pertussis toxin, and a peripheral neuritis upon genetic ablation of the T cell costimulatory molecule B7-2 (3, 4, 5). Many humans with, or at risk for, T1D also have other underlying autoimmune susceptibilities (6). Hence, it is clearly a concern that any interventions designed to block the original development, or the subsequent functional activation of diabetogenic T cells, not enhance any other autoimmune proclivities to dangerous levels.

Although ultimately shown to be overly simplistic (7, 8, 9, 10, 11, 12), a previously popular hypothesis was that the pathogenicity of {beta} cell autoreactive CD4 T cells could be dampened by manipulations which skewed the types of cytokines they produced from a Th1 (primarily IFN-{gamma}) to a Th2 (primarily IL-4 and IL-10) profile (reviewed in Refs. 13 and 14). However, at the time this hypothesis had not yet been vigorously tested, we initiated several different approaches to produce NOD mice with an impaired ability to mount Th1 responses. One approach was based on the finding that in mice, CD4 T cells are characterized by a strict reciprocal relationship between the production of IFN-{gamma} and expression of the receptor for this cytokine (15). As a result, CD4 T cells expressing a functional IFN-{gamma} receptor cannot produce IFN-{gamma}. The expression of a functional IFN-{gamma} receptor is regulated by the induction of the {beta}-chain signaling subunit that pairs with a constitutively expressed {alpha}-chain subunit (15). Because of this, mice can be rendered deficient in Th1 cytokine production when a CD2 promoter is used to induce constitutive expression in all T cells of an IFN-{gamma} receptor {beta}-chain transgene (abbreviated here as IFN-{gamma}RB Tg) (16).

During the time we were generating a stock of NOD mice expressing a IFN-{gamma}RB Tg (designated NOD.IFN-{gamma}RB Tg), significant evidence accumulated indicating the pathogenicity of {beta} cell autoreactive CD4 T cells cannot be strictly compartmentalized on the basis of currently defined Th1 and Th2 cytokine production profiles (7, 8, 9, 10, 11, 12). Thus, while NOD.IFN-{gamma}RB Tg mice were indeed found to be Th1-deficient, we were not surprised to find they continued to exhibit robust pancreatic {beta} cell autoimmunity. However, unexpectedly, NOD.IFN-{gamma}RB Tg mice were also found to develop an early lethal paralysis. The current report provides an initial characterization of the NOD.IFN-{gamma}RB Tg stock indicating that their paralysis results from a CD8 T cell-dependent autoimmune-mediated myositis (AMM) syndrome. Hence, in addition to the deleterious interventions described earlier, great caution should be also used when considering Th1 inhibitory protocols to prevent T1D in susceptible humans, as this too can sometimes enhance other autoimmune proclivities to dangerous levels.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Mice

NOD/Lt mice are maintained in a specific pathogen-free research colony at The Jackson Laboratory (Bar Harbor, ME). T and B lymphocyte-deficient NOD-scid mice (formal designation NOD-Prkdcscid) have been previously described (17) and are maintained at the N11 backcross generation. The previously described (16) TG2 line of C57BL/6 (B6) mice carrying seven copies of a CD2 promoter driven IFN-{gamma}RB Tg (formal designation TgN(CD2-Ifngr2)) was imported to The Jackson Laboratory. Characteristics of IFN-{gamma}RB Tg expression are also described in this previous study. The transgenic insertion was then congenically transferred to the NOD genetic background. All IFN-{gamma}RB Tg mice used in the studies were derived from a single N8 backcross generation progenitor that was found by our previously published methods (18, 19) to be fixed to homozygosity for genetic linkage markers delineating all known diabetes susceptibility (Idd) loci of NOD origin. During the course of establishing this congenic stock, backcross segregants carrying the IFN-{gamma}RB Tg were identified by screening genomic DNA prepared from PBL by PCR with the primer pair 5'-GCACGTGGTTAAGCTCTCG-3' and 5'-TGTCTCTGTGATGTCCGTACA-3'.

Assessment of diabetes and insulitis development

T1D development in the indicated mice was assessed by weekly monitoring of glycosuric values with Ames Diastix (kindly supplied by Bayer, Diagnostics Division, Elkhart, IN). Values of >= 3 were considered indicative of T1D onset. Pancreases from mice assessed for insulitis development were fixed in Bouin’s solution, and sectioned at three nonoverlapping levels. Granulated {beta} cells were stained with aldehyde fuchsin and leukocytes with a H&E counterstain. Islets (at least 20 per mouse) were individually scored as follows: 0, no lesions; 1, peri-insular leukocytic aggregates, usually periductal infiltrates; 2, <25% islet destruction; 3, >25% islet destruction; 4, complete islet destruction. An insulitis score for each mouse was obtained by dividing the total score for each pancreas by the number of islets examined.

Histological assessment of myositis development

Dissected paraspinal segments and rear legs from the indicated mice were fixed overnight in Bouin’s solution, and then treated for an additional 24 h with Cal-EX decalcifying solution (Fisher Diagnostics, Fair Lawn, NJ). Paraspinal segments were cross-sectioned and rear legs were sagittally sectioned at 5-µM intervals and then stained with H&E for light microscopic analyses. For immunocytochemical analyses of infiltrating leukocyte populations, tissues were snap-frozen in OCT, and 5-µM cryostat sections stained with biotinylated Abs specific for either CD4 T cells (GK1.5), CD8 T cells (53-6.72), B cells (B220-specific Ab RA3-6B2), macrophages (F4/80-specific Ab MCA497), or granulocytes (Gr-1-specific Ab RB68C5). Ab staining cells were developed with the chromogen diaminobenzidine conjugated with streptavidin (Vector Laboratories, Burlingame, CA). Controls consisted of sections stained with the secondary conjugate only.

