SEARCH

SEARCH BY CITATION

Keywords:

  • autoimmunity;
  • diabetes;
  • salivary gland;
  • regulatory T cell;
  • thyroid

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Approximately 80% of female wild-type non-obese diabetic (WT NOD) mice spontaneously develop diabetes, whereas B-cell-deficient (B−/−) NOD mice are resistant to diabetes. B−/− mice are also resistant to other spontaneous and experimentally induced autoimmune diseases, including arthritis, systemic lupus erythematosus, Sjögren syndrome and thyroiditis. Under normal conditions, activation of self-reactive T cells in the periphery is limited by CD4+ CD25+ natural regulatory T (Treg) cells. B−/− NOD.H-2h4 mice, normally resistant to spontaneous autoimmune thyroiditis (SAT), develop SAT when Treg cells are depleted, suggesting that Treg cells are preferentially activated when autoantigen is initially presented by non-B-cell antigen-presenting cells. To test the hypothesis that increased Treg cell activity in B−/− mice contributes to their resistance to other autoimmune diseases, WT and B−/− NOD mice were given anti-CD25 to transiently deplete CD4+ CD25+ Treg cells. The WT and B−/− NOD mice given anti-CD25 developed diabetes much earlier than WT mice given rat IgG, whereas rat IgG-treated B−/− mice did not develop diabetes. Treg-cell-depleted mice had increased lymphocyte infiltration of the pancreas, salivary glands and thyroid compared with controls given rat IgG. These results are consistent with the hypothesis that resistance of B-cell-deficient NOD mice to several autoimmune diseases is due to the activity of Treg cells.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

B-cell-deficient (B−/−) mice are resistant to most spontaneous and some experimentally induced autoimmune diseases, including diabetes,[1-4] thyroiditis,[5-7] arthritis,[8, 9] Sjögren syndrome[10-13] and systemic lupus erythematosus.[14, 15] In addition to producing antibody, B cells can function as antigen-presenting cells and may amplify or sustain responses initiated by other antigen-presenting cells.[1-3, 6, 8, 14-17] B cells are important for developing ectopic germinal centres found in sites of inflammation in thyroiditis and arthritis, where they can promote T-cell activation and cytokine production.[18-20] The mechanisms underlying the resistance of B−/− mice to many autoimmune diseases are not yet fully understood. Recent studies suggest that a lack of regulatory T (Treg) cells in wild-type (WT) mice results in increased autoimmunity caused by a skewing of the effector T-cell pool towards the follicular T-cell subset.[21, 22] Through interaction of cell surface proteins or secreted factors, B cells could limit the generation, activation or function of Treg cells.[23-26] Under normal conditions, activation of self-reactive T cells in the periphery is limited by CD4+ CD25+ natural Treg cells.[27-29] Previous work by our laboratory and others suggests that B cells could limit the activation of Treg cells, given that Treg cells show increased activation in the absence of B cells.[7, 30, 31]

Our previous studies showed that B−/− NOD.H-2h4 mice, normally resistant to spontaneous autoimmune thyroiditis (SAT), develop SAT if CD25+ Treg cells are transiently depleted.[6] These studies suggested that if autoantigen is initially presented to autoreactive CD4+ T cells by non-B-cell antigen-presenting cells, then Treg cells are preferentially activated and activation of effector T cells is inhibited. In contrast, if Treg cells are absent, effector T cells will be activated, and autoimmune disease will develop. We hypothesized that increased Treg cell activity could be a general mechanism that could explain, at least in part, the resistance of B−/− mice to many autoimmune diseases. To test this hypothesis B−/− non-obese diabetic (NOD) mice, which are resistant to diabetes, were transiently depleted of Treg cells by administration of anti-CD25 antibody.[1-3] The results indicate that anti-CD25-treated B−/− NOD mice develop diabetes similar to that which develops in anti-CD25-treated WT NOD mice. Interestingly, Treg cell depletion also altered the kinetics of diabetes development, with much earlier onset of disease in both B−/− and WT mice. Treg-cell-depleted B−/− NOD mice also had increased salivary gland infiltration and increased incidence of thyroiditis. This increased Treg cell activity appears to have a role in the resistance of B−/− NOD mice to several autoimmune diseases.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References
Mice

The WT NOD mice and B−/− NOD mice were provided by Dr Habib Zaghouani (University of Missouri School of Medicine, Columbia, MO) and The Jackson Laboratories (Bar Harbor, ME), respectively. All mice were bred and maintained under specific pathogen-free conditions in our animal facilities at the University of Missouri. All animal protocols were approved by the University of Missouri Animal Care and Use Committees. Only female mice were used, and mice were age-matched for each experiment.

