Enhanced production of early lineages of monocytic and granulocytic cells in mice with colitis

Abstract

The bone marrow (BM) is a large, highly active, and responsive tissue. Interestingly, little is known about the impact of colitis on hematopoietic functions. Using dextran sodium sulfate (DSS) to induce colitis in mice, we identified significant changes in the BM. Specifically, cells of the monocytic and granulocytic lineages increased nearly 60% and 80%, respectively. This change would support and promote the large infiltration of the gut with neutrophils and monocytes that are the primary cause of inflammation and tissue damage during colitis. Conversely, the early lineages of B and T cells declined in the marrow and thymus with particularly large losses observed among pre-B and pre-T cells with heightened levels of apoptosis noted among CD4⁺CD8⁺ thymocytes from DSS-treated mice. Also noteworthy was the 40% decline in cells of the erythrocytic lineages in the marrow of colitis mice, which undoubtedly contributed to the anemia observed in these mice. The peripheral blood reflected the marrow changes as demonstrated by a 2.6-fold increase in neutrophils, a 60% increase in monocytes, and a decline in the lymphocyte population. Thus, colitis changed the BM in profound ways that parallel the general outcomes of colitis including infiltration of the gut with monocytes and neutrophils, inflammation, and anemia. The data provide important understandings of the full impact of colitis that may lead to unique treatments and therapies.

Ulcerative colitis (UC) and Crohn’s disease (CD) are chronic inflammatory bowel diseases (IBD) that affect more than 1.4 million Americans (1). Both genetic and environmental factors contribute to the hyperactivity and dysregulation of the immune system associated with colitis. IBD is characterized by recurrent bouts of illness associated with symptoms of weight loss, diarrhea, anemia, rectal bleeding, and abdominal pain (1, 2). Past studies demonstrate that gut infiltration with neutrophils, monocytes, and macrophages is a major contributor to the gastrointestinal tissue injury and inflammation noted in IBD (3–7).

Although many facets of immune function in the IBD patient and in animal models have been investigated, almost nothing was known of the impact of IBD on the bone marrow (BM). In particular, it was not known whether the ability of marrow to produce new leukocytes or red blood cells each day was altered by UC or CD. This question was of interest because the marrow is a large, active tissue that produces billions of new cells each day and is often impacted by changes in nutriture, stress, and disease (8–10). Indeed, it was known that UC changed the rate of maturation of marrow osteoblasts, which contributed to the thinning of bones (11). Thus, the purpose of the research reported herein was to determine the impact of dextran sodium sulfate (DSS)-induced colitis on hematopoietic function, with a focus on erythropoiesis, lymphopoiesis, and myelopoiesis (the production of neutrophils and monocytes).

Myelopoietic processes in the marrow were of particular interest because neutrophils are among the first cells to be found in mucosa of the gut, especially in the case of UC (3–5, 7). They secrete large amounts of chemokines that subsequently cause increasing numbers of monocytes/macrophages to enter these areas, creating sustained injury and inflammation (3–5, 7). Indeed, cells of the myeloid lineages become so prevalent that they are found in large numbers in the stools, being used in the past as one method for determining the pathology of the various forms of IBD (12). Therefore, understanding how forms of IBD affect the production of these cells in the marrow was also important to understanding the etiology of this disease.

Although CD and UC are categorized as autoimmune diseases, it is known that modifying the epithelial barrier, for example through modulation of β-catenin, causes a Crohn’s-like colitis (13). Studies in senescence accelerated mouse P-1 (SAMP) mice, which display ileitis similar to Crohn’s disease, also suggest that defects in epithelial permeability may be the primary source of ileitis susceptibility in these mice and that inflammatory responses occur as a consequence of this initial barrier break (14). Therefore, to carefully assess how changes in barrier function impact hematopoiesis, we used a DSS mouse model of colitis. DSS treatment induces colitis by increasing gut epithelial barrier permeability, which is thought to allow gut bacteria to activate immune responses and promote colitis. The data will show that colitis made substantial changes in lymphopoiesis in both the marrow and the thymus with major losses in pre-T and pre-B cells within 15 d. Likewise, there were substantial reductions in erythrocytic lineages that would contribute to and extend the anemia in DSS-treated mice. Conversely, cells of the monocytic and granulocytic lineages increased significantly in the marrow and peripheral blood of mice with colitis, which could promote or sustain inflammation and tissue damage in the gut. These important findings clearly show that colitis significantly alters hematopoietic processes.