Flow cytometric analysis of leukocyte subsets

Proportions of various splenic leukocyte subsets in the indicated mice were determined by multicolor flow cytometric techniques (FACScan; BD Biosciences, San Jose, CA). The mAb 145-2C11 specific for the CD3 component of the TCR and conjugated to a green fluorescent FITC tag was used to detect T cells. These T cells were then further characterized for CD4 expression using the mAb GK1.5 conjugated to the red fluorescent tag Cy3.18-Osu (Cy3; Biological Detection Systems, Pittsburgh, PA), or for CD8 expression with the mAb 53-6.72 conjugated to a red fluorescent PE tag. B lymphocytes were detected with a FITC-conjugated goat polyclonal antiserum specific for mouse Ig (Southern Biotechnology Associates, Birmingham, AL). A FITC-conjugated Gr-1 specific mAb (RB68C5) was used in combination with a PE-conjugated Mac-1-specific mAb (M1/70) to delineate macrophages/dendritic cells (DC) and granulocytes. Macrophages/DC stain with the Mac-1 Ab only, while granulocytes costain with both the Mac-1 and Gr-1 Abs.

Cytokine production assays

Splenic leukocytes from the indicated mice were suspended at a concentration of 5 x 106/ml in the previously described tissue culture medium (20) containing 10.0 µg/ml of the mAb 145-2C11 (BD PharMingen, San Diego, CA) capable of activating T cells by binding the CD3 component of the TCR. To drive the anti-CD3-stimulated CD4 T cells into a Th1 or Th2 mode, the medium in some cultures was further supplemented with either 50 U/ml rat recombinant IFN-{gamma} (kindly supplied by P. van de Meide, TNO Primate Center, Rijswijk, Netherlands) combined with 5.0 µg/ml of the murine IL-4 neutralizing mAb 11B11 (BD PharMingen), or 500 U/ml murine recombinant IL-4 (BioSource, Camarillo, CA) combined with 10.0 µg/ml of the murine IFN-{gamma} neutralizing mAb XMG1.2 (BD PharMingen). The cells were incubated for three days at 37°C, harvested, washed, and assessed for proportions of viable T cells by the flow cytometric techniques described above. Viable T cells were then reseeded at a concentration of 1 x 106/ml into tissue culture dishes that had been precoated as previously described (21) with 3.125 µg/ml anti-CD3. After 24 h of secondary stimulation, culture supernatants were harvested and subsequently assessed with commercially available ELISA kits (BD PharMingen) for levels of IFN-{gamma}, IL-4, and TNF-{alpha}. The cytokine concentrations were then normalized to the number of viable T cells in each of the secondary cultures.

Splenocyte adoptive transfer into NOD-scid recipients

Aliquots of 107 splenic leukocytes from paralytic NOD.IFN-{gamma}RB Tg donors were injected i.v. into 4- to 6-wk-old NOD-scid female recipients. The recipients were then monitored for the development of hind limb paralysis over a 20-wk follow-up period. Upon the development of hind limb paralysis, or at the end of the 20-wk observation period, splenic leukocytes from the recipients were typed by flow cytometry as described above to assess the extent of repopulation by various types of donor-derived cells. The extent of leukocytic infiltration in paraspinal regions from some NOD-scid recipients that developed paralysis following repopulation with NOD.IFN-{gamma}RB Tg donors was also assessed histologically as described above.

T cell depletion studies

Starting at 4 wk of age, the indicated numbers of female NOD.IFN-{gamma}RB Tg mice received biweekly injections of mAbs capable of depleting CD4 T cells (300-µg doses of GK1.5) or CD8 T cells (200-µg doses of YTS169). Controls received 250-µg doses of rat IgG (Sigma-Aldrich, St. Louis, MO). All mice were monitored for subsequent AMM development. At the onset of disease, or upon reaching 20 wk of age, splenic leukocytes from mice treated with the various Abs were typed by flow cytometry as described above to assess the extent of CD4 or CD8 T cell depletion.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
NOD mice expressing an IFN-{gamma}RB Tg develop a lethal paralytic myositis in addition to T1D

B6 mice carrying a CD2 promoter-driven IFN-{gamma}RB Tg have been previously described (16). Compared with T cells from standard B6 mice, anti-CD3 stimulated T cells from the IFN-{gamma}RB Tg B6 stock (TG2 line) were characterized by a virtually complete suppression of Th1 responses as assessed by IFN-{gamma} secretion. No unusual pathologies were reported in B6 IFN-{gamma}RB Tg mice. Hence, we were initially curious whether congenic transfer of theIFN-{gamma}RB Tg to the NOD genetic background would result in the same severe impairment of Th1 responses as observed in B6 mice, and if so, what effect this had on T1D development. This NOD congenic stock was first shown by our previously described techniques (18, 19) to be fixed to homozygosity for genetic linkage markers delineating all known Idd loci of NOD origin. It had been previously reported that on the B6 background, the IFN-{gamma}RB Tg induced no alterations in the relative proportions of any leukocyte subsets (16). However, as shown in Table I, following congenic transfer to the NOD genetic background, the IFN-{gamma}RB Tg-positive segregants were respectively characterized by ~3- and 2-fold reductions in the proportion of splenic CD4 and CD8 T cells compared with transgene negative segregants or standard NOD mice. Conversely, the proportions of splenic B cells, macrophages/DCs, and granulocytes were significantly higher in the NOD IFN-{gamma}RB Tg-positive segregants than in either set of control mice. Similar results were observed when splenic leukocyte subsets were evaluated on a total cell basis.