Antibody treatment and experimental design

Mice were given two 0·5-mg injections, 1 week apart, of rat anti-mouse CD25 monoclonal antibody PC61 (BioXcell, West Lebanon, NH) beginning in most experiments 10–11 days after birth.[6, 32] This treatment results in depletion of all detectable splenic and circulating CD25+ Foxp3+ cells 1 week later. The Treg cells gradually return over the next 6 weeks and are present in normal numbers by the time mice are 8 weeks of age (our unpublished results). Age-matched control mice were given rat IgG at the same dose and injection schedule. Blood glucose levels were tested starting at 8 weeks of age. Pancreas, thyroid and salivary gland were harvested when mice were aged 16 weeks, sectioned and stained with haematoxylin and eosin as previously described.[5]

Assessment of hyperglycaemia and diabetes

Assessment of blood glucose levels was done using test strips and an Accu-Chek Advantage monitoring system (Roche Diagnostics, Indianapolis, IN) starting at 8 weeks of age. Mice were tested weekly until 16 or 30 weeks. A mouse is considered diabetic when the blood glucose level is 300 mg/dl for two consecutive weeks.

Assessment of pancreatic islet infiltration

Pancreata were harvested from WT or B−/− NOD females given anti-CD25 or rat IgG, fixed in 10% formalin, and embedded in paraffin. Four sections per pancreas were stained with haematoxylin and eosin and analysed by light microscopy. Insulitis scoring was performed according to the following criteria: severe insulitis, 50% or more of the islet area is infiltrated; mild insulitis, < 50% of the islet area is infiltrated; peri-insulitis, infiltration is restricted to the periphery of islets; and no insulitis, absence of cell infiltration.[33]

Assessment of thyroiditis

Thyroids were collected and one thyroid lobe from each mouse was fixed in formalin, sectioned, and stained with haematoxylin and eosin as described previously.[5, 34] Thyroid histopathology was scored for the extent of thyroid follicle destruction using a scale of 0 to 4 + as described previously.[5, 19, 34] In brief, a score of 0 indicates a normal thyroid, and 0 + indicates a few inflammatory cells infiltrating the thyroids and/or mild follicular changes. A 1 + severity score is defined as an infiltrate of at least 125 cells in one or several foci, and a 2 + score represents 10–20 foci of cellular infiltration, each the size of several follicles, with destruction of up to one-quarter of the gland. A 3 + score indicates that one-quarter to one-half of the thyroid follicles are destroyed or replaced by infiltrating inflammatory cells, whereas a 4 + score indicates that more than half of the gland is destroyed. As described previously,[34] thyroid lesions in NOD.H-2h4 mice reach maximal severity 7–9 weeks after mice are given NaI water beginning at 2 months of age; lesions are chronic and remain relatively unchanged in severity for several months.[19, 34]

Assessment of salivary gland infiltration

Salivary glands were harvested from NOD females, fixed in 10% formalin, and embedded in paraffin. Sections were stained with haematoxylin and eosin and analysed by light microscopy. Lymphocytic foci per submandibular section were defined as an infiltrate consisting of > 50 cells.[35, 36]

Statistical analysis

Statistical analysis was performed using GraphPad Prism software version 4.0 (GraphPad Software Inc., La Jolla, CA) with the non-parametric Mann–Whitney test. A value of P < 0·05 was considered statistically significant.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

Depletion of CD25+ Treg cells results in early onset of diabetes in WT and B−/− NOD mice