Results

Confirmation of Disease in Mice.

Mice, 6 wk of age, were separated into two weight-matched groups (22.4 ± 1.4 g for controls and 22.6 ± 1.6 g for colitis mice). The colitis group received 1% DSS in drinking water, whereas controls received regular drinking water. After 15 d, DSS-treated mice lost 1.8 ± 1.8 g (4.7%) of body weight, whereas controls gained 1.1 ± 0.9 g (8.0%), resulting in an 11.1% lower body weight for DSS-treated mice compared with controls (Fig. S1A). By 8 d, the majority of DSS-treated mice showed signs of rectal bleeding and soft fecal pellets. DSS-treated mice displayed significant histologic changes in cecums (Fig. S1B) including structural changes marked by longer and distorted crypt structure (scored at 4 ± 1.3 compared with 1.2 ± 1.0 for control mice), hyperchromasia, and few goblet cells (Fig. S1C). Consistent with disease, inflammatory cells were elevated in the cecum lamina propria of DSS-treated mice, with higher numbers of lymphoid aggregates (1.4 ± 0.7 compared with 0.4 ± 0.7 for control mice), a greater lymphoid cell score (5.2 ± 1.2 versus 1.4 ± 1.6 in controls), and the presence of neutrophils. The number of mucosal ulcers, a sign of disease severity, increased threefold in colitis compared with controls. The ulcers were moderate but in some cases extended into the submucosa. The overall disease score (the combination of parameters noted above) was significantly higher for colitis compared with control mice (Fig. S1D). The significant DSS-induced elevation of levels of IL-1β, TNF-α, and IL-6 mRNA in the proximal colon further demonstrates inflammation/colitis (Fig. S1E).

Impact of Colitis on Hematopoietic Processes in the Marrow.

Colitis had a significant effect on the composition of hematopoietic cells in the BM, causing changes in the proportion of cell lineages (Fig. 1). The largest affect was seen in lymphocytic cells, which were substantially reduced by colitis. DSS treatment caused a 70% decline in marrow lymphocytes, from nearly 26% of the total marrow cells in controls to only 7.7% in DSS mice. This result represents either a decline in the production and maturation of lymphocytes in mice with colitis or enhanced elimination of lymphocytes from marrow via cell death and/or egress to peripheral sites (see Fig. S2 for typical flow cytometry plots; refs. 8 and 15).

Fig. 1.

Changes in BM cellular distribution in control mice (Upper) and DSS-treated mice (Lower). Data were obtained by phenotypic labeling of BM cells with anti–Ly-6C and anti-CD31. Lymphoid cells, Ly-6C⁻/CD31⁺; erythroid cells, Ly-6C⁻/CD31⁻; monocytes, CD31⁺/Ly-6C^hi; granulocytes, CD31^−/lo/Ly-6C^med; mixed progenitors, CD31⁺/Ly-6C^med (Fig. S2). Results shown are mean percent of total BM nucleated cells ± SD. One representative experiment of three is shown. Controls, n = 10; DSS-treated, n = 11 (*P < 0.001).

The erythrocyte compartment was reduced from 16.3% to 9.8% (down 40%) of nucleated cells. These data are particularly noteworthy given that anemia often accompanies colitis, and DSS treatment of rodents is known to cause anemia (16–19). Indeed, in blood of DSS-treated mice, significantly fewer red blood cells, reduced hemoglobin, and a lower hematocrit was observed, indicative of the beginnings of anemia (Table S1). Of note, colitis did not change the total number of nucleated cells within the marrow [38.0 ± 6.1 vs. 37.0 ± 5.0 (×10⁶) in control versus colitis mice, respectively]. Thus, the deficits in cells of the lymphoid and erythroid lineages represented absolute declines in cell numbers, not just changes in the proportion of cells present.