View this table:
[in this window]
[in a new window]
  		Table I. Expression of the CD2 promoter-driven IFN-{gamma}RB Tg alters the relative proportion of various leukocyte subsets in NOD micea

 
Given their reduced T cell numbers, it might be anticipated that NOD IFN-{gamma}RB Tg mice would develop T1D and/or other autoimmune pathologies less frequently or more slowly than nontransgenic controls. However, it was unexpectedly difficult to determine whether the rate of T1D development was altered in NOD.IFN-{gamma}RB Tg females, because almost all of these mice developed a lethal paralytic syndrome before 10 wk of age. Histological analyses revealed that paralysis in NOD.IFN-{gamma}RB Tg mice is associated with development of a massive leukocytic inflammatory response that initiates exterior to the lumbar region of the spinal column (Fig. 1A). This inflammatory lesion then progresses toward the cervical region of the spinal column, as well as into the peripheral limbs (Fig. 1B). The inflammatory response appears to mediate the destruction of muscle, and not peripheral nerve tissue (Fig. 1B), and thus is classified as a myositis syndrome. Such myositis lesions are not observed in age-matched standard NOD mice (Fig. 1C). Early onset of severe myositis did not prevent the codevelopment of an extensive and destructive pancreatic insulitis in NOD.IFN-{gamma}RB Tg mice (Fig. 1D). This was further illustrated by the finding that there was no significant difference in the mean insulitis scores of 9-wk-old female NOD.IFN-{gamma}RB Tg (2.15 ± 0.77, n = 4) and standard NOD mice (2.14 ± 0.51, n = 4). The fact that pancreatic {beta} cell autoimmunity was not attenuated in NOD.IFN-{gamma}RB Tg mice was also demonstrated by the fact that three females which survived to between 8–12 wk of age developed overt T1D in addition to the paralytic myositis syndrome.



View larger version (159K):
[in this window]
[in a new window]
  		FIGURE 1. Inhibition of Th1 responses through IFN-{gamma}RB Tg expression promotes development of a paralytic myositis, and does not inhibit {beta} cell autoimmunity in NOD mice. A, Presence of an extensive leukocytic infiltrate, and muscular destruction, exterior to the lumbar region of the spinal column in a paralytic 8-wk-old NOD.IFN-{gamma}RB Tg female mouse. Arrow depicts the outer bone plate of the vertebral column. B, Extension of the leukocytic infiltrate, and muscular destruction, into the right rear leg of the same paralytic NOD.IFN-{gamma}RB Tg mouse. Arrows denote peripheral nerve bundles containing relatively few infiltrating leukocytes, compared with the high levels found in muscle. C, Lack of leukocytic infiltration and muscular destruction in the lumbar paraspinal region of standard 8-wk-old NOD female mice. D, The continued development of extensive, destructive insulitis in longer surviving NOD.IFN-{gamma}RB Tg mice, indicating no attenuation of {beta} cell autoimmunity.

 
It is important to note that the myositis syndrome was absent in the original B6 donor stock of the IFN-{gamma}RB Tg (data not shown). Hence, the genomic insertion site of the IFN-{gamma}RB Tg or its level of expression does not independently induce a paralytic myositis syndrome, but rather can only do so through interactions with other genetic variants present in NOD, and not B6, mice. Furthermore, in addition to the sialitis lesion that also develops in standard NOD mice, it was found that oophoritis was occasionally present in NOD.IFN-{gamma}RB Tg females. However, there was no obvious leukocytic infiltration of brain, liver, kidney, lung, or the large intestines in NOD.IFN-{gamma}RB Tg mice.

T cell secretion of IFN-{gamma} and TNF-{alpha} is strongly suppressed in NOD.IFN-{gamma}RB Tg mice

As noted above, the highly expressed TG2 variant of the IFN-{gamma}RB Tg virtually eliminates Th1 responses in B6 mice. Hence, we tested the extent to which this IFN-{gamma}RB Tg could suppress Th1 responses in the NOD strain. To do this, we first compared the ability of T cells in the spleens of standard NOD and NOD.IFN-{gamma}RB Tg mice to secrete Th1 or Th2 cytokines when restimulated with anti-CD3 following a primary activation under culture conditions designed to initiate either type of response. As noted in Table I, there are 2- to 3-fold less T cells in the spleens of NOD.IFN-{gamma}RB Tg than standard NOD mice. Hence, the cytokine production data was normalized to the number of viable CD4 T cells placed into each of the secondary cultures.

Following an initial activation under Th1 priming conditions, CD4 T cells from the NOD.IFN-{gamma}RB Tg stock produced 10-fold less IFN-{gamma} than those from standard NOD mice upon secondary anti-CD3 stimulation (Fig. 2A, left panel). Similarly, under these culture conditions, less TNF-{alpha}, another indicator of a Th1 response, was secreted by NOD.IFN-{gamma}RB Tg than standard NOD T cells (Fig. 2A, right panel). On the nonautoimmune prone B6 background, expression of the IFN-{gamma}RB Tg did not alter the induction of Th2 responses under the proper priming conditions (16). In contrast, following initial activation under Th2 priming conditions, 10-fold lower levels of IL-4 were secreted by anti-CD3 restimulated CD4 T cells from NOD.IFN-{gamma}RB Tg than standard NOD mice (Fig. 2B).



View larger version (18K):
[in this window]
[in a new window]
  		FIGURE 2. Expression in NOD mice of a CD2 promoter-drivenIFN-{gamma}RB Tg suppresses Th1 and Th2 cytokine secretion under conditions that normally promote such responses. Splenic T cells underwent primary anti-CD3-mediated stimulation under the conditions described in Materials and Methods that were designed to promote a Th1 or Th2 response. Viable T cells recovered from the primary culture were enumerated by flow cytometry, and then restimulated for 24 h with plate-bound anti-CD3 alone. The secondary culture supernatants from cells initially activated under Th1-inducing conditions were analyzed by ELISA for levels of (A) IFN-{gamma} (left panel) and TNF-{alpha} (right panel). Secondary culture supernatants from cells initially activated under Th2-inducing conditions were analyzed for (B) IL-4 levels. Data represent the mean cytokine concentration ± SD normalized per 105 viable T cells in two splenocyte pools from each strain. Each splenocyte pool was prepared from two mice. Similar results were obtained in a second experiment.