Approximately 80% of female WT NOD mice spontaneously develop diabetes,[37] but B−/− NOD mice are resistant to diabetes (Fig. 1a).[1-3] To test the hypothesis that B−/− NOD mice would develop diabetes after transient depletion of Treg cells, female NOD mice were given two injections of anti-CD25 (0·5 mg) 1 week apart beginning 10–11 days after birth.[6] This resulted in depletion of all detectable CD4+ FoxP3+ CD25+ Treg cells. They remained depleted for 2–3 weeks and gradually returned with normal numbers of Treg cells present by 7–9 weeks of age (data not shown). Blood glucose levels were determined weekly starting at week 8.[6] None of the B−/− mice given rat IgG control developed diabetes by 16 weeks of age (Fig. 1b). However, B−/− NOD mice given anti-CD25 developed diabetes earlier than WT NOD mice that did not receive anti-CD25, with half of the Treg-cell-depleted B−/− mice being diabetic by 11–12 weeks of age compared with week 21–22 in rat IgG-treated WT mice (Fig. 1a,b). WT NOD mice depleted of Treg cells by anti-CD25 also developed diabetes earlier than WT NOD mice given rat IgG (Fig. 1c), as previously reported.[38] The experiments had to be terminated at 16 weeks because of the severe diabetes in anti-CD25-treated WT mice. Controls were terminated at the same time to directly compare inflammatory cell infiltration in both groups. These results suggest that Treg-cell activity, at least in part, is responsible for diabetes resistance in B−/− NOD mice. Given that transient depletion of Treg cells results in earlier onset of disease in WT NOD mice, Treg cells are apparently able to delay diabetes onset in rat IgG-treated WT NOD mice but are ultimately unable to control the disease. In B−/− mice, however, Treg cells are apparently able to suppress diabetes completely.

image

Figure 1. Depletion of CD25+ regulatory T (Treg) cells results in early onset and increased incidence of diabetes. Blood glucose levels were determined in unmanipulated wild-type (WT) or B-cell-deficient (B−/−) non-obese diabetic (NOD) mice starting at week 10 of age. Graph represents percentage of diabetic mice over time; n = 15 (a). B−/− (b) or WT (c) NOD mice were given two injections of 0·5 mg anti-CD25 antibody 1 week apart to deplete CD4+ CD25+ Treg cells, or rat IgG as control. Blood glucose levels were determined weekly starting at 8 weeks of age and continued until 16 weeks of age. Experiments were terminated at 16 weeks because of health issues in diabetic animals; 14–15 mice per group.

Download figure to PowerPoint

Treg-cell-depleted mice have increased infiltration of inflammatory cells and islet destruction in the pancreas

To determine whether Treg cell depletion resulted in increased pancreatic lymphocyte infiltrate and islet destruction, pancreata of 16-week-old anti-CD25 or control rat IgG-treated WT and B−/− mice were stained with haematoxylin and eosin. Islets from anti-CD25-treated mice had increased infiltration and islet destruction; whereas islets of the majority of rat IgG-treated mice had little infiltration (Figs 2 and 3a). There are also fewer islets in anti-CD25-treated mice, suggesting that many islets have already been completely destroyed (Fig. 3b). Surprisingly, even though diabetes develops early in B−/− mice given anti-CD25, there are still a few islets with no insulitis (Fig. 3a). Pancreata from anti-CD25-treated B−/− mice showed increased intra-insulitis compared with control B−/− mice (Figs 2 and 3a). No islets lacking infiltration were seen in pancreata from anti-CD25-treated WT mice, and the numbers of islets exhibiting severe intra-insulitis are greatly increased in the Treg-cell-depleted WT mice (Fig. 3a). It is likely that the reduction of total islet numbers in both anti-CD25-treated groups is a result of their increased islet infiltration and destruction, as completely destroyed islets are not detectable (Fig. 3b). These results suggest that there is less islet destruction in Treg-cell-depleted B−/− mice than in Treg-cell-depleted WT NOD mice but islet destruction in both groups is sufficient to result in diabetes.

image

Figure 2. Depletion of CD25+ regulatory T (Treg) cells results in increased lymphocytic infiltration of multiple tissues. B-cell deficient (B−/−) or wild-type (WT) non-obese diabetic (NOD) mice were given two injections of 0·5 mg anti-CD25 antibody 1 week apart to deplete CD4+ CD25+ Treg cells. Control mice were given rat IgG. At 16 weeks of age the pancreas, salivary gland and thyroid were harvested and stained with haematoxylin and eosin. Three representative mice are shown for each group. Arrows indicate lymphocytic infiltration. Pancreas sections show intra-insulitis in anti-CD25 groups, and peri-insulitis in control groups. Magnification: 100 ×.