Conversely, increases in the number and proportion of cells of the myeloid lineages (monocytes, granulocytes) were noted in colitis mice. Monocytes increased more than 58%, from 9.8% of nucleated marrow cells in controls to 15.5% of cells in DSS mice. Likewise, granulocytes in marrow increased from 34% to 60% in DSS mice, an almost 80% increase (Fig. 1). These increases in cells of the myeloid lineages represent significant changes in the composition of the marrow and, therefore, its activity. These increases are also an important facet in the etiology of colitis that has been overlooked. It should be noted that the mixed progenitor population also declined nearly 40% in DSS-treated mice (Fig. 1). Because the composition of this group was not determined here, its potential impact if any is unknown.

Because of the large decline in the lymphocyte compartment, the BM B-cell subpopulations were further examined. This discovery is important because in adults the majority of developing lymphocytes in the marrow are of the B lineage. We found that the B-cell population as a whole declined 70% in mice with colitis, representing a substantial decline in lymphopoiesis. Much of the decline in the marrow B-cell population was due to the loss of the pre–B-cell subset, which showed decreases in both large and small pre-B cells as a proportion of the residual B-cell population (71% and 24% decreases, respectively; Table 1). By contrast, late pro-B cells more than doubled in proportion. A 44% increase in mature B cells was also observed, whereas the proportion of immature B cells modestly declined (Table 1). Thus, pro-B cells and immature-mature B cells were surviving somewhat better than pre-B cells, which is not surprising because these cells are often more resistant to stress because of expression of anti-apoptotic Bcl-2 proteins. These proteins are lacking in pre-B cells undergoing Ig gene rearrangement (20), suggesting that apoptosis, which is ongoing in the pre–B-cell lineages, might be accelerated in DSS-treated mice. This change was the case for pre-T cells developing in the thymus of DSS-treated mice, as will be seen.

Table 1.

Phenotypic distribution of bone marrow B cells

Sample	Control†, %	DSS-treated mice, %
Pre Pro-B cells‡	3.9 ± 0.3	2.6 ± 0.9**
Late Pro-B cells	10.7 ± 1.2	23.6 ± 3.0***
Large Pre-B cells	11.0 ± 1.3	3.2 ± 1.6***
Small Pre-B cells	50.0 ± 2.2	37.9 ± 6.0**
Immature B cells	12.0 ± 1.0	10.2 ± 0.9**
Mature B cells	4.3 ± 0.5	6.2 ± 3.3

**P < 0.01

***P < 0.001.

^†Data shown is percent of B220⁺ population (26.2 ± 2.3% of nucleated marrow for controls, 8.0 ± 1.6% for DSS-treated mice).

^‡Pre Pro-B cells (B220⁺ Gr-1⁻ CD43^hi IgM⁻ CD19⁻); Late Pro-B cells (B220⁺ Gr-1⁻ CD43^hi IgM⁻ CD19⁺); Large Pre-B cells (B220⁺ Gr-1⁻ CD43^−/low IgM⁻ FSC^high); Small Pre-B cells (B220⁺ Gr-1⁻ CD43^−/low IgM⁻ FSC^low); Immature B cells (B220⁺ Gr-1⁻ CD43^−/low IgM⁺ IgD⁻); Mature B cells (B220⁺ Gr-1⁻ CD43^−/low IgM⁺ IgD⁺).

Impact of DSS on Thymopoietic Processes.

To complete the depiction of the impact of DSS on the primary immune tissues involved in lymphopoiesis, the thymus was investigated to compare the observed decline in B-cell genesis in the BM. The average control thymus weighed 48 mg, whereas that of colitis mice had atrophied to 28 mg. The total number of thymic cells lost was ∼43 million in DSS-treated mice, dropping from 76.4 ± 23 million present in controls to 33.1 ± 22 million in colitis mice (Table 2). This change represented a 57% decline in the number of cells in the thymus. It was of interest to investigate the distribution of T-cell subpopulations within the remaining thymocyte population. Phenotypic analysis showed the CD4⁺/CD8⁺ thymocytes in DSS mice had declined approximately 60% in absolute numbers of cells (Table 2). Similar to the case of pre-B cells, DSS treatment also caused substantial losses in the proportion of pre-T (CD4⁺/CD8⁺) cells, which declined to 75% of the residual pool of cells (Table 3). However, similar to more mature B-cell populations, modest increases were noted in the proportion of mature CD4⁻/CD8⁺ and CD4⁺/CD8⁻ T cells, and a 50% increase was noted in early CD4⁻/CD8⁻ T-cell precursors (Table 3), although overall their absolute numbers had declined (Table 2). This result is therefore analogous to the findings for changes in distribution of early B cells, because the better surviving cells have antiapoptotic proteins.