 
It is unlikely that the decreased IL-4 secretory capacity of NOD.IFN-{gamma}RB Tg mice contributes to their myositis susceptibility through the loss of some immunoregulatory mechanism because a similar syndrome is absent in our previously described NOD.IL-4null mice (12). This conclusion was also supported by determining the natural Th1 or Th2 proclivity of NOD or NOD.IFN-{gamma}RB Tg CD4 T cells by subjecting them to primary and secondary stimulation with anti-CD3 alone. Similar to what was observed under the Th1 biasing conditions, there were significantly lower levels of IFN-{gamma} (~20 fold) and TNF-{alpha} (~10 fold) produced by NOD.IFN-{gamma}RB Tg than standard NOD CD4 T cells stimulated with anti-CD3 alone (Fig. 3, A and C). However, when stimulated with anti-CD3 alone, neither NOD.IFN-{gamma}RB Tg nor standard NOD CD4 T cells produced any IL-4 (Fig. 3B). Hence, while expression of the IFN-{gamma}RB Tg can suppress Th2 cytokine production by NOD CD4 T cells under artificial conditions that promote such responses, it would appear this sort of reactivity is normally absent. Collectively, these results indicate that the NOD.IFN-{gamma}RB Tg stock has been rendered Th1-deficient, while also retaining the NOD strain characteristic of poor Th2 responsiveness.



View larger version (10K):
[in this window]
[in a new window]
  		FIGURE 3. The Th1-deficient NOD.IFN-{gamma}RB Tg stock also fails to generate Th2 responses under unbiased activation conditions. Splenic T cells underwent primary and secondary stimulation with anti-CD3 alone. Viable T cells were enumerated by flow cytometry before secondary stimulation. The secondary culture supernatants were analyzed by ELISA for levels of (A) IFN-{gamma}, (B) IL-4, and (C) TNF-{alpha}. Data represent the mean cytokine concentration ± SD normalized per 105 viable T cells in two splenocyte pools from each strain. Each splenocyte pool was prepared from two mice. Similar results were obtained in a second experiment.

 
Myositis develops in virtually all NOD.IFN-{gamma}RB Tg mice, but with faster kinetics in females than males

Development of the paralytic myositis syndrome in NOD.IFN-{gamma}RB Tg mice is associated with an obvious loss in body weight. Hence, we determined whether the first sign of a loss in body weight could be used as a marker for the onset of myositis in NOD.IFN-{gamma}RB Tg mice, and thus aid in determining the kinetics of disease onset. A total of 10 NOD.IFN-{gamma}RB Tg female mice all exhibited an initial loss in body weight between 6 and 10 wk of age, and in each case were histologically characterized by the presence of paraspinal myositis at this time (data for three representative females depicted in Fig. 4). Similarly, eight of nine NOD.IFN-{gamma}RB Tg male mice exhibited a loss in bodyweight associated with paraspinal myositis development between 8–19 wk of age. Therefore, as summarized in Fig. 5, paraspinal myositis develops in virtually all NOD.IFN-{gamma}RB Tg mice, but with faster kinetics in females than males.



View larger version (120K):
[in this window]
[in a new window]
  		FIGURE 4. An initial loss in body weight correlates with the histological appearance of paraspinal myositis in NOD.IFN-{gamma}RB Tg mice. Body weight was measured weekly in 10 female and 9 male NOD.IFN-{gamma}RB Tg mice. Mice were killed when a loss in body weight was first observed and assessed histologically for the presence of paraspinal myositis. Data depict the presence of paraspinal myositis in three representative NOD.IFN-{gamma}RB Tg females showing an initial loss in body weight at 6, 8, and 10 wk of age. Paraspinal myositis was present in all mice at the time a loss in body weight was first observed.

 


View larger version (12K):
[in this window]
[in a new window]
  		FIGURE 5. Paraspinal myositis develops in virtually all NOD.IFN-{gamma}RB Tg mice, but with faster kinetics in females than males. Data depict the accumulative proportion of paraspinal myositis development as a function of age in the same female ({square}, n = 10) and male ({blacksquare}, n = 9) NOD.IFN-{gamma}RB Tg mice described in the Fig. 4 legend.

 
Myositis in NOD.IFN-{gamma}RB Tg mice is associated with infiltration of CD8 T cells and macrophages

Immunocytochemical analyses were performed to assess what leukocyte subpopulations were present in the myositis lesion of NOD.IFN-{gamma}RB Tg mice. There was minimal background staining of paraspinal tissue treated with the secondary conjugate only (Fig. 6A). In contrast, primary staining with specific mAbs revealed significant numbers of CD8 T cells (Fig. 6B), and very few CD4 T cells (Fig. 6C) among the leukocytes comprising the myositis lesion in NOD.IFN-{gamma}RB Tg mice. However, macrophages also comprised a prevalent cell type in the myositis lesion (Fig. 6D). Neither B lymphocytes or granulocytes were detected within these lesions (data not shown).



View larger version (135K):
[in this window]
[in a new window]
  		FIGURE 6. Paraspinal myositis in NOD.IFN-{gamma}RB Tg mice is associated with infiltration of CD8 T cells and macrophages. Leukocyte subpopulations comprising the paraspinal myositis lesion of an overtly paralytic 16-wk-old NOD.IFN-{gamma}RB Tg male were assessed by immunocytochemistry as described in Materials and Methods. Data depict the myositis lesion stained with (A) the secondary streptavidin conjugate only, (B) CD8-specific mAb, (C) CD4-specific mAb, (D) macrophage-specific mAb. No staining was observed with B lymphocyte- or granulocyte-specific Abs.