Download figure to PowerPoint

image

Figure 3. Pancreatic infiltration after regulatory T (Treg) cell depletion in wild-type (WT) or B-cell-deficient mice. At 16 weeks of age, pancreas was harvested from mice treated with two, weekly injections of anti-CD25 or rat IgG starting 10–11 days after birth. The islets were scored as no insulitis, peri-insulitis, mild intra-insulitis or severe intra-insulitis as described in the 'Materials and methods': (a) depicts the percentage of total islets from each group that were given each score. (b) represents the number of islets of each score and the overall total number of islets per group.

Download figure to PowerPoint

Salivary gland infiltration and thyroiditis develop when CD4+ CD25+ Treg cells are depleted

B-cell-deficient mice are also resistant to other autoimmune diseases, including Sjögren syndrome[10] and thyroiditis.[5, 6] WT NOD mice develop both of these autoimmune diseases, and we hypothesized that depletion of Treg cells in B−/− NOD mice should result in both Sjögren syndrome and thyroiditis. The results are consistent with this hypothesis, as there were increased foci of lymphocytes in the salivary glands of Treg-cell-depleted B−/− mice (P = 0·023) compared with rat IgG-treated B−/− controls (Figs 2 and 4a). There is also a significant increase in the foci present in salivary glands of anti-CD25-treated WT NOD mice (P = 0·023) compared with rat IgG-treated WT controls.

image

Figure 4. Depletion of CD25+ regulatory T (Treg) cells results in significant increase in salivary gland infiltrate and thyroiditis in B-cell-deficient (B−/−) mice. B−/− or wild-type (WT) non-obese diabetic (NOD) mice were given two injections of 0·5 mg anti-CD25 antibody 1 week apart to deplete CD4+ CD25+ Treg cells. At 16 weeks of age the salivary gland (a) and thyroid (b) were harvested and stained with haematoxylin and eosin. (a) The number of foci of lymphocytic infiltrate in the salivary gland were counted and (b) thyroid infiltration and follicular destruction were scored as described in 'Materials and methods'. Depletion of regulatory T cells resulted in significant increases in salivary and thyroid infiltrates in both B−/− and WT mice. *P < 0·05, ***P < 0·001.

Download figure to PowerPoint

NOD mice can develop spontaneous thyroiditis.[39-41] In the related NOD-H.2h4 strain, essentially all WT mice develop SAT when given NaI in their drinking water.[19] B−/− NOD-H.2h4 mice are resistant to SAT development,[5, 6] but they develop SAT after depletion of CD4+ CD25+ Treg cells.[6] It was therefore of interest to determine if thyroiditis would be increased in WT or B−/− NOD mice given anti-CD25 to deplete Treg cells. To address this, thyroid glands from the same groups of mice were harvested, fixed, sectioned, and stained with haematoxylin and eosin. Both WT and B−/− mice develop thyroiditis when Treg cells are depleted, but the incidence and severity of thyroiditis was minimal in both groups of mice that were not Treg-cell-depleted (Fig. 4b). Mice treated with anti-CD25 showed increased infiltration of multiple organs as well as diabetes. Whereas only 16% of control WT mice had coincident lymphocyte infiltration of the pancreas, salivary gland and thyroid, more than 70% of WT mice treated with anti-CD25 had infiltration in all three organs. None of the control B−/− mice had significant infiltration in all three organs, but 80% of anti-CD25 treated B−/− mice had coincident infiltration of all three organs (Table 1). The results show that depletion of CD25+ Treg cells leads to the development of multiple autoimmune diseases in WT and B−/− NOD mice.