Table 2.

Numbers of CD4/CD8 subsets in thymus of control and DSS-treated mice

Sample	Total Thy (×106)	CD4−/CD8− (×106)	CD4+/CD8+ (×106)	CD4+/CD8− (×106)	CD4−/CD8+ (×106)
Control†	76.4 ± 23.1	5.16 ± 1.2	63.6 ± 19.9	2.55 ± 0.73	4.96 ± 1.5
DSS‡	33.1 ± 21.9	2.85 ± 1.5	26.5 ± 18.5	1.27 ± 0.67	2.39 ± 1.5
Change, %	−56.7**	−44.8	−58.3**	−50.1**	−51.8**

**Thy, thymocytes; P < 0.01.

^†Control (n = 10) and mice fed 1% DSS in drinking water for 2 wk (n = 11) were used.

^‡Data shown is number of thymocytes in each population per mouse ± SD (× 10⁶).

Table 3.

CD4/CD8 subsets as a proportion of total thymocytes

Sample	CD4−/CD8−	CD4+/CD8+	CD4+/CD8−	CD4−/CD8+
Control†, %	7.06 ± 1.2‡	83.0 ± 1.4	3.35 ± 0.32	6.48 ± 0.68
DSS§, %	10.7 ± 4.4	75.4% ± 11.5	5.17 ± 3.2	8.73 ± 4.1
Change, %	51.6**	−9.16	54.3	34.7

**P < 0.05.

^†Control (n = 10) and mice fed 1% DSS in drinking water for 2 wk (n = 11) were used.

^‡Data shown is percent of total thymocytes ± SD.

^§Total cell number in thymus of DSS-treated mice declined 57%.

Subdividing the latter group (CD4⁻/CD8⁻ thymocytes) into double-negative (DN) populations, e.g., DN1–4, did not reveal any noteworthy shifts in the proportion of these subpopulations (Table S2). Overall, DSS treatment caused an almost 60% decline in the number of cells in the thymus, with the greatest loss in absolute number of cells noted among the CD4⁺/CD8⁺ cells.

Underlying Cause of Reduced Thymic Lymphopoiesis.

The large loss in early lineages of lymphocytes in the primary immune tissues suggested the changes in DSS-treated mice might have accelerated apoptosis among these vulnerable precursor cells. To assess this possibility, thymocytes from control and DSS mice were cultured to determine apoptosis rate. Because the thymus contains large percentages of CD4⁺CD8⁺ pre-T cells (∼80%), known to be vulnerable to apoptosis, these cells were selected for analysis of potential changes in degrees of apoptosis. Fig. 2 shows more apoptotic cells among CD4⁺CD8⁺ cells of DSS-treated mice, with the proportion of cells undergoing apoptosis being 38% higher than the rate of death observed in control thymocytes. This sort of increase over time could certainly create the greater losses noted in this population. There was no significant change in the apoptosis rate of CD4⁻CD8⁻ precursors or more mature CD4⁺ or CD8⁺ T cells in the thymus of either group (Fig. 2). Thymocyte apoptosis can be induced by elevated glucocorticoid levels; however, analysis of serum corticosterone in DSS-treated mice showed no significant changes compared with controls at this point in time (Fig. S3). In sum, apoptosis contributed to the observed thymic atrophy and losses of CD4⁺/CD8⁺ cells during colitis.

Fig. 2.

Apoptosis of thymocytes from DSS-treated mice. Thymocytes were harvested from control and DSS-treated mice and assayed for apoptosis after 8 h in culture. Data shown are percent apoptosis of major subpopulations of thymocytes as determined by DAPI staining. Thymocyte subpopulations were delineated by labeling with anti-CD4 and anti-CD8 antibodies. Each circle represents percent apoptosis for one mouse; bars represent mean apoptosis rate. Controls, n = 8; DSS, n = 9 (*P < 0.01).

Composition of Peripheral Blood.