 
Splenocytes from NOD.IFN-{gamma}RB Tg mice adoptively transfer paraspinal myositis and paralysis to NOD-scid recipients

The fact that the paraspinal myositis lesion of paralytic NOD.IFN-{gamma}RB Tg mice contains CD8 and a fewer number of CD4 T cells, which could both only be activated in an Ag-dependent fashion, suggested this syndrome may have an autoimmune basis. However, macrophages that provide a non-Ag-specific phagocytic function, also comprised a highly prevalent cell type in the myositis lesion. This latter finding suggested the myositis lesion could also represent a leukocytic infiltration that is a secondary phagocytic inflammatory consequence of muscular degeneration originally induced through nonimmunological mechanisms. To distinguish between these two possibilities, we determined whether immune cells (splenocytes) from NOD.IFN-{gamma}RB Tg mice could transfer paraspinal myositis and paralysis to T and B lymphocyte-deficient NOD-scid recipients. As shown in Fig. 7A, splenocytes from paralytic NOD.IFN-{gamma}RB Tg donors successfully transferred this syndrome within 18 wk postreconstitution to all (14 of 14) NOD-scid recipients. FACS analysis confirmed repopulation by donor type CD4 and CD8 T cells in all NOD-scid recipients at the time of disease onset (data not shown). Furthermore, at the time of disease onset the NOD-scid recipients were all characterized by paraspinal myositis lesions similar to that observed in the original paralytic NOD.IFN-{gamma}RB Tg donors (Fig. 7B). These results conclusively demonstrate that the paralysis which develops in Th1-deficient NOD.IFN-{gamma}RB Tg mice does not result from an inherent muscular disorder, but rather from AMM in the paraspinal region.



View larger version (87K):
[in this window]
[in a new window]
  		FIGURE 7. Splenocytes from NOD.IFN-{gamma}RB Tg mice adoptively transfer paraspinal myositis and paralysis to NOD-scid recipients. Aliquots of 107 splenic leukocytes from paralytic NOD.IFN-{gamma}RB Tg donors were injected i.v. into 4- to 6-wk-old female NOD-scid recipients. A, Data represent the cumulative proportion of NOD-scid recipients developing paralysis as a function of time postrepopulation with NOD.IFN-{gamma}RB Tg splenocytes. FACS analysis confirmed repopulation by donor type CD4 and CD8 T cells in all NOD-scid recipients at the time of disease onset. B, Presence of extensive myositis in the paraspinal region of a NOD-scid mouse that became paralytic following repopulation with NOD.IFN-{gamma}RB Tg splenocytes.

 
AMM development in NOD.IFN-{gamma}RB Tg mice is CD8 T cell-dependent

We next analyzed whether CD4 and/or CD8 T cells were essential for AMM development in NOD.IFN-{gamma}RB Tg mice. To do this, it was determined whether AMM development was impaired in NOD.IFN-{gamma}RB Tg female mice injected biweekly from 4 wk of age with mAbs capable of deleting CD4 or CD8 T cells. As shown in Fig. 8A, AMM developed in all four NOD.IFN-{gamma}RB Tg females injected with the control Ab (rat IgG). Similarly, while slightly delayed compared with the control group, all five NOD.IFN-{gamma}RB Tg females treated with the CD4-specific Ab also developed AMM. In contrast, AMM failed to develop in any of the four NOD.IFN-{gamma}RB Tg females treated with the CD8-specific Ab. As shown in Fig. 8B, flow cytometric analyses of splenocytes from anti-CD4- or CD8-treated NOD.IFN-{gamma}RB Tg mice documented specific deletion of the appropriate T cell subset. Although not ruling out a lesser role for the CD4 subset, these data demonstrate that CD8 T cells are essential to the development of AMM in NOD.IFN-{gamma}RB Tg mice.



View larger version (27K):
[in this window]
[in a new window]
  		FIGURE 8. CD8 T cells are essential to AMM development in NOD.IFN-{gamma}RB Tg mice. A, Female NOD.IFN-{gamma}RB Tg mice were injected biweekly from 4 wk of age with either anti-CD4, anti-CD8, or a rat IgG control Ab and were simultaneously monitored for AMM development. Compared with the control group, AMM development was only significantly suppressed in the anti-CD8-treated group (p < 0.01, Kaplan Meier lifetable analysis). B, At the onset of disease, or upon reaching 20 wk of age, splenic leukocytes from mice treated with the various Abs were typed by flow cytometry to assess the extent of CD4 or CD8 T cell depletion.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Our current findings indicate that a stock of NOD mice in which Th1 cytokine responses are virtually eliminated by a CD2 promoter driven IFN-{gamma}RB Tg develop an early lethal paralysis which results from an AMM pathology. Furthermore, the few (three) NOD.IFN-{gamma}RB Tg females that did not previously succumb to AMM all developed T1D. These results add to the growing list of evidence (11, 22, 23) that Th1 responses are not essential to the development of T cell-mediated autoimmune T1D in NOD mice. More importantly, our current results indicate that in addition to their inability to inhibit T1D, some protocols which suppress Th1 responses in NOD mice can actually enhance other, normally less pronounced, autoimmune pathologies present in this strain. Indeed, the sort of myositis that is fulminant in the NOD.IFN-{gamma}RB Tg stock, can occasionally be detected at much lower levels in standard NOD mice (2). Many humans with, or at risk for, T1D also have other underlying autoimmune susceptibilities (6). Hence, our current results should raise a cautionary note about the possible use of Th1 inhibitory protocols to block T1D development in high risk humans, because this could conceivably exacerbate some other underlying autoimmune proclivity. Such concerns should add to those raised by the previous observation that inducing a Th1 to Th2 cytokine shift among effectors contributing to experimental autoimmune encephalomyelitis can trigger lethal anaphylactic shock (24).