Table 1. Regulatory T cell depletion results in lymphocyte infiltration of multiple organs in the same mice
GroupaDiabetesbIslet Inf.cSal. Inf.dSAT Severitye
  1. a

    Wild-type or B-cell-deficient non-obese diabetic (NOD) mice were given two injections of anti-CD25 1 week apart beginning 10–11 days after birth. Control mice were given rat IgG at the same dose and injection schedule.

  2. b

    Blood glucose levels were determined weekly starting at 8 weeks of age. Symbols represent diabetic (+) or non-diabetic (−) mice at 16 weeks of age.

  3. c

    The presence of islet infiltration was determined by haematoxylin and eosin staining. Levels of infiltration are graded as follows: −, not detectable; +, weak; ++, moderate; +++, strong.

  4. d

    The presence of salivary gland infiltration was determined by haematoxylin and eosin as described in 'Materials and methods'. The level of infiltration is graded as follows: −, not detectable; +, weak; ++, moderate; +++, strong.

  5. e

    Spontaneous autoimmune thyroiditis (SAT) severity score was determined as described in 'Materials and methods'. SAT score of each mouse is listed. n.t., no thyroid tissue harvested.

Wild-type control
A+++n.t.
B+−/+n.t.
C++1
D−/++0
E++0
F+1
Wild-type anti-CD25
A+−/++++0
B++++2
C+++++3
D++++++3
E+++++2
F+++3
G++++2
B-cell-deficient control
A++0
B+0
C0
D+0
E1
B-cell-deficient anti-CD25
A+++3
B++++2
C++1
D+++4
E++++3

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

B-cell-deficient NOD mice are resistant to diabetes.[1-3] In addition to their role as antibody-producing cells, B cells can function as antigen-presenting cells. Diabetes susceptibility in NOD mice is known to involve a requirement for B cells as antigen-presenting cells.[1, 2, 4, 42] Antibody-mediated depletion of B cells inhibits diabetes onset.[3, 4] Given our previous results regarding the importance of Treg cells in determining the resistance of B−/− NOD.H-2h4 mice to SAT,[6] the current study was undertaken to test the hypothesis that transient depletion of CD4+CD25+ Treg cells following administration of anti-CD25 antibody would result in spontaneous diabetes in B−/− NOD mice. Both WT and B−/− NOD mice were given anti-CD25 to transiently deplete CD4+ CD25+ regulatory T cells. Treg cell depletion using this protocol is transient, and they had returned to normal numbers before beginning blood glucose monitoring. The results indicate that depletion of Treg cells in B−/− mice results in diabetes development, whereas rat IgG-treated B−/− mice do not develop diabetes. In fact, overt diabetes developed as early as 8 weeks in one animal, the first time-point for blood glucose testing. Previous studies suggest that Treg cell depletion also greatly alters the kinetics of diabetes development in WT NOD mice.[38] The results in Fig. 1(a) and (c) show that average diabetes onset is at 11–12 weeks in WT mice given anti-CD25 compared with 21–22 weeks in control mice. The loss of resistance to diabetes in Treg-cell-depleted B−/− NOD mice and the earlier development of diabetes in Treg-cell-depleted WT NOD mice could result from an increase in diabetogenic effector T cells. Alternatively, the removal of Treg cell suppression may simply allow those islet antigen-specific effector T cells that are already present to become pathogenic. Although our studies did not address whether diabetogenic effector T cells were increased following Treg cell depletion, others have shown that Treg cell depletion allowed T cells with known islet reactivity to undergo increased proliferation in pancreatic lymph nodes,[38] and depletion of CD4+ FoxP3+ Treg cells leads to increased proliferation of islet antigen-specific BDC2.5 CD4+ T cells in the pancreas.[43, 44] However, the percentage of mice that eventually became diabetic (~ 75–80%) was similar for both control and Treg-cell-depleted WT NOD mice, suggesting that if effector cells are expanded after Treg cell depletion, this results in earlier disease onset but does not lead to an increased incidence of diabetes.