To ascertain whether the changes observed in the BM and thymus were reflected in the periphery, differential counts of blood from control and DSS-treated mice were evaluated (Table 4). In control mice, lymphocytes represented 80% of the leukocyte population, declining in DSS mice to 56% (Table 4). This result suggests that the decline in the lymphopoietic compartments of the marrow and thymus had translated into declining numbers of circulating lymphocytes. This outcome of course could compromise immune defense against infections. Likewise, the substantial increase in myeloid cells in DSS-treated mice was reflected by a 2.6-fold increase in the percentage of circulating neutrophils and a 60% increase in blood monocytes (Table 4). This finding was surprising given the study was only of 15 d of duration. The total white blood cell count in DSS-treated mice did not significantly change. Thus, the enhanced proportion of cells of the myeloid lineages noted in the marrow was reflected by real increases in peripheral blood. Increased blood neutrophils and monocytes indicates that cells were migrating out of the marrow in large numbers, thereby adding to infiltration of the gut of mice with colitis, exacerbating and extending the ongoing inflammation. As discussed, the gut of DSS-treated mice was heavily infiltrated with neutrophils, whereas none were readily visible in control mice.

Table 4.

Blood differential counts in control and DSS-treated mice

	% of white blood cells
Treatment	Neutrophils	Lymphocytes	Monocytes
Control†	14.2 ± 3.9‡	80.0 ± 4.8	3.15 ± 1.7
DSS	37.3 ± 11.8**	56.1 ± 12.1**	5.01 ± 4.1

**P < 0.01.

^†n = 6 for controls, n = 8 for DSS.

^‡Data shown is % of white blood cells ± SD.

Discussion

Marrow is found throughout the skeletal mass of the body including arms, legs, ribs, sternum, and skull. Collectively, the marrow represents a huge mass of tissue that is almost always affected by disease but often ignored in the scientific literature. The marrow no doubt requires a sizeable amount of nutrients to replenish cells that are needed daily by the billions. It was already known that malnutrition, especially deficits in zinc or protein calories, substantially altered hematopoietic processes in the marrow, while initiating rapid thymic atrophy (9, 21). Moreover, stress levels of endogenously produced glucocorticoids also alter marrow function and cause thymic atrophy, in part due to initiation of apoptosis among pre-T and pre-B cells (8, 22). If one considers the known outcomes of colitis, which includes weight loss, anemia, stress, reduced assimilation of nutrients, and accelerated numbers of myeloid cells in the gut and stools, it seemed highly probable that colitis would alter BM functions.

The severity of colitis can vary depending on a variety of factors including mouse strain and age. For example, 129/SvPas mice have low susceptibility, C3H/HeJ have high susceptibility, and C57BL/6 mice display an intermediate response to DSS treatment (23, 24). Similarly, young mice tend to display a more severe response than older mice but are a model for IBD effects in growing children. Herein, we show that treatment of young C57BL/6 mice with a low dose (1%) of DSS causes gastrointestinal inflammation (colitis) that significantly altered hematopoietic functions in the marrow and greatly changed the composition of the BM.

Perhaps the most salient finding was the increase from 34% to 60% in granulocytes (80% increase) and the 60% increase in monocytes in the marrow of DSS mice (Fig. 1). Given the size of the marrow in the body as a whole, this result represents a large overall increase in the numbers of cells of the myeloid lineages being produced in the body of colitis mice. This change is a clear demonstration that colitogenic changes in the body enhance myelopoiesis, which would contribute significantly to the ongoing tissue damage and inflammation in the gut (3–5). The larger numbers of these cells noted in the peripheral blood suggest that cells of the myeloid lineages are moving from the marrow into the periphery (Table 4).

Anemia and loss of weight accompany IBD diseases (1, 2, 17). A 2011 study indicates that 42% of adults with colitis are anemic (17). Moreover, 18–62% of individuals with colitis experience weight loss (25). Bleeding, ulcerative lesions, and bloody diarrhea are associated with loss of iron and can affect nutrient absorption as well. With these collective changes in iron, vitamins, and other nutrients as a bout of colitis proceeds, it is not surprising that we observed a decline of 40% in the proportion and relative numbers of cells of the erythrocytic lineage in the marrow of DSS-treated mice (Fig. 1). Evidence of the onset of anemia is apparent in the periphery where significantly reduced numbers of red blood cells, blood hemoglobin, and hematocrit are noted (Table S1). Reduced nutriture alone can reduce erythrocytic lineages in a manner similar to colitis (9, 26) and would contribute to anemia.