The ability of their splenocytes to adoptively transfer the syndrome to NOD-scid recipients conclusively demonstrated the paralysis which develops in Th1-deficient NOD.IFN-{gamma}RB Tg mice does not result from an inherent muscular defect, but rather from an AMM disorder. Immunocytochemical analyses indicated the most prevalent type of Ag-specific immunological effector in the myositis lesion of NOD.IFN-{gamma}RB Tg mice was CD8 T cells, with CD4 T cells present at much smaller numbers. No B lymphocytes were detected. Furthermore, the depletion of CD8, but not CD4 T cells, abrogated AMM development in NOD.IFN-{gamma}RB Tg mice. Thus, it seems likely that the AMM disorder in NOD.IFN-{gamma}RB Tg mice is largely driven by CD8 T cells which recognize MHC class I bound peptides derived from muscle proteins. Because muscle cells express MHC class I molecules (25), they may be destroyed in NOD.IFN-{gamma}RB Tg mice by their direct presentation of autoantigenic peptides to CD8 T cells. Interestingly, polymyositis in humans also appears to be largely mediated by CD8 T cells (26, 27), indicating NOD.IFN-{gamma}RB Tg mice may represent a new and accurate model of this disease. Although originally thought to not play a primary role, it is becoming increasingly clear that MHC class I-dependent CD8 T cell responses are also essential to the initiation of autoimmune pancreatic {beta} cell destruction leading to T1D in standard NOD mice (28, 29). Hence, a generalized NOD strain-specific tolerance induction defect(s) may underlie the development of autoreactive CD8 T cells contributing to either T1D or AMM, with the latter process being enhanced through some unknown mechanism when a Th1 deficiency is induced by the IFN-{gamma}RB Tg.

Macrophages, which function in a non-Ag-specific fashion, represent another prevalent cell type within the paraspinal myositis lesion of NOD.IFN-{gamma}RB Tg mice. One possibility is that these macrophages are merely present to phagocytize cellular debris produced through the autoimmune destruction of muscle by CD8 T cells. Conversely, macrophages may directly contribute to muscular destruction in the NOD.IFN-{gamma}RB Tg stock through the secretion of various cytokines, similar to the way they have been proposed to contribute to pancreatic {beta} cell destruction in standard NOD mice (30, 31).

Although fully differentiated T cells from adult NOD.IFN-{gamma}RB Tg mice are deficient in IFN-{gamma} production, we cannot rule out an earlier role for this Th1 cytokine in the development of AMM. This caveat is partially based on the finding that naive Th0 precursors of both Th1 and Th2 cells naturally express the IFN-{gamma}RB component, and its genetic ablation also impairs differentiation into the Th1 lineage, but through a different mechanism than that induced by the IFN-{gamma}RB Tg (15, 32). Hence, it appears the initiation of Th1 differentiation actually requires IFN-{gamma}-induced signals, but this process is arrested if the T cell maintains IFN-{gamma}RB expression as would occur in the NOD.IFN-{gamma}RB Tg stock. It is possible that such T cells which have failed to fully differentiate into IFN-{gamma}-producing Th1 effectors, but yet required early IFN-{gamma} signaling events during their initial development, may be important contributors to the AMM syndrome in NOD.IFN-{gamma}RB Tg mice. The early source of IFN-{gamma} required under this scenario could either be provided in an autocrine fashion by T cells that then subsequently lose their ability to produce this Th1 cytokine due to continuous expression of transgenically encoded IFN-{gamma}RB molecules, or alternatively by another cell type, such as NK cells. Supporting the possibility that IFN-{gamma} production may actually be required at some point during the development of AMM in NOD.IFN-{gamma}RB Tg mice is our observation that a similar syndrome does not develop in previously described stocks of NOD mice made genetically deficient in IFN-{gamma} or IFN-{gamma}RB (11, 12).

The suppression of Th1 responses through expression of the IFN-{gamma}RB Tg does not independently induce AMM development, but rather does so through interactions with other genetic components characterizing the autoimmune prone NOD background. This was demonstrated by the absence of AMM in the nonautoimmune prone B6 stock to which the IFN-{gamma}RB Tg was originally introduced. Outcross studies of NOD mice with several different nonautoimmune prone strains, including B6, have identified multiple polymorphic genetic loci conferring susceptibility or resistance to T1D (33, 34). A number of these overlap genetic loci associated with susceptibility or resistance to other autoimmune diseases (35, 36). Thus, it appears that NOD mice are characterized by a genetic constellation providing generalized susceptibility to multiple autoimmune syndromes, and among these, myositis development is amplified for unknown reasons by the Th1 suppressive effects of the IFN-{gamma}RB Tg. Furthermore, if this scenario is correct, it would be predicted that at least some portion of the genetic loci of B6 origin found to confer resistance to T1D should also induce protection from IFN-{gamma}RB Tg-induced AMM. Such genetic segregation studies are currently underway.

In conclusion, our current results demonstrate that a manipulation which inhibits Th1 cytokine responses in NOD mice induces a CD8 T cell-dependent AMM syndrome resulting in early paralysis, while also failing to exert T1D protective effects. This outcome indicates that altering cytokine balances can exacerbate the development of other autoimmune effectors which are normally present at lower levels, or would develop at a later time, than those causing T1D in NOD mice. The fact that to date no Th1 suppressive strategy has exerted an T1D protective effect in NOD mice (11, 22, 23), and the one in the current study actually exacerbated another lethal autoimmune syndrome, should be taken into account before considering Th1 inhibitory approaches to block progression to overt T1D in high risk humans.