In addition to being resistant to diabetes, B−/− mice are resistant to many other autoimmune diseases, such as thyroiditis,[5, 6] Sjögren syndrome,[10, 11] arthritis[8, 9] and systemic lupus erythematosus.[14, 15] Some WT NOD mice develop lymphocyte infiltration of salivary glands and thyroid glands characteristic of Sjögren syndrome and thyroiditis.[39-41, 45-48] Salivary gland infiltration was significantly increased in Treg-cell-depleted B−/− mice compared with controls (Fig. 4a) (P = 0·023). Although there was no significant difference between the numbers of lymphocyte foci in Treg-cell-depleted B−/− mice and WT rat IgG-treated controls (P = 0·133), mice that received anti-CD25 had larger foci of infiltration (Fig. 2). The data in Fig. 4(b) show a significant increase in thyroiditis severity in B−/− mice that were treated with anti-CD25 (P < 0·0001), suggesting that Treg cells play an important role in suppression of thyroiditis in B−/− NOD mice as shown previously for B−/− NOD.H-2h4 mice.[6] A significant increase in thyroiditis severity in WT NOD mice treated with anti-CD25 was also observed (P = 0·033). It is interesting that Treg cell depletion greatly increases the incidence of autoimmune infiltration of multiple organs in WT NOD mice, given that Treg cell depletion has little effect on thyroiditis severity in the related WT NOD.H-2h4 mouse.[6, 7]

B cells can function as antigen-presenting cells and are capable of modulating T-cell responses.[49-51] B cells can promote T-cell activation and cytokine production in ectopic germinal centres found in the sites of inflammation in thyroiditis and arthritis.[18-20] However, the mechanisms underlying the inhibition of autoimmune diseases through B-cell depletion are not yet fully understood. In the light of recent studies, it seems likely that one mechanism could relate to B-cell interaction with Treg cells.[4, 6, 30, 31] Regulatory T cells are involved in modulating the activation of autoreactive T cells in the periphery.[27-29] We have previously shown that B−/− NOD.H-2h4 mice, normally resistant to SAT, develop SAT when depleted of CD25+ Treg cells.[6] In addition to being important for effector T-cell activation, B cells could regulate the suppressive function of Treg cells. In B−/− mice, this regulation of Treg cell function is lacking, and the Treg cells are therefore able to fully suppress the autoimmune effectors. When Treg cells are depleted, their effect is removed and autoimmune disease can develop. Alternatively, the activation of effector T cells that occurs when B cells are present may result in a decreased ability of Treg cells to inhibit autoimmunity.

The results presented here demonstrate that depletion of Treg cells results in development of multiple organ-specific autoimmune diseases in B−/− NOD mice. Treg cell depletion also greatly altered the kinetics of diabetes development, resulting in earlier onset of diabetes in both B−/− and WT animals. Others have previously shown that B-cell depletion in NOD mice leads to increases in functional Treg cells, and that Treg cells play a critical role in protection against diabetes.[4, 31] Those results are consistent with our earlier reports indicating that Treg cells are responsible, at least in part, for the resistance of B−/− NOD.H-2h4 mice to development of SAT and for suppression of SAT in mice given anti-CD20 to deplete B cells.[6, 7] The current studies agree with those earlier studies, and extend them by showing that Treg cell depletion results in a coincident lymphocyte infiltration of pancreas, salivary glands and thyroid in most anti-CD25-treated WT and B−/− NOD mice (Table 1). Depletion of Treg cells apparently leads to increased activation of autoreactive T cells that mediate diabetes, thyroiditis and Sjögren syndrome even though Treg cells repopulate after only a few weeks. In WT mice, autoreactive effector T cells can be activated despite the presence of Treg cells, but when B cells are lacking, Treg cells more fully suppress autoreactive effector T-cell activation. This agrees with recent results in our laboratory indicating that Treg cells from B−/− NOD.H-2h4 mice are better suppressors of SAT than Treg cells from WT mice and that Treg cells that return after Treg cell depletion differ phenotypically from those present before depletion, and are less capable of suppression (unpublished data).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References

This work was supported by National Institutes of Health grant RO1 AI 076395.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. Disclosures
  9. References