As is evident, DSS treatment greatly altered lymphopoiesis in the marrow and thymus. In the marrow, lymphoid lineages declined 70%, from about 26% to 7.7% of nucleated cells in the BM (Fig. 1). The pre-B cells that are prone to undergo apoptosis experienced the greatest losses (Table 1). The somewhat greater survival of progenitor and immature-mature B cells might result from increased resistant to cell death. Greater survival of pro-B cells as observed here has also been noted in malnutrition and stress and may allow for rapid regeneration upon normalization of conditions (21). Analogous findings for alterations in lymphopoiesis were noted in the highly atrophied thymuses of DSS-treated mice where the number of developing T cells was decreased by nearly 60%—verification that colitis significantly alters both primary tissues. As was the case for marrow B cells, there were substantial losses in the numbers and proportion of CD4⁺CD8⁺ pre-T cells (Tables 2 and 3) in colitis mice. On a percentage basis, the depletion of double-negative and single-positive populations was not as extensive as that for CD4⁺CD8⁺ cells (Table 3). Similar results including thymic atrophy in DSS-induced colitis have been observed previously, including large losses in CD4⁺CD8⁺ thymocytes after treatment of mice with DSS (27).

The changes in the proportion and absolute number of T cells and B cells could also be due to reduction in the production of growth factors and cytokines needed to promote their development and differentiation. However, the disproportional losses among pre–T- and pre–B cells, which have low expression of antiapoptotic proteins and are prone to cell death, suggested that enhanced apoptosis might be occurring. Indeed, thymocytes from DSS-treated mice exhibited a 40% increase in the percentage of cells undergoing apoptosis at day 15 (Fig. 2). Over time, this rate of death would certainly contribute to thymic atrophy and the loss of thymocytes.

Thymic involution and its associated loss of CD4⁺CD8⁺ thymocytes during colitis have been reported by others (27, 28). Interestingly, the use of RU486, an inhibitor of the synthesis of glucocorticoids, partially reduced thymic atrophy in DSS-treated mice, suggesting a role for stress (28). However, in our studies, we did not observe an increase in glucocorticoid levels at 15 d. In a different scenario, memory T cells adoptively transferred to immunodeficient SCID mice demonstrate a role for memory T cells in the continuance of colitis (29). Transferred CD4⁺CD45RB^hi T cells induced colitis 6 wk later (29). Similar results were obtained by using lymphopenic mice. The memory T cells produced G-CSF and GM-CSF, which the investigators assumed might account for the greater number of Gr-1⁺ granulocytes found in blood and BM. Certainly the retention of such T cells from bout to bout of colitis could also extend granulopoiesis and contribute to more intense bouts of colitis.

Understanding the changes created by colitis in the marrow provides opportunities to revise current therapies in favor of more effective ones. Natural and synthetic glucocorticoids are the most extensively used drugs for the treatment of UC and CD (30), but they do not prevent future bouts of disease. Moreover, it is recognized that prolonged treatment with glucocorticoid can increase opportunistic infections and even lead to organ damage (30, 31). The overt increase in myelopoietic cells in the marrow noted herein among mice with colitis presents yet another important problem. It has been repeatedly shown that glucocorticoids whether provided orally or generated endogenously during disease and stress substantially enhance myelopoiesis (8, 10, 32). Moreover, it has been known for years that promotion of myelopoiesis in primary BM cultures can readily be accomplished by adding Gc (32). Furthermore, glucocorticoids are notorious for initiating apoptosis among pre-B cells in marrow and pre-T cells in the thymus (8, 22). Although treatment of UC and CD patients with glucocorticoids may reduce disease symptoms temporarily, they may also inadvertently promote even greater myelopoiesis that could prolong a bout of the disease. Additional reductions in lymphopoiesis created by these steroids might well increase the likelihood of opportunistic infection. These changes add to the growing concerns about the efficacy of the wide use of glucocorticoids for treatment of IBD (31).

Erythropoietin used in conjunction with iron supplementation could be used to promote erythropoiesis. Interleukin 7, which promotes lymphopoiesis among pro- and pre-T and B cells, could be used if its production by stromal cells is compromised. Although often more difficult to implement, antibodies that would neutralize myelopoietic factors, such as granulocyte-macrophage colony stimulating factor (GM-CSF), might reduce the elevated production of macrophages and neutrophils. Indeed, a combination of these therapies might reduce the length and intensity of bouts of colitis, offsetting the need to use glucocorticoids while improving patient outcomes.

In sum, colitis-like conditions substantially altered hematopoietic processes in the marrow. However, to be identified are potential changes in cytokine levels, growth factors, and stromal cell activities within the BM that may contribute to the observed changes in composition. Similarly, changes in activity of intestinal cells and/or mesenteric lymph nodes undoubtedly will have a great impact on the changes occurring both in the bone and the colon. Future analyses need to also identify the sequence of events occurring over time. It would also be important to understand how frequent bouts of colitis might affect hematopoietic function. Moreover, it will be interesting to see whether these findings are reproducible in different rodent models of colitis that work through other inflammatory responses (Th1 vs. Th2) and in humans. Taken together our studies contribute to the further understanding of these issues to provide much-needed novel therapies.

Materials and Methods

Mouse Model.

C57BL/6 males (6 wk old; Harlan Laboratories) were given 1% (wt/vol) DSS (36,000–50,000 MW, catalog no. 160110: MP Biomedical) in sterile water for 15 d to induce colitis. Mice were housed in a 12-h light/dark cycle room at 23 °C and provided standard chow from Teklad. Body weight was monitored throughout the study. All protocols were approved by the Michigan State University Animal Use Committee.

Cecum Histology.

To monitor and confirm colitis induction, a scoring system was developed based on previous reports (33–35) (SI Materials and Methods). To further confirm colitis, RNA was extracted from the proximal colon to assess cytokine expression levels as described (11).

BM Harvest and Immunophenotyping.

BM from control and DSS-treated mice was harvested and processed as described (10, 22). BM cells were labeled with anti–Ly-6C (clone ER-MP20), anti-CD31 (clone ER-MP12; AbD Serotec), and anti-Ly-76 (clone Ter119) antibodies (10) to delineate five major hematopoietic subclasses (SI Materials and Methods). The five populations identified were as follows: erythroid (CD31⁻Ly6-C⁻Ter119⁺), lymphoid (CD31⁺Ly6-C⁻Ter119⁻), granulocytes (CD31⁻Ly6-C⁺Ter119⁻), heterogeneous monocytes (CD31⁺Ly6-C⁺⁺Ter119⁻), and committed progenitors (CD31⁺Ly6-C⁺ Ter119⁻) (8, 15). Determination of composition of B lineages in marrow was done as described (26, 36) and is outlined in SI Materials and Methods.

Thymocyte Harvest and Immunophenotyping and Quantitation of Apoptosis.

Thymuses were processed and labeled as described (37). Briefly, thymocytes were labeled with anti-CD4 and anti-CD8 (eBiosciences) to determine major subsets of thymocytes. For CD4⁻/CD8⁻ subsets, cells were labeled with anti-CD25, anti-CD44, c-kit, anti-CD4, and anti-CD8 antibodies. Cells were gated on CD4⁻/CD8⁻ cells, and delineated as follows: DN1, CD44⁺/CD25⁻/ckit⁺; DN2, CD44⁺/CD25⁺/ckit⁺; DN3, CD44⁻/CD25⁻/ckit⁻; DN4, CD44⁻/CD25⁺/ckit⁻ (SI Materials and Methods). Apoptosis of thymocytes was detected flow cytometrically after culture of isolated cells as described (36, 37) (SI Materials and Methods).

Flow Cytometric Analysis.

Flow cytometry was performed with a Becton Dickinson FACS Vantage flow cytometer or a BD LSRII flow cytometer (BD Immunocytometry; DIVA 4.1). Unlabeled samples and single-color labeled cells served as color controls. Data were analyzed by using WinList software (Verity Software).

Blood Analysis.

Blood was collected from anesthetized mice by cardiac puncture. A complete blood count with differential test was performed by the Diagnostic Center for Population and Animal Health at Michigan State University for analysis of red and white blood cell counts, hematocrit, and blood hemoglobin, and for differential counts of white blood cells.

Statistical Analysis.

Data were analyzed by using the Student t test. Statistical significance was set at P < 0.05. Unless otherwise stated, means ± SD are reported. The number of mice per treatment group was n = 8–11 unless otherwise stated. All mouse experiments were performed two or more times.