   Footnotes
 
1 This work was supported by National Institutes of Health Grants DK51090 and DK46266, Cancer Center Support Grant CA34196, and by grants from the Juvenile Diabetes Research Foundation. Back

2 Address correspondence and reprint requests to Dr. David V. Serreze, The Jackson Laboratory, 600 Main Street, Bar Harbor, ME 04609. E-mail address: dvs{at}jax.org Back

3 Abbreviations used in this paper: NOD, nonobese diabetic; T1D, type 1 diabetes; AMM, autoimmune-mediated myositis; DC, dendritic cell. Back

Received for publication April 10, 2002. Accepted for publication December 17, 2002.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Serreze, D. V., E. H. Leiter. 2001. Genes and pathways underlying autoimmune diabetes in NOD mice. M. G. von Herrath, ed. Molecular Pathology of Insulin Dependent Diabetes Mellitus 31. Karger Press, New York.
  2. Leiter, E. H.. 1990. The NOD mouse meets the "Nerup Hypothesis": is diabetogenesis the result of a collection of common alleles present in unfavorable combinations?. P. Vardi, and E. Shafrir, eds. Frontiers in Diabetes Research: Lessons from Animal Diabetes III 54. Smith-Gordon, London.
  3. Baxter, A. G., A. C. Horsfall, D. Healey, P. Ozegbe, S. Day, D. G. Williams, A. Cooke. 1994. Mycobacteria precipitate an SLE-like syndrome in diabetes-prone NOD mice. Immunology 83:227.[Medline]
  4. Winer, S., I. Astsaturov, R. K. Cheung, L. Gunaratnam, V. Kubiak, M. A. Cortez, M. Moscarello, P. W. O’Connor, C. McKerlie, D. J. Becker, H.-M. Dosch. 2001. Type 1 diabetes and multiple sclerosis patients target islet plus central nervous system autoantigens; nonimmunized nonobese diabetic mice can develop autoimmune encephalitis. J. Immunol. 166:2831.[Abstract/Free Full Text]
  5. Salomon, B., L. Rhee, A. Montag, B. Soliven, J. Arcella, A. M. Girvin, S. D. Miller, J. A. Bluestone. 2001. B7-2 costimulation protects NOD mice from cryptic autoimmune peripheral neuropathy. J. Exp. Med. 194:677.[Abstract/Free Full Text]
  6. Thorsby, E., D. Undlien. 1996. The HLA associated predisposition to type 1 diabetes and other autoimmune diseases. J. Pediatr. Endocrinol. Metab. 9:75.
  7. Akhtar, I., J. P. Gold, L.-Y. Pan, J. L. M. Ferrara, X.-D. Yang, J. I. Kim, K.-N. Tan. 1995. CD4+ {beta} islet cell-reactive T cell clones that suppress autoimmune diabetes in nonobese diabetic mice. J. Exp. Med. 182:87.[Abstract/Free Full Text]
  8. Pakala, S. V., M. O. Kurrer, J. D. Katz. 1997. T helper 2 (Th2) T cells induce acute pancreatitis and diabetes in immune-compromised nonobese diabetic (NOD) mice. J. Exp. Med. 186:299.[Abstract/Free Full Text]
  9. Zekzer, D., F. S. Wong, L. Wen, M. Altieri, T. Gurlo, H. von Grafenstein, R. S. Sherwin. 1997. Inhibition of diabetes by an insulin reactive CD4 T-cell clone in the nonobese diabetic mouse. Diabetes 46:1124.[Abstract]
  10. Poulin, M., K. Haskins. 2000. Induction of diabetes in nonobese diabetic mice by Th2 T cell clones from a TCR transgenic mouse. J. Immunol. 164:3072.[Abstract/Free Full Text]
  11. Serreze, D. V., C. M. Post, H. D. Chapman, E. J. Johnson, B. Lu, P. B. Rothman. 2000. Interferon-{gamma} receptor signaling is dispensable in the development of autoimmune type 1 diabetes in NOD mice. Diabetes 49:2007.[Abstract/Free Full Text]
  12. Serreze, D. V., H. D. Chapman, C. M. Post, E. A. Johnson, W. L. Suarez-Pinzon, A. Rabinovitch. 2001. Th1 to Th2 cytokine shifts in NOD mice: sometimes an outcome, rather than the cause of diabetes resistance elicited by immunostimulation. J. Immunol. 166:1352.[Abstract/Free Full Text]
  13. Tisch, R., H. McDevitt. 1996. Insulin-dependent diabetes mellitus. Cell 85:291.[Medline]
  14. Delovitch, T. L., B. Singh. 1997. The nonobese diabetic mouse as a model of autoimmune diabetes: immune dysregulation gets the NOD. Immunity 7:291.[Medline]
  15. Pernis, A., S. Gupta, K. J. Gollub, E. Garfein, R. L. Coffman, C. Schindler, P. Rothman. 1995. Lack of interferon {gamma} receptor {beta} chain and the prevention of interferon {gamma} signaling in Th1 cells. Science 269:245.[Abstract/Free Full Text]
  16. Tau, G. Z., T. von der Weid, B. Lu, S. Cowan, M. Kvatyuk, A. Pernis, G. Cattoretti, N. S. Braunstein, R. L. Coffman, P. B. Rothman. 2000. Interferon {gamma} signaling alters the function of T helper type 1 cells. J. Exp. Med. 192:977.[Abstract/Free Full Text]
  17. Prochazka, M., H. R. Gaskins, L. D. Shultz, E. H. Leiter. 1992. The nonobese diabetic scid mouse: model for spontaneous thymomagenesis associated with immunodeficiency. Proc. Natl. Acad. Sci. USA 89:3290.[Abstract/Free Full Text]
  18. Serreze, D. V., H. D. Chapman, D. S. Varnum, M. S. Hanson, P. C. Reifsnyder, S. D. Richard, S. A. Fleming, E. H. Leiter, L. D. Shultz. 1996. B lymphocytes are essential for the initiation of T cell mediated autoimmune diabetes: analysis of a new "speed congenic" stock of NOD. Igµnull mice. J. Exp. Med. 184:2049.[Abstract/Free Full Text]
  19. Chiu, P. P. L., D. V. Serreze, J. S. Danska. 2001. Development and function of diabetogenic T cells in B cell deficient non-obese diabetic mice. Diabetes 50:763.[Abstract/Free Full Text]
  20. Serreze, D. V., E. H. Leiter. 1988. Defective activation of T suppressor cell function in nonobese diabetic mice: potential relation to cytokine deficiencies. J. Immunol. 140:3801.[Abstract]
  21. Christianson, S. W., L. D. Shultz, E. H. Leiter. 1993. Adoptive transfer of diabetes into immunodeficient NOD-scid/scid mice: relative contributions of CD4+ and CD8+ T lymphocytes from diabetic versus prediabetic NOD.NON-Thy1a donors. Diabetes 42:44.[Abstract]
  22. Hultgren, B., X. Huang, N. Dybdal, T. A. Stewart. 1996. Genetic absence of {gamma}-interferon delays but does not prevent diabetes in NOD mice. Diabetes 45:812.[Abstract]
  23. Kanagawa, O., G. Xu, A. Tevaarwerk, B. A. Vaupel. 2000. Protection of nonobese diabetic mice from diabetes by gene(s) closely linked to IFN-{gamma} receptor loci. J. Immunol. 164:3919.[Abstract/Free Full Text]
  24. Pedotti, R., D. Mitchell, J. Wedemeyer, M. Karpuj, D. Chabas, E. M. Hattab, M. Tsai, S. J. Galli, L. Steinman. 2001. An unexpected version of horror autotoxicus: anaphylactic shock to a self peptide. Nat. Immunol. 2:216.[Medline]
  25. Dudley, M. E., D. C. Roopenian. 1996. Loss of a unique tumor antigen by cytotoxic T lymphocyte immunoselection from a 3-methylcholanthrene-induced mouse sarcoma reveals secondary unique and shared antigens. J. Exp. Med. 184:441.[Abstract/Free Full Text]
  26. Hohlfeld, R., A. G. Engel. 1994. The immunobiology of muscle. Immunol. Today 15:269.[Medline]
  27. Bender, A., N. Ernst, A. Iglesias, K. Dornmair, H. Wekerle, R. Hohlfeld. 1995. T cell receptor repertoire in polymyositis: clonal expansion of aggressive CD8+ T cells. J. Exp. Med. 181:1863.[Abstract/Free Full Text]
  28. DiLorenzo, T. P., R. T. Graser, T. Ono, G. J. Christianson, H. D. Chapman, D. C. Roopenian, S. G. Nathenson, D. V. Serreze. 1998. MHC class I-restricted T cells are required for all but the end stages of diabetes development in NOD mice and utilize a prevalent T cell receptor {alpha} chain gene rearrangement. Proc. Natl. Acad. Sci. USA 95:12538.[Abstract/Free Full Text]
  29. Amrani, A., J. Verdaguer, P. Serra, S. Tafuro, R. Tan, P. Santamaria. 2000. Progression of autoimmune diabetes driven by avidity maturation of a T-cell population. Nature 406:739.[Medline]
  30. Nerup, J., T. Mandrup-Poulsen, S. Helqvist, H. U. Andersen, F. Pociot, J. I. Reimers, B. G. Cuartero, A. E. Karlsen, U. Bjerre, T. Lorenzen. 1994. On the pathogenesis of IDDM. Diabetologia 37:(Suppl. 2):82.[Medline]
  31. Amrani, A., J. Verdaguer, S. Thiessen, S. Bou, P. Santamaria. 2000. IL-1{alpha}, IL-1{beta}, and IFN-{gamma} mark {beta} cells for Fas-dependent destruction by diabetogenic CD4+ T lymphocytes. J. Clin. Invest. 105:459.[Medline]
  32. Lu, B., C. Ebensperger, Z. Dembic, Y. Wang, M. Kvatyuk, T. Lu, R. L. Coffman, S. Pestka, P. B. Rothman. 1998. Targeted disruption of the interferon-{gamma} receptor 2 gene results in severe immune defects in mice. Proc. Natl. Acad. Sci. USA 95:8233.[Abstract/Free Full Text]
  33. Wicker, L. S., J. A. Todd, L. B. Peterson. 1995. Genetic control of autoimmune diabetes in the NOD mouse. Annu. Rev. Immunol. 13:179.[Medline]
  34. Leiter, E. H.. 1998. Genetics and immunogenetics of NOD mice and related strains. E. H. Leiter, and M. A. Atkinson, eds. NOD Mice and Related Strains: Research Applications in Diabetes, AIDS, Cancer, and Other Diseases 37. R. G. Landes, Austin.
  35. Vyse, T. J., J. A. Todd. 1996. Genetic analysis of autoimmune disease. Cell 85:311.[Medline]
  36. Wandstrat, A., E. Wakeland. 2001. The genetics of complex autoimmune diseases: non-MHC susceptibility genes. Nat. Immunol. 2:802.[Medline]



This article has been cited by other articles:


Home page
 J. Immunol.Home page
H. Bour-Jordan, H. L. Thompson, and J. A. Bluestone
Distinct Effector Mechanisms in the Development of Autoimmune Neuropathy versus Diabetes in Nonobese Diabetic Mice
J. Immunol., November 1, 2005; 175(9): 5649 - 5655.
[Abstract] [Full Text] [PDF]

Home page
 J. Immunol.Home page
S. M. Lieberman, T. Takaki, B. Han, P. Santamaria, D. V. Serreze, and T. P. DiLorenzo
Individual Nonobese Diabetic Mice Exhibit Unique Patterns of CD8+ T Cell Reactivity to Three Islet Antigens, Including the Newly Identified Widely Expressed Dystrophia Myotonica Kinase
J. Immunol., December 1, 2004; 173(11): 6727 - 6734.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Serreze, D. V.
Right arrow Articles by Cox, G. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Serreze, D. V.
Right arrow Articles by Cox, G. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS